Weekly Northern NJ Meet, Sundays 12:00

[Emre]

So: the M20 with billet rockers, crank scraper and accusump (apparently the I-J crank scraper is insufficient, according to several posts) should be able to use all its stock RPM to rev limiter and get from april to november.
Conversely, the M20 stock will blow up.
Equally, it would appear, an S52 (not M50, not M52, not S50) with a pan baffle and a 7000RPM or below rev limiter, running 15W50 or 20W50 oil, should be able to do the same and get to november.
Conversely, the other options in 24v will blow up.
Perversely, by the by, the 'delicate' Euro S50B32 will suffer a lot less (otherwise, among other things, I'd be at my third motor in my euro car).
And mind, in both cases, we're talking normal cars in HPDEs. Not racing slicks with Gruppe N suspension and cages in real races.
I learned my lesson: see you at the track - with a 964. The 318/S52 will make a very nice DD.

Whichever engine you choose to run on the track will have it's own quirks and niggles that you'll have to deal with. If you think an S52 is unreliable and expensive, wait till you get your hands on that 964

Beat the crap out of any engine and it will fail. Drive the car within the parameters of its design and it will last. I have many seasons of hard track use on my current M20B25. It's bone stock with over 150K miles on the clock. No crank scraper. No Accusump (yet). No fancy synthetic race lube. I've had zero issues with it.

An M20 was simply not designed to rev up to 7K rpm...no matter what any internet chip salesman might tell you. Besides, the power drops off after 5500 rpm. Without high lift/duration cams and a totally different intake manifold, there's very little point in pushing it beyond 6K rpm to begin with.

[FredK]

How did your S50 US blow up, anyway?

There are a few ways that S50USs blow up. Excessive revs, bouncing off the rev limiter for a little while, oil pump nut backing off, running a non-factory crank nose harmonic damper, or dropping a valve in <10/95 production engines, which used top valve spring plates that were fairly soft.

[nmlss2006]

Emre,
the problem is, I know 964s. And the short version is, given what the E30 has cost me, given what I'm learning about the *actual HPDE profiles* for these cars and given the relative performance levels, I'll take my chances.
I understand if I beat the crap out of any engine it will fail - however, by any reasonable standard, I was NOT beating the crap out of the S50.. and it failed and noone can tell me why. I drove and drive a GT1 block 996 much more aggressively with 200 more HP than it should have stock and a 7200 RPM rev limiter... and the engine is still there. Heck, I run an Euro S50B32 in europe about six times as hard as I did the S50 here, and the engine is still there. Granted, I bought the car new and it has 80kmi now, so mileage is certainly a factor, but this is the data I have to work with.
As to the M20: I agree that it was not designed to go up to 7000 RPM. That is fine. But it was designed with a stock rev limiter of 6250RPM - IIRC, it was even higher in EU cars. And you're telling me that going to the stock rev limiter in the relatively limited period of time that we see in an HPDE - that is the other point, an HPDE is in many ways milder than me driving up a mountain pass in europe AND certainly shorter - may cause issues. Allow me to be disheartened.
Now, I will grant you that there is one other fundamental difference: noone I know in europe would dream of taking an engine with 200k km - 125kmi - on the track and I certainly was very VERY skeptical about it myself. However, everyone's assurances were absolutely unassillable: 'don't overrev it, mind the oil level and you'll be fine'. I was - for 8 x 25 min?
A final point about the M20: I'd have to see a trans scheme, but if power is still above what you get after you shift, it's worth going to the rev limiter - and AFAIK, at least in 3rd and 4th that is the case. Does anyone have a *factory* dyno plot for the M20 stock?

