ML20070E212: Difference between revisions
StriderTol (talk | contribs) (StriderTol Bot insert) |
StriderTol (talk | contribs) (StriderTol Bot change) |
||
| Line 2: | Line 2: | ||
| number = ML20070E212 | | number = ML20070E212 | ||
| issue date = 03/31/1981 | | issue date = 03/31/1981 | ||
| title = Residual Stress Improvement by Means of Induction Heating | | title = Residual Stress Improvement by Means of Induction Heating | ||
| author name = | | author name = | ||
| author affiliation = ISHIKAWAJIMA-HARIMA HEAVY INDUSTRIES CO., LTD. | | author affiliation = ISHIKAWAJIMA-HARIMA HEAVY INDUSTRIES CO., LTD. | ||
Latest revision as of 00:55, 24 May 2020
| ML20070E212 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 03/31/1981 |
| From: | ISHIKAWAJIMA-HARIMA HEAVY INDUSTRIES CO., LTD. |
| To: | |
| Shared Package | |
| ML20067D329 | List: |
| References | |
| EPRI-NP-81-4-LD, GL-81-04, GL-81-4, NUDOCS 8212170111 | |
| Download: ML20070E212 (354) | |
Text
Residual Stress improvement by Means of Induction Heating g
EPRI NP-814-LD Keywords: Project T113-5 IHSI Research Report Residual Stress March 1981 BWR Piping Stress Corrosion Welds Pipe Remedies Prepared by Ishikawajima Harima Heavy Industries Co., Ltd.
Japan l
l l
l l
ELECTRIC POWER RESEARCH INSTITUTE l
8212170111 821203
- PDR ADOCK 05000
<tc I 74 _ _ _ - _ _ _ _ _ _ _ _ _ _
N t:
Residual Stress improvement by Means of Induction Heating N P-81-4-LD Research Project T113-5 Research Report, March 1981 Prepared by ISHIKAWAJIMA HARIMA HEAVY INDUSTRIES CO., LTD.
Japan Edited by Anthony Giannuzzi Robert Vaile Nuclear Power Division Electric Power Research Institute Published by BWR Owners Group and i Electric Power Research Institute
- l 3412 Hillview Avenue i
Palo Alto, California 94304 EPRI Project Manager A. Giannuzzi Nuclear Power Division
PREFACE This report presents results of an extensive analytical and experimental study performed by Ishikawajima Harima Heavy Industries Co., Ltd. (IHI) in Japan to qualify and verify the Induction Heating Stress Improvement (IHSI) process for boiling water reactor piping. The program activity was funded entirely by IHI.
The role played by the Electric Power Research Institute has been to provide for public release of this information.
The IHI activity described in this report has focused on providing a sound basis for anticipating residual stress improvement resulting from the application of IHSI; identifying essential and non-essential variables associated with the process; verifying the effectiveness of the process both analytically using finite element methods, and experimentally by means of residual stress measurements: and establishing process parameters from which procedure specifications for field implementation can be prepared. This document is extensive in scope and exhaus-tive in detail. The report will be invaluable as a reference document to be used by utilities, architect engineers, vendors and regulators in understanding and evaluating the IHSI process.
The main body of the report presents a summary of the work performed by IHI in the development and verification of the IHSI process. The appendices which follow report on the analytical and/or experimental detailed studies which justify the s ummary. A considerable editorial effort has been devoted to rephrasing the main body of the report and the appendices so as to clarify important features which otherwise might have been ambiguous as a result of language differences. These editorial changes have been approved by IHI. No attempt has been made to alter the technical content or the report structure. Where no ambiguity exists in the original document submitted to EPRI, the original translation was preserved. The figures in this report were prepared by IHI and are reprinted herein in the form in which they were received.
iii
CONTENTS Section Page 1 INTRODUCTION 1-1 2 CONCEPT OF IHSI 2-1 3 TEST APPROACH AND TEST RESULTS 3-1 3.1 Preliminary Test 3-1 3.2 Parametric Studies for 4-Inch Pipe 3-1 3.3 Demonstration Mock-up Test for 12-Inch Pipes 3-2 3.4 Demonstration Mock-up Test for 24-Inch Pipes 3-2 3.5 Effect of Repeated IHSI Trea tment 3-2 3.6 Relaxation Studies on IHSI 3-3 4 COMPUTER CALCULATIONS 4-1 4.1 Preliminary Studies 4-1 4.2 Effects of a Geometrical Transition 4-1 4.3 Parametric Studies 4-2 5 EVALUATION OF PARAMETERS 5-1 5.1 Temperature Difference 5-2 5.2 Temperature Distribution (Heating Duration) 5-6 5.3 Coil Width 5-11 5.4 Coil Setting tocation 5-14 5.5 Maximum Heating Temperature 5-14 5.6 Prequency of the Electric Supply 5-18 5.7 Pipe Size 5-20 5.8 Cooling 5-22 5.9 Coil Input Power 5-23 5.10 Geometrical Configuration 5-25 5.11 Summary of Parameters 5-32 v
r- .i<
L
- i
(
l'
[ .h Section Page 6 CVALUATION OF OTHER EFFECTS OF IHSI TREAhMENT; '
6f1 y 6.1 Sensitization Ef fects 6-1 6.2 Mechanical Properties ,
-; 6-1 ,
s 7 FIELD APPLICATION ; 7-1 7.1 Estimate of Input Power c 7-1 7.2 Trial Heating 7-2 7.3 Estimate of Pipe Inner Surfade Theperature 7-6 7.4 Examples of Field Application of the IHSI Process 7-7) 7.4.1 Preparation - 71 13 g 7.4.2 Trial Heating 77-13 7.4.3 Actual Heating , 7-13 7.4.4 Results and Evaludtion 7-13 8 PIPES WITH SMALL PRECRACKS 8-1 8.1 Objective 8-1 8.2 Test Details and Test Results 8-1 8.3 Computer Calculations for Precracked Pipes 8 8.4 Evaluation 8 ,
8.5 Conclusion . 8-107 j 9 RELAXATION 4 9-1 9.1 Relaxation at Operating Temperatures 9-1 9.2 Relaxation Resulting From Applied Axial Stress 9-1
'1 10 CONCLUSIONS 10-1 1
i Vi
t I
t:
ILLUSTRATIONS 1
Figure Page 2.1 Outline of Heating and Cooling Process 2-2 2.2 Stress-Strain Diagram 2-3 2.3 Concept of IHSI 2-5 2.4(a) Temperature History Curve (4-Inch Pipe) 2-6 2.4(b) Through Thickness Residual Stress Distribution T Sec. Af ter Heating 2-7 2.5 Temperature and Residual Stress Distribution Along Pipe Axis 2-8 2.6 Residual Stress Distribution Through the Thickness 2- 9 5.1 Ef fect of Temperature Difference Between Outer and Inner Surface on Residual Stress of 4-Inch Pipe 5-3 S.2 Ef fect of Temperature Difference (Theoretical Calculation) 5-4 5.3 OR - 6T Relations (Inside Surf ace Residual Stress Presented as a Function of Through Thickness Temperature Gradient - Theory and Experiment) 5-5 5.4 Ef fect of Dif ferent Through Thickness Temperature Distributions on State-of-Stress 5-7 5.5 Temperature Distribution Through the Thickness at Different Heating Times 5-8 5.6 Ef fect of Heating Duration on Temperature Distribution (Fourier Number) 5-10 5.7 Resultant Stress Distribution with Step Temperature Change in Axial Direction 5-12 5.8 Ef fect of Coil Width on Inner Surface Residual Stress 5-13 5.9 Effect of Heating Coil Location on Inner Surface Residual Stress 5-15 5.10 Coil Setting Location Guidelines 5-16 5.11 T-T-SCC Diagram for 304 Stainless Steel in Oxygenated Water at 250* C (CBB Test, 20 ppm DO, 310 hr) 5-17 5.12 Effect of Current Frequency on Temperature Distribution and Sta te-o f-St re ss 5-19 5.13 Ef fect of Prequency on Residual Stress 5-21 5.14 Ef fect of Cooling on Through Thickness Temperature Distribution 5-24 vii 1
- - -- __________________.__.)
Figure Page 5.15 Analytical Model for Temperature Distribution 5-27 5.16 Residual Stress Distribution for Gently Sloping Thickness Transition 5-28 5.17 Calculated Model and Temperature Distribution for Nozzle to Pipe 5-29 Transition 5.18 Residual Stress Distribution for Nozzle to Pipe Thickness Transition 5-30 5.19 J-l Residual Stress Distribution for Nozzle to Pipe Transition 5-31 5.20 Residual Stress Distribution for Pipe to Elbow Weld . 5-33 5.21 Analytical Model for Pipe to End Cap Joint 5-34 5.22 Residual Stress Distribution for Pipe to End Cap Weld 5-35 6.1 Hardness Distribution in 12-Inch Pipe Weld Joint 6-4 7.1 Estimation of Coil Input Power 7-3 f.; Estimation of Steady State Condition by Trial Heating 7-4 7.3 Temperature Difference Between Outer and Inner Surface of a Pipe- 7-8 7.4(a) Secord of Field Application of Induction Heating Stress -
Improvement 7-9 7.4(b) Record of Field Application of Induction Heating Stress Improvement 7 8.1 Propagation Behavior of Notch 8-3 8.2 Cracking Observation at Notch Tip (R-53) 8-4 8.3 Cracking Observation at Notch Tip (R-45) 8-5 8.4 Calculated Residual Stress Through the Thickness of 12-Inch Pipe With a Flaw 8-6 8.5 Calculated COD at Maximum Heating 8-8 8.6 Relation of COD and CTOD, to Deflection in Three Point Bend Ttst 8-9 8.7 COD and Stress Distribution 8-11 9.1 Residual Stress Distribution Comparing Three Heat Treatments to an As IHSIed Pipe 9-2 9.2 Residual Stress Analysis by Two Element Models 9-3 9.3 Load-Strain Curve of Welded 14-Inch Pipe (Experiment) 9-5 viii
TABLES Tatle Page 5.1 Summary of Pararreters for Optimized Ef fect of IHSI 5-36 6.1 Mechanical Test Results 12-Inch Pipe Weld Joint 6-3 1
I i
l iX l
APPENDICES Appendix Title
~A-1 Improvement of Residual Stress Pattern in Pipe by H1qh Frequency Induction Heating Preliminary Tests A-2 Improvement of Residual Stress Pattern in Pipe by HFIII (Second Teport)
-- Parametric Survey
~
'A-3 Improvement of Residual Stress Pattern in Pipe by HFIH (Third Report)
-- Demonstration Mock-Up Test (12-Inch Pipe)
Basic Study for 12-Inch Pipe Computer Calculation by FEM A-4 Improvement of' Residual Stress Pattern in Pipe Weldment (Fourth Report)
Pipe With Small Pre-Crack A-5 Improvement of Residual Stress Pattern in Pipe Weldment (Fifth Report)
-- Ef fect of IHSI Repetition
.A Relaxation Study on IHSI A-7 Effect of Plastic Deformation by IHSI and Subsequent Heat Treatment Up on IGSCC Susceptibility of Type 304 Stainless Steel A-8 Some Analytical Case Study of IHSI Ef fect of Geometrical Transition A-9 Analytical Studies of Some IHSI Parameters Temperature Difference
-- Coil Width FYequency xi
Appendix Title A-10 Residual Stress Improvement for 24-Inch Pipes A-11 Estimate of Input Power for IHSI' A-12 ' Trial Heating Technique of IHSI A-13 Estimate of Pipe ID Temperature During IHSI A-14 Analytical Studies of IHSI. for a Pipe With a Small Pre-flaw A-15 Residual Stress Improvement by Heans of Induction Heating
( A) Article Prom IHI Engineering Review, Vol. 11, No.'4) i xii
Section 1 INTRODUCTION Intergranular stress corrosion cracking (IGSCC) of austenitic stainless steel is believed to occur when three factors, material sensitization, stress, and environ-ment exceed some threshold simultaneously. It is believed that weld residual stress is one of the most important components of the tensile stress which contri-butes to IGSCC. A new technique, Induction Heating Stress Impovement (IHSI) has been developed with the capability of lowering the residual tensile stress on the inner surface of a pipe in the heat-affected zone near a weld, the region suscep-tible to IGSCC. This new technique can be simply explained by the fact that cour pressive residual stress is induced by plastic flow on the inside surface caused by a large thermal stress resulting from a large temperature difference across the pipe thickness. This large stress is produced during induction heating of the pipe outside surface and simultaneous cooling of the inside surface of the pipe.
During heating the pipe yields in tension on the inside surface and upon cooling of the outside surface, compressive strains are produced on the inside surface.
Many basic experiments and full-sized mock-up tests on the actual austenitic stainless steel pipes, 4-inch, 10-inch, 12-inch, and 24-inch, and calculations using finite-element techniques have been conducted in order to determine appro-priate heating conditions and verify the effectiveness of IHSI. As a result of these studies, the key parameters which have a significant effect on IHSI have been determined.
Based on the studies mentioned above, IHSI has been successfully applied to piping in several boiling water reactors in Japan, both operating and under construction, to mitigate the occurrence of IGSCC.
Recently, some additional experimental and analytical studies have been conducted which demonstrate that a small initial crack in a pipe will not propagate during IHSI and also that the residual stress is effectively improved in this case.
1-1
1 In this report, the test details and calculations are briefly described and the key parameters resulting from these investigations are discussed in some detail.
The test results and calculations for a pre-cracked pipe are also explained. The
- details of test conditions and calculation methods are explained in the attached i appendices.
i a
t i
~
~I 1-2
, . . , ,, ,, , ., ,_ _ m - .,m,. _ _ _ . _ _ ._
Section 2 CONCEPT OF IHSI The objective of IHSI is to introduce a large temperature difference between the inner and outer surfaces of a pipe in order to produce sufficient thermal stress to induce some plastic flow and consequently to obtain a compressive residual stress at the inner surface of the pipe. The large temperature difference can be introduced by heating a pipe from the outside with an induction coil while cooling water is supplied simultaneously to the inner surface of the pipe. This process is shown schematically in Figure 2.1.
The mechanism for the stress improvement is, briefly:
When a linear temperature gradient is produced across the thickness of a long pipe, the thermal stress induced can be approximately written as follows:
o = 2 (1-v) (1) where a: Thermal stress at each surface, both axial and circumferential as Linear thermal expansion coefficient AT: Temperature difference between inner and outer surfaces E: Young's modulus V: Poisson's ratio When this thermal stress is below the yield stress (e.g. , point 1 in Figure 2.2),
the stress is relieved when the temperature dif ference is removed, and no residual stress is produced. However, if the thermal stress exceeds the yield stress (e.g. , point 2 in Figure 2.2) , the subsequent removal of the temperature dif ference changes the stress-strain condition from point 2 to point 3. Thus compressive residual stress is produced.
2-1
4 WELOMENT INDUCTION COIL HEATING ZONE ,
PIPE S' N t wisisisi s s i si sin - visis/ / // / / / sis
/
4
\/
!-l e
~
COOLING
! WATER E
b* i
~
0%,,,,,W....,,riss~ m\
iE 4
Fig. 2.1 OUTLINE OF HEATING AND COOLING PROCESS 2-2
c
/
/
E
- /
N j
a 8
H /
/ @/
Q/
3
- 6
- s kt n HE W L STRAIN (1/2)o AT b$
RM NN a: p G
4 2.2 STRESS-STRAIN DIAGRAM 2-3
Figure'2.3 illustrates the stress distribution, deformation and temperature distribution in the pipe element during heating and af ter treatment. As a plane-sterin condition is always maintained through the thickness, each fiber on the inner surface is elastoplastically elongated to the equilibrium point during heating, therefore becoming longer than its original length. Af ter the treatment, the fiber is compressed to approximately its original length. (equt11briu4n posi-
. tion) , thus producing compressive residual stress on the inner surface, and vice versa on the outer surf ace.
Figure 2.4(a) shows a calculated temperature history for a typical IHSI treatment and Figure 2.4(b) shows the stress distrihntions corresponding to some typical points in the temperature history.
Figure 2.5 shows the temperature and the residual stress distribution in the longitudinal direction of_ pipe which was induction heat treated by a coil of a finite length. The residual stresses on the inner surface of the pipe under the coil are compreusive. Figure 2.6 shows the calculated residual stress distribu-tion through the thickness at the coil center. The residual stresses remain compressive for approximately 60% of the pipe thickness.
2-4
OUTER SURFACE -cy MEAN HIGH (a) DURING f HEATING 8 INNER SURFACE +0y AT' oy: YIELD STRESS PLANE STRAIN CONDITION FREELY EXPANDED CONDITION OUTER SURFACE TENSION (b) AFTER -
I
! l TREATMENT ' I INNER SURFACE COMPRESSION AT=0 DEFORMATION EM N TURE STRESS DISTRIBUTION DISTRIBUTION Fio. 2.3 CONCEPT OF IHSI 2-5
E-
-' oa ,
600 gegPCE r 400 @ '
l _
p (
y D2 2
. 200 x s h
\ x m 355 0 10 20 30 40 50 60 SEC r
! TIIE (SEC)
.i I.
3 1
I i
Pig. 2.4 (A) TEllPERATURE IIISTORY CURVE (4 IN. PIPE)
J' c
d 1
2-6
. . ~ _ _ . . . . - _ . _ . _ . . - , _ . - _ _
OUTER SURFACE (kgf/mm2) 20 -10 0 10 20 30 -30 10 0 10- 20 30 ,
. . . . . . . . . . . . 1 I f f f f f f I t f I t INNER SURFACE
@T = 0 SEC @T = 0.6 SEC' 20 -10 0 10 20 30 20 -10 0 10 20 30
, . . . , i , i . , , ,
\
l i t i 1 I I I I f f 1
@T = 35 SEC @T = 40 SEC 20 -10 0 10 20 30 -30 10 0 10 20 30
. . i i i . . i i . . .
G
)
, , / , , , , , , , , ,
@t = 45 SEC -@ T = 355 SEC Fiq. 2.4 (b) TIIROUGli TIIICKNESS RESIDUAL STRESS DISTRIBUTION TSEC. AFTER HEATING 2-7
1 125 y HEATING COIL
\ h TEST PIPE (R-21)
U 1
+ 600 -
) , e : CALCULATED
- a -
^
i g -
o : EXPERIMENTAI.
I 400 -
g : OD 3: -
y --- ID g 200 -
m - -* - - -+- - - o . . , s N
e r- - - - -- -o -- _ _ _ , ,,s*,',D 3, 0 20 40 60 80 100. 120 140 160 180 200 220 240 7 ~ DISTANCE FROM COIL g (mm) ,
E Dtr
~
' ~
u A Ws -
~
va u 10 x's % ..e-4 a6
<=
oS DISTANCE FROM COIL g (mm)
- 5o 0
$U 0 20 '40 60 80 100 120 140 160 1M 200 # 2'25' 240 a
g _ et
- i. /s w //
N -10 -
--- O D. 2~ O
. 3_0_ '. - -.. ~ . . / ,/
j -20 -
Fig. 2.5 TEMPERATURE AND RESIDUAL STRESS l
DISTRIBUTION ALONC PIPE AXIS i.
^
5 l
2-8
21.4 . OD PIPE SURFACE k' '
J f
17.12--
12.84__
8.56--
li o---- : AXIAL STRESS
$ e : HOOP STRESS S 4.28 .
!2 0
-10 :c e ID PIPE SURFACE f 1 I I 1
-20 10 20 30 STRESS (kgf/mm2)-
Fig. 2.6 RESIDUAL STRESS DISTRIBUTION THROUGli THE THICKNESS 2-9
Section 3 TEST APPROACH AND TEST RESULTS Many tests were conducted to develop IHSI parameters and some of these have been previously reported.III In this section, all the tests and the test results, including what has not been reported previously are briefly explained. The detailed information is described in the appendices.
3.1 PRELIMINAlf TEST To evaluate the effectiveness of the IHSI process, two austenitic stainless steel pipes, each 4-inch Schedule 80, having one circumferential weld at the center of the pipe, were subjected to an IHSI treatment and the residual stresses were determined by conventional strain gage and cracking tests in boiling 42% MgCl 2 solution. It was shown from these data that IHSI produces a relatively large compressive stress, or at least less tensile stress, on the inside surface of the pipe throughout the weld heat af fected zone. The details are presented in Appendix A-1 3.2 PARAMETRIC STUDIES FOR 4-INCH PIPE A parameter survey test was conducted to define improved heating and cooling conditions for the IHSI process. The primary variables considered were electric power frequency, heating time, temperature difference, and pipe size. The results showed that the temperature difference through the pipe thickness was important to the final state of restdual stress, and that frequency and heating time were of less importance.
Reference:
(1) Appendix A-15 Residual Stress Improvement by Means of Induction Heating.
3-1
The data for the 10-inch pipe suggested that large pipes require a longer heating-coil to optimize the residual stress results. The details of these experiments are presented in the Appendix A-2.
3.3 DEMONSTRATION MOCK-UP TEST FOR 12-INCH PIPES Prior to the field application of IHSI to 12-inch diameter recirculation primary loop riser piping in a boiling water reactor, a series of demonstration mock-up tests and some fundamental tests were conducted to determine the optimum heating conditions and to verify the validity of the IHSI process.
The results demonstrated that IHSI can be applied successfully not only to straight pipe but also to pipe-to-elbow joints and further, that no detrimental effect of this process on microstructure or mechanical properties occurs. The details of these tests are presented in Appendix A-3.
3.4 DEMONSTRATION MOCK-UP TEST FOR 24-INCH PIPES Two 24-inch pipes, one with a thickness transition and the other just a straight pipe, were tested to demonstrate the validity of IHSI for larger pipes typical of the main recirculation loop piping in a boiling water reactor. The test results demonstrated that there were no difficulties in applying IHSI to larger pipe. The test details are presented in Appendix A-4. In addition to these tests, three full size mock-up tests were conducted in a joint utility-fabricator IHSI develop-ment program. The results of that program are presented elsewhere.
3.5 EFFECT OF REPEATED IHSI TREATMENT It is desirable to make trial IHSI runs under laboratory conditions to determine the input power level and duration needed to produce the desired temperature regime in each size of pipe. From one set of such trial runs the effect of repeated IHSI treatment can be deduced. It appears that repeated IHSI treatment produces only slightly better residual stresses than a single treatment. The details are presented in Appendix A-5.
3-2
3.6 RELAXA PIOri STUDIES ON--IHSE Relaxation behavior af ter IHS t was checked on 4-inch Schedule 80 pipes. Three pipes given an IHSI treatment were subjected to three different stress-relieving heat treatments (500* C x 24H, 400* C x 24H,- 300* C x 24H) and the residual stress was measured. . tie results showed that no significant' difference in residual stress was produced by the stress relaxation heat treatments of any of the pipes. The details of this investigation are presented in Appendix A-6.'
4 t
s t
I i-t k .'
I 3-1
,, . ~ . , - , .,--.c, - - .- - ,,. ,,,n -- - , . ,-. ..-r, , , . - - .,,-,,..,-,.-.._,.-..:.-
Section 4 COMPUTER CAICULATIONS 4.1 PRELIMINARY STUDIES Thermo-elastoplastic stress analyses were conducted for two finite-element models using the ANSYS computer code. One model was a smooth straight pipe to compute the stress distribution across the pipe wall and compare it with experimental' data. The other was a pipe with a weld 60 mm from the coil end.
The material physical properties and the stress-strain curve were adjusted as a function of temperature, and a kinematic strain-hardening rule was applied to the model. The analysis produced a temperature distribution and residual stress along the pipe axis which compared closely with the experimental data. The calculated residual stress distribution through the pipe thickness indicated that for appror-imately 60% of the wall thickness, as measured from the inside surface, the resid-ual stress was compressive following the IHSI treatment, assuming initial residual stress (before IHSI) of zero.
The results of the second analysis showed that the different mechanical properties of the weld metal, and the coil location, had no significant effect on the residual stresses.
The details of the computer calculations are presented in subsection B of Appen-dix A-3 of this report.
4.2 CFFECTS OF A GEOMETRICAL TRANSITION Computer calculations were performed for different geometry pipe transitions using the finite-element code ITEPC-II (a thermo-elastic plastic stress analysis program developed by IIII) . For this calculation, the physical properties and stress-strain relationships were varied with temperature and an isotropic-strain harden-ing rule was applied.
l l
4-1 4
The calculated results showed that the geometric effects on residual stresses were not significant.
The details of the calculations are presented in Appendix A-8.
4.3 PARAMETRIC STUDIES The effect of temperature difference, coil width and electric power frequency were studied by finite-element methods using the same program and the same mechanical and physical properties as described in paragraph 4.2.
The results of this study showed that temperature difference and coil width pro-duced a significant effect on the residual stress distribution when they were smaller than the critical values and that frequency had no significant effect.
The details of the calculations are presented in Appendix A-9.
4-2
'Section 5 EVALUATION OF PARAMETERS The effectiveness of IHSI is influenced by many parameters; however, sufficient residual stress improvement is promised if the key parameters, which significantly af fect the residual stresses, are properly controlled.
In this section, all relevant parameters are discussed from the following v iewpoints:
- 1. Which are the key parameters?
- 2. In what range should they be controlled?
It can be readily observed from Equation (1) on page 2-1, that a thermal stress sufficient to cause plastic flow aan be obtained for any material if the tempera-2(1-v) ture difference in larger than #
Ea Y*
Therefore, tamperature difference is one of the key parameters to be controlled in improving the residual stress distribution. It must be noted that Equation (1) is valid only for the case in which a sufficient length of pipe is uniformally heated for a sufficient time to make the temperature gradient through the thickness constant.
For these reasons, the heating coil length and temperature distribution across the thickness also become key parameters in the IHSI process.
- rnese f actors, which are the key parameters, and the controlling ranges of appli-1 cat
- inn are discussed in the to11owing paragraphs.
J 5-1
l 1
1 TEMPERATURE DIFFERENCE 5.1 Thermal stress produced during heating is proportional to tempe ature difference (AT) across the pipe wall [ Equation (1)]. Accordingly, it is expected that the plastic strain produced during heating and the resulting residual stresses will increase with increasing AT.
Figure 5.1 presents experimental results for 4-inch Schedule 80 pipes which were treated at different AT. Pipe R-15 produced a residual stress (o R) f -17 kg/mm2 at AT = 322'C and pipe R-17 produced -8 kg/mm2 at AT = 132* C, which is in good agreement with the expected results. This OR - AT relationship is valid however only when residual stress is equal to or less than the yield strength of the pipe.
As illustrated in Figure 5.2, when the residual stress corresponds to a larger AT than 4cy (1-v)/Ea, yielding occurs at point 3a during cooling and the residual stress becomes y-c , that is, the residual stress is always -cy even if AT > 4cy (1-v)/Ea.
OR - AT relations obtained experimentally and analytically are shown in Fig-ure 5.3. In this figure, the abscissa describes AT normalized by 4ay(1-v)/Ea.
Both curves indicate that the residual stress is almost constant if AT > 4ay (1 -v) /Ea.
It may be concluded from the above discussion that the temperature difference should be larger than 4c y(1-v)/Ea to obtain the maximum residual stress effect.
If one assumes the following material parameters for an austenitic stainless steels o:
y 25 kg/mm2 v 0.3 E: 1.98 x 04 kg/mm 2 a: 16.4 x 10
-6 mm/* C 5-2
30 30
= = HEATING COIL I
TEST PIPE 600 --
o : R-15 U -
e : R-17 400 _
$ - _ _ - : ID N
200 -
e*- ' No 0
40 20 0. 20 40 60 DISTANCE FROM COIL g (mm)
O N
l 10 -
DISTANCE FROM COIL g (mm)
~
m 40 m 0 ,A
- # 'e [
0 20 60 s m i d *
-10 O
m f O w a o M
- O j -20 -
H O N
-30 -
Fig. 5.1 EFFCCT OF TEMPERATURE DIFFERENCE BETWEEN OUTER AND INNER SURFACE ON RESIDUAL STRESS OF 4 INCH PIPE 5-3
2ay -
f
/
, /
/
i .
/
w /
ay .
i a /
w h
l o
/
/ / e
/ aT = 2(1-v) ,
2ay ,
s@
E E*
v3 N
o . _ _s' 0Y 3 h
8 m
DURING ISill TilERMAL STRESS 0 = -
AT-2 -v) f TO 03TAIN MAX. COMPRESSIVE STRESS l
c > 2cy
- I ~"'
., ST > -
2cy % 216*C i
Fig. 5.2 EFFECT OF TEMPERATURE DIFFERENCE (TIIEORETICAL CALCULATION) s-4 j i
AT/ I doy (1-v)/Ea]
O 1,. 0 2,. 0
^
R N
$ -5 O 41N SCII8O ei m oo A 12IN SCH100 0
N m
& O O 8
y 5 O A
(a) EXPERIMENTAL RESULTS AT/[4cy(1-v)/Ea]
O 1.0 2.0 t t g COIL g b s E ;
=
\\\\\\\\ }
N 1 m --
ka O : :
y su (b) ANALYTICAL RESULTS Fig. 5.3 aR - *T RELATIONS (Inside Surface Residual Stress Presented as a Function of Through Thickness Temperature Gradient - Theory and Experiment).
5-E
the required minimum temperature difference (ATmin) can be calculated as follows:
I 4 x 25 (1-0.3)
T =
- " 4 ~
1.98 x 10 x 16.4 x 10
= 215.6* C This is the minimum temperature difference expected to produce the maximum resid-ual stress.
Note: The material constants at 20* C were used for the above calculation because they give a larger AT than those at higher temperature.
5.2 TEMPERATURE DISTRIBUTION (HEATING DURATION)
Thermal stresses produced at the inner surface of a pipe depend on the temperature profile through the pipe thickness. As mentioned earlier, if the temperature gradient is linear, the thermal stresses are given by Equation (1). However, a temperature distribution similar to a thermal shock as shown in Figure 5.4(a) induces very small thermal stresses at pipe inner surface even if the same AT as that in Figure 2.4(b) is attained.
The temperature distribution in Figure 5.4(a) and (b) correspond to rapid heating and steady state heating respectively.
In order to obtain the steady state condition, the pipe should be heated for a suf ficient time; thus, heating time becomes a key parameter.
In order to estimate the minimum required heating time, an idealized model as shown in Figure 5.5 may be considered. When an infinite plate, which has uniform 5-6
I To To
)
i 1
Ti Ti
^ -
-ErxAT * ~ E:x6T "o "
(1 - v) 0 2(1-U).
ai = 0 c i = 2 ( 1 -- v )
(a) RAPID llEATING (b) STEADY STATE IIEATING Fig. 5.4 EFFECT OF DIFFERENT TilROUGli THICKNESS TEMPERATURE DISTRIBUTIONS ON STATE-OF-STRESS i
5-7
~
C t
@ {
o E* -
IN w~
0.
5 N E4 _
To s
T==
T=T i
T Ti T=0
=
t i
i n at .
Tx = (To - Ti)( Et 1 .E I-II -1 e (nr F u n=1 n x Sin ""* ] + Ti Fig. 5.5 TEMPERATURE DISTRIBUTION TliROUGH THE TIIICKNESS AT DIFFERENT llEATING TIMES 5-8
temperature Tt across the thickness, t, is suddenly heated to To on one surface, the temp,trature distribution across the thickness af ter an elapsed time of T is given by the following equation.III T
x =AT(*t
-- 2,y (-1)"~l ,-(nn) u n yxsin t
] +' T i 2)
T,T,T:
o i x 'hmperature at x = 0, x = t and x = x, respectively as Temperature diffusivity l na Natural number AT = To -T g It is obvious from B l uation (2) that the second term in the parenthesis shows the viriation from the steady state condition and that if the Fourier Number, aT/t2 ,
becomes large, the second term can be neglected; that is, aT/t is2 the key parameter to obtain a steady state temperature distribution. In Figure 5.6, the calculated temperatures at mid-thickness during IHSI are plotted against Pburier Number. The calculations were made on 12-inch, 20-inch and 28-inch pipes which !
were heated from the outer surface using an induction coil. For all cases the temperature (71/2 - TL )/(To -T) tbecomes approximately constant when at/t2 > 0.7.
) Based upon the above discussion, the heating time T should be controlled so that the F'ourier Number becomes larger than 0.7.
For example, given the following thermal diffusivi y and pipe thickness:
as 0.016 m2 /h t 0.04 m (28-inch pipe) l the required minimum heating d.iration can be calculated as follows: i
- a 0.7 (0.04) 0.016 i
= 252 nec.
Referenen: (1) Y. L1tsutch " Theory of Heat Conduction," Kyoritsu, Japan, p.45 (1956).
2 5-9 l
, - - - _ . - . . - , . - _ . , - - _ . _ - , , . - , . . _ ._..,-._,-._.,.-_.m_,._,._ .__,_.-.m_. _g , . . . , , _ _ _ - . .
0.7 -
012" x SQt 100(CAI4UATICN)
- O 0.6 - O c A- -C O -O O 20" x SQt 100(DmERImfr)
A 28" x SCH 100(CAIInATICN) o ----huwavt 00GATIa4 0.5 - ~ ~~ ~~~-~~~~~~~~-~~
o b s
m e 0.4 - f TO TO
, Tl/2 w h o i b E* 0. 3 - r o .
, 'I/2 i
f t/2l Ti t/2 Ti a: DEREL DIITLSIVITY 0.2 -
f t t t: TIIOCESS 4 Q' t: IIEATI!O TIME I/ ,
IAPCE II)LRIER FO.
0.1 -
0 '
O.5 0.7 1.0 1.5 FOURIER NO. = a t/t*
Fig. S.6 EFFECT CF HEATING DURATION ON TEMPERATURE DISTRIBUTION
5.3 COIh WIDrli Since the pipe is heated by a coil of finite width, a temp"rature dif ference in the axial direction occurs near the coil end. This situation is schematically illustrated in Figure 5.7. The effects of a step temperature change in the axial direction can be approximately expressed as the effect obtained when pipes with radii dif fering by AR are connected together by a shear force (Q) . The ef fect of the shearing force can be expressed using elastic analysis as follows:
2 M = -D dx (3)
~*
= o sin Bx where 3(1-v) 8=4 22 1.285 1
/ Rt /Rt__
Mg reaches a maximum at Sx = 0.8 (x = 0.62 3/Et) and exponentially decreases as x increases.
Hg acts to concel the thermal stress due to the temperature difference across the pipe thickness.
Therefore, the coil width should be great enough so that the ef fect of one coil end is not superimposed on the other. As a result, coil width becomes a key parameter and if it is larger than 3 /5t, the maximum effect of Q is not increased by the ef fects of the other end. Since the above discussion is based on elastic analysia, a series of experiments and elastic plastic analysis r;y finite element methods were performed to investigate this effect in more detail. The results of this study, presented in Figure 5.8, are consistent . ith the above discussion; namely, if the coil width is greater than 3/5t, the residual stresses obtained are approximately constant.
An an examplo, if the following parameters for pipe radius and thickness are given, L = 40 mm R = 3 56 mm (28-inch pipe)
, 5-11
HEATING COIL Z 7 8 [2. _- _ _0_ h
-D V N /--- N k___
ta o
x/ M I ,
y a o
Ed Ed kra #
k n
x/ M L
$ x/ M -
E y (L/ M 43)
Fig. 5.7 RESULTANT STRESS DISTRIBUTION WITH STEP TEMPERATURE CHANGE IN AXIAL DIRECTION 5-12
COIL [
COIL WIDTl!
30 -
+' .
]1 i
~
'h -a PIPE 20 " l 3
~
o AXIAL (10" x SCH 80)
S t o -. y AXIAL (4" x Scil 80)
H EXPEAIMENTAL*
y , a AXIAL (12" x Scil 100)
O HOOP (12" x SCH 100)
- 8 re AXIAL (12" x SCII 100) g u gxg CALCULATED <
g p ,
"c o '$ N
' h 'gs ,'A m ,.
b \
-2 0 - b o
8
=
g .N.N _
m 8 i i . .
0 1 2 3 4 5 6 COIL WIDTH / / lit Fig. 5.8 EFFECT OF COIL WIDTH ON INNER SUR{ ACE RESID'fAL STRESS 4
then the minimum required coil width can be calculated as follows:
L = 3/Rt = 3/356 x 40
= 358 mm 5.4 COIL SETTING LOCATION It is clear from Figures 2.5 and 5.9 that the residual stresses at all inner sur-faces of pipe under a coil can be improved by IHSI. Therefore, if the weld heat affected zone (HAZ) is located within the area covered by the heating coil, resid-ual stresses in the HAZ are expected to be improved.
The RAZ of a normal weld is generally within 15 mm (or t/2) from the weld center line (mid-plane) as illustrated in Figure 5.10. Therefore, the coil should be positioned so that the weld center is located at least 15 mm (or t/2) inside the coil end.
Note: In many cases, the weld center line should be located much further inside the coil to obtain a larger temperature difference across the pipe thickness.
5.5 MAXIMUM HEATING TEMPERATURE In general, as the maximum temperature increases, the temperature difference, AT, across the pipe thickness becomes larger. The effeci of AT, however has already been taken into account in paragraph 5.5 Therefore, maximum temperature is not a key parameter from this perspective but should be restricted so as not to deterio-rate the pipe material properties.
For austenitic stainless steel, the maximum temperature should be determined so as not to sensitize the pipe material. Considering that the maximum heating time for IHSI is only a few minutes, it is clear from Figure 5.11 that the temperature of 550*C is sufficiently low for Type 304 stainless steel.
As a result, the temperature on pipe outside surface should be limited to less than 550*C.
5-14
125 _
!! EATING COIL TEST. PIPE (R-21)
U o
600 -
N e
.C : CALCULATED
$ _ 2 _ 7%
b ~ '
k o : EXPERIMENTAL w 400 -
& : OD Ei e -
--- : ID 200 -
e - - +- - -e- - -e _
\
~
0- - - - ~ ~ -o - _ ,,,,, 's ,
0 I ' ' ' I ' I
' *-N' g ' ' I '
0 20 40 60 80 100 120 140 160 180 200 220 240 DISTANCE FROM COIL g (mm) x Ne O Os v2 e'*
$ 10 e m
u e%*'sj of aG
~
DISTANCE FROM COIL g (mm) e ac u
$$ 0 ' ' ' ' ' ' - ' ' N* '
~~
$$ 0 20 40 6U 8b lb0 120 140 160 180 2$0520 240 ym -
,e ,i a0 o V
$~ _10 -
~ -Q 30 x O. - Q ~ __ _,O_ -- ' _ _O ,/
e, _ _, . - -.- o o n _o , / '
8
-20 -
O i
Fig. 5.9 EFFECT OF llEATING COIL LOCATION ON INNER SURFACE RESIDUAL STRESS 5-15
IIEATING COIL t/ /
SENSITIZED ZONE 5
\\QSi]
\ \
\
j l
5
\ 'NN /
/
\ \ / /
15 mm OR t/2 WIIICIIEVER IS LARGER Fig. 5.10 COIL SETTING LOCATION GUIDELINES 5-16
)
8 7
9 1
(
5 6
) 1
~"N m ,
- , (
p 7 2
,TH P
d e
u s
E t t D a u N n 0 j 0
0 K e i 5
0 C g G y -
m 9
< 5
> % 0 x u 0 1 o k 0
5 ~2' n o
< 'U' i h
s N%
M ) o W ' I X
A j
)
r l
e e
h r B M t 0 o s
w - ...
- N E H
T h
(
e m
s s
e 1
3 O w
, a t
o m
~ - W O
H i
t n
l i
n D
m a
K S a p .
o t p T S i s R t 0 ,
E a 4 2 i wi N B z 0 h M i 3 ,
s 1
U t t a r
"- N E
i s
n e
f o
s k e A t
H g N S
.} .
- x .
.{
T m a
r B M B
C :
,x
[ s g) (
0 a( 1 e
r- y i C c
- .[{1 2
d n
-- 7' T .-- 0 e C 5 r
/ C 2 e S f f(\ - t e T a R C T r C e t
S N a N o 1
1 1 w g N. N 0 5 g
i 0 0 0 0 0 F 0 0 0 0 0 9 8 7 6 5 UE o tgo c # n j- 7gm yi3
5.6 FREQUENCY OF THE ELECTRIC SUPPLY
.l The heating depth, S, (current penetration depth) for the induction heating process is given in Equation (4).
l S=b "
(cm) (4)
.pf x 10 '
~
l where p Specific magnetic permeability f: Frequency (Hz) p Specific resistance (O . cm) f The temperature profile through the pipe thickness resulting from induction heating is given by the-following equation.III WS T =T -
p
(-1+h+e'*!) (5) where T x, To: Tempecature at position x and at outer surface respectively.
Atp Thermal conductivity of pipe Wg Heating density at outer surface x: Distance from outer surface J ,
Equations (4) and (5) show that the temperature profile across the pipe thickness depends on frequency, f, and the profile changes from linear to a convex (para-bolic) shape as the frequency changes from = to 0. This situation is illustrated in Figure 5.12. For the parabolic temperature distribution, the thermal stress on the inner surface of a pipe is given by Equation (6) 4 2MT o (6) 3(1-v)
Reference:
(1) Appendix A-15 Residual Stress Improvement by Means of Induction Heating.
5-18
~
% /
To S/t=0.8 S/T==
S/t=0 Ti t _
(a) TEf1PERATURE DISTRIBUTION t'
S/t=0
\ __ S/t=0.8
//
/ S/t==
/
/
/
, / c
/ /
/ /
/- /
(b) STRPSS STRAIN REIATION FOR VARIOUS TCf PI'RATCP.E DISTRIBUTIOf!
!' i g . 5.12 EFFECT OF CURREtJT FRE')UENCY ON TE!!nERY2U al: DISTRIflUTION AND STAT.;-OF -STRESS 5-19
Comparing Equations (6) with (1) demonstrates that a very low frequency (parabolic temperature distribution) provides a thermal stress which is 33.3% higher on the inside surface than the linear tAmperature distribution. Therefore the parabolic I
temperature distribution seems to improve residual stress more than the linear distribution. Since the minimum required temperature difference was determined based on a liner temperature distribution, it is expected that the residual stresses induced by a parabolic temparature distribution will be similar to that provided by a linear temperature distribution. This effect is illustrated in Figure 5.12.
In order to verify this expected result, elasto-plastic calculations for various S/t were performed using finite-element techniques. The results of these calculations are presented in Figure 5.13. As mentioned above, no significant difference in residual stress is observed in the range S/t from 0 to a.
Therefore, it can be concluded that large S/t produces some additional margin in temperature difference but this margin is not converted to additional residual stress improvement when compared to small S/t as long as the minimum required temperature difference is maintained.
5.7 PIPE SIZE Two parameters are important in pipe size evaluation, thickness, t, and diameter, d.
The effect of diameter has previously been considered in the discucsion of coil width effects on residual stress.
The effect of thickness has also been accounted for in the prior discussion of coil width and heating times and the effect of S/t was accounted for in the deter-mination of the minimum temperature difference required to produce an optimized residual stress distribution.
Pipe size, therefore, need not be restricted as long as the coil width and heating time are properly controlled.
5-20
) ) )
z z z H l i t i
K K K 0 0
0 3 = 1
=
f f f .
