New oil control solution for Race/ORR/SC/Stroker/GTS/GT/CS/SE/S4/S3 928 Engines
05-03-2011 | 06:56 PM
#1
Thread Starter
Archive Gatekeeper
Rennlist Member
Rennlist Member
Joined: Sep 2002
Posts: 17,715
Likes: 2,884
From: Irvine, CA
New oil control solution for Race/ORR/SC/Stroker/GTS/GT/CS/SE/S4/S3 928 Engines
Back in January, I was lucky to be able to see Andy G’s ITB’ed stroker monster on the engine dyno (see https://rennlist.com/forums/928-foru...o-results.html, which revealed a lot about oiling dynamics in the 928 engine – the Cliff notes version is that even in an engine sitting still on a dyno cart, sustained high revs results in oil delivery to the heads, overwhelming the block drains’ ability to drain them. This may be because the drains are just too small, or crankcase pressure or oil foam blowing up the drains is inhibiting return, but whatever the root cause is, the heads fill with oil.
After more a minute at 6000 rpm, Andy’s engine (which has breather hoses on all four VC elbows, rising up to the sides of the ITB airbox- a much greater vertical rise than stock), all four breather hoses shot liquid oil into the airbox at the same time. The heads were thus filled to the level of the breather extensions. Think about the volume of the 32V head, double that, and you have a sense of how much oil is NOT available to the pickup at that moment…. Lots and lots of race engines and autobahn motors have gone kablooie because of oiling failures, and Porsche either wasn’t able (or more likely, willing, by 1991/2/3) to do anything about it.
Anyone who has done an intake refresh on a 32V car not driven by a dentist’s wife knows what an oily mess the inside of the intake is- all three engines’ intakes I’ve R&R’ed had oil pooled in the lower plenum and oil soaked runners. Greg says the ’85-’86 cars are even worse, if only because there’s so much ‘flat’ surface area for oil to collect on. Here’s a lousy pic (mostly blocked by the elbow fitting) of some of what drained out of the front breather hose after pulling the connector out of the MAF elbow. There was a nice puddle in the bottom of the lower plenum, too. Maybe from the 3-way hose on the filler neck(?)
Porsche clearly knew that they didn’t get the breathing right, given all the S4+ intake breather revisions over the years:
87-92: All crankcase venting via the base of the filler neck, returning either directly to the elbow MAF, or in and out the passenger valve cover, or (small amount) via the 3-way hose to the lower plenum.
Early 93: Block off the filler neck base, big port off upper filler neck to right front valve cover elbow, then out the right rear elbow to the MAF elbow.
93-95: Upper filler neck directly to MAF elbow, separate right rear elbow hose to MAF elbow, and ‘crossover’ hose, to try to vent the driver’s side head. Clearly insufficient for a 300, let alone a 400 (or 500) hp motor….
So the goals here are multiple:
1. Improve the oil pickup in the pan
2. Reduce windage by cutting down on the amount of oil that the crank sucks out of the pan.
3. Reduce crankcase pressure
4. Keep the heads from filling up with oil, uncovering the oil pickup
5. Keep oil from reaching the intake system, decreasing the octane of the fuel and increasing detonation.
And for me personally, 6. fit it all in a 928 engine compartment without compromising A/C, windshield washers, and have it look to most eyes like a stock engine compartment.
How to do address all this? If you’re Greg, you start by dry-sumping the heads.
Greg spec’d out a scavenge pump built up out of sections that provide adequate output, and fits above the A/C heat shield and ahead of the #1 header tube. The windshield washer tank in the passenger side fenderwell gets pulled to make space for a Provent (with its 1” ID hose connections, necessary to slow vapor velocities but ugh, what a PITA to work with), and the breather hoses are routed through existing filler neck and washer wiring harness grommets. The washer wiring harness was pulled up to the cowl, and a 2-wire jumper made to connect the ‘normal’ washer connector to a retrofitted intensive pump tank. Everything can go back to stock in 2-3 hours, if need be.