Frank:
my S50 grabbed the #1 rod bearing, spun it and cracked the block. The top end seems fine (and will be for sale shortly...). The oil levels were fine. I did not mechanically overrev the engine. It was a ~116kmi engine that had a *very* well documented maintenance history AND looked absolutely clean. The only mods that were done to it other than removing the fan/fan clutch and replacing the oilpan for obvious reasons were a KAmotors (from here) CF intake to get cool air into the engine and a MarkD chip. I was running BP premium gas FWIW (I usuall run Shell Vpower, but unfortunately up in the sticks I was lucky to find unleaded... :) ).
I am a bit confused by your statement that 'bouncing off the rev limiter for a little while, ' will cause the engine to blow up. I assume that's an exaggeration or something I'm misunderstading.
My fundamental issue at this point is as follows: let's leave the M20 aside for now, other than the considerations above, because if I am going to run a BMW at the track, it's going to be this 318 with an S50/S52 in it.
People have been taking these engines to the track for 13 years now. I will not dispute that the *reputation* of these engines is more or less faultless: 'keep the oil level up and you'll be fine unless you misshift'. It has been argued that the E34 oil pan actually is a BETTER pan for tracking because it allows for less 'sloshing' room and therefore for less pickup problems, in theory. And yet, when you actually go TALK to people (Stickley, Yaskin at Bavarian, several others) you get: 'it shouldn't happen! But hey you know, I have this customer who... and this other customer who...'.
Therefore, now I'm back to seeing it as I did before: <60-80kmi engines, with a very proven maintenance history AND driving history are probably fine. The rest are at the very least unknowns, and I won't go to the track with an unknown.
So we're left with rebuilding an engine: here it seems to cost $7-9k and a RMFD one from BMW runs $12,900. I won't go into the fact that $8k to rebuild an engine with stock parts is lunacy: the US seems to have accepted that $90/hr is a perfectly fine rate for mechanics even when their hours seem to be comprised of 15 minutes instead of 60. I will simply remark again that at this pricepoint there are many better cars which are definitely happier at the track. Unless someone can persuade me otherwise: I am listening.

[kishg]

shit happens.. however, the statistical evidence is against you. there's no empirical data that shows s5x is inherently unreliable on the track. as i said, i've seen numerous cars running 20-30 days a year for many years without issues. going to stock rev limiter for short periods should be fine on an m20. problem is most of us have chips :)

[FredK]

my S50 grabbed the #1 rod bearing, spun it and cracked the block. The top end seems fine (and will be for sale shortly...). The oil levels were fine. I did not mechanically overrev the engine. It was a ~116kmi engine that had a *very* well documented maintenance history AND looked absolutely clean.

Allow me to put on my "armchair mechanic" hat on. ;)

Were you able to look at the other rod bearings for wear patterns? I assume you were using 93 octane with a 93 octane MarkD chip?

Do you monitor oil pressure? How much of an overfill do you run? Did you ever get an oil analysis?

I am a bit confused by your statement that 'bouncing off the rev limiter for a little while, ' will cause the engine to blow up. I assume that's an exaggeration or something I'm misunderstading.

My friend hammered an S50US off the rev limiter right before the stop box at an autocross for a few seconds. This caused light piston-to-valve contact. The head had to come off and get rebuilt.

Therefore, now I'm back to seeing it as I did before: <60-80kmi engines, with a very proven maintenance history AND driving history are probably fine. The rest are at the very least unknowns, and I won't go to the track with an unknown.

There aren't many guarantees on a 60-80k mile engine with a maintenance history. What really entails a maintenance history? Oil changes and a timing chain tensioner change every 50k miles?

So we're left with rebuilding an engine: here it seems to cost $7-9k and a RMFD one from BMW runs $12,900.

True. You could get lucky and pick up an engine from a racer who has either logged a short amount of time on an engine, or a spare, for a bit less than that. Or, you can do a refresh yourself, though this takes a bit of effort.

A Varioram costs around 8K used. I would hate to find out what a rebuilt costs, especially one done by a reputable Porsche engine builder.

[nmlss2006]

See, that's again where we agree to disagree: if 'the statistical evidence is against me' then shit does not happen. If shit happens - and this is precisely what I have come to accept, that the community agreement is that these engines are fine because 'shit happens' but in reality, if you expect the engine to last you're in for a bad awakening - and it happens during HPDEs on stock cars, then from my perspective these engines are 'inherently unreliable though you may get away with it perhaps even most of the time'. 'Reliable' means that 'shit does NOT happen' as long as you're within operating parameters and 'shit happens occasionally' if you skirt the edge. That is the industrial definition and, BTW, the definition that would cause mechanics' skulls to remain intact: a lot of friends with a lot less patience than I would have simply FED the engine to their mechanic: you do not blow up an engine on your first exit to the track or if you do, whoever put the car toghether foots the bill and thanks God that you don't ask him compensation for the time you wasted.
It's sort of like saying that computing fruit is reliable :) :) :).

[nmlss2006]

Allow me to put on my "armchair mechanic" hat on. ;)

Were you able to look at the other rod bearings for wear patterns? I assume you were using 93 octane with a 93 octane MarkD chip?

Do you monitor oil pressure? How much of an overfill do you run? Did you ever get an oil analysis?

