Jeff...I'm sure Grady can weigh in here...but maybe 3.o displacement weighs in as well? A mechanic I trusted completely once told me that he thought 2.8 was about max. for an MFI system with "hot cams". The MFI on race cars had different pumps than those on street cars...just my understanding, anyway...larger pistons in the race car pumps, different pump cam profiles, etc.

Now, back to Grady, who surely knows...

Just one more thought...Oregon and Washington have alcohol in pump fuel these days.

I pulled the rings out of the grooves to get a better look, and sure enough the top compression ring was broken on #1. It actually spun freely all the way around the piston, not binding anywhere in the groove, and did not separate into two pieces until I pulled gently on the end to start it out of the groove. Anyway, all of the ring lands and grooves look fine. The skirts look fine, with some light scuffing top and bottom where one would expect.

The rod bearings also look fine, but will be replaced anyway. That's a no-brainer at this point. I'll upgrade to Raceware bolts at the same time, since I now have to toss the stock bolts anyway.

The top of cylinder #4 is typical of the other 4 remaining good ones. No visible damage. The heads (#2 and #3 shown) look fine as well. I had already pulled the gaskets from the grooves, but they all came out freely in one piece. Well, except for one...

Oil is Brad Penn 20/50. The base gaskets are the stock .025, for a total of .050 of gasket on each cylinder. I found my notes, and they say I set the deck height at 1.45mm. Towards the high end (lower compression), but removing one of the shims would have set it too low. The cam bolts were still tight, the pins were still in the cam gears, and the valve lash checked o.k. at TDC for #1 before I tore it down.

I'm pretty sure the MFI pump was properly timed. It was the last time I put it back on, and the belt was still tight, so with no way of checking it now, I'll just hope it was in time. No reason why it wouldn't be.

The oil lines and front cooler were very hot. They were actually steaming from a puddle I ran through on my way through the pits. So, I'm assuming the thermostat in the filter console was opening. It's a "T" console opened up to take the stock thermostat ('72).

That's what I've got for now. Please keep the ideas coming. Paul, thanks for the words of encouragement. I'm actually not even all that "down" about all of this; it's costing me more time than money. A cylinder, a head, some rings, rod bearing and bolts, a handful of gaskets and a long Saturday puts me back on the road. I'm just curious at this point, and hoping to spare some one else this trouble. And next time, I probably wouldn't be quite so philosophical about it...

With all respect, could it be time for 2 plugs per cylinder? FWIW

Hi Jeff,
Damn, sorry that happened to your engine. However, I find your analysis along with all the other guys interpreting very fascinating and education. In medicine we would call this an "M&M" or morbidity and mortality report. But you have included the "autopsy" photos which are very cool. Thanks for sharing this with all of us.

I have missed about 6 months of the Saturday breakfast clubs (because I am training for a marathon and our group runs are every Sat. morning.) My race is this weekend and I was hoping to see your engine at the next Sat. meeting, but now this happened to you. Damn!
James

Jeff-
You said that "nominal" CR was 9.5:1. Did you measure this, or are you relying on what the pistons were supposed to be? I ask because, when I built my motor (JE pistons, see avatar for "before" configuration) I did measure, and found that the pistons that were supposed to be 9.8:1 were more like 11:1. This was in contradiction to what was marked with felt pen on the box! I sent them back to have a bunch of material milled from the domes. I'm glad I checked!

Jeff,

The mystery deepens. I don’t see obvious signs of overheating or detonation. It may be there but I
can’t see it from your images. I do see some other issues.

Most important is the failure of #1 head gasket, the damaged head sealing surface on #2 yet the
still perfect sealing surface on #3. That sorta indicates the entire 1-2-3 head/cam housing assembly
was firmly attached at the #3 cylinder and loose at the #1 cylinder (with the blown gasket).
Is this a cause or result?

Lets discuss the images individually.

1336a
http://forums.pelicanparts.com/uploa...1209435665.jpg
Here is one of the rod bearings (all three look the same).
The marks here (red arrow) probably is from construction debris in the oil system. Not germane here – I think.
The area (yellow ovel) doesn’t show any evidence of damage but I think this needs very careful
investigation on all six bearings.



