Engine Bearing Failure- Disassemble Heads Too?

[KTL]

Agreed the full engine swap is theoretically easier. But that means buying a fully rebuilt engine. Because any used engine I buy, i'd be hard pressed to not open it up to make sure all is fairly good before racing it. I've seen too many people get "good" used engines and shortly thereafter the swap, uh oh needs a rebuild for some reason.

Yep yot my first engine work rodeo but admittedly is my first engine blow rodeo. Definitely got bucked off the bull and stomped good before the clown could distract it!

[lindy 911]

Kevin,

How many hours were on the engine when this happened? Was the engine cold? You seem like the kind of guy that would warm an engine up before he ran it hard; true? I can't see how the filter caused this. Are you sure your piston scoring didn't plug the filter and then starve the crank? Just wondering, I'd hate to see this happen again and not have the filter be the cause.

[KTL]

Engine had very few hours. 3-4 hrs tops. Ran it in July for a race weekend and then it crapped out first time out on the track mid Sept after 4 laps. Ran it in the garage & around my neighborhood, around track property before taking it on track in July.

July day 1 maiden voyage I fiddled around and only got out for the race- 40 min at reduced load. All seemed good. Monitored temps, oil level and catch can. No blow by or any driveability problems. Day 2 was practice (15 min) qualify (15 min) and 40 min. race. HOT weather for the afternoon race. REAL hot. Not a cloud in the sky & high 90's + humidity. Engine ran great and no signs of distress. I actually "won" my class. It was great. Only negative was a considerable leak from cam seal that had to be fixed but wasn't black-flaggable. Temps were high at 230-240. Attributed temps to weather, new engine, removal of engine oil cooler and needing better flow behind center front cooler.

Found distorted engine filter after July weekend. Didn't like what I saw. Appeared to me that filter was a restriction since it's a rather small filter element w/no bypass. So I reinstalled the oil cooler since I was into the engine anyway to fix cam seal leak. So that left me with the typical oil tank filter as my only filter.

Fast fwd to Sept. Went out for qualify, cool temps but car was warmed up a bit at my paddock spot and also sat idling on grid for several minutes. Temp was up to ~170F (according to my nice new Autometer oil temp gauge installation) on grid departure. First lap was easy going. Came up to 190F and dropped the hammer. Power was fine for 3 laps and then I felt some grunt loss. Then some hesitation. Power continued downward and I could see oil pressure wildly fluctuating with throttle input on the mech oil pressure gauge. Should have shut it down on track and got towed in. That stupidty likely caused a lot more damage. But since I had oil pressure I felt I could limp it back to my paddock spot

Upon disassembly the oil pressure relief & bypass pistons came out cleanly w/no resistance & they're the right springs & pistons. Cyl piston scoring is hard to analyze. There's clearly bearing material embedded. Piston-cyl clearance was checked during the build. 0.0025"+ was meas'd for each set of P&C. No rings came out damaged and the ring lands + piston perimeter looks good around ring lands.

Maybe should have changed oil right after initial startup & neighborhood drive (20min. initial break in for cam security) instead of waiting until after race weekend. But I would think that the harm done by "dirty" break in oil would have killed the engine during operation under crazy hot conditions, not at the following event? When I changed oil before doom event, oil had its share of stuff in it but nothing alarming. Drain plugs looked as expected when draining 1st dose of oil out of a newly built engine.

Reason I think the filter did it is because of the distortion I initially found first time. If crank was not getting enough oil due to restriction (crank is next in line after t-stat (removed) & oil filter housing) then its doomed. There's no bypass in the filter to ensure the crank is still getting ample oil, be it clean or not clean.

My theory is that since my oil pressure is detected from the oil warning light port (IDIOT light is the appropriate term here....) it would still see suitable oil pressure? Because the restriction is downstream of the oil pressure takeoff. The restriction would create overpressure but the bypass and relief ports are right there by the IDIOT light port and it would seem they were doing their job, bypassing to the case sump and re-routing to the pressure side of the pump.

[lindy 911]

At the first oil change you say "oil had it's share of stuff in it"; please elaborate. My builds show zero "stuff" at first change and every change thereafter for that matter. If your oil showed tiny flakes, something was giving up. There should be no sign of metal at all. And if the drain plug had anything on the magnet the engine was hurt already.

[Lapkritis]

A filter will grey during engine break-in... that's what I experience and took from his post as well.

[BGCarrera32]

Ugghhh, man that's awful.

Given the sheer amount of stuff that's floated through there I'd be going through the heads, the oil cooler t-stat, you name it.

If I was in the zipcode I'd be over to help.

[KTL]

My experience has been that new builds can show a little bit very fine sparkly stuff since things are wearing in- cams & rockers, rockers & rocker shafts, chains & sprockets for instance. I always find a miniscule bit of fuzz/sludge on my drain magnets, even on the street car, during oil changes.

But I certainly don't disagree that something was giving up. Because my entire engine failure theory revolves around the filter choke making the engine slowly give up! I pulled apart the small filter and it's got some significant darkness to it, as well as some particulates that would have concerned me.

I did some tracing of the oil circuit paths on the engine case itself last night to confirm where/how the oil goes from the pump to the crank. Here's how I see it:

1. Oil enters pump from S-hose & hard pipe beneath engine-mounted oil cooler. Even with my oil filter housing installed, this is of course still the case.

2. Oil is sent thru the pump and discharged out the opposing port on the pump, slightly offset in the pump housing, and is sent thru the adjoining passage in the case. This passage contains the "J" pipe that turns the oil 90 deg vertical and up the pipe to the bottom of the t-stat cavity. This is the pipe that commonly needs to be patched with JB Weld where the aluminum casting around the pipe is rather thin on the exterior of the case and in the oil cooler recess.

3. Oil goes up the pipe into the bottom of the t-stat cavity and then is sent to the hot or cold window in the t-stat, obviously depending on temperature. Or even both windows at an intermediate temperature. "Or" is the important word here. More on that in a minute.

Note that the passage adjoining the pressure side of the pump for the J-pipe is also intersected by the horizontal safety pressure relief piston which dumps overpressure into the sump. This oil is then picked up by the pump sucker screen & sent back to the oil tank.

4. Beyond the t-stat cavity, be it the cold or hot side, is a galley that feeds oil to the vertical bypass valve which reintroduces oil to the inlet side of the oil pump. Also in that galley is the port on top of the case that triggers the low oil pressure IDIOT light.

5. Since I had this t-stat blockoff in place, the oil is ALWAYS directed to the hot side of the t-stat cavity. There is no "OR" to the cold side in with this piece in place. With a t-stat, as it gets warmer, some oil is going to the hot side and some cold. W/out the t-stat, there is no way for the cold side of the circuit to be fed.

So when the hot side of the t-stat cavity is permanently engaged, CRANKSHAFT oil is provided only from the oil cooler/filter path. I can't emphasize this enough. The supply to the crank comes from nowhere else. Backed-up oil that can't get pass the hot side is either short-circuited at the safety valve (and dumped to the case sump), or is sent down to the vertical bypass valve which reintroduces oil to the inlet side of the oil pump. In either path, oil is not reaching the crankshaft.


Use of the 993 diagram helps better illustrate this, since my engine was basically set up like the 993 in terms of engine case oil plumbing. The old 3-D engine diagram also helps illustrate this.






Yep the entire oil system needs to be cleaned as well as possible. Appropriate case plugs will be removed, piston squirters replaced, all oil lines cleaned with pigs or replaced as needed (almost all oil lines on this car are braided flex line), oil coolers junked, t-stats disassembled & cleaned with new regulators installed, oil tank will be cleaned and shelved. Now is as good a time as ever to install the Peterson oil tank that's sitting in my basement!

[BGCarrera32]

Sorry if you said already, but what are your plans with the pistons? With all the skirt scuffing I would be really leery of reusing those...

Also, sending a PM.

[KTL]

Yeah my posts can be tiresome to read for sure!

Pistons with a little scuffing can be cleaned up IMO. But obviously the scuffing needs to be investigated as to what caused it. In my case, they're not only scuffed, but they have bearing material embedded in them. Hard to get that out w/out destroying the piston precision.

Not going to risk it so these are conversation pieces/desk paperweights.

[lindy 911]

Kevin,

When you installed the console, did you also shim the pressure relief as required? William Knight said that if you replace the cooler with the filter console you will loose 20 PSI oil pressure. Without the shim in the pressure relief on the bottom of the case it will make very little pressure.

[KTL]

Lindy,

Thanks for the pointer on the overpressure spring adjustment. I did not install that 964.107.532.00 shim/spacer ring. Honestly I look at the parts diagrams quite frequently and have never noticed that before. Hmmmm........... Had I seen it, i'm sure I would have asked about it's function. But i'm surprised i've not seen other people mention the oil filter console installation needing that spacer.

Odd thing I found in the parts diagrams is that all 3.6 engines have the shim spacer. Even those w/out the oil filter console (964 3.6). I suppose the thought is that if the shim is in place, even w/out the console (and no oil cooler, since the 964 3.6 has no engine mounted cooler) it's not a problem. Only when the spacer is not present & an oil filter is installed does the lack of shimming present a problem?

