ML17261A814

Beam Spread Correction.

ML17261A814
Person / Time
Site:	Ginna
Issue date:	03/15/1979
From:	Mcgaughey W SOUTHWEST RESEARCH INSTITUTE
To:
Shared Package
ML17261A806	List: ML17261A805 ML17261A813 ML17261A814 ML17309A414 ML17309A416 ML17309A417
References
NUDOCS 8901130200
Download: ML17261A814 (18)

Text

BEAN SPREAD CORRECTXON

'P. C. NcGaughey Southwest Research institute San Antonio, Tezas 78284 Beam spread measurements 0'73 have been a Section XI requirement since the 197 Summer Addenda, Paragraph T-4460, became effective.Section XI 1977, para-graph XVA-2232(a) refers to Section V, Article 4 for ultrasonic examination rules. The latter Paragraph T-431.3 requires beam spread measurements at intervals no greater than 3 months. The beam spread measurement technique is in the nonmandatory appendiz as Paragraph B-60, so other techniques may be used.

iVo use of beam spread measurements is specified in the Code. Unoffic'al but g enerally n recognized justification for continuing the requirement include (a) identification of beam variables to assist in selection of a search unit for a reexamination at a later date, (b) to have the informat'on avail-able when it is determined how to perform beam. spread correction of indica-tion dimensions.

During the February 1979 Tnservice Ezamination of the Rochester Gas ana Electric Company's Robert E. Ginna plant, a complete mechanizea ezamina-cion of the reactor pressure vessel weld was accomplished. These examina-tions disclosed the presence of several ultrasonic reflectors due to imperfections in the weld or associated base metal. :!ost of these reflectors were readily determined to be within the AS'ode acceptance standards.

Therefore, further evaluation of these reflectors was not necessary and they were recorded for future rererence. One indication in the nozzle-to-shell weld of 'Bozzle N2/4ras of an amplitude requiring further evaluation of its significance. I&order to perform a fracture mechanics evaluation of the significance of this reflector, the true size of the flaw was required.

A process of evaluation of the ezamination data was undertaken in an attempt to determine the size and nature of this reflector. Also, controlled ezperi-

~ents were performed to conzixm the accuracy or the theoretical calculat'ons and considerations.

Recorded mechanized ultrasonic examinations permit reading indicat- ons'0% .

DAC ana greater. The 0', 45'nd 60'xaminations from the inside surface of the vessel as well as the 45'xamination from the nozzle bore show no record of the indication detected with the 15'ngle beam longituainal wave from the nozzle bore. The plane of maximum sensitivity is the same for the 15'ozzle bore and the radiographic examinations. This is supported by the fact that no ultrasonic response from the imperrection was obtained with the 0', 45'nd 60'xaminations performed on the vessel

. wall ox from the 45'xamination also performed from the nozzle bore. These observations indicate that. the reflector is directional and is oriented essentially perpendicular to the 15 longitudinal wave sound beam. Tf the reflector was a rough and faceted crack, ultxasonic inaications would have been obtained from the other ezamination beams which also interrogated this area. Xf the reflector were more globular, or rounded, responses to these other examination angles would also have been noted. Based on ezperience n vessel fabricator shops, this type of thin planar slag has been noted on other occasions. Furthexmore, a xeview of the fabrication raaiographs 890k l30200 88i227 PDR ADOCK 05000244 6 PDR

of this weld show the presence of entrapped planar slag. This was confirmed by several Level III reviewers of the radiographs.

Therefore, it was concluded, that the reflector is thin, smooth, and in a plane perpendicular to the 15'eam. The writer nas witnessed the excava-tion of a similar indication in a fabrication shop. The shop-excavated indication was not detected with the 0', 45'nd 60'ltrasonic examinations but was repaired due to a clear radiographic indication. Excavation revealed a 1/32-inch thick, smooth slag inclusion measuring 1-inch throughwall by 3 inches long near midwall at the fusion line of a nozzle-to-vessel weld.

Due to the similarity of the nondestructive examination responses, believed that the indication in the Ginna reactor pressure vessel (RPV) it is nozzle-to-vessel weld N2$Vis a thin, smooth slag inclusion in a plane perpendicular to the beaih of the 15'ozzle bore examination.

Heat affected zone (HAZ) cracks were detected in three nozzle-to-shell welds in the 1972 preservice of the Hatch Unit 1 RPV. These cracks were detected with 45'nd 60'xaminations from the outside surface of the vessel and subsequently excavated and repaired. These cracks were con-firmed with 0', 45 and 60'xaminations from the inside surface of the vessel, 10'xamination from-the nozzle bore and by metallographic examin-ation during excavation. Neither record radiographs nor radiographs taken on site revealed the HAZ cracks. The multifaceted nature of such cracks breaks up the reflection so that several small indications of multiple planes are observed. Sizing such reflectors to the 50% DAC limits of the indication works well without beam spread correction of indication dimen-sions. Indications having unknown orientation and identity should be sized to the 50% DAC limits without beam spread correction. However, based on Southwest Research Institute's (SwRI) experience with flaw indicat'n sizing, beam spread correction should be used on some flaw indications.

