All "IMS failures" are not IMS failures! Carnage within!
 

Victim: 2001 M96 engine pulled from a Boxster S. Mileage reported to be 45K

Catastrophic Failure

MOF #14

We have discovered another MOF (Mode Of Failure) related to the M96 engine. This one is our most recent tear down that occurred with a 45K mile engine. It is one of the most radical failures that could occur with an engine and we have gained a good understanding of why it occurred. More importantly we are in work already on another "overkill" part to solve this issue that we can apply during our engine updates for our complete engines and component kits.

This is interesting, because this engine fooled us, it also fooled a Porsche Dealership and another independent Porsche shop that had diagnosed the engine prior to it being sent into us. The engine had all the classic symptoms of a seriously failed IMS that had resulted in valve timing alterations, thus colliding the majority of the valves with the pistons, a true nasty affair. When the intake manifold was pulled we immediately saw carnage (broken valves and chunks of piston in the intake ports!) and since I love to dissect engine failures I stayed late to see what was inside this beast.

As the engine was torn down I was expecting to see the IMS shredded, but every time we turned the engine while removing the valve train the only symptom that the engine had was altered valve timing, this was odd. Typically when an IMS fails this badly the shaft shears in half and "parks" the valve train that operates from the opposite end of the IMS drive, located at the flywheel end of the engine (I am trying to explain this so you guys can understand!) and that results in all the things we had witnessed. When this happens the IMS will not turn, so the oil pump stops functioning and 3 cylinders worth of valves stop actuating, resulting in a big mess.

This is where it gets interesting:
When the IMS fails to the point of allowing piston collisions with the valves the IMS is most always sheared.. But when this engine was rotated fore and aft the IMS was responsive to the change in crank position and the typical damage to the oil pump that results from IMS failure was not present.

How could an engine have such a radical failure, with all the same initial symptoms as the IMS failure and have an IMS that was responsive to crank position changes, intact and still functional??? Read on...

When the first cylinder head was removed the engine still had all the symptoms of an IMS failure, but when the second head was removed that cylinder head was found to be perfectly intact with zero piston/valve interference and no damage. The engine was still spotlessly clean inside and showed no signs of abnormal wear on the "intact" bank of cylinders.

As we went deeper it got interesting as it seemed the valve timing was radically off on both sides of the engine, but all the timing chains were intact, nothing was broken and all was in decent condition. At this point of the autopsy we knew that when the case was split we'd find something we hadn't seen before.

When the case was split the IMS came out in good condition without any oil inside it and without any signs of bearing failure or wear. We immediately removed the crank carrier and then we saw the culprit..... The crank/ IMS drive chain tensioner pad was non existant and the tensioner was broken in half! We then dug through the rubble and found the broken pieces and started putting the pieces together to figure out what happened.

Why the valve timing changed:
Because the main drive chain for the IMS lost all tension when the "tensioner" sheared in half, thus allowing the IMS to stop rotating and "parking" the valves. When this occurred a couple of cylinders had open and partially open valves that then collided with their pistons and that resulted in scattered parts, a loud bang and chunks of piston being emitted from the tail pipe...

What we learned:
Something we had never really paid close attention to was the shape and thickness of this tensioner through the various years and models of the M96 engine. We pulled this piece from several engines and started comparing them in shape, interchange, composition and design. We immediately noted that the the early 2.5 tensioner was thicker across the area where this tensioner had broken and that the early unit also had a hardened steel contact surface for the chain adjuster to ride against. The later 3.2 unit had been made thinner through the area where the breakage occurred and had a PLASTIC surface for the chain adjuster to contact.

The part number from the two parts were the same, but one was a .4 while the other had a .5 suffix.

So, why was this part changed? Why would anyone ever replace a hardened steel wear surface with something comprised of plastic, when even the steel wear surface does typically wear....

This is a MOF that we have never seen before and never heard of occurring. After seeing this failure and it's symptoms I believe that some of the "IMS failures" that are diagnosed without an engine tear down are actually this MOF, or at least something similar to it, that cannot be thoroughly understood without complete dissection of the engine.

That said, we are already in work creating a two new components that will solve this MOF issue. First we are making a billet aluminum tensioner unit made to use the OE tensioner pad, this will replace the lightweight, cast aluminum factory unit. This will incorporate a larger, tool steel wear surface much like the earlier 2.5 unit pictured below. Making the wear surface larger will increase the contact patch that the chain adjuster sees, thus increasing service life.

