AOS/CC discussion about air oil separators and catch cans...

[bigBADbenny]

So I have continued the researches, with much success! Not wrong but not exactly right either.

The thrust of the research is to try to improve the schema for a dual can setup.

We see too many installs that simply dont work effectively.

Anyway the info is in the FSM for all to see, I'll compress images and post them asap.

[bigBADbenny]

So the relevant section w/ diagram in the FSM states that the forward breather ports on the heads are fresh air inlets from the third from rear port on the turbo inlet pipe. See below in the thread for the FSM page...

This allows fresh air to circulate through the valve covers, crankcase and sump, and consequently the pcv system.

The pcv Y hose is fed from the block at the big port below and right of the turbo compressor and tee’s off to the intake manifold (where the PCV valve is located) and the rearmost, upper port on the turbo inlet hose: where it draws a strong vacuum under boost.

This all has ramifications for running an efficient dual can AOS/CC setup, but its still my strong opinion that using the central CCV port that feeds the so called balance pipes, and also using the #3 port on the inlet pipe as the inlet AOS/CCV return, are both the wrong approach, even if its the currently accepted way.

Eg most aftermarket inlet pipes delete the rearmost port.

[Yowie]

Yes a single can w/pcv delete is blocking/deleting everything stock bar the rearward heads balance ports and CCV, and running the 3 sources to a single can with vent to inlet or ( ) air (or exhaust)

Very interesting. I've given it some thought. Ramblings follow.

To comment on the obvious, with no PCV you miss out on the strongest source of vacuum (manifold vacuum) and have to make do with mere "ported vacuum" ahead of the turbo inducer. Commonly-accepted wisdom is that the PCV setup on a road car is for emissions reasons.

Emissions aside, low pressure and/or a mild vacuum in the crank case is a good thing to:

(a) reduce oil leaks;

(b) increase piston ring seal;

(c) reduce internal wind resistance for the crank & rods;

(d) reduce oil mist in the crank case and reduce the oil clinging to the trailing side of the crank & rods.

Many of the performance advantages above (and any serious % horsepower improvement) only occur when a scavenge pump (with a dry sump system OR just to pull gas out of a wet sump crank) are drawing a strong vacuum on the crank case at all different rev points.

[Summarised from "Four Stroke Performance Tuning" 4th Edition by A. Graham Bell, pages 545-547]


On a road car (under manifold vacuum conditions a lot of the time) I submit that benefits (a) to (d) above will still be present, but won't help on-boost performance.

"Vent to atmosphere" seems to be ok for competition engines, but not as good as a strong source of vacuum at all rev points via scavenge pump. I would think that the more limited vacuum draw of the pre-turbo pipe would still give some benefit (compared to vent-to-atmosphere) under boost conditions. An experience racer/tuner might be able to tell us if the crank vacuum benefits offset the detriment of a closed-loop system for a high performance application.

From the same book on page 551:

"Any wet sump competition engine should have at least a 1" hose connected into each rocker cover and another 1" hose connected into the block"

There is no mention in the book of blocking the crank vent and relying on rocker cover breathers only. Just to speculate, it could work fine or perhaps sub-optimally under some conditions. On a performance engine it seems to me to be a retrograde step to reduce the gas flow (in exit points and total exit-hose cross sectional area) out of the engine. I also wonder if the factory balance of pressure and oil drainage is negatively affected by blocking the crank gas exit path.

---

I've heard of crank case vacuum being generated via exhaust, although not for cars with a turbo or any form of muffler. Seems to be a naturally aspirated race-engine-only style of setup.

[bigBADbenny]

So apart from the aim being an efficient 2 can setup, the other stated aims include:

Keeping varnish out of the PCV valve for reliability.

Maintaining sufficient crankcase vacuum to assist turbo oil drain.

Improve or maintain oil control in the valve covers, help prevent valve cover oil surge in high G corners.

Keep blow-by out of the inlet tract, to reduce chances of detonation.

[bigBADbenny]

Yes absolutely regarding actual crankcase vacuum: Colonel Red Racing on LGT has proven that running 1.5 dry sump oil scavenge stages on the lower valve covers and around 15psi of crankcase vacuum at wot improved everything: oil control and oil temperature especially, in extreme track conditions.

https://legacygt.com/forums/showthread. ... ost5996008

On my stock PCV system my07 GTB, crankcase pressure is generally well within +/-1psi, and can be seen to drop with load, indicating that I have no major compression issues.

[Yowie]

The set of Perrin catch can instructions at this link...

https://www.perrin.com/images/products/PSP-ENG-607.pdf

...seems to show a dozen different ways to connect catch cans and a few different Subaru engine layout variants. They mention the "PCV delete" and "vent off the rocker covers only" options with a few covering caveats.

The diagram on page 11 is a handy guide to the difference between the balance ports (valve covers & crank) and the venting ports (" ").

[bigBADbenny]

That’s brilliant and well up to date.
Its great to see they say leave the balance pipes well alone for that type of engine.

I say if you see a lazy install using the central CCV balance breather: call it out, unless it’s eg a single can setup with PCV delete.
Even then I’d be favouring the sump breather/block PCV feed.

[Yowie]

I would think that the block balance vent would be harder to get to than just using the rear/drivers side vent that is relatively easy to get at (at least on an SH9 forester).

Conversely, it would be easy to "grab the wrong hose" if a person wanted to run a catch can off a rocker cover.

