It's a great day- We're pulling our engine for teardown

[AN23]

Yesterday, I pulled our RV-8 (IO-360 Angle Valve) into the hangar for an oil change. Engine running great, great oil analysis, very low oil consumption and very high 70's compressions. Drained, refilled and cut the filter which was about as clean as I've ever seen. I went home happy. Today, I went back to the hangar to wash down the engine and do a leak check test run. After the wash down, I was using the compressed air to dry down the engine when I noticed one of the large (thru bolt) #3 cylinder studs was broken. On closer examination, the smaller stud next to it was broken also. After searching for just a few minutes with a magnet, I found the large nut and thru bolt piece. The engine has 403.3 hours TSMO and was overhauled at a reputable (now closed) facility. Despite some local opinions that it's possible to just replace the stud and through bolt and go on, I called two very well known experts and expressed my concern that the thru bolt provides clamping force on the bearings and it has obviously lost that clamping force. Both experts agreed that it should be torn down and inspected/repaired. It's coming off and heading to an engine shop next week. Contrary to what we like to joke about, my wife fully supports the expense.

So, why do I say it's a great day? First, I am in a partnership with my son (A-10, now Delta) and he and I agree that safety is the absolute #1 priority and we can split the cost. Next, we both have the resources to do this without missing meals. Lastly, and most importantly, we caught this before something happened that would forever change a real joy in our lives (RV-8) into a potentially catastrophic event.

Please, please pay special attention when you're doing your inspections. From this point on, I will look at every visible cylinder bolt, every time I have the cowl off.

 

[Mike S]

Good catch, not only for the safety aspect of this, but the IRAN is going to be a lot less $$$ than if it had grenaded.

 

[AN23]

Thanks Mike. I hadn't thought of it from that angle but you are absolutely correct!

 

[blueflyer]

Any idea as to why the bolts broke off?

 

[AN23]

No way, short of getting a metallurgist involved, to know for sure but improper torque at buildup is a common cause.

 

[Freemasm]

Can't be 100% sure without a first hand analysis of the HW; however, the flat fracture surface is a classic example of a fatigue failure. This suggests the fasteners experienced reversing loads from under tension. Initial under-torqueing or material between the mating/contact surfaces. Upon removal,

For my own curiosity, I'd check the contact surfaces for evidence of paint/primer. Also, I'd perform torque checks/recording the results on the remaining fasteners; that jug and others.

Count your blessings. This was pending (see added pic)


Edit = If you want, send the nut/broken fastener to me. I have access to a world class lab. It would be a favor by them and no guarantee but they've always been helpful before.

 

[jrs14855]

Engine

Several decades ago there were some issues with quality control on cylinder studs.
I have no proof but I also believe that some people doing field overhauls do not even know about the "torque plates" that are a required part of crankcase assembly and cylinder installation.
Of course quality torque wrenches are a must. No Harbor Freight!!!

 

[danny]

"Great" day..

Really good catch and the decision to tear down as a no brainer is a good one IMO. I had this happen to a C86 and ignored it to the point that the case would "wow" out with every revolution. I didn't tear down until the oil pressure was too low from worn crank bearings. Ended up destroying the cam and damaging a bunch of repairable things.
"A pat on the back" for doing the right thing.
danny

 

[jliltd]

No question this is the right decision. You likely have fretting where the case halves come together. The crankcase shop can take care of any such damage.

 

[AN23]

Can't be 100% sure without a first hand analysis of the HW; however, the flat fracture surface is a classic example of a fatigue failure. This suggests the fasteners experienced reversing loads from under tension. Initial under-torqueing or material between the mating/contact surfaces. Upon removal,

For my own curiosity, I'd check the contact surfaces for evidence of paint/primer. Also, I'd perform torque checks/recording the results on the remaining fasteners; that jug and others.

Count your blessings. This was pending (see added pic)


Edit = If you want, send the nut/broken fastener to me. I have access to a world
class lab. It would be a favor by them and no guarantee but they've always been helpful before.

Thank you! I will take you up on your offer. Could you please PM me your shipping address and I'll get it right out to you.

Rob

 

[AN23]

Cylinder studs broken

Here's a couple of close up photos of the thru bolt. One end has a concave surface and the other has a rounded end. Striations are obvious.

 

[Freemasm]

Micro fatigue cracks for the first 90-95% until ultimate rupture. Classic.

PM sent

 

[Canadian_JOY]

Where is the "Double Like" button for this discussion thread?!?!?!

This thread is full of goodness. Good luck in finding an issue on the ground rather than in the air. Good luck in actually seeing the broken part. Good luck in having true experts to consult. Good decision making in plotting the safest path forward.

