I've been reading about this system recently and will be having a go on my car. The basic principle is to have a tapping in the exhaust connected to the crankcase vent with a check valve between. The system uses the negative pressure pulses that occur in the exhaust to draw crankcase fumes into the exhaust, and even create a modest vacuum which can have mild benefits such as improved ring seating, reduced oil consumption, and marginally smaller windage losses. Looks like an interesting principle as it could be pretty much something for nothing. It could even negate the need for a catch tank without the engine having to ingest oil vapour, and get rid of the nasty smell of oil vapour that's hanging around my car when the engine is running (currently have a catch tank venting to atmos) Any thoughts?
I too have looked into this. From what I`ve found, it appears to be more efficient at higher revs and placement does seem quite critical to the operation of the system. I ended up venting my catch system back into the inlet, not perfect, but proven and it works. I don`t have the unpleasant oil smell I used to get when it was vented to atmosphere. We use the same principle to create suction at work, it works well, but needs a good velocity of the liquid/gas flowing past the tube to work well.
What happens with venting to anywhere other than atmosphere when the catch tank fills up for some reason?
I'd've thought the check valve would soon fail (shut?) in that application. Any oil mist that gets past the separator is going to tend to 'set' on the *hot* innards of the check valve, isn't it?
I'm going to use the EGR boss on the no.3 header, which is about 150mm from the port, so hopefull there should be some good velocity there, I'll be interested to see how the performance changes at different engine speeds though. Sorry, I don't follow This system doesn't necessarily need a catch tank though, you can just feel the vapour into the exhaust where it will burn off. Fair point, I'm planning on placing my check valve higher up the system away from the hot bits! Should be fine as long as the oil doesn't degrade it's internals.
I didn't like the idea of possibly hot oil vapor going into the exhaust system with that idea. You should not need anything more than a factory style setup really. A well baffled catch tank and breather should separate any oil from the breather pipes. The catch tank having a drawn outlet back into the inlet tract before the throttle body should not be sucking in oil if this is done right. This takes care of high load and full throttle breathing. There should also be a small check valve connected to the inlet plenum after the throttle as this takes care of drawing vapor during idle and high vac conditions such as throttle lift off. Just my 2p's worth. Gurds
Yeah that's certainly a drawback, my problem is that it's a twin carb setup so there isn't anywhere to plumb it back into the inlet
My mistake, I thought there was a catch tank in the system you posted. Clearly there isn't. With Nige's catch tank feeding back in to the inlet, if the catch tank ever fills up, then the overflow would just end up in the intake. I have no idea how the check valve works though, so that may ensure this doesn't happen.
On my ITB`s, I tapped into one of the runners and vented the catchtank there. On F/I, my Catchtank at Rockingham was a stock jobbie, fed from the crankcase and vented to atmosphere. This overfilled and leaked oil out of the breather each session. Gurds had a rocker modified with baffle and take off which I now feed to a catchtank I made, this has internal baffles and stainless oil-separation mesh. 3 inlets (2 from rocker, one from crankcase). 1 outlet to the pre-turbo inlet, this is bone dry with NO sign of oil vapour (admittedly after only 6 laps... lol) 1 drain from the catchtank to the sump. Any oil that builds up in the catchtank drains back into the sump straight away, emulsification should be kept to a minimum as it doesn`t sit in the catch tank, it drains immediately
The type of check valve I've ordered is basically a sprung diaphragm that lifts off a seat and allows flow in one direction, but seals in the other. Like a simpler version of this: The vapour should just pass through the device. Some of it will stick to the insides though.
Somewhere at home I've got an almost-new check valve off a Vauxhall Omega's SAI system you could have for nothing if you want. That's connected directly into the exhaust manifold via a foot or so of ali(or maybe steel?) pipe in the original application, so seems like it would suit? Looks like the top bit here. Edit: Obviously that's designed to survive 'fairly hot', but only used with fresh air throughput. Filtered even. And only when the engine's cold.
Interesting, was this on the valve side of the butterfly, or the atmospheric side? Thanks for the kind offer Pete, that would be great. That valve looks much the same as others I've seen used in this context. PM'd
I'm wondering whether there would be enough velocity (and therefore, pressure) in the Venturi-effect created to open a spring-seated check valve? I think a non-sprung ball-type, that would simply be shut by any reverse flow would be better. Also no seals or moving parts that could be affected by the oil.
That's a good point. EZ Pete is kindly donating a check valve originally intended for the secondary air injection system on a Vauxhall engine, he tells me this uses a reed valve arrangement, rather than the sprung type I've pictured above. The nice thing about the Vauxhall part is that it's intended to work in the same environment, injecting air into the exhaust during the negative pressure phase, so should be fine in this application. The only difference is that I will plumb it into the crank case rather than fresh air.
This picture quite nicely illustrates the negative pressure period that this system is meant to utilise
It's on its way matey. I was just thinking about this and the Vx SAI application. A good hint that it will work for you is that there is a second solenoid-controlled vacuum-powered isolating valve upstream of this non-return valve, nearer the pump. This has to be opened actively by the ECU when it wants to pump the secondary air into the system. I believe this must be there to prevent air getting passively sucked into, rather than pumped into, the exhaust system when the SAI is inactive. Any such air would confuse the bejeezus out of the oxygen sensors when in closed loop operation.
Right, I managed to get this all assembled and installed last night. Here it is, reed valve unit (thanks Pete), M22 to M14 adapter, 10mm copper pipe flared into M14 connectors. This attaches nicely to the M14 EGR boss on no.3 header that I wasn't using, making it a really easy installation I may well redo the pipe to fit better in the future, but this is fine as a proof of concept. Here's a vid of it running at idle. There's a pleasant gurgling sound as it sucks in air and definite negative pressure when I put my thumb on it. [video]http://i243.photobucket.com/albums/ff181/ez-does-it/Pan-evac/25906D63-1016-4256-8984-15098CB5D20B.mp4[/video] After going for a drive I can happily say that the nasty smell from the catch tank is all gone. There were some other subjective improvements but they could have been down to the improvement in weather since I last drove it. I'm pleased overall though as it's lighter and probably cheaper than a most catch tanks, and doesn't need emptying. The reed valve is metal and ought to be very robust as this is very similar to it's intended application. I'm now waiting on a T-piece so I can plumb it into the datalogger and see if it generates any meaningful vacuum during normal/WOT driving, results to follow.
Nice one. I could see a case for making the copper pipe a bit longer, to put the valve farther from the hottest bits, but what you've got there should certainly do as proof of concept. Edit: And a very handy view of how the brake lines are meant to look along the firewall, I was going to do a search for something like that. Mine have been done rather less neatly at some point in the car's past.
