Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SDETERS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Cummins 5.9 rod separation - root cause? 1

Status
Not open for further replies.

Lou Scannon

Automotive
Feb 11, 2003
2,979
See this link for photos of a Cummins B5.9 connecting rod that failed by separation of the cap from the rod. It is pretty self evident that the failure mode was one of the capscrews backed out over a period of engine operation. The question is, what was the root cause?
This is from a fairly seasoned engine, approximately 225,000 mi, so it is not a case of infant mortality. There is no evidence or reason to think that the bottom end of the engine was touched since its manufacture, or that the engine was subjected to any untoward abuse such as extreme over-revving. None of the other connecting rods show any sign of distress. The other connecting rods were disassembled with an air wrench, so break-away torque data are not available, but the mechanic reported that the break-away torque seemed normal for the remaining rods.
Has anyone seen this before? Any theories for the root cause?


I forgot what I was going to say
 
Replies continue below

Recommended for you

Exceeded fatigue life.

Over torqued at factory.

Faulty bolt.

Any of the above may have been only to a minor extent so it took years to crack then fail.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &
for site rules
 
Actually, nothing cracked. It appears that one capscrew gradually backed itself out, such that the rod cap was cantilevered off the other capscrew. The bearing and the cap mating surface were getting hammered at this point, as well as the piston crown & valves, which were colliding due to the loss of deck clearance. In the final failure the remaining capscrew appears to have been fretted or sheared out of the rod, after it had been bent to the angle shown.

I forgot what I was going to say
 
I concur with Hemi. You can tell the bolt backed out over time due to the flattened threads around the whole circumference. The bigger question is why it took so long before it came out? I don't see any cracks on the bolt, but you would probably have to look very close with a light to see one. It would be interesting to see if the loose bolt has yielded like Pat suggests.

ISZ
 
A bur under the head of the bolt that slowly flattened out and eventually left the bolt loose.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &
for site rules
 
ISZ and Pat, good input, I'll take a closer look and report back.

I forgot what I was going to say
 
the way the first 2/3 of the threads are flattened is interesting. Maybe the cap "gatoring" on the remaining bolt for a few dozen cycles buffed them down.

The surviving rod bolt sure gave it the good fight.

Dan T
 
I took close-up photos of the bolt flange underside and the bolt seat on the cap. They are posted to the same album as before; here is the link again to save you scrolling back up:

In the center of the respective photos of the cap and the capscrew flange, there is a tiny gouge, in the 3 O'clock position (deliberately oriented for the photo; it is not conclusive that these marks are related to each other.

One interpretation of these marks is that a hard foreign particle was trapped between the capscrew flange and the cap when the capscrew was installed and torqued, which as Pat suggested, may have falsified the torque and hence the preload.
A difficulty with this theory is that one would expect more of a ring-shaped scratch on the surfaces, created during the final torquing of the capscrew, rather than just a single pit.

Any other theories?

Tmoose, I fully agree with your comments!

I forgot what I was going to say
 
I have a handful of pictures from CAT engines failures that are similar, there are a couple of areas I would look more closely at.

I'd like to see some better pictures of the bolt head face, there appears to be some signs of fretting and would be nice to have a better look.

The bolt seat on the cap shows an uneven contact pattern, subtle but there, there are a couple of smaller scratches that run radial, the elongated pit you see may have been caused by a larger piece of debris under the bolt head. The machine marks in this area indicate a rougher surface finish, could be the piece of debris caught a slight edge and the bolt head forced it into the rod.

The parting face of the con rod also appears to have a fairly rough surface finish, the machine marks running parallel to the rod bore seem fairly pronounced. Take a look at that area with some magnification, about 10X, see if you can determine signs of longer term fretting. I know its hard now since the cap got knocked around a bit.

I have seen similar failures, all higher time, most of which we found signs of either a poor surface finish that yeilded in long term use and reduced joint tension. We did have a group of bolts that had poor surface finish on the bolt head face area, causing a similar problem, all near the 8-10k hour mark. We also had a batch of REMAN rods that had similar failures we traced back to a surface finish problem on the parting face, was hard to tell visually, but depending on engine load factor can make a difference.

Hope that helps.
 
had 2 5.9l failures similar up here, tech said that the head of the bolt failed under load then rod cap opened, didn't see the parts myself, just what i was told
 
CSE, great input. I will take a closer look at the areas you mentioned and report back.

I forgot what I was going to say
 
Pursuant to CatServEng's input, I added 8 photos to the album:

They are tagged to indicate the specific location being depicted.
The machining quality of the rod/cap parting surface does not look that great. The machining marks indicate tearing rather than a clean even toolmark. The last 4 photos are of one of the non-failed rods. The same issue is apparent, but to my examination not as bad as the failed rod. It's hard to be sure that the difference is not merely due to the different history of the failed rod once the separation began to occur.
As a check, I looked at one of the main caps and the mating surface is about the same as the rods'.
So, is this machining quality typical, or below par? Is it a tool issue, material issue, or procedure (e.g. cutting speed/depth) issue?

I forgot what I was going to say
 
Based on your additional pictures, the surface finish of the parting face of the bearing bore on the failed rod looks the worst on the side of the rod the bolt appears to have backed out on. The unfailed rods have much less pronounced machining marks. A poor surface finish could lead to a large number of asperities, that in service, depending on load factor, could lead to them flattening out and reducing the joint loading.

While it's hard to firmly declare a root cause based on what we can see here, so far there are some things we can say don't appear to be the root cause.

The bolt that backed out does not appear to have been over torqued on assembly, the threads don't looked rolled like you'd normally see with an over torque, and there is no necking or plastic deformation. The polishing on the threads is typical of a loose bolt, at least based on the reference material and experience I have.

Nothing in the pictures I can see indicates a fatigue type failure, the metal surfaces aren't discolored so it looks like there was plenty of lube present while in operation. Because this was limited to a single rod and the other rods don't seem to show early signs of a similar problem it doesn't look like there was an overloading, poor driving habit like excessive lugging, or some other overall engine problem.

I'm not sure if these parting faces are machined or ground, but they sure look not so great to me. A rod sure takes a lot of abuse in a modern engine, and a lot of forces are transmitted thru them, doesn't take much of a problem to eventually cause a failure. It looks like either the cutter was getting dull or the grinding wheel was loaded.

If the rod parting faces were moving in relation to each other they would likely either show signs of fretting or polishing. I can't think of any failure mode that would make machine marks appear more pronounced.

My two cents worth, hope it helps. Mike L.
 
Thanks again for your insightful analysis. Your theories and arguments seem to fit well with the evidence at hand. There is no question of warranty or liability in this instance, but I'm sure the owner will be grateful to be able to rule out some of the potential root causes you mentioned, and have a workable theory for what actually happened.

I forgot what I was going to say
 
When I worked at a shop that rebuilt rods we used a Sunnen rod and cap grinder to gently and precisely remove material from, and create a finish on the rod and cap parting faces. It referenced off a side clearance face, and the original parting face.

OEM passenger car rods varied Enormously in quality of parting face finish. For a time Some SB Chevies were a torn up mess. the accuracy of the parting face geometry and other features could also be completely haphazard. Some FORD rods (consistently) had the OEM parting face generated at an angle to the bolt holes and parting face such that when the newly machined parts were assembled , before full torque, there might be a 0.015-0.030 inch gap on one edge. IN retrospect the way the face "cleaned up" was a clue there was trouble, because the first Sunnen cut was only on one edge, but much heavier than expected because the height above the grinding wheel of the cap-to-be-machined was referenced on a center line. Pretty quick we learned to back the wheel down, for the first cut, just in case. Also, those angled rodz-n-capz were shimmed to cut and maintain their original "mostake."
 
The discoloration of the big end! Did it happen before or after the bolt loosened?
 
An excellent question. But we'll never know the answer for sure. I'm open for conjecture. I guess a few thousand hours of lube oil infiltration due to loss of parting face preload could explain the darkening.

I forgot what I was going to say
 
Maybe I should have explained my opinion on the rod discoloration a bit better, and the assumptions made. Didn't see any pictures of the bearings, but looking at what you provided, along with your story, the iron doesn't look like a loss of lube and subsequent bearing failure. Usually if a bearing beats out of the bore after running with too little lube the load area of the rod also gets hammered pretty good.

While it looks like the bearing spun in the bore, the rod bearing bore surfaces from the pictures don't look like it did spin for long and as the rod opened up things got bad rapidly. Also looking at the crank pin can be helpful, as it can give some clues as to what happened, but sometimes the follow-on damage is too severe.

Getting the parts in hand and doing a really good lookover is the best way. When I did this stuff all the time many times it was a tiny detail that lead you to the exact root cause. And a lot of times the actual root cause couldn't be determined, but we did have a pretty good list of what we didn't think caused the failure.

Back when we were serious about applied failure analysis, a single rod bearing failure would typically generate about 40-50 pictures and a 12-20 page report. The failed rod and bearing (both shells front and back) both adjacent rods and their bearings, failed crank pin and both adjacent crank pins, and adjacent mains if determined they may have evidence to contribute, top of piston, piston pin and rod small end bearing, and if the bolts were suspected, a good number of pictures of them. But do that much anymore.

Hope that helps,

Mike L
 
Mike, thanks for your latest response, but I don't think I understand completely where you're coming from. I was not implying loss of lube oil pressure, only oil getting into the parting face area due to loss of preload, thus causing or contributing to the darkening of the parting face area. Agreed, once preload was lost, bearing was toast.
8D investigations, a good methodology. It always seemed to me that the last D, "Recognition for the Team", was only window dressing and always just got lip service, so to speak.


I forgot what I was going to say
 
I was trying to answer the question Dicer posed. No worries. I used to really like doing failure analysis. At one time CAT had probably one of the best programs for applied failure analysis, great training and materials, and a good community to share info in. Cost savings and the desire to not really know what was going on sometimes killed it. It's still around but pretty much in name only.

It was an interesting exercise, thanks.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor