I started to post this as a reply in the "Causes of M97.21 Engine Failure?" thread, but that thread is pretty long in the tooth, and is (was?) really focused on discussing the causes of engine failures. There is information about various modifications sprinkled through it, but I thought it would be useful to assemble people's experience with different solutions in one place. So...
For new readers on this topic: the issue is lubrication related engine failures in Cayman/Boxter gen 1 engines, brought on by high lateral g conditions. This usually (but not always) means a modified car (suspension, sticky tires) on a track, and is particularly problematic with sustained high g left hand turns. The "Long Version" is available at the thread referenced above. Causes of M97.21 Engine Failure?
I'm getting ready to move to R compound tires, and I think it would be naive to do so without trying to make the engine more reliable under high g conditions. The question is, what modification to make???
I'm taking a pragmatic approach to the problem: I don't need to solve the design issue (though it might be an interesting exercise). Rather my goal is to solve my immediate problem: this engine model seems to have an history of failure under conditions to which I may very well subject the car. So my goal it to find the best way to reduce the risk of engine failure, and I really only care about effectiveness and cost. I think most members are probably of the same mindset.
Of course, understanding the underlying problem would help significantly in assessing modifications. This is my frustration with Porsche, since they are by far in the best position to provide meaningful data about the failure mechanism. I don't mind spending some money to upgrade the reliability of the car for track use, but some guidance on effective modifications would be nice. Alas, Porsche has made their position clear and we get no help.
So it seems there are a few known options out there people have tried:
As far as I can tell, this is a remote accumulator that will theoretically help maintain pressure in the oil pump discharge line. If properly configured, it should probably be able to cover for brief periods when the oil pump is not maintaining pressure (perhaps due to no oil at the pickup, or highly aerated oil at the pickup). I would need to work through a diagram of the oil system and how the accusump connects to be sure.
Probably $1500-$2000 depending on the installation
Straightforward install but need to find a location for the canister
Auxiliary scavenging pump that appears to draw oil from the front of the right side cylinder head and return it to the sump. Theory seems to be that oil accumulates here, particularly during left hand turns, and the stock scavenge pickup is not effective in removing it (because it's too far to the rear? doesn't have sufficient capacity?). Again, it would help to better understand the oiling system to assess this approach. Some people have reported great success with this system, including people who race the cars in Europe. Also I believe Porsche has offered a similar modification kit for some engines in the past.
Very Expensive: probably $3000+ to install
Porsche will immediately void your warranty if this is installed (which is a risk anyway if the car is tracked)
Deep Sump / Windage Tray
I'll put these together since the setup I think looks most likely to be effective is the Mantis extended sump / windage tray combo. Seems like it should help in two flow directions since the tray is mostly closed: (1) oil below the tray (and at the oil pickup) should be less aerated since the heavier, better flowing liquid will drain down through the tray and baffle trap doors more readily than froth, and (2) the limited open area should act to dampen the lateral shifting of oil in transient high g turns. [In a sustained skid pad session eventually the oil would reach an equilibrium angle in the sump and it would not help, but in typical road course transients it seems it should slow the migration of oil side to side. I do wonder whether any science or testing went into the area and location of the holes in the windage tray, or if they just guessed...]
This probably only delays the onset of the problem, but may be sufficient for typical road course transient conditions. Seems it should help at any rate. Wish I had a good way to test it....
Simple install; probably a 2 hour job
Relatively low cost (<$1000)
Has anyone tried other modifications with any success?
What about people's experience (success/failure) with these?
Last edited by wgn; 05-08-2010 at 11:45 AM.
Reason: Added link
I am not sure we need yet another thread on this, but..
One important modification is the retrofit of an older 996 oil separator / defoamer / swirl pot (many names) in place of the 997 part installed. This modification should help remove the air that is trapped in the oil (foam) as the oil returns from the heads. Note: this is not the same as the AOS fix descibed elsewhere.
The part in the orginal PET catalog is a 996 part (996 107 080 54) and looks just like the part in the drawing. The as built though is a 997 part (on the left) that looks much different. The more recent PET shows the 997 part.
If these pots are installed in conjuntion with the Mantis windage tray (5 more minutes to the job) then the tray must be modfied to allow the pots to pass through. Not a big deal.
Here is the "as built" sump (courtesy K-Man S):
And here is an photo of how the car was specficed in the original PET (Boxster engine shown courtesy Pelican Parts):
Also, if you actually want to fix the problem, and not just band aid in a solution then you need to modify the crank and rods:
My car has raced for over a year with the above modifications, but without an accusump or TPP kit without isssue. YMMV.
Does this means that you install (2) each of the 996 P/N (996 107 080 54), on at each side of the oil pick up suction tube?
Yes. Easy job with the sump plate off. A bit more involved if you are installing a deep sump and windage tray as the tray needs to be modified to allow the swirl pot to pass through (need to cut a hole).
This is definitely a couch tuning question as I do not even come close to understanding all these systems. But all engine failures seem to have occurred WITHOUT the oil pressure light coming on. Therefore, will the Accusump help at all? Will it discharge oil since there is no detectable oil pressure drop? Does the Accusump "measure" oil pressure differently than the oil pressure sensor?
This is definitely a couch tuning question as I do not even come close to understanding all these systems. But all engine failures seem to have occurred WITHOUT the oil pressure light coming on. Therefore, will the Accusump help at all? Will it discharge oil since there is no detectable oil pressure drop?
Does the Accusump "measure" oil pressure differently than the oil pressure sensor?
I believe that the Accusumo does not have any meassuring device, it works by pressure differential, (is just a hydraulic accumulator)
For example if your engine normal operating oil pressure is @ 45 psi, this pressure fills the Accusump with oil (against an internal rubber bladder) @ 45 psi, and whenever the oil pressure drops below 45 psi the Accusump will push back the oil into the engine immediately by the pressurized rubber bladder.
The theory is really simple, the trick is to ensure that you have a proper installation (leak free) and to determine how do you want to trigger the signal to open the electric valve solenoid on the accusump, unless you want the accusump next to your seat (like a race car) with a manual operated valve...