The pump is only one small part of the ‘kit’, in conjunction with many other pieces, and there is a ton of modding needed to get everything to fit and work in harmony. The oil pan, intake, and valve covers have to come off for various fittings and additional breather hoses to be routed. This solution is definitely one to consider during an engine-out rebuild or major refresh. And getting the plumbing functionally right and aesthetically pleasing (to me) took the better part of a couple of days, much of which was Greg cutting, drilling and milling chunks of aluminum into spacer, adapter and fitting ****. The next kit will be easier. J
So how’s it work? The belt-driven scavenge pump pulls vacuum on pickups in the heads, in an rpm dependent fashion. At idle, the pump is doing essentially nothing, but as you can see in the video below, different pickups begin sucking oil as low as 2500 rpm. By the time your engine is turning 4000 rpm, all corners of both heads are sucking a lot of oil-air mixture and returning it directly to the pan, right above the screen (we retrofitted an early pan screen and pickup setup, modded to work with the pan spacer).
Oil being scavenged from driver’s side pickup at 3600 rpm:
Video of same. (Sorry, the flashlight obscures the flow until 0:10 or so) :
As far as the breathing goes, there are now two separate crankcase breather hoses that vent to each head, from which vapor is sent to the Provent in the fender. Separated oil is returned via the scavenge pump, and the air returns go back to the passenger side elbow MAF. After an hour of rather aggressive driving, the lumens of both air returns are bone dry- no oil residue at all. So no more oil ingestion! (Honestly, the Provent may be overkill (especially for the street), but it is there to make sure that no way can any oil get into the intake system. On Andy’s engine, even with the scavenge system working (but no oil-air separator, yet) there is a bit of oil mist coming out of the breathers at sustained runs above 6500 rpm.)
There are many benefits to this setup- Obviously, you decrease oil consumption, and you keep the intake clean and dry. Dry intake means no oil mist in the intake charge, avoiding dilution of the air-fuel charge, allowing more aggressive timing before running into detonation. I’d guess it’s probably good for emissions as well. You also keep the heads from filling at higher rpm, which not only keeps oil in the pan, it keeps your exhaust cams (spinning at 3000 rpm or so) from being submerged and further whipping up the oil.
Of course, Greg’s engines also incorporate windage trays, pan spacers, and all the other oiling tricks to keep the pickup submerged. But the real innovation was the observation on Andy G’s ubermotor that the heads fill with high rpm - Of course dry sumping the pan keeps the engine together, ‘cause there’s enough reserve to fill the pan regardless of how full the heads are, and I’d guess a big-enough Accusump is effective too, but none of that addresses oil ingestion through the intake, nor the root problem of insufficient oil return to the pan at high rpm. Greg’s solution takes care of both, quite elegantly, I think. And it doesn’t look like much, under the hood, unless you know what you’re looking for.
Pics:
Disassemble a perfectly good 2500 mile engine:
Stroker oil pan off showing windage tray and headers:
New oil pan fittings:
Valve cover scavenge fittings, driver’s side:
Underside of scavenge pump showing relation to #1 header tube:
The accessory belt routing is a little tricky
:
Oil pump return line into the pan:
I have put about 2500 miles on my stroker since last July- in that time it has used about 2.5 quarts of oil- some of that is probably from the break-in period, but I do seem to spend lots of time above 5000 rpm . If you shut it off shortly after a hard run, wait a couple of minutes and restart, you get a big chuff of oil smoke out the back. This does not happen after ‘normal’ driving, so there’s clearly rpm-dependent oil ejection into the intake . And it doesn’t happen now no matter how I hoon around in the car, so I believe that it was due to oil ejection into the manifold, and not leakage past valve guides, ring blowby, etc…Below is is what the inside of the Provent return line looks like after an hour on and off the freeway at uh, supra-legal speeds. No more oil in this intake, thankyouverymuch…
The final product, installed. (We put one of Greg's new fuel supply line in too, WWWAIT. Better insulated from valve cover/head/coolant hose heat, easier pass side VC removal, etc. )
FWIW, I will have the clear hose segment at SITP on Saturday and can demonstrate the rpm-dependent oil scevenging to anyone who is interested.
After more a minute at 6000 rpm, Andy’s engine (which has breather hoses on all four VC elbows, rising up to the sides of the ITB airbox- a much greater vertical rise than stock), all four breather hoses shot liquid oil into the airbox at the same time. The heads were thus filled to the level of the breather extensions. Think about the volume of the 32V head, double that, and you have a sense of how much oil is NOT available to the pickup at that moment…. Lots and lots of race engines and autobahn motors have gone kablooie because of oiling failures, and Porsche either wasn’t able (or more likely, willing, by 1991/2/3) to do anything about it.