( ( (
h 6
0 cm
=_
9 S m 0
= = = .4 N
s s s I =
8 o , a 2 t t
/
M0A$m x .
S M0M$a 5 1 S
/ E R
o l' T 5
1 S
/ L
_ . 14 D o A 5
1 U
- /
3 D
_ 1 I
A o S 5
1 E
-p ./2 1
R N
A o oa
' o
/
1 5
5 1
1 O
Y C
N 1
/ E 0
1 U
A O 5
Q 1 E
./ R 9 F 5
1
/
F
,p O 0 T 5
C
.J/
1 7 E F
Q F E
, 6/
5 A 3 1
/ 1
- 5 .
O 5 5 A 1 .
/ g 4
5 i 1 F
/
3 5
1
, /
2 5
1
, /
1 A' N
UFeom 0 0 0 0 0
0 3 2 1 0 aO mz- 1 2
3 7 EEy n t5 $gg' H$e
,4 "-
5.8' COOLING The through-thickness temperature difference resulting from the induction heat treatment process is easily derived by substituting t for x in Equation (5) and subtracting the value from To as presented below:
WS ~
AT = T -T y =
4
(-1 + +e ) (7) p One observes from Equation (7) that cooling variables do not affect the through-thickness temperature difference. Werefore, cooling variables are not key parameters.
It is necessary, however, to remove the heat produced by induction heating from the pipe inner surface in order to protect the pipe from overheating.
Heat convection from the pipe to the water is given by q = hm (T -T) (8) where q: Heat flux at pipe inner surface hm Mean heat transfer coefficient T, Water temperature Consider the heat conduction in the pipe wall, q=A ( 9)
Dif ferentiating equation (5) at x = t and substituting Equation (9) into Equa-tion (8), produces
~
T -T = (1 - e ) (10) 5-22
l l
I where hm is given by the Dittus - Boolter equation as, hm e 0.023 *P Ua (11) 0 r d .2 v0.8 where Aw: Thermal conductivity of water di 'Inside diameter of pipe v: Coefficient of kinematic viscosity P
r Prandtl Number U=: Water flow rate Temperature distributions across the thickness for various cooling conditions were calculated using Equations (5), (7), (10) and (11), and have been plotted in Fig-ure 5.14. It can be easily understood that the temperature at outer and inner pipe surfaces are reduced as the cooling water flow rate is increased. In other words, higher water flow rate enlarges the allowance between T and o the maximum temperature limit, maintaining the same temperature difference through the I thickness.
4 5. 9 COIL INPUT POWER a
Coil input power is a function of
- 1. Heat. generated per unit length: [W e
-2x/S dx o
- 2. Coil width: L
- 3. Pipe diameter: D
- 4. Efficiency
- n
. The coil input power can be represented functionally as follows:
-2x/S i
P= f(fW e dx, L, D, n) (12) o i
6 1
5-23
l EFFECT OF FLOW RATE 600 (CONSTANT HEAT FLUX) -
600 SCO 8" * * !" ~ '
pfx10 St CURRENT PENETRATION DEPTH (cm)
U: PERMEABILITY f: FREQUENCY (CYCLE /S)
U RESISTIVITY (Q.cm) a m
400 X 400 _
EI N
@ b Es m$
h FLOW $
@y RATE (U.)
@ bM tc H (m/s) h t4 y 200 ~
O g 200 $0 e8~
5 N En
@ 1.0 @ bU
@ 1.5
@ 2.0 G g
h 2.5 \
N (35*
~ _ _ _ _C)_
@ 3.0 0
0 IS 2S THICKNESS (L) (mm)
Fig. 5.14 EFFECT OF COOLING ON TIIROUGli THICKNESS TEMPERATURE l DISTRIBUTION l
l 5-24
L i
i The temperature difference between the pipe outer and inner surface is a function of
- 1. Heating density at the outer surface: W o
- 2. Ileating depth S
- 3. Pipe thickness: t
- 4. Thermal conductivity: A p
i as shown in Equation (7) .
l The temperature difference can be presented functionally as follows:
Er = g(W , S, t, o
A)p (13)
Comparing Equations (12) and (13), one observes that the coil input power P and the temperature differences AT are not uniquely dependent upon the same paraar i eters. That is AT can be changed by varying S, t, Ap , L, D and q even if P is maintained at a constant value, and therefore, it to impossible to control Ar by solely controlling P. Therefore to control AT one should measure AT directly.
Af ter determining, S, e, Ap, L, and D, we can, however estimate AT from coil input power P by assuming an appropriate n. This technique will be discussed later.
5.10 GEOMETRICAL CONFIGURATION The previoud discussions have been based on the induction heat treatment process applied to a uniform straight pipe.
j In actual plants, however, there are many geometrical transitions in the vicinity of a weldment. Typical transitions include:
- 1. Counterbore for weld edge preparation
- 2. 4>zzle-to pipe transition
- 3. Pipe-to-elbow transition l 4. Pipe-to-cap transit ion
, 5-25
In this paragraph, the effect of these transitions on the residual stresses pro-duced by the IHSI process are discussed.
- 1. Counterbore for Weld Edge Preparation The pipe end is usually counterbored to provide reasonable fit up for welding. As a result a thicknesa transition existo near the wold joint.
In order to investigate the effect of such a thickness transi-tion, a representative model having a gently sloping transition was evaluated using finite-element techniques, as shown in Fig-ure 5.15. The results of this evaluation are shown in Fig-ure 5.16 and are compared to a smooth straight pipe. It is clear that the same amount of residual stress improvement is expected for the pipe having a gentle thickness transition as for a smooth pipe.
- 2. Nozzle-to-Pipe Transition Since a nozzle has a thicker wall than the connecting pipe, there usually exists a steep thickness transition. This steep thickness transition effect was also evaluated analytically by finite-element methods. The finite-element model is shown in Figure 5.17 and the results are presented in Figure 5.18. All internal pipe surfaces under the coil exhibited an improvement in rooldual stress distribution resulting from the IHSI process.
This type of thickness transition in the 24-inch demonstration mock-up test joint, J-1 and the residual stresses measured fol-lowing the IHSI treatment are plotted in Figure 5.19. The results demonstrate that the weld residual stresses on the pipe inner surface following an IHSI treatment are compressive in agreement with the calculated results.
It can be concluded from both the analysis and experiments that even a steep thickness transition has no significant effect on IHSI results.
- 3. Pipe-to-Elbow Transition For this case, the heating coil will usually be placed so that the weld joint is located near the coil end in order to avoid interference of the coil with the elbow. 'But it is necessary that the heating coil covers a range of approximately 2t (t:
thickness of elbow) from the weld center in order to obtain a sufficient temperature difference between the outside and inside surfaces of the elbow at this location.
5-26
I t
l l
I
. I (.6. 0 137.85 l-1 OIUWE *
(mq FHOM HODEL 400 121 //'/ / // / / 126 cEnrEn /
' ' 5-./
175 '
o e
i / 3 l D us 150 91 '96 Z WATER TEMP. /
l 20*C
@f
/.
125 85 ' 90 -
WATER VEIDCITY /
0.72m/SFC f
/
112 79 /48
/
102 73 .
78 90 67 lillf 72 i
4 80 61 lll& 66 70 55 60 on 60 49 54 p
1 /
4 48 43 48 9 #
' ./
36 37 42 p
5 30 31 36 a-i 24 25 30
/
/
i 18 I? 24 ,
}/ 1 12 13 18
/
6 7 12 O ~
@ #6 1
R 0 343,o ,/ / ff /. (mm) 142.13 146.41 153.69 154.97 159.25 rig. 5.15 ANALYTICAL MODEL FOR TEMPERATURE DISTRIBUTION t
i 1
5-27
}
40 -
125 1
30 - n a vg pboy %
e -c' s h* 12IN SCH100 9g g b, 'N t 20 - l te ' -*- AXIAL STRESS AT ID
't 4 ,
's , \ l g4
~
E
- ,# ' b jd'
\
" " " 8'**SS ^' '
AXIAL STRESS AT OD
's s 1 -
N 10 -
---o---
8 t 'g
.nc g
1 e
~
s ---o--- HOOP STRESS AT OD Y k 0 '. D ,6N9#3h %_ g _ ,
$ h '
l'00% / s g. ' .M* 3'0 0 460
^ U U E I"I 125 7, /* [# '
/.-rfo - pi %
12IN SCH100 mM a O AXIAL STRESS AT ID EEEme" ,
N' ' # HOOP STRESS AT ID
---S--- AXIAL STRESS AT OD
~~'@~~ HOOP STRESS AT OD Fig. 5.16 RESIDUAL ' 'ISTRIBUTION FOR GENTLY SLOPING , TRANSITION
1 "4 400 _.7HEATING COIL 4 6 1
k?N -,.-. ..
a L _ _ __ _ __ __ _
i 45*
I ANALYTICAL MODEL
.N
\nr.
I f E8
\l\
9s N , 400-l l \ N\n a
N l
t, , x g 100 l I l
m
\ ,3x \[,hN200 5 500 100 i
w , e xA ,, \ y300400 00 - - -, ,- -- -,
400300,-200 n
dixd, N{g N Nji- - h: !
s'
- -,r .,
, [
C
,c,
.6 ' , '3} ',' ff!gr fc ' ,c, e oc rs c (%,%
e ic--e: ,.'
n es
--r.
1,1 g:
n uz-r-
~3
- s_ :
cs o
1 j i 4
WELD JOINT "
m s
n 1
i'
- Fig. 5.17 t CALCULATED MODEL AND TEMPERATURE DISTRIBUTION FOR NOZZLE TO PIPE j TRANSITION I
i
? -
y -+,9-- - ,
t -rm_a_ -
i
- AXIAL STRESS AT ID os
, 's
' s c
30 - i
,o r ,e ' o. _, . .-o- - -o - - o HOOP STRESS AT ID
/ i g i i g g a f g g
- --o- - - AXIAL STRESS AT OD ll s . . - - L 's HOOP STRESS AT OD 20 - *s ' * - -e -' '
-e '# 'g 's - - + - -
\ 'es yl s
i
\
g \
is g s 30 g \
10 - N f 18 g E /$ \
o N d /f N N 'o' t o.
/ COIL ,.os x
U, g_ , e ' .f I l \
s s 'n -
A ~~-o
- 9 o w-cr
, g < .A _o/ -100 1
ld0 2d0 300 i
400 50[
i (mm) s u) <
w ca (4 WELD SEAM E*
m o
\o ^
O O -
e F
Fig. 5.18 RESIDUAL STRESS DISTRIBUTION FOR NOZZLE TO PIPE THICKNESS TRANSITION
369 49 ,
measured icint ifEATING COIL
[
v x x
- F 4 -
1 m
COOLING WATI:P_ -
O measured (OUTSIDE) 8 0 a.
C D calculated (INSIDE)
(: 600_
- )
O~
F 400- -
a 5
200- -
0 O O 50 60 4'O 20 0 50 40 60 8'O 100 l'20 140 l'60 DISTANCE FROM SEAM CENTER (mm)
O ID flOOP STRESS
- 40_ 4 ID AXIAL "
~
3m -
E-2 0.. 4 -
$ 10- -
N m 0 , ,
a 10 30 50 150 C. -0100 l -
a ,
2gr \" -
w x
~30-rig. 5.19 J-l RESIDL';.L STiiESS DISTRIBUTIO'; FOR NOZZLE TO PIPE TRAt;SITIO!;
e, - 31
The bend radius of the elbow is usually 1 x D to 2 x D or more, where D is the nominal pipe diameter. The thickness t is usually smaller than one tenth of D. Therefore, the heating area of 2t is much less than D. In this case, the assumption j may be made that the heated area of the elbow is an extension of the mating pipe and that the same residual stress improve-ment can be expected. Figure 5.20 shows the experimental results for a pipe-to-elbow weld in a demonstration mock-up test. The residual -tresses were reduced to less than 5 kg/mn8 tension as expected.
- 4. Pipe-to-End-Cap An end cap, usually hemispherical in configuration, is some-times welded to a pipe. In order to evaluate the applicability of IHSI for a weld of this type, computer calculations were performed using finite-element techniques. The calculated model and the results are presented in Figures 5.21 and 5.22 respectively. It is anticipa*ed, based on these modeling results that the residual stress of the pipe-to-cap weld can be improved.
5.11
SUMMARY
OF PARAMETERS The aforementioned discussions are summarized in Table 5.1 with the recommended controlling ranges. It is noteworthy to recall that the maximum ei.ect of the residual stress improvement process is expected when the parameters are maintained within the recommended controlling range and that the IHSI effect does not disap-pear even if these parameters are below the specified value; but the effect may be reduced from the optimum value.
5-32
190 60 ,
llEATING COIL l g l (R-2h ONLY)
I 4 # r m o PIPE-20 -
10 10
,s', - ~,
'g _ ~~
d
] 10 - (I) HOOP
- A3 WELD (R-25)
$~ l NU H<
0 " '
'4 $ ,
$S
- a SS -10 -
/
n' x / /
/lIHSI (R-28)
/
-20 -
/ / // .
10 0 10 20 DISTANCE PROM !.* ELD { (snm)
_ 20 -
e-~~t's s I
N (II) AXIAL s k - N s
s s
$ 10 AS WELD (R-25) N d -
N G l//
$N
~
10 0 /j' 10 ' / / /20 a$ _ !4// / / 77*
jg /
/ DISTANCE FROM WELD { (rnm)
~I ~
.' /
IHSI (R-28) /.
/
-20 rig. 5.20 RESIDUAL STRESS DISTRIBUTION FOR PIPE TO ELBOW WELD S-33
250 330 _
V////////////////////////////////////////>\ HEATING COIL f : Pr '_ ,- .F%
F-- l: ': T +'t
]+,
l -
1 b-5. _ I-[ E INiDIN Z DIRECTION f WELD JOINT
- RESTRAINED WATER TEMP. 20*C \
w Um 0.3 m/SEC. .~
\\
- s. U l
ll i
'/////////
R DIRECTION RESTRAINED i
Fig. 5.21 ANALYTICAL MODEL FOR PIPE TO END CAP JOINT 1
i
40 HEATING COIL I I 30 - _ _ . _ AXIAL STRESS AT ID STRAIGilT PART ROUND PART
_o- If00P STRESS AT ID
- ~~*'
20 ,e
.o.g,- Ag , AXIAL STRESS AT OD s
,o- o -o. ---o-.
A-E. ,/ -' '
- g 8
.f: I
..o*
s t
--.-o--. IlOOP STRESS AT OD
$ s' , * '
"4 5 10 -
g , .- ,
i,'
8 l '.,9e 0 I #
w s
w a
0 68 7J00 - ~ ~AD 8',
" '30'O 400 500 WELD SEAM i
s'*w'/
o,' -g M u
-o g 'o -
, -e.
Q w -e
$ -10 M 'N
' r.
o N
p'*
/
D C 11 4 -
Fig. 5.22 RESIDUAL STRESS DISTRIBUTION FOR PIPE TO END CAP WELD
'T Table 5.1
SUMMARY
OF PARAMETERS FOR OPTIMIZED EFFECT OF IHSI tlumber Parameter Controlling Range Remarks 4cy(1-v) 1 Tamperature Difference AT >
Ea 2
2 Heating Duration T > 0.7 h 3 Coil Width L~ > 3/E 4 Coil Incation whichever is greater x > 15 nun or f 5 Maximum Temperature To < 550'C 4
6 Frequency -----
Control is not required 7 Pipe Size -----
Automatically controlled by other parameters 8 Cooling -----
To be sufficient to get enough AT 9 Cbil Input Power -----
To be sufficient to get enough AT 10 Geometrical Configuration ---
oy Material Yield Strength (kg/mm 2) v: Poisson's Ratio E: Young's Modulus (kg/mm2) as Thermal Expansion Coefficient (mm/*C) t: Wall 'Ihickness of Pipe (mm)
R: Mean Radius of Pipe (mm) t at Temperature Cbnductivity (mm 2/sec.)
x: Distance from Weld Center to Coil End (mm)
T:
o Pipe Outside Temperature (* C) 5-36
Section 6 EVALUATION OF OTHER EFFECTS OF IHSI TREATMENT In this section, other effects resulting from an IHSI treatment are discussed.
The IHSI treatment is one of the heat treatments of pipe, so it is necessary to consider changes in metallurgical and mechanical properties _resulting from the application of such a process.
6.1 SENSITIMATION EFFECTS The pipe inner surface, which is the region of concern for IGSCC resistance, is continuously cooled by water during the IHS! process, so the temperature at the pipe inner surface will not rise above 100*C even in the extreme case. Because this temperature is lower than the reactor operating temperature, this temperature will produce no deleterious ef fect on the sensitization kinetics of the pipe inner surface. As for the pipe outer surface, the maximum temperature of the IHSI pro-cess is limited to less than 550*C as specified in paragraph 5.5. During the IHSI t re s tment, the heating time is limited to several minutes. Figure 5.11 presents a Time-Temperature - SCC diagram for Type 304 stainless steel obtained by CBB test III in 250*C dater with 20 ppm dissolved oxygen for 310 hours0.00359 days <br />0.0861 hours <br />5.125661e-4 weeks <br />1.17955e-4 months <br />. The data demonstrate that no SCC susceptibility results from this process. Consequently, sensitization is not expected to occur during the IHSI treatment.
6.2 MECHANICAL PROPERTIES The IllSI treatment produces a small amount of plastic strain at the inner and outer pipe surfaces. In this section, the effect of that plastic strain is discussed. Since the maximum temperature at the outer surface is limited to less than 550* C and water is used as the cooling medium, the maximum temperature clifference between the inner and outer surface will be no greater than 550*C.
Re f e rence: (1) M. Axashi and T. Kawamoto, Boshoku - Gijutsu, 27, 165 (1 978).
5-1
Thermal strains produced by such a AT can be approximated by using the following equation.
c 0.5 a AT
~
= 0.5 x 1.8 x 10 x 550
= 0.005 No detrimental effects are expected from this plastic strain since this strain is very small compared with that produced in usual plastic work. An example of a mechanical test on a 12-inch pipe weld joint is presented in Table 6.1 and Fig-ure 6.1. No significant differences were observed on the weld joint tension test, root bend test, side bend test and hardness when comparing the IHSI weld with the non-IHSI weld. Consequently, no detrimental mechanical property effects are anti-cipated when employing the IHSI process.
4 6-2
TEST ITEM CONFIGURATION JOINT NO. IHSI T N TH * "^ " 9 (kgf/mm2) ROOT BEND SIDE BEND TEST TEST BASE METAL DEPOSIT. M STRAIGHT PIPE R-25 W/O 57.1 (MIN.) W/O. DEFECT W/O DEFECT 198 213-
, 58.7 (1%X.) W/O DEFECT '0.3'x 1*
e ELBOW R-28 WITH 57.7 (MIN.) . W/O DEFECT W/O DEFECT
& -215 224
.59.7 (ImX.) W/O. DEFECT W/O DEFECT
- } WELD DEFECT SIZE (mm)] x NUMBER Table'6.1 MECHANICAL TEST RESULTS 12-IN. PIPE WELD JOINT
12IN PIPE A ELBOW 260- _. I + R-2 IDE "ji ,
.'s } AS-WELD 2 -o-- R-25 INSIDE
, "t C - R-28 OUTSIDE y 24 7 B ) WITH IHSI o . -wk- R-28 INSIDE
^
22a '
t I" s i l,s,*ii \ 's, x- ,s
, s i
s E
g 20c //
\A y w
).
,/ t' ~ </ '
< s A ,.
- %, ,e-
's
~
18 & '
s /~ ~ A M
N.,'
2mm 160-BASE METAL DEPOSIT METAL
~
(SUS 304)
Fig. 6.1 HARDNESS DISTRIBUTION IN 12 IN PIPE WELD JOINT
Section 7 FIELD APPLICATION When applying this process to the field, it is necessary for the confirmation of IHSI validity that key parameters are controlled within the specified range. Two key parameters, coil width and coil location, are verified prior to IHSI applica-tion. The other three key parameters are controlled during the IHSI process.
It is required to get a temperature difference greater than the minimum required while maintaining the maximum outer surface temperature not to exceed 550*C. The tamperature difference between the outer and inner surface of a pipe and the maxi-mum temperature of the pipe outer surface depend on the coil input power so we must choose a suitable input power. The required coil input power which is neces-sary to produce the required temperature difference is estimated initially. Trial heating is performed to confirm that the power produced is suitable.
Usually the temperature of the pipe surface cannot be measured directly. Cons e-quently, an estimate of the tamperature of the pipe inner surface is required.
The method for performing that estimate follows.
7.1 ESTIMATE OF INPUT POWER As mentioned above, the coil input power determines the time temperature relation-ships when the input power is larger than required, the heating duration will be less than required since the outer surface temperature reaches 550*C too rapidly. When the input power is less than required, the temperature difference between inner and outer surface will be insufficient. So the optimum coil input power should be selected.
7-1
The required coil input power Pc is proportional to AT, D, L/t where AT: Temperature difference D Pipe diameter L 0011 width t: Plate thickness Test and calculation results relating the above parameters are shown in Fig-ure 7.1. Required generator power Pg is obtained from the quantity Pc/n, where n is the efficiency which must be determined experimentally. For example, the required generator power Pg for heating of a 28-inch pipe is calculated as 340 KW as shown below.
AT = 400*C Pc = 2 90 kW from Figure 7.1 L = 450 mm Pg = 2 90/0.85 = 340 kW D = 707.7 mm where, 0.85 is the value of the heat transfer t = 37 mm efficiency 7.2 TRIAL HEATING The coil input power is fixed after the trial heating. The trial heating is per-formed to verify that the coil input power determined from Figure 7.1 is proper.
The trial heating is performed such that the maximum outside surface temperature is less than 280*C (operating temperature), therefore, minimizing material prop-erty effects. In Figure 7.2, an estimating method of long time heating condition from trial heating data is presented.
Since a sufficient width of the pipe outer surface is covered by the heating coil and the pipe inner surface is cooled by the flowing water, it is possible to pre-dict the temperature by idealizing this situation as follows:
Assume an infinite length plate whose initial temperature is 0*C. When heat is introduced at one side wnich is under insulated conduction and the other side is maintained at O'C, the temperature of the heated side af ter T seconds will be given by the following equation.
7-2
Pc : COIL POWER (KW)
A: Calculated AT : TEMPERATURE O: Mens >1 red DIFFERENCE (*C)
D : PIPE DIA (m)
L : COIL WIDTH (m)
Pc (kw) 300 - t : THICKNESS (m) 200 -
0 0
100 -
-'- i i i
1 2 3 x 10 3 AT.D.L t
Fig. 7.1 ESTIMATION OF COIL INPUT POWER 7-3
tot / Tom N
1 0
~
- j- 'o o o o ooooo o o'
-0.5
.4
, , , I , . . , ,1 . . . i . . . i e i 250 200 150 0.2 0.3 0.4' Tot - Ti(Tw) ( C) 50 x\_
\
~ t=20 t =25 t=30 i
100 -
7_
Oss 4 T(SEC) d,
- o Fig. 7.2 ESTIMATION OF STEADY STATE CONDITION BY TRIAL HEATING 3
s 7-4 f
.I
l l
l
, 2n-1)2 2 apr 2 2 T = 9k- I x (1 - e ) (14)
A t n= 1 p 2a (2n-1)2 2
~5 t
where T: Temperature q: Heat input 4 A: p Thermal conductivity of the pipe a: Thermal diffusivity t: Plate thickness n: Natural nunber T: Heating time Tb make Equation (14) dimensionless, divide by the value of TT== which is obtained from Equation ( 9) as Tt== = at an infinite time following heating p
1 TT=T 9 " TT==
_ ( 2 n- 1 )2,2 a1
= S- ""1 Y x (1 - e ) (15) n (2n-1}
2 The value g in Equation (15) has the meaning (Measured temperature of outer sur-face - Water temperature)/(Predicted maximum temperature of outer surface - Water temperature) . Figure 7.2 presents the graph of g. The value g is plotted versus Bourier Number in quadrant A of Figure 7.2. The ordinate of quadrant B represents heating time of the trial heating. The lower abscissa of quadrant D represents the measured temperature of the outer surface minus the water temperature and the upper abscissa represents predicted maximum temperature of outer surface minus water temperature. An explanation of the use of Figure 7.2 is presented below.
For a given set of experimentally determined trial heating parameters,. locate in Figure 7.2 the measured value of T on the ordinate of quadrant B. Move horizon- -
tally to the line for the plate or pipe thickness of interest. One may interpo-late for intermediate values of the plate thickness. From this intersection point. nove vertically upwards to the. line for g in quadrant A of Figure 7.2.
From t als intersection point move horizontally lef t to the measured value of the measured temperature of the outer surface minus the water temperature of the lower abscissa of quadrant D obtained in the trial experiment. The intersection point of this horizontal line with the family of curves representing the upper abscissa 7-5 e- 3 y - - - - - - -- _
v . - . , . . - - - .
of quadrant D produces the predicted maximum temperature of the outer surface minus the water temperature, thereby obtaining the predicted maximum temperature of the outer surface to be produced during the IHSI treatment.
l 7.3 ESTIMATE OF PIPE INNER SURFACE TEMPERATURE The effect of IHSI is influenced strongly by the temperature difference between the outer and inner surfaces. When IHSI is applied at a reactor site, the temper-ature of the outer surface can be measured using thermocouples but the temperature of the inner surface cannot generally be measured. So it is necessary to predict the temperature difference exclusively from outer surface temperature data and cooling parameters. The estimation method for obtaining the temperature of the inner surface is explained as follows.
The temperature difference between the outer and inner surfaces is given by the following equation.III AT = T, - T1 3
WS
= (-1 + t+e ~ !) (14) p where Wg Heating density on outer surface S: Heating depth Ar p
Thermal conductivity of the pipe t Plate thickness Reference (1) Appendix A-13 Estimation of Pipe ID Temperature During IHSI.
7-6
j The temperature difference between the outer surface and the water is given by the following equation.
AT2 = Ty - T _
~
=
{2 p
(-1 + +e !) +g (1 + e !)} (15)
I where hm Heat transfer coefficient From Equation (15), W o is obtained and when obtained, oW is substituted inte Equa-tion (14), and the temperature difference (T o -T)1 is obtained. It is very dif-ficult to do the above calculation at a reactor site. Figure 7.3 is presented to display graphically the above mentioned calculation, and is explained as follows.
Locate in Figure 7.3, the measured value of (T o - T,). Move horizontally to the line for the plate or pipe thickness. One may interpolate for intermediate values of plate thickness. From this intersection point move vertically upwards to the line for the value of plate thickness in the first quadrant. From this intersect tion, move horizontally left to the ordinate and read the value of (T -T).
o 1 7.4 EXAMPLES OF FIELD APPLICATION OF THE IHSI PROCESS In the case of :.he field application, it is necessary to record and to evaluate the results. Pour steps are required to obtain good results.
The first step is preparation. The second step is the trial heating to. confirm the key parameters. The third step is the actual heating. The fourth step is evaluation of the results. Figure 7.4 shows some examples of the IHSI results for actual primary loop recirculation piping.
7-7
1 I
}
(*C) 24 A Sch. 24B,5ch.
~
500
[ .100 p
[ 100
! 7
%128x Sch.100
=To-Ti !
310 '
/ '
300 200 - ,
100
/
4 0
30 40 50 60 TO 80 = 10'Fcal/dh wo 100 To-Tw 200 300 350 400 t
\
i
, -12 BxSch.100 500 \
h ] l
. 20B= Sch.100 24Bx Sch.100 Fig. 7.3 TE!!PERATURE DIFFERENCE BETWEEN OUTER Allo INNER SURFACE OF A PIPE
+
1 4.
7-8
OCT. 11, 1978 WELDING JOINT NO.
PLR J ESSENTIAL VARIABLES REQUIRED MEASURED SA376 MATERIAL TP304+ Sa240 TP304 COIL WIDTH I// Rig 2.7 3 288 nun 290m f P8 + P8)
TEMP. ON OITTER 456'C HEATING gyg 2 SURFACE at h TIME 0.8 "> 159S 300S ROOM TEMPERATURE 27 *C DI OF -"
ATL 3 203*C 343*C TEMP. 36 *C
. . OmR COOLING SURFACE 550'CN 483*C
^
- VEIDCITY ^
1.18 m/s OF COIL (T/2,'15cun) >15mm 128mm 288 DOWN STREAM . _60 160 l
If f 4#- #####l
._is .'
[
10,,,1,,g, MAX. 483*C
.-500'c
-2
_4T 4
-2 a
3-1 i -300
- 200 60sec. -100 i
Fig. 7.4(a) Record of Field Application of Induction Heating Stress Improvement 7-9
, , ~ , - , , - - , , - - - , - , - - - -
288 _
=
60 - -
160 _
b k h O \ N./ N y e @
' )
# m C
500 - 0 0 400- O n 300-b 200-l 100 - 100 200 300 400 nm Fig. 7.4 (a) (continued) 7-10
l l i. OCT, 10, 1978 hILDING JOINT PLR B - NO. ESSENTIAL VARIABLE REQUIRED MEASURED MATERIAL
- COIL WIDTH L/Mg 3.8 .2 345mm 380mm (P8 + [N TEMP, ON OUTER "^ "
S *C at/T 2 3 0.8 SURFACE at TIME * *> 146S 243S ROOM TEMPERATURE 25 *C I 288T EM . = COOLING TEMP. 37 'C MAX. TEMP. ON OUTER SUFFACE 550*C") 542*C VELOCITY 3.27 m/s W A ION OF >g$,, gL 4;, 190 190 42 1
, // ! // !! // /l (o s' A *:( ' ) Y k$ h
- llo so l DOWN STREJM max.542*c-i
, 3-1
- 500 i .
4
-2
- 400 4
2-2 2-1
- 300
- 200 i
60 sec. l
- 100 l
Fig. 7.4(b) Record of Field Application of Induction Heating Stress Improvement i, i 1 7-11
b 190 190~
= =
\( N -Q q OO GO(/3
- ~
G) g 10 80-
= = =
i 4 i O . 'C o ! 500 - 400 - 0 300 - 5 200 - 100 - i i , 100 200 300 400' nm i Fig. 7. 4 (b) (continued) 7-12 g
. _ , _ = - , - - _ . - - . . _ _ - . _ . _ _ _ . - - - . - . , . -
7.4.1 Preparation - Preparation is performed to check that the following items will be in the con-trolled range or have a suitable'value.
- 1. Planning
- a. Coil width
- b. Power (equipment . capacity)
- 2. . Preparation for heating
, a. Coil setting
- b. Thermocouple location to check the outer surface temperature
- c. Cooling condition (water temperature, water . velocity) 7.4.2 Trial Heating A trial heating is performed to check the following items below a maximum tempera '
ture of 280*C.
- 1. Suitability of power source
- 2. Adequate temperature difference in the controlled range
- 3. Hoating duration in the controlled range 7.4.3 Actual Heating Af ter confirmation by the trial heating process, the actual IHSI heating to 550*C is performed.
7.4.4 Fesults and Evaluation
- When the actual heating is performed, the results will have to be checked and evaluated. The following three key parameters are to be evaluated.
- 1. Heating duration-
- 2. Temperature. dif f erence, which is calculated by estimating the inner surface temperature
- 3. Maximum temperature at outer surf ace and temperature .distribu-tion in the axial direction, which is made graphically to con-
- firm that a good residual stress distribution is generated.
i 7-13
l i l i Section 8 PIPES WITH SMALL PRECRACKS 8.1 OBJECTIVE When applying IHSI to an operating plant, there is concern that a small crack may be present which is undetectable by inservice inspection using ultrasonic examina-tion techniques. In order to apply IHSI to an operating plant, it must be demon-strated that ti.e residual stresces are improved and that detrimental effects are not produced by IHSI even in the presence of small flaws. In.this section, the ef fectiveness of IHSI is examined experimentally and analyti-
- cally for the situation in which small indications may exist.
8.2 TEST DETAILS AND TEST RESULTS A test was performed to confirm that the residual stress at the inner surface is improved by IHSI even if small cracks exist. Here we assume that a small indica-tion is identical to a small crack. The test sequence is shown below.
- 1. 4-inch pipe is girth welded
- 2. Precrack formed by electron discharge machining 3a. IHSI treatment 3b. th treatment 4a. Immersion in boiling 42% MgCl2 s lution 4b. Immersion and oye penetrant examination NOTE: The depth of the precrack was chosen as 25% of the wall thickness. This 4
crack depth is nearly twice that of the calibration notch of the ultra-sonic test standard specimen. This depth is large enough to be detected by the ultrasonic test, so this assumption appears to be very conservative. 8-1
Figure 8.1 presents the results of the MgC12 cracking test on the precracked pipe with and without IHSI. As shown in the figure no stress corrosion cracks were observed after the immersion of the test pipe into Mget2 solution after IHSI treatment, but many cracks were observed on the test pipe without IHSI. Figure 8.1 also shows the stress corrosion cracks initiated at the artificial notch in the horizonal direction. Figure 8.2 and Figure 8.3 show the section of the artificial notch after the MgCl 2 test. The artificial notch was sectioned at the center of the notch length. No cracks were observed with IHSI, but several cracks have propagated in the depth direction at the notch tip of the test pipe without IHSI. 8.3 COMPUTER CALCULATIONS FOR PRECRACKED PIPES Assuming the existence of a small crack at the inner surface of a pipe, finite-element method calculations using the program ITEMP II, IEPTC II were performed to investigate the inelastic behavior of the crack under transient loading during the IHSI treatment. The analytical model is shown below. Pipe Size Crack Shape Case Nominal Outer Analytical No. Size Diameter Thickness Depth Direction Models (mm) (mm) (mm) 1 12-inch 318.5 19 1.4 Circumferential Crack Axis ymm. Sch. 100 i 2 2.4 3 4.2 In the above analyses, three different crack depths were selected. One was 2.4 mm 4 (12.5% of pipe thickness) and another was about 1.2 mm and the third was about 4.8 mm. The 12.5% deep crack corresponds to thr. reference hole diameter of an ultrasonic test standard specimen. Figure 8.4 presents the residual stress distribution for the axial stress near the tip of each crack following IHSI treatment. The residual stresses showed the minimum value near the crack tip and became gradually deeper as the crack depth increased. 8-2
1 (a) EDM - s\ps 0 G 0 [EDM: ELECTRIC ) 'c 6 0's\ s J3; DISCHARGE - - 9 9'g .
~
'\ MACHINING / ~~f U , ,0 <,
wa.r.
* [_ .
',[-c - T .-
.w i
(b) EDM e
'v' 5
\ MgcI2
- ; l'II;;,,Ilitishg ,
,,jaj 3 ,&,
~ ._ -
. ' i Propagation -
i 1
*ttf 2-i (c) EDM
, 'v" O \ 1HSl x lIlliItjig,'ll :1 '
,Zitsglit Mgc I 2 - -- : ;
v- Ulill No Propagation o gg' Fig. 8.1 Propagation Behavior of Notch 8-3
~'
'. (, - '
s->
, , . y . . a, . - .k - y,;j r *e . m, -
,.g kh,?bN * #
Y' I. . - ,
,<j? - .
. .k .3Z: .
- .;j..z. .. :. Y.Q , - g.3,N * : .' ? p, :u .i ' l~4~.. K Q<,., wg.
g-:
.. . .7-
'. M , A 1
t h,; N
., , y 3 f i<,f ,, .$,i
.: -en
, N;'i,';; ,
' P WELD DEPOSIT R-53 45
. .-*e. e
-- .. - y_g y- 4.. u..
4 . y%. '??. . , p% . , ,.,
,- .& $.-4, . . . 3 I t 7/_,..-
W .
. {. , \
t .i ' N, *1t--
- JL i r. .
'ig' &p.jW%
+N SN 4.Q .:f4
, ::-l'l> Ef. ,,,s *$',: k. R?;jy-;~;&.g $
6 9' ty '.,9 j ar
~ l U 6 *,. /-- yMs g ..w ,
Ja .
.i ,
.v h . . k . -Q g ;.,,? f~k 7.
~~ '
aWr~c . . ' qa,h' s
. . _o .-
l l R-53 225 j I' IG . 8.2 CRACKING OBSERVATION AT NOTCH TIP (R-53) i MAG x 100 8-4
g .
.m ,, .
et .s.. . 6
. . r7 . ., ,.
l l 1.'.4., -
.~. y .,w
?. .
t ,e . 4
. :1.: GU.. ,., s a
.. 4..f ,"- ....
- o. s .. --
- - - < ../
........v.,
- t. .e: & 3, ,
. .. . . - . , m, .- .
, . ~ .,, f ' . * , .-*
.~..*.j.., .
< j
, ' ~ ." >~. , -
n
, ' * [py . .a., ; .-_w.k
.. . ' i , ,' i *c ' ', ; f . WELD DEPOSIT
,~, *.,
~. .
*? 9r 4, *4an . -; p,'.T .; %
< ..Qgey~e4M;-fi y m*. *
. p T'y ~M_(
d.,f,.+el%y
,. .yqt< , p _ A .g/c-4 . .y-.. '
h'*. v~ $4gb$d'i d;.2.?'. i
.'s - ,
e.+v 3 d,pr -[- <x , y.. s .pl* [w.g. p,'b,l1. r .- *i ~. QPL -
.f. b *L..'. . ,,. a., ,.
* , lp ,e. '
f /d. p.s s. "- ;'*
, a s '. ^'.? -
- 7. pr . 7. :
x.q
,?. i -,pY. s:'1F.
- t. . .. .-
9:%' 4)%, . . 4. . f. -.se ' .:N:.., , '.,
'~
A. * *--
.b . c,u_xt 4..
ym~*);1* c ..u ;%. 3 a. ..W
~
4 . .- y ,,. s n e- .. f y..y s s. W L b{ A;q N.b
-. ~. .*. n. .
L1.T..g-(c, '/- ck N.' w [. tyNI 7 eM 5N.
-d,L y
. - s'
. g gv. g s
N7 N:"
- g.F.g%.:.
#7%}- j ,'F=..--
e, -
- g. ,
- y. \g. %,; .;,
.y c ,V 1* -_. .
-fd ,ff W ?':h Q j_\, < m # .' * ,.
-f
, u; n.g% . ..
- n .Y u .y . - ,7; .-W : a,% . ' _
.x -
.1, m <
. ;, . . , y 7% -
.y.~ 3,. 0_. % ' ~%y y ,5* "
W:/ ,i. 1- ,.'"
. NW% T ' * .&'.\ h -~ ~~~$
- I' x --- . ,-
1,e.: .
,.. c)p'4-3.a-{* % 4%W' 4eT., g.
.. %. t, 9. 13 n, *-
ro 5! 's ~ M '
'y%Q
,a w
2
< . ,, ?- f '. u.I..
., . . u - . , esse 2-- :v- ,y{,, A s.t g, ' ~'t' . ; T, -
?- frF. ~ /' p~, '
R-45 135*
-. Sr. e- .NN. y - ==4 .
. ve -
.** # e s 7 ~ 's ' ,M, 'A a 4 0 3'7 y,j '
..; f AW
, ,T. ' a '.. ..
... [c. ,.-q'd* S. .'.I'M_.,,-
l pa- *,
.M f. ,.
. hp _' og _ys R-45 315*
, FIG. 8.3 CRACKING OBSERVATION AT NOTCH TIP (R-45) i MAG. x 100 8-5
CRACK DEPTH l - 1.4mm
- - - - - - 2.4mm 4.2mm 50 o
0 g gf
/r l l l
.[ l l l1
\ l l
f'i/
/
s
, I e I i 1
l'
-50 -
g,l- -- II I
-100 0 5 10 15 19 WALL THICKNESS (mm) i l
FIG. 8.4 CALCULATED RESIDUAL STRESS THROUGH TifE TAIICKNESS OF 12 IN. PIPE WITH A FLAW 1 8-6
l Figure 8.5 shows calculated COD (crack opening displacement) at the maximum heat-ing condition during IHSI. As shown in the figure, the COD increases as the ' initial crack depth increases. In order to maintain pipe integrity during IHSI for the pipe with an initial crack, it is necessary that crack extension does not occur. To determine the critical CTOD (crack tip opening displacement) for Type 304 stainless steel, a
~
three point bend test was conducted. The test piece was 20 mm in height and 10 mm in width. The bending span was 80 mm. The crack depth was 2.4 mm and 4.8 mm. Figure 8.6 shows the relationship between COD, CTOD, and the deflection of the , . test piece. The observed CTODs are shown to be of similar value for the two crack sizes. The critical CTODs were about 1.22 mm and 1.30 mm for the crack depths of 2.4 mm and 4.8 mm respectively. 8.4 EVALUATION The cracking test in 42% MgCl2 solution and finite-element calculations showed that IHSI can improve the residual stresses at the crack tip as well as on the pipe inner surface even if there is an initial circumferential crack. From the finite-element calculation, CTOD at maximma temperature during IHSI is about 30 x 10-3 mm for 2.4 mm crack depth. On the other hand, the critical CTOD is about 1.22 mm for 2.4 mm crack depth. From reference (1), the critical CTOD is about 3 mm. It is clear from the comparison of these CTODs that .the safety factor is quite adequate. In reference (2), an interesting calculation is described. The purpose of the calculation was to deterline if with the presence of a small crack at the inner surf.tce of a pipe, IHSI could be applied without detrimental ef fects and if under the operating conditions compressive residual stresses existed in the vicinity of the crack tip. Re fe rences: (1) Mechanical Fracture Predictions for Sensitized Stainless Steel Pip-ing with Circumferential Cracks, Section 4, EPRI NP-192 Sept. 1976.
'2) G. Yagawa, et. al. , EPAS Finite-Element Program for Analysis of Nonlinear Behavior of Nuclear Power Piping, 5th SMiRT, Mb1K, 19'79.