Please do - more than anything else, I am looking for ANSWERS, at the present stage.
The other rod bearings are being taken apart. The concern is that the crap that went through the bearings subsequent to failure of #1 may have affected them *after* the fact, rendering a post-mortem more difficult.
And yes, I was running 93 oct - with a 91 oct MarkD chip, per his words, so detonation shouldn't have been an issue. Ambient temp was in the 50s and it was cloudy, so ambient heat was not a contributing factor.
I was not monitoring oil pressure, I don't have an oil pressure gauge in the car. From previous experience and given the speeds I was going, I assumed that this was VERY far removed from being a factor.

My friend hammered an S50US off the rev limiter right before the stop box at an autocross for a few seconds. This caused light piston-to-valve contact. The head had to come off and get rebuilt.

See, this shouldn't happen: the stock rev limiter *should* be low enough to prevent this rather comfortably. The rev limiter IIRC is at 6900ish and valve float should occur above 7500, I'm told.

There aren't many guarantees on a 60-80k mile engine with a maintenance history. What really entails a maintenance history? Oil changes and a timing chain tensioner change every 50k miles?

Perhaps not, but there are a lot more, I'd say, than in an engine with 150kmi on it: all the tolerances are going to be closer to factory spec, metal fatigue is going to be a lot lower and wear on bearings/seals is going to be lower.
As to 'maintenance history', the best situation is factory recommended maintenance done on time *every time*, in the hopes that that is indicative of a particularly attentive mindset. Other oil changes of which the US market is enamored may be good but generally offer little benefit *with proper driving and decent, modern oils*. One of these days, by the by, someone will explain what people are thinking when they put mineral oils in modern engines in 2009, but that's another question for another time.
The problem is, the *driving history* is what matters more. If an engine has been broken in properly, subsequent to that never saw >1/2 throttle and 40% max RPM before water AND oil temps were above 80C, has always had good oil in it and has seen very little commuting/city driving, then its condition at 100kmi will be materially better than an engine that has had its oil changed RELIGIOUSLY with mineral oil every 1500mi... and has been taken to redline every cold winter morning that it's been driven, sitting in commuting traffic for hours at a time as a bonus.
The problem is, when you buy a used engine... you just don't know what happened to it. So either the engine is indeed 'inherently reliable' and it shrugs off the HPDE environment even in less than ideal wear conditions, or it is not and you must accept that... the engine won't be reliable.
I have come to understand that my expectations are inherently different from the majority: I won't accept this kind of behaviour from my street car, much less so from a track car that cost about 10 times what E30s seem to go for these days. If my mechanic tells me 'you're going to be fine, see you in 5000mi' I *expect* to have no problems for those 5000mi. Not 'shit happens'. Not 'whoops'. None. And for that reason, I will pay the $90/15minutes that seem to be en vogue, grousing all the way - but you better believe that the moment that 'whoops' happens, I will be back and it won't be pretty. At that pricepoint, there is no room for 'whoops': I can hire engineers with 10 years of experience for less money and they will fully understand that 'whoops' means 'very fired and do NOT use me as a reference if you care for your career'.

True. You could get lucky and pick up an engine from a racer who has either logged a short amount of time on an engine, or a spare, for a bit less than that. Or, you can do a refresh yourself, though this takes a bit of effort.

A Varioram costs around 8K used. I would hate to find out what a rebuilt costs, especially one done by a reputable Porsche engine builder.

Kevin @ UMW will sell a GT1 long block for a turbo, checked by him after being sent in by Porsche, at $24k as of last spring. The M64s are about half that. These are hard numbers that I have in writing, by the by.

[kishg]

See, that's again where we agree to disagree: if 'the statistical evidence is against me' then shit does not happen. If shit happens - and this is precisely what I have come to accept, that the community agreement is that these engines are fine because 'shit happens' but in reality, if you expect the engine to last you're in for a bad awakening - and it happens during HPDEs on stock cars, then from my perspective these engines are 'inherently unreliable though you may get away with it perhaps even most of the time'. 'Reliable' means that 'shit does NOT happen' as long as you're within operating parameters and 'shit happens occasionally' if you skirt the edge. That is the industrial definition and, BTW, the definition that would cause mechanics' skulls to remain intact: a lot of friends with a lot less patience than I would have simply FED the engine to their mechanic: you do not blow up an engine on your first exit to the track or if you do, whoever put the car toghether foots the bill and thanks God that you don't ask him compensation for the time you wasted.
It's sort of like saying that computing fruit is reliable :) :) :).

sorry, but your not making much sense. there is no 100% reliability in anything.. see how much of a money pit a porsche will become when you start tracking it 20-30 days a year as I do. statistically, flying is much safer than driving, yet planes still crash. go figure..

