ML19270B537
| ML19270B537 | |
| Person / Time | |
|---|---|
| Site: | Perry, Catawba, Harris, Grand Gulf, Byron, River Bend, Vogtle, Duane Arnold, San Onofre, Comanche Peak, Rancho Seco, Midland, Bellefonte, 05000000, Washington Public Power Supply System, Shoreham |
| Issue date: | 12/16/1983 |
| From: | Mathews C TRANSAMERICA DELAVAL, INC. |
| To: | Novak T Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19270B538 | List: |
| References | |
| NUDOCS 8312200154 | |
| Download: ML19270B537 (19) | |
Text
Transamerma I','""l",;'af,,*;',;,'o",b,io, gglgggl 55o 85th Avenue P.O. Box 2161 Q
Oakland, Califo,nia 94621 (415) 577 7400 tb December 16, 1983 Mr. T. M. Novak Assistant Director for Licensing Division of Licensing Of fice of Nuclear Reactor Regulation Nuclear Regulatory Commission Washington, DC 20555
Subject:
Standby Diesel Generators at Nuclear Power Plants
Reference:
Mr. T. M. Novak's Letter of December 1,1983
Dear Mr. Novak:
The Users' Group gave us a copy of the three-page list of nine questions at the November 30, 1983 meeting. This is the same list sent to us on November 29, 1983 by your Mr. R. Caruso. We are today sending the Users' Croup answers to the nine questions. Sinte the content of the nine questions is essentially the same as the list submitted by the referenced letter, we are sending you a copy of our answers to the nine questions asked by the Users' Group.
We trust that these nine answers are responsive to the list submitted by the referenced letter. However, if there are any additional questions, please don't hesitate to call. We intend to cooperate fully with the NRC and the Users' Group to answer all your questions.
Very truly yours, k
x C. S. Mathews Vice President and General Manager CSM/WVD/pn 8312200154 831216 k
NE9_QUESIl0NS Describe the history and evolution of crankshaft design of DSR-48 0 #1 diesel generators.
The DSR-48 diesel engine crankshaf t was developed from the DSR-38 A #1 engine which has been in production since the early fifties.
The DSR-38 was developed from the "Q" engine which was in production since the thirties.
The "O" engine had a 10" diameter crankpin and 11" diameter main Journal and was rated at 260,327 and 360 rpm. The R-8 engine started with an 11" diameter crankpin and 11" diameter main journal (11" x 11")
and changed to 11" x 13" and 12" x 13" during the course of evolution, from 300 rpm originally to 327, 360,
- 375, 400, 425 and 450 rpm. The first DSR-48 engine was built and shipped in 1969.
It was rated for 400 RPM operation. The first 450 RPM DSR-48 engines were built in 1975.
0 #2(a) - What prompted you to change the size of the crankpin after the Shoreham engines were built?
A $2(a) - TDI changed the crankpin diameter to achieve higher torsional stiffness.
This change to the engine was made to give broader capabilities as a driver for different applications, such as pump and marine drivers.
- Further, the change is a part of the evolutionary process.
For example the crankpin of the "RV" had been changed from 12" to 13" the previous year for the same reasons and not because of a problem with the 12" pin shaft.
0 #2(b) - When was the decision made to change the cr.n' pin size?
e A #2(b) - The drawing f or the 12" crankpin crankshaf t no. 03-310-05-AD was dated 2/4/75.
D #2(c) - Why was the crankpin fillet geometry changed?
A WP(c) - The "RV" crankshaf ts have a 3/4" fillet. When the change was made to the crankpin diameter of the "R" engine, TDI made the fillet radius change to again commonality in design between the R-48 and "RV".
The commonality is desirable from a manufacturing standpoint.
D #2(d) - When was LILCO informed of the change in crankpin size?
A #2(d) - Immediately following the crankshaf t failure at LILCO.The requirement quick supply of new crankshafts dictated the 12" diameter pin for a shaft be supplied because it was the only shaft immediately available.
What is the TDI mechanism for informing its customers of problems of 0 #3 product improvements? Does TDI use a technical information letter approach or its equivalent?
The TDI mechanism for informing it's customers of problems or product A #3 improvements is the Service Information Memo (SIM) program.
The SIM is to a Technical Information Letter with the additional advantage of an index system, which allows the collected SIMs to form a fourth volume of the Instruction Manuals.
Additionally, TDI informs nuclear plant customers of " potential defects" as required by Federal Law 10 CFR 21.
O #4 8 - In its report on the crankshaf t failure, LILCO's consultant noted that 4(a) the forcing function used by TDI in its torional analysis changed significantly between 1975 and 1983.
A #4 &
The torsional analysis for LILCD used the forcing functions which were 4(a) in the TDI computer program data base in 1974. We made two changes to the forcing functions values in 1975. The second change made in 1975 was used until 1977. In 1977, we made minor refinements to the forcing functions and these remain in use today.
C #4(b) - Why did this change occur?
6 #4(b) - The analytical resulte from the torsional analysis are verified by torsiograph tests.
Since the intention of the analysis is to accurately predict the natural frequencies and stress levels of the diesel generator system, input data to the computor program is
- adjusted, so that the calculations result in agreement with the test results.
The changes to the forcing functions are steps taken to match calculated or predicted values with those obtained from tests.
Our current forcing functions predict slightly higher stress levels than measured.
D #4(c) - What effect does this chagne have on any other components of DSR-48 engines?
A #4(c) - None.
The changes to
.he forcing functions were made to get the computer assisted computation to accurately predict the actual behavior of the engine generator shaft mass elastic system.
0 #4(d) - What forcing functions were used in the design of other TDI engines (such as the DSRV-16-4s) in nuclear service?
A #4(d) - The following tabulation shows what forcing function groups were used for all the TDI engines for nuclear service, CONTRACT CONTRACT TORSIONAL HARMONIC NUMBER NAME REPORT COEFFICIENT
__________________D91E________________0B992 74010 LILCO 7/18/74 1
75041 S. C. E.
6/27/75 1
75005 KU0SHENG 4/25/75 1
74046 CP&L 1975 2
74033 MP&L 9/15/75 3
75017 DUKE-CATAWBA 3
75051 C. E. I.
7/9/76 3
75084 WPPSS 8/16/76 3
75080 TVA-BELLEFONTE 4/27/76 3
76001 T.U.S.I.
1/5/76 3
77031 CONSUMERS 4/26/77 3
74039 GULF STATES 5/3/77 3
77024 TVA-STRIDE 8/16/77 4
8/1/78 4
78006 MAANSHAN 6/22/78 4
81015 S.M.U.D.
9/10/81 4
HARMONIC COEFFICIENT GROUP 1
1974 TO 1975 HARMONIC COEFFICIENT GROUP 2
1975 HARMONIC COEFFICIENT GROUP 3
1975 TO 1977 HARMONIC COEFFICIENT GROUP 4
1977 TO CURRENT U_9_B_5_9_N_1_G___G_9_E_E_E 1_G_1_E_N_T
_(_l,@_I_1_N_Q BEB199._____Zi_:_ZE___
__ZE
____ZE_:_ZZ_____77------
LISIING_EB95__LILG9_______G_E_1_L-M P-1_L______EIBIDE_
60Bd9NIG_____EB992.1._____QB992_E_____GB992.3 _____EB99E_i
.5 11.00 90.88 97.00 155.45 1
20.62 89.78 94.34 94.21
- 1. 5 19.00 94.88 100.70 129.21 2
24.06 45.43 42.53 42.61
- 2. 5 20.20 62.38 65.61 71.51 3
19.97 14.84 16.57 16.52
- 3. 5 16.70 38.91 40.61 42.72 4
13.30 29.04 20.25 27.62
- 4. 5 9.85 12.48 12.73 12.72 5
7.30 9.21 9.39 9.38
- 5. 5 5.65 7.01 7.14 7.14 6
4.18 5.55 5.68 5.68 6.5 3.29 4.39 4.49 4.49 7
2.66 3.60 3.69 3.68
- 7. 5 2.23 2.98 3.05 3.04 8
1.87 2.46 2.52 2.52
-3_
8.5 1,61 2.20 2.26 2.26 9
1.42 1.92 1.97 1.97 9.5 1.25 1.50 1.53 1.52 10 1.11 1.25 1.27 1.27 10.5 1.00 1.13 1.14 1.14 11
.91 1.01 1.02 1.01 11.5
.82
.88
.89
.89 12
.74
.78
.79
.79 0 #4(e) - Have these forcing functions changed?
A #4(e) - The current Tn values have been in use since 1977.
0 #4(f) - Please describe the development of the forcing functions for each TDI diesel in nuclear service.
A #4(f) - The forcing functions are derived from Fourier analysis of the torque vs crank angle diagram for one cylinder. These forcing functions are subsequently adjusted to correlate the analytical results with test results as already noted.
Since all the TDI engines for nuclear service are rated at 225 bmep and 450 rpm, (except for S.C.E.) which has a lower rated RV-20-4, the forcing functions are similar.
0 #5(a) - What does TDI view as the reason for the Shoreham crankshaft failure?
A #5(a) - Site operating stresses approximately equal to the endurance limit caused high cycle (10.