1494a
http://forums.pelicanparts.com/uploa...1209435736.jpg
The #4 cylinder shows compression leakage past the head gasket on the intake side.
That is very unusual.
It appears there is some starting on the bottom (exhaust) side also.



1525a
http://forums.pelicanparts.com/uploa...1209435795.jpg
This appears to be #2 head. Note the damage to the sealing surfaces. This damage appears to
be from ‘pounding’, meaning the mechanical beating between the head and cylinder as if the
head(s) were loose.



1555a
http://forums.pelicanparts.com/uploa...1209435896.jpg
Here are both #3 and partially #2 heads. Note the perfect head-to-cylinder sealing of #3 (teal arrow)
and the beating of #2 (red arrow).



1229a
http://forums.pelicanparts.com/uploa...1209435963.jpg
This is the underside of the #1 piston I assume.
If there had been serious overheating I would expect discolored oil deposits here (brown oval).
This is where knowing the type of oil is important. This appears way too clean for heat to be an issue.

That said, the compression leak (blown head gasket) may have prevented this piston from getting hot
enough to show any heat. What do the other pistons look like on the underside?



1262a
http://forums.pelicanparts.com/uploa...1209436031.jpg
I assume this is the #1 piston.
Is this (red arrow) a crack or a simple scratch?
The color on the piston above the top ring indicates to me too much piston-to-cylinder clearance.
Clearly the top ring is doing its job as there isn’t any indication of combustion gasses below the ring.

If you re-install the top ring and go around with a feeler gauge, I suspect you will find equal clearance
and no ‘tight spots’.



OK, so where does this lead?

I’ll propose another theory: The heads got loose starting with #1 cylinder.

WHY? Were they not torque properly? Is there something else at work here?

After the engine was built and run, were the heads re-torqued? If so, how (sequence)?
What other intervening maintenance?


Do you have the 4-5-6 side apart yet? If not STOP and lets plan some tests.


Best,
Grady

Your original conclusion seems the highest probability.

How did you set the A/F's originally?

Did you just accept what the pump builder set it at on with the bench test?

With a factory set up, it is adjusted at cruse. With a factory space cam that is designed to match the motor, this is adequate. However, that may not be enough with a custom set up.

Did you verify your A/F's on a dyno, with a EGT, or wide band?

I second the injector test. Any diesel shop can do this.

Sorry Grady, too late - the other side is apart already. It has some of the same marks on the sealing surfaces, but not nearly as pronounced. #5 and #6 look pretty much like the top of the #4 cylinder shown above. The heads on #4 through #6 show evidence of some pounding as well. Thanks for pointing this out; your practiced eye is seeing things in my photos that I'm not seeing in real life. I'm assuming this needs to be cleaned up before it goes back together.

All of the pistons look like #1 on the underside and the top side (yes, that is #1 in the photo, and that is only a scratch from the end of the ring). No discoloration on the bottoms, and pretty brown down to the top ring groove. According to the paperwork that came with them, piston to cylinder clearance was set at .015".

The heads were re-torqued at 1,000 miles, following the same sequence used in assembly that is outlined in Wayne's rebuild book. Quite the pain in the rear end while the motor is in the car, by the way. Anyway, they were initially torqued to 24 ft lbs, and re-torqued to the same. Actually, more like "checked" - none needed re-torquing. The other intervening maintenance was the usual valve adjustment and oil change, first done at 100, then 500, then 1,000 miles. Oil changes are at 3,000 after those first three.

I keep going back to the burned plugs and discolored exhaust; along with the high oil temps. It was clearly running hot, but the whole cooling system appears to have been functioning properly (front oil cooler and engine cooling fan). Those JE pistons do have a very wide top land, and a very thick dome. I wonder if that was enough to eliminate the other symptoms of detonation and high heat - the pinched down top groove or the oil burnt to the bottom.

I used an LM1 with a wideband sensor to set it initially. I then verified the results on a local chassis dyno. The A/F was running between about 11.5:1 and 12.8:1 across the range at full throttle, and about 11:1 at part throttle cruise. The exhaust pipes were a very dark gray, or even black. After running the small M&K's at this track day, they came out a very light gray. Nowhere near as sooty as I'm used to seeing from MFI. Thanks for the tip on the diesel shop; I will find one and get the injectors tested. Where would I find the spec for opening pressure and flow?

