engine failures with odd root causes 1

[ivymike]

okay, I was talking to an engineer at work today, who has about 11 years more experience than I have, and most of his experience is with larger engines than I'm used to (his smallest engine experience has been ~250mm bore diameter).

We were discussing instrumentation of a test engine to correlate some FEA and other models. In a roundabout fashion we got to a mild argument about the things which might drive high predicted stress around the injector hold-down region of the head. He suggested that crankshaft damper selection might be to blame. I tried not to say anything to offend him, but what he was describing was so foreign to me that he might as well have been telling me that his backpack has jets and he is an intergalactic bounty hunter as a side job. Boba Fett proceeded to tell me about some head gasket failures he'd solved years earlier, wherein the root cause was.. get this... ratcheting rotation of the head caused by pressure waves in the exhaust manifold.

Either idea by itself would normally have made smoke come out of my ears... the two of them together just made me want to discuss something - anything - else.

So, anyone with "big engine" experience ever come across such odd root causes?

 

[MikeHalloran]

Coupla jobs ago, I went to investigate reports that our exhaust pipes were vibrating hard enough to shake the whole boat. ... Not a boat in the sense you're thinking; a big yacht, well over 100 feet long, with a pair of big V16 engines. So I had the cap'n take her out of the harbor and demonstrate. Sure enough, the pipes were shaking, as were the engines, and the rest of the boat, at a couple of Hz, strongly enough to be felt everywhere. Our pipes were well made of good stuff, but they're not strong enough to rock a big engine, which was what they appeared to be doing.

Turns out the boat had originally been equipped with some monster V24 engines that were removed because they kept breaking cranks or something. It was also equipped with wet mufflers sized for the monster engines. The mufflers were not changed because they were more than big enough.

... which was exactly the problem.

[ A wet muffler comprises a vertical tube extending nearly to the bottom of a chamber, into which the engine exhaust gas, and the seawater that's been used to cool the engine, are sprayed. Once the engine starts, it blows any pooled water out of the chamber. The gas passing radially in around the bottom end of the lift tube picks up the water and carries it out with the gas as a mist, cooling the gas and suppressing the engine noise. ]

In this case, the gas flow from the big engines was not enough to blow all the water out of the lift tube that had been sized for the monster engines, so there was a big slug of water, rising up in the lift tube, not quite breaking into a mist, and falling back down into the chamber, a couple of times a second.

Out of the roughly 450 boats that I've worked on, just that one had an exhaust system that was too big.


Mike Halloran
Pembroke Pines, FL, USA

 

[ivymike]

ha. nice one. that's like the time Thomas the Tank Engine got fish in his boiler... except real.

 

[tbuelna]

ivymike,

There are lots of weird loading conditions (thermal and mechanical) in recip engines, especially in diesels.

The strangest one I know of is the very damaging cavitation errosion effect at the coolant side surface of diesel engine cylinder liners. Apparently, the surface cavitation is caused by the acoustic response of the liner structure due to the combustion pressure pulses. Somewhat similar to the increased stresses in your cylinder head structure due to its vibratory response when coupled with other vibrations in surrounding structures.

How would you like to be the engineer assigned with the task of solving a vexing problem like the one you described? Especially if your job depended upon solving that problem? Yikes!

 

[IceStationZebra]

"The strangest one I know of is the very damaging cavitation errosion effect at the coolant side surface of diesel engine cylinder liners. Apparently, the surface cavitation is caused by the acoustic response of the liner structure due to the combustion pressure pulses."

That isn't as uncommon as you may think. One of my professors in college passed one around the room - looked like deep rust pits. I believe it is most prevalent when using wet liners. As the cylinder fires and deforms the liner it sends out pressure waves, which can cause gases to momentarily come out of solution forming bubbles on the liner, which then explosively implode as the pressure rises again. One solution I have heard of is a coolant additive, that I think lowers the surface tension which helps to prohibit the bubble formation.

Some of the old Interational 466 engines had a coolant filter. A guy I worked with at CaseIH told me that it was really installed as a method of slowly releasing the additive into the coolant.

ISZ

 

[ivymike]

How would you like to be the engineer assigned with the task of solving a vexing problem like the one you described?

uh, that's been my job description for nearly my entire career. I gather that you wouldn't like it?

I've never thought of liner cavitation as a strange phenomenon, to be honest. I guess it just depends on what sorts of things you're used to seeing/working with/testing for...