06 to 10.
07) fatigue failure of the crankshaft.
D #5(b) - What conclusions has TDI drawn from the LILCO failure report?
A #5(b) - The operating stresses in the 11 x 13 cranksaft were essentially equivalent to the endurance strength and results in high cycle (10 E 06 to 10 E 07) fatigue failure.
The failure is effectively an unfortunate endurance limit test. Even a small reduction in stress (perhaps only 2 or 3 percent) would have resulted in unlimited life.
Since the 12" x 13" crankshaft is subjected to significantly reduced stresses it will result unquestionably in a shaft that will give unlimited life.
In these matters we are in complete agreement with the LILCO/ Failure Analysis Associates report. However, we do not feel that the FaAA analytical analysis, particularly the finite element model (Sec.
6 o'
report) is necessarily satisfactory. It fails to predict the actual r ate of stress measured by Stone & Webster (Sec.
- 4) and it fails to satisfactorily predict the crack location and direction. The crankshaft stress analysis is inadequate and therefore does not fully explain the reason for failure.
TDI is currently engaged in its own stress analysin program, which is expected to yield a more accurate analytical model and a clear understanding of the stresses which caused the failure.
0 #5(c) - What actions has TDI taken or does TDI plan to take for Shoreham and other plants as a result of the Shoreham crankshaf t failure?
A #5(c) - TDI plans to continue its investigation into the reason (s) for the Shoreham crankshaft failure in accordance with the outline given in the discussion presented by Mr. Greg Beshouri (attached). The results f these investigations will be published at the appropriate time and mace available to all interested parties.
D #5(d) - Does TDI plan to prepare a report of its own regarding the Shoreham crankshaft failure?
A #5(d) - TDI will develop a formal report containing it's views on the reasons for the failures.
Much of the report will be developed using understandings gained from the R&D studies outlined above.
0 #6(a) - Describe how TDI design calculations are reviewed and independently verified?
A e6(a) - Calculations performed by design engineers are reviewed, signed and dated by the Manager of Design Engineering.
Designs which rely on calculations in which assumptions cannot be verified are subject to experimental testing by the Research and Development group. In some instances the Manager of Applied Mechanics will also review the result. Some components are subjected to testing on a shaker table, if practical.
D #6(b) - What detailed stress analysis of the crank web and pin were performed?
A #6(b) - No detailed analysis were done on the crankshaft other than the crankshaft was designed to American Bureau of Shipping Rules, as detailed in the attached excerpt from the rule book.
TDI has successfully used such rules as a design standard for 45 years. The R-48 crankshaf t was developed from the "Q" engine with 10" x 11" (10" diameter crankpin and 11" diameter main journal) to the first "R" engines with 11" x 11" crankshaf ts then 11" x 13" to the current 12" x 13" configuration.
Could the problem with the crankshaft have been detected during D #7 initial torsiograph testing at the factory?
No.
The total vibratory amplitude measured was only + or
.50 deg.
A #7 which equates to a stress of 5314 psi. Th9 portion attributed to the
+ or
.43 deg. or 4570 psi, well within the 5000 fourth order was psi allowed by DEMA for single order contribution. (The stress required to break the crankshaf t was more on the order of + or - 30 to 35 ksi.)
D #8(a) (i) -LILCO has also identified problems with failures of the diesel engine connecting rod bearings. We understand that they have provided you with a copy of their initial report on the subject. (a) What does TDI view as the reason for the Shoreham bearings failure?
A #B(a) - Four of the forty bearing shells were reported to be cracked, only one of which had a significant crack through the edge of the top shell.
The small piece 4-7/16" long and 11/16" wide at the thickest point was Jacked apart from the main body of the bearing shell for study.
None of these shells had failed to the extent that the clearances were opened up nce did any of the shells result in damage to the crankpin.
A photo-micrograph of this broken bearing showed porosit y ranging from 0.01 to 0.03 in diameter. In addition, the material was found to be below standard for elongation.
An examination of the fracttre surface with scanning electron microscopy identified some of these voids as the apparent crack initiation locations.
In compression the porosity would not pose a problem.
- However, the ovet hung bearing arrangement resulting from a 1/4"
~
chamfer on the connecting rod as shown in Figure 1.1 (attached) in conjunction with the normal yawing of the crankshaf t, put the I.D. of the bearing into tension.
The surface porosity acted as stress int ensifiers and with the poor material elongation characteristics, initiated a crack.
This was clearly a material rather than design problem as evidenced by the fact that more than 300 cylinders of this connecting rod arrangement are in operation, many of which have operated for more than 25,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> without bearing problems.
O hB(b) - What action has TDI taken to ensure that new bearing will not fail in a similar fashion?
A #B(b) - In regard to LILCO and other R-48 engines installed in emergency standby service, the crankshaft is fitted with a connecting rod which has a smaller 1/16",
chamfer on the edges. Figure 1.2 (attached) shows the bearing is fully supported. Even though there may be some porosity in the bearing shell material, the shell is in compression and therefore minor porosity would not be detrimental.
Chemical and physical properites of casting lots or heats are tested to verify compliance with requirements.
In addition, TDI does a visual inspection for porosity of each shell during manufacture.
0 #8(c) - What other TDI engines in nuclear service use similar bearing material?
A #8(c) - All of the nuclear and commercial engines which TDI manufactures contain bearings using identical B-850-T5 bearing material. This material as a 6% tin content aluminum alloy. We purchase castings from Aluminum Company of America and perform all machining and plating operations at the Oakland facility.
O eB(d) - What action has TDI taken for other engines to preclude their failure in a similar fashion?
A #8(d) - The bearings in all other engines in nuclear service have connecting rod and bearing arrangements shown in Figure 1.2 and 1.3 (attached).
Each provides full support for the bearing shell. The bearing shell is in compression, both from crush and operating
- forces, with no portion of the shell in tension. Therefore, if bearing material contains minor porosity, as all castings do, the loads present will not act with the stress intensifiers and result in Cracks.
What controls has TDI provided on bearing material in the past?
D #8(e)
A #8(e) - The purchase order for bearing material has required the supplier to furnish a Certified Material Test Report (CMTR).
This was a requirement in 1974 and still is. The CMTR is reviewed for compliance to the material requirements. All bearings are inspected visually for porosity during the manufacturing process.
O ue(f) - How did and does TDI ensure that bearing material meets its tpecifications?
A #S(f) - In 1975 TDI initiated it's own bearing material sample testing program to check checmical and Physical properties against specification and the CMTR supplied by the vendor. This program remains TDI's standard practice.
D #6(g) - What other experience has TDI had with connecting rod bearing failures, of any kind, in any nuclear or non-nuclear installations?
A #8(g) - TDI customers have encountered occasional babbitt fatigue. It has the appearance of small worm holes in the surface of the babbitt. In addition several users have suffered the results of faulty reinstallation, dirt ingestion and abuse which have resulted in bearing failure.
D #8(h) - What procedures does TDI use to ensure that bearings and journals are properly designed and manufactured?
A $8(h) - TDI has been designing, developing and building engines since before 1938. The intervening years have provided considerable experience and knowledge regarding what constitutes a properly designed crankshaf t journal and mating bearing, such as L/D ratio, surface finish, babbitt thickness, etc. In addition, we work closely with the bearing material vendors regarding the bearing design.
All of this information culmanates in a design that is translated into detail drawings for manuf acturing. The DA department ensures conformance to the drawing requirements through TDI's 10CFR50B program. The bearing material vendor provides Certified Material Test Reports (CMTR's) for each casting heat which are review for conformance to the drawing requirements and are verified by TDI's own Chemical & Physical test for each casting heat.
D #8(i) - Deteribe any problems you or any of your customers have encountered with the use or manufacture of aluminum bearings with babbitt overlays.
A #8(i) - Users have occasionally encountered babbit fatigue in the bearing overlay. This may occur if the tin content in the babbitt is too low, resulting in a weaker babbit.
The composition of the babbitt is monitored quite closely.
TDI has initiated a chanDe in the babbitt composition to further improve the f atigue resistance. This calls for the inclusion of 2.75 - 3% copper in the S. A.E. - 19 babbitt. With the TDI bearing design, babbitt fatigue or even complete babbit overlay loss does not result in any sort of catastrophic bearing failure that might cause the engine to stop functioning properly. TDI has also occasionally encountered porosity and low elongation charact erist ics in the aluminum castings used to manufacture the bearing shells.
D #9 -
LILCO as also identified problems with cracks in almost all of the piston skirts at Shoreham.
A #9 -
This statement is incorrect. There has been only one piston at LILCO which has been identified as having a crack. The examinations being conducted at the site are using an " eddy current" inspection process which TDI and it's Metallurgy Consultant considers not suitable for e x rninat i on of cast nodular iron surfaces. This eddy current process has predicted linear indications in the piston skirts which in most cases may be nothing more than the grain boundaries within the nodular iron structure.
D #9(a) - Describe the stress analysis and testing that has been done by TDI in the development of type AF, AN and AE pistons.
A #9(a) - AF, R ', and AE pistons have been subjected to many experimental test programs to reveal the patterns of stress and temperature existing in the assembly.
The tests included studies of thermal distortion, effects of corobust ion pressure and inertia forces. Finite element analysis (FEM) was attempted on a crown, however the technique proved to be less than adequate.
Piston attemblies of the AN and AE type were successfully run for 687 hours0.00795 days <br />0.191 hours <br />0.00114 weeks <br />2.614035e-4 months <br /> in the experimental RS-V12 engine at 514 rpm and a power level of 937 BHP per cylinder to support the results of the static and ar.a lyt ical studies.
Nuclear standby generator diesels are rated at 450 rpm and 609 BHP per cylinder. Therefore the test work subjected the pistons to considerably higher operating stresses than the pistons used in any standby engine.
C #9(b) - Has TDI or any of its c'.:tomers encountered similar or different problems with piston cracking?
A #9(b) - The crack reported by LILCO is the first such crack identified and modified ("AF" style piston skirt which has been reported on the manufactured in accordance with design requirements. There are 252 modified "AF" piston sRirts operating, which have accumulated in excess of 1,772,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of successful operation.
-B-
The "AN" style piston has experienced several field failures, which have been attributed to high residual stresses not removod by a stress relief process.
There have been no reported failures of the "AN" style piston which have been stress relieved and properly m ach ir.ed.
There are 1374 "AN" style pistons opeatatng which have accumulated in excess of 2,760,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> of successful operation.
The "AE" style piston is the latest TDI R-4 piston design and incorporates prior R-4 design and operating experience and new design knowledge we have gained through our R-5 engine test program. The "AE" piston design has been successfully tested in our R-5 test engine at 514 rpm and 302 BMEP and has acquired in excess of 7000 operating hours in a 16 cylinder 7000 kw engine in the field.
D #3(c) - Has TDI modified its piston skirt design to improve stress levels in the area of the bolt holes?
A #9(c) - As part of a continuing program of product performance and reliability improvements, TDI has modified the piston skirt design to improve stress distribution in the area of the fastener holes and in the circumferential mid rib blend to the wrist pin boss.
D #9(d) - How and when were these modifications made?
A #3(d) - Primarily as a result of the studies referred to in the answer to quest ion 9a, TD1 concluded that a more massive boss around the bolthole would better diffuse forces to the piston pin area.
Calculations alsc verified that the protection afforded the fasteners against cyclic loading could he achieved with only 13 believille washers instead of the original 26 washers.
On August 10,
- 1982, piston skirt 02-341-04-AE was released for production. It required that a change be made to the corebox in which the mold for the piston skirt interior is formed. This change provided the more massive bosses around the boltholes and precluded the manufacture of earlier designs.
9-
_U OL'$.iBil ElEEEE_969x1E15_EEEE365 Greg E<echouri, Recearen Engineer
~ T_i;MHO.:.
nath tnc follure of the 11" x 12" cranksnaf t s in t ne LI6.CO D5R :48 (5/N 79 M /.1)er.gine, TDI ir.s t r at ed a strett analysis prograrn (including physical t e s,t a n;, a '.t' en !ytical rocce]Ing) witn tSe o:Jective of determining tne stresses e r..
thcar tocret in an 8 tnrou 11" v.13" crankthaf t in orcer to identify the e t..v.
cauren of tr.e failure of tne L.I.C0 snaf ts. In ecdition, this program as raore scp11stacated input for future crankshaft stress a n'; e r.t c c to crosace a i: Myst: erc c e n i g r..
~ 3 5f.d.5 ate:n.c.w Scacu
.c r te tne a ra t iat on of pnyoical t est in;;, an ext ensive review of the
.s
.11.;e 11;ciat>re as concuct ec.
From thic review we ceterminet, as
(
.r:;t.,
t-a t rie :.ni f t in so vice is suojectati to a corcolex, cynamic state 3
a...nt: t t r e r. :. c a. ~w s ey to succettf vf. st r ens analysts is an underttancing c - t orce of eacn strers: componer t end now tnece incividual components aco "ht cuaa:nec ctress tate.
The 11 terat ure incicatec the netetsity of s'
ran s.;e t e.. 2 r.,.
The tecnnical papers c:so were a good source of v.% t.. i. t. r re.L>rchert het vtc;
- 1. regarc to g,ege typc, length
~
.s..
li_;. M '3 d ia.d_3 G iu r c : c '.,
lnt 1ccratvre review conf i,.e: t-tne need for steaan gage
- t ;.