Jeff,

You did everything ‘right’. Something went wrong.
We really need to find ‘what’.

I generally agree about the indications of over-temperature. IF that is the case, then why?

Is the distributor still in place? Can we determine the ignition timing advance? Once the distributor is out, we can check its function on a test bench.

The other issue is the MFI system. Could something failed there somewhere? I agree with checking the nozzles. That is easy. Could something have stuck in the pump?

Looking at the head images you posted, they all appear to have the same ‘gun metal gray’ color. That is usually characteristic of detonation but not definitive. Do all six appear the same color?


Tell me more about the sequence of overheating. Was there any indication on the first day? Did it occur first thing on the second morning or later in the event? {{VERY IMPORTANT}} what was done between when it ran properly on the first day and the temperature problem appeared on the second day? Nothing is insignificant!

What fuel? What else? Was the 911 secure? I want you to think of everything.


Lets assume your original theory is correct. How do we confirm that? Clearly we need to have the pump and nozzles inspected. I suspect we can enlist one of the primer pump specialists to do a quick test. Nozzles are easy, even I still have a tester. Any diesel shop can do this.

A clue that it isn’t one nozzle or MFI pump cylinder is the #1 cylinder appears as the others.

This can just as well be an ignition issue. Do we absolutely know what the ignition timing was during this time? Is it possible to have multiple firings before normal ignition timing? Ignition advance of 40-90º can have similar effects. I have a holed 2.0T piston to prove it.

Lets discuss gasoline. Please take several gas samples in sealed containers. We may want to submit them for testing if all else fails.
What do you know about the trail and source of the fuel? Is it possible that the fuel you probably added for the second day was low octane or contaminated? We need to carefully examine the contents of the fuel tank. Water and gas can separate out. The MFI system stirs everything up.

I’m still concerned with the apparent compression leakage from many cylinders. We need to confirm with measurement that your pistons are really 9.5:1 CR as installed. If they are somehow 13:1, that would be a possible explanation.

I think these are the major clues:
The engine ran hot.
There was a change between the first day and the second.
It appears the heads came loose.
All six cylinders appear the same.

Please add others….


I feel we need Sherlock Holmes and Dr. Watson here.

Without absolutely, positively knowing what happened here, we can’t be sure it won’t happen again.

Best,
Grady

Do they still sell leaded race fuel. If you ran that you would get a white exhaust.

Maybe some one has the injector specs. I do not have my old MFI stuff anymore.

More important is that they all flow the same. They operate at about 230 psi and I think they first open around 180 psi or so but I do not remember. The diesel shop is used to this. They cam test for opening presure if you want. If they are close, no problem.

I would mark the one injector on the cylinder that went bad and let them know you want a quanity of flow test. You can mark each one if you want.

Have them do a test first. Then a cleaning and retest if you want to know if you had a bad injector and get them ready for next time.

Your note of going from a failed old muffler to what is basically a straight through pipe should be significant.

Another thought, did you put your O2 on one side or both.

I'd bet you are correct in your initial assessment about a change in muffler. MFI is known to be sensitive to muffler choice.

First, sorry to see that you're having a problem, especially after that outstanding journey of discovery on the MFI pump. Also, I want to acknowledge that you all have much better MFI know-how than I do since I'm an MFI noob.

However, one bit that I haven't seen mentioned is the fuel boost pump. If you are driving the engine hard - track - with a boost pump that is marginal, is it not possible that your flow demands exceeded the capacity of the boost pump and thereby leaned out an otherwise perfectly functioning MFI pump? Combine that with the leaning effects of a different exhaust and you could create a very lean condition. Ditto for the tank sock (do you have one?) and MFI filter.

Eh?

Hello Jeff
would you tell us more about the assembly :
were the cylinders the same height, heads too ? were the head mating surface with cam tower right aligned (top), a head gasket couldn't have been crushed during assembly ?
Hope this helps with my non academic technical english, i've seen a similar event with unaccurate top end assembly, hot gazes ruined the head mating surface through a bad gasket...
I think your AFR's were right if this are the numbers you said (at least rich enough)
and congrats to M. Clay analyses...

Grady, this happened at a single day event. I'm sorry; I probably wasn't very clear about the time line on this. The previus track event was three weeks prior, under similar weather conditions. The car ran consistantly at about 210 degrees that whole day with no issues whatsoever. It had the failing Leisritz on it that day. All plugs came out with that kind of sooty MFI rich fouling on them, so all appeared normal. The only change made between the two events was the mufflers. That's why I keep going back to that.

The distibutor is still in place. I'll check the timing when I get it back together, but it looks like the body has at least not moved. The rotor can be advanced by hand, and springs back just fine. It was totally rebuilt, with all new bearings and whatnot, by Barry Hershon just prior to assembling the motor. He has been at it for a long, long time, and certainly knows his stuff. I would be surprised if anything failed on that end. Even if it did, and stuck at full advance or something, that shouldn't matter. I asked him to set it up for full mechanical advance, and to have full advance come in very early - like 2,000-2,500 rpm. The motor never sees below about 3,800-4,000 even in our tightest 2nd gear corner, so it's running on full advance all the time. I set that at 33 degrees, and I'll check to make sure it is still there when it's running again.

Jim, you might be on to something. One of my friends that runs a local vintage series with a carburated 2.0 had similar symptoms occur last summer, on a fresh rebuild. It burned the ends off of several spark plugs and lost compression in a couple of cylinders. His compression loss was through failed rings, and did not damage the heads or cylinders. His did show significant pounding on the outer rod bearings (piston side of rod). He traced it down to a failing electric fuel pump. It delivered full volume until it heated up, then slowed to a trickle. So, it would pass the "beaker test" on the first beaker, and fool all of us. By about the fifth beaker in rapid succession, it was only running at about 10%. So, I'll look at the supply side and check my fuel pump flow. The filter itself was brand new with the new motor. The car only sat for a week between dropping the 2.4 and installing the 3.0. It had been driven several times a week for the five years I have owned it, so I don't believe it's a plugged screen (from sitting for an extended period of time).

All six heads do appear about the same color. The pump is about a five year old rebuild from Gus that was never installed on anything. It's spotlessly clean inside, and everything functions very smoothly. It responds exactly as one would expect to the various mixture adjustments. All of the hard lines had pressure in them when I took them off. The injectors themselves were an eBay find; advertised as freshly cleaned and tested. While I've purchased other items from this seller, and they have always been as good or better than represented, these are looking like a possible "wild card". The motor did, however, run exceptionally well for 5,000 miles and a couple of track days with them. Plugs always looked identical from all cylinders; that kind of sooty black characteristic of slightly rich MFI. So I believe they were all functioning properly, but I will get them tested. I've got a couple weeks to kill waiting for parts anyway.

A couple of us drove down to the track together and filled at the same Chevron station. Nothing in the other car indicated a problem with the gas. It was just a normal single day club lapping event, so it was never left unnattended between filling and running. Couple hundred folks milling around the pits all day, so it's unlikely some one would have fooled with it. Anyway, it's just straight pump unleaded premium; 93 octane in our area. If my compression is higher than the 9.5:1 advertised, that may be a problem. I'll call JE this morning (I need some wire lock rings anyway) and verify the part number with them. When it's back together I'll get a cranking compression number.

The other thing I think I'll do is to go ahead and tune it with the new M&K sport muffler that's on it's way. Then I'll put the failed Leisritz back on, and see if it's plugged up enough to make it go super rich. Then I'll put the little track mufflers on and see how much leaner they make it run, if at all. That will either validate or invalidate my assumption about the muffler change.

Jeff-
To reiterate, I wouldn't trust JE about the CR. I'm sure that if I had called them, they would have said that I had the right thing (what was indicated clearly on the outside of the box). You can measure your CR by purchasing a calibrated burette and making a plexiglass window to attach to your heads. I don't know much about SC heads, but I do know that others can vary significantly in combustion chamber volume, so it would be tough for a piston maker to determine what your CR might be, particularly when the variable of deck height is added.

Just getting it on the table.

I think after a rebuild, one is supposed to re-torque the heads with some miles.

Fuel in oxygenated here in California. Might be good to add a little high-octane at the track for some insurance.

Were the heads cleaned up or fly cut? Did you cc the heads to check compression with the assembly?

Still vote mufflers. Add an EGT for insurance.

Hmm....no scruf on the piston crown yet hole was almost torch-welded trough head seal? Weird.

