ML13322B141
| ML13322B141 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 05/02/1990 |
| From: | Nandy F SOUTHERN CALIFORNIA EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML13322B142 | List: |
| References | |
| NUDOCS 9005040032 | |
| Download: ML13322B141 (31) | |
Text
I2 Southern California Edison Company 23 PARKER STREET IRVINE, CALIFORNIA 92718 F. R. NANDY TELEPHONE MANAGER OF NUCLEAR LICENSING May 2, 1990 (714) 587-5400 U. S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555 Gentlemen:
Subject:
Docket No. 50-206 Emergency Diesel Generators San Onofre Nuclear Generating Station Unit 1
Reference:
Letter dated November 21, 1989, from Charles M. Trammell (NRC) to Harold B. Ray (SCE), TDI Diesel Crankshaft Cracking Problem This letter responds to your safety evaluation on the crankshaft crack propagation problem for the San Onofre Unit 1 emergency diesel generators. The safety evaluation was enclosed with the reference, which made the following observations:
SCE must demonstrate that a 10 mil deep flaw can be detected with a high degree of confidence by eddy current testing (ECT) or the crankshafts should be replaced.
SCE should review and revise (as appropriate) its diesel generator oil maintenance procedures to assure that oil quality is properly maintained and impurities which could contribute to the cracking problem do not exist. The safety evaluation referred specifically to the adverse effects of impurities such as sulfides.
Engine start-stop cycles with engine speeds maintained less than 200 rpm may be excluded from the number of start-stop cycles permitted by license condition 3.L (1) provided SCE can demonstrate that the crankshaft stress levels for this situation remain less than the steady state stresses.
Along with resolution of the above items, SCE should consider requesting an amendment to license condition 3.L. This amendment would eliminate the need for a specific license condition by transferring the requirements of license condition 3.L to the plant Technical Specifications.
The first three of the above items are the subject of this letter and are addressed below. SCE agrees with the recommendation in the fourth item and is 9005040032 900502 FDR ADOCK 0500026
Document Control Desk May 2, 1990 preparing an amendment application to implement it. This amendment application will be submitted separately.
Adequacy of ECT Technique Your safety evaluation requires SCE to demonstrate that the ECT technique is capable of detecting 10 mil deep flaws with a high degree of confidence.
Failure Analysis Associates (FaAA) has conducted the necessary testing and concluded that 10 mil deep flaws can be readily detected. Enclosed is FaAA's report on this topic (Letter No. FaAA-SF-90-02-03, Revision 1.0, dated April 19, 1990).
FaAA used a carbon steel block with similar electromagnetic properties to the material of the San Onofre Unit 1 crankshafts. A standard size oil hole (15/16 inch diameter and 7/16 inch fillet radius) was machined through the center of this block to resemble the oil hole in the crankshaft.
Artificial flaws or notches of known dimensions were machined into the oil hole fillet using electro-discharge machining (EDM).
These flaws had depths of 5, 10, and 20 mils and length to depth ratios of 6:1 and 2:1.
Flaws with the higher 6:1 length to depth ratio resembled more closely the cracks discovered at San Onofre Unit 1 in 1984. The width of the notches varied between 4 and 5 mils.
The eddy current probe used had a sensing spot size of 0.1 inch diameter and operated at a frequency of 2 MHz. The probe was manipulated over the flaws by means of a standard test fixture specifically designed for the inspection of oil hole fillet radii.
The results showed that flaws to depths as small as 5 mils can be readily detected.
Diesel Oil Procedures Since crankshaft cracks as deep as 1/4 inch were reported previously at San Onofre Unit 1, the safety evaluation discusses the possibility that oil impurities, which result from engine operation or oil manufacture, may have contributed to crack initiation and propagation. Your letter recommends that SCE review and make appropriate revisions to its diesel oil maintenance procedures.
The diesel oil sampling, testing, and analysis procedure has been reviewed.
No evidence was found that the diesel oil used by San Onofre Unit 1 accelerates stress corrosion cracking or bulk corrosion processes. A review of the oil sampling and analysis methods shows that the total base number and water content of the oil are within acceptable limits. The crankshaft material has the recommended hardness value and yield strength to resist corrosion in a sulfide bearing environment. Key oil parameters are trended.
Based on this information, no changes to the procedures are considered necessary.
Document Control Desk May 2, 1990 Start-stop Cycles with Engine Speeds Less than 200 RPM Your transmittal letter states that start-stop cycles associated with engine speeds less than 200 rpm may be excluded from the maximum limit of fifty provided by license condition 3.L(1), if the crankshaft stress levels for this situation are demonstrated to be less than the steady state values. This statement represents your conditional acceptance of a request made by SCE to exempt engine operation below 200 rpm from start-stop restrictions. FaAA has analyzed the no load operating stresses up to 200 rpm and has concluded that they are less than the full load steady state values.
FaAA analyzed the crankshaft stresses during idle speed startup (0-150 rpm),
idle speed operation (150-200 rpm), and coastdown from idle speed. These stresses were compared with the crankshaft stresses previously calculated for full load (6000 KW) steady state operation. In all cases, the stresses were less than the full load stresses.
Enclosed is FaAA's report on this topic, describing the method of analysis used and the results achieved (Letter No. FaAA-SF-90-02-03, Revision 1.0, dated April 19, 1990).
Based on these results, start-stop cycles associated with no load engine speeds less than 200 rpm need not be counted toward the requirements of license condition 3.L(1).
If you have any questions or require additional information, please contact me.
Very truly yo -rs, Enclosure cc: J. B. Martin, Regional Administrator, NRC Region V C. Caldwell, NRC Senior Resident Inspector, San Onofre Units 1, 2 and 3
ENCLOSURE
Failure Analysis Engineering and Scientific Services 149 Cornonwealtn Drive. P 0. Box 3015 Menlo Park. California 94025 Associates, (415) 326-9400 Telex 704216 Fax (415) 326-8072 FaAA-SF-90-02-03 Revision 1.0 April 19, 1990 Mr. David Pilmer Southern California Edison Company Nuclear Engineering and Construction 23 Parker Street Irvine, California 92718 Re: Response to NRC Safety Evaluation Report Issues
Dear Mr. Pilmer:
This letter provides responses to the issues raised by the NRC in their Safety Evaluation Report dated November 21, 1989, regarding the crankshafts in the TDI DSRV20-4 diesel engines at San Onofre Nuclear Generating Station (SONGS), Unit 1. Specifically two issues were identified by the NRC: 1) reliability and sensitivity of the eddy current inspection method, and 2) evaluation of the crankshaft stresses during an engine start up to 200 rpm.
ISSUE: Demonstrate that SONGS Unit 1 eddy current inspection technique is capable of detecting 10 mil deep flaws with a high degree of accuracy
RESPONSE
The ability of the eddy current inspection technique to detect 10 mil deep flaws in the crankshaft oil holes was demonstrated by development and evaluation of a reference standard for the SONGS Unit 1 crankshaft oil holes.
The reference standard was created from material similar to the crankshafts at SONGS Unit 1. A standard size oil hole (as specified by TDI drawings) was bored through the center of the reference block, and oil hole fillet radii were machined on each end of the block.
Artificial flaws of known dimensions were machined into the oil hole fillets utilizing electro-discharge machining (EDM). Artificial flaws with depths of 5, 10, and 20 milswith a length to depth ratio of 6:1 and 2:1 were machined into the simulated oil hole. The higher aspect ratio was evaluated, since the cracks discovered in the main journal oil holes at SONGS Unit 1 in 1984 had high aspect ratios. Although the method of crack removal did not allow exact determination of crack depth and thus aspect ratio, based on the observed lengths and estimated depths, an aspect ratio of 14 is representative of the cracking observed at SONGS Unit 1. The width of the EDM notches was between 4 and 5 mils. The flaws were placed 70 degrees into the fillet radius (0 degrees being on the journal surface) to correspond with the location of initial cracking observed in 1984.
Failure Analysis Associatese Inc.. a Wholly Owned Subsidiary of The Failure Group, Inc.
UNITED STATES EUROPE CANADA Boston Detroit Houston Los Angeles Miami Phoenix Son Francisco Seattle Wasmington DC D1sseldorf Vancouver
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 2 Eddy current response of each flaw was evaluated in order to demonstrate the sensitivity of the eddy current test method for the inspection of crankshaft oil holes. The eddy current probe used for this work was an FaAA 10OR type, radius tip, shielded probe with a sensing spot size of 0.1 inch in diameter. with an operating frequency of 2MHz. The test instrument was a smartEDDY 3.0 computer-based, portable eddy current instrument. The probe was manipulated over the cracks by means of a standard test fixture specifically designed for inspection of oil hole fillet radii.
Eddy current inspection of the reference standard clearly detected all six artificial flaws with a high signal-to-noise ratio.
The literature surveyed is inconclusive with respect to the ability of EDM notches to represent actual fatigue cracks which may form under service conditions. The main geometric feature of an EDM notch which differs from an actual fatigue crack is the degree of opening that exists between the faces of the crack. Support is found in the literature to demonstrate that the crack opening does not have a significant effect on the magnitude of the eddy current signal. It is also found that in certain instances the signal from a natural crack is smaller than that from an EDM notch. EDM notch widths are typically between 3 and 5 mils, whereas a natural fatigue crack is expected to be significantly tighter. The crack response is reduced when two fracture surfaces touch and an electrically conductive path results. Conductivity of this path is dependent on whether or not a sufficient insulating layer exists between the crack faces to produce a high resistive path to the current in comparison to the path within the metal. For the case where a sufficient insulating layer is present, an EDM notch will act as a good simulator for actual fatigue cracks. If a crack is tight and if some amount of shorting of the crack occurs, then the eddy-current response will not be as large as from an EDM notch of the same size.
If a crack were to initiate in the crankshaft oil hole environment, some combination of oil, oxide, and air must exist in the crack in order to provide the insulating layer necessary for detection.
FaAA believes that an oxide layer would form to provide a sufficient insulating layer for purposes of eddy current testing.
The calibration response used in the past eddy-current inspection of oil holes are quite conservative. They are based on the response from flaws with a 2:1 length to depth ratio compared with the much larger aspect ratio characteristic of cracks in the oil-hole. The study reported here demonstrated that the eddy current response from flaws with a 6:1 aspect ratio is three times larger than the response from the calibration flaws.
The analysis method used to predict the inspection interval assumes high aspect ratio cracks by utilizing a 1DOF crack model [16]. If a 10 mil deep crack Failure Analysis Associates
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 3 with a 2 to 1 aspect ratio did occur, a 3DOF crack model would be appropriate resulting in a much greater inspection interval.
FaAA has found the eddy current inspection technique to be the most sensitive technique for detection of crack-like indications in crankshaft oil holes. This technique has been used extensively for the inspection of crankshaft oil holes in work performed for the TDI Diesel Generator Owner's Group.
Details of the eddy current evaluation are provided in Attachment A.
ISSUE: Engine start-stop cycles with engine speed less than 200 rpm need not be counted towards the 50 start-stop limit provided it is demonstrated that the induced crankshaft stress levels remain below full load steady state values RESPONSE: The purpose of this study was to evaluate the stress levels during a 200 rpm idle speed test and verify that they remain below 6000kW load steady-state operating stress levels. Steady-state analyses were performed to ascertain the effect of dwelling at an engine speed during the test. Transient analyses and test data were used to evaluate stress levels during the test startup and coastdown. The lumped inertia and torsional spring model, used for previous steady-state and transient analyses of the SONGS Unit 1 crankshafts, was used for the current work.