Secondly we are making the wear surface "button" compliant with the OE tensioner, so those that have the early tensioner can replace their wear surface with a larger, stronger part if they don't choose to utilize the entire billet tensioner that we are creating.

I'll never trust one of these OE tensioners again, we'll be applying the billet units to ALL our engines effective immediately and all builds are currently halted while this component is being made.

The key to avoiding problems like these are to make the parts heavier duty than they "need" to be on paper... I call it "overkill Engineering" and it's what keeps things from failing... Its nothing more than classic hotrodding being applied to these newer, robotically assembled, mass produced engines.

We are now going through cores and doing research trying to specify when Porsche changed this tensioner.... I'll report back when we have concrete evidence.

Now for the pictures... More will be posted today on the "reliability" page at www.flat6innovations.com

The pictures speak for themselves, its mass carnage of an unreal kind.. The kind of stuff that makes it to our "trophy Shelf" filled with offerings to the Gods of Speed...

I had more pictures to post, but the software limited me to five..

Here are the pics of the two different tensioners. The broken unit came from the failed 3.2, the intact unit came from a 1998 2.5 Boxser engine.

Note the wear surface and the thickness of the tensioners.

WOW, Jake, great write up.

Any evidence in you opinion that this damage was abuse/neglect related or more of an assemtly flaw?

45K does not sound like a lot of miles for a quality engine.

The rest of the engine showed no signs of abuse or neglect... The engine failed at normal driving speeds, around 70 MPH.

It kind of hard to eye ball the thickness difference of the tensioners but I guess there is. I one thing I did see was the pot metal at the break. Those tensioners must be put under a tremendous amount of stress (I think of watching a motorcycle chain under power and then letting off) and to use crap metal is crazy.

So Jake, you think this could be the cause of some of the failures? Because of the carnage in a block after one lets loose that it simply is being overlooked and the blame is put on the IMS? If thats the case the fix seems so much easier to remedy.

Yowzers! Great write-up Jake, even I could understand that. And the pictures, wow, what carnage. Pictures are not just worth thousands of words but dollars too!

Is this a part that an average joe with average joe tools can replace on his own?

No way, this involves engine removal and breakdown.

Thanks!
I tried to leave all the techno-babble out as much as possible and post at a comprehension level that those who haven't been inside this engine could understand.

Its a minute difference, but it does exist. Its so small that I had overlooked it until this instance.

The whole part is "pot metal" and crappy.

Unreal tension, especially when the vario cam comes in and everything in the top end of the engine tightens up.

Absolutely.It already has failed in this engine and I am sure it's not the first, but its the first time we have seen it.

The symptoms are exactly the same as an IMS failure, thats why it fooled everyone that had diagnosed it, even me.

Without TOTAL disassembly this would be misdiagnosed as an IMS failure. Since a majority of the engines are sent to Porsche for reman without ever being opened up, everyone **THINKS** that they had an IMS failure, when it really wasn't.


NO!!! It's not! To uncover this tensioner the ENTIE engine must be torn down, if you notice the tensioner attaches to the crank carrier, the first assembly installed into the engine when building it and the last assembly removed at tear down. There is no way to apply our updated part without removing every bolt from the engine, unfortunately.

We are doing a metallurgical survey on the broken part to see what it's comprised of, so we can look for the traces of this material when we pull oil samples from customer's engines orfor those who are part of our oil monitoring program. This will help diagnose issues before they occur.

With one of our engine assembly kits of updated components and our tools it could be done by following the DVD we are working on now... Otherwise this is a near impossible level of repair for a DIY project.

We klearn something new everyday...

If I understand what I'm seeing correctly, it looks like a overloading failure, maybe due to a manufacturing defect or fatigue. Not a wear issue. I wonder if you'd ever get a warning from an oil analysis.

It does appear that the unit was overloaded, but all the up stream components that actuate it were in fine condition..

It appears that the tensioner wear pad may have worn and then been ripped off and that created the tensioner to operate at an odd angle that could have made it break..

An oil sample could seek out the wear materials from the tensioner pad in the oil and then if the chain is going metal on metal before a failure the elements of the tensioner would be in the oil as well..

Its a long shot, but we are already observing so many oil samples its just one more thing to add during analysis.

The newer tensioner being thinner with the plastic contact patch doesn't help anything.

I wonder if that's a design flaw that would be easier to work up the old class action law suit on? Imagine a recall to replace those... :eek:

Quote: "If thats the case the fix seems so much easier to remedy.

NO!!! It's not! To uncover this tensioner the ENTIRE engine must be torn down, if you notice the tensioner attaches to the crank carrier, the first assembly installed into the engine when building it and the last assembly removed at tear down. There is no way to apply our updated part without removing every bolt from the engine, unfortunately."

Well, Jake, you don't have to answer my now apparently stoopid question over at RennTech! :o

Hey Jake, I know I am speaking for others on the fourm when I say I am sure glad you are in our corner with these engines. Keep up the good work!!!!

Ed

:cheers:

Thanks for the kind words.. I like to make things work and I prefer to work with things that are problematic, like this engine... We are full steam ahead making components to eliminate these issues.

As for a class action law suit, I am a hands on kind of guy and nothing thats settled in court is going to solve the issues this engine has.. I stay completely away from anyone who even mentions the word lawsuit...

If I can't fix it in my shop, I want nothing to do with it... The last time I checked a Gavel wasn't a tool developed for engine repairs :-)

Jake, you are the Stephen Hawking (who could explain concept of black hole to common man) of M96. Thank you for your time effort in diagnose these engine. I have 2001 Boxster S and it is right at 43k miles mark. I am holding my breath now. :-)

Don't be afraid.. Just drive it. If it breaks, then deal with it.
A lot of these engines fail, but a lot more don't...

Thanks for the info. Too bad its not going be to an easy DYI preventative project for most guys, but its good to know that when and if it needs to be rebuilt, then it will be 'better than new'.

Until that time comes, as Jake said, just drive and enjoy the damn thing.

Not to worry Jake, your advice from Excellent magazine article helps. I do drive my Box year-round even at -35C here in Canadian winter. I guess if it breaks, I will be talking to you for a new engine. :-) Cheers and thanks again.

Now that's the right attitude!
Jake maybe I missed it but was this off a trip or a manal trans?

Jake, I don't think they use gavels all that much in suits, but maybe Lovebunny can weigh in on that one.

This engine came from a Boxster with a manual tranny.... Owned by a Female that bought it new.

I don't like the idea that the guide is a wear item. They should be rollers instead. I built the attached in CAD. More rollers are better but this conveys the idea. If you want, I can help you develop the part with CAD and finite element analysis. Send me a PM.

Jake..
 

Did one of the 3.2L tensioners survive the damage intact? If so, could you post a picture of the 2 side by side.

Is it fair to say the "newer" design is less robust than the old 2.5L design? If so, what would you guess the cost of material saved might be between the 2 parts?

Thanks

Interesting
 

Now we have another weakness to worry about! This may just be a rare failure I hope!

This design also has some issues as the roller axles are a failure point and the contact patch of the rollers would result in point loading. The drive chain is also a very hard composition that would create some challenges in metallurgy development and heat treat of the rollers, it would be easy to create more issues with chain wear with an arrangement like this. I do appreciate the concept and your input.

We do solid modeling and FE analysis in house along with LN Engineering on all the update components, but I do appreciate your willingness to assist us with the design of the updated component.

Very simple, add shallow teeth to the rollers and they turn with the chain. Like I said, it's just for the concept. Point loading is what a bearing is made for. This is a very simple problem to fix.

We have to be very careful not to create a new problem while trying to solve another... This is something we fight all the time as we work to create solutions and what seems simple, generally never is.

Lots of the things inside this engine has simple fixes, but unfortunately this is not an area that is easy to access or inspect as it takes 6 hours of labor to access this component.

If we decide to experiment with bearings I'll certainly contact you...

Another occurrence
 

The second occurrence of this mode of failure in 10 months.. Yet again this one was misdiagnosed as an IMS bearing failure.

The IMS bearing is intact and has no symptoms of failure, but upon inspection we found a chunk of the IMS tensioner paddle wear pad in the oil sump. This sparked a boroscope interrogation of the IMS area through the IMS tensioner bore after the tensioner was removed.

The engine is not totally apart, but after I did a thorough inspection with the boroscope I can see that the tensioner paddle is missing it's plastic wear button and the entire wear pad has been ripped off and chewed up by the IMS drive chain. The remnants are laying at the bottom of the crankcase beneth the IMS drive as depicted here in these somewhat blurry photos from my boroscope from deep inside the engine.

The early engines had metal wear buttons for the IMS tensioer to rest against, that was too expensive so around 2001 this was changed to PLASTIC and when it wears thin and fails this sort of failure occurs since a ton of 'slack" exists in the IMS drive chain which causes the chain wear pad to be eaten up, resulting in an engine that has variable valve timing in a very bad way. The billet chain tensioner paddle that we have made with LN Engineering includes a STEEL wear pin that will resist this sort of failure; needless to say I use this in EVERY engine we build.

This engine is more than likely totally nuked. My forecast is that the majority of the pistons have had collisions with the majority of the valves and that most everything internally looks like 3 mile island.

More will be known when I complete the autopsy and fill out the failure report. For now it's clear to see that the IMS and chain are intact but that the wear surfaces and plastic wear button are FUBAR.

More pics later

BTW- Here are some shots of the billet chain tensioner paddle vs the new, revised Porsche component.. Theirs is still plastic and cast aluminum.

Go overkill.

Here ya go.. These pics are of the newest Porsche revised part Vs the LN/ Flat 6 overkill component. Ours is the shiny, heavy duty one, theirs is the cheap cast aluminum one.

Shiny doesn't mean better, if you feel the difference in the weight of the parts alone you'd know the difference.. And you can't buy a new one with a STEEL wear pad for the IMS chain tensioner!

Thank you for keeping us educated Jake.

"The Replacement"
 

I'm pretty certain Jake has done more M96 autopsies than anyone - anyone that is other than the guy(s) who work for Porsche AG tearing down blown up M96 motors - and those guys aren't talking.

This is the second failure he's seen like this and it occurred in an '03 M96/24, three chain motor with approx. 60K miles. There's not much that this 986 owner could have done to prevent this. Tic, Tic, Tic Boom!

Jake's contribution to this thread included photos of "The Replacement" tensoner paddle that he's developed along with Charles Navarro at LN Engineering. He gives you a description of the differences between the two. I have the first motor that this "Replacement" part was installed in. What Jake doesn't say in the thread is what was actually involved in fabricating the part. Once this problem was identified (N=1, now 2), there was an active decision not to rebuild my motor until this part could be perfected, because if this part breaks, there is nothing left of your M96 to rebuild.

It took three prototypes and nearly 4 months of work to get it right. Only then was my 3.6 upgrade reassembled - The guys are persistent! When I looked at the LNE website, the part cost $249 - this is an ounce of prevention.

Engine development is an iterative process. You build it, you drive (track) it, you break it, you analyze it, you fix it. Then you do it all over again. This mode of M96 failure isn't a common problem, but it is a problem. Any M96 rebuild need this "Replacement" part.

Well I hope having a 2000 3.2 with a steel rest pad rather than plastic will keep me a bit safer than most!

Nice idea with the rollers! One has to consider the chain shape produced with them especially the speeds we are talking about; I would guess flat would be better. Any bearing dia that small is going to be doing some wicked rpms assuming no skidding/wear; one will need to ensure strength/lubrication at these sites...increasing the bearing dia will weaken the the arm at points; all in a limited space.

From what I can see the LN arm has a greatly increased web area which would give increased strength. I am surprised the Porsche arm fails as it has. It must be a combination of poor castings or choosing to put a casting where one shouldn't and shock loading resulting from the rest pin failling......just a guess without seeing the set up of the engine

When the factory changed the IMS drive chains in the '01 model engines they also changed the metal tensioner contact pins to the plastic type...

Phil's engine had a metal wear pin because it was a 2000 model Boxster S, the only "S" with the metal contact pin and the old style IMS drive arrangement.

I probably won't have time to tear this core down for a while, but when I do the failure report with pics will be posted.

Thanks much appreciated.........Jake you must have been an aircraft mechanic or turbine engine specialist. Nobody else usually utters the word boroscope....

Yep, worked with T58 and T64 GE Turboshaft engines as a Marine when I wasn't crewing CH 46E Helos..

Phil calls it a bronchoscope :-)
He has been trying to find me an even more bad ass unit than what I have now.. This new one does video and allows for narration as well, its really sweet!

Chain Tensioner
 

Jake

Is it possible that the plastic button let go first and then the ware of the guide increased?

:cheers:

That is exactly what happened.

I am working on an extraction procedure for the plastic wear component that can be replaced in the car, engine assembled. LN and I already have the metal wear surface made that can replace it as a retrofit.

I can't wait to see this piece of magic