I intercepted the rocker cover gas just before it hits the pre-turbo pipe. I see some Perrin comments on one of their alternative options:

"Connect each valve cover vent separately to each of the AOS valve cover vent inlet ports. This method is what we recommend to
provide the best venting while under high lateral G forces"

[bigBADbenny]

Huge info! ^^^

Can’t wait to draw a line under this research and pull it all together!

Also: the famous LGT coloured vacuum diagram thread:

https://legacygt.com/forums/showthread. ... 225p2.html

[bigBADbenny]

I see some Perrin comments on one of their alternative options:

"Connect each valve cover vent separately to each of the AOS valve cover vent inlet ports. This method is what we recommend to
provide the best venting while under high lateral G forces"

I’m unsure about this.
Subaru states in the FSM that these are strictly fresh air inlet ports and supports the claim with a diagram as well.

That’s why this can is always clean (unless the valve covers fill with oil in high g cornering).

You can see fresh air enters the valve covers and enters the crankcase via the heads oil drain galleries.

[bigBADbenny]

So the valve cover breather baffles must be there to prevent oil surging out of them and into the inlet.

I guess all the different hose sizes and restrictors in the system help to create a pressure differential in the PCV system with the majo vacuums being drawn at the PCV return in the inlet and the actual PVC valve in the intake manifold.

Another thread on LGT mentions that the heads/CCV balance pipe runs alongside the turbo coolant hose supply pipe in order to help evaporate sump condensation...

[Yowie]

Subaru states in the FSM that these are strictly fresh air inlet ports and supports the claim with a diagram as well.
That’s why this can is always clean (unless the rocker covers fill with oil in high g cornering).

The collected contents of my valve covers catch can disagrees.

There is certainly outwards-venting flow from the valve cover vents under some conditions.

At high vacuum (with low blow-by and with the PCV drawing hard on the crank case) there is probably some airflow:

(i) from the pre-turbo pipe (under very mild vacuum);

(ii) back through a catch can in my case;

(iii) into the breather ports in the valve covers;

(iv) through the engine insides and/or balance hoses;

(v) up through the crank vent, (through my crank catch can in my case); then

(vi) through the PCV into the inlet manifold.

(under high vacuum conditions there is probably also some airflow from the pre-turbo pipe to the PCV via the F-shaped branch in the crank/PCV system)


Conversely, under boost conditions the PVC is shut, there is more blowby and the vents in all three locations are sending dirty air to the pre-turbo pipe via various paths.

As the little Mexican girl says "why can't we have both?" (airflow directions for the valve cover breathers). The pasted workshop manual page doesn't rule out 2-way traffic.

[bigBADbenny]

To add to the complexity, there's also the loose-built engine factor.

Would be interesting to use a PCV as a one-way valve on the cam cover breather return.

Changes all my references to valve covers as per FSM.

[bigBADbenny]

On my stock PCV system my07 GTB, crankcase pressure is generally well within +/-1psi, and can be seen to drop with load, indicating that I have no major compression issues.

This test was done with a boost gauge (AEM 30-4900 wbo2 failsafe) connected to a fitting in a spare oil cap.

The scatter graph, made in MegaLog Viewer, shows AFR (X) vs crankcase pressure (Y) and AFR as colour (Z).

[Yowie]

(1) So the valve cover breather baffles must be there to prevent oil surging out of them and into the inlet.

Almost certainly. There are also baffles inside the engine just before the crank vent for the same reason.

The idea is to deal with blow-by gas but keep the oil in the engine as much as possible. Obviously these baffles and even my catch cans are not perfect funk filters or some oil wouldn't be present on the "clean side" hoses after my catch cans.

I wonder how many efficient catch cans in series it takes to remove all oil residue & water vapour? It's a shame Mythbusters are not going anymore.

(2) I guess all the different hose sizes and restrictors in the system help to create a pressure differential in the PCV system with the majo vacuums being drawn at the PCV return in the inlet and the actual PVC valve in the intake manifold.

I would think so. I'm still intrigued by the factory restrictions in the crank F-shaped splitter and what is best practice (factory/some/nil restrictions) for a warmed-over road car. There are too many variables for my experiments to yield any answers.

(3) Another thread on LGT mentions that the heads/CCV balance pipe runs alongside the turbo coolant hose supply pipe in order to help evaporate sump condensation...

Interesting. That makes sense. Sump water vapour is a really big deal on E85 but still a factor on petrol (especially for "cold start 2 min drive to train station" usage where it can't boil off).

This is why I am not a fan of "drain back to sump" catch can setups, even if heated by coolant water. A trickle of liquid water into the sump (eg on cold start) cannot be a good thing. Andrew Hawkins of the Motive videos reports no issues with drain-back catch cans on E85, so I'm probably wrong to be such a sceptic.

(4) To add to the complexity, there's also the loose-built engine factor.

Yep, more blow-by.

A lot of those drag & dyno competition cars just vent-to-atmosphere from the catch can on my understanding. Probably better for their application for a lot of reasons including octane preservation under maximum effort runs.

(5) Would be interesting to use a PCV as a one-way valve on the cam cover breather return.

Preventing which direction of flow?

On a related note, I'm interested in using a OCV on the hose between the pre-turbo pipe and the PCV so the PCV cannot draw clean air from the pre-turbo pipe under manifold vacuum conditions. I can stop worry about the restrictor issue then.