A 150 to 200 rpm idle speed test is sometimes performed after maintenance and rework of the engine. During this test, the engine is started and the engine speed is manually increased by the operator until an engine speed of 150 to 200 rpm is reached. The engine is run with no load between 150 and 200 rpm for approximately 20 minutes and then shut down. During the transient portions of this test, the engine passes through the first mode 10th order resonant speed of 119 rpm. The engine speed versus time data for the startup varies and is not necessarily repeatable since the speed of the engine is controlled by the operator. The engine speed is not controlled during a coastdown. Throughout the duration of the'test, the engine speed does not exceed 200 rpm (SONGS, Special Engineering Procedure SO1-SPE-712).
Steady-state harmonic analyses at engine speeds between 120 and 200 rpm (in 10 rpm increments) were performed to determine the effect of dwelling at a particular speed. This condition could occur either during the manual startup procedure or while running at constant speed with no load. No-load pressure harmonics calculated from a cold compression curve with a peak pressure of 450 psi (developed previously for the coastdown analysis) were used to simulate no-load operation at a particular speed.
For all engine speeds Failure Analysis Associates in
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 4 evaluated, the amplitude of torsional stress was below the maximum full-load steady-state stress level of 3185 psi at 6000kW load.
Transient analyses were performed to evaluate stress levels developed during the startup portion of the 200 rpm idle speed test.
Four different startup conditions were evaluated, starts with a 20 second duration up to 150 and 200 rpm and starts with a 30 second duration up to 150 and 200 rpm. For all starts a linear increase in speed was assumed. The cylinder pressure loading on the crankshaft was estimated from the same technique used for a previous study of the impact of slow starts on the SONGS Unit 1 crankshafts. An initial crankshaft start angle of 0 degree was assumed for all startup conditions.
Results of the analyses indicated that the linear startup to 150 rpm in 20 seconds produced the highest stress amplitudes. Stress levels during this start were below the peak stress amplitude during 6000kW load normal operation.
During recent torsiograph testing at SONGS Unit 1, data were collected during a 200 rpm start. Preliminary data reduction indicate that the test data are consistent with the input used for and the results obtained from the analytical model.
The effect of initial start position on the crankshaft response amplitudes was quantified by performing analyses with different initial start positions for the startup condition that produced the highest stress levels (150 rpm in 20 seconds). The following initial startup positions were evaluated: 0, 45, 90, 180, 360, and 540 degrees. A 22% variation in response was observed as the initial crankshaft position changed.
For all initial start positions evaluated, the maximum stress amplitude was below the 6000kW load steady-state stress levels.
Coastdown response levels during a 200 rpm idle speed test were obtained from previous test data and analyses on the SONGS Unit 1 crankshaft. Stress amplitudes as the engine passed through the 10th order resonant speed were below the 6000kW load steady-state stress levels.
Evaluation of the stress levels during the 200 rpm idle speed test indicates that the crankshaft stresses remain below the 6000kW load steady-state stress levels throughout the duration of the test. Details of the 200 rpm idle speed analysis are provided in Attachment B.
Failure Analysis Associates nc
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 5 This work was performed in accordance with FaAA's Quality Assurance Operating Procedures for nuclear safety-related equipment. If you have any questions, please feel free to call me at (415) 688-7210.
Sincerely, W'~
k 1/
Paul R. Johnston, Ph.D.
Principal Engineer Lisa M. Shusto, P.E.
Senior Engineer Failure Analysis AssociatesI
Mr. David Pilmer FaAA-SF-90.02. 03 4/19/90 Revision 1.0 Page 6 ATTACHMENT A VERIFICATION OF EDDY CURRENT INSPECTION SENSITIVITY EDDY CURRENT REFERENC The ability of the eddy current inspection technique to detect 10 mil deep flaws in the crankshaft oil holes was demonstrated by development of a reference standard for the SONGS Unit 1 crankshaft oil holes. The reference standard is a 3 x 3 x 3-inch carbon steel block with a standard size oil hole (a5/16-inch diameter and a 7/16-inch fillet radius) machined through the center of the block.
The material used was a remnant from an actual crankshaft of similar dimensions. This material has similar electromagnetic properties to that used for the crankshafts at SONGS Unit 1. A schematic of the reference standard is presented in Figure Al.
Artificial flaws of known dimensions were machined into the oil hole fillet utilizing electro-discharge machining (EDM). Artificial flaws with depths of 5, 10, and 20 mils and a length to depth ratio of 6:1 and 2:1 were machined into the simulated oil hole. Flaws with the higher 6:1 aspect ratio were evaluated, since the cracks discovered in the main journal oil holes at SONGS Unit 1 in 1984 had high aspect ratios. The width of the EDM notches varied between 4 and 5 mils. Flaws in the oil hole were placed 70 degrees into the fillet radius (with 0 degrees being on the journal surface) to correspond to the location of initial fatigue cracking discovered in the main journal oil holes at SONGS Unit 1 in 1984 [14, 15]. The actual EDM flaw dimensions are summarized in Table Al.
CALIBRATION BLOCK EVALUATION The eddy current response of each flaw was evaluated in order to demonstrate the sensitivity of the eddy current test method for the inspection of crankshaft oil holes.