I believe what William says about the 20 psi. I recall my mech oil pressure gauge reading around 60 psi and noticed that was down ~20 psi from where I saw it pre-rebuild. I ignorantly dismissed it as no-big-deal because I assumed the lower pressure was attributed to the lower viscosity 30W break in oil.

But I don't think it will make very little pressure is necessarily the case? I always had oil pressure but likely not enough. Whoops, there's another contributor to my failure. I suspect by the time I reinstalled the cooler (next race) it was too late and the damage was already done.

[lindy 911]

This actually explains a lot; piston squirters would not operate correctly resulting in scuffed skirts. Low pressure at the crank and rods mean failed bearings. The resulting crud in the oil then packed up the filter. I think we're on to something.

[Walt Fricke]

Well, something not well understood (certainly by me) is at work with oil pressure. My 2,7 with a turbo pump and a not very remote oil filter in place of the engine mounted oil cooler never had an issue. It had the oil pressure update (sand cast case), but I don't recall just which screw cap I had. There is a difference in the depth of the internal drilling depth depending on the cap you have.

Then I bought a 2.8, and ran it with an external oil filter also instead of a cooler. Don't recall the cap (I should look). No issues - again a sand cast case with the modification.

Then I built a SS2.8 with a GT3 pump using the early Carrera/Turbo case with the 6 bolt crank journals. Oil pressure issues out of the box - pressure at startup and idle nice and high, rose some, then quit well below 60 psi. I investigated these changes Porsche made in these parts, and posted it here (if I knew how to scarf up addresses I'd post that, as I included pictures and facts about different springs, depth dimensions, etc).

Added the spacer. Put car back on dyno. Pressures better - higher all around, but still not climbing as I'd expect given.

Then had oil into exhaust from #5 and 6. Plus a bit of aluminum. Still haven't gotten around to pulling engine and taking off those heads. Looked in with borescope but nothing I could identify going on, though something was for sure.

Wondering if somehow the initial low pressure (which I deemed good enough for a few chassis dyno runs - might have been big mistake) wiped those bearings, and when back on dyno the wiped bearings (oil temps never rose) caused the already tight piston/head and valve/piston clearances to close and something hit. We'll see.

But this is the first I have heard about changing out the engine oil cooler for a filter, remote or console mounted, or just directing the oil directly to the distribution galleries, could cause a 20 psi pressure drop.

The diagrams put both pressure regulators - the pressure setting system (where the late cars added a shim) and the emergency relief piston - between the pump and the thermostat. Which should mean that they open based on whatever pressure restrictions are downstream. If oil is diverted directly to the distribution system, that system cannot show a lower pressure than what the regulator piston/spring sees, can it?

After all, that's what happens with the thermostat in place - oil bypasses the cooler.

Now put a filter in place of the cooler, and disable thermostat system (lock it open, machine it, or replace it with the 935 part or replica. First assume the filter is less of a restriction than the cooler. How could that lower pressure downstream?

Now assume filter is more restrictive. Which Kevin thinks may have been the root cause of his disaster. Obviously the filter cartridge suffered from overpressure, though I am unsure whether this was at stage 1 or stage 2. These can be distorted by too much pressure for them to handle, or by a lesser pressure if too much debris gets into them, raising their resistance.

If I understand the fluid dynamics right (and I may well not), the filter, like the oil cooler, acts like a resistor in series in an electrical circuit. Which would mean the filter is going to lower the pressure (voltage) seen immediately downstream from it. Once the pressurized oil enters the distribution system, nothing (or nothing much) is in series any more - rather the various oiling points act in parallel, and that isn't going to drop the pressures they see.

Then there is where you measure oil pressure. If it is before the filter in this system, you'd see peak pump pressure. If after (especially at the later stock position by the fan housing), you'd be reading the pressure the bearings and squirters and so on see.

So I can see how a filter can affect things. But I don't recall that the spacer for the main spring entered the equation until the GT3s. Might not remember that correctly, though. But if I do, that would mean the stock 3.6 filter must flow enough, doesn't it? Or were, perhaps, the cam tower line restrictors introduced on the 964 motors to boost the main pressure as much as to reduce the amount of (excess) oil sprayed onto the valves and lifters and cams?

[WinRice]

Originally Posted by Walt Fricke

Then there is where you measure oil pressure. If it is before the filter in this system, you'd see peak pump pressure. If after (especially at the later stock position by the fan housing), you'd be reading the pressure the bearings and squirters and so on see.

Looking at the two oil diagrams, notice how the oil pressure sender is before the cooler (if I'm reading that correctly, kind of hard to read) on the older system, but after the filter on the 993. Porsche probably knew this and moved it.

I have a filter console on the shelf, so this hits home.

Sorry about the carnage, KTL!

[Walt Fricke]

There is old, and there is older. I'm pretty sure that Porsche moved the pressure pickup location when it added an oil pressure gauge. The flywheel side pressure takeoff was used for the idiot light switch, so a pressure sender had to go elsewhere. Which was up by the fan. The first ones used a banjo fitting on the right cam oil line. Then that was switched to the big steel block.

Unless, at some point, the flywheel end location was used with a dual sender - idiot and pressure. VW used those in some applications, and they all come with an M10x1.00 thread.

In any case, the change of location was well before the advent of the 993s. (to some of us, alas, those are new, not old).

[Walt Fricke]

Looking at a variety of oil system diagrams, I conclude that all of them from the time an oil pressure guage was introduced read oil pressure after the engine oil cooler (or its replacement filter on the 993, or the filter and then water/oil heat exchanger on the GT3).

If my analogy of pressure to electricity (voltage) is correct, basically the pressure should be the same right at where the oil enters the main gallery, or at the rear (far) end, because at that point everything downstream is in parallel.

Basically, because just as there is resistance depending on the length and diameter of wires, which lowers voltage the farther away from the source you get, so also (if I have this right) is some pressure in a fluid reduced due to pipe wall friction (or whatever it is called in hydraulics). But my instincts (and visualizations of passage sizes) tell me such losses in pressure have to be very small: the main gallery is wide, and relatively short also.

Anyway, it appears an obstructed filter, in a thermstatless system like Kevins, should show up as distinctly lower pressure.

[WinRice]

Originally Posted by Walt Fricke

Looking at a variety of oil system diagrams, I conclude that all of them from the time an oil pressure guage was introduced read oil pressure after the engine oil cooler (or its replacement filter on the 993, or the filter and then water/oil heat exchanger on the GT3).

Anyway, it appears an obstructed filter, in a thermstatless system like Kevins, should show up as distinctly lower pressure.

Agreed. Went home last night and looked over better diagrams and the sender was located after the cooler/filter. So if the filter was a major restriction it should have shown a major pressure drop. The oil galleys are large enough to not cause a major restriction or pressure drop, if they are clean and clear.

[kyhunting]

Had the same thing happen to me as well. Canton filter was basically crushed on the inside. New engine and of course it died at Sebring on its first outing. Sebring is only 13 hours from home. What a wasted trip $$$. I no longer use the Canton filters.

[Walt Fricke]

Ky - My experience tells me that the "long" Canton filter flows enough oil to support the output of a Turbo oil pump, at least if it is not clogged with debris. However, it might be subject to crushing if the oil is too cold relative to the RPM, even if it is not holding a significant amount of small (small enough to get through the pump screen) debris.

And I don't see how crushing, alone, will reduce its flow capacity. It might affect filtering, by tearing the glued ends out of their steel caps.

It seems as if a rod bearing failed, producing a lot of debris. Which escaped filtration due to a crushed Canton filter (rereading, it looks like a Canton was also used as the primary filter on the oil tank console?). Changing out the engine mounted filter for a cooler would not have done anything about the debris already in the oil tank. Since the tank is beyond the filter, when the motor was run again, without the secondary filter, nothing prevented the debris from getting even deeper into the brearing system. Eventually, splash oiling deposited debris on the cylinder walls (piston scoring).

Long ago, a team at a race told me what the failure sequence looks like: oil temps rise, engine loses power, oil pressures drop, engine fails. This was in the conext of when to take a race motor out of service to replace bearings which have exceeded their expected racing life.

I think Kevin's description of the final fatal day goes pretty much like this except for the pressure drop. If an oiling passage gets clogged, you won't see a pressure drop. A friend had a crank with one rod oil passage blocked with part of an O ring. Never lost pressure, but things got so hot that after returning to pits with some obvious distress to motor and stopping engine, it could no longer be turned with the starter. Carillo rod was deep blue.

But I think that lunching a motor is apt to be caused by something other than what goes on in the filter - at least in the long one. And the crushing, if the motor has lunched, could be a result instead of a cause.

If I read the descriptions correctly, the tank filter was not checked when the engine filter was replaced with the cooler?

So I'd hold open the possibility that Kevin's problem was caused by something else which prevented some bearings from working properly.

[KTL]

I don't think the shim in the pressure relief spring is the definite answer for my problem.. Note what I mentioned in one of my previous posts

Beyond the t-stat cavity, be it the cold or hot side, is a galley that feeds oil to the vertical bypass valve which reintroduces oil to the inlet side of the oil pump. Also in that galley is the port on top of the case that triggers the low oil pressure IDIOT light. My mechanical oil pressure gauge is connected to this idiot light port

So as I understand it, the relief spring in question here is the one I describe above. When this valve is activated, the oil being bypassed is returned to the inlet side of the pump. It is parallel to the t-stat housing and is therefore independent of whether the t-stat is open or closed, just like Walt said. This bypass valve is always seeing what the oil pump is producing before the entry to the oil cooler/oil filter console. So the pressure should be as high as whatever the bypass spring allows. I had good oil pressure with the old SC pump before I rebuilt the engine and installed a 964 pump. I would get 100 psi+ at startup, 80-85 during high rpm operation- you can see it operating here in the spot where the clock usually goes




Incidentally I noticed when I first started the engine after the rebuild, I thought the oil pressure was a little bit low around 60 psi. Like I said, I attributed that to the 30W oil. However, I think I still should have seen some high oil pressure at initial startup. Then when I changed the oil (20W50 racing BP) and also had reinstalled the oil cooler and also the internal t-stat, the gauge showed big pressure like I was accustomed to at startup- zings up to 100psi and I am very cautious with the throttle during this warmup time.

I do understand that the shimming of the valve is going to allow more pressure to be made before the valve opens and bleeds off the excess back into the inlet side of the pump. But I agree that a filter vs. an oil cooler should have around the same amount of pressure drop.

Since I had the neato RSR t-stat bypass installed, the oil is ALWAYS directed to the hot side of the t-stat cavity. This similar to a 964 or a 993 since they do not have a t-stat. They simply have a cap on top of the t-stat location. The 993/964 cap is also where those engines tap into the oil circuit to get oil temperature and pressure. So 993 the diagram is a bit misleading as to the oil pressure sensor location.

So when the hot side of the t-stat cavity is permanently engaged, CRANKSHAFT oil is provided only from the oil cooler/filter path. I can't emphasize this enough. The supply to the crank comes from nowhere else. Backed-up oil that can't get past the hot side is either short-circuited at the safety valve (and dumped to the case sump), or is sent down to the vertical bypass valve which reintroduces oil to the inlet side of the oil pump. In either path, oil is not reaching the crankshaft.

So it may be that I wasn't getting enough oil to the crank on account of a restrictive filter and the addition of the 964 pump w/out the necessary shim on the pressure bypass spring. Another thing to consider is the cam oil line restrictors. I don't think it's a coincidence that these "magically" appeared the same time as the 964 pump in the 964 turbo. Note that these restrictors are not necessary on the 964 or 993 since those engines have internal cam housing oil supply.

These restrictors help overall engine oil pressure by reducing the oil sent to the cam housings. In turn, it would seem that this pressure increase would be required since the 964 pump has a bigger scavenge side? My thinking is that the combination of the cam line restrictors and the additional spring shim account for the added scavenge capacity of the pump. Meaning, you have to keep enough pressure on the supply side so that the scavenge side of the pump doesn't starve the system?

Here's another little twist to my story. I had the cam restrictors in place when I had the filter console installed. Then I chose to install the old cam fittings when I replaced the oil cooler. Reason I did this is I was thinking about other information i'd heard about how with today's oils not being what they used to be, more oil in the cam housing is not necessarily a bad thing. But if the 964 pump wants to see those restrictors, as well as the pressure bypass shim, that could have been a contributor as well.

Yes the oil system had two Cantons on it initally. One small filter on the engine filter housing and one on the oil tank. When I first changed the oil, the small filter was slightly deformed as previously pictured. The big filter was fine in terms of structure, but dark in color. So when I reinstalled the cooler, I also put a new filter cartridge in the big tank filter.

I will agree that I think my failure was not solely the small filter being a restriction. I think it was a contributor. What i'm thinking is the small filter was only slightly damaged due to cold oil flow thru a filter with not a lot of flow capacity. However the oil was straight 30W break-in oil. But there was also some sticky assembly lube in the mix as well. So that red glue could have contributed to the filter restriction as well. Agreed that the ultimate failure of the big filter was bearing debris causing it to implode. Because that filter had no signs of implosion during the first oil change.

Bottom line is i'm not giving up on the 964 pump and filter housing. I'll put the cam line restrictors in, use the 964 pressure bypass spring shim, and i'll use Mahle filters. Problem solved? We'll see.

Thanks to everyone for their contributions on this thread. I GREATLY appreciate the input to help me diagnose & understand what went wrong & why. This place rocks.