Et th
- cpt.n n; cf our inv e ci. i;at i on, Stone and Webster (LILCO
..r
'. : n t :. ' -i c :. 4. i o c,. :Oraattet t.nta r c i vet to concucting dynamic strain tjage
. et.. r ;, a a;.,
fe;. % te very cifficuit to follow because of tne anyriad of tr.....' r t e.. tion m.:;em:. etsociate: with freceeney modulated (FM telemetry (a
..'..t cf t ret :..i t t a n;,
strain inf ormation via ratio waves from the operating e.. u f '. ).
. c e ore, ve elected t. per form static strain ge;;e testing on an f
. '...../. c e r.;. nu :4 v.t r f atalit y in the booe it would cori.plement tne S&w
.. %n. c
. c ;.t. r.g.
T ra t static tecting wat tiesi gt.ed to provice informatron re..:.. t. :, t o i.. e :.f rc '. er.c verify the festability of the cynamte test cata.
N t e:t a rg, wet cone c n a TDI Fecearen and Developteent engint-with a 11" x
- " c cr. w ft.
r.e crane:;,ha f t 12 sin 11er to but not identical to tne 11" x 13" theft wnacs f ct :ed a.t.I C3. Tne cranxsmaf t of inte unit was statically loated c it.x. r *. e c y arc.a c f ccc c. I n t ne 5t h through 8th tnrow of an R-48 engine rated II: ::. E PO' a 4M R:?'.
~'.r' cr e.na re f t ;o..cc fr om gt L :ressure (le:s inert ia), torcue transmission
.2 1.19 :.
.lbretier. wtre first sisiu;atte indepencently. Tney were tnen
- .t ;.s. i n Es.er:. c i f f er ent co n02 nat ions to ceterr. ine tne resulting stress.
rom
.'1r c:tc 4
qer.o r
- .el e t t er. v.at obtained of which it is possible to creclet en ;ncft L rn;
.~c r eny comosnaticn of Derid i ng a nc' torstonel a:
i
- s., i. c t..
2Ed 6222 LOG The cr an :,5af t was outje: tee to toroue in such a enanner as to samulate
- :n i c ne '. st'estet frv n t raro.r it tcd toroue and f rom torstor.al vibrat ion, and to x c. r.;, %? cer, s a raulat i n;, ger prer,ture ( l e s t.
2nertia).
T%e r.ut e. s a r y tercic was gereratec by fatting cylinders No.
286 t ;ea tr. e vo; u ne: witn spacers anc 0-rang seals (see Figures 1 & 2), and tnen
.n t : s u r i r.;.
t her, with cIl with tne two pittons located at 240 ceg. anc 120 ceg.
T.~TC re;::-:t i ve; y.
E.
. ;at.In ber di ng force was ge r. erat ec oy staa l i n g and pressur)zang
.. 2 :.c t '
- e.
3 with the p :t:n at ~DC and 10 cegrees and 20 A: DC.
E.:. e n Cr N jttygg p
~9e t**c::en generate: cy t t r cive and ber. ding forces were measurec Dy e t :l t unc e t y ;c: ttrain p r ea:,.ocatta on tne ho. 3 crankpin fillet ar.d on tne
.y
.p. n
( s t.
7: La 3 2 i
4).
- csttter, E throurn E locatec in tne fillets
.. c re s. i r ur ti ra:nc arn-t,c;r principle cirection. Rosettes A & F mear ureo
... ar d ter.cin; or t,e ce: 'are c f t ne f ree put c.f t ne pin. Corcoarison of f4 t
(#cre part of t r.c p a r.)
witn E tarough E (fillet), yielos the stress t.
m
.n 2 en e'et : of t ne crar r.oln-fillet-wes configurat ions.
- ..
- t c. n (ca.. s o t e r t.c p;et uce:) located on t9e cylindrical surface of tne veriftet tie ectue". torcae indurec in t ne syst era.
2 n. n p;.. n t :
- i. '. '
c. t ' '. e
.ce rut, yi.ler t5 rec
- 3;e tyre of 0.12D" (3 mm) effective as,,
o.i ".'c; '. e t by i c e o-%ct ure;i ent s (P/N CE4-06-12 W R-120).
".5 il_ i A % 2 Nee.:, sac.r. w n "Itct titu lc.o: by prettering cylinders 2 8 6 only iri 300 in:ce.co. t s f rom e to 300 psi yleicing torques up to 2,360,000 in.lef.
3.n e
- <r:c.ng at TD; wtt. t9en rirulc.tec ey pressuring cylinder ho. 3 only, 1*.
G? ;.
Incrarente f rom 0 to 1600 psi, representing a maximurc peak firing
.'. e :.. t e 46 erc::: of S?3 psi, (Notc tnat tne primary and seconoay inertia
'erers 0
' it pt tor, anc connecting roc assembly oppose tne firing pressures 4
rr Ove ' i e: vivaler.t ef f ect c.f loering t ne fi+ 1ng pressures by 377 osi wnen
- .