Yup, that's a lot of goodness indeed. Well done!

 

[AN23]

Through VAF members, I potentially have professional metallurgists looking at it. I’ll update you guys once I hear anything!

 

[David Paule]

Micro fatigue cracks for the first 90-95% until ultimate rupture. Classic....

I agree.

Dave
(old retired structures engineer)

 

[terrye]

Great catch and great pictures. Looks to me like a fatigue failure due to either initial undertorque, or relaxation/loosening in service perhaps due to flange to case contamination.
I'll suggest ALL NEW FASTENERS for the rebuild.

 

[TASEsq]

We had this exact same failure in our aerosport io-375 with 178 hours on the clock. I can’t quite recall but I think we broke a couple of studs and one through bolt.

Bit of a long story but the short version was the original engine spec was for a forward prop governor. Aerosport could not get a new engine case so used an overhauled one. As it turns out, when an engine reaches life and is pulled apart and overhauled to a new engine, the case is inspected etc and machined flat. The studs are not removed as their removal is very difficult without damaging the case. The case is punched with a mark and is only allowed to be overhauled x number of times (I think it’s 3 but don’t rely on that as fact).

Now, our case was one of these old ones. Had been overhauled 3 times already - who knows as what engine. Probably a stock io-360. Now aerosport gets this old case, since it’s experimental, makes it flat and builds it into our engine. Now we are high compression etc the studs are working a lot harder than before. I asked do they preload the studs with more torque when they increase the compression to account for the extra cylinder pressure but couldn’t get a straight answer. If they didn’t do this, or if at anytime in the 3 x lives of the engine the studs were not perfectly preloaded, the studs would be getting load cycled.

So in our case these studs have seen 3 lives already then we go and really start making them work. It’s no surprise they failed.

The case was toast after it happened - the shop could not get the cylinders back on flat. They installed a cylinder off another engine (a known flat cylinder) and a feeler gauge could be inserted at some spots under the cylinder base. As the case had been overhauled and flattened multiple times the deck height was on the limits and it couldn’t be machined again.

We took all this to aerosport who were great. They said the “old management” had some questionable practices and they would never do that now. They sent us a new engine case at their cost and we rebuild (bulkstripped) the engine - bearings etc we’re replaced as required.

So maybe check your build logs on your engine to see if an overhauled crank case was used!

 

[AN23]

Thank you! I will take you up on your offer. Could you please PM me your shipping address and I'll get it right out to you.

Rob

Thanks Trent!

In our case, we know the history of this engine. It came off of a Mooney with 2183 hours before teardown and rebuild so it has had one overhaul prior to this event. I'm sending it to Triad and I trust those guys to carefully look at all of this.

 

[lr172]

The fact that both of the broken fasteners were on the same jug, it would seem the issue is more likely related to that cyl vs the fasteners themselves. Could be an issue with contact area, flatness or torque or various other things. I would also be looking for issues that could have been putting more stress on that cylinder. It is likely not coincidence that two studs broke on the same side of the same cyl.

 

[Freemasm]

I'll state again that a torque check of all cylinder fasteners would be advisable before shipping the PP off.

 

[AN23]

Thank you! I will take you up on your offer. Could you please PM me your shipping address and I'll get it right out to you.

Rob

Reference the above about checking the torque on the other fasteners....

Just curious. Why if the engine is going to get a complete tear down?

 

[Jpm757]

Curious, what compression ratio?

 

[AN23]

Standard 8.7/1

 

[Brantel]

Of course quality torque wrenches are a must. No Harbor Freight!!!

I would say that a properly calibrated torque wrench is a must. Many reviews out there that prove calibration is more important than brand.

 

[Freemasm]

Reference the above about checking the torque on the other fasteners....

Just curious. Why if the engine is going to get a complete tear down?

The failure is a result/symptom, not a cause. It could potentially help with a root cause determination; i.e. a material or installation quality issue. That could help determine if others here that had OHs around the same time/utilized the OH facility/component OH shop/whatever have reason to suspect their PPs.

This type of failure would take a significant amount of luck to catch before rupture even if accessible. Your courteous review of the torque findings, OH location/dates, related yellow tag info, etc, shared here could prompt someone with similar component history to perform some needed "due diligence". ALso , anal engineers like myself would appreciate your findings even if there's little chance the PP builds share any similar history.

Just a thought. There's so much to learn on this site as long as info keeps flowing.

 

[AN23]

Got it! Once we have the engine to a point that I can get to all of the cylinder bolts, I will check the torque and report.

 

[Freemasm]

I'd set your torque wrench IAW the attached and torque in the "tightening" direction and monitor for any nut movement.

