M42 Head Gasket Blown, Project Thread

[bmwman91]

So far no updates. I have been extremely busy and MM doesn't get much done during the big holiday season. Now that we are past all of that, I have been communicating with them about formulating a game plan for the rebuild. I finally have enough time to give proper attention to this, so I will hopefully have more to post in the near future.

[Jordan]

Hurry up slacker

[bmwman91]

Ha, yeah I've been feeling the slack. Too many other projects, including wrenching on a 2006 Jetta TDI that I got from Craigslist a few months back. The E30 has spoiled me...wrenching on FWD cars is a much bigger pain in the butt than a 30 year old 4 cylinder in an engine bay that can fit a V12! Still, it is nowhere near as awful as a clutch job on a Mini Cooper S.

[bmwman91]

OK, here's the plan. MM has my engine apart and got a good look at things. As far as they can tell, there was no damage to anything else other than the head gasket, which had managed to fail between all cylinders, as my peek with the borescope indicated. Cylinder 3 was the worst offender, having totally blown out into cylinder 2, and small leaks between the others. The primary cause is likely to be the failed damper/trigger wheel which caused some significant misfires last year, and it is possible that inconsistencies in head bolts used at the time compounded things. The engine now has a new damper wheel, and MM now uses beefier studs with larger shanks and a finer thread pitch which apparently leads to extremely consistent installation torque behavior. Anyway, the head needs to get a thorough inspection and pressure test since the HG failure seemed to be between cylinders, but coolant and oil were mixing. Fingers are crossed that the HG was just lifting or combustion pressure was pushing stuff where it did not belong, rather than a crack in the head.

Cylinder bores were ~0.001" out of round, which is not great but not awful, and had none of the damage that indicates detonation. Pistons and rods were still good, but I am having those tossed and the lighter currently-used stuff is going in. Yeah it's a mild cost increase, but I sort of want everything on the bottom end out and replaced, and my employer's stock is riding the current ZIRP bubble so I might as well take advantage lol.

I am going to be heading out to Missouri sometime in the next week or so to visit MM for a couple of days, go over what all is being done, make sure that a couple of things that weren't done right back in 2012 do not get repeated, and just enjoy watching some of my engine being built.

[digger]

MM are great at upselling lol.
i'm not sure how missfires cause a HG failure but nevertheless get some pics of the new parts going in and the build process if you are able.
If the head is scrapped then time for a nice ported one

[bmwman91]

I am guessing that the HG install was probably compromised from the start, and the misfires finished of an already weak interface. That's the only explanation that seems plausible for the major failure between 2 & 3 and the smaller ones elsewhere. It could be anything though, hence the block and head getting thorough inspections and re-machining (if not damaged). I am just stoked to get out there to observe how they put it all together, and with any luck I'll meet up with Sssquid tuning to shake hands and whatnot since I've been working with him on Motronic reverse engineering stuff on and off for a couple of years now.

I'm sort of glad that I've had a few month of non-running E30 life since it has given me the change to start digging deeper into the Jetta TDI that I bought and make a plan for how to make it a little more fun of a DD (mainly, ditching the sluggish DSG timebomb for a manual). The Pumpe Deuse engine in it (BRM) is sort of a sweet spot between the original VE pump engines (ALH, etc) and the later common-rail ones...no DPF, no Dieselgate drama, 27000PSI injection pressure like the CR ones, without the insane complexity of the CR's, but a good deal more tunability and ECU-controllable stuff than the earlier ones. PD engines did have some notorious cam/lifter wear issues, which are mainly a function of people not using 505.01 rated oil and not changing it. Then again, this is "that" / A5 generation of VAG vehicles where every single plastic trim clip inside the cabin will break as soon as you sneeze and make rattles everywhere. It's a solid DD though, and it is amusing as shit to smash through forest backroads with the crappy M+S tires the PO left on it. Rwoooooooobbbbbrrrrrrrrrreeeeeeeeeewwwwwwrrrreeeee eeeeeoooowwwwwwww.

[digger]

what i mean is that missfires arent hard on the head gasket, high pressure and temps are (excluding the mechanical clamp aspects). missfires dont produce much of either (its like the opposite of knock) but a tune that is suboptimal wouldnt help if it wasnt rich enough and too much timing.

whats the bore on the HG? ive never like the M20 being smaller than the bore on the MM setup as it exposes more of the fire ring to heat. 87mm bore with 91 centres doesnt help

[bmwman91]

Back when the damper went (months before the HG blew), I had been driving the car hard. So, mid-high RPM/near-max ignition advance and full load, with bogus crank position readings. While I really only heard the misfires, I have no doubt that it was also detonating. Here's the reasoning for that.