8-7
(( ! h ( f O@p ) 5 b, t 1 r N '# N 40 1 O b
- a. 1
\
30 \ 20 % i U 10
'*4*#g y k
, 1 0 1 2 3 4 19 WALL THICKNESS (mm) j FIG. 8.5 CALCULATED COD AT MAXIMUM HEATING f
f l 8-8 4
, . . - , . - , , - - - - . -- .-,,- -. ---~ -,- ,
1 3 v
/
l 2 j / ~ / l M / t / o G a / 09
'/CTOD
/
h / / Of / a , h 1 / / $ i
/ /!
g l / / y ,0 /
/
/
/
d
/ //
0 1 2 3 4 5 DEFLECTION (mtr ) FIG. 8.6 PELATIOr4 OF CnD A'JD CTOD TO DEPTECTION T N TliftFF. POINT REND 't rg7 ti- 9
P/ E n f
't x 10- 'm - o - RESIDUAL COD -o- IHSI + IN FRNAI, PRESSURE (100 kgf/cm')
30 -
---o--- COD AT MAX. TEMPERATURE - o- AS IllSI -
O s (DURING IHSI)
's s 2 = 0.05 s ,
\
20 - s s M s 50 o 's
- N. g c-o s s % s $
10 - o g a R $ 0 o' E s
- i. -
0 M-7,o. 0 0 0.5 1.0 1.5 2.0 2.4 oV DISTANCE FROM INSIDE SURFACE 4*
}
k
' /
RESIDUAL COD 1 3 i , 4 t STRESS IN THICKNESS DIRECTION Fig. 8.7 COD AND STRESS DISTRIBUTION
- R-11
1 Section 9 i RELAXATION i 9 9.1 RELAXATION AT CPERATING TEMPERATURES There is concern that the residual stresses induced by IHSI may relax during long , servica at high temperature, but liutle data exist which can be used to estimate the relaxation behavior at temperatures near 300*C. A test was conducted to determine the relaxation behavior of residual stresses due to heat treatments at 300* C, 400* C and 500*C for 24 hours following an IHSI treat-want. Figure 9.1 shows the experimental results comparing these heat treatment conditions to an IHSI treated condition. No apparent difference was observed among these data, showing that the residual stresses will not disappear or c..'nge due to service temperatures. 9.s RELAXATION RESULTING FROM APPLIED AXIAL STRESS It is well known that residual stresses are relieved by overstressing. This sit-untion is illustrated for pipe af ter IHSI using the two-element mode'- presented t in Figure 9.2. The pipe is assumed to be composed of two elements A and B which have tensile and compressive residual stress respectively. The stress and strain relationships of A and B are shown in Figure 9.2(b) . Curve "C" shows the relationship of composite material (A + B). 4 i
- when the pipe is strained to less than point 1 and then unloaded, the strain will recover entirely and no residusi strain will remain. If the pipe, however, is j strained beyond point 1, such as to point 2, the permanent strain
- c remains af ter unloading. In this case the residual stress is relaxed to point 4. Using this diagram, the remaining residual stress af ter removing axial stress is plotted 9-1 l
vs.
R-15 as 1HSled R-73 af ter 300*C24H P-74 after 400'c24H R-75 after 500'C24H 30
~
~ E ~A 20
* ' f'* g' 5
'f, ' ,
% - A
%e c
M
~
~
0 y -~
- n g"
-10
/'/..
Q w
-20 .-
h ,~-C" "h
-30 _,
-40 0 20 40 60 80 100 120 DISTANCE FROM HEATING COIL CENTER (mm) a 30 E
ho 5 20 , _ , n N R:--n - n
- '~, ,**
s ' ~. . :."
~ -
N o Y E # A g -10 , p s W--- -- a / .- f d'4 N. - E ' E$ -20 ,*
-30 0 20 40 60 80 100 120 DISTANCE FROM HEATING COIL CENTER (m)
Fig. 9.1 Residual Stress Distribution Comparing Three Heat Treatment Conditions to an As IHSied Pipe 9-2
FREELY EXPANDED I
/ \
_.. + > B % U
.a PIPE WALL ELEMENT pgpg IN PLANE STRAIN COND.
(a) TWO ELEMENT MODEL y (A) @ w Oy ~ ve Q &=BT cg -
@\ p
/
/
_ _ g STRAIN o (b) STHESS STRAIN CURVE
-cy -
D ta AXIAL EXTERNAL STRESS O$ CY-% OY xm i
'l
-:c
~
(c) RESIDUAL STRESS rio. 9.2 RESIDUAL STRESS ANALYEIS BY TWO ELEMENT MODEL 9-3
versus axial stress in Figure 9.2(c) . The results show that the residual stress is gradually reduced to zero as the external stress increases. i , Figure 9.3 shows a load-strain relationship measured on the inside surface of a 14-inch, welded pipe at room temperature.III The area very close to the weld exhibited lower strain than the base metals that is the strese-strain behavior of e the material near a girth weld was very different than that of the base metal. This suggests that stress relaxation of such a region near a weld will be differ-ent and possibly smaller than in the base material. Testing to confirm the above mentioned hypothesis is r.ow in progress. References (1) P190, JWES-AE-7807, Japan Welding Engineering Society, Nov. 1978. 9-4
--- w--.e w , , , -- . - - - - . ,- , . , . - -
Y _ 800 Omm 8mm s*
\/ O 30mm _
a O C -, a y 500-A A ~ 1 b i x - NOTE ax TIIE NUMBERS ON TIIE CURVE SilOW TIIE DISTANCE FROM WELD CENTER a i i i i l i , , , I i , 0.5 1.0 AXIAL STRAIN (%) FIG. 9.3 LOAD-STRAIN CURVE OF WFLDED 14 IN PIPE (EXPERIMENT) 9-S
l l l Section 10 CONCLUSIONS Many basic experiments, full size mock-up tests and elastic-plastic finite element calculations were conducted to evaluate the IHSI process and the results lead to the following conclusions.
- 1. Key parameters which have significant effect on IHSI are f a. Temperature difference between pipe outside and inside j surfaces l b. Heating duration
- c. Coil width
- d. Coil location in relaxation to weld seam
- e. Maximum heating temperature
- 2. A simple method to predict coil input power and pipe inside surface temperature was developed and has made IHSI a simple and reliable technique.
- 3. IHSI is valid not only for smooth pipes but also for pipes having preflaws which are not detected by NDE.
- 4. The beneficial residual stress effects of IUSI (imposed com-pressive residual stress) will not disappear during normal plant operation.
10-1 ee
l A-1 IMPROVEMENT OF RESIDUAL STRESS PATTERN IN PIPE BY HIGH FREQUENCY INDUCTION HEATING
- PRELIMTNARY TESTS -
l ( i t o.e -
.) Summary Preliminary studies to improve the residual stress pattern were performed using the High Frequency Induction Heating Technique. Three 4B Schedule 80 SS pipes, each having one circumferential weld joint, were tested. The following preliminary resulta - ra obtained: High tensile residual stresses in the inside surface of . pipe weldment can be reduced by High Frequency Induction Heating. I 2 w - - - - - - .&*-,
- ., y , y - . - - - - ,.---;7
I l
- 1. Introduction This paper presents experimental results for the High Frequency Induction Heating (HFIH) treatment of weld joints of 4B Schedule 80 austenitic stainless steel pipe. The parameters evaluated in this study include:
1 (1) Max. temperature (2) residual stress Post test metallography was performed on the induction heat treated samples. The test matrix for this experiment is shown in Table -1.
- 2. Theorstical Considerations A long circular pipe heated on its OD and cooled on ID will produce a radial temperature gradient and the maximum thermal elastic stress will be
; given by the following equation, outside Ea AT o max = 2(1-v) inside AT Where E; Young's modulus-a; Coefficient of thermal expansion AT ; Temperature difference V ; Poisson's ratio l
1 1 1 The stress is proportional to the temperature difference AT. When AT becomes greater than 20 (1-v)/Ea, a nutx exceeds the yield strength and plastic flow will take place. i . 3
In this case, the axial stress is distributed as illustrated in the figure below.
-es ?
% l / )
) N \ l / \
l +6; M
~'
cy ; Yield strength , AT Stress distribution Temoerature distribution Terminating the heating and cooling process produces a uniform temperature distribution in the pipe. The thermal stress will be relieved, but some residual stresses will have been produced in the plastically deformed region of the pipe by the prior heating and cooling operation, o l +
\ U l l
L (! ) ( /
/ -
i
/ AT=0 W
- 3. Experimental Approach 3.1 Material Chemical composition and mechanical properties of Type 304 SS pipes are listed in Table-2.
4
i 3.2 Weld Edge Preparation Joint geometry of the specimens is shown in Fig.-l. 3.3 Welding Conditions Table-3 summarizes welding conditions and processes. 3.4 Thermocouples Fig.-2 illustrates the location of the thermocouples on the pipe surface used to measure the temperature history during HFIH. 3.5 Heating and Cooling System Equipment used for HFIH and cooling of tLe pipe inside surface are illustrated in Fig.-3.
- 3. 6 Measurement of Heating and Cooling Cycles Heating and cooling cycles were measured at the points shown in Fig.-2 and recorded using a pen recorder.
3.7 Measurement of Residual Scresses The residual stress of pipes was measured by using strain gage and also checked by 42% boiling Magnesium Chloride Solution (MgC1 )* 2 3.8 Metallography Each cross-section of the weld joint was etched in 10% oxalic acid solution and examined by conventional light microscopy. t 5 I
- - -. - - . -,-v- -- - - -w. ..- _
_. - _- . = j
- 4. Results and Discussions i
4.1 Maximum temperature i The distribution of the maximum temperature produced by HFIH is plotted in Fig. -4. 4 4.2 Residual Stresses Profile and Crack Pattern l l Fig.-5 and Fig.-6 present the effect of HFIH on the residual stress l i profile of 4B SS pipe inside surface. . High tensile residual stresses on the pipe surface are observed to be reduced by HFIl!. Photo No-1 and 2 show the cracking pattern of 4B SS pipe treated in 42% boiling MgC1 s lution. 2 No cracking was observed on the ID of the pipe with HFIH while many cracks were observed on the ID of the pipe without HFIH. 4.3 Meta 11ography Photo No-3 shows the microstructure of the 4B SS pipe weld joints treated with HFIH. a No significant microstructural effects are observed since the maximum J temperature is lower than 5500C.
- 5. Conclusion i
Residual stresses at the inner surface of a 4B SS pipe can be improved by HFIH.
.?k) significant microstructural effects are observed for the 4B SS pipe >
when HFIH is applied. 6
, - , - . , -- -,,,,-y. . - - - - -- - -- - - - - .y, .- - - . . ,,--, --c -,
TABLE-1 HFIH TEST MATRIX (PRELIMINART TEST) MT'L H.F.I.H CONDITION *2 HAX. TEMF. ITEM OF ANALTSIS CERTIFICATE AFTER FIPE JOINT WEIS IN BEAD CEN1ER ('C) TP-No' a
*3 RESIDUAI. STRESS I E kW A V 00 10 A T (00-ID) TTD lh MICR0 m mo R-1B 540 168 372 O- / O /
R-2 45 SCH.80' 5M 170 M6 O O f O q R-3 % 8) 42 120 400 340 88 252 O O f O
///// / // o /
=0TE -
*1 WEtalleG CONDITIONS PRIOR TO N.F.1.H. ARE SHOWN kN TABLE-3. .
*2 DETAILS OF H.F.1.ll. CONDITIONS ARE ILLUSTRATED IN TABLE-4
*3 TTD (TIME-TtetRATURE DISTANCE CURVE)
TABt.E-2 CHEMICAL COMPOSITIONS AND MCHANICAL PROPERTIES OF BASE MAT 2 RIALS i gg gt CtfEMICAL COMPOSITION (Ut %) M!CHANICAL PROPERTIES MATERIAL
- C St Mn P S Ni Cr gg y 2) (gfy 2 3 y
'* 1,71 3,33 , TTC5618 0.05 0.58 0.023 0.007 9.2 19 24 57 70 i
4 e f 4 T
nAa l l ( \H
\ s S
co.
- m. .c o V U3 rA s 2:
% H 2
, 9 '
f 150 _ 150 _ 300 a Joe +2.So ~
; 210.4 1.6R MAX.
I m
- e. 4' o
+ 44 . e.
N N't ' c0 o n i m 40 0 cn 10' MAX. o u i DETAIL "A" FIG-1 CONFIGURATION OF WELD PREPARATION l 9 i
TABLE-3 WELDING CONDITION TP-NO. PROCESS POSITION HEAT INPUT (KJ/cm) COOLING METHOD AVERAGE 9.9 % 10.7 (lP % 3P) 10.2 R-1B 21.6, 24.1 (4P, SP) 22.9 g 9.9 % 10.9 (lP % 3P) 10.2 R-2 23, 23.7 (4P, SP) 23.4 GTAW (.A) 1G NATURAL 9.7 % 11.3 (lP % 3P) 10.3 22.7, 23.6 (4P, SP) 23.2 9.6 % 10.5 (lP % 3P) 10 R-4A 23.5, 24.7 (4P, SP) 24.1
i HFIH COIL , i 11 2 1 n. J .i m 1 i
) o nn 1 #N 11 I l n i
dr
'r
/{
\ w
{ il i N* 30 10 Ny
'%zo, 4
Fig. 2 LOCATION OF THERMOCOUPLES 11
l , 150KW MG 400KVA __ TEST PIECE TRANS PRESSURE
\ GAUGE e
N
==
e_ .
=d N, .. CITY WATER i
f 1 E b - FLOW METER SPRAY C"N
=
HEATING COIL i SCUPPER l I 5 FIG-3 H% TING AND COOLING EQUIPMENTS x t
i 600-8 R-2 (OUTSIDE) 500-U R-18 (OUTSIDE) 400-
$ U U o o
$ x "
h-3 (OUTSIDE) $ I 7 7
" 3"
. t 3 a o
c4
, - N R-18(INSIDE) 7 7 $
% M 11 20b o R2 (INSIDE) b W
100-s #' - x '~~sy N
\ xx R-3(INSIDE) 0 l I I 0 10 20 30 4l0 5l0 DISTANCE FROM WELD CENTER (mm) k HEATING COIL
/
TEST PIECE (4IN S.80 SUS 304TP) l CL Fig. 4 Distribution of Maximum Temperature Produced by HFIH l 13
CIRCUMFERENTIAL DIRECTION 4U l l AS WELDED (P-4D) N / 20 "SW hk A / y* \ t L j)
$ -20 '
$E o VM' \ HFIII AFTER PIPE
$$ JOINT WELD (R-2)
-40 0 20 40 60 80 AXIAL DISTANCE FROM WELD CFNTER (mm)
DIRECTION 4 I I
~
AS WELDED
~m (P-4D) lN ~/
gg 20 28 as \ / , /,
!! k __
/
/
kN g; -20 NQj7 g MH liFIli AFTER PIPE
$ JOINT WELD (R-2)
-40 0 20 40 60 80 DISTANCE FROM WELD CENTER (mm)
\ 4IN SCH.80 SS PIPE k V j IIEATING COIL v,,/ :,,,,,,,1,-//n Fig. 5 EFFECT OF IIFIII ON RESIDUAL STRESS OF 4IN SS PIPE 14
n-I i CIRCUMFERENTIAL DIRECTION l 40 j l
^
AS WELD a w (P-4D) 20
=
eim mS
$5 0 m \
g N$ Hz
-20 1m @H N
liFIH AFTER PIPE j JOINT WELD (R-3)
-40 0 20 40 60 80 AXIAL DISTANCE FROM WELD CENTER (mm)
DIRECTION 4 I I AS WELD (P-4D)
~ ~w 20 HW V' Ds \
ES
'w 0 $O U
NbM HFIH AFTER PIPE yZ JOINT WELD (R-3) N I 5 -40 0 20 40 60 80 DISTANCE FROM WELD CENTER (mm)
\ 4IN SCH.80 SS PIPE A"
v--1, // //111/ W " Fig. 6 EFFECT OF HFIll ON RESIDUAL STRESS OF 4IN SS PIPE 15
;,_ -.. a--u 2.h. , . _ _ . . ,
l i l
$ bg g g== g 94s v: . .,
- r. .3. . , ,
%f
'* y (
_.1 ' [.
.- ~ 3, ?
g it ? . c . n f 4 {4 ,
~ uc.a i -
; s4.
I '" *
+.
l
}. g
~,
l AT .,
- w
%~ ' 44 4 '
3 . $! w w I
. :. . . .. ~ . . . - .- .c. -
PHOTO-1 CRACK PATTERN OF 4 IN. SS PIPE WELD SPECIMEN (WITH HFIH)
1, l
= -
1-h, l l G e U e ,-
.g
~
d, I A l' . l 4 Pif0TO-2 CRACK PATTERN OF 4 IN. SS PIPE WELD SPECIMEN ( AS WELD) i
m _. ,
' ^
; } { Il l-
{l
.i, y ),
. {-
.p
..a! t.i ....
c' .
;f h m ~;
N - 3
' Ljf .
e, t: {'? y [, I .g
~
. - .l 4 E
.{
fc'.
- . ..i 4 - +.
y -
.O ljh% I.g x ,.o \$ .'j; A? O
$ M '$i ? =
a V ' _ .I b,
~_ .-
;! P YF '! 'i '
. "jf M43 l@ ^
i e4 .F.
- 1;.
3 j y Ei ' y _-h ] , .s ,
~
!- -g; <,--
)
ina ' g ,. b l-i e ' .j p y 3 iQ 7 s , y a-M @ :7 -3 ... ,.
, -5. e ? -t 3, , s,--::
yj l ...'
' ' -) :
; i: ' ,E w
I l' il [MI
> o % };;
g,
~
~. . i 5b 5$
i - PHOTO-3 MICRO STRUCTURE OF 4INSS PIPE IN HFIH AFTER JOINT WELDING (X 500) 18
i i I A-2 IMPROVEMENT OF RESIDUAL STRESS PATTERN IN PIPE BY HFIll (SECOND REPORT)
- PARAMETRIC SURVEY -
Summary
..l' It was reported previously that the residual stress p'attern of a pipe weld joint could be improved by high frequency induction heating'and simulteneous water spray cooling on inside surface of the pipe.- ,
4 This paper describes the results of a parametric survey to find a better heating and cooling condition for improvement of the residual stress pattern. .,, Many short pipes were treated with various heating and cooling conditions.- The residual stresses were measured by the conventional strain gage method. The main results are: (1) Frequency and heating time had no significant influence. s (2) A larger through wall temperature difference produced,a larger compressive stress. , 4 4 ,< U. / 'a IF
. n v * ,
b h
\ j' s.
N g # 2 N<
/
A
- 1. Introduction It was reported previously that the residual stress pattern of a weld joint in pipe could be improved by high frequency induction heating (HFIH) and water spray cooling.
The main objective of HFIH and water spray cooling is to obtain a large bending stress (beyond the yield strength) induced by the large temperature difference through the ws11. The bending stresses may be influenced by the many factors such as heating depth, heating time, heating temperature, temperature gradient in axial direction, etc. The influence of these factors has been investigated. This paper reports the results of the test mentioned above.
- 2. Experimental s
2.1 Material , The chemical composition and the mechanical properties of the pipes used-are'shown:ln Table-1. The pipes were cut into short pieces and subjected to a stresa relieving heat treatment (900 0C x 2H).
- 2. 2 Heating and Cooling -
Heating was performed from the outside surface by HFIH using a single coilwithvariousconditionssoshdaninTable-2. The temperature on the'inside and outside surfaces experienced during the treatment was measured.by means of conventional A.C. thermocouples. ( f I 4 3
,.n . -
-.y.
i
, Y_ h'\
/- .r'-- ) -
'i
,t
,. h :t"
< e ,:g
?
vi '
- 3. [' i t - }(
ResultsandD{scussion ~
- g. _
3.1 lleating Depth ' 'k
- j. : - ,~s .
.) . ; ,r; qi In order'to study the e,ffect of heating depts R-12 was heated using. I Y a frequency of 8.4'idiz while the others were heated /using a frequency of I 2 1diz. k
/ ,fe
-g f' -
The results are shown in Fig. 1. The inside surface temperature of-y R-12 was lower}th'an that of R-15 because the heating; depth of R-12 was shallower
, I\ ; . . ,
, a 'I %
that that: of R-15. There is, however, no.significant difference.in residual 1
. ., i a i .
7 , stresa between R-12 and R-15. This results from the fact that theitemperature i, difference through the wall of R-15 is sufficient to obt.ain the maximum effect.
- r.
il t
% ,p7 E: -
A5 3.2 Temperature Difference hf4: ' ,
?: >
,0
.R-13 was heated to a somewhat higher temperature and R-16 and 17 were 1
'. k heated to lower temperatures in order to st Ady the effect of heating terperature i'
f. V (teaperature difference) on residual-stress. *
, +.
Fig. 2 shows the results of the test. R-13'and R-15 have greater t '.
. temperature
- differences resulting in more compressive stresses on thespipe ,;
wm .g ID than d-16 and R-17. .: I D ,- 4 a
.), ' ;. b .3 J3,-1 j
3.3 ~ IIcating Duration ( , }; 9., .IA'"
) 'O Q7 ; li
} R-14 was heated for a longer time than the others in order to dnvestigate[ j
. .1 5 ..
the'effect of heating duration on residual stresses.3 The results are'shown ,- 4 ,. . , p, .' inFig(3andnosignificanteffectisobserved.' e, '
,,. , (
') -
g .,
! < s .J ! i ,
41 /
, 1r
< $ l
, r 4
C ,
** 4 (,' ,.
1 -?
+
.I -- , . - , -a
s li ( 3.4 Effect. of Pipe Size 1 5
/.. ~
t ' I! ' , R-18 was prepared from 10B Schedule 80 pipe in order to investigate the 3k i 'teffect of pipe size on residual stress. i b t f-? , The results are shown in-Fig. 4. A greater temperature difference through W p .< g ' wall was obtained on R-18, but smaller residual stresses were observed. The 9 - 4<' effect is explained as follows. 3, 4 i The temperature gradient in the axial direction is not favorable for this f' ~ residual stress improvement, bTeause it cancels the effect of the temperature difference in throrgh thickness direction. i Genera 11y, the effect of axisynsnetric loading in a pipe can be neglected !' if the location of interest is remote from the loading point by a sufficient distance. This distance can be estimated as (1.5 % 2.5) 4II, where "a" is the mean radius and "h" is the wall thickness. Therefore the region where the residual stress must be improved should be at least'l.5 4fi away from the coil end, namely, a larger' pipe should be heated using-a wider coil.
- 4. Conclusion i 'TN results obtained in this test have produced the following conclusions:
f (1) Heating depth and heating duration have no significant effect (. 'l on the residual stresses in this test. f (2) Larger temperature differences through wall are favorable to pro-i !! ducing a' larger compressive residual stress. (3) Heating a wider band of material will be required for larger pipe. i Reference l (1) S. Timoshenko et al: " Theory of Plates and Shells," McGraw-Hill, i (1959) pp. 469-471. 1 5 1 g
TABLE 1 CHEMICAL COMPOSITION AND MECHANICAL PROPERTIES SPEC 1 IED MECHANICAL CHEMICAL CONPOSITION (%) PROPERTIES C Si Mn P S Ni Cr 0.2 oU EL REMARKS max, max. max, max, max. 8.00 18.00 * * (%) PIPES 08 1.00 2.00 .040 .030 11.00 20.00 21 53 35 fT 48 )
.06 .59 1.75 .022 .006 9.40 18.60 28 62 65
[2
" .06 .53
( 35 ) 1.76 .018 .006 9.20 18.80 22' 55 70 R-19
- STRESSES ARE IN Kg/mm
HEATING COIL
,0 TABLE-2 IlEATING CONDITION vuuyfffg PIPE
/
~
15 ~% \ ' SPRAY vfGUN TYPE A IIEATING PIPE FREQU- HEATING MAX. TEM. MAX. TEM.
*2 200 AT TF METil0D SIZE ENCY DURATION (OUT) (IN) .
(*1) (INCll) (kilz) (sec.) (*C) (*C) '(*C) R-12 8.4 19 530 105 425
-13 16 676 250 426
-14 102 520 192 328
-15
^
21 562 240 322
-16 23 322 167 155 -17 25 247 115 132 -18 10 29 519 118 401 Notes: 1 Heating method are illustrated on the right.
2 AT is the temperature difference between outside and inside surfaces of pipe.
0 30 HEATING COIL TEST PIPE 800 - (4IN S/80 SUS 304) U o R-15 (2KHz) - - 600 - g N> e e R-12 (8.4KHz) o - OUTSIDE w 400 - INSIDE Ed
. o - -o s 200 - \
s _. o'~ h so _ _ ,_ _a . e ,e I I I l l l 40 20 ) 20 40 60 80 DISTANCE FROM COIL CENTER (mm) 10 -
~
" e
.~
N U e Ig ,, 7 / a5 ae
-1o _
- g/o/
m / gZ if 0 g~ s
-20.- [t o
o a: Fig. 1 EFFECT OF HEATING DEPTH ON INNER SURFACE RESIDUAL STRESSES 8
30 30 HEATING COIL j i
/ TEST PIPE (4IN'S/80 SUS 304) 800 ~ f
~ _
@ 600 -
\a F -13 o
g o R-15 w _
$
- R-16
$ 409 _ ,
k R-17 N ~
$ #N 200 - *[
k E'Us s, s' INSIDE 0
' I I l l 0 20 0 20 40 60 80
' DISTANCE'FROM COIL-CENTER (mm) 10 -
^
i k y, 0 D O w ;; h xs o x x'x */ g
~'
. g 8'
~
* ,(p o N f-
$~ ,
4 o ' 5
~
i Fig. 2 EFFECT OF MAX. TEMPERATURE ON INNER SURFACE RESIDUAL STRESSES i 9 i-
30 30 IIEATING COIL TEST PIPE 800 -
) ')g (4IN S/80 SUS 304) l C
o R-15 (21 sec.) 600 - e k A> , e R-14 (102 sec.) OUTSIDE
$ 400_
Z y 0 e --- INSIDE
.-o
>h N I 200 - , o -*~ ,,,_,_0
'O . ..o 0
l I I l l l 40 20 ( 20 40 60 80 DISTANCE FROM COIL CENTER (mm)
@Q C' 25 ea % ,
O " E -2 0- [bo a s Fig. 3 EFFECT OF IIEATING TIME ON INNER SURFACE RESIDUAL STRESSES 10
l 30 HEATING COIL I I y SUS 304 TEST PIPE g _
- o R-15 (4IN S/80)
~
600 - l, o R-18 (10IN S/80) OUTSIDE 400 - INSIDE g. g 200 - N o I >~ ~ ~ =O
\
~ _
,- -o 0
! ! l l l l 40 20 C 20 40 60 80 l DISTANCE FROM COIL CENTER (mm)
; 1, _
N h 0 N I-
~Q *-ll--*x.N.
o
/
$N -
. #1 ,, N8 3-d [< > Q" o 0 m Fig. 4 EFFECT OF PIPF THICKNESS ON INNER SURFACE RESIDUAL STRESSES s 11
, - _ _ _ , _ . - ,. . . ~ . . . ,
A- 3 IMPROVEMENT OF RESIDUAL STRESS PATTERN IN PIPE BY HFIH (3RD REPORT)
Summary It has been reported previously that induction heating is very effective in improving the residual stress pattern on the inner surface of a welded l I pipe. A series of demonstration tests were conducted on 12 inch schedule 100 stainless steel pipes and elbows prior to the application of this process to an actual plant. The residual stresses with and without induction heating were measured by the conventional strain gage methods: Cracking tests in boiling 42% MgCt2 solution were also conducted for this demonstration. The main results were: (1) High tensile residual stresses were lowered or shifted to compres-cion by HF1H. (2) No cracks were observed after the cracking test in MgCt 2 n the inside surface of the weld joints treated with induction heating. Conversely, many cracks were observed on the weld joints prepared without induction heating. In addition, the effect of the heating ccil width and cooling water rate was investigated using 12 inch pipe; also some calculations were conducted to study the residual stress distribution in through thickness direction and to check the effect of weld metal. 2
i l
- 1. Introduction It had been reported previously that induction heating is very effective in improving the residual stress pattern on the inner surface of a welded pipe.
These data were based on tests conducted mainly with 4B Schedule 80 pipe. It is necessary prior to application to larger pipes to show that induction heating produces satisfactory results on larger diameter pipe. Consequently, a series of basic survey and demonstration tests and numerical calculations were performed. This paper describes primarily the results of the demonstration tests. Other results are described in subsection A, B, and C.
- 2. Experimental Procedure 2.1 Material
> 12B Schedule 100 austenitic stainless steel pipe and elbows were welded as shown in Fig.1 to simulate the actual piping condition. The mechanical properties and chemical compositions are listed in Table-1.
2.2 Welding
\
The first three layers were welded by the GTAW process and the remainder by SMAW in the SG (horizontally fixed) position to duplicate the field condition. The lower portions of the three joints, No - R - 26 and 28 were gauged and re-welded to simulate repair welds. The details are shown in Table-2, 3
2.3 lleating aad Cooling Devices The general arrangement of the heating and cooling devices and the test l model are shown in Fig.-2. The model was fabricated on a restraint frame to simulate the installed piping condition at the site. The butt weld seam was located 60 mm from t coil edge to simulate a pipe to elbow configuration. The details of heating and cooling conditions are summarized in Table -3.
- 3. Results and Discussion 3.1 Temperature Distribution The temperature distribution in the mid portion of the coil was very uniform if the pipe thickness was approximately uniform ssee Fig. 3). Where there was a large thickness transition under the coil, the temperature varied with the thickness (see Fig. 6). This is easily understood by the following l basic equation.
00-01 q-A ,
......................... (1) where, q : Heat flux A: Thermal conductivity e : Wall thickness 60,0i : Temperature at outside and inside surfacas respectively In this case q and A are nearly constant. Therefore so-01 is inversely proportional to t, that is, Oo soes down and 01 goes up as t becomes small. This behavior is also demonstrated in Table 3.
1 n 4
3.2 Residual Stresses The residual ~ stresses were measured using conventional strain gage techniques. I The results are presented in Fig. 4 to 6. The'high tensile residual , stresses in the as welded condition were lowered or shifted to compression ' by IHSI. The highest tensile residual stress after IHSI is approximately kg/mm for R-27 and 28, but no tensile stresses were measured for R-30 which ' exhibited the largest temperature difference. (a) Effect of repair weld In order to evaluate the effect of repair welding, a simulated repair weld was performed in the interval between 157.5 and 202.5 degrees' on two weldments, No R-26 and R-28. R-26 which had not been treated by HFIH was exposed to Magnesium chloride solution and developed many cracks near the repair weld since high tensile residual stresses were present i (see Fig. 7). On the other hand, the residual stresses on R-28, which had been HFIH treated after repair welding was evaluated by conventional strain gage technique. The results are shown in' Fig. 4. The residual stressr.s in-the repaired portion of the pipe in the vicinity of the repair were not significantly different from the as welded and HFIH locations of the pipe. These results indicate that although repair welds induce high tensile 7 residual stresses to the pipe ID, the stresses are lowered and possibly shifted to compression by HFIH. (b) Effects of restraint
- The mock-up containing weldments R-24 through 27 was fixtured in 4
a restraint frame simulating a worst case plant condition. R-24 and R-27 ! 5
*?-
i . I
had the thinnest cross and consequently were the most severely constrained. The results of the HFIH treatment presented in Fig. 6 and 7 show that the residual stresses were effectively improved. One can approximate the restraint effect by utilizing the following simple model. The free thermal expansion of pipe due to HFIH is 6 = aL(Ta - Tr) where : Coefficient of linear thermal expansion + a = 1.8 x 10 nun /aun*C L: Heating width = coil length = 250 aun Ta : Average temperature across the thickness 350*C Tr : Room temperature = 20*C Therefore . 6 = 1.8 x 10 -5 x 250 (350 - 20)
= 1,49 mm The axial force in the mock up induced by local heating is given by (see Fig. 13).
24 . EI . 6 P7 p where E: Your.g's modulus I: Moment of inertia of section 1: Piping length (see Fig. 13) and results in 0 p , 24 . 2.1x10 x2.171x10 3 x g,49 5 (1.247x10J)3
= 1.68 x 100 kg 6
i J The minimum cross sectional area (at R-24 and' 27) is, A = x.(318.5-15) 15
= 1.43 x 104 mm 2 The axial thrust stress induced during HFIH is a =f=1.17kg/mm 2 This stress-is very small compared to that introduced by the through wall temperature difference, that is, the restraint effect is negligible.
Generally, the pipe restraints (support) are located at such incremental distances that the thermal stresses produced normally are within allowable limits and, therefore, the restriction of any free end displacement due to local heating can produce only small stresses. 3.3 42% MgCE 2 Cracking Test Cracking tests in boiling 42% MgCE 2 s lution were conducted on pipes R-24, 26 and 29, and the results are shown in Fig. 7 and 8. Many cracks were observed in the highly stresses areas, such as the-outer surf ace or inner surface of the untreated joint but no cracks ~ were observed on the inside surface of R-24 and 29, which had been treated by IHSI. This observation indicates that the high tensile residual stresses on R-24 and 29 are reduced or changed to compression at points near the weld on the inside surface of the IHSIed pipes. 7
3.4 . Mechanical Properties l
)
Tensile, bending and hardness tests were conducted on R-25, 28, 30 ! l and 33 to investigate the difference in mechanical properties between
-joints fabricated with and without IHSI.
These results are shown in Table 4, Fig. 9 and 10. The tensile strength af ter HFIH was a little bit ( 0.4 to 1.0 kg/mm ) 2larger than that before HFIH. A small increase in hardness at the outer and inner surfaces was also observed in the HFIH treated pipes but no hardness differences were observed elsewhere. The results of these measurements are presented in Fig. 9 and 10. These results show that mechanical properties do not change axcept for a slight increase in tensile strength and hardness near pipe surfaces. This observation results from the fact that the plastic strain induced by HFIH is 0.5% or less as presented below. The maximum temperature of a pipe outside surface is monitored and controlled so that it doesn't exceed 550 C. The cooling water temperature is higher than 0 C. Therefore, the maximum temperature difference between the outer and inner surface of the pipe is less than 559 C. , 4 Assuming a linear temperature distribution across the pipe thickness, the thermal strain at the inside surface cth is approximately given by 1 cth " 7 a AT i
= 0.5 x 1.8 x 10-5(550-0)
= 4.95 x 10 -3
, 8
Consequently, the maximum strain at inside pipe surface is about 0.5%. This I strain level is too small to produce a substantial difference in mechanical properties. 3.5 Microstructure The microstructure of the weld joints with and without IHSI were observed using a 10% oxialic acid etchant. Figure 5 shows the microstructure of pipes R-25 and 28. No significant differences were observed. This observation results from the fact that the strain due to HFIH is not so large as to produce a big difference in microstructure.
- 4. Conclusion A series of demonstration tests were conducted to demonstrate the validity of the HFIH process for 12 inch pipes. The following conclusions were obtained from these demonstration tests.
(1) High tensile residual stresses were lowered or shifted to compression not only in 4 inch pipes but also in 12 irc.:. f'7 c (2) This observation was verified by the cracking tests in boiling 42% MgCt solution. 2 (3) The pipe restraint (restriction of free end displacement) produced no significant changes to the residual stress improvement process. (4) The repair welded portion of pipe subsequently HFIH treated demonstrated a significant improvement in residual stresses. (5) Little difference was observed in mechanical properties and micro- ! structures between pipes which were treated and pipes which were not l treated using the HFIH process. 9
Table 1 Chemical composition and Mechanical Properties chemical Composition I Mechanical Properties Heat L C St Tens 11* g* Mn F S Ni Cr II*I A Elongeth max. man. max, Point Stren8th Remarks 0.08 max. man. 8.0% 18.0 % 21 53 30 C 1.00 2.00 0.040 0.030 11.00 20.00 (ks/mm2) (kgf 2) (2) D58412 .06 .56 1.71 .022 .006 9.25 18.10 C .06 .56 1.74 .020 .006 9.20 24 58 70 P!PE W 18.10 o L .05 .55 1.69 .023 .005 DS9218 9.20 18.45 C .06 .54 1. 72 .022 .006 9.25 23 57 M 90' LONC ELS$ 18.35 L .0). *5 1.75 .024 I~ .008 9.77 18.88 SAFE END C .033 .54 1. 79 .024 .013 9.15 26.1 54.5 67.5 18.86 (FORGINC) I SPECIFIED VALUES SHOW MINIMUN.
.m -
2 -. ew=
TABLE 2 DETAILS OF WELDING CONDITION
- WELDING CONDITION BEAD SEQUENCE PASS PROCESS
[ YO CE H(A lNj }PUT g'C) POSITION ME 182F304) 98b125 R-24 1%3 CTAW 12 7.6%10.4 MAX.37 r Pi g 100 _ 4s9 SMAW 73b106 21%22 14 %34.2 66 . R-25 I DITTO dim 90*g U (128 S/100 4sg SMAW 75s105 21s22 14.2%34.9 54 R-26 1%3 CTAW 98s125 12 8.6%10.6 47 DITTO DITTO DITTO 4s9 SMAW 80%105 21s22 14.3s30.5 63 u.
'itHULATED 1%$ CTAW 95%110 12 7 411.7 112 REPAIRED REFAIR l AREA ,
g i 6s16 SMAW 80s 93 21%22 9.9s22.3 110 jgg. T .)02.5' 1s3 CTAW 98s125 12 8.7% 9.1 41 R-27 SAME AS R-24 SAME AS R-24 SAME AS R-24 4%9 SMAW 80s105 21%22 13.8s28.7 69 R-28 SAME AS R-25 DITTO
- es9 80s110 DITTO SMAW 21%22 12.9s30.7 72 ilMUI.ATED 1%3 CTAW 100%115 12 11.3s13,9 115 4 %11 SMAW 80% 90 2W22 17.6%29.3 118 R-29 DtTTO SAME AS R-24 SAME AS R-24 4% SMAW 80s110 21%22 13.1430.9 54 SAFE END 1%) 0 CTAW 90M 00 12 8.ls!O.3 62 (ASME SA182 F304) 4M0 80M15 SMAW 22 20.7s37.9 98 g 1%) CTAW 90415 12 7.540.4 49 R- Dt m D"
49 SMAW 85% 95 ~22 16.2 M 8.6 76
i TABLE 3 TEST MATRIX .
*1 HEATING CONDITION ANALTSES J0llf! NO. TNICK F v " JU-POWER udp INC COOLINC COND. *2 *? 83 *4 a5
- T C61L hA]IM TI VEu)C ETY ENT/
T* (T T1) TTD NaC12 S.C MECKA. MICRO. EXITft: (*C) ('C) R-24 15 3 300 250 179 0.5 394 281 Q Q - - - R-25 19.25 - p - - O O O R-26 19.25 # % - Q - - - R-27 15 3 300 250 176 0.5 400 236 O - O - - R-28 19.25 3 300 0.5 250 177 514 395 0 - O O O F-29 19.25 300 3 250 178 0.5 514 386 O O - - - R-30 28.5 3 300 250 116 0.5 5% 477 Q - O O O R-n 28.5
.~
o o. o
*1 WALL THICKNESS AT JOINT *3 TTDs TIME TEMPERATURE DISTANCE CURVE
*2 To AND T1 SHOW THE OUTSIDE AND *4 NgC12 : CRACKING TEST IN 42'l BOILING MgC1 IHS1DE TEMPERATURE RESPECTIVELT 2 SOLUTION
*5 S.G RESIDUAL STRESS HEASUREMENT BY STRAIN GAGE To C (HEATING Coll)
\ 5 c
W -
_. . ~ - _ TABLE-4 MECHANICAL TEST RESULTS TEST ITEM TENSILE BENDING TEST HV MAX, (HV-10kg) MATERIAL owr go, IHSI FACE SIDE RASE MTL, WELD MTL. R-25 NO MIN 57.i N.D. N.D. PIPE (SUS 304 ) MAX. 58.7 N.D. 0.3x1 205 213
+
R-28 YES 247 192 12BS/100) SUS 304 , , , U MIN. 55.5 N.D 1.0x1 SAFE END R-33 NO 205 238
, MAX. 5 7.3 N.D. N.D.
SAFE END R-30 YES MIN. 55.9 N.D. N.D. (ASMESA182 F304) MAX. 57.9 N.D. N.D. 194 216 41 N.D.: NO DEFECT
INDUCTION COIL R-25 ) (190) / (R-24, 600 R-27) 90* ELBOW SUS 304 0* 60 so/, l /I e , 3
-/ 9 o
o i 270*-1 90* O E N f y 180* gSFEEND
- SA182 F304 [,
12" S/100 SUS 304TP ( ' R-26 318.54 R-24 (S/100) e 4 0
.7 cn
^
/
60*15* 5< R-27 30*12 3404 l STUB { ASME m SA182 F304
?.
_, a f - J % $ 210.4 1.6 DETAILS OF JOINT (R-24, R-25, R-26, R-27) Fig. 1 MOCK UP FOR 12" PIPE DEMONSTRATION TESTS (R-24, -25, -26, -27) 14 6
7 8 2 i MG ,' g i 1 CONDENSER 3 4 TRANSFORMER Fig. 2 INDUCTION HEATING DEVICE MARK NAME FUNCTION REMARKS 1 MOTOR GENE. 300 kW 3 kHz 2 CABLE 50 m 3 CONDENCER 8.9 pF 4 TRANbFORMER 1000 kVA 14:1 5 LEAD 180W x 10001 6 INDUCTION COIL 250 W 7 BAFFLE CYLINDER PIPE 8" og 8 RESTRAINT FRAME 9 FEED WATER PIPE 2" 15
250 _ 20 HEATING COIL
, 65 g __
60 = POWER : 300lM HEATING : 177SEC TIME
)
g o Q, -- I
-1 -~
b 1 . COOLANT -- E (COILE) (0.5m/SEC.) 600__ OUTSIDE
- - - - -a - -a
. /
- 's g
$ 4 00-- ' N
/
[ \3
\
4 \ 200-- INSIDE \
/ 3
/
#_~
- o - ._/_._ \
/
_,___'o O l l l ! ! ; : ; l ; ; ; 120 100 80 60 40 20 0 20 40 60 80 100 120 140 l DISTANCE FROM COIL CENTER (mm) t
- Fig. 3 TEMPERATURE DISTRIBUTION OF R-28, A PIPE TO ELBOW JOINT 16
v i fa le I M PI (12" S/100 SUS 304 TP). 90'I2aCW(SUS 304) s
!! EATING COIL @ /j.,,
(R-28) 10 10 10 N 25 - WITHOUT IHSI 20 - a
,,' - * ~
ir WITH IHSI l 15 - 4
/
\
"%'s
- 1 e i
/
,5 10 - a L ll R-25 R-28Am
- e v 30*
a <> mj 5- f > ai
- 4 0 75*
mm s ia 0 " u--
-'"~~ ~-
sq
" " ^
My e o' 165' . Ay W ~~~ }l
$o QU e it s f of p 210' mm " /
mm o N s '
/
n d' a a 255' WU ir % a* o 1> a a 300*
-15 _ e , 4 Il
n M 345' os " 8h s,- [ 4 jp', -o _-
- REPAIPG AEA 10 0 10 20 DISTANCE FROM WELD CENTER (mm) 25 - "
4 ~ ~ 4%" . 20 _ s u y
~ " N
~ 15 - o <e s
' il s p % Dcn 10 - ,_at, s a y M \ m ** ~~ h 0 !,' b b N M_ 1;
'N" gy o
,- % s .o se ox / N 05 m
c / \ b
\ st
$ \
o N Ir N
's
-2 5 -
Fig. 4 RESIDUAL STRESS OF R-25 NAD R-29 17
IIEATING COIL ' (R-30) ie E SA182 F304 / [ 10- 0 COOLANT (0.5m/sw .) N WITHOUT IHSI' 5- WITH IHSI fI l _0 R-33 R-30AZDL'2II D & J -5 d , , O'
~3 'a , ' ,Y e , y$ 1F' ' % , , ,, , - e e 45' ag "
ri ,, yy - a a 90' ad ) gg -20 ,, j N- e o 135' ou , < ' hh '
, / ,
, g 180*
- o. -
s'
/ / Wi' e o 225' f
a a 270' 10 0 10 DISTANCE FROM WELD CENTER (mm) 25- ,-o ,_ 20- i s' <> e, a g g D 15- .
~
10- f 's ,
\ !
5-
$3 , (
"6 d s
gg -0 o o- e M
>l e /
N . Fig. 5 RESIDUAL STRESS OF R-30 AND R-33 j 18 _ . _ . . . _ --..___ _ -~ _-- _
r- _g.. !
- g
~
l s r s , 1 A j * % y I f
* \x. e v.
i
.. i.
a 20 60 l 190 1
- l. .s . .
4 { % s
-i e
- f '
ff ' a % NL -
- w2 l
ur
/\i/ _l E 4
_' 'U Nl 65 .q , ,
- COOI. ANT p -*
f(HEATING COIL) - HEATING COIL STUB g (0.5m/SEL1, 7 g r%.-
'^ w w
ASME 12,,[hh -), i POWER r 300KW SA182 F304 J t . t HEATING 176SEC e).[- TIPE
*-1 e
.. 4, a, (
c ., %
,. i + ., i s
+4 g a .'
.
- i I
( C $.h . s ., , [ r, .p
* *
- r i - i i sa i .
, 80 60 40 a i 20 i 0 20 40 60 a a a i 80 100 120 :(40 160 180 200 220 6 e i
\
DISTANCE FROM WELD CENTER (m:al-- & -4 . L t 25 -
'+ hl 8
ry - 9, . - 20 o CIRCUMFCRENTIAL J
,f --
~
- o AXIAL s 15 -
s . 5 u ,
& 10 -
Jd ,
,* j m 5 -
m ,4 (* g4 -i e = %.
$ 3v > c -
,- " al f HOOP (0*) -
m sa i I I s _ '.\ sc + t t A' g' :e , < 3 N AXIAL (0*) \ 20 - '
-] 8 s ,,
N,
'a
+
g .,m, i i-
~. <
Fig. 6 RESIDUAL STRESS OF-R-27 ..N --
\ ',
l a y- - j , %
"s %
i e '. - 4 3
% , * .N A,
2 -
'N 'p g. s
"' i-
, 19 , E.
,g '(
t %
. Y, r
__ . . , - - . , . , _ - . L
o?4 -
.,.-; , 2
)
1+ ._ . 4 . t e , i
% e T
,/ f , y, q ., . - ,
- r.,
.,74 - - 1 . . .
x.s o,,
- . 1
-.,5 r
,.s ....::. .. .: r.., __. , ,,.., ,i f:
.s.y ~s
~
- u. .
t . .m. c
. .;,M4
' ^
-A l'
~ 1STR.24 . . . n. .a,u.
x A_ ,
- ,% c M.
~
j un,- :~ 2 -
- m. ?nf,w., l*
4 []
' 'l t ;,y ,
[ r y ,, h s.
'A . \;5 e?.
Qq ;w e
- v. n . f.
u.s,: - -n-
;..) . .: ,
u ,.*.,3, J .
' ,% ; .>r *; z ':
/ ;. y *1 r ,
, -- , . _ - . . f, g..,,
{.) , ' l '$Rff
%; +; quy . N.Q.Q ,, .
- S g = O ~ ' , q~! '
Q.
}Wiar:dras a cn M
. .g
, - r
,g c
~. .,4 ,.
- ; , ya. *
. . 4.: . .., ,
-- ~ :. ,, ,
, .- . :.; ~- - :
_3,
/; 7.2a;; ~w
.<t ,m... -d .' c. .~ : , . ~J.
e
. W q..
ne k $'>j} r , .
~ ^'7 -
'QK.
- d. ' fjyp . - ~
-+r 2 '
' ;,6 ., .
, r.. *y..,
-.. ' = . , .
+..%
y
* ),g,r, ,..
,pn 3'_'
~ ..s~ % .
o g , , , ,. ~, - -
~
n:> m
,.g . a< .
. ,c , ,..
-. , , +
- 3. n I
;9
. i
? , ,
)
i i INSIDE SURFACE OUTSIDE SURFACE l FIG. 7 CRACKING PATTERN OF R-24 AFTER IHSI (42% BOILING Mg Cl2) 1
si ! v, 1 i i l
* (-
[ r t HEATINC CilIL l l
% N
~ 47'
' 2' A X e p e
._\:s, _ __ _v _ . . _ _ w s ,l "
s?o ,L 20 .j ' A B C o I a
# il' TP-N O. A g , R- 2 ; '. . ; R-3 a 0
- s. a n '
MAX. TEttP. 536*C
>r, 3 40 . *C
, 8 e
, - t . A . ,. .g
' ,ie
- c ,. .y- r-f, , j V{ ..y.-,. .e s
'.- , s. . .
,. a ap . ,-
m J. [ ;- \ ,, \- tijl e ,,a ,,.
,r (s 1 .s s
.; j',ly~ } 'Y, $
f,.,a;'kf[Ur fp*. / 'y ym, y o ~. g
', ' . ,,v .
v a 1.d *. ?.A f U
}fg[a',,Qfy.)%y , ,/ . n l.' p; \, g ;1l \ % y .y e -
..A .
f.*h
'l' y 'f' / , , ,f.. ; -}N ~ L. lbi , lKff . .
,L .s y n * .
o y e W~ < c .. u w . o e B _ % h 3 M g C $ c i W
< m o
c 4 h i e 1 #
.i o
} 80 S
c H * * . , .
. * .. x
* * . E l ,,
C 4 1 0 m a C- .. . 4 g
~.
Note
- Max. temp. indicate in head conter of pipe (outside surface) 21
A N l O
' ' / A R, A B 260 - - 260
~ l 1 B
240 - o - 40 e R-33 OUT } (AS WELDED) R-33 IN
- - R-3 0 OUT -
} (WITH IHSI)
-- A - - R-30 IN ,
220 - - 220 A _ 2mm iN _ m
/ N A,\ s X
/ /
o 200 __ N ga', ' 200
, A # \/ , #
_, ,e ,4s / N ,, / . - 180 -
- gc ^ ~ ~ ^'<
*[A ' A
, A'
'O
[ 4
- 180 A f,] p, A-g'$ ' -
160 - 'o o
' 160 BASE METAL WELD METAL (ASME 140 -
SA182 F304) _ 14o A B Fig. 9 HARDNESS TEST RESULT :'OR R-30 AND R-33
__ .__ . _ _ - - - -_.- . . . - ~ _ _ . _ . - . _. -. - -_ . - .- _ _ . - . . - _ _ _ SAFE END (ASME SA182 F304) _ o m. 200 - - 200 40
. R-33 (AS WELDED) 180 - - - * - - R-30 (WITH IHSI) - 180 E
s i g ,
, , 5- - -a n 4
160 - 's s ,5 . .
.*' - 100 E' i
, x . ,$ e x_ _ _ - s ,L* =< , .
- s a' . . ,
!. 140 - - 140 OUTER INNER i 120 - SURFACE SURFACE ' _ 120 ) l i I Fia, 10 THROUGH THICKNESS HARDNESS DISTRIBUTION FOR R-30 AND R-33 l l -.
r ' ~~
; i t
, .. xx- i
! 'y?;.1., s .
e
.e
. .t.
4 1 5, : 4 , . M: Reg I - i I !
... i N : ,' a
~~, !
L,. '.L.
)a/l Q. U. :
n . 4,v S!
. . l . n ,< r)l D
- p\ m
- l. /
,. n -
i, i m o r- you t l j . pa . .- I 4 s m e g m.92-
- m. 1 i --
D
- i. , o
['sl
'., s. k g3
- fi j
- O
, ,. g E
l a g-
- y m o a
e
.a / n.
o 8 ' y i 3- mh - ^ l E S ' ' (:l y g . 'N s . ,.' . M. .,
.i . . N. .- ,
N.-
. ,e. .
i o m I .
\-
8 7,. li - o i s- -
- u. .1 .u dg ,.., . ,
0, & l,ty@w? . sl'c.?fl
, r.
?.$hgh'. -
l
. + j . f's-/L ,f . - A ,,v.
i',s .. '
- s n . *J ,%6 ,.
l r r g
'$ c '?l;!A
- f.4,f+. ,,y..
<n k,z -
;;n o r A.; ;.1 a ,
. t
~
y l r=4
' f ; Liv ; ..
~
~,f ,'. H
', . .t }.' .,g' 't tyy
<. i:
' ?r'; .Q+.:>;.,-Q:'.t
$ . .g 1 9. :,. . ' l Ls., ..
.-Q,1.4 s
[.8. 3 2,.f..fi ?i. ,
.'/
- f *j' e- * ,
-! / .Q.sI::%. y;th ' J.aw*'# ,
*?
O
\%. .4 3 3 ..~ .q*. s
;:: n .y. . .,;.f *,* % * ^ +, , .
.m ,t.
s 1 g y- z., m . ,;, , . . . i., w
<t;{p.
- p. ':gy;:::.,.p-y :'M
. : : > '. ; ". ^ei.. .4:-
', W: oc. : ...f. .)p..,
. r' 7s '>!;: ?
*t.: ' - q,%.f,x p.t. .s ;. -. ?,k--
S y b+\.
. . . . .w y :. .
. .. o,, wg . +. .x. h . r . . ,*ig; ,* ;s
,-i g:n:v _i.ypl ; .~,9:h..
;<,... .r3p;g.,, * .;..: u.
., . ...t. c : l/. . .., .> ;- . y, .: ,s .
s
,f..g{.* *
- l,A' 8 2 'ls /, *a.. (*'v. .l ** ,,, * ' ~ . t f.
_1
. ,~, *:
l^v
? ',Ik, . .Q,
- f l '. ./ . ~ , , ,
- c;; :y7.d
, ;,q. lyy^r. ;
.;we[;;Q{'i.z
. n ._(awam;; . , G ....a ..
..p..,..
. t f. . a. .. . .
! 00LX 009X 00tX 009X i
L._.___.___. (a'Ian sv) 29k (IsIII su.Ia) 02-8 l I 24 l d
i i' h \** i l i
' . ' ~ .7 / . ,
, N'~) *.':k. *'
~
oQ
\
.w -
g- ..
, ,s. . . . .
~\.
P;
%' *4.,. -
.7 .,
c / s w . .,f r, w - ;
! .} I ...
hf., '*_ u
%$,A ,/M s
. . g,.
b 4 k d.i1 1 o m
-( ,,
! .__4 0 W '~ %g, .
w O
]
s
,ys
. 9m#,, .
m
+ .
,.. . z
~*-L g,a. m. .\ s. H g.
. an ~.
- y. p j -
s w o
,a y 2 .e pj
-5 v '
g ". ) N. ; ', -
- id 3
o . , . . . .
'. : . e
"' l -l5 ' i
/
f
- c. %-l. 8 m
. .. '.t .g . - -
iY N L9
- m. . _ _ . - 1 H , L-4p__ ,
a
/ .;.q F;.;;~: oq _
-. .s. p uy,.), . , ,\ ,s. <s
- a m
.& . ; ' y' v .: 7is?'$ *. ^<; %
. 7 ..:
y@,.l:4.. y\ %- S N ,s ::e . gi ., 3
.' t. ' *
.. .C ' ', ' '
~l N 0 -}$' a.
m
,e$q?:3.g. .N..;.A7p,.. .. .
m n y 8 ', .. w ; . s / _- i
,, .. m
@ 8/
- s , ,
,. " ,4 "
g ', x - g w -
,.se g
s,m. . C '+ y ~ *W~ . * .sR - ,* ~
-'h,.L ";
Lw%;:~WL5, =*
.al;r+s.ntyp@ffa r \
, '. , y..d A$
.~2 ~
h:5.Mioyyy ' ' '
%. % m sw :,vsn
. a.v. f . .:1 - - 4:
k.,.
~
I' ,, 4', I *
% "ENh , . . . .., f .
+ :..w u .- k'[h.k an,z ?
, .- a
.- . o
[6) D. .- - - c'. C inq%}[v2B 19
.n.m.psw.m. .
- &&R?j%,id 90Mii:f5 x .-
"*L'.f
',i..s- .4 .c
;#W4NTNf3 'y No * * ', 'I' *
- a s .
=tiTDqy *y %"' . s . ~ ry'g'd[w[T+,*.$;'s SE5fd.'.?d f
@' ~ ' ? *.' *
?.
- k'h.'N h * ~
._ -i.~= F n -i -
00tX 00SX OotX 009X _......_,...,w . . - . . . = - - - -+v-**-.e--*=+ '~**%++-- (aun sv) 99-8 (IsiiI tura) OE-8 I_ - - . - - -.- - - . - - . - .- 25
I-
+
'j P I i o l f _ _ _ _ _. ~~ ' \ 5
$ I
\'
i l a E j
= = gl 6 = Eat (Ta - Tr)
I I g i I I I r
~
g 6
. 777/777 6 N P d s o-N N 1 '
62 P / o i 6 = 3EIg3 1 l r X 6= 2 t* Mt P 2 Mt 2E3 51
- 5I PL 3 M g2 =6 3EI _ 2EI P1 2 ,
2ME 2EI El M=fPt 4 PE 3 _ 1 Pt 3 = 6 3EI 2EI 4 (8-3)Pt 8 P= h.6
= 6 24EI i
i Fig. 13 FORCE INDUCED BY LOCAL HEATING i I 26 4
- w, e -we- -
, a~-o w , y r, re~ w r- --~w --r - --,.-~---w~- - - --
~ . - - . -. . . - ~ .
p SUBSECTION A i i BASIC SURVEY ' r The effects of cooling water flow rate and cc : . width on the residual stress
, distribution of an IHSI treated pipe were studied experimentally. The results
! of that study are presented below. t
- 1. Material d'
- SUS 304 (AISI TYPE 304),12B Schedule 100 pipe was used in this study. ,
The mechanical properties for this material are shown in. Table.l.
- 2. Stress Relief Treatment
} All pipes were stress relieved for 2 hours at 9000C to relieve the pipe fabrication residual stresses.
- 3. Heating and Cooling Details !
J Heating and cooling details for these pipes are presented in Table A-1.
- 4. Results (Fig. A-1,2,3) l 4.1 Effect of Flow Rate (R-21 and R-22)
(a) The temperatures on the inside and outside surfaces were reduced as the flow rate increased. 1 (b) No significant effect of flow rate on the residual stress was observed. 4.2 Effect of Coil Width (R-22 and R-23) (a) The uniformly heated zone of pipe R-23 was smaller than pipe R-22 ] consistant with the differences in coil width. 1 i (b) The stress improvement region became small as the coil width was reduced, but degree of stress improvement at the coil denter did not change markedly. 27 e
. - ..~e- - , _ - - - - . - - - - ,
._ - _ . _ . . _ _ _ . . _ . - . ._ _ . . . _ - .. _- -- _ _. _.._- .___~ _ . . _ . _ .__
f j i TABLE A-1 HEATING PARAMETERS AND RESULT OF TEMPERATURE MEASUREMENTS 4 HEATING PARAMETERS MAXIMUM TEMP. (*) *1 A1 i TP-No. FREQU- POWER WIDTH HEATING COOLING COND. i ENCY (kHz) (kW) CO L (mm) (sec) VEIDCITY ENT EXIT (m/sec) (*C) (*C) (*C) h h h h h. h h h h hh
- E-21 3 300 250 ISO 0.72 21.5 22.5 22 220 311 526 537 334 102 98 81 -
435 R-22 3 300 250 157 0.43 22.0 24.0 19 207 301 523 548 358 126 127 77 60 422 S R-23 1 250 200 125 0.43 22.0 24.0 18 199 293 535 558 387 122 122 67 34 436
*1 The measured points are shown below.
3 ) HEATING COIL SUS 304TP 12IN S/100 h (4) h (2) h[ l i , ., l t/2 Y_,_ b6 7 t/4 Y, t/4 8
,10 h _
, 4 -
_ . ._ , __ _ _ _ _ . m _ _ _ _ _. - - i J b 125 POWER:300kw l # !!COIL EATING {TIME HEATING 180Sec i
' p TEST PIPE
/ ) (12" x Sch100) 800- COOLANT (0.72m/sec)
~
' OUTSIDE SURFACE
,i w 600 m
O /~a 2 w
- 400-5 b
, , INSIDE SURFAC
- 200-
\
4 1 g i ;
-o., ' N 4.,
'A
. . . . i e i . . . . .
) 20 40 60 80 100 120 140 160 180 200 220 240 DISTANCE FROM COIL CENTER (mm)
~
g *
~ o- AXIAL } OUTSIDE 20- lI ---a--- liOOP SURFACE
+ 15- l 6- -o y 10- . \, 5~ 0 a: N , % -. a .. - N A l A. /"',
~ . . .-y
, ,...o'
,*/ '.
.* *'s. ,l ,
H
- r. * ' 'a.....'
, n INSIDE
---.o---
_ . . .--HOOP AXI AL } SURFACE l Fig. 2 A-1 MAXIMUM TEMPERATURE AND R.,SIDUAL STRESS DISTRIBUTION FOR TEST PIPE I, l 29
POWER 300kw
!!Co1L EATING I TIME HEATING:160Sec g g
* - TEST PIPE
)
i (12" x Sch100) 800- COOLANT (0.43m/Sec)
, _ a _ OUTSID7 SURFACE .o * - 600- -o INSIDE SURFACE y n E
2 w 400_ o. b e . 4 200- \. .
's N o
,O~.. ,'~
~....~._
i i i a i i e i i i i i O 20 40 60 80 100 120 140 160 180 200 220 240 DISTANCE FROM COIL CENTER (mm) 25-
-o- AXIAL } OUTSIDE SURFACE
--- A ---HOOP p 20- g 6 ,
o 15-o en - 5 E 10
\
m 5- '\, . -a E a \ . ,-
,0 N / 'o*' $ ~5 -
f a#'s o' o D .
- 0 8
*M'Y'a !
= o u , ..f '
'O.N.
--o g-o /',"'c'
- 15 _
oW m,
% a.. ,, s
, ----o--- AXIAL } zyg ggg
, g _. a_.- HOOP SURFACE Fig. A-2 MAXIMUM TEMPFRATURE AND RESIDUAL STRESS DISTRIBUTION OF PIPE R-22 30
. - _ - - - - , _ = . - . - . - .
100 HEATING COIL I POWER:250kw . HEATING 120Sec [ TIME TEST PIPE (12" x Sch100) 800- e : - G 600-4 W W 400-e
. A x '
_ a - OUTSIDE SURFACE
-o - INSIDE SURFACE 8
%g \
a a s e a N '.':s e a s * *
- O 20 40 60 80 100 120 140 160 180 200 220 240
. DISTANCE FROM COIL CENTER (mm) 25- -o-a AXIAL } OUTSIDF
{ - -a HOOP SURFACE 20- f -a% _o N a 15- o d 10-0 5-M \ h '--f f=.TI .
$ s," \
a o s -. S y 3o-O'N/,4.p
.$.D>pA . *. ..e
/ 300 ob
-20 ----o---
AXIAL INSIDE
. a .. HOOP FACE Fig. A-3 MAXIMUM TEMPERATURE AND RESIDUAL STRESS DISTRIBUTION FOR TEST PIEP
.. 31
SUBSECTION B FINITE ELEMENT COMPUTER CALCULATIONS Computer calculations were performed to obtain the through-thickness stress ) l profile for the pipe and to evaluate the experimental results.
- 1. Finite Element Program Characteristics Program name: ANSYS Strain hardening rule: Kinetic strain hardening rule
- 2. Boundary Conditions and Calculated Models Fig. B-1 Model for temperature calculation Fig. B-2 Model for stress calculation Fig. B-3 Finite element model Fig. B-4 Heating pattern Fig. B-5 Stress - strain relation at various temperature Table B-1 Material properties
- 3. Results 3.1 Temperature Distribution (Fig. B-6,8)
(a) The calculated temperature distribution displayed a shape similar to the experimental results. (b) The calculated temperature difference (AT) between the outside and inside surfaces was somewhat smaller than the experimental value. 3.2 Residual Stresses (Fig. B-6,7) (a) Calculated residual stresses were very close to the experimental values. (b) The stress improved area on inside surface was about 25 mm larger than the coil width. (c) The depth in which the compressive stress remained was approximately 60% of the wall thickness. 32
s Hwe'wW u n w{ x x = N N C N N c N D
, l e N E a s N T c z N A K m N L m U 7 N S
- N N 0 w I 1 N N
x N 6 E N
} 4 P N c . I N e 4 P N s N
/ C T T m* N N
22 E 72 I N L S . C N I I 0 I N O X ( F N C A E N 5 O N G 7 E C N N 3 P I I T R N T P N E N A A F N E L S N H O N N. O C A R T w\= N N A T A E N N U H N N H
/
N N /
/
N N N H N N H
/
5 2 N 1 N N N M s N N M N H N N /
/ , =
o U
%g mp. . olw t > Mj 5>" 3OO$ yo:c
- 9n 85 g ,tl jC r
w"
i s i i Z
. O i _ 192.5 J.
~
21.4 , l I 4
' w o
i n v2 1 o v2 N H f o m o r-a m , t i .
- I:
O m N ,, i H I i l 2 oo -a
-R
/////////
4 i t i Fig. B-2 ANALYTICAL MODEL FOR STRESS 1 I l 34
. - _ - , . _ .-. . _ - . . . . ~ . . . . . . . _ , - - . - - . - - . . , - , ..
Z o e e . _ _ _
%~y l
it
- : : : : o a o o o o o <> a e, o o.
- : : , = =
= = = = =
4> t> <> 46 th (p lb il <> tb ID 4b
, : : .c : e e : : : :
el er <> 41 46 ID 4> d> 4D 4> 4> ( >
- : : : =
4> il 4> q> 46 di m ; ; ; ; lm
. _ _ _ _ n ; e, <> di o o o I o ti <> e 4> El , dp 4> 4> di 41 4>
c - - - - - N a, ei o ai q> 0 N e = ='
= =
dp 46 EP o di al 4> li il (> 45 di E* O : : : : : ; = : : : : H 3 m ii <i e, si i, <> m A H H O : : : e mi O 7% ii o ,, o <> o # I = = = = =
,, ,, ,, ,, ,, 4,
= = = = = = y Fig. B-3 FEM MODEL 35
l-m o o co - M h _20 )to u M i
' o e-4 o x
' w
-15 i 2 i
O H E4 I r - - -- = z I _10 W i 0 l i B I d x l _J - 5 g i I_ w I Ed z
, H p_ _ .a tr w
- 2) 3 i
i )
- 21.4
=
WALL THICKNESS i i I t i Fig. B-4 INTERNAL HEAT GENERATION l 36 4
- .. . ._ ,_ . - , - - - - - . . _ . . _ _ , ~ . - - _ . . . . _ - .
]
0
' 1
~
C C C 0 _0 0 0_ 0 3 4 5 0
=
C-2
=
T T 5 E 7 R
'0 U T
A R E P M E
) T
( T c A E N V I R A U R C 5 T i 0 S N I - A
- R T
S
- S S
E R
~
T S
- 5
. B 5 2 .
# 3 0 g i
F - I/ d - - 0 0 O 2 1 _N hNw j_ gem 4 w
* ,i ! 3 ! 1 ;, : ,i1 a
i 125 HEATING POWER: 300kw g g COIL I HEATING t180Sec TIME TEST PIPE (12" x SCH100) 800- COOI. ANT (0.72m/sec)
.O OUTSIDE SURFACE ~ ~ "'~ CALCULATED
}
- ~ ~ * - -
600- / su g ll u -. -a- -- L--a ---
- o. 400-5 f* INSIDE SURFACE
\ -a- EXPERIMENTAL I 200- o-- - o-- e ---o .,
l VALUE i i i e i
- Qw.
0 20 40 60 80 100 120 140 160 180 200 220 240 DISTANCE FROM COIL CENTER (mm) 25-
-o- AXIAL } EXPERIMENTAL 1 - ll , #.6., , -- a-- HOOP VALUE 1 o-1& .
m.J 0 s 5- OtfrSIDE SUPTJCE
' *f
* \ Y, y, .- a,w -
~ 5- .
3 th. g g h ^* . I I N -S w l-e-o AXIAL yCALCULATED
--a -- HOOP ME Fig. B-6 EXPERIMENTAL AND CALCULATED TEMPERATURE AND STPESSES FOR PIPE R-21 38
- q. .
s 4 d OUTSIDE SURFACE t 1 j -o- AXIAL j t HOOP i i [ 7 1 i INSIDE SURFACE I k i e i ;
-30 10
-20 -10 0 20 30 l
l RESIDUAL STRESSES (kgf/mm2 ) e f Fig. B-7 CALCULATED THROUGH-WALL RESIDUAL STRESSES 39 i
. , , - - , .. -e-- . , - . . , , _ , - -
, , - . - , . - . , . , ,_v, . - . . - ..-.~.-,.,-c-
750 250 N p INSULATED
- /
l 2x
.. . _ ,_, .._, Uoo = 0. 7 2m/s T = 22 *C Wo=58.3 kcal/cm3 h S = 9.6mm N
4 ('C) 600 500 160Sec N a: 180Sec E EN s/
!' D N N e
300
~
N\
^NN q w N A
40Sec 20sec 100 OSec N N I 21.4mm (THICKNESS) Fig. B-8 CALCULATED THROUGH-THICKNESS TEMPERATURE , DISTRIBUTION 40
I TABLE B-1. MATERIAL PROPERTIES USED IN FINITE ELEMENT MODEL TEMPERATURE THERMAL SPECIFIC THERMAL CONDUCTIVITY HEAT DIFFUSIVITY (*C) (Kcal/ca-s *C) (Kcal/kg *C) (cm2/s) 20 34.7x10 ' O.107 50 35.7 0.112 0.0390 100 37.2 0.117 0.0399 150 38.9 0.122 0.0409 200 40.5 0.125 0.0420 250 42.1 0.127 0.0432 300 43.7 0.130 0 0439 350 45.4 0.131 0.0447 400 47.0 0.132 0.0453 450 48.6 0.134 0.0462 500 50.3 0.135 0.0470 550 51.8 0.137 0.0478 600 53.5 0.139 0.0487 650 55.1 0.141 0.049 7 700 56.7 0.143 0.0506
~
750 58.3 0.145 0.0514 800 59.9 0.148 0.0523 TEMPERAT'JR2 YCUNG'S MODULUS POISSON'S RATIO DENSITY (*C) (kg/cm2) (kg/cm2) 20 1.98x10 ' O.260 8.03x10 8 50 1.97 0.264 8.02 100 1.95 0.270 8.00 150 1.91 0.274 7.97 I 200 1.88 0.278 7.95 250 1.84 0.281 7.93 300 1.79 0.284 7.90 l 350 1.76 0.288 7.88 ( 400 1.72 0.292 7.86 450 1.67 0.296 7.83 500 1.63 0.300 7.81 550 1.58 0.304 7.79 600 1.53 0.308 7.77 650 1.47 0.314 7.74 l 700 1.41 0.318 7.72 750 1. 36 0.320 7.70 800 1.31 0.324 7.67 41
- - - , - e -
TABLE B-1. - continued l i TEMPERATURE THERMAL EXPANSION THERMAL EXPANSION l (INSTANTANEOUS) (MEAN FROM R.T.) l 1 ('C) (1/*C) (1/*C) 20 16.4x10 ' 16.4x10 8 50 16.7 16.6 100 17.1 16.8 150 17.5 17.1 200 17.9 17.3 250 18.3 17.4 300 18.6 17.6 350 19.0 17.8 400 19.3 18.0 450 19.8 18.2 500 20.2 18.4 550 20.5 18.6
'600 20.9 18.7 650 21.3 18.9 700 21.6 19.0 750 22.0 19.1 800 22.4 . 19.2 4
42
._. .=. . . . . - _ _ . _ _ . _ . . _ .
I ! SUBSECTION C i ! COMPUTER CALCULATION BY FEM (II) Computer calculations were performed to investigate the effect of the weld F deposit on residual stresses produced by the IHSI process.
- 1. Finite. Element Program Program name: ANSYS i
l Strain hardening rule: Kinetic strain hardening rule 3
- 2. Boundary Condition and Calculated Model
- Fig. C-1 Model for temperature calculation i
Fig. C-2 Model for stress calculation-Fig. C-3 Finite element model Heating pattern: Fig. B-4 Material properties: Table B-1 i The yield strength of the weld metal was assumed to be 140% of the base metal at every temperature.
- 3. Results l
l Fig. C-4,5,6 present the principal results No significant change from Fig. B-6 and 7 in residual stress distribution was observed when the weld metal is added to the finite element model. l t i i i I I 43 I
. 1' ,' " .
4 i e wve oit ug i d n HziC5sMO E ,l
)
C 5 E. . 3 .
= .
T, c e . s .
/ E .
m P . 0 O 3 L . S ( - S . I . X T '0 . A _ N 1 . A . E L P O . I P O C
# . />
/
/
/
\.
i
/
/ .
/
/
2 m m _ , e i
/ 0 0
, . 5 5
. . / 2 7
/
/
. /
. /
. /
. / -
1 mHc. oIH >2>M 5 n $ : roc i4 mM$aC M. s-
- 1 4 ' 4';i4 ,
t
i i I
@ECS 2.sw - p.
Z i n i ! 21 4 .
\ ; _ _19.25 4
j i _. A% WELD METAL F
/ : R i
1 1 i JM
//////////
4 t l t f I i 4 1 Fig. C-2 ANALYTICAL MODEL FOR STRESS t 45 i
._ s . _, . _ . . _ .. . . _ . . _ _ . _ _ . -. _ _ _ . _ _ _ _ _ . _ _ _ _ _ , . _ . _ _ _ _
m ns m
..e w .
m
. .. .. c.
m NS
. en .es 955 in us in
.p. .e .s. as
.g ..g is l
Y hw . . . .
, Fig. C-3 FEM MODEL
.0
. s. . . .
.j ./-;.
- s. .. . .
46
lj i E E C C A E A F C F R A R U F U S R S U ) R S R E E m T D N ( U I N 0 . G M I 3 .0 2
- - _ R o o g E T
- - _- N
- - E
- o- .
C
- - g,-3' -
L o L I .\oo ' I O N Z O o\o\ o C O C - I G M T N ,N O U B I o o 0 R s o. i 0 F I T 1 R A g - E T E H o y' o C N S I sC -
.. A D
/
- _ s'l o o
- T o_ S E m5 1 I R 3 3 -
D
, 0 . U 6 .
- T m - A
\ m _ -
- R 5l E o
ii
/ .
2 1 o o- P
- - M 0o 0- ,D'0 0 0 0o E 0 0- T
, 4- .3 2
;'l, R
0 0 o o- - M 0 D o 0 E _U c.Eua o o- T
- - - A 1 m - - - L 1 n 0 O o U 5 - ' - C 2 -
- L
. 1 ~~~ ~
- - A
- o C
~ - ,
/
wl ~ ~
- 4
- o -
~ - C
) 'o *
"o o g s- - i 0
[o 'o o
, ,0 1
F o o/. o,
/ /o o
0
,0 2
u4 7 lllll1
l ' l l l il,lI,1\ I I 1 s
/
= M
--' ~ O >
R F m 7 F, L k 1 A A ? TL SI I O IO X O D A H h[C
# o- 4 -
E C A
,s' ,n-
' F R
E U C A , c .[ o S F ' , E R ~ - D U .
'- s I S
E D r,
'a S
N I S I I e E S ,s 1 l T U O b
- 0 0
1 u\ S S E R
. " - T S
) "
. s A, a L
A U S m
- ' ,u,
~ aw95 d* D I
m-2 S o S/ ,o E
,WFms Ef R R g Tk
.,7 u S(
- 1, D
- o. E 3 0 0
1 =0 0 3 A T 3 M, - .
,*- 1 0 -
9, 2 - L _ - - - U R C
,, y c.
C L A C
.--. o.. - -
m . c 5 s - o. - u -
. - C
- a
- o, ?.
g i F
}
0 0
=z.
, , 'l
,'a 0- c i-
**88' k
o
's i .
e
'9 .
bw
R A
\:/
l A J Z e AXIAL
- HOOP l R 19.25nm 1/\
o I / OUTER SURFACE 15.40-oo-11.5)#
./o/
./ o/
ao
/7
/ 3.85--
y/
./ /
o
/
INNER SURFACE i
. .? . . .I . -30 -20 -10 0 10 20 30 (kgf/mm2)
STRESS PERPENDICULAR TO SECTION A-A Fig. C-6 Ct.LCULATED RESIDUAL STRESSES i 1 49
f A-4 IMPROVEMENT OF RESIDUAL STRESS PATTERN IN PIPE WELDMENT (4TH REPORT)
SUMMARY
In order to verify the effectiveness of IHSI (Induction Heating Stress Improvement) in the presence of some crack like defects near the weld, a cracking test was conducted on two pipe weldments in boiling 42% magnesium chloride solution. Two small cracks were introduced by electric discharge machining on each pipe and.one weldmont was heat treated using IHSI prior to the MgCl ra king test. 2 Liquid penetrant tests and microscope observation following the MgCl 2 exposure revealed that no crack had initiated on the artificial crack treated by IHSI. However, the untreated pipe exhibited some cracking. i 2
I
- 1. Introduction It had been previously reported that the residual stress pettern of a pipe weld joint can be altered by high frequency induction heating on the OD surface and simultaneous IC ;ooling. This treatment is named IHSI (Induction Heating Stress Improvement).
In the actual plant in service, some defects are anticipated to exist near the weldment and therefore it is one of the most important studies to
- verify the validity of IHSI in the presence of some defects.
In this study, IHSI was applied to a pipe weldment with notches machined' by means of electric discharge. This paper describes the results of the cracking test conducted on the notched pipes by exposing them to boiling 42 wt% MgC1 solution. 2
- 2. Experimental (1) Material The chemical composition and mechanical properties of 4B schedule
,i 80 austenitir ;s steel used in the experiment are given in Table 1. (2) Welding Two test weldments were produced by automatic gas tungsten pulse-arc welding. Their welding conditions are given in Table 2. (3) Notch Preparation Two notches per one test pipe were prepared by electric discharge machining. Their dimension and location are given in Figure 1. The deptn of the notch was prepared to be about double the allow-t able planar surface indication of IWB in ASME Code Sect. XI, in Service Inspection that is 2 x 12.5% wall thickness. The aspect 3
ratio was made to be between 0.20 and 0.25. l (4) IHSI IHSI was carried out on one test pipe under normal conditions and ! the temperature profile at maximum heating condition is given in Figure 2. No treatment was performed on the other pipe. (5) Cracking Test Both test pipes were irmersed in boiling 42% MgC1 solution for 2 48 hours. Af ter this treatment, their inner surface was observed by liquid penetrant examination. Micro-photographs near the notch tip were taken in order to check the existence of SCC in the direction of wall thickness.
- 3. Results and Discussion Distribution of maximum heating temperature during IHSI is given in Figure 2. The significant temperaturc difference, AT = 20Y (1 - v)/Ea, was obtained in the range exceeding the width of the coil.
The results of liquid penetrant examinations are shown in Figure 3. Fig. 3 (a) shows the notch by electric discharge machining. Fig. 3 (b) and Fig. 3(c) show the results of liquid penetrant examination for pipes with IHSI and without IHSI, respectively. No cracks were observed on the inside surface of the pipe with IHSI while many cracks were found on those without IHSI. Fig. 4 and Fig. 5 show the results of the microphotographic examination near the notch tip. No SCC was observed at the notch tips in the pipe with IHSI, while significant cracking was found on those without IHSI. It may be concluded that the effect of IHSI for a pipe weldment with cracks is significant for mitigation of SCC caused by welding residual stresses. 4
- 4. Conclusion The results obtained in this study led to the following conclusions:
(1) IHSI treatment can also improve the stress around the notch near the weldment. (2) The application of IHSI to a weldment with an existing crack is found to be significant in preventing IGSCC nucleation as well as propagation of an existing crack. 5
- - ._. . . _-. -~ , _ . . -
J I 4 Sch.80 SUS 304 TP C Si Mn S Ni Cr Specified Yield Point Tensile Elongation Strength men. max. man. max. men. man. 21 53 35
"'8'** 0.Os 1.00 2.00 0.030 11.00 20.00 (ks/mm2 ) (gg f,,2) (2) m TTC 3750 0.05 0.50 1.65 0.006 9.10 13.20 29 59 66 Table 1. Chemical Composition and Mechanical Properties of 4 inch Schedule 80 Pipe Used in This Evaluation.
1
I V m co M 8 d i d S 150 E 150 _ 4 N 8 " " Position Layer I* e t H, p Welding Rod ( C) R-53 50 7 % 10 1 70 9.9 fT 1G 2%3 60 % 170 8 % 10 6.1 % 9.1 d1.2 E308 1 70 R-54 4%5 180 N 230 10 % 11 1 70 15.4%19.9 Table 2 Wel_ ding Condition
0 9 5 4
// =- i n
o 1 t X 5
- c O1 O
0 8 0 e S ( i 9 A A s 5 e 2 P h 2 T c t 4 o 0 N 3 M
- S D 0 U E 7 S 2 0 f 8 o h n c o S i t
N a I r 4 u g i f n
# o C
\
1 7
~A
, A V g
_ i F 5_ 0 0 3 0 u
$N
50 50 Heating coil 140 kW, 30 sec. 4IN Sch.80 SUS 304 TP
? N/ l 4
600 - - 600 U
\
- o 400 - .u 400 #
s 5
$ e A EP 200 -
R B
\ - 200
,~, x n(
l I I I 40 20 0 20 40 4 Distance from coil C Fig. 2 TTD Curve (TP-No. R-5 3) 0 9
l (a) EDM \s S Ogg\ ., 5- \ 0s\\ [EDM: ELECTRIC 'N 0, 0'g1 DlSCHARGE '
'~; 9 '4-( MACHINING / ~ ad 's , ,' - ~^
a- ,
- < :q '
cri G .f _ _ m
'*w]l.q; ;
V (b) EDM i e M' '
~
~
Mgcl2 :- lIrg'Istigl'!!!tig :
,'%I
~ t s ;l'a ig '
Propagation - -
'w. .. N e -
111
*tG .,
(c) EDM ' l ~. lHSl
= I ' ' ' 'I t t i ,l 1111, : ,,' ,
IZ11,,litg,
~
i Mgcl2 s 1
,- m . ..
V' .ll[Ill}' l m l l No Propagation
~
o g' C 1 l Fig. 5 Propagation Behavior of Notch l 10 W I She _ _ _ _ _ - __ - ,,, - - - -_ - - - = _
i 1
' - q: .ig ,,,, e; ~*a- .'.
; s)o'.ho~, 'R. 4.a
--- ~ , f..,,e;. . <-
i,'Szc -(,;;
'r. .pyy%'
, ..,s .
~-
\ 1 4 s:g ..Q g s ..
e , ,. /. . _ .
~. ..
i .,
*.3. %'4 -
W ,, -m > 31 *' " ~ < -,'~,44.~f3. 4 .7 i _ . _\ . _ . . .J .c
. , g. , . / g,gv; .9 - -
- g. _p
.. . g .. , .C.AA-~- cl
. . x ,
s L 7,,4 ,i , r.'.;
, i
, -, g ,. , f. . ,', .e 4 t.
WELD DEPOSIT R-53 45' g* . s% - " ,
';;'~ .
y h* pe
. .._ .- - - z ' ,'p: .
~j, l 2 g m ,A; v2 ~. ,
^p$
% v '_ h f TI.V E,.
.~ _
. . x ,,.
y ..T' K,
.. 3. ..
's &- *%
.e p e- %g - -
A y ,#
~
!iN r.
. ,', 4. ' c;w. , s,.: . - 7.
- e. kT-L'.4 uy u,
-~
. J;, _ .u'a .:,:y w:..'%g h f
s
\~h e . ..... :a .- - - .: .
- l l 4. E ,*), > f .'; - - h l' ,*
- l .: ~
^
T .' d%-% .J : . . .r #
?$'! , 't ,n ~{Q . ..' U ,_'h{ y,f a
R-53 224* FIG. 4 CRACKING OBSERVATION AT NOTCH TIP (R-53) MAG. X 100 11
, g , n
; ,-~ .
.:p.r-I ". ...e,': *r..
. .L -
,g- _ s y,i -
.; g- - 3 .
.. 1 , .
- e. , =.. -
....-w ,
. . . .- .. o. .
~ ~ .
,.', . - / -
- , %' 17/ds e
....,,L'/*
. . . . WELD DEPOSIT
#4
* .. ' as . = , ., 5. .. ,
,5 er
- g
* &, . ^%., %' d .
. _* '4
'A <g
- j *,,,8w 1 -
, 4. s= , * *
- g.
..*,,T
* == * * . -p g , . *g g;
} g.g*, ,.'. '.,p, r,,. ****
%,.. . . s .t s
#- = *=V g, ',e, . , 3 - ,,"',,p L . k , f *.
* ., .s g,f -
I*g..
,- N ,,1
- 4 - *.
p .. # -, ' t v v . v.,. * .
.' +
,p c, i g..
&, n.
- f ** ' *,g ,
* *. K'sj ,
[* "
%.n.- -
r _
,or,...
'Ny,# *
- D .M. .s.s a
m ,.
, w
.y.
,g 4
r~ ;y>
~
x ,* ? , Y d2 ./ * '? . : , g . . *. . 4 E' 9,) *, * * , q. . - f, ,. c ; f 4.,? gs ,, y + 1, ,,. . , - ,., , ,
.{d.
o c! .. .-c ' ,k , N. . fl. ., ~***{ : .-
**T.
, A'
- $y 4 .*d *p 4
. *; 7 ',*~,c y .~, .,.
4 *
'(['.
v
. /[*** f "J.
v s?
' . .*(
6~. , , J . v4 - ;, ',g . v. .; , ~*'.,k
.y _-
- e. y , ~;tr;5.
. .y A.
,1
. g ?
t ,. , F -
%,a -, t
. ..,p= * " g ., ,
L. F .-.. y;ms'" 5 g .,e . 1
- 7y g, ,,LY,.,. .y ,.* ! .,** , ,
s, @ ,6/ 2
$.-% 4 ;-e ,,' . -. .
~ ,g
.+y f*... .? - e- +. .-.~
g
- ~
s yv ,
- (, ~ $^ *,c..- --
%.ig* .W,' .
.i '};. , e ... h . , ;y.M,'s, ',."Q.,' f, '
f' 's s, .- 1
- W
.g u.e . . , . .
,. n-sm. .a r . , . A..- .
- n. N. '=.
.r. >~,,t_- , ~.,,c,
+C...
- 4. ...-
4 s.. g. ... n. ' a . , m. m. ' r
, ., , 4 .
3 '-
=.t.,.-
,:g , .y ,';
. s. y M T,;
, , ,. m _.
n r. .- ~i ** J. s, , 3 , _ w" . u ,
? T~ - M-...h *7. ,h.
. ', [. +> w j,j ,. ,-,' h. N. .-(
- . - , ;? , .1.
. ~ ,
'q : < < , p
~
. m .: ...... %c,'64
,qy 3&s .pf ,.
.-i % ,__ - ..2.
'y,.,,'a-f '. .
A ' ~
& _;.';G n .
- 4. ~ ' _ ~ {. W :" . * *
.% . i . }'7 * "'T'?4 41i-W.'4.,. ? 't aN. ~'
_ L-_ pen
- Y p # = g s'.g 2 . A . *. ' --(i, . c,o. . '- >
,k .- r.;
C ..