 

[PJGD]

Not a particularly difficult problem to diagnose, but amusing anyway: Some 40 years ago, working at Dennis Brothers, we built mid-size fire engines with the Jaguar 4.2L XK engine in them. One fire engine still under warranty was trucked back into our Service Department because the engine had failed; the noises suggested that at a minimum the rod bearings were shot.

It turned out that the sump was virtually out of oil, so that readily explained the failure mode, and yet the fire station records showed that the oil level had been checked regularly and it had never needed a top-up! Investigation showed that the brazed-on stop washer on the oil dip-stick was loose and so every time the dip-stick was reinserted in the tube and pushed home, it moved the washer a little further up the tube allowing the stick to go deeper into the sump. Thus the rate of oil depletion roughly tracked the increased insertion depth of the dip-stick making it look as if the oil level was fine until eventually there was no oil left!

Interestingly also, we got through a lot of those engines on first start-up in the new chassis; on initial crank, valves would hit pistons and trash the cylinder head. Obviously, the failed engines would have to be returned to Jaguar. We understood that the root cause [so we were told] was that the engines were end-of-line tested at Browns Lane on natural gas, after which the SU carburetors were fitted and the engine stored prior to shipping. Supposedly, running on NG dried out the oil on the valve stems so that they would stick open. That does not sound entirely plausible to me, but that is what we were told.

PJGD

 

[Lou Scannon]

Crazy stories, PJ

 

[evelrod]

PJGD, your story sounds strangely familiar. Not big engine, but a little two cylinder Onan 4.0 generator. Again, the exact same thing, well almost. As I was told, the engines are run up on natural gas and the reason one of my exhaust valves was stuck open. This was an 80's era engine and it had less than an hour on it. They did not replace it but instead just reamed the valve guides and put it back together. I probably should have demanded a new unit, but all turned out well enough. I still have the thing and it operates well enough after all these years.

At the time I thought the factory rep was nuts. I guess it's too late to apologize...;o)

Rod

 

[dicer]

I would say the high predicted stress around the injector clamping area would be, material thickness, and heat stress. How can he suggest the damper? Is the loading cyclic with the combustion position?

And the head gasket failure if what he says was true, is not the real cause. Its a very basic design flaw. Any guesses what I'm thinking?

 

[ivymike]

on the gasket, my thought was that the thermal expansion of the manifold, coupled with firing pressure taking some of the clamp load off the joint, was a more likely cause than the exhaust pulses, but this was a long time ago and I wasn't there. The head apparently did not use anything but friction to prevent rotation, after redesign with a few pins it ceased failing gaskets. I would assume it had unit heads (one per cylinder). What were you thinking?

his comment on the injector area was much like the remark above "...the increased stresses in your cylinder head structure due to its vibratory response when coupled with other vibrations in surrounding structures."

I know that's not the cause of the (modeled) high stresses, because the FEA didn't include anything remotely of the sort. It wasn't even a "dynamic" model - it was a fatigue analysis done with a few specific load conditions applied. It wasn't worth bringing that level of detail into the discussion, here or there (in both cases I was just a bit stunned someone would suggest that the injector hold-down area might have an important vibratory response - seems that the important modes for such a chunk of metal would have MUCH higher freqencies than the relevant excitations). I had started the discussion at work along the lines of "here are some of the things that may need to be checked in the model to make sure it better matches reality" and got a response along the lines of "these are the things we'll look for in the engine test to see where the stresses are coming from." (we've not broken a single one in this area in 20+ years of production)

 

[rmw]

Ivymike,

I deal with young engineers every day who weren't even born when some of the experiences that I relate to them that I have had that drive my current designs.

But, that said, I am kind of like Ronald Reagan when it comes to this type of person you are struggling with - trust but verify.

rmw

 

[dicer]

Right on IvyMike, Pins.

 

[Metalguy]

>"I believe it is most prevalent when using wet liners. As the cylinder fires and deforms the liner it sends out pressure waves, which can cause gases to momentarily come out of solution forming bubbles on the liner, which then explosively implode as the pressure rises again."<

Not gasses; the bubbles are empty--vacuums. Just like cavitation on the back side of props and impellers.

I had to solve a broken turbo cast alum. impeller breakage problem on a V20 7,000+ HP diesel. Hmmm, the engine also had suffered a broken exhaust rocker arm too. Talk about low cycle fatigue cracking on nearly every vane---pulsation city right thru the intake tract.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein