This is a long one...
Dan/tunapie got me thinking more about the loss of oil through the valve cover vents experienced on various turbo and BBSC cars especially because his has been leak-down tested and shown to have very low true blow-by. I too had something like it happen when I ran mine at Gateway. Prior to that I hadn’t seen any oil, but there I burped out a bunch. Others have also said that hard runs on the street don’t push out a drop, but on the track they lose a lot. Now I’ve got a theory or two about what happens and I’d like others to kick it around and look for the fatal flaw.
This is based on the 3.0 engines which have the PCV valve on the front cover. There were changes to the 3.2 which may or may not have the same problem
The NSX block has huge oil return drain passages leading from the head down to the pan. There are four down each side of each bank (actually, some of the inner ones lead to the main bearings rather than dumping into the pan). These passages are what create those large ribs down the sides of the block and under the intake manifold. Since the engine is a 90 degree V the outer passages “hang” off the lowest sides at 45+ degrees. They are much larger at the top than the bottom. Interestingly, where they empty into the pan the rear outer passages are quite a bit smaller than the front outer ones. (14mm vs. 17mm which is a big difference in area) Makes me wonder why. Anyway, on the rear valve cover the vent pipe comes out high and close to the inside of the V. Under the valve cover you find a flat plate in the shape of a U, creating an air chamber under those conspicuous bumps in the cover, with the vent pipe out one end. The plate is sealed and solid from the vent tube up to the far end, across the end, then back down the other side, which seems to rule out a simple path for the oil. In the center of the far side (outside of the V) are ten 8mm holes for air to pass between the engine and vent chamber. Just one of those is located on the extreme outer/lower edge while the rest are clustered slightly inside/above of center, so I’d guess the one acts as a drain for any oil that does splash up and through the others.
Now imagine sustained high revs and therefore high oil pressure and flow to the head for the cams etc. I need to examine an engine and the drawings more closely, but I think the oil that actuates VTEC then also flows into the head below the cover. (can’t think where else it would go) If so, sustained high RPM means lots of extra oil that needs to drain back down.
Switching to the front valve cover vent for a minute, it normally vents to the intake under vacuum. That vacuum actively draws pressure out of the crankcase, so simply routing the line to atmosphere reduces it’s ability to relieve the pressure. I think that’s a contributing factor to our problem, though I’m not sure how big of a factor. It may be that as oil flow gets very high, oil builds up under the valve cover unable to drain back down for several reasons, one of them being that pressure below in the crankcase. Add boost to the engine and even with good piston rings the volume of gasses past those rings must rise somewhat, increasing the amount of venting required at the same time pressure relief is compromised.
Now add to that high G’s in various directions, sloshing around all that oil up top and possibly centrifugal force “holding it up” in the drain passages. With all that happening the result may be excessive “pooling” of oil under the valve covers, and when you get enough under the rear one it sloshes up through the baffle holes and is carried around to the vent. Belch!
One possible way to decrease the pooling might be to increase the venting capacity of the front cover so it can relieve crankcase pressure and allow oil to drain faster. So a much larger vent tube there may help, allowing the drains to do their job better. I’m not real confident that the pressure is the main thing slowing down the drains, but it’s a simple enough mod for starters. I think the amount of oil escaping might be reduced with a few strategically placed drain holes in the inner/upper side of the cover baffle, allowing oil to run back out before reaching the vent tube. I can’t think why Honda didn’t do that to begin with. But that would still leave lots of oil pooling up there, which may be a problem for other reasons
I’m wondering if the same thing contributes to the oil starvation failures seen on high banks. That’s generally assumed to be caused by oil in the pan being forced to one side by G loads, leaving the pickup to suck air. But if the NSX sump is well designed that shouldn’t happen when you have the proper amount of oil. However, if a bunch of the oil is trapped under the cover and in the return passages then the risk of starvation increases as though you were low on oil.
Another possibility is that under high G loads the oil in the pan may be sloshing over and under the drain outlets, effectively obstructing them and slowing the drain flow from above, but I need a closer look at the pan and baffle arrangement to see if that’s possible. If so, some baffle mods down there might be part of the solution.
Looking for input and ideas.
Dan/tunapie got me thinking more about the loss of oil through the valve cover vents experienced on various turbo and BBSC cars especially because his has been leak-down tested and shown to have very low true blow-by. I too had something like it happen when I ran mine at Gateway. Prior to that I hadn’t seen any oil, but there I burped out a bunch. Others have also said that hard runs on the street don’t push out a drop, but on the track they lose a lot. Now I’ve got a theory or two about what happens and I’d like others to kick it around and look for the fatal flaw.
This is based on the 3.0 engines which have the PCV valve on the front cover. There were changes to the 3.2 which may or may not have the same problem
The NSX block has huge oil return drain passages leading from the head down to the pan. There are four down each side of each bank (actually, some of the inner ones lead to the main bearings rather than dumping into the pan). These passages are what create those large ribs down the sides of the block and under the intake manifold. Since the engine is a 90 degree V the outer passages “hang” off the lowest sides at 45+ degrees. They are much larger at the top than the bottom. Interestingly, where they empty into the pan the rear outer passages are quite a bit smaller than the front outer ones. (14mm vs. 17mm which is a big difference in area) Makes me wonder why. Anyway, on the rear valve cover the vent pipe comes out high and close to the inside of the V. Under the valve cover you find a flat plate in the shape of a U, creating an air chamber under those conspicuous bumps in the cover, with the vent pipe out one end. The plate is sealed and solid from the vent tube up to the far end, across the end, then back down the other side, which seems to rule out a simple path for the oil. In the center of the far side (outside of the V) are ten 8mm holes for air to pass between the engine and vent chamber. Just one of those is located on the extreme outer/lower edge while the rest are clustered slightly inside/above of center, so I’d guess the one acts as a drain for any oil that does splash up and through the others.
Now imagine sustained high revs and therefore high oil pressure and flow to the head for the cams etc. I need to examine an engine and the drawings more closely, but I think the oil that actuates VTEC then also flows into the head below the cover. (can’t think where else it would go) If so, sustained high RPM means lots of extra oil that needs to drain back down.
Switching to the front valve cover vent for a minute, it normally vents to the intake under vacuum. That vacuum actively draws pressure out of the crankcase, so simply routing the line to atmosphere reduces it’s ability to relieve the pressure. I think that’s a contributing factor to our problem, though I’m not sure how big of a factor. It may be that as oil flow gets very high, oil builds up under the valve cover unable to drain back down for several reasons, one of them being that pressure below in the crankcase. Add boost to the engine and even with good piston rings the volume of gasses past those rings must rise somewhat, increasing the amount of venting required at the same time pressure relief is compromised.
Now add to that high G’s in various directions, sloshing around all that oil up top and possibly centrifugal force “holding it up” in the drain passages. With all that happening the result may be excessive “pooling” of oil under the valve covers, and when you get enough under the rear one it sloshes up through the baffle holes and is carried around to the vent. Belch!
One possible way to decrease the pooling might be to increase the venting capacity of the front cover so it can relieve crankcase pressure and allow oil to drain faster. So a much larger vent tube there may help, allowing the drains to do their job better. I’m not real confident that the pressure is the main thing slowing down the drains, but it’s a simple enough mod for starters. I think the amount of oil escaping might be reduced with a few strategically placed drain holes in the inner/upper side of the cover baffle, allowing oil to run back out before reaching the vent tube. I can’t think why Honda didn’t do that to begin with. But that would still leave lots of oil pooling up there, which may be a problem for other reasons
I’m wondering if the same thing contributes to the oil starvation failures seen on high banks. That’s generally assumed to be caused by oil in the pan being forced to one side by G loads, leaving the pickup to suck air. But if the NSX sump is well designed that shouldn’t happen when you have the proper amount of oil. However, if a bunch of the oil is trapped under the cover and in the return passages then the risk of starvation increases as though you were low on oil.
Another possibility is that under high G loads the oil in the pan may be sloshing over and under the drain outlets, effectively obstructing them and slowing the drain flow from above, but I need a closer look at the pan and baffle arrangement to see if that’s possible. If so, some baffle mods down there might be part of the solution.
Looking for input and ideas.