Anyone who has done an intake refresh on a 32V car not driven by a dentist’s wife knows what an oily mess the inside of the intake is- all three engines’ intakes I’ve R&R’ed had oil pooled in the lower plenum and oil soaked runners. Greg says the ’85-’86 cars are even worse, if only because there’s so much ‘flat’ surface area for oil to collect on. Here’s a lousy pic (mostly blocked by the elbow fitting) of some of what drained out of the front breather hose after pulling the connector out of the MAF elbow. There was a nice puddle in the bottom of the lower plenum, too. Maybe from the 3-way hose on the filler neck(?)
Porsche clearly knew that they didn’t get the breathing right, given all the S4+ intake breather revisions over the years:
87-92: All crankcase venting via the base of the filler neck, returning either directly to the elbow MAF, or in and out the passenger valve cover, or (small amount) via the 3-way hose to the lower plenum.
Early 93: Block off the filler neck base, big port off upper filler neck to right front valve cover elbow, then out the right rear elbow to the MAF elbow.
93-95: Upper filler neck directly to MAF elbow, separate right rear elbow hose to MAF elbow, and ‘crossover’ hose, to try to vent the driver’s side head. Clearly insufficient for a 300, let alone a 400 (or 500) hp motor….
So the goals here are multiple:
1. Improve the oil pickup in the pan
2. Reduce windage by cutting down on the amount of oil that the crank sucks out of the pan.
3. Reduce crankcase pressure
4. Keep the heads from filling up with oil, uncovering the oil pickup
5. Keep oil from reaching the intake system, decreasing the octane of the fuel and increasing detonation.
And for me personally, 6. fit it all in a 928 engine compartment without compromising A/C, windshield washers, and have it look to most eyes like a stock engine compartment.
How to do address all this? If you’re Greg, you start by dry-sumping the heads.
Greg spec’d out a scavenge pump built up out of sections that provide adequate output, and fits above the A/C heat shield and ahead of the #1 header tube. The windshield washer tank in the passenger side fenderwell gets pulled to make space for a Provent (with its 1” ID hose connections, necessary to slow vapor velocities but ugh, what a PITA to work with), and the breather hoses are routed through existing filler neck and washer wiring harness grommets. The washer wiring harness was pulled up to the cowl, and a 2-wire jumper made to connect the ‘normal’ washer connector to a retrofitted intensive pump tank. Everything can go back to stock in 2-3 hours, if need be.
The pump is only one small part of the ‘kit’, in conjunction with many other pieces, and there is a ton of modding needed to get everything to fit and work in harmony. The oil pan, intake, and valve covers have to come off for various fittings and additional breather hoses to be routed. This solution is definitely one to consider during an engine-out rebuild or major refresh. And getting the plumbing functionally right and aesthetically pleasing (to me) took the better part of a couple of days, much of which was Greg cutting, drilling and milling chunks of aluminum into spacer, adapter and fitting ****. The next kit will be easier. J
So how’s it work? The belt-driven scavenge pump pulls vacuum on pickups in the heads, in an rpm dependent fashion. At idle, the pump is doing essentially nothing, but as you can see in the video below, different pickups begin sucking oil as low as 2500 rpm. By the time your engine is turning 4000 rpm, all corners of both heads are sucking a lot of oil-air mixture and returning it directly to the pan, right above the screen (we retrofitted an early pan screen and pickup setup, modded to work with the pan spacer).
Oil being scavenged from driver’s side pickup at 3600 rpm:
Video of same. (Sorry, the flashlight obscures the flow until 0:10 or so) :
As far as the breathing goes, there are now two separate crankcase breather hoses that vent to each head, from which vapor is sent to the Provent in the fender. Separated oil is returned via the scavenge pump, and the air returns go back to the passenger side elbow MAF. After an hour of rather aggressive driving, the lumens of both air returns are bone dry- no oil residue at all. So no more oil ingestion! (Honestly, the Provent may be overkill (especially for the street), but it is there to make sure that no way can any oil get into the intake system. On Andy’s engine, even with the scavenge system working (but no oil-air separator, yet) there is a bit of oil mist coming out of the breathers at sustained runs above 6500 rpm.)