The evaluation was performed according to FaAA's procedure NDE 11.3, Revision 2: "Eddy Current Inspection Procedure - Main Journal and Crank Pin Oil Hole" (see Attachment Al). The eddy current probe used for this work was an FaAA 10OR type, radius tip, shielded probe with a sensing spot size of 0.1 inch in diameter. The operating frequency was 2MHz. The test instrument was a smartEDDy 3.0 computer-based, portable eddy current instrument. The probe was manipulated over the cracks by means of a standard test fixture specifically designed for inspection of oil hole fillet radii.
The results of the evaluation are summarized in Figure A2. The actual eddy current signals as the probe passed over the artificial defects are presented in Figures A3 through A14. For each notch size evaluated, the eddy current data are presented in two modes.
The first mode is an impedance plane Failure Analysis Associates n
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 7 presentation where the axes are calibrated in units of percent change in impedance (the change in impedance is referenced to the null or balance point).
Typically the X and Y axes represent the resistive and reactive components of the probe's impedance, respectively. For the current work, the liftoff is rotated to display horizontally, and to the right, in line with the X axis.
Thus the X and Y axes represent the components in the rotated coordinate system.
With the liftoff signal rotated to display horizontally, a crack-like indication would be displayed predominantly vertical and downward, in line with the Y axis. The impedance plane plots are presented in Figures A3, A5, A7, A9, Al l, and Al 3. The vertical and downward change in signal is clearly visible for all notches evaluated.
The second mode of presenting the data is a time base representation presenting the same data as the impedance plane plot. The rotated X and Y components of the impedance plane display are presented as a strip chart trace. The upper trace showing the liftoff or X direction and the lower trace showing the Y direction. Due to the signal rotation, a crack like indication would be displayed on the Y axis. The time base plots are presented in Figures A4, A6, A8, A10, 1A2, and A14. The eddy current inspection of the reference standard clearly detected all six artificial flaws with a high signal-to-noise ratio as evidenced in the Y axis strip chart plots. This study indicates that flaws to depths as small as 5 mils can be detected with the appropriate inspection method and equipment.
The eddy current trace obtained while testing an oil hole surface is significantly different that the familiar "figure eight" trace obtained while testing steam generator tubing. The main difference is due to the type of probe used to perform the testing. Tube testing is normally performed using a differential bobbin probe which produces the characteristic "figure eight" trace, whereas the oil hole fillets were evaluated with an absolute, shielded, focused pencil style probe producing the type of signal shown in the impedance plane plots. In addition to the probe type, the material type of the component being evaluated affects the trace. The change in phase angle as a function of crack depth is insignificant for thick wall ferromagnetic materials (e.g. steel) as compared to a nonferromagnetic thin wall tube (e.g. inconel).
The eddy current evaluation was performed by a certified FaAA, ET Level Ill inspector. The most commonly used procedure for certification of NDE personnel is ASNT's "SNT-TC-1A - Recommended Practice for Qualification and Certification of NDE Personnel". This document specifies that certification is to be conducted according to a written practice set up by the employer of the inspector. FaAA's written practice "NDE 2.1 - Certification of NDE Personnel" is in full compliance with the national practice.
Failure Analysis Associates Inc.
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 8 LITERATURE REVIEW Eddy-current inspection is a nondestructive inspection technique used to detect surface and subsurface flaws in components. The literature suggests that eddy current testing is beginning to move to a level of confidence where flaws can be not only be detected but fully characterized [12].
The eddy current inspection technique is based on principles of electromagnetic induction. An electric coil/probe in which an alternating electromagnetic field is present is placed adjacent to the part to be inspected.
This applied alternating electromagnetic field causes eddy currents to flow in the part as a result of electromagnetic induction. The amount of eddy currents flowing in the part is determined by the electrical conductivity of the part as well as the frequency and amplitude of the applied electromagnetic field. The eddy currents in the part are monitored by a sensor as the coil/probe passes over the part to be inspected.
When the probe passes over a defect or other discontinuity, the eddy current flow is distorted, changing the monitored signal.
Thus, only those defects which disturb or alter the normal eddy current flow patterns are detectable. The eddy current setup is adjusted with calibration standards duplicating the test material in geometry as well as electrical and magnetic properties [1, 13]. EDM notches or machined slots are typically used to simulate cracks (13].
The literature surveyed is inconclusive with respect to the ability of EDM notches to represent actual fatigue cracks which may form under real service conditions [1-13]. The main geometric feature of an EDM notch, which differs from an actual fatigue crack, is the degree of opening that exists between the faces of the crack. Support is found in the literature to demonstrate that the crack opening does not have a significant effect on the magnitude of the eddy current signal. It is also found that the signal from certain natural cracks are smaller than that of an EDM notch. EDM notch widths are typically between 3 and 5 mils, whereas a natural fatigue crack is expected to be significantly tighter. The crack response is reduced when two fracture surfaces touch and an electrically conductive path results. Conductivity of this path is dependent on whether or not there is a sufficient insulating layer between the crack faces to produce a high resistive path to the current in comparison to the path within the metal. For the case of a fatigue crack where a sufficient insulating layer is present, an EDM notch will act as a good simulator for an actual fatigue crack. If a crack is tight and some amount of shorting occurs, the eddy-current response will not be as large as from an EDM notch.
Failure Analysis Associates 1i0
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 9 If a crack were to initiate in the crankshaft oil hole environment, some combination of oil, oxide, and air must exist in the crack in order to provide the insulating layer necessary for detection.
FaAA believes that an oxide layer would exist to provide a sufficient insulating layer for purposes of eddy current testing.
The 10 mil deep inspection criteria was based on analyses developed using a crack model that produced shapes similar to those observed in 1984. The two to one aspect ratio was chosen for the inspection criteria prior to the observance of cracking at SONGS. The cracks observed in the SONGS Unit 1 crankshafts were long, with lengths up to 3.5 inches.