Under most conditions, a reflector can be sized by using the rules contained in Appendix 1 of the 1974 Edition of Section ZI of the ASIDE Boiler and Pressure Vessel Code. In this case, these rules are not appropriate.

Appendix 1 sizing rules are based on using 45'nd/or 60'ngle beam (or other angle beams separated by at least 15') examination techniques. Such examinations were applied to the vessel inside surface but did not detect the subject reflector and the weld in effect passed the Code-reauired UT examinations. Even though it was not a Code requirement, an angle beam longitudinal wave was applied to the bore of the nozzle and directed perpendicular to the axis of the weld. This additional examination was performed in the interest of maximizing the effectiveness of the weld interrogation. Directing a beam perpendicular to the major reflecting plane of a weld-related defect results in a high degree of reflectivity from an imperfection. This technique, which is more sensitive to planar reflectors at the weld interface than the typical technique described in Appendix 1, can be applied only because of the unique geometry surrounding the nozzle-to-shell weld. Because the search unit movement during'he nozzle bore scan is essentially parallel to the plane of the flaw, some of the examination parameters which enter into the flaw-size calculations des-cribed in Appendix 1 are not available. Since reflection amplitude an d parallel search unit movement are the only parameters available to be used in determining flaw size, supplemental considerations must be e fected.

Recognizing that the difference between the measured and the true reflector size can only be determined by considering the basic sound beam properties, a brief summary oi the physics of ultrasonic beam propagation which defines the beam spread and sound pressure amplitude distribution within that beam axe as follows:

The beam spread angle $ from the beam axis to the edge of the total beam is calculated from the equation, d total arcsin 1.22 X/D

=

Where X ~ wavelength~ V/f V = velocity in the material in millimeters per microsecond f = ezamination frequency in cycles per microsecond D = dimension of the txansducer in millimeters The beam spread angle to the 50% point is calculated using the equation, 4

arcsin .56 V/fD and to the 20% point 4

arcsin .92 V/fD A 2.25 8Hz, 3/4-in. dia. seaxch unit producing a straight beam longitudinal wave in steel has a V/fD 5.89/2.25(.75)25.4 = .1374, therefore f50~4

= 4.4'nd 20"/0 Since these angles are computed for the beam azis to the 50% and the 20%

levels, the 50 to 50% and 20 to 20% angles double. That is, half of the energy is outside the 8.8'one and 20% of the energy is outside the 14.6'one.

In order to obtain beam measurements on the same basis as the computed beam spread angles, we would need to use various sizes of disc reflectors (such as flat-bottom holes) to reflect the 50 to 50% portion of the beam. Side-dxilled holes are a better geometric simulation of suspect reflectors and are more convenient reflectors for angle beam calibration since they are equally rezlective to various beam angles and modes of wave motion. The nonmandatory Code technique for beam spread measurement uses the 50 to 50%

DAC response from the side-drilled hole. (This is different from the 50%

~ of total beam computation and the measured angles are different from 'the computed angles.) In this case the investigati'on metal path is greater than 3T/4 metal path. so the responses from the T/2 and 3T/4 calibration reflector were used for beam spread measurement. Figure 2 shows spread for the 50 to 50% DAC points, while similar plotting gave an 1/2'eam 11'eam spxead for the 20 to 20% DAC points as measured on side-drilled holes. Such a beam detected and 50% DAC sized the Figure 1 indication which gave 117% DAC maximum amplitude from a reflector in nozzle-to-shell weld N2. in the Ginna RPV during the ~~farch 1979 inservice examination.

As shown in the top frame of Figure 1, the indication was recorded on a cixcumxerential scan of the nozzle-to-vessel veld from the nozzle boxe with the 15'ngle beam longitudinal wave at 7.64-inch metal path in rexerence position 19.14 inches and on successive circumferential scans at rexerence positions 19.40, 19.62, and 19.87 inches. These scan increments increase the metal path to a rexlector perpendicular to the 15'eam by the relation-ship:

AW = Sin 15'scan increments): that is,

~ ~ Sin 15'.26 in.) = .0673 in.

= Sin 15'.22 in.) = .0569 in.

~ Sin 15'.25 in.) = .0647 in.

Adding these HPs and comparing the computed HPs for a planar reflector to average mmamum amplitude metal paths for each of the three successive scans gives the following:

Range of Maximum Amplitude Deviation Computed Comouted Wi Average Metal Path iso. of Areas from Ave. <Anus Ave.

7.64 in. 7.64 in. (Base Reading) (1) 7.71 in. 7 '3 in+ (4) +.01 .02 in.

7.76 in. 7 '5 in ~ (3) +.01 +.01 in.

7.83 in. 7.83 in. (4) +.01 .00 in.

The +.01-inch range of deviation from average calculates to a range of reflector plane angles of 15'3'or the eleven maximum amplitude metal path readings. This substantiates our conclusion that the slag inclusion is located at the weld-to-base metal interface in a plane perpendicular to the 15'ngle beam longitudinal wave.