We were discussing this on 930 forum. Some knocked piston pictures coming:

930 piston, knocked dead:

http://i139.photobucket.com/albums/q...ltedPiston.jpg

Waterpumper knock with JE pistons, head looks like this:

http://forums.pelicanparts.com/uploa...1209408040.jpg

But then, piston looks like this:
http://forums.pelicanparts.com/uploa...1209407353.jpg


Hard to say but you might have had some sort of mechanical malfunction which let piston/head clearance grow and subsequent leak severely disturbed AFR's inside combustion chamber and then torched trough the crack in the seal. If piston look healthy then there might be another explanation...

This is of course, pure speculation from my point of view. Let the experts decide.

Jeff,

Before you start on a ‘final assembly”, it would be wise to mock-up a single piston, cylinder and head and measure the combustion chamber at TDC. You don’t need the cam housing. You need to position the engine so the sparkplug hole is up. This is somewhat ungainly but actually pretty easy.

I grease the valve contact area and grease the cylinder before I bring the piston up to the top. I also grease the head gasket. Wipe off any excess.

I use a 100 ml burette and regular Stoddard solvent.


I’m beginning to suspect that you may have much higher CR than 9.5:1. Actual measurement is the only way to tell.

Too high CR might agree with many of the clues.
The ‘gun metal grey’ color in the combustion chamber.
All six look the same.
The burned electrodes on the sparkplugs.
The problem happened after refueling (you may have had just enough octane prior) and your friends were probably dealing with their 8.5:1 CR.
The lean effect of changing the exhaust.
I didn’t say so before but I tend to use ignition advance in the 30-32º range for an engine like yours.

I Don’t see any clues that are counter to too high CR.


I recommend you measure the CR asap. If you find you have 11:1, you haven’t wasted any time on assembly and these pistons. This probably should have been done during the original assembly but that is water under the bridge.

If it turns out this is detonation from too high CR, we need to look at every other component that could have been damaged. The rods come to mind.

You are going to need to repair (at least) the damaged #2 cylinder and head. This will give you time to get new pistons fitted to cylinders (talk to Steve Werner at Rennsport) and measure the CR for them.

Best,
Grady

Jeff, Sorry to here about your problems. Thanks for sharing the info.

Grady, You are amazing. I have learned so much from you and the members on this board. Thank You :)

Something has been nagging at me since we started this analysis. Something Grady has mentioned several times. Something that should be a "no brainer", and one of the simplest operations in engine assembly. Head nut torque. In my efforts to consider every detail, no matter how seemingly insignificant, I had one of those cold chill, "oh crap" moments this afternoon at work. Now I don't know what to think.

Now just about any numbskull, including me, can read a torque wrench and torque a nut. But what if the torque wrench is off? In an effort to "do this right", I borrowed a clicker type 1/2" drive torque wrench, forsaking my beam style I've used for years. The 23.4 ft lbs called out for the head nuts is way at the low end of the adjustment range for this wrench, which I thought nothing of at the time.

My "oh crap" epiphany came today when I was shopping on line for one of my own. I saw a lot of click type torque wrenches that said right in their product descriptions that they would not guarantee accuracy below a certain percentage of their adjustment range. You've got to be kidding - why even put numbers in a range where it is acknowledged to be inaccurate?? And if a brand new one isn't accurate in its lower range, what about one with some miles on it?

So I'll be retrieving that very wrench from my buddy tomorrow. I need to test it to put my mind at ease. I'll torque a few different nuts and bolts throughout its range and see how the values compare to my beam type. There is probably nothing to it, but, what if in my quest to be as precise as possible, I actually used the inferior tool?

I almost didn't post this, feeling kind of silly about such wild speculation. Maybe I'm more just "thinking out loud", where no idea is a stupid idea. Or, as Grady put it, no detail is insignificant. We call this "thinking outside the box" at work. If the components look correct, if there is no obvious answer jumping out in front of you during their inspection, then well, maybe they really are just fine. Maybe one of the very simplest process steps failed for some reason.

Boy will I feel stupid if there turns out to be anything here. Maybe the real lesson will be "stick with the tools you know". We'll see. I'll report on this tomorrow. I'm just waiting for parts now anyway, so this will give me something productive to do. It will answer that nagging voice, anyway.