The method of removing the significant cracks was to bore the oil hole at an increased diameter and reinspect the oil hole for indications. This process was continued until no defects were detected using the eddy current test method. This method for removing the cracks does not allow for exact determination of the depths of the observed cracks making it difficult to determine exact aspect ratios. The last indication was removed when the oil hole diameter was increased from 15/16 inch to 1-1/2 inches yielding an approximate crack depth of 0.25 inch [14, 15].
This indicates that aspect ratios as high as 14 are representative of the cracking observed at SONGS.
The calibration response used in the past eddy-current inspection of oil holes are quite conservative. They are based on the response from flaws with a 2:1 length to depth ratio compared with the much larger aspect ratio characteristic of cracks in the oil-hole. The study reported here demonstrated that the eddy current response from flaws with a 6:1 aspect ratio is three times larger than the response from the calibration flaws.
The analysis method used to predict the inspection interval assumes high aspect ratio cracks by utilizing a 1 DOF crack model [16]. If a 10 mil deep crack with a 2 to 1 aspect ratio did occur, a 3DOF crack model would be appropriate resulting in a much greater inspection interval.
FaAA has found the eddy current inspection technique to be the most sensitive technique for detection of crack-like indications in crankshaft oil holes. This technique has been used extensively for the inspection of crankshaft oil holes in work performed for the TDI Diesel Generator Owner's Group.
Failure Analysis Associates Inc
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 10 Table Al.
Summary of EDM Flaw Dimensions Flaw Dimensions (mils)
Flaw Number length depth width 1
11 5
5 2
19 10 5
3 40 20 5
4 27.5 5
4.5 5
60 10 5
6 120 20 4
Failure Analysis Associates in.
3" Flaw No.
Main journal surface1 1200 Typ 3"
3" 3
2 EDM Notched Location 7/16" Radius (Typ) 15/16" 3"
Diameter 4
6 5
Figure Al: Schematic of Reference Standard for Songs Crankshaft Oil Holes
0.7
/
0.6 0.5 6:1 Aspect Ratio C.,
O
- 1ApetRai E 0.4
/
0
.1 1
2 2
2 0
a)0.3
,/
- 0.
-, O 2:1 Aspect Ratio 0.21/
0 5
10 15 20 25 Flaw Depth (mils)
Figure A2. Summary of Eddy Current Evaluation 000241
0.30 Inpedance
+ +Filters (Sac)
+-
+
+
+
+4
+
4 x-h i = OFF 0.18+
+
+
+
+
s-hi = OFF
+-
+
+
+
4-
+
+
+-
Rttos' 4-
+
+
+
4-R1 =198.50 R2 = 0.00 E S
-R3 0.00 4-
+
+
-0.180.0
-0.3D6-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Frequencies
+
+
+
+
+/-
+
+
+
+
1)2.00MHz I.... + ---- +.
- 2) 0
-Hz
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
1)0.0 Uolt 2)l =0.020ol AutoStore OFF AAM
-0.30
+
1
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change In Impedance Figure A3: Impedance plane plot of eddy current response for a 5 mil. deep x 10 mil. length EDMV notch.
1)20092M z
Inpedance SETUP #1 2.0 -
+
+
+
+
+++s u:
2.0
+
+
+
+
+
Filters (See) x-lo =.020 1.0
+
+-
+
+
++
+
+4-1
.2 x-hi = OFF
+
+
+
+
+
+
+
RX 1
4-lo
=.020 0.0......
.....4. 4-..... 4.
1-Z
+
-hi = OFF a+
+
+
+
+
+
+
+
0 Rotations (9)
-1.0
-+
+
+
+
+
+
+
+
I RI1a 198.50
+
+
+
+
+
+
+
+
+
R2=0.00
-2.0 -
+
+
+.
+4
+
+-
R3=0.00 MI I
I I
I R4=0.00 0.040
+-
+
+
+
+
+
+-
R5.0.00
+-
+
+
+
- 4.
+
+
+
0.020
+
+
+
+
4
+
+
+
Frequencies 0
~
+
+
+
+
+
+
+
LZ
+-
+-
+-
+
4
+
+
+1) 2.10 MHz Y
0.0..--......-..
+.
+.
- 2) 0 Hz Drive Levels 4+
+
+
+
+
+
+
+
-0.020 -
+
+
+
4+
+
+
+
+
+
1)3.0 Uolt 2)0.0 Uolt
+
+
+-+
+
+
+
+
+
-0.040
-+
+
+
+
+
4
+
+
4-LAM AutoStore OFF ALARM Time (sec.)
0 4
8 12 16 20 Figure A4: Eddy current response of 5 mil. deep x 10 mil. Length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
=4 a
0.30 Inpedance SETUP #1
+
+
+
+
+
+
+
+
Filters (Sec)
+
+
+
+
+
+
+
+
X-10 =
.020 x-hi = OFF 0.1813
+
+
+
+
+
+
+
+
.- lo =
.020 4 4 W-hi = DFF Rotations (0)
R1 =198.50 0.060--
+
+
+
+
t
+
+
+
+..
.+
=R2
=0.00
-4+
+
+
+
+
+
+
R3 = 0.00 E
R4= 0.00
+
+
+c.
+
+
+
R5 = 0.00 e -0.060-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Frequencies
+
+
+
+
+
+
+
+
- 1) 2.00- MHz 2 0 Hz
-0.18-
+
+
+
+
+
+
+
+
+
DriveLevels
+
+
+
+
+
+
+
+
1)3.0 Uolt
- 2) 0.0 Uolt
+
+
+4
+
+4
+
+
+
).