A study was conducted, to demonstrate the appropriateness of beam spread corrected reflectox sizing. Essentially this study consisted of placing a flat-bo'ttom hole reflector in the calibxation block and comparing the measured size with the known reflector size. Zn this test, the geometry of the nozzle ezamination area was simulated, the same or similar search unit and wedges wexe used, and similar records were taken. Other controls exercised in the study to assure appropriateness of the comparison are as follows:

1. A flat-bottom hole was drilled at an angle so that the search unit-wedge combination used in measurement of the flaw indication produces a beam perpendicular to the, flat-bottom hole at a metal path within +10% of the flaw indication metal path.

2. The flat-bottom hole was located in a calibrat'on block so it does not interfere with subsequent calibrations.

A3-12

3. The selected calibration block had ultrasonic coupling conditions similar to the examination and had the same diameter side-drilled holes as used in calibration for the examination.

4. Calibration was performed on the block for comparing the flat-bottomed hole response to the maximum response of the flaw indica-tion.

5. The flat-bottom hole amplitude response did not deviate from the flaw indication response by more than 2 dB.

6. 50% to 50% DAC measurements were made on the flat-bottom hole in the through-wall and length directions as the search unit was moved toward and across the reflector.

7. The 50% to 50% DAC through-wall dimension of the flat-bottom hole minus the flat-bottom hole diameter was demonstrated to be the through-wall spread correction.

8. The 50% to 50% DAC length dimension of the flat-bottom hole minus the flat-bottom hole diameter was demonstrated to be the length beam spread correction.

9. The flaw indication through-wall dimension 2a minus the through-wall beam spread correction was shown to be the beam corrected flaw indication dimension 2ac.

10. The flaw indication length dimension R minus the length beam spread correction was shown to be the beam spread corrected flaw indication Rc.

11. The beam spread corrected flaw indication dimensions ac and Ec as shown in Figure 3"were used in computing a/R ratio and the a %

of t of the indication for comparison with the allowable indication limits applicable to the indication location.

En this instance, it was possible to simulate the examination 'condition and vessel component geometry and demonstrate the effects of ultrasonic beam spread on determining the size of a reflector. This particular reflector was well suited for this exercise because its orientation was established with an unusually high level of confidence by interrogating it in more ways than is usually possible. All of the information accumulated relative to this reflector gives a high level of confidence that its true size, orientation, and character are as reported and that the practice of basing its size on the projected 50% DAC limits corrected for beam spread is appropriate.

Also, paragraph IWA-2240 allows for alternate techniques to be used in lieu of the Code specified techniques ifit can be demonstrated to the satisfaction of the inspection specialist that the alternate techniques provide results which are equal or superior to the Code specified techniques. This was done and the alternate sizing techniques used are therefore in compliance with the requirements of Se tion ZI of the ASME Code.

W.

w( b~

C.'"fcGaughey, Staf ngi r Attachments (3)

305'

+53V 0 /OP g ".

0- -- .. f.7~a .'

6

~

)

t~t.g>o,g J

(4 ~'( )f gall L 5LA/VS gw" 44"

$,)

ga.=83 l7QI 2<

P ALL04: )BATE ALLOHA

/ j /

/GAEL I

yiC I

) I

A3-14 I ~

8EAlvi SPPiAO Odl bYDIA Fl.Af'5(DE DAIL/ED hC'/ES I ~

1$

)

~Sf I

0 f~ I f

69 64~

50 ':I/Zi:50;:

0/ a,", o:

'~T

~ I

. I ~

'->-t '

+

-:...:=..: -78498 975';:,:

~ 'c 527/-/./G+.375

= cf..52 )"

I I ~

I 5@a ~

I

'I

.:: FIG////'EZ E9/v/ SPPEAOZfOI/CTION OFCO/'EPOS/TE INPlCATIONOINEI/Sf/I/IS Dwk'j 128 C(i'(OfifAP7't Qdi('C.

<<~ ~

~ '" i I

I

~ ~

q,o I

315$

)~'g I

3I5'l0 I J;-452I A

I

~ I

/2A =.03$ ~ gl I ~85 I

I ~ \

• f

,32N"=Pa L' 0' 305'300' a =.JBD3"=%78%oF 3045'/YozzAE

~ ~

ALlDW-ABI E .'::. I .. SECTIDk'IEW p ~

~ I 1

I I I tAtELO CEITTERLIAIE SECTIDAI e;

VIEll',

I P I A

. PY /IIPSkliz~F ioV- 4ElI) NI)ICNIOiV;,

ESSFL

'BNT:0'/5' I

I Sic)t< JWN>lux 7~<ÃflaulI II'ite

+"~~~~ &zF

)

/

F@N'E3 Brest 5msn0Cwevu /

(; 'Dt'7-RRI r iAT<in IfJWnu 7/ ill,')u'l":.,E, 2/4' )

' kPP

-i.;....

IhIFT Ai.;,

I ATTACHMENT 4 FRACTURE MECHANICS EVALUATION OF INLET NOZZLE INSERUICE INSPECTION INDICATION