There are many benefits to this setup- Obviously, you decrease oil consumption, and you keep the intake clean and dry. Dry intake means no oil mist in the intake charge, avoiding dilution of the air-fuel charge, allowing more aggressive timing before running into detonation. I’d guess it’s probably good for emissions as well. You also keep the heads from filling at higher rpm, which not only keeps oil in the pan, it keeps your exhaust cams (spinning at 3000 rpm or so) from being submerged and further whipping up the oil.
Of course, Greg’s engines also incorporate windage trays, pan spacers, and all the other oiling tricks to keep the pickup submerged. But the real innovation was the observation on Andy G’s ubermotor that the heads fill with high rpm - Of course dry sumping the pan keeps the engine together, ‘cause there’s enough reserve to fill the pan regardless of how full the heads are, and I’d guess a big-enough Accusump is effective too, but none of that addresses oil ingestion through the intake, nor the root problem of insufficient oil return to the pan at high rpm. Greg’s solution takes care of both, quite elegantly, I think. And it doesn’t look like much, under the hood, unless you know what you’re looking for.
Pics:
Disassemble a perfectly good 2500 mile engine:
Stroker oil pan off showing windage tray and headers:
New oil pan fittings:
Valve cover scavenge fittings, driver’s side:
Underside of scavenge pump showing relation to #1 header tube:
The accessory belt routing is a little tricky
: Oil pump return line into the pan:
I have put about 2500 miles on my stroker since last July- in that time it has used about 2.5 quarts of oil- some of that is probably from the break-in period, but I do seem to spend lots of time above 5000 rpm
. If you shut it off shortly after a hard run, wait a couple of minutes and restart, you get a big chuff of oil smoke out the back. This does not happen after ‘normal’ driving, so there’s clearly rpm-dependent oil ejection into the intake . And it doesn’t happen now no matter how I hoon around in the car, so I believe that it was due to oil ejection into the manifold, and not leakage past valve guides, ring blowby, etc…Below is is what the inside of the Provent return line looks like after an hour on and off the freeway at uh, supra-legal speeds. No more oil in this intake, thankyouverymuch…The final product, installed. (We put one of Greg's new fuel supply line in too, WWWAIT. Better insulated from valve cover/head/coolant hose heat, easier pass side VC removal, etc. )
FWIW, I will have the clear hose segment at SITP on Saturday and can demonstrate the rpm-dependent oil scevenging to anyone who is interested.
Last edited by Rob Edwards; 05-10-2011 at 09:48 PM.
05-03-2011 | 08:07 PM
#5
Rennlist Member
Joined: Feb 2008
Posts: 2,742
Likes: 0
From: Lillington, NC
Engine looks stock in last photo. Excellent work. What are the kits pricing at? I'm due an intake refresh soon and would really like to put this kit on to finally address one of my GTS' oiling issues.
Trending Topics
05-03-2011 | 09:03 PM
#8
Nordschleife Master
Joined: Apr 2003
Posts: 9,600
Likes: 34
From: Abbotsford B.C.
Looks great,
judging by the components (daily engineering pump?), and mods to the pan I think $2500-$3500 possibly as high as $4000. Seeing as how I know Greg claims he never charges as much as he should.
judging by the components (daily engineering pump?), and mods to the pan I think $2500-$3500 possibly as high as $4000. Seeing as how I know Greg claims he never charges as much as he should.
05-03-2011 | 09:55 PM
#11
Chronic Tool Dropper
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Joined: Oct 2001
Posts: 20,506
Likes: 549
From: Bend, Oregon
For those drooling at home, the workmanship on this stuff is absolutely top-notch. Only what you would expect. There's so much that you don't see, a bit of a requirement here in smog-**** Kaleefawneeah.
Rob suggested that he wil be doing some, uh, Sharktuning on the way up to SITP, so there will be a very real-world test of the system just by checking the dipstick before and after.
Rob suggested that he wil be doing some, uh, Sharktuning on the way up to SITP, so there will be a very real-world test of the system just by checking the dipstick before and after.