[FredK]

Please do - more than anything else, I am looking for ANSWERS, at the present stage.
The other rod bearings are being taken apart. The concern is that the crap that went through the bearings subsequent to failure of #1 may have affected them *after* the fact, rendering a post-mortem more difficult.

Powder/shavings getting into bearings will have a different wear pattern than toasted rod bearings that have been loaded excessively and continuously over time. Was there any "knock" prior to things letting go?

And yes, I was running 93 oct - with a 91 oct MarkD chip, per his words, so detonation shouldn't have been an issue. Ambient temp was in the 50s and it was cloudy, so ambient heat was not a contributing factor.
I was not monitoring oil pressure, I don't have an oil pressure gauge in the car. From previous experience and given the speeds I was going, I assumed that this was VERY far removed from being a factor.

The tune was safe assuming fuel and air were predictably and reliably being placed inside the combustion chamber. Many things can frustrate this from happening.

For fuel, leaning the mixture and increasing combustion chamber temps can cause problems. Bad injectors, clogged fuel filter, weak pump, fuel starvation can cause lean running.

For air, unmetered air could cause problems. Basically--vacuum leaks.

Perhaps not, but there are a lot more, I'd say, than in an engine with 150kmi on it: all the tolerances are going to be closer to factory spec, metal fatigue is going to be a lot lower and wear on bearings/seals is going to be lower.
As to 'maintenance history', the best situation is factory recommended maintenance done on time *every time*, in the hopes that that is indicative of a particularly attentive mindset. Other oil changes of which the US market is enamored may be good but generally offer little benefit *with proper driving and decent, modern oils*. One of these days, by the by, someone will explain what people are thinking when they put mineral oils in modern engines in 2009, but that's another question for another time.
The problem is, the *driving history* is what matters more. If an engine has been broken in properly, subsequent to that never saw >1/2 throttle and 40% max RPM before water AND oil temps were above 80C, has always had good oil in it and has seen very little commuting/city driving, then its condition at 100kmi will be materially better than an engine that has had its oil changed RELIGIOUSLY with mineral oil every 1500mi... and has been taken to redline every cold winter morning that it's been driven, sitting in commuting traffic for hours at a time as a bonus.
The problem is, when you buy a used engine... you just don't know what happened to it. So either the engine is indeed 'inherently reliable' and it shrugs off the HPDE environment even in less than ideal wear conditions, or it is not and you must accept that... the engine won't be reliable.

You can bypass this requriement by opening up the bottom end of an engine and refreshing the bearings.

I have come to understand that my expectations are inherently different from the majority: I won't accept this kind of behaviour from my street car, much less so from a track car that cost about 10 times what E30s seem to go for these days. If my mechanic tells me 'you're going to be fine, see you in 5000mi' I *expect* to have no problems for those 5000mi. Not 'shit happens'. Not 'whoops'. None. And for that reason, I will pay the $90/15minutes that seem to be en vogue, grousing all the way - but you better believe that the moment that 'whoops' happens, I will be back and it won't be pretty. At that pricepoint, there is no room for 'whoops': I can hire engineers with 10 years of experience for less money and they will fully understand that 'whoops' means 'very fired and do NOT use me as a reference if you care for your career'.

Mechanics are like doctors. You have good blood pressure. You have low cholesterol, and no family history of heart problems. It's not outside of the realm of possibility that you could suffer a heart attack. So, when they slap you on the ass and tell you OK, good to go, you're as strong as a bull, they're not saying, you will DEFINITELY not suffer a heart attack over the next year. They're saying, statistics show you aren't predisposed to getting a heart attack.

Now, fine, a mechanic is not nearly a noble profession, or skilled profession, as a doctor. However, they look at the limited amount of data that is presented to them, as well as patient feedback, to determine relative risk levels of allowing you to exit the clinic's front doors.

When you pay the equivalent of $360 an hour to a mechanic ($90/15 mins), you aren't paying him to mic the bearings, check wear particles with an oil analysis, use dataloggers to find exhaust gas temperatures on individual cylinders, etc. That is like expecting to try to find out if you have a brain tumor by a doctor tapping your knee to test your reflexes. You're paying him to look for leaks, change the oil, or whatever other job you took it to him to get done.

[BrewCity11]

Could possibly be the longest consecutive posts in r3v history.
My head hurts. I can haz cheezburger?

[kishg]

Could possibly be the longest consecutive posts in r3v history.
My head hurts. I can haz cheezburger?

nah.. the bicycle thread has this one beat.. you have pizza, no need to cheezburger ;)

[nmlss2006]

sorry, but your not making much sense. there is no 100% reliability in anything.. see how much of a money pit a porsche will become when you start tracking it 20-30 days a year as I do. statistically, flying is much safer than driving, yet planes still crash. go figure..