- < :'.ar 6 at 450 rpc.
Ine r.,
my o"eE:uring cylinders 2, 3 and 6 aporoorlately, combinec torsion e r.d cent.r.; reertzentin; the actual stress state was simulatec.
"w : n o orn:Ing ent comnined bencing anc t orsion test s were receated witn
.. a r.: c r '.c. 3 pitton lect.tec at 13 anc 20 cou1 valent cegrees ATDC.
,... e#
.. l.. L.
.'a.4 3 4 3 7 *. e,* rain Eig2 tota were recuted to maximum and minamum princtoie r
- r e = :. a n: :er a nci;' e direct ior.c.
2
~n ou bei22ng load anc torque cata, general matherratical expressions were c o r s '. c :'
for nou;nal ctreLsts in the free part of tne pin due to bencanc loads t.. t..r c.' e.
In at: I t a cr',
stress concer.t r a t i on factors were calculatec for
,+';ous f;ilot locations.
Pcm tne combined ctreet cat a, a general analytical technique (using
- . v ' t.
circit) for calculeting corebaned stress oue to a given bending load and
. re m war generatec.
It wat tnen ec.nfit raed tnat this techntave could oe n ! : c-J an re.orte, i.e., given a certain comolnec stress state, the bencing
- 1. ? :
a'.d t orc;ue creat ar g this strebs could be calculated.
Li. r.;,
t'u n technic;ue the cyna:cic stress cata taken by Stone & Weoster on Et
- C I,
~D1 5/A 74011, at Ll_CD were t sen DroWen cown into com3onents of cynalaic torcue creating tne stress, resulting in a
- . n s,.. c :ending load a nc.
-!e0-uncer tt e nc'2 r.[
of t nc dyneroc st ate of stress cornaonents. hed tne static
- ting not beer ccicuttcc, it w.alc not have been postab;e to satisfactorily
- .. a it-r t.e c'yne.u c h t t eba d at e t a k en by 5d n'.
.as-a..
./. q,v m..--.
a- :
3: t ?
c vf t9e
- e j i ut o cf t se Ll CO CranR0naft' and a recuirement to
~
c.ce.y vi.cccc'.anc..e restor's for tne feilure and given tne success to cate of
~
,'r.
It U e.
'1ct 9 0c t of C r a r., h a e #1 stres0 analyLib a?D1190 t o tne 11" x 13'
- tv0 c olc.i.11 t t L o.'rEelvEE to an on-going Crandsqaf t stress analysis
.e'L, we 3..
~' 1:
- c.. ;, r e n..
- i...,. ; c e. ; r, t wc-cor.:..) ; rce r.t a ry c it eCt i ons, analyt1 Cal anc
. ;r
...E.t er. i i' i g. UJ.
V.i ot e ; w '. e,t
- 111 :e conovet ec. on ell cran <s7af t configurat ions to
- e r.
r.e
' n ea cvr.c ent rat ion factors in torston anc bencing. Concurrently, a a.
0.
e.
",31cn prettets torc;ue anc cencing loao w11.1 De tenerateo. Tne oncer.s. r ; 1on f ac*. ors an:] dynamic Dencing loac anc torque calculations
'e m;
n.,
a s v ; t. e 2 r.:.v t for a seconc' mccol wn2cn will calculate crankshaft
.. cec _
v c ro r., ent ic. 'i n i t celculation will t,en ce ver1 fled by cynamic strain
,,.. ; e
!. c n c un coleneo et en. eof t configut ations. Once tne ste ess calculation
- .. c
- . r c at w ri f 2 e:.,
treet ocn ttaa stress and exact operating factors of t.
ecy f..
er.y crenc:. eft config/tation can se calculated.
r a ccc tics to provicint cotign Infcreation for new engine prograrns, the cc.tvletion p-ocecute will De u s t.c' to confirm and refine tne less to;'uct:ceter, ra:>re contervative stress calculation procecurcs currently useo on
-4 c: aca en;1ncs a n nuclear t ert:ce.
- bvi:ut:y, the oyn;<mit tett on tne 12" x
13" crank.chaft, current l y t r s*. a. i e r i n DE R-6 6 er.g1not, will oe en imp rtant step in t his prograre. Static t e n t n:,
raet t t;rc Oc contucted of. tqis snaf t in order to prooerly cecioner tne c;,;u ;c cet: Locttcc to ocerue from tests planned for late tnis year.
='
O5t N'd.
bk 11 dit F.s Q
/
<?V b
k u"
J N
1
\\\\
c x.