~l
+
+
+4
+4
+/-
+4
+
+
+
LFI AutoStore OFF ALARMS 0.30I I
I I
I I
I I
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change in Impedance Figure A5: Impedance plane plot of eddy current response for a 10 mil. deep x 20 mil. length EDM notch.
0024 k
Inpedance SETUP 01 2.0
+
+
+
+
+
+
+
Up
+ Filters (Sec) x 0
+
+
+
+
+
+
+
X-I0a.020 x-hi = OFF
+4
+.
+
+
+
+
+
W i-lo =.020 0.0..-----
.+-.-+-.-----+
-.- +---.
+-hi = OFF
+
+
+
+
+
+
+
+
0 Rotations (6)
W
-1'.
+
+
+
+
+
+
+
+
+
I BR1 a198.50
+
+
+
+
1-
+4
+
+
+4 R2=0.00
-2.0
+
+
+
+
+
+
+
+
20
-+R3=0.00 Z
II R4=0.00 0.24
+4
+
+
+
+4
+
+
+
R5=0.00
+
+
+
+
+
+
+
+
y 0.12
+
+
+
+
+
+
+
+
Frequencies
+
+
+
+
+
+
+-
- 1) 2.00MHz 0.0..-----
+
+-------+-- --... +-.----+.
-+
+----..+........
- 2) 0 Hz
+
+
+
+
+
+
+
Drive Levels
-0.12
+
+
+
+
+
+
1)3.0 Uolt 2)0.0 Uolt
-0.24
.+
+
+
+
+
+
+
+LR AutoStore OFF ARM Time (sec.)
0 4
8 12 16 20 Figure A6: Eddy current response of 10 mil. deep x 20 mil. length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
GW4 b
0.30 Inpedance SETUP #1 Filters (Sec)
+
+
+
+
+
+
+
+
+
X-I0 =.020 0.8
+
+
+
+
4
+
+
+
+
-h=OF x-hi = OFF 0.18
+
+
+
+
+
+
+
+
+
lo =
.020 Y-hi = OFF Rotations (0)
RI = 198.50
- 0.060-
+
+
+
+
+
+
+
R2 = 0.00 c
+
+
+
+
+
+0 R3= 0.00 CL
~~........................
.............---- R4 = 0.00 E
+
+
+
+
+
+
+
+
R5= 0.00
-0.060--
+
+
+
+
+
+
+
+
c
+
+
+
+
+
+
+
+
Frequencies
+)
+
+
+
+
+
+
+
1)2.00MHz
.4........
- 2) 0 Hz
-0.18
+
+
+
+
+
+
+
+
Drive Levels
+
+
+
+
+
+
+
+
- 1) 3.0 Uolt
+
+
+
+
+
+
+
+
2)0.0 Uolt
+
+
+
+
+
+
+
+
- AutoStore OFF ALARMS
-0.30 i
I i
i I
j IllIgg
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change in Impedance Figure A7: Impedance plane plot of eddy current response for a 20 mil. deep x 40 mil. length EDM notch.
OD924 L
Inpedance SETUP #1 2.0-
+
+
+
+
+
+
+
+
Filters (Sec)
Xx-lo
=.020 x
+
+
+
+
+
+
+
+
x-hi = OFF 4
+
+
+
+
+
+
+.
w-lo =.020 0.0
+.-.-
+----.+-.
--- +.
Y-hi = OFF
+
+
+
+
+
+
+
+
0 Rotations (*)
W
-1.0
+
+
+
+
+
+
+
+
+
I RI1z 198.50
+
+
+
+
+
+
+
+
R2=0.00
-2.0--
+
+
+
+
+
+
+
+
R3 = 0.00 Z
............ 4....
.4.... 4-..........4 0 O W
I R4= 0.00 0.24-
+
+
+
+
4-
+
+
+
+
W
~R5 a0.00
+
+
+
+
+
+
+
+
0.12.
+
+
+
+
+
+
+
+
+
Frequencies
+.0
+
+
4
+
+
- 1) 2.00 MHz Y
0.0
+
+--.----+
+...
...... 2) 0 Hz
+
+
+
+
+
+
+
+
DriveLevels
-01 1)3.0 Uolt
-0.12
+
+ +
+
+
+
+
+
+
2)0.0 Uolt
-0.24
+
+
+
+
+
+
+
+
+
AutoStore OFF I
I I
I I
I I
I I
Time (sec.)
0 4
8 12 16 20 Figure A8: Eddy current response of 20 mil. deep x 40 mil. length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
0004 C
0.30 Inpedance SETUP #1
+
+
+
+
+
+
+
+
+M
++
+
+
+
+
+
+
Filters (Sac)
+
+
+
+
+
+
+
+
x-lo =
.020 0.18
+
+
+
+
+
+
+
+
+x-hi
= OFF e-lo =.020 u-hi = OFF Rotations (*)
R 1= 198.50 S 0.0601
+
+
+
+
+
+
+
R2 =0.00
+
+
+
+
+
+
+
e -
R3= 0.00 E
R4= 0.00 E
R5 w 0.00
-0.060
+
+
+
+
4-
+
+
+
+
I
+
+
+
+
+
+
+
+
Frequencies
+,
+4
+
+
+
4
+
4 1)2.00 MHz S..............
4.......
- 2) 0 Hz
-0.18
+
+
+
+
+
+
+
+
+
DriveLevels
+
+
+
+
+
+
+
+4.
1)3.0 Uolt
+
+
+
+
+
+
+
+
+
2)0.0 Uolt
+-
+
+
+
4
+
4+
+
+-AAFI AutoStore OFF ALARMS 0.30 I
Ri
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change in Impedance Figure A9: Impedance plane plot of eddy current response for a 5 mil. deep x 30 mil. length EDM notch.