Having the damper wheel go loose was delivering a mix of plausible tooth pulses, but with increasingly incorrect phase. The rubber ring that connects the outer toothed part to the hub normally allows a small amount of rotational shear to occur since that is how these dampers work by definition. Rotational shear is the main stress that the rubber ring is subject to as the crank accelerates in both directions during and between power strokes. In my case, more than 180° of that rubber connection was severed, and probably close to 270° of it could be considered compromised, so the angular disparity between the toothed portion and hub was likely enormous as the crank accelerated during power strokes. The measured crank position would be severely retarded during the productive part of the power strokes as the crank accelerated in the forward direction, and it the position would be severely advanced in between peak power strokes as the crank slowed and the outer ring overshot. Once misfires and/or detonation got into the mix, I think that the effect would have gotten worse and essentially been a positive feedback loop until the Motronic identified that the crank speed was varying too much to be plausible. I know that later Motronics would identify misfires by measuring crank acceleration after each ignition event, but M1.x did not have that functionality (or any of the 1MHz 8051-based ones). Things have to get really wonky before it figures out that something is wrong.

This is what the damper looked like when I removed it.

https://www.r3vlimited.com/board/forum/e30-technical-forums/engine-drivetrain/m42/386826-m42-won-t-idle-some-backfiring-edit-photo-dump-of-broken-stuff

The block is bored to 87.5mm, although MM now uses 87mm, and the M44 HG is from an engine with an 85mm bore. Sssquid has worked on a number of MM 2.1L M42's now, and as far as he heard nobody else has had these issues. You make a good point about the fire ring being subject to greatly increased thermal abuse like this.

[ForcedFirebird]

I was going to mention detonation blowing the gasket before reading all the latest posts. I had an m20 grenade in a similar fashion, except it was a loose crank bolt that caused the key way to get gouged out. When discovering the issue, I put a digital angle finder on the crank just to see how bad the variation was, and was seeing like 25° advance and 18° retard. Detonation was so bad in #1, it ate a small hole in the piston crown, and there was clear evidence of all the cylinders detonating (aluminum dust everywhere).

[bmwman91]

It sort of irritates me that BMW put the 60-2 wheel on the damper. The M44 & other later engines have it on the back of the crank where it is rigidly attached, which seems like an inherently better location. Of course, I also wonder if maybe the older, slower Motronics would have had a problem with keeping up with a rigidly mounted one since it would see a lot more high-frequency acceleration components from individual cylinders firing, so they put it on the damper to get a mechanically low-pass filtered signal for crank position. The internal timer/counter in the SAB80C515 MCU runs at 1MHz, which is still "fast" compared to the engine's mechanical processes, but maybe Bosch found that direct/rigid mounted position wheels caused issues. All speculation, of course.

[digger]

Back when the damper went (months before the HG blew), I had been driving the car hard. So, mid-high RPM/near-max ignition advance and full load, with bogus crank position readings. While I really only heard the misfires, I have no doubt that it was also detonating. Here's the reasoning for that.

Having the damper wheel go loose was delivering a mix of plausible tooth pulses, but with increasingly incorrect phase. The rubber ring that connects the outer toothed part to the hub normally allows a small amount of rotational shear to occur since that is how these dampers work by definition. Rotational shear is the main stress that the rubber ring is subject to as the crank accelerates in both directions during and between power strokes. In my case, more than 180° of that rubber connection was severed, and probably close to 270° of it could be considered compromised, so the angular disparity between the toothed portion and hub was likely enormous as the crank accelerated during power strokes. The measured crank position would be severely retarded during the productive part of the power strokes as the crank accelerated in the forward direction, and it the position would be severely advanced in between peak power strokes as the crank slowed and the outer ring overshot. Once misfires and/or detonation got into the mix, I think that the effect would have gotten worse and essentially been a positive feedback loop until the Motronic identified that the crank speed was varying too much to be plausible. I know that later Motronics would identify misfires by measuring crank acceleration after each ignition event, but M1.x did not have that functionality (or any of the 1MHz 8051-based ones). Things have to get really wonky before it figures out that something is wrong.

This is what the damper looked like when I removed it.


The block is bored to 87.5mm, although MM now uses 87mm, and the M44 HG is from an engine with an 85mm bore. Sssquid has worked on a number of MM 2.1L M42's now, and as far as he heard nobody else has had these issues. You make a good point about the fire ring being subject to greatly increased thermal abuse like this.

Honestly if not already id be getting them to use a torque plate, on pretty much every stock bmw bottom end there are visual signs of differences of wear patterns near the bolt holes and taking that much out wont be helping ring seal.

Plus i'd also want an MLS gasket, if the stiff part of the fire ring doesn't sit properly on the block because there is a radial difference of 1.25mm the seal is weak because the local contact pressure is reduced

on the m20 its not so bad with 86 vs 85.5 gasket as the radial difference is 0.25 mm

[ForcedFirebird]

Actually the other dampers are the same. There is still rubber between the reluctor and crank. One of my clients actually found the limits of an m20 damper - they don't take to 8,000 RPM in door-to-door competition....

Post # 81: https://www.r3vlimited.com/board/for...70#post9874970

[digger]

i think he means the toothed rings that hard mount to the crank inside the block

[ForcedFirebird]

i think he means the toothed rings that hard mount to the crank inside the block

Derp.

[bmwman91]

Yeah, those ones, like the M44, M52, M54, etc.

Hopefully MM is using a torque plate when they bore/hone. I'll find out soon enough. I'll also re-visit the MLS HG topic with them.