E
\\
()
T
\\
t
%u T
g V
J
\\
u i
o 3
o s
er
()
6 NEh 0
1 i
e b./
1..
LER b$
l be 4
e F/GURE &
.SEALwc, OF CVL/NOGKS /t.7, 3 & 4 cirAKANGE VOLVA1ES F/LLGO wsTH CNctxt Oil p
i f
IIEAD
)
J i
HEAD l
l j
L M00/F/EU CROWN
'l~C - RING, P/Af toA9E'6 wsrn-L P/N Jr-c.U*-otr 3TANDARO SK/KT sPACEK PLATE A 55M1'O TO AA/
p/sr An-M36
'~
0-xtwGeo cacwAre c.ot.ownf ar4MPALO I
\\
en 3MRT 4 5 Skl.
}
O'
'V
\\
n
~
zi TVP/c4L c H/NOER,
.D TPP/C44, CPt /A/CElz$
No. 3 AT TDC
^'04 46 &
AT a% A TDC.
N 6 / 3 0 *.A TO C Gsrx sof.v/st e
FIGURG 3 - C,A GE LOC 41~/ oat A/O.3 GA4A/K PbV, VtEwsp pgo,3 UNOEp StOE OF P/Af f 4-d ouRN AL
- 3 CRANKPIN oncies or RosETTEE N
ROT N TIO N
~-
8
's %Wh !
3 7
wt
,),,,c]P t y y'c1
, OF:
C.RAN K EMP PT
- p*
/
/ ( FIN
'I rr +
b_.
R0stT7E'A
{
$t t.tT G AG ES g
TYPE r.ALI D.TRtPtW e 5,.--
)',
/0/r3/83
e s
t h
Q >-
.O/
q L..
6k J
&/
~
c-i
/
U i M *n '
/,s' w ] :2 l'
E
. s= =
a a
i
,/, /
{yg
,k,,, 4_ _g g-
-u w-a,
_2---.
///
g
/9f Q
/
7 J
/
$E
'/ /
bk E
\\,!.
/,,,, y g
=
w
.g a
~.,
M yf a
n e
h i,i/
e g
5 s
y
.:-lV:
- --. - =,
I T
- -t e4
,?/,
,v-
- 7. g i.
o
.L.J v
y N
j
,s us' y
\\
Q C
t AF V
E 0
5 4
- g. E e y
- (
WMI".o aG u.
o<
O rv
.C E *b b
Y O
a e 4
E a
.b D
ea s
k R
@u
~
d 2
9 1?>
- $.8 B e 30 s
d t
1[
Fis. 4 7
CRANK 5 HAFT STRESS ANALYSIS PROGRAM ANALYTICAL EXPERIMENTAL i Generate i
4 Forfore e I Dynemac s
- Statae i Torque &
6 e Testa e
e Bendano Model i
I...........e 4
e s
e y T.M vs 6 I Any Crankahaft I
e
- Stress Concentrataon i Factore ht. Er 8
8.............qy..............
8 a
Strema e
All Crankshafte I
Model e
i e
8 F vs e YAnyCrankahaft 4
e 4
F vs e Predact Maxaaus e
..............y
............Ii Stress & F.5.
e i
i Select Crankahafts t................e e
i i
Proven Analytacal I Vertfacetton i
Calculataon Procedure i Dynaeac' Tests i For Stresa Analys6e 8...............
4 of Any Crankshaft.
GMB/wam 12/5/83 F/ce UAE S-FLOW GNAA7" cA4wp:SMA N.sTPESS M4Lt'/5 Tis. 5 a
BEARINCn COM PA RISON CONNEC TING ROD 03-340 O C
,hx45'CHrx
, x45* Cur.
BRG. SHELL r,
03 340-03 0A
%,N A = 6 9.62 /NS 2
'Nw A (SUPPORTED)s 65.50 INS II~D I A.
BEARtWG L oAo
- 4 215 F.
R (II" CRAMKPIN)
Fis i. :
CONNECTlHG ROD 03-340 A A
,$ s 4 5' C H r.
W Jh-A 45
- CHF.
~
'~""
A
- 75.95 INS 2(tts.g',
12"O f A.
BEAE'iys toAo. 3(.25 Ps R clz" CRANKPIN)
Fic.i.z.
CONNECTING ROD 02-340-II-AJ
,jx 4 5 *cHT.N 545*CHF.
/
gg BRG. SHELL 02-340-04-AG A s *l8.t7 INS l { \\ t9.57.)
~
^-
brARING LOAD E970 PSI 13DI A.
gy (I5" CRANKPIN)
Fic,. i.3