00124 M
Inpedance SETUP 01 2.0
+
+
+
+
+
+
Filters (Sec) x ~
+
+
+
+
+
+
+
F-0=.2 X 1.0
+
+
+
+
4
+2 x-hi = OFF
+4
+
+
+
+
+
+
Y-lo =.020 S0.0
+-...
4-+.4---.-****..*..+
+-----+........
z y-hi = OFF
+
+
+
+
+
+
+
4 Rotations (0)
W
-1.0..
+
+
+
+
+
+
+
+
+
RI1= 198.50
+
+
+
+
+
+
+
+
R=00 R2=.0
-20
+
+
+
+
+
+
+
R3=0.0D I
I I
R4=0.00 0.24
+
+
+
+4
+
+
+
+
W
~R5=a0.00 0
+
+
+
+
+
+
+
+
Y 0.12-
+
+
+
+
+
Frequencies
+
+
+
+
+
+
+
- 1) 2.00 MHz 0.0.. -- +... -
+-........
--....-+.----.
+ -....
- 2) 0 Hz
+
+
+
+
+
+
+
+
Drive Levels V.
- 1) 3.0 Uolt
-0.12
+
+
+
+
+
+
+
+
+
- 2) 0.0 Uolt
-0.24
+
+
+
+ +
+
+
+
+
AutoStore OFF ALARMS Time (sec.)
0 4
8 12 16 20 Figure Al 0: Eddy current response of 5 mil. deep x 30 mil. length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
024 d
0.30 Inpedance SETUP #1
+ 4
+
+
4-
+
+
+
+
Filters (Sac)
+
+
+
+
+
+
+
+
x-lo =.020
+
+
+
+
+
+
+
x-hi = OFF 0.18
+
+
+
+
+
+
+
+
-lo =
.020
+
4
+
+
+Rotations()
R-h = 19FF5
+
+
+
+
+
+
+
+
+ +
+
+
+
+
+
+
R1= 198.50 0.060
+
+
+
+
+
+
+
+
R2= 0.00
+
+
+
+
+
+
+
+R3=
0.00 4.......
R4= 0.00
+
+
+
+
+-
+
+
+
R5=0.00 E
-0.060
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Frequencies
.C
+ 4
+
+
+
+
+
+
+
- 1) 2.00 MHz S............
- 2) 0 Hz
-0.18
+
+
+
+
+
+
+
+
Drive Levels
- 1) 3.0 Uolt
+
+
+
+
+
+-
+
+
++
+
+
+
+
+
+
- 2) 0.0 Uolt 4
+
+
+
+
+
+
+
ALARMS AutoStore 0FF
- AM
-0.30 11 a
I 1
g
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change In Impedance Figure Al 1: Impedance plane plot of eddy current response for a 10 mil. deep x 60 mil. length EDM notch.
00M4 N
Inpedance SETUP *1 2.0
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Filters (Sec) 1.0-
+
+
+
+
+
+
X-10 =.020
+
+
+
+
+
+
+=
OFF x-lo
=.020 Z
x-hi = OFF
+
+
+
+
+
+
+
+
W y-lo =.020 X
0.0...---.
4-.
.4 -
4
.+ -
y-hi = DFF Rotations (a) 1.o 4
+
- 4.
+
+
4.R1=198.50 r+
+
+
+
+
+
+
+
+
R2=0.00
-2.0
+-
+
+-
+
+-
+
+
R3= 0.00 I
I I
1 I
R4=0. 0 0.24
+
+
+
+
4-
+
+
+
.4 R5=0.0D
+
+
+
+
+
+
+
+
+
0.12
+
+
+
4
+
+
+
4 Frequencies SR 01)2.00 MHz V 0 0
) O H tj
+
+
+
+
+
+
+
+
Z 0.12--~~~~)
Hz
+
+
+Feqece X
Y ~0.0 ------
- + -- + - -
+-
+
...........+....
+)
0 H
+
+
+
+
4 Drive Levels
-0.12
+
+
+
+
+
+
+
+
+
- 1) 3.0 Uolt
-2) 0.0 Uolt
+
+
++
+
+
+
-0.24 -
+
+
+
+
+
+
+
+
ALARMS AutoStore OFF ARM I
a I I
I I
a I
I Time (sec.)
0 4
8 12 16 20 Figure Al 2: Eddy current response of 10 mil. deep x 60 mil. length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
0.70 Inpedance SETUP #1
+
+
+
+
+
+
+
+
Filters (See)
+
+
+
+
+
+
+
+X-0
=
.020 x-hi = OFF 0.42
+
+
+
+
4-
+
+
+
+
lo =
.020 t
y-hi = OFF
++
+
+
+
+
+
+eh F
Rotations (0)
+9
+
+.
+9
+
+
+
+
R1 = 198.50 0.14
+
+
+
+
+
+
+
+
0.4
.R2=
0.00
+
+
+
+
R3= 0.00
-E
+
+
+
+
+
+
+
+
R=00 C
R5 =0.00 o
-0.14
+
+
+
+
+
+
+
+
c
+
+
+
+
+
+
+
+
Frequencies
+
+
+
+
+
+
+
+
- 1) 2.00 MHz
..... 4......
4........
...4
.4.........
- 2) 0 Hz Drive Levels
-0.42-
+ +
+
+
+
+
+
+
+
DieLvl 1)3.0 Uolt
- 2) 0.0 Uolt
+
+
+
+
+-
+
+
+
+
AutoStore OFF ALARMS
-0.70 i
i I
I f11
-2.5
-1.5
-0.50 0.50 1.5 2.5
% Change in Impedance Figure Al 3: Impedance plane plot of eddy current response for a 20 mil. deep x 120 mil. length E DM notch.
