ML20087B764

From kanterella
Jump to navigation Jump to search
Part 21 Rept Re Eight Camshaft Locating Studs in Diesel Engine Found Broken & Three Studs Found Cracked.Caused by Fatigue Loading & Torsional Vibration.Justification for Continued Operation Encl
ML20087B764
Person / Time
Site: San Onofre  Southern California Edison icon.png
Issue date: 10/11/1991
From: Matthew Mitchell
MKW POWER SYSTEMS, INC.
To: Murley T
Office of Nuclear Reactor Regulation
Shared Package
ML20087B753 List:
References
REF-PT21-91 NUDOCS 9201130277
Download: ML20087B764 (5)


Text

. _ . . _ . . _ _ . . _ _ _ . . _ - _ _ . _ _ . _ _ . _ . . _ _ . . - _ . _ . _ . - . _ _ _ _ . . . .

MKW NMs3MiffMBhnc. 1 1

L.r.... ..

j i

1 October 11, 1991 '

i Mr. Thomas Murley Directcc - Office of Nuclear Reactor' Regulation 11555 Rockville Pike Rockville,-MD 20852

Subject:

Potential Reportable Defect for EMD 645 Series - -

Camshaft Locating Studs:(EMD #9034672) s

Dear Sir:

[

On 10/10/.91, MKW Power Systems received notification, upon inspection:of San Onofre Unit #3~ diesel engine,-that_one:'(1) ,

of the eight (8) camshaft locuting-studs.(EMD #8034572). was found broken. Three (3) other studsiwere found cracked.-

The camshaft flange which contains the studs is located at-the drivegear end of the-left bank camshaft.

Inspection by Southern California; Edison reveals that the-cracks were caused'by fatigue loading. It'was also found that the bolt materialJcontained a concentration of sulfur '

stringers which may have. reduced the-bolt's: mechanical properties. _ Inspection of the right bank. camshaft.atuds revealed no evidence of cracking.

Preliminary evaluation of the SCE report (attached) by EMD has yielded the belief that-the: cracks were the resultrof '

thei5th harmonic' torsional 1 vibration-(previouslyLreported.by MK/PSD- on 5/10/88). translated to the--left-bank camshaft.

This is contrary to the conclusion 1ofLthe'SCE1 report.-

I j These studs have been manufastured usingEthe same mechanical process since 1939 without any knownLfatigue. cracking failures. Thereforer:we#believe thatLthis failure =is- -

specific to the SCE m San'Onofre diesellgenerator unitsLand is a result'of the15th' harmonic problem;which1is currently ,

j .being corrected on site'.

Upon receiptoof2 final o resul ts f rom lEMD ,1a. f ollow-up_- report .

will be issued.

L Yours very-truly_,-

u MKW POWER-SYSTEMS, INC.

P/W /ne

! M. Vann Mitchell Quality Manager MVM:jb-Post Office Box 1028 e Rocky Mount. North Carohna 27802-1928. Phone: (919) 977 2720 9201130277 91101p TWX: (510) 929-0725 e FAX: (919) 446-3830 PDR ADOCK 03000361 S PDR

.- _ ~ _ . , - _ . _ _ . _ _ , _ . . . . . . ,,__,_..._ - ._ , _ ,. _ .., _ . _ _ . _ ,, _ ,_ _ _

, v. u, . , ... a , . , .~u .. c s a n .~. . ~ , a w.v ,

'N h) lp ,

    • ' ann.

.r . - - [d;~.kp)l,'$7h u-

~-~

[N (K ~

g r

r CCI - 4 l99h

, - .I _j! ---

i w ._.

- = .; -

O $.?!!$l'KX SOUTHERN C A LIF O RNI A EDISON SAN ONOFRE hUCLEAR GENERATING STATION STATION TEC11NICAL FAX NO: (714) 366 6249 VERIFICATION: (714 36S.931S Number of pages (including coicr shect)

Date: ./0[ 9/ . - ,

IOl MO , _

rrmn: Will M ._ k _ _

uc,saac: _._ . . . _ . _ . __

__ .. 5%oa3 foor .

r Ak k sa

-- - - - - . ~ . .

e.- .96 6-MW

  • -YW O' . M. .

i i l

l .. . . . . . _ . . _ _ . , , ....__ _ . __. . . . _. ._, .__..._.__ .. . _ , .

,pg, m , , , ,gp,,,,,.p 496.,p.*4 e.e.=w, q,yg,, .

8"*8"*" PNe e 4 .6 e.-s 6ee

[1ki) From: DILL STROM at ROOF 3 10/3/91 6tO7PM (7787 bytes: 147 in)

Tot SHERRY KUNZ at AWS2 cci CHONG CHIU, BRI AN KATZ, ROBERT PLAPPERT at AWS, MIKE SHORT at AWS, 1 JAMES REILLY at NESL4, MOSTAFA MOSTAFA, BILL STROM, HOWARD SCHUTTER at AWSS, JOHN MIRSCM at AWS5

Subject:

3ggg - final version for NCR dicposition j

______. .____l__JCO___.... __.----- Message contents ------------ -----------------

NCR 9110 0002 JUSTIFICATION FOR CONTINUED OPERATION FOR 3 GOO 3 IDENTIFIED DEFICIENCIES During an inspection of 3 GOO 3 cam shaft, one of the four cam shaft flange dowel bolts were found broken and the other three bolts were found cracked. The cam shaft flange that contains degraded dowel bolts is located at the drive gear end of the left camshaft.

FACTS AND EVIDENCE As a result of the identified deficiencies,-the following facts and evidence were collectedt (1) The total run time for the degraded dowel bolts _is about 164.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> since their original installation. The total number of starts i

the degraded dowel bolts have experienced is Approximately 95.

(3) The material of all the degraded dowel bolts is AISI 1144 ETD (high temperature drawn) steel which is manganese carbon steel.

(3) There are no other deficiencies idantified in 3 GOO 3. All the bearing cap bolts are tight (preload greater than-28 ft-lbf with a-l original 32 ft-lbf preload). All the anchor bolts on the " pork chop" l

bracket (provides support between the camshaf t and drive gear) are l tight. No cracks were found on the dowel bolts in-a flange next to the flange containing degraded bolts (4) Based on past operating experience', both 2 GOO 3 and 3 GOO 3-(not.

2 GOO 2 and 3G002) have experienced several " pork chop" bracket failures. The cause of the-failures was diagnosed-by the Diesel Generator vendor to be 5th order torsional harmonics which only exist- ,

l in the 20 cylinder engine. The brackets in all four diesel l generators were improved in 1988 and 1989. The improved j brackets have not failed since the. modification.

(5) The 5th harmonic torsional vibration problem will be corrected with appropriate design modifications for 3 GOO 3 in the upcoming refueling outage (February, 1992).

(6) The flange bolts in the flange near_the gear on the right hand side esm shaft (note that the flange that contains broken and cracked bolts is on the left hand side cam shaft) were examined by dye penetrant.:No cracks were detected.

FRACTOGRAPHIC EXAMINATION OF THE DEGRADED BOLh3

~

l The fracture surfaces of the broken bolt and three cracked bolts were examined under Scanning Electron Microscope. The results of the examination revealed the following facts:

(1) The failure mode of the broken bolt is a fatiguo crack followed by overload. There is a small zone (about 20% of the total crosc-sectional area) containing striation marks that are characteristics of fatigue fracture. 50% of the fracture surface was covered by a fatigue related failure, consisting of fracture along sulfide stringers in the longitudenal direction coupled with fatigue in the transverso direction. This fact l

indicates that the bolt material was severely degraded with a high j density of sulfur stringerb, which in turn had reduced the transverse l

mechanical properties. The remaining 30% of the fracture surface was a ductile overload.

(2) The failure mode of the three cracked bolts was pure fatigue. When the bolts were removed, the break-away torque had been reduced from a nominal 32 ft-lbs to 12 ft-lbs on the two bolts which exhibited substantial fatigue cracks. The third bolt had a fatigue crack about 5-10% through the diameter, and had a break-away torque of 30 .it-lbs.

The number of beach marks (which indicate how many interruptions of sjectrum loading occurred) on the fracture surface were counted. ,

There are 20 to 25 beach marks, indicating that there have been 20 to 25 stops of spectrum loading after crack initiation.

(3) AISI 1144 dovel bolts have been used in the EMD diesel generators since 1939 with a very good track record. Moreover, in the past, l

there have been several bracket f ailures, and none of the dewel bolts fractured. This fact indicates this broken bolt is a unique failure.

MOST LIKELY FAILURE SCENARIO 1

Based on the facti collected above, the most likely failure scenario is statt.d belows (1) One of the flange bolta installed in 1989 contained a high density of large, long sulfide inc,1usions. This is mainly a result of inadequate mixing of the molten steel during the fabrication l process. This results in an anisotropic material condition which produces substantial decreases in transverse mechanical properties, such as ductility, fatique life, and fracture toughness. The cracked flange bolts also contained sulfide inclusions to a lesser extent.

1 (2) The problem bolt fatigue fractured as a result of vibration normally encountered by the cam shaft gear. EMD (DG vendor) believes that the level of vibration is very low because the

" pork chop" bearing is very rigid and it reduces the gear vibration te a very low magnitude at the location of the problem flange. EMD also believes that the AISI 1144 dowel bolts are adequate for their application.

(3) once the problem bolt fractured, the flange contact stress is no longer symmetric, resulting in a larger cyclic stress.

(4) The larger vibration stress results in fatigua crack of the remaining three bolts.

__ _ . ._ - _ _. _ _ .- .- . __ . ~. _ _ . _ . - _. ._ . _

4 9b

?

ROOT-CAUSES OF THE FAILURE The most likely root causes of the f ailure is _ inadequate material 'of the 4 brokon bolt. It=contains many large sulfur-stringers that: tend to p mduce '

substantial decreases in fatigue life and fracture toughneung CORRECTIVE ACTIONS ,

Install new bolts that_contain small and low density of sulfur i stringers. The. specific actions taken are stated belows (2) New AISI 1144 bolts were ordered. .

-(2) Two bolts were taken from the same shipment. . Meta 11ographic analysie-confirmed that the material contains small uniformly-distributed - ;l sulfur stringers anc anLabsence of large, long sulphur inclusions.

'(3) New bolts from the same shipment were. install'ed..

(4) -After installing,thefnew bolts, DG 3 GOO 3=was run for one hour.

Torque checks confirmed :there was--- no. loss of preload. ~

f OPERABILITY ASSESSMENT Tho above stated corrective actions will eliminate _ recurrence of

-tho identified-failure.

+

-Morcover, the past operation experience has'shown that:in-many past p -bracket fracture events, these AISIL1144 bolts; sustained:the.high __ -

. vibration (much higher thanithat with an1 intact?new design bracket) .

I -without fatigue'or overload-' fracture.-This factisuggests-that AISI-  !

1144 material (if-fabricated properly)Lis adequata as dowel bolt -

.matorial. .

9 6

?

-*w,.m * -- ,-. --w,, . m.,, ry ,e v er- - ww osey-v v w y v- wi +

  • W' r 5 eov +t - twer r ~ - -- m -r b + r r w-+-