. ' 7,. . y6- y. ;5 ,_. R-S4 y3$o
.. <r e- . - ^
4 . A, e
., tr., ,/
u par .
.u.h ,p c < . , A ., $
}..:7 .K. a.s , *
+.
.- : ;J ... ...,.R. . . y , m, > <. .,
w M ,
'.'. - ~ 4 *
. --f,W. 4./) p*. , . 'g t.
** u :. M
- .'-:n..
1 R-54 315* 1 l FIG. 5 CRACKING OBSERVATION AT NOTCH TIP (R-54) 12
A-5 IMPROVEMENT OF RESIDUAL STRESS PATTERN IN PIPE WELDMENT t
SUMHARY When IHSI is implemented in an actual plant, some trial heatings become necessary to confirm the heating parameters. In this investigation, the effect of repeated IHSI treatment was investi-gated using 12 inch pipes. The residual stresses produced by multiple IHSI heating were compared with a normal single IHSI heat treatment. 'The results indicated multiple heating produces slightly more compressive residual stresses than a single IHSI heat treatment.
- 1. Introduction It has been well documented that the residual stresc pattern of a weld joint in pipe can be improved by high frequency induction heating and simultaneous water cooling. This method has been names IHSI (Induction Heating Stress Improvement) .
When IHSI is applied to actual pipe weldments, it sometimes becomes necessary to make a trial heating in order to determine the hottest location and to determine the heating power to be used. This paper describes the results of tests carried out to study the affect of multiple heating cycles on the residual stresses.
- 2. Experimental The experimental sequence used in this investigation is shown in Figure 1.
2.1 Material The chemical composition and mechanical properties of 12B schedule 100 austenitic stainless steel used in the experiment are shown in Table 1. 2
f Table 1 CIIDtICAL COMroSITIONS ANG MF.CilANlt:AL PROPERTIF.S 12" Sch.100 SU5304 TP Spectfled C St Mn P S Ni Cr field Point Tensile Elongation Strength max. max. eaa. max. man. 8.00 18.00 21 min. 53 min. 30 min. Charge No. 0.08 1.00 2.00 0.040 0.030 %t1.00 s20.00 (kg'/ 2) (kg) 23 (g) TfD 5156 0.06 0.59 1.84 0.020 0.050 9.65 18.70 22 55 74 2.2 IHSI IHSI was performed in the sequence as shown in Figure 1 and Table 2. The welding conditions are shown in Table 3. For R-71, the initial 10 heatings simulated the trial heating to check the hottest point and the electric power to be used, and the last heating simulated the normal IHSI heat treatment. The first four heatings of R-72 simulated imperfect heatings which might result from disconnection of thermo-couples or other reasons. A similar heating sequence to that described above is considered unavoidable when IHSI is applied to actual pipe weldments. 2.3 Residual Stress Measurements Residual stresses on two IHSI'ed pipes were measured by conventional strain gage methods.
- 3. Results and Discussion Figure 2 and Figure 3 show the maximum temperature distribution during 10th and lith heating cycle, respectively for R-71.
For R-72, Figure 4 shows the maximum temperature distribution during 5th heating cycle. The residual stress measurement results for R-71 and R-72 are shown in
- Figure 5 and are compared with results for R-21 on which one normal 1 SI has l 3
been' conducted. The residual stress at the coil center of R-72 was the maximum, and R-21 was the minimum, among three while the temperature differences were the same. It is clear, therefore, that repeated heating produced larger compressive residual stress than single heating. This phenomenon is ' thought to be caused by cyclic hardening characteristic of 18-8 austenitic stainless steel. A similar effect was also found on R-71.
- 4. Conclusion j The results obtained in this study lead to the following conclusions.
(1) Repeating IHSI further improves the residual stress pattern of i a welded austenitic stainless steel pipe. (2) The preliminary heating proc-as contributes to improved residual stress for an IHSI treated pipe. 4
i i l i 1 (1) R-71
- Max. Tump.x Repetitive No.
300*C= 10 + 500*C= 1* 12IN sch.100 SUS 304 (2) R-72 RESIDUAL STRESS TP-NO. R-71, 72 500'C = ,5* _ MEASURDEENT 1 m Fig. 1 Test Sequence ' I l
TAht.E 2 IHSI HEAT TREATIGlNT CONDITIONS e
' I" I "
. Frequency Power ""'" "E '"' ""4"#"""'" (
TP-No, Materla! Repeat Duration Cull Cooling T Ngid No. (kitz) (kW) (sec.) (mm) OD 10 (og,-I D) I 1 180-200 60 298 66 232 2 150 100 317 83 234 3 1 30 145 293 79 224 4 120 180 283 76 207 5 1 30 -171 295 105 180 R-71 s/100 6 130 180 280 76 204 7 3 1 30 162 250 Ci 295 76 219 Spray 8 130 180 295 73 222 (3m3 /Hr) 9 130 180 297 76 221 to 1 30 180 300 74 226 11 250 180 491 76 415 1 230-240 180 478 81 397 2 240 ISO 496 82 414 g.72 3 240 180 493 79 414 4- 240 180 499 80 419 5 240 180 499 79 420 R-21 435 Reference
I. L V so. 2 8 __ J i n 3 150 150 O l 1 Welding Current Voltage Inter Pass a Method y,,ggg'" g,I Heat input (A) (y) 7; . (KJ/ca) elding b d R-53 1 60 % 150 7 % 10 1 70 9.9 Auto CTAW
- 1C 2%3 60 % 170 8 % 10 1 70 6.1 % 9.1 d1.2 E308 R-54 4%5 180 % 230 10 % 11 < 70 15.4 % 19.9
_ Table 3. Welding Condition
j 4 125 125 _ R 12IN Sch.100 SUS 304TP l , 600_ _ 600' U 400_ _ 400 S s a N
- 200 _ [. :200 N
\ , ,
O *
- 0 i i i i i i 120 80 40 0 40 80 120 DISTANCE FROM COIL CENTER (mun)
Fig. 2 TEMPERATURE DISTRIBUTION (R-71, IMSI REPEAT NO.10) ! 8
t 125 125
=
,/
12IN Sch.100 SUS 304 600- - 600 N
- 400- ,o
-400h>
I -
. e n 200- <a
- 200 I ha 0 5 "
# 6 1 1 o
1 1 120 80 40 0 40 80 120 DISTANCE FROM COIL CENTER (mun) Fig. 3 TEMPERATURE DISTRIBUTION (R-71, INSI REPEAT NO.ll) I l 9
1 i l 4 125 125 i 12IN Sch.100 SUS 304TP i j 600 - 600 U
' ~
400 -. 400
. 2
, 5 I
200 - b 200 N I M- , 0 NN 0-1 1 5 1 I i 120 80 40 0 40 80 120 i DISTANCE FROM COIL CENTER (nun) i Fig. 4 TEMPERATURE DISTRIBUTION (R-72, IHSI REPTAT NO.5) i l 10
-_ _ ~ . _ _ _ _ - _ _ _ _ . - _ . _ - - _ . _ _ _ - _ - _ - _ . - . _ , . - _ . - . . - ~ . _ . -
.~ . .__.. . . _ , _ . -. _. _ _
i-4 i I t 12IN Sch.100 SUS 304 / o e R u i1 ID 4 p",'-g 9s-~s 20~ y s o R - 72 CD
~
b W'- o --- a 3
'g_, -
. - 21
- 10~
%'.o-- -o,.~4 3 AXIAL
"*. \
",o 1 * 's' 's' -o .
s M . y.
- r N 0 m
/ . .
l m
-10~ ~
I-se .- , m m/ s-c '*
-20~
k 1 . ! l ' l l 0 40 80 120 160 200 240 30-DISTANCE FROM COIL CENTE" (mm) 4 4l
*Og ,0-*.
s
- i
's 20 -
l
.:,x:- .. v.
i s
; o N%
i 's
's N, gs i y 10-ko i ff00P N -
es, , - 'e..,,,,,,,,,e
.a 0
\ o o
% D.
l " m *~ l I Q ,. I 1o-
; /
\ / .'
l V* n y,. *
/
L.W
~'"~
i
,/
i 2.- v I Fig. 5 RESIDUAL STRESS DISTRIBUTION (12IN) L l l 11 i i Y a _.m , -e,s -m-_e,.-- ---- um --u . , . , - . . - ' ~ , - ,+. g e- +-m4 -y .p w w- p. - - . . - - -_ _ - - -- - - - , --,3 - ..
l A-6 RELAXATION STUDY ON IHSI I I i l
SUMMARY
Some concern exists that residual stresses produced by IHSI may be relaxed by temperature and/or external stresses during plant operation. The relaxation behavior due to temperature and external load were separately investigated in i this study.
)
- Three IHSI treated pipes were subjected to heat treatuents of 300 C, 400 C and 500 C for 24 hours respectively and another pipe was stressed by axial load exceeding yield strength. The residual stress measurement after these treatments revealed that the residual stresses were hardly lowered by thermal treatments up to 500 0 but were easily relaxed due to axial stresses exceeding the yield strength.
The above mentioned tests were conducted on pipes without circumferencial welds. Weld joints are expected to respond somewhat differently from base metal and an additional test on a weld joint is strongly requested.
- 1. Introduction Stress remaining in material is gradually relaxed with increasing temperature rises and/or with time. It is, therefore, of concern that the 4
residual stress induced by IHSI may be relaxed after long service at high temperature. In this study, two kinds of experiments were carried out. They were designed to check the relaxation behavior of IHSI'ed pipe due to elevated temperature and to axial stress.
- 2. Experimental 2.1 Material The material used in this experiment was type 304 austenitic stainless steel and its chemical composition and mechanical properties are given in Table 1.
2
2.2 Preparation of test pipes Four test pipes, R-73, 74, 75 and 76, were prepared fc,r this experiment. These pipes were 4 inch Schedule 80 and 300 mm in length. 2.3 IHSI IHSI was performed for these test pipes using the parameters described in Table 2 so that the pipes were subjected to a temperature difference of approximately 300 C. As an example of the temperature distribution during , IHSI, R-75 is presented in Fig. 1. The other test pipes were also IHSI'ed in the same manner as R-75. 2.4 Relaxation induced by high temperature To estimate the relaxation effect of BWR piping in service, (300 C for 40 years), three heat treatment conditions were selected. As shown in Table 3, the test pipes were' heat-treated at 300 C, 400 C or 500 C for 24 hours in an electric furnace in air. 2.5 Relaxation induced by axial stress In order to investigate the residual stress relaxation behavior produced by axial loads, a tensile test was performed on IHSI'ed pipes. The sequence of the tensile test is shown in Fig. 2. The configuration of a test pipe is i shown in Fig. 3. Tensile testing was performed at 290 C by using an Amsler type universal tester. The displacement-of a test pipe was measured by high-temperature strain gages fixed on the inside of the test pipe. 3
2.6 Residual stress measurement l 1 Residual stresses of the IHSI'od pipes were measured by conventional ' l strain gage methods after heat treatment or tensile tests had been performed. I
- 3. Results and Discussion The results of the residual stress measurements for pipes, R-73, 74 and.
75 are shown in Fig. 4 and are compared with E-15 which had not been subjected to the relaxation heat treatment. R-15 has been similarly IHSI treated with the other pipes as shown in Table 2. No significant difference in residual stress was observed among the pipes tested. In order to estimate the temperature effect for a longer time, for example, the design lifetime, the Larson-Miller equation was used for cxtrapolation purposes. The calculated service times corresponding to 400 and 500 C for 24 hours are 15 years and 8 x 104 years, respectively as shown in Table 3. These data show that the residual stresses produced by IHSI do not disappear nor vary during 40 years at service temperature only. Figure 5 shows the load-elongation diagram obtained in the pipe loading test. The test pipe was unloaded to zero axial load after the naminal stress reached 0.5ay, oy and 1.29ay. Using the residual plastic strains, the residual stress after unloading were estimated by a simple two element model as shown below. Figure 6 (a) illustrates the concept of the two element model. Element A and B correspond to the outer and inner pipe wall. The , inner wall has a compressive residual stress and vice versa. i 4
i Figure 6 (b) shows the stress-strain relation of elements A and B assuming that the elastic perfectly plastic relationship holds. Curve "C" is composed of curves A and B, and shows the . tress-strain relationship'of the composed element, A plus B. The residual stress is maintained at the initial level unless the aprarent stress goes beyond point (1) on curve "C", because the stress goes back elastically to point 0 and no plastic strain remains. If the stress goes to point (2) which is beyond point (1), the residual stress is decreased to level (4), because a plastic strain of c 3 remains in the composite element after the axial load is removed. If the axial load reaches the yield strength, the residual stress similarly becomes zero. The residual stresses corresponding to 0.5 oy and 1.0ay loading were estimated in the same manner as mentioned above but in this case the stress-strain relations obtained by the pipe loading test and small tensile test were used as those producing the composite element 04 + B) and element B respectively. The initial residual stress (at zero axial load) was assumed to be -oy. which is nearly equal to R-15. The residual stresses on the pipe after 1.29ay loading were measured and are p~atted in Fig. 7 with the estimated values. . Figures 7 and 8 suggest that the axial external load exceeding oy lowers the residual stress to approximately zero but it never shifts to tensile stress. The previous discussion is based on data for a pipe without a weldment. , The stress-strain relation at a weld joint is considered to be much different i from that of base metal because of strain hardening and deformation during welding. Therefore, the relaxation behavior of a welded joint will be different from that described above and experimental studies on a IHSI treated joint is recommended. 5
- 4. Conclusion The results obtained in this study lead to the following conclusion.
(1) The effect of IHSI is reduced insignificantly by BWR service temperature. (2) The effect of IHSI on a pipe without a weldment nearly disappears due to an external load exceeding cy. (3) The effect of external loading can be qualitatively explained by a two element model. (4) It is recommended that a tensile test be performed on an IHSI treated pipe weldment because the stress-strain behavior of the weldment may be substantially different from that of base metal. Reference (1) Y. Ando et al, Local Structual Behavior and Safety Analysis in Nuclear Piping (Vol. 3) , Japan Welding Engineering Society, 1979. 6
Table 1 Cl!EMICAL COMPOSITION AND HECHANICAL PROPERTIES SPECIFIED C Si Mn P S Ni Cr max. max. max. max. max. 8.00 18.00 0.08 1.00 2.00 0.030 0.030 - 11.00 - 20.00 TTC-3750 0.05 0.48 1.65 0.026 0.005 9.20 18.20 w SPECIFIED Room Temperature 290*C YIELD TENSILE ELONGATION YIELD TENSILE CilART ELONCATION POINr STRENGTH POINT STRENGTH NO. 21 53 35 (kg/mm2 ) (kg/mm2 ) (%) (kg/mm2 ) (kg/mm2 ) (%) TTC-3750 29 59 66 17.7 45.1 46 l
3 3 0 I S - K 1 R 5 A 4 M - E E R C T I I E C h E 7 1 6 2 2 4 1 8 5 2
) 2 3 2 2 3 C F
- I
( D E E R D 0 5 0 1 0 U I 4 8 3 3 4 T S 2 1 2 2 2 _ A N R I E _ P M E T E D I 7 6 6 3 2 S 8 9 1 8 6 T 4 4 5 4 5 U O N O I T I D E ) N T . O A c C R e 5 s l W / 2 l O m _ Si L ( I F 2 e N l GO b NI ) a I T . 0 0 8 0 1 T TA c 3 3 2 3 2 AR e EU s HD( R E ) 0 0 0 0 W W 3 3 3 3 O k 1 1 1 1 P ( Y _ C N E ) 3 2 U z Q H E k R ( F 3 4 5 6 5
~ 7 7 7 7 1 o
p R R R R R o o
! i li 1 ' :
4 iI
HEATING COIL 50 50 PIPE 4IN SCH. 8 0 R-75
}
i l 20 3
- o o
_ g E4
- @0 @
m
) -
4
- 5 e
200 i i i i i x:s i5 , 40 20 0 20 40 DISTANCE FROM COIL CENTER (MM) 1 i Fig. 1 TEMPERATURE DISTRIBUTION DURING IHSI 9
.- - . - . - - - - - . - .- -. -=.
i y e
~
E C N E U Q E S E G M N I I T D A O
- L 2
g i F Y Y Y 0 o 0 5 5 1 0 m $:a8m ACZs2oZ m_ go
6 2 G U h) L G N I D C% A h O L
. T E S -
G E A T G G N N I I . A D R A - T O S L l 0 0 3 0 6 9 R O F E v.
! P I
P T S , E T 3 g i 0 F 8 E P H _ I C P S T N S I E 4 T f hJ.
=
- 1 ! !'+; l ,
R-15 as IHSI'ed
-- R-73 300*C x 24H
---- R-7 4 400*C x 24H R-75 500*C x 24H 30
/ - - -
$tMs;s
- i b
- D 10 s, l '
[ M 0
,,N I
$ I g ---
m s ~10 8 , - 4 20 . - - " , ~ ~- -' -~- '*- - . a g g O
-30 s,
,e
,,'a=*,,,
-40 0 20 40 60 80 100 120 DISTANCE FROM HEATING COIL CENTER (mm)
,", 30 6
b 20 m m
%--.,, l 's W 10 's s N
m ,
- ' \. -h ,-
N
- 0 x -
4 A l G U
-10
- 7
.-W'p~ *
~
$ 20 j'
-30 0 20 10 60 80 100 120 DISTANCE FROM HEATING COIL CENTER (mm)
F19. 4 RESIDUAL STRESS DESTRIBUTION i 12
80 LOADING g UNLOADING s l.29aY 60 )
)
17.7 kgf/ttm 2
% oY -
f u i
@ f 3 40 -
I n i
" l a
'l
$ )
0.5ay - I 20 ; I u I I
' ' I I I
' I O
I lE I 0.5 1.0 1.5 STRAIN (%) 4 Fig. 5 LOAD-STRAIN CURVE OF PIPE WITH IHSI i 13
1 l FREELY EXPANDED
~%
A - 9 .. . - , B =v I
'C'l PIPE WALL ELEMENT IN PLANE STRAIN COND. PIPE 4 .
(a) TWO ELEMENT MODEL w g Y (A) b p dII 1@ ,
/
-c- h STRAIN
-C o (b) STRESS STRAIN CURVE
-c y 8, AXIAL EXTERNAL STRESS 01-0o oy I
AO
!E HE -
-C o (c) RESIDUAL STRESS Fig. 6 RESIDUAL STRESS ANALYSIS BY TWO ELEMENT MODEL 14 4
l
APPLIED EXTERNAL LOAD 0.5 Y Y O
^
a l[
~
/ -MEASURED k /
* /
m -10 /
$ ESTIMATED BY FIG. 8 a: /
U m
.3 /
gl 4 -
/
5 /
/
m AO7-E E
-20 4
Fig. 7 RESIDUAL STRESS AFTER UNLOADING
.,.u.. J Y
f 4 I i I
- =
g P TENSILE SPECIMAN 4 m 15 - x
~
ll
// -s //
E // y 10 lj o 0. soy
// UNLOADING d / //
d I // c. 5 -
/
f //
/ //
/ //
r // , , , ,
.1 / 0.2 0.3 0.4 0.5 STRAIN (%)
N k -5 -
- s I. 8 6
en , y -10 - a: Ed to ! p +- et g -15 - H (n J Fig. 8 ESTIMATION OF RESIDUAL STRESS AFTER UNLOADING 1
- , 16 i
Table 3. HEAT TREATMENT CONDITION LARSON-MILLER PARAMETER P = T (A + log t), A = 20 NO. TEMPERATURE TIME CALCULATED TIME .; (*C) (IIRS. ) (YEARS) 1 U R-73 4 300 _ i R-74 400 24 15 )
- R-75 500 80000 4
J I 4
l I i A-7 EFFECT OF PLASTIC DEFORMATION BY IHSI i AND SUBSEQUENT HEAT TREATMENT UPON IGSCC i 3 SUSCEPTIBILITY OF TYPE 304 STAINLESS STEEL ]. f a 4 [
- l l
SUMMARY
The effect of plastic strain induced by IHSI on the level of sensitization was studied on 4 inch Schedule 80 stainless steel pipes. Two artificial cracks were prepared in the heat affected zone of three welded joints. One joint was subjected to IHSI and a sensitization treatment while the other two were not subjected to IHSI. The degree of sensitization was determined by micro and macro-etch tests, and creviced bent beam tests (IGSCC Test). No difference in level of sensitization among the three joints was observed.
- 1. Introduction It is well known that plastic deformation generally accelerates sensiti-zation produced during subsequent heat treatment and makes stainless steel more susceptible to IGSCC. IHSI, which is a remedy for IGSCC, is a method to introduce a compressive stress caused by plastic strain on the inside pipe surface, c.nd therefore produces concern that it might make stainless steel more susceptible to IGSCC. The effect of plastic strain induced by IHSI on the subsequent low temperature sensitization characteristics was studied for pipes with and without small artificial cracks.
Corrosion tests and hardness tests were performed so as to examine the above described superimposed effect. ) I
- 2. Experimental 2.1 Material l
Type 304 stainless steel with 0.07 percent carbon content was used in this experiment to provide the worst material condition. The chemical composition and mechanical properties are given in Table 1. 2
2.2 Preparation of test coupons Three test coupons were prepared for this experiment. All these coupons were 4 inch Sched. ale 80 pipe of nominal size and 300 milimeters in.. length. They. were butt-welded by shielded metal arc welding following automatic gas tungusten arc welding (GTAW + SMAW) . High weld heat input (20 KJ/cm) was selected to provide a hecvily sensitized condition. Four notches were produced by electric discharge machining at a point which was nearly 3.5 milimeters away from the weld fusion line on the inside surface of each test coupon.- The notch size was 2 mm in depth and 8 mm in width as shown in Figure 1; this size defect is expected to be easily detected by the usual ultrasonic techniques. 2.3 Heat treatment and IH3I The sequence of the heat treatment and IHSI is shown in Table 2. The first heat treatment simulated the in-service thermal condition before IHSI. The second heat treatment was intended to investigate whether the plastic deformation produced by IHSI accelerates the subsequent low temperature sensiti-zation or not. The low temperature sensitization heat treatment was conducted for 72 hours at 450 C. This corresponds to about 30 years at 288 C assuming
- i. an activation energy of 46 kcal/mol. This IHSI condition is the usual one as shown in Table 3, 2.4 Specimen removal and tests The specimens were removed as shown in Figure 2. In order to survey the superimposed effect of plastic deformation and sensitizing heat treatment, the following four kinds of tests were conducted; (1) Micro-Vickers hardness test o
(2) Chromic acid etch test i 3
(3) Copper sulfate-sulfuric acid etch test (4) Creviced Bent Beam test
- 3. Results The result of chromic acid macro etch tests is shown in Figure 3. There is no significant difference among the three pieces as concern the sensitization zone as shown by the schematic hatching zone in Figure 3 (d) . Figures 4, 5, and 6 show the results of the chromic acid micro etch tests on 8 points neum the notch as illustrated in Figure 7. Figures *4, 5 and 6 correspond to the structure of coupon 1 (AW + lst LTS), coupon 2 (AW + lst LTS + 2nd LTS) and coupon 3 (AW + lst LTS + IHSI + 2nd LTS) , respectively. There was little i difference in degree of sensitization among each examination.
4 The result of the copper sulfate-sufuric acid etch tests is shown in Table 4 and Figure 8. Although a slight difference in extent of sensitization was found among the test pipes, the apparent effect of plastic deformation or low temperature sensitization was not observed. Micro-vickers hardness distribution is shown in Figure 9. Open circles show the hardness number measured along the line 2.5 mm away from the pipe inside surface without a notch, and solid circles show the one with a notch as shown in Figure 10. Comparing coupon 2 (without IHSI) with coupon 3 (with I!!SI) , the hardness near the fusion line of coupon 3 was found to be a little higher than for coupon 2. The hardness increased rapidly near the tip of the i notch. The result of the creviced bent beam test is shown in Table 5 and Figure 11. The difference in history for each coupon does not correspond to the variation of the maximum crack depth presented in Table 5. The crack i I 4 4
distribution was practically the same for each specimen. A typical crack depth distribution (coupon No. 3) for a test piece with cracks is presented in Figure 11. No deep cracking (SCC) was observed near the notch.
- 4. Discussion Plastic strain followed by heating is generally known to accelerate sensitization. As IHSI inducca plastic strain through thermal strain, stain-less steel pipe may become heavily sensitized as a result of subsequent LTS.
In this investigation there was, however, no remarkable difference among each test pipe in condition (1), (2) and (3) . This reveals that the additional treatments following welding plus the 1st LTS have little effect on sensitization, that is, IHSI and the 2nd LTS did not change the sensitization state. This may be the result of the fact that thermo-mechanical effects of welding and the 1st LTS were dominant for seeding and subsequent sensitization of the test pipes. It was observed that hardness near the weld metal increased for the test pipe with IHSI compared to the other. This result showed that IHSI could effectively produce plastic deformation on the pipe. Figure 9 shows that the hardness af ter IHSI increased rapidly at the notch tip. This observation corresponds well with the calculated strain concentration phenomena presented in Figure 12. "This calculation was conducted on a notched 12 inches schedule 100 pipe using a finite element method (IECT-II; an IHI computer code) . The maximum strain at the notch tip is approximately 3 percent. It has been reported in reference (2) that a plastic strain of 64 significantly enhances sensitization. It should be expected, therefore, that severe IGSCC at the notch tip would be observed in the CBB test. No crack, however, was observed in this region and the maximum IGSCC was observed at a distance 8 mm from the notch 5
l as shown in Figure 11. The following can be considered as the reason for this i discrepancy: The plastic strain in the weld HAZ was calculated by FEM to be inore than 4 percent. On the other hand the additional strain at the notch tip produced by IHSI was considered to be smaller than 3 percent because the artificial crack had a finite width (= 0.3 mm) while the calculation was based upon zero width.
- 5. Conclusion-The above discussion leads to the following conclusions.
(1) Neither IHSI nor the 2nd LTS were effective in accelerating sensitization for stainless steel pipe, namely, the welding and 1st LTS were dominant. (2) The plastic strain at the notch tip after IHSI would be less than 3 percent and had little additional effect on the sensitization produced by welding and subsequent LTS.
References:
(1) E. Friedman: ASME Paper 75-PVP-27 (2) Y. Ando et al, Localized Structural Behavior and Safety Evaluation of Nuclear Power Plant (No. 3), Japan Welding l Engineering Society, Nov. 1978, pp. 38-40. l l 6
TABLE 1 CllEMICAL COMPOSITION AND MECllANICAL PROPERTIES C Si Mn P S Ni Cr YIELD TENSILE ELONGATION 4 POINT STRENGTH max. max. max. max. max. 8.00 18.00 21 53 35 0.08 1.00 2.00 0.040 0.030 - 20.00 (kg/mm2)
- 11.00 (kg/mm2 ) (x) 0.07 0.48 1.52 0.027 0.001 9.65 18.65 30 60- 64 1
1 4 i .
L . r lll; ~ S S T T L L -. d d n n 2 2
+ +
I . S l , i Y l . R O + T Y S S S S R I T T T O H L L L . T S t t t - l i s a s l l l l - L A
+ + +
M D D D R E E E E D D D H L L L T E E E . W W W S S S A A A . 2 E L l l . . A T ~. O N N 2' O 1 ) l' U O C ~ ~ o c 4l ,I! i;;i!!' 4 ; ,, ii : 4 :!] a11i ;I- d ' .
i f TABLE 3 1HS1 CONDITION i FREQUENCY POWER HEATING FLOW RATE TEMPERATURE (*C) ! e DURATION (kitz) (kW) (sec.) (m/sec.) OUTSIDE INSIDE
- l DIFFERENCE 3 150 35 2.5 513 130 383 l
i 1 2 i i
.s .- _ - _ - - _ _ _ - _ _ . . -
4 J l l TABLE 4 SENSITIZATION REGION.(STRAUSS ETCH)
. FRO.M FUSION W E COUPON NO.
r A B A-B i 1 5.0 11.0 6.0 2 3.5. 10.5 7.0 3 60. 11.5 5.5 i 12 3 i i [ }& \ , 1 A
^
N B 1 , NLHBER IN A30VE FIGURE CORRES?0NDS TO NLH3ER IN FIGURE 3. 1 A: DISTANCE TO STRAUSS ETCH 3EGINNING FRO 11 FUSION LINE B: DISTANCE TO TF.DlINAL ?RCM l l l 4 i i e } 10
. - . -- - . - . . -, -. - ~ , - - . - . . . . - - - .. . - , _ . . . - . . . . . . .
TABLE 5 MAXIMUM CRACK DEPTH (CBB TEST) COUPON NO. MAXIMUM CRACK DEPTH NOTCH (x10-6,) YES 83 1 YES 1039 NO 991 No 1589 YES 1097 YES - 2 NO - NO 982 YES 153 YES* 999 3 No 1800 NO 1779
*: CRACK DEPTH DISTRIBUTION IS SH0'4N IN FIGURE 11 11
ev
- i, 8 =
SLIT WIDTH - O.3 FIGURg I UOTCH CCNyyGURATION 12 , l
0 315 45 E i A 4
'T 270 - -
] -
90
. A D B
C A C D A 225 135 1 180 1 POSITION TEST ITEM A CBB B STRAUSS ETCH C HARDNESS D MACRO E MICRO FIGURE 2 SPECIMEN REMOVAL i 13 y- .- -a. -. y
f
/.
e 4. . ..w Aab - . wws .u. 3 A (a) AS WELDED + lst LTS i i c.v fm P s p u- __.L,e . (b) AS WELDED + lst LTS + 2rd LTS g t .
*r d
y* j ( 5;' ' ? .I % i _.,,;; : ; 5 ,, .
- i' '
. ,: x wnaan: 1: ; s. s xaun?ss x[: ,
, (c) AS WELD ED + l s t LTS + !!!SI + 21i LDS l 1
\_ t, ,
1 4
~
l t SENSITI ZATION ZONE _ .
'i i
(d) SCHFMATIC FICURE
^\
FIGURE 3 CHROMIC ACID MACR 0 ETC I TEST RESULTS s
\
14 ' s l 4L , j
'\
I mm from inside .' mm from inside
\ \ l \ 'W % ,. fs /( ; ')....,.m : ,..
t y \- ~ . .%. ' ,
//4' Y /(
)'f) (/.g,/ >.-- -{\
. [
u - ,
). "-
\, ,-
i N, , l ,s
~ -
s) i s .
, , i .
%' s\ . \ y .'
~
' f_'/,'.'7/ ,,,s r'
y
. . ' \ \
> ,e l W (, C Y;7 3;% ,
f,.gf.1 ,. ' f. , j .. $, . 3 N. .\ </- t_ p d .;,. , ), ,nf
!(
' I
- 3
, . ~ . , ., -
e
~
,>., f ,. ,_ .N a,,
\
,, V ;- N, .. g
..s.....
;D . r_. ,-
~
f^/ '~
/./ ,;.,
' . v; / g. -
~
/_r./ r-s7 _/,,>
m.q m
/
Q. ,i c u\ L./
- s. .. .
m..
~
y , .e/ y- ,
.. I s /,k!( Nr V~~
i.
. 1 .o 1 I
( 3 s. ,
<(g
~
1
. .u- y ;, v. ... C _ cc. ;v..>
ss ,
' /
w l l s - . .
, 4 .,
(( N ? -, e [
.:r' .
4, ,
.s s- .
9 y y,<
- p; -
FIGifRE 4 CHROMIC ACID MICR0 ETCH TEST RESULTS 15
.# e , -, .-
N\ 1 1 mm from inside 2 mm irom insidi l 'N N -. -...-_...a
-_. .f
)
* \ /
~ ,
e s -
,s . \/' ,/[,
. s i** s'9
'j
.,'e\
./ .g \%
- we
, ,/
) - /
.I s ) ,
I '~ % ( j \\. 7/ < ~~. .
/ /
l , s , ,
,, <w',r , s
" i i
/ N, } .
s s (' %-- >
}4 -)
,r % . [
[ 4 .,, ' .5 s gs - m e ( r y' '
\ 3- N. i 2
( ', ' ',b ".j [ (
- yw
. y. . '3 ,
% 5. i .
/ ]'
,y \p / 'g . 4 t' s g
( . xt
. . . . . . . . , . . . _ ~ .....-.......,....-.-..-.4
- p =**.-==. 4 ,
v'.
, I , , .
- e ,
s , ,
, , 3.- . .; t -
/
! eg 1
. .., ~- /K . l ',t '
'l 3 r,/[7 , / ,
'/ ', NNi , . J' /
u / - y;- ~ "- ]
.g.
, -3
.. f- I.h N '.
lN r.
. s<
nc ! . --<l f*' f N,,i y '-p - \ }' N '\ f' h
^
,.~Q ~ L, x myg
,x) -
pad _r
' 1; ~ -
6
$n4.;
/j- x..> U '
- q q~.; '
/
/'/ [t <
7 m, \ FIGURE 5 CHROMIC ACID MICR0 ETCII TEST RESULTS 16 f
N N L l 1 mm from Inside 2 mm from inside
-s -_ . - - - - - -
\
,,s.
x :gf . ~
. N K.; -
7.-
'O
[ _ 1 .
\x. , , , .f t
' I i ~ 4 i
}' * .-
, j.. 7 t y- . ,
- v. ,
=; / )L
,. > . ; .. s ..
o .
- n. i
" ' ~
'+ \
i, . . 4-
*r.
lst '. ( s '. .
'\ s # %.- N *
/, ~ ;n, ,
.y.;. . ). b. . u .. l.%....
,...~
.\
g
.(. / 1. J, . 'i 4 ;
.fu l
.. N.. i.
. . ,. i e ,. _
s.
.s- y ,
2 ,, , . , - .
. s s g
' . ',l ' ','2 P
( \, ,
, g. ,
) . .f 3.
. ,. t. _ . . ' ___
~
~,
- - . ,M -
- ~
i f '
)
/ -J g.4[kgQ -e
...,7-(^'\.,
% e
/
) ,- .
3 J , J
.s
/--s
/' '
w g[/ .
.w )- y,\ ~/f '
7 3 N/j ' s.' g w8 s 3 g r , > p / i.) ,
,"' . . , ;3 .
'$c f
.)
4
,t /. . % .,5' . ,.; J .
T._/'/},.....,p/g Y,/(( i N . ' [I
- + '
./. , ,.
- r s .~ .
.- 7'.
- t '
4 l %,'.- . 2
. "-r~ .$ m\' ,, / - 4
\ r* /' .
i
\' 3 -'f"%.Ip ' >'
]
s-\ ' f
, ,s -
L
&< .f9 ~~ ' Q ~
f.'
l l 1 .-
U % V ' FIGURE 6 CHROMIC ACID MICR0 ETCH TEST RESUI.TS 17
l l l l l l
# N s i s m " N 3 _3_ _3_
FIGURE 7 MICRO INVESTIGATED POINTS 18 l ~ 1
I 2 3 y- . y- s
,p . 3 , ,
. c . . '/ .s: .. .
u '.
/% Y
/
AS WELDED ..; . ,. ri
- 71-
+ .
Wy ' _.. . ; -7, . c-[.i_ 3 ._ '._.. Bs. < 1_.d M , s L'. s g. -I ,? ' w' 1st LTS - t -
. '. . - s i
-' .,.y .
-% \ ,,. _
;'. vJ ,
-; s ...,,~ ,.
J. Q.>. ,
~. ..
- L.
2 (. , -
,.._....w 1
,L _
._ .s a c ~ r. . .
~ '. . . .
= -
w AS WELDED < . . ' '....
. - .. M. .
- p- ,
\.' -
( . . x' . - - - , , g s -/
, ) ,),.s
+ ,
,- s ~
) ,
.~^
1st LTS . 7 7' . ,,'. . e-.,, s 3, ~'./ [
/
t ,(f , , , ' -
. j e **g . -
. p..
. ' .e
;. , j. __ . c a ;.,
. (, .: 1
+ -
.?
2Nd LTS ,,' -
.- . i
- q. , " 7 g ', '- . .- + 4 - -
. . e< , ,
. ..,. m'~ .
../ ' '
A... /. u y . ' /. . " . . . AS WELDED .- \,
.. 2. y i. , . .. .-{ .
~
.~. .
,' ,)j ,. )., , . , - -- r. t ,;
a:
.. ~.
+
m: . N
'- "I.
1st LTS - J t . . . . .
... +,. - J. , y,, . -: /3. , 7- ,,,
+ y - -
IHSI 1,, .. . %. K . . {'. ..
~.~ - ' ,~ f ;
-. .. L
+
.\
. /, ! ' l
'ty '
e 2Nd LTS 4r . FIGURE 8 SURFACE INVESTIGATION FOR STRAUSS ETCHING
N - H _ V 5 M 1 I H S
!M I
I R S E T - H I I F ) S A m O H ( m T I I W S A E H T L E U I U F N D O ,W
. I H H R _ L E T T A G I I E 0 N N W WN 1 O A I H oO?
S U C F N fW ' M O R F O I T U B I
- s E C
N A N O I T R T S I T I D S S 4 ,, I O S D P S E
^ 5 W N A D L R I
-[
I ' 1 F A H 9 g i F N C 0 0 0 0 0 2 1 3x n S zm>:lE l ! ,.
-f
)
2000 DATA OF CRACKS OVER 50pm . 1000 l l l WITH IHSI I -- -- WITHOUT IHSI
- 500 l I i W I I l S I l E 200 gl l 2 I! '
4 g iI i g 100 1! u ll 5 II U I 50 llI l l-I
- I I 1 I 1 I I I I I l 1 20 I I l-l l 1 1 I I I I i 10 0 5 10 15 20 25 DISTANCE FROM FUSION LINE (mm)
FIG. 10 SCC. DISTRIBUTION CHANGE DUE TO IHSI (CBB TEST) i I h i 21
6 0.030 ,, 0.025 -- 0.020 -- o Z H g 0.015 -- M E4 b a
$ 0.010 --
5 8 0.005 -- O i
-0.005 ' ' -
1 5 10 15 20 NOTCH DISTANCE FROM PIPE INSIDE (mm) FIGURE 11 MAXIMUM EQUIVALENT STRAIN DISTRIBUTION 7 BY IHSI(CALCULATED BY FEM) 12 INCHES SCHEDULE 100. NOTCH DEPTH =2.4 mm 22
A-8 SOME ANALYTICAL CASE STUDIES OF IHSI
SUMMARY
i l In order to verify the validity of IHSI for various cases, the effect of some geometrical transitions and also of welding residual stresses were analytically studied using a thermo elastic-plastic finite element method. All of the mechanical and physical properties inputted were temperature dependent. The analytical results reveal that the following geometrical transitions; (1) thickness transition produced by weld edge preparation, (2) large thictness transition such as nozzle to pipe, (3) transition in dia-meter such as reducer, and (4) pipe to end cap transition have no significant effect on residual stresses and sufficient IHSI effect is expected. Further simulated welding residual stress also have little effect on the final residual stress. l 2
TABLE 0,1 MECHANICAL AND PHYSICAL PROPERTIES TEMPERATURE THERMAL SPECIFIC THERMAL CONDUCTIVITY HEAT DIFFUSIVITY (*C) (Kcal/cm-s *C) (Kcal/kg *C) (cm2 /s)
-6 20 34.7x10 G.107 50 35.7 0.112 0.0390 100 37.2 0.117 0.0399 150 38.9 0.122 0.0409 200 40.5 0.125 0.0420 250 42.1 0.127 0.0432 300 43.7 0.130 0.0439 350 45.4 0.131 0.0447 400 47.0 0.132 0.0453 450 48.6 0.134 0.0462 500 50.3 0.135 0.0470 550 51.8 0.137 0.0478 4
600 53.5 0.139 0.0487 650 55.1 0.141 0.0497 700 56.7 0.143 0.0506 750 58.3 0.145 0.0514 800 59.9 0.148 0.0523 TEMPERATURE YOUNG'S MODULUS POISON'S RATIO DENSITY (*C) (Kg/cm2 ) (kg/cm2 ) 6 ~3 20 1.98x10 0.260 .8.03x10 50 1.97 0.264 8.02 100 1.95 0.270 8.00 150 1.91 0.274 7.97 200 1.88 0.278 7.95 250 1.84 0.281 7.93 300 1,79 0.284 7.90 350 1.76 0.288 7.88 400 1.72 0.292 7.86 450 1.67 0.296 7.83 500 1.63 0.300 7.81 550 1.58 0.304 7.79 600 1.53 0.308 7.77 650 1.47 0.314 7.74 700 1.41 0.318 7.72 750 1.36 0.320 7.70 800 1.31 0.324 7.67 i 3
1 l i TEMPERATURE THERMAL EXPANSION THERMAL EXPANSION (INSTANTANEOUS) (MEAN FROM R.T.) (*C) (1/*C) (1/*C)
-6 -6 20 16.4x10 16.4x10 50 16.7 16.6 100 17.1 16.8 150 17.5 17.1 200 17.9 17.3 250 18.3 17.4 300 18.6 17.6 350 19.0 17.8 400 19.3 18.0 450 19.8 18.2 500 20.2 18.4 550 20.5 18.6 600 20.9 18.7 650 21.3 18.9 700 21.6 19.0 4
750 22.0 19.1 800 22.4 19.2 l l 4
, c '
- =
0
.'1 C CCC C
0 000 0 _ 0 0 50 3 34 5 2 n = _ T T 5 7 0 S E
- I E T R
E P O R
. N P I
A L 5 R A
,'0 T C S I N
A H C E M l o
~
g i F 5
,2 O
h l 0 0 0 2 1 D pSNwCd mM$$ m l 4 <l- )
I l i
- 0. . Gene ral -
In this case study, the following conditions were assumed. a) Mechanical Properties i - S t re s s-S t rain Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fi g . 0.1
- Y o un g ' s Mo dulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tab l e 0.1
- Po is son ' s Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tab le 0.1 b) Physical Properties
- Th e rmal Con d uc t ivit y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tab le 0.1
- Specific Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T ab l e 0 . 1
! - The rmal Di f f us ivi ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T ab le 0.1 i - Heat Trans fer Coef ficient : hm AW Pr0*4 p= 0*8 hm = 0.023*
i 20.2 60.8 Aw : The rmal . conductivity of water d: Inside diameter of pipe v: Coe f ficient of kinematic viscosity Pr : Frandtl's number y= : Water flow rate c) Heating Density : W W=We 0 WO : Heating density on outer surface 1 j S: Heating depth (9.6 m=) I x: Distance from outer surface ! d) Soundary Condition As shown in Fig. 2 unless otherwise specif fied. e) Work Hardening i Isot ropic-wo rk hardening law 6
h f) Cosputer Program Temperature distribution. . . . .ITEMP II (Developed by IHI) Elastic plastic st ress. . . . . . . IEPCT II (Developed by IHI) ' analysis i b i9 F l 1 i 1 1 i I e I 6 i. i e I i i 7 4
. _ - - - . .. , . , . . - , - . . . - . . - - - - . _ . . , - . - - - , - - ..n-.._, .- - , , . ., , , - - . , - . . - - - - ,c,.- - ,- -- .. , , , - - - n -- -.