Likewise if you have an electronic torque wrench (most will record the highest torque it applies), you can check for breakaway torque in the loosening direction. All of these should exceed the original torque value. Others will hopefully chime in with other ideas. Thanks and keep us informed.

 

[ronrapp]

So maybe check your build logs on your engine to see if an overhauled crank case was used!

Speaking of build logs, I emailed the company that rebuilt the engine on my airplane and asked if I could get a copy of the work order. The reply back was "no, sorry". No further information. I'm not sure if they simply don't want to provide that info or if they discard if after a certain amount of time or what.

I'd love to have that info if it was available.

--Ron

 

[AN23]

Broken cylinder studs

Thanks to a generous VAF member, we are hopefully going to get a metallurgist to establish the type of failure. If, as expected, it shows a fatique failure then, the effected cylinder hold down bolts were, for some reason, experiencing a cyclic load greater than the load (preload) seen by the bolts from installation torque. According to the excellent Sky Ranch Engine Manual, this could be caused by insufficient torque, excessive torque, incorrect lube of fastener or contamination of the mating surfaces (paint, washers, etc.). Any of these can cause elasticity at the joint. Here's where it gets interesting. I went back to the hangar and carefully looked at each of the visible cylinder hold-down fasteners. Something didn't look right. There is what appears to be washers under each nut! I searched the overhaul manual, parts list, etc. and cannot find ANY mention of washers for this assembly! Someone please correct me if I am wrong. It's a IO-360 angle valve with standard compression pistons.

The engine is going to get removed next week and taken to a respected overhaul facility for teardown. Before it leaves, I am going to confirm that it does, indeed, have washers under each nut. The engine is seventeen years since overhaul but relatively low time. It has not been touched since overhaul. I will NOT try to go back to the overhaul facility (I was wrong, they are still in business) but once I confirm that there are incorrectly installed washers present, I will come back here and identify both the timeframe and the facility that installed them so that others can check their engines.

 

[lr172]

Thanks to a generous VAF member, we are hopefully going to get a metallurgist to establish the type of failure. If, as expected, it shows a fatique failure then, the effected cylinder hold down bolts were, for some reason, experiencing a cyclic load greater than the load (preload) seen by the bolts from installation torque. According to the excellent Sky Ranch Engine Manual, this could be caused by insufficient torque, excessive torque, incorrect lube of fastener or contamination of the mating surfaces (paint, washers, etc.). Here's where it gets interesting. I went back to the hangar and carefully looked at each of the visible cylinder hold-down fasteners. Something didn't look right. There is what appears to be washers under each nut! I searched the overhaul manual, parts list, etc. and cannot find ANY mention of washers for this assembly! Someone please correct me if I am wrong.

The engine is going to get removed next week and taken to a respected overhaul facility for teardown. Before it leaves, I am going to confirm that it does, indeed, have washers under each nut. The engine is seventeen years since overhaul but relatively low time. It has not been touched since overhaul. I will NOT try to go back to the overhaul facility (I was wrong, they are still in business) but once I confirm that there are incorrectly installed washers present, I will come back here and identify both the timeframe and the facility that installed them so that others can check their engines.

Washers are not specified there. Not an ME, but have my suspicion that there is not a huge issue having them there, assume they are hard (grade 5 or better). However, I do wonder how much the washer will change the torque reading for a give bolt stretch. Given the proper procedure is to torque wet with a fairly thick oil, it likly isn't much of a difference. Given two fastener failures in the same area and no others, kind of points to a mating issue, like a chunk of metal or paint between flange and case.

 

[Freemasm]

Thanks to a generous VAF member, we are hopefully going to get a metallurgist to establish the type of failure. If, as expected, it shows a fatique failure then, the effected cylinder hold down bolts were, for some reason, experiencing a cyclic load greater than the load (preload) seen by the bolts from installation torque. According to the excellent Sky Ranch Engine Manual, this could be caused by insufficient torque, excessive torque, incorrect lube of fastener or contamination of the mating surfaces (paint, washers, etc.). Any of these can cause elasticity at the joint. Here's where it gets interesting. I went back to the hangar and carefully looked at each of the visible cylinder hold-down fasteners. Something didn't look right. There is what appears to be washers under each nut! I searched the overhaul manual, parts list, etc. and cannot find ANY mention of washers for this assembly! Someone please correct me if I am wrong. It's a IO-360 angle valve with standard compression pistons.

The engine is going to get removed next week and taken to a respected overhaul facility for teardown. Before it leaves, I am going to confirm that it does, indeed, have washers under each nut. The engine is seventeen years since overhaul but relatively low time. It has not been touched since overhaul. I will NOT try to go back to the overhaul facility (I was wrong, they are still in business) but once I confirm that there are incorrectly installed washers present, I will come back here and identify both the timeframe and the facility that installed them so that others can check their engines.

