ML17256A979
| ML17256A979 | |
| Person / Time | |
|---|---|
| Site: | Ginna  |
| Issue date: | 05/18/1982 |
| From: | ROCHESTER GAS & ELECTRIC CORP. |
| To: | |
| Shared Package | |
| ML17256A978 | List: |
| References | |
| NUDOCS 8206010640 | |
| Download: ML17256A979 (20) | |
Text
Steam Generator Evaluation Ginna Steam Generator Tube Failure Incident January 25, 1982 R. E. Ginna Nuclear Power Plant Docket, No. 50-244 Addendum 2
Laborator Fati ue Testin Revision 0
May 18, 1982
FATIGUE TESTING OF STRUCTURAI LY DEGRADED MODEL 44 STEAM GENERATOR TUBING (7/8" O.D.
x 0.050" WALL)
MAY 1982 0196s:10
Fati ue Testin Pro ram The purpose of this testing program is to determine the fatigue charac-teristics of structurally degraded tubes such as seen in the Ginna-B Steam Generator.
For structurally degraded tubing, the predominant cyclic loading results from fluid interaction (with or without lateral load impacts).
There-fore, the tube response of prime interest is of a high cycle fatigue nature.
However, in light of the time constraint, the actual testing performed was of a low cycle fatigue nature.
An attempt was made to accelerate the fatigue damage by specifying dynamic vibration amplitudes significantly higher than those a structurally degraded tube in the steam generator would experience.
In the following the results of high amplitude low cycle testing will be correlated to analytically predicted tube responses under low amplitude high cycle conditions typical of flow-induced vibration loadings.
Three types of tests were performed in conjunction with the fatigue testing program:
Strain Survey Tests Basic Fatigue Tests Impact Fatigue Tests The strain survey tests were performed to experimentally'etermine the relationship between deflection and stress at various locations for a length of steam generator tubing between=the tube sheet and first sup-port plate with a locally degraded region.
Strain gages were installed at the point of maximum tube deflection and at locations in the degraded region.
The specimen was then vibrated at different amplitudes and the strains measured.
Stress values were calculated from the measured data.
0196s:10
~
~
~
In the basic fati gue tests nominal 0.875 inch 0.D.
x 0.50 inch thick tubes were mechanically degraded locally near one end and set in a tube sheet simulation (collar) at that end and a tube support plate simula-tion at the other.
The tube end conditions approached approximately either a "fixed-pinned" or a "fixed-fixed" situation.
The structurally degraded section was typically 2.0 inches long with its center 4.0 inches from the tube sheet simulated end.
Two damaged configurations were tested:
flattened (to simulate a full collapse) and kidney-shaped (to simulate a partial collapse).
The flattened shape was achieved by clamping the 2 inch section of the tube in a vise, and squeezing the tube.
The kidney-shaped section was one in which half of the tube circumference was made to nestle into the other half.
For the impact fatigue tests a length of tubing between a tube sheet and first support plate simulation with a locally=-degraded region approxi-mately four inches from the tube sheet simulation was vibrated such that it would impact against a fixture causing an impact type load.
To accomplish this a chisel with a rounded tip was positioned near the end of the degraded region so that the specimen would hit the chisel on each vibration cycle.
The chisel was instrumented with a calibrated semi-conductor strain gage to obtain a reactive load measurement.
Test Equipment and Setu P
Figures 1 through 4 show the basic test setup used.
A vibration exciter was attached to the tube with a small hose clamp approximately 12 i nches from the simulated first support plate end of the tube.
Deflections were monitored with a linear variable differen-tial transformer.
The deflection was set by adjusting a micrometer to the specified (single amplitude) deflection.
The tube was then vibrated until it just touched the micrometer.
The LVDT output was then observed and maintained at that amplitude for the duration of the test.
An instrumentation block diagram for the basic test setup is shown in Figure 5.
0196s:10
Two different support configurations were used for these tests, a
fixed-fixed and a fixed-pinned condition.
The faces of the support blocks were 51.8 inches apart for the fixed-fixed configuration with the degraded portion of the tube centered four inches from one face.
In the fixed-pinned case the distance between the support fixtures was 57.8 inches.
The fixed support consisted of a four inch long block through which a
0.89 inch diameter hole was drilled to accept the tube (test specimen).
The first two inches of the tube were rolled into the block.
The pinned support restraint consisted of a.030 inch thick stainless steel sheet through which a hole was dri lied to accept the tube, which was epoxied in place.
After these fixtures were attached to the test
- specimen, the assembly was bolted to a strong back; an 8 x 8 inch wide flange beam.
In order to expedite testing, the strong back for the start of the first test consisted of 1 x 6 inch steel bar which was bolted to a bedplate.
The tubes were oriented so the axis of the minimum cross section of inertia of the degraded region was parallel to the strong back.
The vibration exciter was attached to the tube approximately 12 inches from the end of the tube farthest from the degraded region.
The direction of vibration was vertical.
In the cases where a constant axial load was applied*, the tube was instrumented with four axial strain gages at 90'ncrements around the tribe, six inches from the simulated first support plate support.
The If tubes in the steam generator are assumed to be axially restrained at the first tube support, an axial tensile load of the order of 1000 lb. exists on a plugged tube in a hot leg wedge area due to tube-to-shell and tube-to-tube thermal interactions during steady-state power operation.
0196s:10
required strain for a 1000 lb. axial load was calculated from the spe-cimen geometry for a modulus of elasticity of 31.7 x 10 psi.
The axial loads were applied with a screw arrangement prior to tightening the second mounting block.
The tube strain was monitored during loading and tightening and was periodically checked during the test.
Specidl fixturing was developed to impose impact loads on the tube in the degraded region.
The impact fixturing was installed on the strong back.
The two major problems in the fixturing design were the need to provide a hard impact surface which would not deform significantly during the course of the test and to provide a means of measuring the impact forces.
In order to avoid time consuming machining and heat I
- treating, a commercially available cold chisel was used to provide the impacting surface.
The sharp end was ground to a blunt surface approxi-mately 30 mils wide.
This chisel was then instrumented with four semi-conductor strain gages.
This type of instrumentation provides very high frequency responses with a minimum of extraneous mass effects.
After the strain gages were installed and pri or to actual impact
- testing, the chisel was mounted in a universal testing machine and loaded against a piece of inconel tubing.
Load vs strain curves were obtained with the chisel centered on the tube, as well as with it offset by approximately
.050 inches, to obtain an indication of the effects of off center loading.
The variation was found to be less than 5 percent.
Three factors contributed to the impact loads obtainable in the test apparatus:
the tube velocity at impact, the mass and stiffness charac-teristics of the tube, and the mass and stiffness characteristics of the backing structure for the impactor or chisel.
It was originallly hoped that an impact load magnitude of approximately 450 lb. could be obtained.
After considerable effort to stiffen and increase the mass of the backing structure a load of 186 lb. was a'chieved.
It is doubtful that further efforts would have made a significant difference.
0196s:10
Strain Survey Tests In the strain survey tests strain gages were mounted at the locations shown in Figure 6.
The strain survey results for a.3 inch peak to peak deflection are found in Table 1.
As can be seen from these results the maximum recorded stress was 31,200 psi tension and the maximum stress difference range was 46,800 psi (fixed-fixed flat degradation with 1000 lb. axial tension).
Also the fixed-fixed configuration results in substantially higher stresses than the fixed-pinned configuration for a given vibration amplitude.
Basic Fati ue Tests Four basic fatigue tests were performed and are discussed below.
A fixed-pinned beam with a 2-inch flat degradation was tested with no axial load.
The natural frequency was 39.3 Hz.
The fatigue testing was performed for 300,000 cycles at peak-to-peak dynamic amplitude of
.05 inch (.05 DA),.07 DA,.09 OA,.12 DA,.16 DA,.20 DA,.25 DA,
.30 DA,.40 OA and 3 x 10 cycles at.50 DA or a total of 6.3 x 10 cycles.
There was no failure.