I CASE 1 THICKNESS TRANSITION PRODUCED BY WELD EDGE PREPARATION MODEL 12 B Sch.100 Thickness 28.15 to 19. 25 :::ra COIL LENGTH 250 mm HEATING DENSITY 0.0583 Kcal/mm3 h (AT.0D) HEATING Dl! RATION 180 Sec. RESULTS . - IMPROVEMENT EQUIVALENT TO SMOOTH PIPE IS EXPECTED. 8
R166.0 R137.85 PIPE Z 400 121 />' /"[// // f /> 126 f S/
/
me **d==* Insulated
</=
/
91 % O $ ' z 6 85 90 _ O o a c y " o e E y 79 > 84 a 5 Or b
$ 102 73 ' 78 a
i a 6, lll/k ,2
= '
61 lll #t 66 55 Illll'60 60 49 lllP 54 48 43 748 9 :
- 37 42 g 3 36 8 ~
3,, - o x 3 25 ' 30 E
- 5 19 r 24 3 -
13 f 18 t 7 r 12 1 3 / L I 6 0 g g -- R R 140.00 #f"##"""## R 159.25 ' Fig. 1.1 ANALYTICAL MODEL 9
4 1 O I T m I m _ S _ N 0 A g 0 m R 4 o T E rr _ P f e HS I t TS P en IE ce WN H nc K T a EC O tl PI O si IH M io PT S Dc
) }
D D D 0
= N O O I I 0 0 I d T 3 U B
I R T S E O e o I D E R U T u A R E P M 0 E G d T 2 2
- 1 g
i O F n
/
/
/
/
/
0 Q,
/
d / i /
/
/
( E /
/
P _ / I E / P _ P / I / n'v T O H A P H
~
/
/
/
/
O T h M S O O
/
/
/
M / y S /
/
/
/
/
- - ~ - -
l E 0 0 0 0 0 0 m 0 0 0 0 6 5 3 2 1 Uo~ ou30o$e
i 0-125 , I I I 30 -
-a-SS-k'O'c\ 9 (
_() o
% m W,g g i
h i N_ ) 12IN Sch 100 H B ege-e--a Q t Q 20 - ' I
- s ,4 AXIAL STRESS AT ID s s i i s
\ ' I s' rs t 8% -o HOOP STRESS AT ID . gk f / \
10 - Y-h # g 's k ---o--- AXIAL STRESS AT OD b \ 'a s ---o--- HOOP STRESS AT OD
-E s
--o i
__2 s e u.
- c. 0 i 4'
'M
- >' h
~-
- E-i
- i 100 \ ,,/* tr ' 200 300 400 (nm) u; o Distance from coil center 125 l
/ R N
~ ~
[ n S
--w N
d' s -
$ 5 AXIAL STRESS AT ID
-3 0 - O HOOP STRESS AT ID
--- G--- AXI AL STRESS AT OD g ---0--- HOOP STRESS AT OD
)
-4 0 -
,1-------------------------
Fig. 1.3 RESIDUAL STRESS DISTRIBUTION
l
.I l
CASE 2 LARGE. THICKNESS TRANSITION (N0ZZLE TO PIPE) 1 MODEL 1 24B Thickness 44 mm i COIL LENGTH. 400 mm HEATING DENSITY 0.038 Kcal/mm3 h (AT. 0.D) HEATING DURATION 240 Sec. COOLING WATER I TEMPERATURE 20*C FLOW RATE 2 m/sec. 1 RESULTS
- SIGNIFICANT EFFECT OF LARCE THICKNESS TRANSITION l AT COIL END WAS NOT OBSERVED.
t i e l 4 1 12 ; I
. . _ _ _ - - _ . ~ . . - , _ _ ---,__,--.=_,._..__m_. . . . - - , - . _ . . , ~ _ . , , - . ~ . , _ _ . , , , . . , . . . . _ ,
L I O C a G mMnM N I T 1 A 3 aeoe1c ,rNM E e H W") L
/
E 0inbtsh 0 1 I 0 /
, D 0 /
O 2 /
. 0s s (i :
r M 1 ii 0 3 A cg# L A C 0 4 0 0 /
/
fr4 2 .eV
, 0 I 5 .
e T . h Y li N L O i% A I _ N h$ A ,
, 3 r n i ,
p e T U D
!i I
R T p d' , nN %6 5 3 S I D E R
. - . n3 U
, ,, - w gr T A
R 0 - E 0 P
, . . s M 4 , n rv - L t
T E N T I O J 1 i . . ar- s t D 2 L E . W g 0 ,, r 0
% a- n i
F 5 0 0 ww'w xxNx 4 0 A 0 4 li030 iee- m 0 2 0 1
' , : s s 7
' ( ( lh o
S S s n ?
' n 0 sr . s 3 - r mn*y C
- . . . . . . - - . . - - . ... -~ . . _ . - ~ .
4 HEATING COIL i i
\
-e AXIAL STRESS AT ID
,o, fA, 30 - f ' o y y ,,'
S
- o. ,, _ . o-- -o- - : HOOP STRESS AT ID g I s
'g ---o--- AXIAL STRESS AT OD
# g
",o- i.-s 20 - e ,8 %,___,..e ..,,,,,e
'g ---o --- riOOP STRESS AT OD i
.. ll < 's E >>
's N\
d D 10 - \ l
.?
- # lS
- s o
's \
\ e' x n,
< . * ' p'/ \ e- ,
, ,hg ,
a , . ' N o,- -190
$ 100 200 300 4 Distance from Weld Center (mm)
WEm g j 4- -.
- e Pig. 2.2 RESIDUAL STRESS DISTRIBUTION
l
, CASE 3 TRANSITION IN DLWETER (REDUCER: 24B TO 12B)
MODEL REDUCER 24B to 128 (FIG. 3.1) COIL DIMENSION SEE FIC. 3.1 HEATING DENSITY 3 0.026 Kcal/mm h (AT 0.D.) HEATING DURATION 360 Sec. RESULTS COMPLICATES DISTRIBUTION OF RESIDUAL STRESS WAS OBSERVED AT BOTH IN A:iD OD. COMPRESSIVE RESIDUAL STRESSES WERE INDUCED ON ALL ID SURFACES UNDER COIL i 1 15
. - - - . - . . ~ .- - . - .. ..,- . - - - - - _ - - - _
161.5 I) 190 280 g 114 93 127 46 -; ilTir es-g __ E* A ,//// 0 6 5 m k E m V.//4) u a. u h
;g_g:
r 5: a: R e m a g n ~ 8x 3 u x v .- .
.. ' f ,; g u . . .
o . 5 WELD JOINT [
=
E -p7 Z DIRECTION RESTRAINED WATER TEMP. 20*C c= 0.72 m/sec. Fig. 3.1 ANALYTICAL MODEL l
I 100 g 300 500 500*C l ~\ i
- 400 i
m ik \ m\ t nNC. is Ntn f ENi@;__ t3 !! " .Q 300 r " LL._Jyv7m z _ $ '00 L
- 5 n
J L ""_ ?C f 3-Nt-1 E!qi_
> 2 -1 c.
i L_ a 1 = ' ' 100 w 4
\
47 ra sie 1 iW F ' Fig. 3.2 TEMPERATURE DISTRIBUTION
.m.. -_ .. . _ _ . . _ . - . . _ . . _ _ _ _ _ _ _ . . . . . - ..- _ .__._._____-.______.__m_._ __.. .m._ m_
l 6 30- f's , "*%
,' %, ,# #'~ s' ~
s
,/ r ,o s* s O, /
A % s, -
,o# .o-~.'mg
, o' 's es 20-f
/
/
/ /(\- o- .o, "
e s g
\
1
- s
/ g o ,/
- s w,
\
y h j D
& /
/
f
'-e-4% 's s 0%
x~ :
' w 3
. m -n 'y j' WELD Q ' --o y Ej 0 i i'
',. p- ! i i
@ -200 ' ' -* 100 - 0 100 - 20
) " Distance from weld center (mm)
~10-
. e e
e O w -e- AXIAL STRESS AT ID
- 0
- 3 0- No g -o- HOOP STRESS AT ID
--o-- AXIAL STRESS AT OD 1
--o-- HOOP STRESS AT OD Fig. 3.3 RESIDUAL STRESS DISTRIBUTION P
= -
r - =
CASE 4 PIPE TO END CAP MODEL 168 PIPE TO END CAP (SEE FIG. 4.1) COIL LENGTH 330 mm HEATING DENSITY 0.0515 Kcal/mm3h HEATING DURATION 150 Sec. RESULTS
- COMPLICATED DISTRIBUTION OF RESIDUAL STRESS WAS OBSERVED NEAR THE END CAP
- COMPRESSIVE RESIDUAL STRESSES WERE INDUCED ON ALL ID SURFACES UNDER THE COIL i
19
1 4 i 250 330
'=
( =\' HEATING COIL V//////////////////////////////////////A o p.r...._. . _ ___ _, .,n. , __ ;. ,, ,_;__ _~~ _ .w; . . -
,. r, N .
T.
,-.a , 3.,.g._ _ .- .
.- ,n y ,n- , ,-
-...-,,..,y=~e-n i,. c-
. = c - "rx.. ~=?:;.w -** m sy r L m d.. . m %, n ,.
e ~
~ ,--.. .-.- , . . . .
@ t= .. : - ':7r
-a
,4 . Y,, M.w.h. *;;s. $
n j e- , , < , .,,. . . j .- Z DIRECTION "N m
- . ,f y <
/ ,/"*)- - '+a 4
. - r;* <
RESTRAINED ,
~ WELD JOINT ,7' * ' zN s,v , , ' - ## '
' f.
,E "7 y "} r
,._ m m ~~- o
'~
e4 r~ WATER TEMP.-20*C . <\, " - m
* ,/r p # U. m
'W.3m/Sec.
\g i 36 1 ~~
1 1
~
F- e. e a r --
. g321.6*b,h)k 8 V ,- - > ;r '. -
J n; ,
, v.
.itt.w
+
r 7 p ,>
-d
* ,'T c'
/ -e* bl1t y 'l I{ .' l,4' ,. f
~Q Q "_ j ,' G, "p'
.I y .5 . ,t '. N 1///////////
?., R DIRECTro.N,e
/4
/' r* . ..
a
,- s ~. ,.
RESTRAINED _ ,- .Jff. . + -
^
, -y %.J ,, f, # _ 134
. l ' I,,y.
[ ~~~r=" " Y , } q. w m y, ,- - . ,- : , x~ . . ..
'an Q,
- / s, s,s f-s "Y.
. ,/ . -
,_ r '/ -
f- -- 9 , Fig. 4.1, ANALYTICA!ri9'QDEL , ~ f,s. > n r.+ G ,,p-
,.~ ~- p p'.- . s /
~,
~
/ *~
/"e .,,,, / f, ,
,)
wj f.' ,,d*. "/ s e
+
.% . .v
,b
,, < , . s s ns. ,. .
.n ,
g l'a %~ -# ., / e
> , ; r-# , *s, j$~7'.
/p
<, .g-
,K sk <y ~. -
t 4 -
, - v q. a, .
4 Y + ='- .- . * [ "- [>
.. , .c ,.
~
r* . w ,, t
,' 'e /
i,, e . . 9
-( < *$
? u* fn 7F r p- y . -
* * - ,p;f l, y^. f ,}40 _
' r . ,f
; c
./ *i- ., ,- * , r's
- f. . _
.._,..,.~..,-.
.-.g-. . - - . . . ,
. , - - - . - . ~ .
~
,. ~ - . .~ -
,; . y
, r v , - g,-
, . . . , - r J
,_ r ;.
A - '
.. y %.. q (. ,
4 .-,,},
,*- s 'T ~ r_ i 't' t ,_~ ~,)
s ,.. g / / n
.,j .
- .s y
y 4 s.
+ -
pf.- .
+ ' ~
r
.p
,, , / .s
! r~ -< > , ,
< f. --
1*
~
g
.?
} '/ j .. ' ,~~- "
~ ' '
[ ,
, , j
(' dr '
~
~. / , ,
~~..
) .~ s
~
. s .
WELD JOINT ~ s 500*C n ,u 400 y
, .a 3
g wa 4n t.,. . . J ".i; '-~ =Ti-- w I " %- 500*C _- ~ s. . . ~~ ! 300 n _. --- --- n .,
.a-n - -
. o
... _.I. uu __ -u.
I u_ <
-J - w 00 200
____.y _u. ..p w .. ; u _ .._ , ': ' _ 14, ' 300
- _. _ yL__ _ ga. _g . rr . _u ,___; ; j t_ __ .y 4 ..,~q _ ;_ %
~
?2 2- -
200 sc' _ o 8
~\
I s,d, 00 m
- @\,., t 4 t
, 330 I._ U/////H//////H///> 3 h i b[1 Eh 46 _ _ _h I Fig. 4.2 THERMAL DISTRIBUTION _ , k +
- - _ - . . . ..-. . _ -. - _ - - . .. - ._ =
40 - HEATING COIL I I 30 - - o . AXIAL STRESS AT ID STRAIGHT PART - - R ., o HOOP STRESS AT ID l 0- ,e, .e,\ y1,o. j og --.e--- AXI AL STRESS AT OD
- I 5 E I 0/ % *' **re~e.S% I i
---o--- HOOP STRESS AT OD
$ 10 - ,,'
& A' e WEID g $
M n h ~~. S *
~ 30'0' 550 [,'_q,,DistancefrmWeldCenter
-o' *- ' -h /
- 00 4'00 4 m N
- g ,- -=3 (mm) y ~-o' e
o M-o Fig. 4.2 RESIDUAL STRESS DISTRIBUTION 4
1 l
-CASE 5 EFFECT OF INITIAL RESIDUAL STRESS MODEL 12B PIPE (318.5"O.D, 21.4 " THICK)
ANALYTICAL SEQUENCE I
- STEP 1 GENERATING INITIAL STRESS BY SIMULATING A SINGLE PASS WELD
- STEP 2 IHSI TREATMENT INITIAL RESIDUAL STRESS TEMPERATURE HISTORY FIG. 5.1
- RESIDUAL STRESS FIG. 5.2 RESULTS INITIAL RESIDUAL STRESSES HAVE NO SIGNIFICANT EFFECT ON THE RESIDUAL STRESS LEVEL AFTER IHSI 23 l
125 l l HEATING COIL WELD A ~P g l1 1200 - INITIAL CONDITION OF p hNN 5 STRESS CALCt:LATION N I 3
< E
.. \ ,
9 1000 , B 2
,o %
l l
$ 800 - 0 E* A OUTER SURFACE AT COIL E u o B INNER SURFACE AT COIL g 600 -
y POINT A 400 - 200~
/
POINT B wu f%. . 0 , , , , , , J,' , , , i 30 60 90 120 150 180 210 3600 3750 TIME 3900(Sec) Fig. 5.1 THERMAL CYCLE OF WELDING AND IHSI
- 40 -
O AXIAL STRESS AT ID i 30 - 0 HOOP STRESS AT ID so o\ l \
\
----*---- AXI AL STRESS AT OD j
$ ----0---- HOOP STRESS AT OD 20 - g i
! 8
^
c I 7 E "g 10 - \, N u dl *s w f 4' .
$ $ pl e9'D ,
O ^# ' #y * *- - - " [ ,l\ l
/ 00
[' ' 2d0 ~ 300 4d H
- I f
9 DISTANCE FROM WELD C,
/
-2 0 -
-4 0 -
Fig. 5.2 RESIDUAL STRESS DISTRIBUTION AFTER WELDING i l
. 40-j A) : AXIAL STRESS AT ID
,,Ad I
30 - I O HOOP STRESS AT ID s I A o ,#' '0-o--o - - - - - * - - - - AXIAL STRESS AT OD (W p'.A e
\
-----o----~ HOOP STRESS AT OD 20 -*g ; s, \,
5 g \ b'e s ss i
~
9 '
.4' - uN s
l 10 - e_.*- g '.N
$ \ N o
d \o
\
's N m 'o u m 0 \ ' o'- - b"'-- =
$ lb0 0' 2b0e 3b0 4d0
$ ("I j
- DISTANCE FROM WELD g j .
-2 0 - m_ g f' = -
.7
-3 0 -
I
-4 0 -
Fig. 5.3 RESIDUAL STRESS DISTRIBUTION AFTER IHSI
I i l l A-9 ANALYTICAL STUDIES OF SOME IHSI PIRAMETERS
- Temperature Difference
- Coil Width
- Current Frequency
- 1. Introduction This appendix presents analytical studies of three parameters, temperature difference, heating depth and coil width for IHSI. In Chapter 5, evaluation of these parameters and brief explanations are provided. In this appendix, the analytical results are explained in detail.
- 2. Temperature Difference In Chapter 5, the required temperature difference is calculated using an elastic perfectly plastic model. Some analyses were performed to determine the amount of temperature difference which was required to improve the residual stresses.
2.1 Pipe Model The pipe model was a 20 in. Sc'iedule 100 pipe. Two coil widths were selected. One was 250 mm (2. 8 /Rt) , the other was infinity. 250 (CASE A)
, = (CASE B) _.
! 9 i i N
I HEATING COIL I PIPE T ( l ! f.
;- ?>
l Coil length of 250 mm represents the lower limit of the coil length. On the other hand, a coil length of infinite length was selected as the ideal Case. 2
2.2 Analytical model The analytical models are shown in Fig. 2.1 and Fig. 2.2. Fig. 2.1 represents CASE A and Fig. 2.2 represents CASE B. 2.3. Analytical condition. (a) Cooling condition Cooling water flows inside pipe. Water temperature is 20 C and water velocity is 0.72m/sec. The hatched regions of Fig. 2.1 and Fig. 2.2 show insulating boundaries. The heat transfer coefficient for the pipe inside was determined as 1500 kcal/m h C. Calculations are shown below. Pr = 7.11 Re = = y h = 3.16 x 10 8 Nu = 0.023 Re Pr= 0.023 x (3.16 x 10 ) x 7.11 "' = 1265 ha a 511 d 0.443 = 1459 kcal/m'h*C Nu : Nusselt number Re : Reynolds number Pr : Prandle number Um : Average water velocity d : Pipe inside diameter v : Coefficient of kinematic viscosity A : Thermal conductivity hm : Average heat transfer coefficient 3
(b) Heating condition As shown in Fig. 2.1, the elements located beneath the heating coil were induction heated. The heating density distribution in the thickness direction is shown below. 4
~*
w = wo e (kcal/am' sec) s= 2 fpf = 9.6 (mm) s : Heating depth p : Specific resistance u : Specific magnetic permeability f ; Frequency x : Depth from outside diameter wo : Heating density at outside surface Wo was selected so that the tempec '-ure difference between the outside and inside diameter would be in the range 130 C 4 400 C. (c) Boundary condition The calculation is performed for one half of the length because the analytical model is symmetrical. Nodal points 1 N 6 in Fig. 2.1 and Fg. 2.2 are restricted in the axial direction. Nodal points 121 N 126 in Fig. 2.1 and 13 % 18 in Fig. 2.2 are tied in the axial direction. (d) Physical properties The temperature dependency of physical properties from 20 C % 600 C were used.
- 1) The thermal expansion coefficient, Young's modulus and Poisson's ratio are shown in Table 2.1 and Table 2.2.
- 2) The temperature dependency of yield point is shown in Fig. 2.3.
4
(3) Calculating condition The calculating conditions are summarized in Table 2.3. 2.4 Calculated results i Calculated results are shown in Table 2.4 and these results are summarized graphically in Fig. 2.4 and Fig. 2.5. From Fig. 2.4 for short coil case and Fig. 2.5 for infinite coil case,. the following items are considered. (1) While residual stress for the short coil case saturates at about 350 C, that for the intinite coil case saturates at about 300 C. (2) Saturated residual stress for the infinite coil case is lower than that for short coil case. (3) The temperature difference calculated by elastic perfectly plastic theory does not correspond to the saturated residual stress but gives satisfactory residual stress results. Some typical results for CASE Al are attached. Fig. 2.6 is the temperature transition curve. Fig. 2.7 is the temperature distribution in the thickness direction. Fig. 2.8 is the temperature distribution in the axial direction. Fig. 2.9 is the residual stress distribution in the thickness direction. Fig. 2.10 is the residual stress distribution in the axial direction. 2.5 Conclusion Whether the coil width is of full length or not, the residual stress becomes satisfactorily compressive. Simply calculating temperature difference corresponding to perfect elastic perfectly plastic theory gives satisfactory large compressive stress results. 5
- 3. Coil Width In Chapter 5 the effects of coil width are graphically illustrated using the data obtained experimentally and analytically.
If the coil width is not of full length, the coil end effects due to the bending moment which is produced by temperature difference in the axial direction are significant so that the residual stress benefit will be reduced. Some elastic-plastic calculations were performed to determine how much width would be required for the coil. 3.1 Objects for analysis The objects for analysis were 12 inch Schedule 100 pipe and 20 inch Schedule 100 pipe. 3.2 Analytical model The analytical model is shown in Fig. 3.1 3.3 Analytical condition The analytical condition is the same as presented in article 2 of this appendix. 3.4 Calculated results The calculated results are summarized in Table 3.1 and Fig. 3.2. These results are the same as in Chapter 5 in the text. From Fig. 3.2, the following is observed. Residual stresses saturate at about 3/Rt coil width. Iloop stress is lower than axial stress but is equal to dxial stress at about S/Rt coil width. The residual stresses are shown in Fig. 3.3 43.11. Some typical results are shown in Fig. 3.12 N3.15 for case No. 5. Fig. 3.12 shows the time-temperature diagram. Fig. 3.13 shows the temperature distribution in the axial direction at maximum temperature. 6
Fig. 3.14 shows the temperature distribution'in the thickness direction at maximum temperature. Fig. 3.15 shows the residual stress distribution in the thickness direction. 3.5 Conclusion As shown in Fig. 3.2, the residual stress becomes saturated and sufficiently compressive when the coil width is about 3/Rt or greater. Therefore, a coil width of 3/Rt is enough to improve the residual stress.
- 4. Current Frequency Current frequency effects on the heating depth, namely the temperature distribution will depend on current frequency. Heating depth (s) is given by the following equation.
*" (ca) 2x \ uf 10-s I
u : Specific magnetic permiability f : Frequency (cycle /s) 0 : Specific resistance (Q.cm) Some elastic-plastic and elastic calculations were performed to eva-luate current frequency, 4.1 Elastic-plastic calculation 4.1.1 Object of analysis Object of analysis was 28 inch Schedule 100 pipe. 250 1 Eo 4
! /
1 " a: O e. The modelled coil width was 500 mm. 500 mm corresponds to about 4.3/Rt. 7
4.1.2 Analytical condition Three current frequencies were considered. Analytical conditions are summarized in Table 4.1. Heating power P is calculated by the following equations. The heating power was selected so that the temperature difference between the outside surface and inside surface is the same value when the heating depth is different. For a current frequency of 0.3 and 3 kHz P= Wo e - x/s dV volume For a current frequency of a kHz P= [ i Q dS
" surface Where wo: Heating density at the outside surface Q: Heat flux at the outside surface 4.1.3 Calculated results The calculated results are shown in Fig. 4.1 & Fig. 4.2 Fig. 4.1 shows the temperature distribution in the thickness direction.
Fig. 4.2 shows the residual distribution in the thickness direction for different heating depths. 4.3 Conclusion The calculation using an elastic-plastic analysis shows that the residual stresses are not affected by the current frequency. The reason why the current frequency gives no effect is that when sufficient temperature difference exists, the residual stress will saturate and the current frequency barely produces an effect. 8
TABLE 2.1 TEMPERATURE THERMAL EXPANSION THERMAL EXPANSION (INSTANTANEOUS) (MEAN FROM R.T.) (*C) (1/*C) (1/*C) 20 16.4x10 ' 16.4x10 8 50 - 16.7 16.6 100 17.1 16.8 150 17.5 17.1 200 17.9 17.3 250 18.3 17.4 300 18.6 17.6 350 19.0 17.8 400 19.3 18.0 450 19.8 18.2 500 20.2 18.4 550 20.5 18.6 600 20.9 18.7 650 21.3 18.9 700 21.6 19.0 750 22.0 19.1 800 22.4 19.2 9
1 TABLE 2.2 TEMPERATURE TilERMAL SPECIFIC THERMAL CONDUCTIVITY H1'.T DIFFUSIVITY (*C) (Kcal/ca-s *C) (Kcal/kg *C) (cm2/s) 20 34.7x10 ' O.107 50 35.7 0.112 0.0390 100 37.2 0.117 0.0399 150 38.9 0.122 0.0409 200 40.5 0.125 0.0420 250 42.1 0.127 0.0432 300 43.7 0.130 0.0439 350 45.4 0.131 0.0447 400 47.0 0.132 0.0453 450 48.6 0.134 0.0462 500 50.3 0.135 0.0470 550 51.8 0.137 0.0478 600 53.5 0.139 0.0487 650 55.1 0.141 0.049 7 700 56.7 0.143 0.0506 750 58.3 0.145 0.0514 C00 59.9 0.148 0.0523 TEMPERATURE YOUl!G'S HODULUS POISSON'S RATIO DENSITY (*C) (kg/cm2) (kg/cm2) 20 1.98x10 ' O.260 8.03x10 3 50 1.97 0.264 8.02 100 1.95 0.270 8.00 150 1.91 0.274 7.97 200 1.08 0.278 7.95 250 1.04 0.281 7.93 300 1.79 0.284 7.90 350 2.76 0.288 7.E8 400 1.72 0.292 7.86 450 1.67 0.296 7.83 500 1.G: 0.300 7.81 550 50 0.304 7.79 600 1.23 0.308 7.77 650 _.(7 0.314 7.74 700 1. l.1 0.318 7.72 750 1.25 0.320 7.70 000 , 1.21 0.324 7.67 l 10
TABLE 2.3 CALCULATING CONDITION (PROGRAM 1 TEMP II I EPTC II) ,
.cA 3E
$N/ek #
Mocel- TExfseeruRE DEMS/TT coEF7/CIENT . w/DTH c c3 (Ved u k Gedia Ac) c~~ ) A/ F42 1 20' O. 0 ss /500 i22 i A2 i O.0452 _ i A3 0.0327, I A4 d.0238 !
^
- A5 00//7 8/ F4 2. 2 0.0 53 _ _ _ _ _
B2 s.s4t2' , 83 0. 0 3 t 7 84 d. 6238 gg , e g ,f f i 11
l TABLE 2.4 CALCULATED RESULTS TEMPERATUR' E D//TERENCE MS/ DUAL SfM3S CA3E CETWEEN ouT3tDE AMO AT
/MS/DE 3dMACES jp3lpE syggCE
( *C ) Al 3& & 7/3 -233 A2 3!3 904 -22.8 A3 298.874- --20 9 A4 /83.3r? -/4.2 A5 9 8. ted --
- 3. s k 8I 367.267 - 2 8.o 82 3/4.346 -27.2 83 2 k 7. 2 / / -2 A.L B4 /83.578 -2 4. b 83 98.2/4- 6. / 4 12
1 1 125mm 275mm 1 , HEATING COIL
/ .
v/ //// / / / / / /// // / / / // / ///1 OUTSIDE SURFACE 12 6 R c - S1 't*fB5$hF 6 'n'S'rT b h % ns" ' si ' n#' ier t et_ its
=
gri h INSIDE SURFACE R Z_ Fig. 2.1 ANALYTICAL MODEL (CASE AlN A5)
H m EE i
%$ om $%4 m i
7 1 3 i 6 2 8 1 4 F i g 2 7 2 2 A 3 3 1 N 5 A L Y T I 3 8 3 C 2 A . L 5 M m O m D E
- t 1 L o 6
( C A S 4 9 E B l B _R 5 s 1 1
)
Z 1 7 o 1 5 0 1 e 1 1 2 8 O$M am $%N a
$Ns %gt %NQt2
[ H E A T I N G C _ O _ I L
- Japan Society of Mechanical Engineer 30 _ e JSME DATA A INTEPOLATED FROM JSME DATA 20 -
aqp=0.03 2' - R D - .3' _ ep=0.01 @ o p=0.006 N - cp = 0 . 0 02 10 - cp =0. 0 01 cp =0.0005 cp=0.0 W m 0 lb0 2b0 3'00 4'00 5'00 6d0 TEMPERATURE (*C) Fig. 2.3 STRESS-TEMPERATURE RELATION 15
AT 2(I~") 20 TEMPERATURE ( C) 0 100 200 300 400 500 c. R N
& v =0.3 d E =1. 9 8 x 10 " kg f/mm2 m
cy =25 kgf/mm m a =16.4x10-6 m/m/ C c: Es e
-2 0 -
\'N~.
m E .. N
-4 0 -
Fig . 2.4 RESIDUAL STRESS VS TEMPERATURE DIFFERENCE (COIL LENGTH : 2. 8 &E) TEMPERATURE ( C) 100 200 300 400 500 i f 1 i 1
- e u
$ -2 0 -
E E4 W N a Ne m -
< l g -3 0 -
m s
-4 0 -
0 Fig. 2.5 RESIDUAL STRESS VS TEMPERATURF JIFFERENCE (COIL LENGTH : I NF 7 sJ ITE ) 16
_ -._ .- -_ ._ _ .m.. . . _. _ . _ . , . . . _ ._ .._ _. _. ._ 4 i
- .LINE
- -MESH l
- : 1 I te. n
- 6
..a i.n is .i n n.si i
^ 37 1
- : '42 x : 85 i
e e' 0 0 90 ' 4003 e t-9 t 4 1-i 3008 o f-s- t G'1-3 4 gjo 0 200{s , i :. 1. ilt-
-i l'
4 i I A 100'
/ '
A y.. , j '. e .
~~ N
=x_
a e O ' ... . .. .. .. .- . . . -
!!nt ini 0.1 0.2 0.3 0.4 0.5 0.6 l
Fig. 2.6 CASE Al TEMPERATURE TRANSITION ,
'i r ; l 1': 4 ', <1 Iii]i1 !2. l! , : ' I l!ii11i!4 : *tj i: .i; 1 4 s
EO&=
* $Eoxg 4 ~ QU- .
1 2 3 4 5 6 0 0 0 0 0 0 _. . 0 0 0 0 0 0 . Oe . _ F i . g
. 6 i .
2 5 7 _ C 1e A 3 S . E _ A _ i T 1 E 9 i , M . 5 P E R A _ T _ U 2, R 6 . E D f I . S _ T _ R 3 . I 2 ODEmHCm mo $<UW B 5 n _ U _ T _ I . O D _ N I S . T T _ H A R N . O C U G E _ H F _ R . T O . H M _ I . C I _ K N N S E I S D . S E S U R F _ A ._ C E _ ( m m
)
c - 1 2 * , ', * '
.. - - - _ . . . . . - - - . . - ---. . - . . . - - . -- . . - _ - . . , . . . ~ - . . . _ . - - , . . . . ~ - .. -..... _.
i 3 4 l s 600-t N-e . oo e--e INSIDE SURFACE l ,A 500- *s e---e OUTSIDE SURFACE s h
- t
\
\
, n 1 J U 400-
- o t
~
\
t i' w s D l I 300- ' t hM 8
'y A n E t w
H \
\
, 200,;::::-- - _ g
\
\
s ill 100- 's s N 0 50
'4- _ . -
100 155 2'00. 250 300 ' 350 400
! i i i 125 COIL WIDTH i ,
Fig. 2.8 CASE Al TEMPERATURE DISTRIBUTION IN AXIAL DIRECTION-
N l 40 - u tr 5 m m R 30 - a , a o E g / m w ,
? #2 a 5 $
* ' m
$ 20 - / w s- --* O t
= u
~ a , a u, m H M m / g m x z w
/
/ O 10 - I
/
S 0 1/5 2,/5 3,/5 / , 43 1
/
/
/
/
/
9
/
/
/
/
/
f
/
/
/
/
~30- .,/
Fig. 2.9 CASE Al RESIDUAL STRESS THROUGH THICKNESS DIRECTION
~
40 - h N
% ~
CD ' d 30 -,wo-+ -*~***+ in, 'r 4 m ,* \ as } INSIDE SURFACE g
\x N 0 0 01
'd
$
- Ig W- - + 0:
20 -
- x'N s } OUTSIDE SURFACE e - - - -e a t.
' A I o ***ww .* -x..x..w .y -x,,y', 1 N c s H s m K i s m s m 10 - g s I % l %, I i p --- 0 s , -is*~~~~ '- x / _ _ ~ 4 1
*~
50 125 160 l'5De - -*'i 2'50 [ -fg 400 COIL WIDTH e , x .I x wxd**x-x-x-*-4--x x_x----x p ! -3 0 - ._____ Fig. 2,10 CASE Al RESIDUAL STRESS
, TABLE 3.1 CALCULATION RESULTS RESIDUAL STRESS COIL WIDTil.
. HEATING CASE NO. PIPE SIZE llEATING TIME
( ) DENSITY AXIAL
) CIRCUMFERENCE (nm) (kcal/ nun h) OUTSIDE INSIDE OUTSIDE INSIDE 1 30 180 0.0583 .-7.8 6.8 21.0 - 5.6 (0.53) 2 96 0 , 0.0583 -0.6 -2.0 20.6 -22.0 (1.70) 3 12 IN Scil 100 190 180 0.0583 17. 0' " --20.0 23 : 0.j -30.0-(2.84) ,t, w 4' 250
/ ,
~
f (' h ;' 7r 180 0.0583 24.5' ~24. 5 ;- 23.5 -29.0' .c w (4.43) - f g + %. I JP , ,
,a e ~
sf , 5 350 180 /.0.0583 ~ 25.5 -21.7 '25 <Jf -27.7-
., (6.21) -
6 . 204 300 0.0452 ~ 13.5 -19.5 29.7 -30.0 (2.3')2
- o. s f ',
e e 7- - 4 50 ' 300 O.0452 19.5 -23.0 "=28.5- -30.3 ;
. . , ,, f (2.84) m %-
' . 20 IN Scil 100 ' u,,,
- 8 350 300
~26.0 26 ~-0 -29.7
-,f, _
'" / (3.98) '
0.0452r -27.0 -
.g,, u a --- --- -
- - - - - 'E--- h';,,U f
*? -
j _ 2- ^ , ' 9 456 -- 300 0.0452 y 27.0 -27.0 26.0 -27.0 (5.19)
.-g g,- *.
;- f ..
, it l~ '<
.- Y .Jj d ,
,o ,
, ' ,[,
."' ra -4 /
['~ f y ,l~ ', - d 'Q ", r
~ -
r , .
~
l / Yf ,Y ^/
. -. = --
z,,._ . g_ ,
/
- c .)... <
{ ' J h; Y' ~), s :. ' l :. a(} 4.p'
,?
) Z (rnm)
Q Q .i 400 - 121 126 400 N N
, / F * :==::: %=:::::::(
- i 115 '120 314
' f ~ '* x y:5
\. 109
@ @ (
'114 -228 T #
.,, == ==:::==== # 200 - 103
@ @ De108 e 200 5: ; ::::::: ===-
h
,, 97
@ 102 M n -175
< .,s '" " ~
~% :
s
- ====
91 % 150 g i s \ 8'a i i @ Ok ! o 7
\ 3 as '90 125 5 3' ., a: @ @ \
) 112
') =
n 79
@ @ s' 4
~#
, @ f 73 78 102 g 100 -
3 g g . g b o 67 2 90 $ u 3
, 61
@ @ ,s 66 80 5
y g- 55
@ @ .' 70
@ @ s 1 [ 49
'60 60 54
., @ @ sx 43 48 48 36 37 31 b b q'6 30 25 30 24 19
@" O 24 s
18 J b b
- 2' W @ j ,
s , b b \ R(mm)
., IN CASE OF 121 37 1 46 1 59.25
/ Fig. 3.1 ANALYTICAL MODEL
% ! s I 'k
..g + 23
?
l l 30 -
- e_ __ _ - _4
~ O l 2 c, - O-- - - - o -
/
~ ? s ' O ~~ ~ _._
x (12IN SCH100) 4 a/ O HOOP
$x f
' ~
/
O AXIAL (12IN'SCH100) a: /
, w 10- /
i Ed
- z g /
M / U / A / H 0 / ID u e O u
,/
j ___ on
- g 4 O us Os '
E 10-a: E4 v1 < A s v2 O = ! E ' O
- J-i l
' I a i l' i .
1 2 3 4 5 6 7 COIL WIDTH //Et Fig. 3.2 EFFECT OF COIL WIDTH i
- 1 i
, __ ,=. _ _-
- l a
% 15 20- s i
\ l
\
\ ( 12 IN. SCH.100 t
t I i g _e AXIAL (ID SURFACE) , I 10- 'l l -o HOOP (ID SURFACE)
- t l g
-.-e- .. AXIAL (OD SURFACE)
! D I , I M* ,8~'4 g --- o--- HOOP (OD SURFACE) s '**' t e g 15 I ,, # ,g,,_ ! u y o 7777 ,e- s---g--===-------- s ]
$ M
/ /g k
," odo "
~'
300 1
.a 4
ODILWIDDV g
- i g g DISTANCE FROM COIL CENTER (pun)
S # Os e\ ,
"e o"/
1 1
~
y m 0%
/
r f O
% %3'
/ O o
*f i
- Fig. 3.3 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION
[ I
. . . - . - .. . . - . . . . .- -- - -. ~- . .- . . .. . - - . _ .
F i i 4 _ 48 I I i 12 IN . SCH .100 a
> s.
20 . t i l
, AXIAL (ID SURFACE)
+ ~ o '
"g i o HOOP (ID SURFACE) > ~ . p-x t ,"
y
. -__e... AXIAL (OD SURFACE)
~
p e *l
,i %' ,,
___o.... HOOP (OD SURFACE) m
- a ;
I! J' ' *~ PJ
- e O ,v 5 *
'E M , '.,_,,=________'___
r--- -- , a 200 300
/ . # .-
t
.a % d 400 ,
5 oj
/ DISTANCE FROM COIL CENTER (men)
S /
.D '.N
$ %'O O
-2 0 - g Fig. 3.4 RESIDUAL STRESS DISTRIBUTION IN AXIAL. DIRECTION
-40 i
__ L e..a.- - .x . 1 - > a - x .,a.- .s_-1. m - - , - .s -. o
- ) -
I 80 i . __I 1 30- k . 1( 12 IN. SCH.100 fo-og 1 s o 20-
$, e- AXIAL (ID SUR7 ACE) L 8 e"* **g, o HOOP (ID SURFACE) s
& 10 e 'e' o..e"p/i s *s- g .. e.-- AXIAL (OD SURFACE) 5 i 'e j l 's s y '
, s - - -o - - - HOOP (OD SURFACE) h o i .,#NNEPR % o- a
~ ~ ' "
B ,/*E 3'0 0 " 40'O 1 o
. . DISTANCE FROM COIL CENTER (mm) .
E :
. /
M m .- of-
/ ,
-3 0 -#
Fig. 3.5 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION 1 i i 27 Y t !
l l 1 12$ 40- 1 I I I
- I
,s _
30- 12 IN. SCH.100
,g. 0"O $ o.,
o 20- ,
-o AXIAL (ID SURFACE)
'g a 7 k, o.,
o - HOOP (ID SURFACE) N 10- 't.e + 'g ---o-- . AXI AL (OD SURFACE) s
& I g U. s --.o... HOOP (OD SURFACE) w 0 5
.f NE= - -
l7 N 100 200
o *# 300 400 4
m j j DISTANCE FROM COIL CENTER (mm) S m 0 a:
/
*.. + :
ed
-30. "
F1 3.6 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION 1 t i 1 28
.. . m . _ _ . _ -
. ~
4
- l l
4 4 4 40-
- 175 3
30 : AXIAL (ID SURFACE) 1 4... , 12IN. SCH100 e****e-o.-e.. g * #.O', ,. s
' L : HOOP (ID SURFACE)
_ T, 1 Ig N 20- g - 1 ---o--- AXIAL (OD SURFACE) l $ s I -- -o - -. HOOP (OD SURFACE) 5 . e- . . m 10- %e . ')it %
's a \ g
's ',
h s 's
*, 't g 0 , , , ,
i a H 100 200 ,-- 7 300 400 I E a DISTANCE FROM CdIL CENTER (mm) - a 5 A c. ; Fig,'3.7 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION
-40
+ i i e 4 29
. , . . - , - - - - . - - . - - - , - - - , . . . _ . . - - _. . - - , . , . , , , - . , - -. ,----.v----- -- - . . , ,
- - - - ~
. 3- : h - ~ -
- " ' 8 ~
- ;. e. ---i
_ 4,t..:. : , _ . .. . . _
-4
..a . ., .. _ . . . .
.lp _ . . . '.g
. l .. .
.. . , . . ,v.
l' ?. . Q .
-a ..
b I/M 1 h i. . i- ,L k I .5 S. .' * . .
, ,1. .' .
I '
.a .' '.) .
f f
. -[ .
l f a. k,.,av. , r " 4-
? -"
. [ i.
t -- ~
;; l 3
,'. r!. f* .i' I; r+
i:. ,
- - T-vi y 'i'r
'* L 2 *:,.4:
e ,
'*- h. F' n' ! 'l '. 1 <ji ,; d :F i:
z,' i
.. , , . s .
,g..- T p,;- p g .e4 je,
. s.
- r n.
. 1, e
4 .
- , i a b- 7,: 4; 2 pi .
{ n. p . :) e yjg, f r t
'. ,. I.
. a. 4.. c.
t 4:h', ; a r. " i.- J; - d ie
; .r.. * .! h .' '
'.! d. r .
iF r
- 8- '4. , . .- ... -t; i i s..pIb;[' s :!e 2:
4P !'4U;-
^ . .
,,.t.
pn
.L. .r i!-
rs- m i- 9 b-ci- L- d. *4 s 4- ' , .; t-- i-a
+
a . e /s.-- r- p- Iq
? --h. -j ti t- :-
r- --
+-f s. A I; L- - -
L M-
. I r- t. e, a n *- wl.,
.,. .. i.i- a 5 '..
'. ... d .
., ~., %
.i. +s, !..n,M) 'T':31.
c 1 , ..
., .. A*ML Go SuRfAct)"_
- 4. .e. ,. .g !... i: ," M.L n.; I a
- . 4 #. 4 .f; f 1
? p f m. . s L' 'r
.E L 4
- 2. " '!L
, 4 4+ . n ..
. . I. :.
%q,'y f g 7 };-f' .f" 9,
i,. 4% '- [ LE d
.Sf 4, -.
-p--2 L -rJ-- _. ,
_f(90f_{tO30ffACE) 41*q % $y, yyQg ,ird LY'- f'.i.- ::-fil h+a.--i-il 3 d' ~-
; # -F r:
LI--- F k P @2 -M 5'i
'4 h
(" I i i'? h 2 E - H-f' ^ .E' ' l-S2$it'" y' " !9 ' -4
- r- r 3 - - -
4.t-O :! ":. .. ' ' .