Interesting observation but as I stated in your PM, I'm not feeling it. There are already threads in bearing but as this joint is primarily in tension; shouldn't be a factor here. BTW, Lycoming has the aforementioned on their case bolts as well. Even if a low grade/quality of washer was used, its stiffness should easily exceed that of the aluminum mating components. That said, non-aerospace grade cr@p that could lose it's properties over time could create a loss of preload. That would be more likely today than 17 years ago so IMO, even less of a possible contributor. Also, I do not know but highly suspect that the configurations that utilize thick base rings have the same torque values. If the washers you discovered were a contributor, I'm not conceiving it. Furthermore, that small change in stack height (it's a relatively thick aluminum joint) should be well within any design margin.

I'll be talking to the NDEs at the lab tomorrow but i"m truly expecting no surprises if they do look at it.

Maybe Mr. Paule from the Republic of Boulder stated he is a retired structures engineer. Maybe he will chime in.

 

[AN23]

I'm certainly no engineer. Maybe the washers had an effect, maybe they didn't. My point is that if these are indeed washers and if they were not specified in the overhaul manual, what else is wrong? My nine year old, cherished grandson flies in this with our incredible son. I'm seventy two years old. I'm probably past my "best if used by date" but they aren't. I'm just stating what I see. Others can use it for what it's worth. Thanks so much for your help!

 

[Freemasm]

Your safe approach, concern for your family, and attempts to understand a tough situation are commendable. You certainly have my respect. Any comments by me are with the intent of trying to be helpful.

Larry added a post while I was typing. To me, his comments add a bit more reason/importance to checking the other fasteners torque values. That of course assumes you want to know any potential root cause. Don’t let annoying people like myself determine your actions. Well past my “best if used by” date as well.

 

[AN23]

I will absolutely check the torque on the others and I'll let everyone know what I find. Thanks again.

 

[AN23]

Just heard back from a very respected lead person at a major overhaul facility.

"No washers allowed"

 

[lr172]

Just heard back from a very respected lead person at a major overhaul facility.

"No washers allowed"

That may be true, but according to the Big L, lean of peak operations aren't allowed either and that doesn't seem to cause case bolt cracking. I would stick to what the ME's say about that. For all you know that statements means they are not required vs prohibited. I am NOT an ME, but would be absolutely shocked to find that washers caused the broken studs. most of the fasteners on your plane have washers. The only real harm that a quality washer can do is change the registered torque for a specified amount of stud stretch that the manufacturer is looking for to set a pre-load. I do agree that both over and under torquing can cause this, as can any type of debris, burr, etc. between the flange and the case. Especially with the alum case that will slowly yield over time and lessen the pre-load. Let's remember that the narrow deck engines have a plate between the flange and the nuts. Other than not spinning, it is functionally equivalent to a thick washer.

 

[hgerhardt]

.................. visible cylinder hold-down fasteners. Something didn't look right. There is what appears to be washers under each nut! I searched the overhaul manual, parts list, etc. and cannot find ANY mention of washers for this assembly!

Those are NOT washers! That is a bearing surface, or boss, molded into each nut, which means the nuts only go on one way. In fact, your very own post #11, in the center photo, clearly shows that boss on the nut.

 

[AN23]

Thank you! As you can see in my post #29, I wasn't fully convinced that they are washers. Your post and my previous photo make sense. I'm going out tomorrow to check the torques on the other accesible bolts and I'llremove one just to absolutely verify.

Thanks again!

 

[osxuser]

If improper torque is not a factor, another common cause for this would be detonation. If you ended up with partially plugged injector and ran the engine hard, you can have this same result.

 

[AN23]

I boroscoped it again today and checked the spark plugs. Absolutely normal. There were not washers installed. As pointed out in another post, the nuts have a built in boss that looks like a washer. I started the removal process today and It will go to Triad next week.

 

[Freemasm]

NDE findings of fracture surface are in the General Discussion Thread

https://vansairforce.net/community/showthread.php?t=216888

Hope being, it gets a little more view traffic than being buried deep here.

Edit = Nevermind. They moved it here. I personally think that is a mistake. The lesson learned just so happens to be on a powerplant.

 

[BillL]

Just an FYI.

Continental reported a failure like this where RTV was used on the barrel flange to case joint. The tiny amount of compressibility (strain/stress range for the stud) was enough to put the cyclic stress range in the fatigue failure zone. The base must be metal to metal - you can check if this is a factor.