A fixed-fixed beam with a 2-inch flat deformation was tested with a 1000 lb. axial load.
The natural frequency was 63.6 Hz.
The fatigue test was run for 200,000 cycles at
.05 OA and 2.0 x 10 cycles for.'3 DA.
No failure of the specimen was observed.
A fixed-fixed beam with a 2-inch long flat deformation, with 0.005 inch deep by 1/4 inch long notches at 2 locations within the degraded sec-tion, was tested with a 1000 lb. axial load.
The specimen was'vibrated at.30 DA and failed after 900,000 cycles.
Failure did not occur at a
notch and this may have been due to the fact that the loading blocks used to deform the tube may have weakened the tube at a non-notched location.
0196s:10
A fixed-fixed beam with a 1 inch long kidney-shaped degradation was tested with a 1000 lb. axial load and vibrated at
.30 DA.
No failure occurred after 6 x 10 cycles.
Results of the basic fatigue tests are summarized in Table 2.
Im act Fati ue Tests The impacting test was performed to simulate the effect of an impact force imposed near a degraded region of a tube.
This was done to simu-late the effect of a foreign object striking a degraded steam generator tube.
In the first test a 2-i nch long kidney-shaped degraded region was selected and a 1000 lb. axial load applied.
The test specimen was vibrated at.30 DA and impact loads of 75 lb., 132 lb. and 186 lb.
applied for 1.8 x 10
- cycles, 1.1 x 10
- cycles, and 1.1 x 10 cycles, respectively.
No failure occurred.
In the second test a 2 inch long flat degraded area was selected and a
1000 lb. axial load applied.
The test specimen-was vibrated at
.30 DA and an impact load of 150 lb. applied for 3 x 10 cycles.
No failure occurred.
Results of the impact fatigue tests are summarized in Table 2.
Conclusions The following conclusions are applicable to the above tests:
o Structurally degraded
- tubing, when subjected to significantly large vibration amplitudes, are not expected to fai l due to low cycle fatigue.
o Based on the results of the strain survey, the worst test configuration corresponds to the fixed-fixed tube condition with a 2-inch flat degradation and loaded axially with 1000 lb. tension.
0196s:10
As applicable to the response of steam generator. tubing under operating, conditions, the following is to be noted.
The proposed ASIDE high cycle fatigue curve extended to N = 10 cycles for Inconel 600 based on E = 28.3 x 10 psi and for temperatures not 6
exceeding 800 F shows an endurance limit of S
= 13.7 ksi.
With due corrections for the room temperature modulus of elasticity of 31.7 x 10 psi, 'Se 14.9 ksi.
In the tests, the maximum stress range based on the strain surveys was 46.85 ks i (alternating stress intensity of 46.85/2
= 23.43 ksi) at a peak-to-peak amplitude of 300 mils.
This occurred for the fixed-fixed, 2-inch flat degraded configuration with 1000 lb. axial load.
Since the stresses for a given configuration are proportional to the applied vibration amplitudes, the peak-to-peak amplitude for the worst case configuration above, corresponding to the high cycle fatigue endurance limit is DA =
= 0.19 inch 14.9 x 0.'3 This DA is significantly higher than an analytically predicted upper bound amplitude of a structurally degraded tube in the steam generator.
Based on the results of the room temperature testing of mechanically degraded tubes, it can be concluded that fatigue failure of structurally degraded steam generator tubing is unlikely under the influence of flow-induced vibration loading alone.
Additionally, failures will not be expected as a result of fatigue due to a lateral impact load acting locally at one location.
However, the potential for tube cracki ng and ultimate severence may exist due to progressive notching as a result of repeated impacting at the same location as evidenced by the examination of tube specimens following the impact testing.
Furthermore, if the impact location was continually and randomly changed, tube cracking and ultimate failure could result due to the combined effects of notching and fatigue damage.
0196s:10
TABLE 1 STRAIN SURVEY RESULTS FOR.3 INCH PEAK TO PEAK DEFLECTION Test
'Cfi Axial Load Resonant
~Fre uency
~Ga e 1*
Gacae 2
~Ga e 3
~Ga e 4
~Ga e
5 Gacae 6
Fixed-Fixed Flat Degradation, 2" long 0 lb.