-NML (OOJOR/Mrs) .
l 4 m -l'J
- a ;A 5+-
-.it - .
t< i ,,; m i 9 a v 4 ?. - p'
- r. . . a'.y::ir e
.~... t- ,, . . . . . y -t* -
- p. . m 3., ,
m , ;..- . c. : g 7 :r: ,- t'- 9- e
-- -- 3j ;a:e-6 -* * --
-3. . ... -.,-
g T. . -J- c. i.n.. ,- p 3-. --"t 7.E . $,a
.3..%,. , -
. . i .
Q Tt i,
~.;' 'T , T,- -NadP W JaAWl(E) 4; T- WT -
f.: * '1; p.. ,, 7 .L ,# ,3
%:.lp i , ;...%
.t.. :
a ,e.
.(,rg- j. a.s. q- i.: . y :.i..: ,. ... ..
m, . - w % ,a :}. J. ;l IL i ;. , c.
.. y .. ", , .o-- -
; _. _ , , 7a:s
_.,,,*t g] m it O jy g, ,, ., 3 3 4 f / '- '% ;' c
~
e.
3'
$ g'--#'
1,:,4.. . , ,.. 9 p .p .a. a.
%..9 q'*.,.!* -.g .- "i
* /
m 4., -t Lf { i:H:: ',t
.. :. ;- :ssw - iH r-- : 4 n 7..]. . --
- :r . ;.p p ps s: pi :.s. .. t sp -
;4 .-
.: * > .. u ?:
Vnl .. p'e
, ,.. .e6 t..
'l:. .- - - .-
e
.f ' " ' I a ;- /4,W
.;. , .1 T . . . .
- w . . . ,
hsrmct jaou cm cwre8 c4 - R9 ji N5NN'T Y 5 /'i!' T MlD :* -d 's: y- +k1,;,r t i-+ 4 i s 2,e,,-*--
.- +
Z U:i'!" f
- r. &
$l 42 tu.
c ,i,y
/=
n .,
.n =g . ; .. .
. J
- r*. e...
n.- m. 1. . &;g . k,,, :r.: ..r. C, a r +2
.. . ,1 ,.,
. 'pr .:
, .r .,
- m .m.;;.
W g g,;. er .I.t:
~ -+-
- c y
-.,- ,r h. .
".*.!-r j! u 3 p- -j r( tr .a:. .
f- :-, 9 n op J,j ,j. 7 n ;. y ?: i: g -y j jt- r. -
- g '-
7' y9
*+
w.a L :j. q,
.. ., g .r - '
- r p-
.. , . . ~ -
_' a . g . p; '
..J - c.p..ni* ; t, .a. ..,.; N m y .
, s t :r . .
p . .. ..c m- .n 4 . y44 r ... .- .:,. ,s
.,- ..g. .1:.: ..u a...g.
n g.,. .; a_ 4- .;t.
~
-n +- :p m. : .' E
;y ": a .. -
- .. , tr ., . , .
;p i t;. c:.
7 M. '
. ".; J ; :- e , -
. . , !,j' i !' ',.
~
in.,. . l.t. .[ [fl h I,
- o* j,.
, g .,
..,y
;. .7
- i. , _
1 .n *! .o. t p. 3 ,a -
.II. '
f.. - - ,
-_. .~.1. .e.... i;. -(. y y.l
'.. t .
~;--+4+ .l.u- . f .,,a 4 H1.i.
~-- . ; 4.
H.Pa'OT,i .7 .p. . 4* .L2.,.
.,o h; p'
,! p o tg t.*,a ;. ..;.-3.-
5, i m
- ..ir g hi j
. .ir A
a
; l "w 5c..M a.; l.8 .:-, h**- ;4j+4r
.. e,
.,4,4:. ;a.
hr-. ..a
.s.
,p. . . . .* 4 in vi ,. :
c r .a se _; .AL 1 12. 7,. Figure 3.8 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION b , y ( Q $ 7$ h l; j p p f;jki} @ g' g Q l$ w Q $ i! p h g i g g g p + [ ( -
,l.T I. ~! .r 1 yg n'.. :f.
, t ; , #
i
- g. 2[r i i i j g i q;
e ; l ; ;
.__j
- a. . .
.a . . . .._,.
yl ._. -_ _w l J .. .:.)
. . , _; . .l ,
.a .
t ., . . . , -.
- i. . . , ., , y
,i' .
* ~ f t t _ .. -. . _ . -
g '
,. !? .. i. . ;.. .; J , ta:
.r l- g--
t, -- > r s t
] ,..,...:. g g m
; ~
. . . . - ~~
1 y- - - mr; -
- 7. -g !.*p.
.g y .
r:., =. r;- .
'l ((-
., c7 .
7 .ly 3,. . p- w p -w - + - - - --- p- -- - - - -4 - --- --
-.,i. ,- - - - : .- -
- p- -. . -
m f7.j
- y. .. ,. l.
' t' t-
['
- r. s . .
3 1 .; . '-. 'i r - E 1 ,4
-Q E v, f s, s.L:f; n.
- c. -
.!l f-- t ..
., a . 5, .
.s L
2 a n -.:-
- -.i-
.i
.l;
.i AN/Il. ( /O OyMAc6 )
l
' ,N -. . .l . .....-, e : . ..
, . . .- f, cop (fo ,3vg y cg) ,
r% a ,. n , ,
,, , c "-
F
,s .
,r - - - -
;" f 1-g
;- m 4 . - . . .N- , .. .. -
-- .: #A//if (OD ,$vR/ ace ) .
, i.,,,,
s.- .p t.. i
.-, :r l.
i ,..
--,:;. ;; .,. .. t
- p. r_ 9
. 'l t .s
* ' k, ' 8
. . .:.. . J.. , .. .
3-
. .I 7
p ._ 7
..t.. ...L. p .:. . .
w l!g$.: -
- r s.
2 v
; , i.
g <
-a- ^
,-~
~ .a 6'
- T, I
~ - .lc* h.
>.* *) '.' 91 r a y hl % 4 r f
r.- y/p d- : '% g .a. jd, , - .' - { [, -w g .-., .-. h th p: ,
.. s' N
.i 7 -
I Y N . #e. I'
- b :.1 '
i'. :I l 4.i-.4
. ..:. .q.
( . [' f ;
~
D/.s A WCE M D M Cot CEWTER &*~)
- 6) ,' l :' ' , ri '
- i: :r , ... ., e' e i.. . .
'[ i
~ " '**
TA-
~
**,k D 'l-T*--' ---
i 1
'a-~~
! O 4
- i. '
.? . .
. t
'.i
.. /"T- 4" t-4'*f - ** ~
T-I 'T - A' "T -
*f ..e
.p. '
l I' * " I 5" f- J k.
)
...f f '
; b
. -r u ,.
, -. -- d
- j. _ -/" .
.. y - ._.e .. . . ..L .. _
3 .
. . , , . .. .7 ,,, ..
7; ,
- y. -
*- -
- e' w, qp ps e.a2weof
- c f -- -
r; r-g;* .g 3. . it;- ; s 7 +l- y --- i_.
+
.a.._.,.__
_ ) -. e i. p Y b. h;. '- N
- 1. i. b ;f ;
). 'j, ' .'.,7,
.I ..
f l..,,,hL,
.f: .g ."
. [...
- h. a -
..a a,;. .a 4 .. p ,. g. ;..
._ . 4
, g ,
... .p.... . a. .
L,P
,r r' as r r. ../ %
4 !r .t .I ,,;-
....! c 9
,.. er , g u .
. Axel closonixo e
., i as ,- -. ; .. -
a )-7, 7 Q.... 3 i
.o :i . . . e.
it- -+ p a- n, -
-ie
- i;; N , v- + ji 3.a ;- 9t- a,.. ,;-.- -i -.
@ .H. r ,v ,m..s "t;' t':fi 3 ;.~s+. 8, i.
1. ur a, -
- r. ,
., ... i.t -
- - M- 9 3
-(; s
.. i.,
*- ~ * -' ' ,- i' W'd
-3
~. r,. ;P.. i. . 3. d;
; i-
., #dcp(/O J
*r ~. r . 8 - -
e- 1. -
.J r\h '-
~1 N- irT i ;-- -l; li -j r r - 42 -
i e-- dr
)--:e;f -
t- -.: 4 ' "; N! it .r ; P- a +--*-en
+ ' .- ,-r~r-
*1 N 3. r ; i r.. a --- *h ~ -
(; 9 ea \ i. n b - ~
- t. e;'n. ...rl' 8l
-t-m e
], 7t; 6:-
p fy -,d .: 4-7 :--
-{ T--{p7 q- -//,/M Qp yfg n 7 -p , , ,
g :. %
~-
m _;, 3.\ .y .g y, ..p ;. a.. a. 4: ..g 2 6
- q. a i;: ..
- 4 ' -
i.. . ' . t . ir 2
.t- J.
f, s ac r
.. t t a t .
]g;:
7,'h .,:
'.**^ .
[OO M /M N 7 3 [. . $,_ Te 7Y
~ '~ ~~
'44
& .. l~[eJ ., .$....8ip t!.
i a ip .h ,I U
. .,, A TI.,.* a.
. y,. u. ,.
i i, N
%._,_ ,t i.. .
(N
-.Q F-l- w g a A ;-
- e l ,i g t to '
. ,; pi t. as he .:i e > .
c.y ,.. M * .3 .c i;. r pkT -'- j.i :. o
-pq 1.
.e r s
- g. ,'4:
, i
;u :j
. lg p,, '. . l ' ..m
. . . . i
... . ,,1 .p ...a .a. 4 .p.- g a; .a e j, ,- .:; i. .- [. *I-.ig. . .;, . .;..
c,M..Pg y ..t 4! . . ,.. .
..g.,;, .; i ..a,.
4 .;
, . -9. .,, i, s.: .e e. 1.. 3/s i:: , , er .O g g,a ;; ; ,
y af .., .
, g y .,
, , /-,;-- - - -r--' t-- '
y t - 3: j j;,, it r F 1 F '88: i .. : a i, ; { y, ,1 . g!, $ . . * ?; gggr gg off giggg (, ) i,! ,h_ Er /.: ,, yi [y4 h: - m E 'i' N 1 p Y. .S w { , Y
-l'. .[ P P D : A, -
4- .n-H
,.r
- n... -.
p --;, ,
. ~. u
( 9 - . g{ h .h. [+ :/.U h,o,j ( y ' $_i n b I . .i. .].
.. .' f;.
I. . .E!
!; ui :(t a.- m i: .
.J :t[
/j h.t F.f. i.
3.
,k
-arn-+
'si44r e- me * , , . .s
.': 4- a- -, -+-6-- +- -- a -- . . s. . .
"; g;;/a:e -og-~. 4 6 f !c p ;,,r.: H -o ht -'ir ll; es [;; in 1 ;n g ,: f.4 ,
u !i . Ei dl [ ; 7
- t: ge x r .r. . ,a,, w a
- i i a(
;- t.
r i ,
,i .e. . .
--4"
,j.,
-c,;- n n,,
E.p
. '. ". . .c ,
r - t",7. .. T: .
. t. rr.
- l. .
7 t ; s-t- t-, . . . r>-p r; 3r. e 9.. J.'s;;i p; q, .3, t; .;t.
- e. p . g, : - -
.{- g ., ... a.. . , .
..tp ..q ,
. ... ....q .lt,!: ,j q
,--" ,, 4 ..r 1- .r, p: ; - 8 -
+-s' ,, ; ,
p, , , .
., '~-
2-'s-
. , T'*-~ -
..c. n. . .
-q; .,r
;, . l
_a
.i o-- . . .
- p. . .-m, : .;.
.o .+ 0 ' , -e-m d. ,. n
.t. ,
- d- i'..
.I!.. ..
!! "t. ~* .
--1 -. . _J' ;}. - :- :--- l' , e,p y. 3 ,..l...
.k.r U--- p 1 (f.
'd . ' . g .h h ..i h ..
.h S. d- .!$ $ .! d' li h E *
. ._ _ .. 1_ . ?
ic FIGURE 3.10 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION r w .-- . . r; 4,.; np...,a,. ijt. s. 1 y . a .y .,. 'e $ .. 1 _r ' t . ,.
,..p". ..r . , .,,a ;,'. ,.' ,.: c.. j.. .., p
. !. 9. . + . .,4.,
#. 4 v ., ,,
. .t.
, :\ ;..
r, ..
..a .. ,. -
.i. , -
. . t -t ,
e
r J l i j ! i
'i 8 *
] . t
- 8
- l.l
] g a
t g
- i- i l
[
! i e i -
.y.. ' . . '.
, fc.
- I ..
i
.p . . . . - .~..
.?
.i l
+
j. r l
.f-i
.!~22Yl ,
i i-m i
. g. .h.
i
..: 6.-
p
. .) e
,e j
l j ,- I-e l . *
)
' I. -). . .- .
7 ., .. ...
'l, ,
l.
. ! i l
- 1
.g[ ., . . .. .
h-p ..a . .. f . .. ! !. .. . .: - .. 2 /N [t.ft M
.l.
. ti .
..' l 9-'
- N. U
.,y..,Ra%W.
,; 4,
.y,q,9,.s) .
.l.
..,. r- -c -.rt gr
-r --
l ..' l
- (= 4 -
. .m - 1- 1 i I-i
. ,. .,; ; .4 g- , 4 .. ..y i
.a! y. .
F. . ..-A, -
- '- b " -- - A-
} :c... . .
p?. y .-'
,1.3%- . . .;. i e
- [a 4, yi. j./
,.L ~
MIM (to symtg 3
- . : .-..' .u .
y:l4 . .
- J . v._ .a %.. . ., ' .
5 u, _ ~
.... . . , , u. .. ...p -3 '
- r. it
- , L 3
.7 .. y ..
g..... 3. 1 .l
*5 h
_. .2 a
.4 8-1.
17 J. . ..i.
,if
.$ g.,
g[ fa 1 i-l5. &" ? ' f $ .;y .hA . .%. i S. .5 L.. _ ,,;*,c l90f) {pp Syg)gg) f 9 L ,,:0 ', i; i:% i:- " M ".!j,4 cl"Cr.';, ' i
- !i i:. [" )";, ': d in t, H.
1;, g q, ' 8[ ,',. , 1, t; ; i,,j . . .. ,g,,g ,, j 1 .h -- g , . qi R. 9. q!, \- i, i
..o-4~ -fa s-f-
++-
r .: r - r -e-- - - - 4- h I- .-' y- p : :. K L -~i m . .
-; R
~
.' ; * .u%. .! . C. .....:.
- .t; ~1 3 ,-- - - --- ~- -
h/N {0D Suffgl)
.. ,f. .I.t.-, . , .I':
it * - e n. .._ . , .
.'~ 5.1, .. , .... .e. ... r. .. . , -@. .. ,i.;.
N. ... 3.,. ,... .i : y
.'.d.y.,
(, jr i,. . . V i iy. n ':i: .( o
; .h
- s. u. .c .,
.'.i.:.,f_ .._ .,
... .. i 1 i : i
..- ..Oe . .//Jof) (00 Svg/EE) r
- u. .
.. ~ ,.. -.
r . .. ..., ,.
- a. . . ..
,,. 'g. . . .. . e .. ,-
li .il ' .., M i.I d.
,. : s 3 '
y! !. . .i'. g, $ "'..: 1 -p pd, '*-. *.: ,! Y }}. j.: w
' ' - il 'y..
q .. ..
- l. . ii- ' a:. 4- 4 :.14 8l i: 9 ,
B " d -
- a. - ^
/ .,' .i -$ - a w
qr tr . .
; .i :
g ,: .ic3 .. ,. Q 6 .g. c. [, ,p. )- d;
- 0. Jl.. f 3 . H .:!. ,,- :! i. .. .pp.h j .f.. '
f.,- h
' ~ , .,. , , . .
k - r- h N L.mlg .,- .a- , 4 *
& ~ :- L A - ,i E >.M
- A
.!ip y 2 .' .,:
' DUTAWG /RM CM G/7ER p (. v 9 .
a .p ;;; t- 2. y o--. ; 7 1 3 . TD h.. ....- .. i
.;$h
- r. . .
kh [. g. ..
..p i,
,. . 5-I "i: r l.
.7 v..
Y
/ ... ,..
, J ,
g; qN J.,~ . J.
... .p. y. 1. c -
- q. k y '
i; . i; - 2.; n .j.. ... ; .,
;.. p
_:. .;j : . . .. g . .. a ;. . rt 1; : :. :8^e .!;t i. i8 , N, 4; pl p i.. -
;-;b ;s. .
.. : / -
y
.' d.'
p;- ;4;. 4p f ...j:.e ip * *t. i- -]' 6 + 1. .i .-. ., i t u3 m-H M .r' *?- te
,t .h.p i
"IF :i, l;
r ., tei '.i ; s e ,,.;.. ,e 1
,, r s
6
,- -r 8 -- J'- -
,0
[: N pg? -
-ep 7"
;- :: .e. .c ;g 7 m -r ir
,r,d..r f. ,-
P e 4'- y rr- ~ -- rt--- r-p .lj +4 4 -ji :
]t .
i 3: . . . .
.;;i i y. q :
.t . !" . a, -
- p. 's. .
- 4 j '.E jr
- t. j' jp et;
;r .cl-pp- pi il 7;;, ,
l;.
., ...],
tv- r in "ip c]; i;; y .n ; ;/ \. b p ', i ;p *
- f p. 4~ -
.t a. o t;- ' , 7 el' y *'
-t- .a 1. ;g y:. ,..7 ,
7 :, , J'
.m
;"I ' . i. "--
- - ;-- H" y -
r";
._ ' . *a'
,El.u. T a-te es 8-EMI6&,.*9.:~
g.4?. t I=- g W-C "' 7 "i T dr if
?
-j .i i ' :Y
,yp _ . . . .
D, :O.. ,9 _a.a .n. # .q... ,. r- -- i -E
' '~
. . . ir ,
" . H-.
y "1 f .E *
. a; au :, :. 1, b n d q "
p
-.H *t ,S.
A.- e e g 7 tp .ib. a:, er te u:; ;. i i.n;i r g-- ..r i 'T I e-8 r --- 8
--4 8- *+- r- * -h ..
- P-r.
; .i d-m .h n- :r p; 2r -' ~r 'l - ,
'4 r. tr r- -- ~
_,. l . . in e 8-p-
- . ...m
.o .
.. y 5 . .
v.- b UL ,. U' FIGURE 3.11 RESIDUAL STRESS DISTRIBUTION IN AXIAL DIRECTION
.2..
e A + 3 f a s~s
'!; + + m. n.. 3 it ..i,., ,e i;- ,4 ., .
J..q .?..m.., #.. . n. .w.. o.. s in . . . .. d-t - i-
!i ' i- fi.
- r. k# r ft 9
.L
-6 ,..ti +
d1. e :- 8i L.
" dh - .-
..r- -
- p. ....
.I j j ,jd.1_ U i J '"' 5 M: Ul; i gU Ik .h h- E' l! Ml -If- '1 -f' fi i i PI ' i' l'i ? " + t e', '
if :!- ij~ S '- 4U ii: 'jl:
- 9
.p.
i h
'" a'
J 175 I
^
600 - 12IN. SCH.100 A' 500 -
~ y o' o OUTSIDE SURFACE A P
~
400 - o
/ \g ---o-- INSIDE SURFACE A '
S \ o (a b
^
I y 300 - b
\
200 - , o. og g- 'o, \
,0 ' 't ,
100 - <0' D N d' "'~o % sf o' .'S . ,'~o - oo No ~~~ o g 'O'-o--og@8. :-4q 0 i i . . i i i i i . i i i 40 80 120 160 200 240 280 320 360 400 440 480 520 TIME (Sec) Fig. 3.12 TEMPERATURE TRANSITION
175 I 600 - A 12IN. SCH.100 A' 500 - 1
- 400 - ---o---A' O
d a x S 300 - N o s x N b 200 ---o- o-o- - o--o- --o--o- o- o ,
'o%
\
\
i \ I 100 - DC
\
l
\os C
- o- - - - - - me o 0 i i i i 100 200 300 400 l
DISTANCE FROM COIL CENTER (mm) Fig. 3.13 TEMPERATURE DISTRIBUTION IN AXIAL DIRECTION
175 1 500- LA -B
*N 12IN. SCH100 A' B' 404
} . e A-A' SECTION 8 L 300- _ _ ,_ ._ B-B' SECTION g . - - * . g g - g ~%' m 204 ~~,% , O. %
- El %
g % ' -. _
,~~
100-l 0 , , , , 0 4.28 8.56 12.84 17.12 21,.40 DISTANCE FROM OUTSIDE SURFACE (mm) Fig. 3.14 TEMPERATURE DISTRIBUTION THROUGH THICKNESS (AFTER START UP 180 sec.)
i 175 I
. A l
t 12IN SCH100 30-A'
- 8 == A-A' SECTION b 20- l
, b AXIAL 8 0 o HOOP w 10-E4 U a n' A o O ' ' ' ' Vt S 1/5 2/5
//
3/5 u) 4/5 5/5 > w l i ) I
-30_
Fig. 3.15 RESIDUAL STRESS OF-A-A' SECTION 1
TABLE 4.1 CALCULATED RESULTS CURRENr HEATN*1 HEA rim ColL COQuMG I 70TA L MEQUENCY DEPrH WOVL 0 T/ME wroth WATER ## p 'g g 0,3 420 50*
- 30. Y We we.0)) , 3/C ggg W 0. 72.f, "
3 1.$ W, = 0,o32 t/2o o + 200 u,
% 0 Q =0.13l
- 3$0
- 19f xw k.//A 38
L l gy gg O S== (f=0KHz) Ok 00_ L ~S =9.6 (f=3KHz) wx mw
$g $@ o S=0 (f==KHz) om 500- 28IN. Sch.100 h
i O[ a t = 40mm O 400- , O A u a $ S o '
@ .300- ,
$ o u
a 200-O 100 ' h O l'0 2 'O 3 'O 40 (mm) ; Fig. 4.1 TEMPERATURE DISTRIBUTION THROUGH THICKNESS
ra $ o< $$ s< H4 MN m o: sm zo ao Hm 0 02 30 -- o s =.= (f=0 KHz) D'
, A s =9.6(f=3 KHz) u 4 #^ os=0 (f== KHz) 20 -- A
- o. o -o #-
~
~
k [/ D A/o E 10 -- m 2
$ 28IN Sch 100 a $ 0 , , , , , , ,
7 , , , t = 40mm a 1/15 2/15 Vis 4/15 5/15 06/15g150f/15 9/1510/15 n/1512/1513/1514/1515/15 x/t
< 6
- ^
/
x o
, D e-Fig. 4.2 AXIAL STRESS DISTRIBUTION THROUGH THICKNESS
a . - l A - 10 RESIDUAL STRESS IMPROVEMENT FOR 24 INCH PIPES
- .. . _ _ - =.
a
- 1. INTRODUCTION In a BWR power station, PLR (Primary loop recirculation) piping is among the most important piping. It is intended that the IHSI technique be quali-J-
fled for large diameter piping. To evaluate this process on large diameter. components, mock-up testing was conducted on a pipe to forging joint.
- 2. Mock-up The mock-up is composed of a 24 inch forging and pipe. Fig. 2.1 shows the configuration of the mock-up. The fabrication method and welding condition are the same as for a fabricated plant. The piping was fabricated by rolling and welding the plate. ,
Welding was performed using standard TIG and SMAW welding te'chniques.
- 3. IHSI The IHSI was performed using the parameters in Table 3.1. These essential-variables satisfy the values in Table 5.1 in this text. Fig. 3.1 shows the heating cycle for IHSI application.
- 4. Test Results (1) Temperature distribution
[ The temperature was measured using chromel-almel thermocouples which were attached to the pipe outside surfaces. Since thermocouples were not attached to the pipe inside, the temperature of pipe inside surface was estimated using the method of App. A-13 ESTIMATION OF PIPE ID TEMPERATURE DURING IHSI and the temperature difference between the pipe outside and inside surfaces was obtained. 1 2
Fig. 4.1 shows the temperature distribution through the thickness which was obtained by measurement and calculation. The temperature distribution in axial direction during IHSI is shown in Fig. 4.2. Fig. 4.2 indicates that a fairly uniform temperature distribution was obtained at the induction coil center while comparatively abrupt cooling was noted at the coil ends. (2) Residual stress
$ Fig. 4.2 shows the residual stress distribution in the axial direction after IHSI. From the figure it appears that the residual stresses became compressive.
- 5. Conclusion This mock-up test was performed using the conditions of Table 3.1. The temperature distribution was well controlled so that the temperature difference was sufficiently large to satisfy the required value. The residual stresses on-the ID became compressive. This result indicates that IHSI can be applied to large pipe such as 24 inch PLR piping.
3
I TABLE 3.1 ESSENTIAL VARIABLE FOR J-l COIL WIDTH HEATING TIME TEMPERATURE COIL LOCATION VARIABLE DIFFERCICE L (m) T (SEC) AT ('C) f. (m) REQUIRED L T VALUE 7 &K a p 2A AT2 2 hay flMAX( t,15) DESIGNED VALUE 370 150 - 300 1 220 1 15 ACTUAL .: 328 40 VALUE 9 *:(4.1) 22P(1.f0) -i NOTE : *1 The value in garentheses indicates the value of L//Rt.
*2 The value in parentheses indicates the value of a T/t#,
i
N O _ __;:E: : ::[: I T
. A
. R U
G I F N O C 4 6 P 1 U 2 K C O M
- - 1
; ; :;: 2
- -[ r g
i F l J u m.om x e.moQy W
369 I HEATING COIL 0 t 40 - l l g E-
.c !
O
\ @ O ; T m, " $ ;IL < v 60 64 140 _
COOLING WATER 600 500 i U b w 400 i r At j 100 x 100 200 300 400 500 600 TIME (SEC) f Fig. 3.1 J-l TEMPERATURE TRANSITION 6
369 40 - measured point
, l !! EATING COIL
- -x v x l
~ hm l m L s COOLING WATER Q measured on OD _
800 - G O calculated for ID a 600- _ se o---- I w
$ 400- _
n x I 200 - _ O O O
$0 6 'O 4'O 2'O O 2'O 4'O 6'O 8'O 1D0 12014'O Ido 180 DISTANCE FRO!1 SEAM CENTER (mm)
O ID HOOP STRESS
^
40- 0 ID AXIAL " ,, A OD HOOP - g- a OD AXIAL " D 3
~
20- AA' ,, M 10-
$ 0 , , ,
a 3 p b e# 5 w a; Fig. 4.2 J-1 RESIDUAL STRESS DISTRIBUTION 7
, 4 41 &M.+4.JMis.-4P.-+44% 4W.hK& 4 M
>N-AsbMe EW+-44 e.ab 6 4 4Mk we e --
*Wl Y 4 f
a ' i h 1 ( i i i .< J t. 1 I 1 ' 1 4 i 4 l.
/
P
. r 4
l 4 ' 4 , L .) l i j '< 1 t t l
; -i l
, i t < 1 .
- . i r 1 I
i r b i i i l , i 4 h P 4 4 d i 1 k I i i. 1 2 1-
i d Y A - 11 ESTIMATE OF INPUT POWER FOR IIISI ( (
s - * ..
\ f \
, . -> r1 t
tE t t
)
+,
/
- 1. Ih"rRODUCTION Vg N
,, ,s As mentioned in Chapter 7, Field Application ; suitahls '
,, 9 input power produces good results. We have developed a method si\ '
x i
.o .
which estimates input power by using an experimentally obtained ,
< ., y j, )i equation. .
- 'i c'
<4 / ' :- ;,'
2 EXPERI! ENTAL PISULTS
=
s Coil input power Pc is given by the followick equation.
'\
W1 -
- B.
11 ,t
~
Pc =.ih AT.D.L (g) g-t ~ 4 + Where, Pc : Coil idput Power , AT : Temperature dif ference .g D : Pipe Diameter 4 L : Coil sidth $ t : Pipe thickness ' r) : Ef ficiency (Experimentally determined) ' f . s The experimental results are summarized in Table 2.1, , I
>, 4 (CASE NO.1 ~ 4) and these data are graphically shown in
.' , 4 Fig. 2.1 using the parameters shown in equation (1). -
i . These results show that there is a linear relationship between~ ? Pc and AT.D.L./t , 4 4 TA k + f i r 4
\ h..
's_ 4 A
P J' 9 2 ' { _* 6
i l l l . l: i 3.' ANALYTICAL RESULTS l l In article 4. current frequency of App A- 9, the input power is calculated. The calculation is performed for a 28 inch schedule 100 pipe. The calculated total heat generation is 200 kw. Coil input is calculated as follows: Pc = 200/0.75 = 267 kw O.75 : Efficiency The data are summarized in Table 2.1 CASE NO. 5. And data points are plotted in Fig. 2 1. . 4 CONCLUSION The required coil input power Pc is proportioned to AT.D.L/t and that value can be obtained by using Fig. 2.1. 5 4 1 3 \ 4 t-4 0 k i c, a 4 i 3
- - - - - - ,.-w- ,
Table 2.1 COIL INPUT POWER PIPE SIZE DIAMETER THICKNESS COIL TEMPERATURE COIL AT.D.L HEATING CASE WIDTH DIFFERENCE POWER t TIME (lN) D .(M) t (M) L (M) AT (*C) x 10 3 (SEC) 1 12 0.3185 0.0214 0.25 345 117 1.45 180 , 2 20 0.508 0.0325 0.24 418 126 1.48 300 3 24 0.6096 0.0389 0.364 470 230 2.88 288 4 16 0.418 0.032 0.273 437 153 1.56 200 5* 28 0.7112 0.040 0.500 360 267 3.20 420
- calculated data by FEM.
A: Calculated by FEM o: Measured Pc (kw) 300 200 O O 100 1 2 3 3 AT. D . L t Fig. 2.1 COIL INPUT POWER 5
A*%_. _
-.r,, -.4*.& 1. _m ._ _ m mm. _ a- J_.__A.:. 44-_ ,ia sa _- - + 4-. . w_.a. - . _ .m 4
d t i A - 12 l TRIAL IIEATING TECHNIQUE OF IHSI 4 .i l i l I i l i e I 1 1
,, , ., a _- - - --< - , , . , - ,--. -
- 1. Introduction It is desirable to standardize the IHSI performance. Suitable heating conditions are developed to satisfy the essential variables.
(1) The temperature difference AT of the outer and inner surface of an IHSI'ed weld joint shall be greater than the required value. The required value is explained in this chapter of the test. (2) The heating duration shall be larger than the required value which is determined so that Fourier No. is greater than or equal to 0.7. (3) The maximum temperature of the outer surface shall be lower than 550 C. The temperature distribution depends on the coil performance for the most part. The desirable coil produces a good temperature distribution which is uniform in the heating range and does not produce small hot spots. Af ter the coil design is fixed, we must determine the input power which is suitable to satisfy the above mentioned conditions. The input power can adjust the following items. (1) Raise or lower the maximum temperature. (2) Make the heating duration longer or shorter.
- 2. Trial Heating Trial heating is performed by heating the pipe until the maximum tem-perature is 280 C so as to confirm that the required AT will be obtained and to adjust the power.
2
1 l Since the heating is under 280 C, repeated heating produces no side effects. Fig. 2.1 was developed for the trail heating. (1) The following data.should be measured from the trial heating. 1-1 Heating time T w c s necessary to raise de maxh 280 temperature to 280 C. 1-2 Weld joint temperature at the time when the maximum temperature rises to 280 C. 1-3 Cooling water temperature tw, cooling water velocity Uco. 1-4 Pipe thickness at the location where the maximum temperature occurs. (2) The following items can be estimated. 2-1 AT value at the weld joint when the maximum temperature is 550 C. 2-2 Whether the input power is suitable. .1
, , Excess power Temperature -
550*C ,
,' , . Suitabic power
/ / ,,
/ s' .
/ /
e' '
- .. -Insufficient power
/
s' ,' , * *
/ ,#
280*C ' ' - ? $ F=0.7 Time 3
(3) Items estimated from trial heating.
' The trial heating is performed .to estimate the following . items.
3-1 AT From Fig. 2.1, "Esthmating AT~by trial heating," if one uses the obtained data t, to tw and plate thickness L, then ATE, the-temperature difference in this tested state condition, can be estimated. In this case the steady state condition
.is considered that condition where the Fourier No. is greater than 0.7.
From Table 2.1, one observes that Fig. 2.1 shows C C % 60 0 C lower than the actual temperature difference. 3-2 Suitability of power level If the estimated AT is smaller than the required AT, the power i shall be raised. i 3. Using the approach of Fig. 2.1 (estimating AT by trial heating), Fig. 2.1 is used as follows: Step 1. Obtain the following parameters from trial heating data. (a) Plate thickness L (b) Heating duration T (c). Measured temperature To - Water temperature Tw Step 2. From. fourth quadrant of Fig. 2.1, Fourier No. is obtained using t and i as shown below. Tourier No. i Ng L 4 i a
, , - - , , .- % 9 mu =,, -. .- _ _ . -
,._,< ~ - . , p.m i , ,,-.-.-o
Step 3. From first quadrant of Fig. 2.1, the g(t) value is obtained. Where the g(t) value denotes a nondimentional temperature. This process is presented in the sketch below. 8(t) I F s - T Step 4. From this intersection point move horizontally left, intersecting the measured value of (To - Tw) of the lower 4 abscissa of the second quadrant, move vertically upwards to the intersection point of the above mentioned horizontal line as shown in the sketch below. s(t) m i J-To-Tw (280*C) f 4 s 5
Step 5. Connect the intersection point and the origin of the coordinates and extend to obtain the intersection point with the upper abscissa of the second quadrant. This value is the predicted maximum temperature of outer surface - water tempera-ature, and the predicted maximum temperature of the outer surface is obtained from this value as shown in the sketch below. To-Tw (T==) r- 7 To-Tw (280*C) .i
- 4. Derivation of g(t)
Begin with an infinitely wide plate whose temperature is O C uniformly
, as the initial condition. When heat is generated at one surface which is insulated and the other surface is maintained at 0 C, the Green function Go(t) of the heat generated side is given by the following equation.
2 Go (t ) = f y e - ( "" ) : 3 ,,,,,,,,,, (1) 6
Where, L -: Thickness '(m) T: Time (h) a: Thermal diffusivity n: Natural number The Green function means that when tne-instant heating source which gen-erates heat and disappears instantly occurs in the substance which is uniformly at 00 C.as the initial condition. Tne Green function gives the temperature distribution in the substance. The Green function after a heating of T hour-is obtained by integrating
=
equation (1) from T=0 to T=t.
,t 2n+2)*22aT t a (1-e 2 T' )
G(c)=)! o Go(t)dt =(2n-11*
=1 2
~
f n. F 2
, , ..... (2)
If the heat generation is q kcal/hm 2, then 0=E-E 1 (1_,-(2n-1)2"2at).... 2" F (3) A T n=1 (2n-1)2,2A 2 T' In order to obtain a dimensionless temp. , divide eq. (3) by the steady state temp. t== which is obtained from equation (3) after i an infinite heating time following the heating application. 9t t e g(t) = g2; 9 = 1 _ 3 2 2F')
= f2 [ 3_t , x (1-e (2n-1) 2 T 2
.... (4) 2
, 7 I i
'"he value g (t) in equation (4) means (Measured temperature of outer surface - Water temperature), becuase the assumption is that the inside surface and water temperature are at 00C as the initial condition.
I f 2 1 1 1 ) ) 4 8
TABLE 2.1 TRIAL HEATING RESULTS Er% uf) #~ T!*MM T~fer b Tgo 7n v ot" T N o.
&fjara a sT $#h"J' T*ZE#-
<>,~> ^Tt .e > c c3 ve) ges? Cc) <. '4 L'O
/ 32 o JP4 JJ7 45 S7 ?SS sos sgo 2 J3. 5 ??f zt4 A so /P4 J SJ Sto 3 2P 42? .?7 o SP 47 270 Sto s/O 4-
___J /. 4 JC/ J3P_ _
/? V7 2?D 4z] 460 S 34 347 JS4 5 i'T ??o 4/7 4Jf
& J4 3rd 270 24 4S /78 JM sor
- NOTE
- T Max : Maximum temperature To : Temperature at outside surface of welding joint T280 : Heating time at.which T max becomes 2800c
5 E 1
- s. -
i i tot / Tom A B O E4
'o o oooo o o o o o' - 0. 6 o m eT mo me4 m o LA T T T T T M MM M' M
- 0. 5
- 'N -
.4 i
. . - 1 i . . . l ....i .. , i 1
250 200 150 0.2 0 .' 3 0.4 Tot - Ti(Tw) (*C)
; t=16 i
~
. \_
\
t=18 t=20 1 ! 50 - I l j - t=25 l . I i j t=30 4 i i 100 - T (SEC) Og e*8
'o 4
1 i Fig. 2.1 ESTIMATION OF STEADY STATE CONDITION BY TRIAL HEATING _ l I 2 1 i 10
I A - 13 ESTIMATE OF PIPE ID TEMPERATURE DURING IHSI
- 1. Introduction The effect of IHSI strongly depends on the temperature difference between the outer and inner pipe surfaces. The temperature on the outer surface is easily measured by thermocouples as to the inner surface of the installed pipe, however, the temperature cannot be measured generally unless some special device is provided. Therefore, temperature estimate for the pipe inner surface is an important issue.
Generally, the heating duration of IHSI is sufficient so that heat conduction can be regarded as steady state, and the coil width is very large with respect to the pipe thickness, consequently, this condition is reduced ec a one dimensional steady state heat transfer problem.
- 2. Theoretical considerations (1) Fundamental Equation Under the steady state condition induced by a long coil, heat transfer in the radial direction is dominant near the coil centre. This reduces to a 1-D steady heat transfer problem for which the funda-1 mental equation is given as follows:
$ +-f w = 0 ... (1) 0: Temperature at point x 1p: Heat conductivity of a pipe g w:~ Heat generation density ,
x: Distance from the outer pipe surface 1 I 2
Heat generation density (w) in induction heating is giv.i as follows: w=wo e ... (2) w:o Heating density at the surface where, S = h [u f x 103 9 : Specific permeability f: Frequency p : Specific resistance (2) Solution Assuming that the outer surface is thermally insulated and its temperature is To , Eq. (1) is solved as follows: o3 2 2 e - To wu p (-1 + 7x + e-2x/S) ... (3) As the temperature at the inner surface is obtained subst)*.or'.ng x = L into Eq. (3). The temperature difference between the outer and inner surface in a pipe (AT =oT -T)i is given as follows: AT = To - Ti= . 44p I-l+S L+e- ) ... (4) (3) CALCULATING THE TE fPERATURE DIFFERENCE OF OUTER AND I:NCR SURFACE . IN A PIPE As shown in Eq. (4) , wo is the only unknown, wo is determined from the following calculation. The heat flux at the inner surface: gy is given by Fourier's law as follown: de 91 " ~\P dx *************************** (5) 3 -
Heat flux in the water at the pipe inner surface: q is given as follows: 2 q2 = h,AT y ...................................... (6) h: m Average Heat transfer coefficient 6T 2 : Temperature difference between the inner surface and fluid (bulk) The differential coefficient of Eq. (3) at x=L is given as follows: de W dx x=L " ~ 2Ap o S ( ~*-2L/S ********************* Under q1 = q2, Eq. (5) through Eq. (7) are reduced to the following: 6T 2 =Ti-Tu = (1-e
~
) .'.................. (8)
Tw: Temperature of Fluid 7 tom Eq. (4) and Eq. (8) . S 2L - 1 AT + AT2 = To-Tw = Wo 2 S ( 5 (-1 + 7 + e 2L/S) + g (1-e2X/S)} From Dittus - Boelter's equation. .................. (9) hmd 0.8 0.4 Nu = = 0.023Re Pr - Aw Au 0.4 0.8 d .2 v 0.8 Pr hm = 0.023 0 U, ................ (10) A: w Thermal Conductivity of Fluid d: Inner Diameter of a Pipe v: Kinematic coefficient of viscosity Pr: Prandtle Number U,: Fluid Flow Rate 4
,m .. , , , -,
_n,, - - , - ~ .
From Eq. (9) and Eq. (10), wo is given as follows: 2 1 Wo = (T o - T w) x 37- 0 .2 0~8 d v 21p (-1 + 21. S_ + ,-2L/S) , 2L/S) 0.0231, Pr o
.4g 0.8(y_,-
............................. (11)
The tc. perature difference of the outer and inner surface of a pipe is obtained by substituting Eq. (11) into Eq. (4), it is complicated, especially at a reactor site, to make the calculation described above for each heating and therefore a monograph to calculate the temperature difference is useful. One examole of such a monograph is shown in Fig. 1. I The temperature difference AT = To - Ti is obtained as follows. Step. 1 Locate in Fig. 1 the measured T -T 9 Step. 2 Move horizontally to the line for the value of the pipe thickness. Interpolation =ay be made'for intermediate value of the pipe thickness. Step. 3 From this intersection, =ove vertically upwards to the horizontal axis. (This point shows ao , the heating densi:y at the pipe outer surface.) Move vertically upwards frem this point to the line for the pipe thickness. Step. 4 From the intersection obtained in step 3, move horizontally left to the ordinate and read the value of ST. 5 i
~ _ . _ -m-. . , ,
Notes: It is noteworthy that w o , and consequently AT, are not sensitive to the pipe diameter, the water flow rate and the phisical properties of water as shown below, and therefore one monograph will be sufficient for the determination of AT at a site. The reasons for the above mentioned statement follow. (1) Effect of pipe inside diameter: d It is clear from eq. (11) that the inside diameter contributes in the formof do.2 to so and the variation of wo owing to the change of d is very small. Consequently, w e can be considered to be almost independent from d. (2) Effect of water flow rate: U. The effect of Um on AT (the te=perature difference through thickness) is shown in Fig. 2 and becomes very small if U. > 2 m/S. (generally, U. at site is greater than 3 m/S) (3) Effect of water temperature: Tg The physical constants of the water affect wo in the from of 908 /(A w x Pr 0 *'). This term decrease slowly as temperature rises. Therefore if the constants corresponding to a lower temperature are used, wo , and consequently AT, can be estimated conservatively. 6
i 1 I (*C)- 24BxSch. 24Bxe, g 100 [ 100 500 7 12B xSch.100
=T -Ti
~
~ A=
p 14 kcal/mh*C 300 r / a Aw= 0.537kcal/mh*C v = 0.7355=10-'m2 /s s = 9.6 x 10-'m 200 e g U.= 3.0 m/s Tw= 35'C ( 100 0 30 40 50 60 7 ) 80 x 10'kcal/m'h I wo 100 To-N 200 - - i 300 350 x' X s 400 \ s
'12 BxSch.100 xs 500 s
-20B"Sch.100
; l
) 248"Sch.100 Fig. 1 TEMPERATURE DIFFERENCE BETWEEN OUTER AND INNER SURFACE OF A PIPE
( I 7
.__ ..- . . . - ~ , . . _ . _ ~ _ . - _ _
i 600 , , , U 500 - ca E y 400 - n. g To= 550*C ga Tw= 35*C cu m o 6 300 - S = 9. 6 x 10-3m - I L = 19 . 2 x 10-3m so g d = 0.278m
$ 200 -
Ap = 14 kcal/mh*C _ Aw= 0.537 kcal/mh*C F = 0.7355 x 10-'m2 /s 100 - Pr= 4.98 _ f f f f 0 1 2 3 4 5 FLOW RATE (m/sec) Fig. 2 EFFECT OF FLOW RATE ON AT
i l r A - 14 ANALYTICAL STUDIES OF IHSI FOR A PIPE WITH A SMALL PRE-FIAW
- 1. Introduction When applying IHSI to a pipe with small preflaw which is undetectable using ultrasonic technique, one must confirm that IHSI is effective for stress improvement and that the small preflaw can never propagate during the IHSI treatment.
In order to evaluate the behavior of a preflaw during IHSI, Finite Element Calculations were conducted as follows.
- 2. Analytical Models and Programs The analytical study of IHSI was conducted for a 12 inch Schedule 100 stainless steel pipe. It was assumed that only the inside surface of the pipe was the heat-transfer surface, and that the others were insulated. The analytical IHSI condition was assumed to be as shown in Table 2.1.
An analytical model for calculating the temperature distribution through the pipe wall and an analytical model for calculating the residual stress distribution through the pipe wall are shown in Fig. 2.1 and 2.1 respectively. In the former analytical model, the pipe wall is divided into nineteen equal parts. In the latter analytical model, the analytical mesh as shown in Fig. 2.2 1 is offset since the constraint condition is changed due to the preflaw and the history of stress-strain near the preflaw is affected by the preflaw. The finite element calculations consisted of transient temperature and elastic-plastic stress analysis. The programs utilized are ITEMP II and IEPTC II which are developed by IHI. The ITEMP II analytical code was developed for the purpose of solving the steady state and *.ransient temperature distribution in
~
an axisymmetric thick cylindrical body and in a two dimensional structure. 2
l The IEPTC II analytical code was developed for the purpose of solving the thermal elastic-plastic creep problem in an axisymmetric thick cylindrical body and in a two dimensional structure.
- 3. FEM calculation results for the temperature analysis The analytical result of the temperature distribution through the pipe wall at maximum heating was obtained as shown in Fig. 3.1. The temperature inside the pipe and the temperature outside the pipe were determined to be 91 C and 368 C, respectively. The temperature-pipe wall curve was nearly exponential, and the temperature distribution through the pipe wall was likely to be steady. The thermal history was obtained as shown in Fig. 3.2. The highest temperature curve, the medium curve and the lowest temperature curve present the thermal history of the pipe outside, the pipe intermediate position and the pipe inside, respectively. The temperature history was calculated 500 seconds after the heating initiation.
- 4. FEM calculation result for the stress analysis The result of the elastic-plastic stress analysis was obtained as follows:
The temperature profile calculated by ITFHP II was the input source for the temperature-load data program. The temperature-load data were input to IEPTC II program, the stress analysis data were calculated. The COD (crack opening displacement) at the maximum heating condition was calculated as shown in Fig. 4.1. As sho a in the figure, the COD increases as the initial flaw depth increases. 3
Fig. 4.2 shows the distribution of the axial residual stress on the extension of each flaw after IHSI. The residual stress showed the maximum compressive value at the flaw tip, and became gradually more compressive as the flaw depth increased. The obtained residual stress is larger than the residual stress that was obtained for IHSI of a pipe without a small flaw. The distribution of the axial residual stress through the remaining pipe wall is shown in Fig. 4.3. The stress distribution for the position at a distance of 2.4 mm from the small flaw was unstable as compared with that for the ordinary IHSI. The stress distribution at a distance of 17 mm from the small flaw ignores the presence of the small flaw. 4
4 r y,75 yy; pipe cKA ck RATro or B"U"[^ E Y IMAT 7&AN NEAT /'/9 /MrAT/"T pytexaE-33 pef'Tt/ C h k'^:tEf7f TM/WW ccfff/C/EACY Teas /TT T/H2 (-m %) ( s)o c _e. wry t *;h>_ t *U (!"E*~'A **) ((**0br +8) (3 2-N
~
CO /9 ?. 4 / ?. S 20 4.oSo<10 o, o Sh? /Pd w ..._ TABLE 2.1 ANALYTICAL CONDITION ] 1
& 1+ a H. .s., + 2 a-- - - . eB-aA------.4* 4 - ,-A 4 JL -.4 -e-,. A d
e'e 140.25 19 g
?! ff 21 14 M fa ?? te M in 11 if 11 18 M M 17 sa BA 40 E 1 ED Z
R WATER COOLING (INSULATED FOR OTHER SURFACES) Fig. 2.1 ANALYTICAL MODEL FOR THERMAL ANALYSIS I 1 i
1
^ r'
~
,': Vy f'L.
-s W
ss p /
/ '
- ^ .
/ l
/ )( )
/ 2
%' ~
+ .
S
/) ' nN.o,&M T
/( / mH s N E
/) // r //>// /f/ / / / / / / // / V /$/ M p
/(
/
/ m o t[
mJ C mJ mJ g - Y E L s s( O
) , ' ' E, /
K
- F
, .O A -.,46 7_ E N R
f, t, R C I
, F
( 5 --
+ o a S 0 " n S 2
E E E R C T C A S A F F R L e, R U A U S U S' D R' R I E S - E N E T N s R U I O R O F 0 L 5 9 - E 9 9 D O N M 4 ^w L r a A t C I
- T
- Y p 3 s L 9 I a A -
N t . A a
- 2 ,
?'
f< a N .- 2 6 7 g i F 0 7 7 7 I
' t N
7 . 9
.6 2
\ /N(-
K
- 1MD i
5 l L - M I rW
&t y .'
3 L g
.Fr- ~
[ 3 ,-
's ' < " ' s ' , T '
C P, '
~[ .
f'l.5 v4 l
~
q ,_. p\l,;,ll,1 I
.[; ,1
.ll wU x
- gwEDO )
I 9 N e 1 O I T C
-- E R
I D E S C S A E F N R K U C 5 S I I 1 I R T E N i l N G I U O M R O H R T F ( E N C O N I A T T U S B 0 I I 1 D R T S I D E R U T A R E P M E T
'S 1 3
g i F RU4 nDm :cws-i 0 0 0 0 0 0 0 0 0 4 3 2 1 E&y6v m
l E T C N A I E
- 0 F O C ;0 R P A 5 )
U F . S Y R C gA U E R W $ S : . S E D ( T<I R U M E E O ylX p N N I M I T 1 0
. ; 0
[ 4 _ E V R U C 0 N
- ; 03
%x^ O I
T I S N A R T M. E R U T 0 A x ;2 0 R E _ P s _~ . M E
- T 0
T ;1 6 3 2 x _ . 3 f = g i 0 F
;0 1
x . x .
. 0
/ 4 O
0 0 0 0 0 0 0 0 0 4 3 2 1 - _U ~ $ $$ . 1lll l
1 i 1 60
$' 50 Q Nby Y *v
- a
$ 40 e
O 5
@ 30 z ?;, 2
'20 k N 10 %
U
- 'J 6 0 1 2 3 4 5 WALL TilICKNESS (mm)
Fig. 4.1 CALCULATED COD AT MAXIMUM llEATING 10
-- -- - CRACK DEPTil 1. 4mm CRACK DEPTil 2.4mm CRACK DEPTil 4.2mm 50 Y
t
~
S N e ni 0 j' m . i l
.: p - l
$ I
/
H l / D l // a i i
/ /
< ' i S /
G i I' .1 ,'/ y I il a -50 -
$ Il M ll 4 I
-100 0 5 10 15 19 WALL THICKNESS (mm)
P'- d.2 CALCULATED PESIDUAL STRESS ON TiiE EXTENTION OF FLAW OF 12 IN. PIPE 11
1 Il I I 4 C T2 P = E I
)
N Z t O 6 6 I
; 0 T C
E R c I D S S E N K C I H T uO<k%om %mE8 4 H G U O R H T ( N O I T U B I R T S I ) D K C S A S R E C R T M S O R L P A U m D m K;I lJ C I S 4 A - E 2 R - R ( C - 0 0 0 0 0 0 0 0 0 WO<k%om xRZ" 4 3 2 1 1 2 3 4 3 4
' gs
- Ob umMfm ,
g a i F g"
1 l I I0 C7 T1 P= EZ I Z t ) 6 6 0 N 3 O I T C
- c E R
I D S S E N K C I a5Ev :csg 1 H T H G U O R H T ( N O I T U B I R T S I D ) K S C S A E R R C T S M O L R F A F U C Dm A I m R S 7
- - C - - - E 1 c 0 0 0 0 R (
a5Em tW3 4 3 2 1 0 0 1 0 2 0 3 0 4 4
^ $U gEsm g
i F O \11 lllIlil 1
Residual Stress Improvement by Means of Induction Heating IHI Ishikawaiime-Horime Heavy Industries Co., Ltd.
UDC 620.194:539.434 621.365.5.023:669.15.018 (c621.039.53 Residual Stress improvement by Means of Induction Heating Tadahiro Umemoto* Shinji Tanaka* A new methodfor improving residual stresses in pipe weldment has been proposed. This method, named 11151(Induction llearing Stress improvement), has been developed to improve residual stresses by creating local plasticflows through the application of high thermal stresses induced when a pipe is heatedfrom the outside by an induction coil and cooled simultaneouslyfrom the inside by water. To obtain a better heating condition and to make a demonstration test, some basic studies were conducted on 304 type austenitic stainless steel pipes. The residual stresses with and without 11151 were measured by the conventional strain gage method. A cracking test was also conductedfor demonstration by boiling 2 42*l MgCl solution. Thefollowing experimental results were obtained, showing that 11151is capable ofsatisfactory improvement of the residual stresses in pipe weldment.
- l. liigh tensile residual stresses were lowered or shifted to the compressive side 2.
1 No cracks were observed on the inside surface near the pipe weldment after the application of filSi Little variation was observed on microstructure and mechanicalproperties.
- 1. Introduction and 12B were used for the former experiment, and 4B and 128 were used for the latter. Elastoplastic calcula-It has been noted that intergranular stress corrosion tions by the fmite element method were also made to cracking (IGSCC) does damage to weld heat affected serify the experimental results. Results of these efforts iones of 304 type austenitic stainless steel pipes in boiling to improve residual stress by IllSI and a discussion of water reactors reducing reactor availability. IGSCC is the applicability of the method to actual plants will be known to occur by interactions of three factors: stress presented in this paper.
(residual stress due to welding, opecating stress, etc.), , , material (sensitization by welding, alloy element con. 2. Principle tent, etc.), and ensironment (dissolved oxygen in reactor It is known unisersally that thermal stress occurs when water, corrosion products, foreign material on inside a difference in temperature is introduced in a substance. pipe walls, etc.). Various measures hase been taken to If a temperature difference is introduced between the sohe these problems. As to the problem of material inner and outer surfaces of a long pipe with linear tem-sensitization, for instance, reduction in carbon content, perature gradient, then the thermal stress produced and battering process of covering inner pipe surfaces between the two points can be expressed by the following with SCC-resisting weld metal, or resolid solution treat- equationm: ment, base been attempted. To improve environmental factors, deaeration operation is being performed in a = Ea J T/2(I -- >) . .
...(1)
I P*" where The induction heating stress improsement (IllSI) is E: Young's modulus a method deseloped to improve stress factors, especially a: Linear expansion coefficient residual stresses in pipejoint weld affected zones. In this 9: Poisson's ratio method a pipe is heated from the outside by an induction JT: Temperature difference between inner coil and cooled from the inside with water simultane- and outer surfaces ously so as to produce an appropriate differential tem- As I ng as this thermal stress is within an elastic perature between inner ami outer pipe surfaces for the regi n, r below the 3eild point (point Q) in Fig.1) of resultant thermal stresses to improse residual stresses. material, the stress is relieved when the temperature The authors conducted a parametric surv:y to deter- difference is remosed, and no residual stress remains mine better treatment parameters (such as. frequency, behir.1 lloweser, if the thermal stress goes beyond the he.iting time, coil wid.h, etc.) and a test to demonstrate yield point (point 2 in Fig.1), the subsequent removal the etTectiseness of the treatment. Test pipes 4B,108, f the temperature difference will cause thermal strain
~
~ '
to occur linearly from point 2 to point 1 on the stress-
- Plant I ngineering Departrnent, Nuilcar Podr Engineering Diu. strain curse (at an angle equal to Young's modulus) and sion then return to the zero point. In this instance, a residual 1
lHI Engineering Review g stress with a reverse sign with respect to the thermal stress that has occurred during the heating remains behind.
} The treatment to be discussed here is a method for i
~
Y introducing a large temperature diference between inner ty 'l' and outer pipe surfaces by heating the pipe from the ig outside with an induction coil attached thereto by a high l 3 y*""* *** frequency induction technique while supplying cooling water to the inside pipe face concurrently (Fig. 2-(a)). l The thermal stress obtained as a result of the temperatu o difference is expressed by equation (1). Under a proper
'I temperature difference condition (JTii:2v,(1-9)/Ea),
# the pipe length becomes shorter at the outside than it Fig.1 Stress. strain diagram originally is due to its yield to compressive force, while at the inside it becomes longer than the original lengi j'L because it yields to tensile force (Fig.2-(b)).The difference between the inside and outside pipe temperatures will
{ ] disappear as the heating is stopped, and the outside pipe area where compressive yield has taken place will I I;t be stretched back to the originallength. The inside pipe I I area where tensile yield has taken place, on the other IIIIIIII ***"' hand, will be compressed to the original length (Fig. 2-(c)). In other words, tensile and compressive residual stresses will be produced in the outside and inside pipe [**d w _. surfaces, respectively. This principle is applied to pipe _,, og, l l weldments with the objective of reducing tensile residual ( q '- r i / [ stresses in inside surfaces of pipe joint weldments, or h (J ( .. : h/ ( otherwise shifting them into the compressive side.
"C.( Q5& g,,,,,,,._
l sr
- 3. Experiment g
3.1 Test sasiples 5**a ****~ Wy 304 type stainless steel pipes of sizes 4B,10B and 12B, rer-o) and stainless forgings of 12B were used as test samples, 7 , %*. _[*' * " g'
- d, the mechanical properties of which were as shown in 9 (
[ l c- L Table 1. Test pipes for a parametric survey were all sub-c- jected to stress relief heat treatment (900"Cx2 hours), EaJT and those for demonstration testing were joint welded
* ) * ""**" 2( t - >) in or by the same way as normally used for actual plants.
E: Young's modulus The demonstration test pipes were combinations of [a a: Linear expansion coemcient straight pipe +an elbow] and [a straight pipe +a nozzle g safe end] so as to conform to pipe configurations in an
" J T>
Ea actual plant. An example of the test pipe used is illustrat-a,: Yield stress ed in Fig. 3. Fig. 2 Principle of IHSI 3.2 Experimental methods Table 1 Mechanical properties ltem Mechanical properties %N N' ' _ N Yield point Tensile strength Elongation Remarks
- Pipe parts ,\gx (kgf/mm8) (kgf/mme) (g) diameteN 4B 70 R-IB, R-2, R-4A 24 57 10B 22 55 70 R18 12B 24 58 70 R21-R23 22 56 72 R-25,26,28,29 Elbow 12B 26.1 54.5 67.5 R-24, 27 ,30, 33 Norzle safe end -
(Note) ' Refers to the test sampic No. or joint No. 2
Vol.11 No. 4 October 1978 I 4.) 6 , e,s (b)o 4 q':h
$ __. $ N96di,c a sed ' " W <d s
K 1 V
=
..tw ,e
( l V
. tw ee i
A) . e 1
** (!K<.de=W *w= c
* .(!4 r. d 6.s
- 4. .weta-so; .
no en j ,,L % /' *" *W i ,,. : w. msun.23 2 2 JHes, (128)
! j . -
g- j'o --- g,
' ^#'
' , we g'HS!'R-2)
E
. i& , . c '" e i 4.4 4 60 .~ o c. 6 i c -.6 c ().
Fig. 3 Configuration of test pipe
.I " '(sh *w < ) g *"(n)a.d
) g *p ***'"'*M y,
g a,,*D) S ,-gwainst u pc -~- l- $[ T.8 pee I y .. .
]<
.g i e
'o c. 6 'c h>
p,
- e.ias. W.1H51(R 2)
N' m = Fig. 6 Strees distribution s!ter application of IHSI (48 k! A em- i demonstration test) { W? '
-Q_
D. (,,**"*'*"***"*~* grasp stress patterns throughout the pipe inside surfaces. Fig. 4 Assembly of IH$l
- 4. Results Shown in Fig. 4 is the setup of IHST used in the ex- 4.1 Temperature distributies periment. With an induction coil placed on the outside Some examples of temperature distribution in the axial of a test pipe, the inside of the pipe was cooled by con- direction during induction heating are given in F%. 5 tinuously running or spraying water. The coil was and Fig. 7-(a). Each sample tested indicated almost the powered for a predetermined time. The test matrix used same temperature rise at the induction coil center while is shown in Table 2. Table 2-(a) gives the test matrix for comparative:y abrupt down curves were noted at coil a parametric survey which was made with various fre- ends.
quencies, heating durations, and coil widths to determine The effect of heating depth was examined by changing induction heating conditions. Table 2-(b) lists the data frequency, the results of which are given in Table 2-(e). of a demonstration test on test pipejoints similar to those A comparison between R-12 and R-15 indicated a higher of actual plant pipes. frequency could produce a larger difference between After induction heating, residual stresses on the joint inner and outer pipe surface temperatures. Another weldments were measured by the commonly used strain comparison between R-21 and R-22 for the effect of gage method. In addition, the test pipes were submerged cooling water flow rate on pipe temperature indicated in 42% magnesium chloride solvent and then inspected that a higher velocity could reduce the temperature for cracking by dye penetrant technique in order to at inner and outer pipe surface more greatly.
.e 6 e. ,. c6 v 4 ,w w ic.i e.4 e.n s, :6 .e.n.c.
a a ies.e..6 ,
.na e
,,.)
* % 'f hdicm eed M ** .w,,,,,,,,, l'5 j .* f
,$. 3*d'Ch8" C#4
'
- P t ! -
- i ; et pe n-m
. g.33 .:R 15:4R S 20) ,i f, , ,,,p f p,. .:k-is,toB s so) h- . :Ceic.b a.d e ,te
- I = --- : 0.s., swe .c. F h.
- l Cw., i.e.ce . : M sw.4 r 4a l p 4a - t usi-g,,,,,
~*i"*M*'*
, ... inn., g,,,4.g.
3a -
, s t p .
) 2nn-j
"- $--b: } 3sa- - A j
bD,v.m. $ be.u c.d5c sh we eu ais t- 49 to
' . se eis .e as &In,5$ 1&isi,NS2b r . ;
$D codc.m ?
+ ) De
- c. 6 cod c.* )
g 10-
. [I t< . o ~. 6
* '*~ ~3 h.. W.acek Dwec,.. .,, .,)3r
}} +
~
g} .[ b a j= N in, f { .f -3 o
- b"**c*.I** w 3 5 -,"~ ry=
gi . ; .
.n. -i- 7. w" -m y l-g - s, -
.l+ ]
_t lg - >~ y +--+ , w
, g Fig. 5 TTD red stress distributions after application of lH51(parametric survey) 3
Table 2 Test matrix (a) Parametric survey g'** l i IllSI conditions l Temperature (*C)* I I l__ l Residual stress" Test , M size Frequency i Coil width Heating time Inner surface JT i (kgflmme; P'P' Cooling method l Outside surface j (kitz) l (mm) (s) ( T.) (Td ( T.- Td l R-12 4 8 S/80 8.4 60 19 Spray method 530 105 l 425 - 12.3 l {
-13 do. 2 do. 16 do. 676 250 426 -12.9 l l
-14 do. do do. 102 do. 520 t 192 328 l - 12.3
-15 do. do. do. 21 do. 562 240 ; 322 l - 16.6
-16 i do. do. do. 23 do. 322 ,
167 l 155 l - 7. 6
-17 do. do, do. 25 do. 247 115 132 I - 7.6
-18 103 S/80 do.
i do. 29 do. 519 l 118 f 401 l - 3.5
-21 128 S/100 3 250 180 Running through method 537 102 435 { - 17.3 f (0.72 m/s) g
-22 do. 3 j 250 157 Running through method 548 126 422 - 17.5 (0 43 m/s) I
-23 do. 3 200 125 do. 558 f 122 ; 436 - 14.7 (b) Demonstration test u
Item IllSt c nditions Temperature (*C)* l l Test Pipe size ' --{~q j~ ' lleating time Outside surface JT P'P' Ning metM (kili) l S(mm) Iw dth (s) (T.) Inner (Ts) surface (T. - Ts) ll R- 1 B"* l 4B S!80 8.4 60 60 l Spray method 540 168 372 I
-2"" do. 8.4 60 60 f do. 536 170 366
--4 A * *
- do. - - - - - - -
-2 5 * "
- 12B S!!00 - - - - - - -
-26"
- l do. -
-28"" do. 3 250 177 Running through method 514 119 395 (0.5 m/s)
-29"
- do. 3 250 178 do. 514 128 386
-30"" ,
do. 3 250 116 do. 556 79 477
-3 3 * "
- i do. - - -
}
(Note)
- Temperature measurements taken at induction coil center 5
" Average measurements of stress in the axial direction on inner pipe surface taken at coit T center I
"* Test pipes for crack testing 2
**" Test pipes for residual stress measurements J
z
Vol11 Wo. 4 Octobor 1978 (a) Maaimum heating temperature distribution i I ; R 23 (II)R 30
'? . . _ 2 ',
*"'" % f ,,,a _ 60 190
-.M.-.*'
~~
! I a.! Induct,oa co,il q" r J
.l -
y l r-- .-;J p $
=-
I $N
^
e
.g ' 2,.! a Ij o,,,
woa,,,co g' / L} g 12H pro gae p.pe c, 1 {W"e%* g Nor:4 nefe end si,-
)f i en,- I 12B Wre'9W P'Po h
{ 4<si- 'l 74e % g ,- Ilacer surface 5
}3.- l/ g j _w y =e
'. C ..&,
j lh aw a, w w .u o &Hene ao tw6 # 14 > j su eu 4u & v ;:u eu eu au juu &&140 Iwlee)A, D.sonce l'om co 1 cow mm). D.seawe from meldment cerwer (msm) (b) Residual stress distribution on inner pipe surface ( I R-2$,R-28 ( H )R 33,R-30 p, ,,,, tod laducte (*I e) 190 t a.y onw
. . c ,c,,on co.1 (sor R-roaty t*'
;Norale sole end
~,
./
!.L. ,4 u ."**
3 p,. _1,,,
, JW Norfe safe ead j - 1 w i
_ kto ~i_l IC
- r Tf j,," "lNl ,e o .-
r -,
-(!)Cirewmf.erent d oct.
e g l 30.(I)Cirev=ferential drecten l \p * *g l As weid R-?S' f w As weld R-33) 3e o .--! I s->- % ,i / e 1 2,, # , 4 j
}
i - WA IHS! R lu 3 lWe (H51 R 2n g ' 24- ** ~ 40 1 i
;o o ;o
;y , ;y ,, D,voace from weto=ent cerwer ' mm)
Drpance trom we< omens cearer' em . f, ('g) g,;,gg,,,,,,
- y. ?~'N. 3-
,(D) Anal drecten\ l l [' # g,
's l As me d M'- N
[ lu "E $ i ,f
.,-~%
E l' l As wWd 'R 33) I e cc e he we e,* ce.nor 'mmi l' '?
. '.~,,,,.o D2,n,, ce. - l' 'W*' ,
i 3 i l , I
, l e W.m 1d3]i a y l j We IHSI',R 30
_y. ! E_y. ! ' l Fig. 7 TTD and stress distributions After application of IH51(12B demonstration test)
@"** 4. -e d w* es.
9, e , as
.e, . l I I, I
i i
?---. '
9-- . - - - -
; ~ .
,,, _-_ -#4- :- o,e-wm, -
i
. ..c , ,.m- . . _ -- w._mc.or -.
7 , s e
.w p Q. w ..
i e6,M-,.e, . . , 180s R - 2 t , '_
) _ - _..-
Fig. 8 Crack pattern of inner surface of pipe joint after 4P/o MgCl2 test (demonstration test) 3
lHI Engineering Review 4.2 Residual stress 4.2.1 Residual stress measurement with strain gage '- # -***""~ Shown in fig. 3 are examples of the distribution of residual stresses in the axial direction on pipes 48,10B " and 128 after induction heating (parametric survey). l l' rom the figure it will be seen that in each case residual stresses on the inner surface, down to the coil end, are l mostly in the compressive side. Some examples of the results of a demonstration test [^ _,,,,,,,,,, s= ffy,,,,,,,,,, j on pipes 4B and 12B which were provided with welded 8 c'-' a' P a' l joints are gisen in Fig. 6 and Fig. 7--(b), respectively. In . J , ,,,,,,,,u, both cases, residual stresses on the inner pipe surface "' 4" " N e" m ! were shifted toward the compressive side when induc- 5"*"""""' I tion heat treatment was performed,in comparison with Fig. 9 Residual stress distribution in the plate thickness l direction (calculated results) l 4.2.2 MgC/2cracking test away from an induction coil. fig. 8 shows the results of the observation ofinner pipe fig. 9 shows the calculated results of residual stresses surface cracking induced by MgCl2. fig. 8-(a) compares in the plate thickness direction. The residual stresses in crack patterns of two 4B pipes; one was induction heat the compressive side amounted to about 60% of the treated and the other was not, and Fig. 8-(b) compares plate thickness. This property illustrates the superiority the crack patterns of 128 pipes. Induction heat treating of IHSI to other stress improvement methods, for in-was proved, in both cases, to be effective in preventing stance, shot peening, by which only at most several such cracks. hundred microns of a surface become residual stresses of the compressive sidem.
- 5. Discussion 5.2 Heating factors and residual stress 5.1 Comparison with calculated results by finite element Heating factors, which are considered comparatively method important, and their effects on residual stress will be Thermal clastoplastic calculations by the finite element discussed here.
method were conducted using R-21 as a model. The re- 5.2.1 Effect of hearing temperature sults of the calculations are compared with the experi- Thermal stress produced during heating is proportional mental results in Fig. 5-(c). The calculated results of to temperature difference (JT) between inner and outer temperature distribution in the axial direction agreed pipe surfaces (equation (1)). Accordingly, it is assumed quite well qualitatisely with the experimental results, that as JTincreases, plastic strain produced during heat-whereas some quantitative difference was observed ing and resulting residual stress will be increased. The between the inner and outer surfaces. One possible effect of JT was experimentally examined, some results reason for the difference is that the calculations regarded of which are shown in Fig 5-(a). R-15 showed a residual heating intensity at each mesh in the pipe thickness direc- stress (an ) of -17 kgf/mma at JT=322=C and R 17 tion as constant. But, in the experimental induction showed -8 kgfjmm2 at JT=132'C, 'Shich agreed well heating, heating intensity was reduced from outer to with the above assumption. This residual stress -JT inner surfaces forming a curve like an exponential func- relationship, however, holds true only when residual tion (equation (2)). As a result, a milder-than-actual stress (an ) is equal to, or less than the yield stress of temperature cune between the outer and inner surfaces material (a,). Under a n >a,, work hardening is the only was obtained for the model calculations. In addition, factor to increase an. The effect of the heating tempera-slightly lower measurements are anticipated in the ex- ture, therefore, cannot be fully relied on to increase periment because the thermocouple acted as a cooling residual stress beyond a,. fin. 5.2.2 Effect ofheating depth As to residual stress distribution, the calculated results The depth of heating by induction heating is imersely agreed sery well for this type of model calculation with proportional to the square root of current frequency. the experimental results, especially on inner pipe surfaces At extremely high frequencies, therefore, only an outer where complete agreement was noted. This indicates surface is, in effect, heated, and in an equilibrium con-that this calculatise method will provide a good means dition, a linear pattern of temperature distribution as of predicting residual stresses on inner pipe surfaces illustrated in Fig.10-(a)is obtained. When an extremely with acceptably high accuracy in the future. The residual short heating period is used, a temperature distribution Stresses on the inner pipe surfaces were found to be pattern will be as shown in Fig.10-(b). In contrast, at almost the same in the range from the coil center to lower frequencies, the heating best penetrates deeply, and 25 mm outside the coil ends. This implies that the residual in an equilibrium condition, a temperature distribution stress improsement is effective esen at weldments slightly forms a pattern like a parabola, as seen in Fig.10-(c). In 6
,_______%-___%.__------ --w-Vol.11 No. 4 October 1978 j** ,,,,, i.t ..
'" ,ty"g'>o. evreP d'.' ow<
we - very similar to that of induction heating intensity dis-
- d. , +.e.o. , ,,w OWh Ws47 dhymm6MMMbud-T '
sirable, in terms of residual stress, for the reasons stated [ i [ T. 1 ( r. i t T. ' above. It is not wise to use an extremely limited heating time, w .e ai+ s-a s~n t*" The relations between heating time and temperature
* * * ' * '*" 5"" '""
distribution in the plate thickness direction obtained Fig.10 Three patterns of temperature distribution in the by IHST are shown in Figs. Il-(a) .tnd (b). The calculated plate thxkness direction results showed that a temperature distribution pattern similar to equilibrium condition was obtained in about such temperature distribution conditions as stated above, 80 seconds on a 21.4 mm thick pipe. A similar pattern thermal stresses (a,4) on inner pipe surfaces can be ex- w s experimentally measured in about 180 seconds for pressed by clastic calculations as followsm: a pipe with 34.5 mm plate thickness. This exemplifies (1) In the case of Fig.10-(a): that heating time depends on plate thickness. Simplify-ing the relationship by using Fourier's numberW as a,4=Ea JT/2(1 -9) y ft: (2) In the case of Fig.10-(b): where, a: Thermal conductivity (m2/h) Assuming that the temperature distribution takes t: Heating time (h) a pattern of a symmetric parabola with respect to an inner pipe surface' L: Plate thickness (m) gives 0.78 for the 21.4 mm thick plate and 0.66 for the a, = Ea JT/3(1->) 34.5 mm thick plate. In addition, it is known that a 2L (3) In the case of Fig.10-(c): thick plate with a uniform temperature of po at t<0, Assuming that the temperature distribution takes when cooled quickly at t=0, till its surface temperature a pattern of a symmetric parabola with respect to becomes O C,is cooled down to about 20% of the initial an outer pipe surface, temperature do as measured at its core, providing that F=0.7. The above data lead to the conclusion that a a,n=2Ea JT/3(1 -9) temperature distribution pattern similar to the equili-Thus, a,a of condition (3)is assumed to be the largest brium condition can be obtained by selecting an ap-and capable of producing the largest plastic strain for propriate heating time for making Fourier's number 0.7 the same JT value. Hence, the resultant residual stress or higher, will be the largest. In contrast, the resultant residual 5.2.4 Effect ofplate thickness stress of condition (2)is assumed to be the smallest. The The effect of plate thickness on temperature distribu-above assumption can be verified by comparing R-12 tion is discussed here first. Assuming that a coil with a with R-15 again. The latter having a larger heating depth large width in relation to a plate thickness is heated till produced a slightly larger residual stress than the former an equilibrium temperature distribution is obtained; which was subjected to a greater JT value but had a then, in the neighborhood of the coil center, the pro-smaller heating depth. This agrees well with the assump- blem is reduced only to that of heat conduction in the tion. radial direction, which can be expressed by a simple 5.2.3 Effect ofheating time equation. A basic equation for the problem is: When heating is applied for a very short period, the 320 1 temperature distribution takes a pattern (Fig.10-(b)) um0 ax2 ,
- r. L*.4 ' ,s l n ,,,,,,,, ,,,j,,,
- l \e, '"
p.
'li ,,,,
p, I,~
+v ..
T j ...
*1,- t w \., ,,, ! _
.l .,. E 1 3,<
l 6 .
' -lN l,s ou j
C _. ____ l a d' l g7i r g . ,. 5 W
' o's * < 6 en 1 4me. ,',. es 4.=pu
- U %i,is, Fig. 1I Temperature distribution in the plate thickness direction 7
IHi Engineerir.g Review where, 0: Temperature at position x which is to be discussed later. J,: Thermal conductisity of a pipe On the other hand, in the case of R-28 and R-30, w: lleating density where a coil of sufficient width was used, a greater differ-x: Distance from outer surface ential temperature between the outer and inner surfaces licating density (tr) by induction heating can be given "".d a better improsement in residual stress were ob-by the following equation *: tained for the latter sample which had the larger plate thick ness (Fig. 7). These results agreed well with the above w = w,e - 2< 8 . . . . . .. ..
. . . .. . . .( 2) assumption.
w here, w,: lleating density on outer surface 5.2.5 Effect of cooling waterflow rare As seen from equation (4), the temperature difference S=y- [p (cm) in the plate thickness direction increases proportionally with heatmg intensity and plate thickness regardless of p: Specific magnetic permeability cooling conditions on the inner surface. Ilowever, in f: Frequency (cycle /s) heating pipes,it is generally suggested that pipe tempera-p: Specific resistance (O cm) ture be held within the safe limits to prevent pipe material When the outer surface is insulated and temperature deterioration. For this reason cooling conditions come is T,. equation (2) leads to into the picture. An equation for heat flux gi on inner pipe surface is derived from Fourier's law as 0 = T, !! - -1+ x + c - 2rs . . ..(3) r de q , = - ) , g . . . . . . . . . . . . . . . . . . . . . . . . . . (5) Substituting x=L in equation (3) yields the value of temperature (T,) on the inner pipe surface. Hence, An ther heat flux 42 from inner pipe surface to a liquid temperature difference (JT=T,-T,) between the outer can be expressed by and inner pipe surfaces is derived from q2 = h. J T2 - . . . . . . . . . . . . . . . . .
..(6) 2 where, h.: Average heat conduction ratio of J T= T.- T,= w,S2 /4 (-l +yL+c-2us ...(4) g;g When two plates of different thickness are heated with JT2: Temperature difference between in.
the same heat input under the conditions of L> Sand JT side pipe wall and liquid is proportional to L, the plate with a larger thickness will Differentiating equation (3) and putting x=L gives produce a larger JT. In addition, the resultant thermal de ~ w,S stress is proportional to JT. Therefore, it is generally ~
I7) assumed that the thicker plate gives the better results.
. dx m " 22,II~#~ u )
Fig. 5-(b), when reviewed from the abose standpoint, As 4 =42 equations (5) to (7) give indicated pipe 10B having a larger temperature difference "*- than pipe 4B but the resultant residual stress was smaller J T2 = 7i- T. . ..(8) 2h. (1 -e-203) .. compared with that of 4B which had a smaller plate where T. is the temperature of the liquid. thicknets. This implies a considerable effect of coil width, From equations (8) and (4)it follows
,.,m e.,
um, c o.. .e w ,,. w., 6.,S. e 6*
+ r . 4..,w
,,, % ,,,, w cu sw t .. - 6.15 no' he.1/.N k,%,
5-MWg . - . - 5* 1 l
$ sut
?.. '=t::'
- - w$ Eo 7
Q p$ I E @4 N. 6 1 1 ! : i
- 1c.) +
* 'l j :no %/d . N -
so =, j 3o- ';;; ,c, W .c p so 7, 3T as /x . 4.f. : e.i. w ie
- 1 j
I la _1 is 3 i a 43. T 4
~{ ! 2a y-' [g 7 N, j j J~o '2 s 6 15 i 3 1 7.4s 3I yI sn 2L_
Q3ys Ja _u Q n us m , o is a o is ze e %owt %.) re. m.o oeewat (Note) Water temperature: 35'C S: Heating depth (=9.6 mm) Fig.12 Effect of flow rate on temperature distribution 8
F-~' Vol.11 No. 4 October 1978 JT, + JTa = T,- T,, = w.S S v 3 - l + 2L3 + e *,) j _ l
\ y M '*
Qn_,-my...........(9) h,,,is given by Dittus-Boelter equationm } Nu b "d ,9.9y3g,,,p,,, - k }' 4 y h., = 0.023 d*2 b8 "" "*""""" g_
,_-r,..
,w where, J ,: Thermal conductivity of liquid .--~,
d: Inside diameter of pipe pig.13 Stress distribution of pipe subjected to step tem-v: Coefficient of kinematic viscosity perature change in axial direction P,: Prandtl number U.: Flow rate ofliquid But clastic calculations can be substituted as a simple way to obtam a guide value; the effect of step temperature in Figs.12-(a) and (b) are shown the results of calcu. change in the axial direction, as shown in Fig.13-(b), can lations using equations (9) and (10) for the effect of flow be expressed as the effect obtained when circular pipes - rate at constant heating intensity and at constant outside with different radii by JR are connected together by pipe wall temperature, respectisely. In the former case, applying shearing force (Q) as shown in Fig.13-(r). the temperature of both the inside and outside pipe walls The effect of shearing force (Q) increases to a maximum dropped as the cooling water flow rate was increased, but at a certain distance from the pipe end, and then dimi-the temperature difference was maintained constant nishes as illustrated in Fig. /3-(a)*. The force exhibits (Fig.12-(a)). In other words, the flow rete, when in- a relatively large effect within the range from the pipe creased, did not change the resultant residual stress but end of 1.5s/Rt where R is the radius from pipe wall did decrease the outer surface temperature, gising a thickness center, and I is plate thickness. Therefore the - margin of safety, within which the temperature will effect at coil ends can reduce when a coil width is not cause the material deterioration in the latter case, 2 x 1.5s'Rt or larger. R-22 and R-23 were tested using a greater temperature difference between the inside and different coil widths, more than 3.0v'Rt in both cases, outside pipe walls was obtained as the flow rate was and showed little difference. But R-18, which was heated increased (Fig.12-(b)), under such conditions a great by a coil of about 1.4 JRt width, produced small residual stress could be expected. In summary, the higher compressive residual stresses. the flow rate the better residual stress improvement is 5.3 Comparison with other stress improvement methods anticipated. ~ 5.3.1 Stress relief annealing treatment The experiment on R-21 and R-22 was made by chang. To relieve residual stress by this method generally ing the flow rate at a constant heating intensity. The requires heating up to about the temperature at which results indicated almost the same tendency as shown in the heated material starts recrystallization. When applied Fig.12-(a). The fact that R-22 was subjected to thermal to 304 type stainless steel this method induces material esposure for a period shorter by more than 10% than sensitiration during heating, sustaining, and cooling R 21 was presumably the cause for the small difference cycles, making the materials less IGSCC resistant. To in the outside wall temperatures of the two samples. remove such sensitiration and reliese residual stresses Naturally, almost the same residual stress resulted be-the material must be heat treated in a solution treat-cause the differential temperature on both samples ment at a higher temperature. It is quite difficult to was nearly the same. apply this treatment to an installed plant with the 5.JA E//ect of w!/ width present technology. A difTerence in temperature occurs in the axial direc- 5.3.2 Tensile food method tion of pipe as well, because a pipe is heated by a coil A tensile correcting machine is one application of this with a finite width. The stress on the inner pipe surface, method. But this type of machine is inapplicable to just below the coil end, is compressed by this temperature large diameter pipe because an extraordinarily brge change and, as a result, acts to cancel the tensile stress mechanical lead is required. Another possible approach produced by the temperature difference between the i; to provide tensile load by pressurizing an inaer pipe inner and outer pipe surfaces. Therefore, the areas surface. This method. howeser, can cause plastic defor-intended for stress improsement should preferably be a mation of the whole piping system, aad in an already sullicient dntance from the areas where drastic axial installed plant, residual stress can occur unexpected!y direction temperature changes take place. To precisely at other than the pip ng being corrected. Furthermore, empute this distance, thermal clastoplastic calcula- Inis method canrot shift po,itively the residual strew tions by the f. nite element method must be performed, on the inner pipe surface to the compressive side. 9 L . .
IHI Engineering Revi=w 5.3.3 Residual stress relief by ribration on pipe weldments into compressise ones or into This method can reduce peak stress, but is incapable of very small tensile stresses wss confirmed by strain efTecting complete stress relieving or positive introduc- measurements with strain gages and by a cracking tion of compressed residual stress on an inner pipe test in a higCl2 solvent. surface. It cannot be applied to a large-scale construc- 2. The above was also confirmed by calculations by tion, because of the limitation of capacity of a vibrating the finite element method. In addition,it was found machine. Neither it is suitable for installed piping systems. that compressive residual stresses covered about 5.3.4 Shot peening 60% of a pipe wall thickness. This method can impart a large compressed residual 3. The authors suggest the following as factors that stress, but depth of improvement is small and the mate- can improve IHSI efficiencies: rial is subjected to a high degree of plastic deformation. 1. A large temperature difference between inner As a result, surfaces of some materials are embrittled. and outer pipe surfaces is preferable. The inner pipe surface must be cleaned after treatment 2. A large heating depth is preferable. to eliminate foreign material that has been inserted into 3. The heating time should not be too short. the pipe. The difficulty of shotting small diameter pipes 4. A pipe with a thick wall can obtain a larger from inside is one other incidental drawback of shot temperature difference between inner and outer peening. surfaces than a thin-wall pipe, if heating intensity 5.3.5 Advantages ofIllSi is constant. IllSI has the following advantages: 5. A larger temperature difference can be obtained
- 1. Improvements in resistance to SCC and corrosion between inner and outer pipe surfaces with a fatigue strength are achieved through positive high flow rate of cooling water running on the shifting of residual stresses on inner pipe surface inner surface than with a low flow rate, providing to the compressive side. that the heating temperature ofinner pipe face is
- 2. The treatment is readily applicable even to in- constant.
stalled pipings as lor.al heating of pipe exerts little 6. Heating width (coil width) should preferably be influence on other members. 3v/Rt or more.
- 3. Residual stresses of any desired level can be obtained by selecting pioper temperature difference REFERENCES between inner and outer pipe surfaces without fear (1) S. Timoshenko et al.: Theory of Elasticity (2nd of material deterioration. edition) hicGraw-Hill (1951)
- 4. The depth of compressive residual stresses is large. (2) A. Takaku and ht. Tokiwai: Basic Study on Some
- 5. 'Ihe treatment can be applied to large diameter hietallurgical Factors and Surface Treatment pipes without much difficulty. Effects of Stainless Steels for BWR Cooling Pipes,
- 6. The treatment is simple to perform as there is no Central Research Institute of Electric Power Industry, need to insert equipment into the pipe interior. Japan, Reports No. 275038 (1976)
(3) S. Yonetani: Occurrence and Countermeasures
- 6. Conclusion of Residual Stress, Yokendo, Japan (1975) p.192 Stress corrosion cracking occurs as a result ofinteractions (4) Y. Katsuto: Outline of Heat Transfer, Yokendo, of three factors-stress, environment, and material. Japan (1964) p. 34 An experimental study was made of residual stress im- (5) The Japan Society of hiechanical Engineers: A provement by induction heating with the objective of Handbook of hiechanical Engineering (4th edition) improving SCC resistance of a plant through reducing (1960) the stress factor. The following is a summary of the (6) S. Timoshenko et al.: Theory of Plates and Shells, findings of the study. AlcGraw-Hill (1959) pp. 469-471
- 1. The capability of IHSI to change residual stresses 10}}