0M o
Inpedance SETUP #1 2.0-
+
+
+
+
+
+
+
Filters (Sec)
+
+
+
+
+
+
+
+
1.0
+
+
+
-+
+
+..
x-hi = OFF
+
+
+
+
+4
+
+
+4 U-la =.020
+
+
+
+
+
+
+
+
R2=.OF Rotations (0) 05 -
+
+
+
+
+
+
+
+
+
+
+R2=
0.00 20
+
+
+
+
+
+
+
Z
+R3=
0.00
- 0.
4 4
4..
2 H
W R4= 0.00 0.56-
+
+
+
+
4-
+
+
+
+
W
~R5
=0.00 0
+
+
+
+
+
+
+
+
0.28-
+
+
+
+
t
+
+
+
+
Frequencies
+
+
+
+
- 4.
+
+
+
+
- 1) 2. 00 MHz Y
0.o.. --- + --.--.--.. +--.--
4-4--+
+.--------
- 2) 0
- Hz
+
+
+
+
+
+
+
+
Drive Levels
-0.28--
+
+
+
+
+
+
+
+
1)3.0Uolt
+
+
+
+.
+
2)0.Uolt
-0.56..
+
+
+
+
+
+
+
+
AutoStore OFF ALARMS I
I I
I I
I I
I I
Time (sec.)
0 4
8 12 16 20 Figure Al 4: Eddy current response of 20 mil. deep x 120 mil. length EDM notch Note: The above data is a stripchart display of the rotated impedance plane data. The rotation is such that liftoff is displayed horizontally and to the right, in the x direction. Consequently, a cracklike indication is predominantly vertical and downward, in the y direction.
COM4f
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 25
REFERENCES:
- 1) ASM Metals Handbook, Nondestructive Inspection and Quality Control, 8th Edition, Volume 11.
- 2) Blitz, J., V. J. Willstatter, S. R. Oaten, and N. T. Hajian, "Eddy-Current Surface Crack Sizing in Steel With High Lift-Off," NDT International, Vol. 20, No. 2, pp.
105-110, April 1987.
- 3) Christner, B. K., D. L. Long, and W. D. Rummel, Measurement and Correlation of Eddy Current Field Interactions with Material and Flaw Dimension Variables," Review of Proqress in Quantitative Nondestructive Evaluation,Vol.
7A, pp. 207-214, 1988.
- 4) Halliday, M. D., and C. J. Beevers, "Sizing and Location of Small Cracks in Holes Using Eddy Currents," NDT International,Vol. 21, No. 3, pp.167-170, June 1988.
- 5) Holt, C. C., and K. D. Boness, "Current Deflection at Cracks -- Some Insights from Modelling," Non-Destructive Testina Vol. 1, September 1987.
- 6) James, W., "Eddy-Current/Nondestructive Testing of Powder-Forged Components," adapted from Powder Metal Parts, Design and Implementation for Economics and Reliability. SAE International Congress and Exposition, Detroit, Michigan, March 1983.
- 7) Legai, R., E. Aznar,. and B. Verger, "Examination of Expanded Areas of Condenser Tubes by Means of Focused Ultrasonic Waves," Non-Destructive Testing, Vol. 1, September 1987.
- 8) McNab, A., and J. C. Hale, "Electromagnetic Crack Detection in Ferritic Steel," Journal of Nondestructive Evaluation Vol. 4, No. 3-4, pp. 165-175, December 1984.
- 9) Poulet, J. P., and M. Grozellier, "Automated Eddy Current Testing System,"
Non-Destructive Testing, Vol 1, September 1987.
- 10) Rummel, W. D., B. K. Christner and D. L. Long, "Assessment of Eddy Current Probe Interactions with Defect Geometry and Operating Parameter Variations,"
Review of Progress in Quantitative Nondestructive Evaluation Vol. 6A, pp. 705 712, 1986.
- 11) Rummel, W. D., B. K. Christner, and S. J. Mullen, "The Influence of Calibration and Acceptance Criteria on Crack Detection and Discrimination by Failure Analysis Associatesn
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 26 Eddy Current Techniques, " Review of Progress in Quantitative Nondestructive Evaluation, Vols. 5A and 58, Thompson, Donald 0., and Dale E. Chimenti, eds.,
Plenum Press.
- 12) Smith, E., "Characterization of EDM Notches and Real Fatigue Cracks in Flat Surfaces Using Uniform Field Eddy Current Technique," Material Evaluation. Vol. 43, No. 13, pp. 1640-1643, December 1985.
- 13) Van Drunen, G., and V. S. Cecco, "Recognizing Limitations in Eddy Current Testing.," Paper, Quality Manufacturing Conference, by Special Projects Div.,
Chalk River Nuclear Labs, November 1981. Atomic Energy of Canada, Ltd.
- 14) "Eddy-Current Examination DG1A Crankshaft San Onofre Nuclear Electric Generating Station August 1984", Failure Analysis Associates, Report No.
FaAA-84-10-02, October 1984.
- 15) "Eddy-Current Examination DG1 B Crankshaft San Onofre Nuclear Electric Generating Station August 1984", Failure Analysis Associates, Report No.
FaAA-84-10-24, December 1984.
- 16) "Evaluation of Transient Conditions on Emergency Diesel Generator Crankshafts at San Onofre Nuclear Generating Station, Unit 1," Failure Analysis Associates, Report No. FaAA-84-12-14, Rev. 1, April 1985.
Failure Analysis Associates inc
Mr. David Pilmer FaAA-SF-90-02-03 4/19/90 Revision 1.0 Page 27 ATTACHMENT Al.
FaAA Nondestructive Examination Procedure NDE 11.3 Eddy Current Inspection Procedure Main-Journal and Crank-Pin Oil Hole Failure Analysis Associates