Nin Stress Psi
-8400
-17400
-8250
-3900
-17400
-22200 57.9 Hz Max Stress Psi
+8400
+17400
+8250
+3900
+17400
+22200 Fixed-Fixed Flat Degradation, 2" long Nin Stress Psi
-2700
-11100 0
+3900
-9600
-15600 1000 lb.
63.6 Hz Max Stress Psi
+18300
+26700
+15600
+11700
+25200
+31200 Fi xed-Pinned Flat Degradation, 2" long 0 lb.
39.3 Hz Max Stress Psi
+3300
+8100
+3400
+2550
+7200 Min Stress Psi
-3300
-8100
-3400
-2550
'-7200 0 lb.
Fixed-Fi xed Kidney-Shaped Degradation, 1" long Nin Stress Psi
-9000
-17550
-12750
-15390
-8250
-12750 59.0 Hz Max Stress Psi
+9000
+17550
+12750
+15390
+8250
+12750 Fixed-Fixed Kidney-Shaped Degradation, 1" long 1000 1 b.
67.7 Hz Max Stress Psi
+17550
+24300
+19800
+22800
+16440
+20550 Nin Stress Psi
-1950
-8700
-4200
-7200
-840
-4950 Refer to Figure 6 for gage locations.
0196s:10
Test Degradation No.
Shape Confi uration TABLE 2 FATIGUE TEST
SUMMARY
Axial Tension Amplitude, Inch lbs.
(Peak to Peak)
Frequency Hz Cycles Tested'esults 1
Flat 2" long Fixed-Pinned 0.
.05
.07
.09
.12
.16
.20
.25
.30
.40 39.3
.3 x 106 (each level)
No failure Flat 2" long Flat 2" long Fixed-Fixed Fixed-Fixed, Notched at transition 1000.
1000.
.50
.05
.30
.30 63.6 63.6 3.6 x 106
.2 x 106
.9 x 106 No failure Failed not at notch Kidney-Shaped Fixed-Fixed 1" long Kidney-Shaped Fixed-Fixed 2" long 1000.
1000.
.30
.30
=-67.7 66.6 6 x 106 1.8 x 106 with 75 lb.
impact force 1.1 x 106 with 132 lb.
impact force 1.1 x 106 with 186 lb.
impact force No failure No failure*
Flat 2" long Fixed-Fixed 1000.
.30 63.6 3.0 x 106 with 150 lb.
,impact force No failure*
- Although no failure occurred, tubes were notched about 25 to 30 mils deep at the location of the impactor.
0196s:10
1:
VIEN OF FATIGUE.TEST SETUP
'r v
\\ ~
I
\\
j pzcgzz 2:
NOTHER CLOSE UP OF FATIGUE SETUP
- 4. 0-28
FIGURE 3:
CLOSE VIEW OF FATIGUE TEST SETUP ur f
'FIGURE 4:
ANOTHER CX'OSE UP OF FATIGUE SETUP 4.0-29
Accelerometer Endevco
'odel 2220C Accelerometer Endevco Model 2220C LVDT
+150 MU RANGE Amplifier Kistler Model 504 Amplifier CEC Model 1-127 Amplifier Kistler Model 504 Filter Ithaco Model 4302 10 lb Vibration Exciter Power Anplitude Filter Ithaco Model 4302 Filter Ithaco Model 4302 Oscilloscope Phillips Model 3234 Variable Oscillator H.P. 5302 Frequency Counter INSTRUMENTATION BLOCK DIAGRAM
I 22 FIXED-FIXED CONFIGURATION TUBE ROLLED IN 1
2 FIXED-PINNED CONFIGURATION i'C sr~i 2gll
(
2 STRAIN GAGE IDENTIFICATION FLAT DEGRADATION FIGURE 6 STRAIN SURVEY-TEST CONFIGURATION KIDNEY-SEIAPED DEGRADATION