No, there isn't. There is a lot more reliability than what I've been experiencing though. And there also happens to be a concept wherein when something doesn't conform to the reliability that everyone is so busy upholding, the person who did the work and took 6 times as long as he should have and charged 10 times as much should say 'hey, I fucked up, let me fix it'. Where I grew up, at least, that's the way it works. I am coming to understand that is very, very much not the case here.

[nmlss2006]

Powder/shavings getting into bearings will have a different wear pattern than toasted rod bearings that have been loaded excessively and continuously over time. Was there any "knock" prior to things letting go?

There was knock from the lifters the day before, but that's a different story and expected.
There was a small amount of noise again that morning, but it went away quickly and I (stupidly, but hindsight is 20/20) thought 'lifters again, I'll go talk to someone after the first run group' because see, I had this notion that the engine *should not have failed*. Finally, there was LOUD KNOCKING right before the failure, for about.. oh I want to say 15s?
And I agree with you on the bearings - we'll see when they FINALLY finish taking the thing apart. Sometime in 2017. The car was supposed to be ready wednesday before last. Still waiting.

The tune was safe assuming fuel and air were predictably and reliably being placed inside the combustion chamber. Many things can frustrate this from happening. (...)

Yes, I am familiar. And none of those 'many things' seemed to have been a factor because the parts you mentioned and many more were new - and then checked before coming to the track, I drove the car some 700 miles on the road and took it back to have things checked TWICE just to make sure we didn't have teething problems.
Besides which - overheating due to lean conditions has a very recognizable effect on the pistons. Which doesn't seem to be present here, at least that was already figured out.

You can bypass this requriement by opening up the bottom end of an engine and refreshing the bearings.

...and seals, guides and rings. And you still haven't dealt with metal fatigue unless you magnaflux everything. So you just rebuilt the engine.. which is fine, but it brings us right back to what I was saying: the engine is only reliable if it has been rebuilt recently or new.

Mechanics are like doctors. (...)
When you pay the equivalent of $360 an hour to a mechanic ($90/15 mins), you aren't paying him to mic the bearings, check wear particles with an oil analysis, use dataloggers to find exhaust gas temperatures on individual cylinders, etc. That is like expecting to try to find out if you have a brain tumor by a doctor tapping your knee to test your reflexes. You're paying him to look for leaks, change the oil, or whatever other job you took it to him to get done.

There seems to be a fundamental misunderstanding here: this was not the typical $500 r3v 'count the bolts and toss in an engine and I'll go to the track' job. Everything short of micing the bearings should've been done, God knows the time and money were there. And as a matter of fact, I could easily maintain that there was even the bugdet and time to sample at least a couple of rod and main bearings.
If the doctor charges me for TWO MRIs but then taps my knee because he says that there is no additional diagnostic advantage and then it turns out I have cancer, well, it's going to be his responsability, isn't it?
It's about expectations: if I had gone to bob the mechanic and told him 'listen, I found this MG A block in a ditch, here are $500, stick it into my car and I'm headed to PIR' it would've been one thing.
If I carefully select a 'known good' specimen of a historically very reliable engine and have it put in a shell with no expense spared, then I expect that nothing will go wrong, especially when I operate it so far within original operating conditions it's not even funny. If for no other reason than because that's how statistics work, unless they changed the shape of Gauss curves while I wasn't looking.
And if, all this nonwithstanding, a failure does occur, I expect an exact postmortem given the work that was done to eliminate variables at installation and given the known history of the engine.
If this were not the case, to borrow Kish's perhaps somewhat misguided parallel, we'd have an aircraft engine failing terminally midflight every hour with no explanation AND excellent statistics on aircraft engine reliability.
Most of my frustration comes from the fact that people (not only on the forum, mind) are desperately saying 'these things happen' and 'these engines are reliable'. But those are mutually exclusive statements, especially, again, given that the case in object was so far from being on the edge or risky it's not even funny. Which leaves us with either one of those two statements being false, or the one in a trillion event that 'shouldn't have happened' and did. But it will be a long hard road to persuade me of that, precisely because... it shouldn't have happened. Everyone is saying that.

[ZM Blue Devil]

Love of God, move this discussion to the Tech side. This is a get together thread.

[nmlss2006]

Yes, we probably should, I apologise. The thing is, as a GTG thread it had kind of died.. I'm done, anyways.