Killer B Motorsport 2022+ FA24 BRZ/GR86 Oil Retention Baffle Development

[vindiesel]

Originally Posted by 900BRZ

Thank you! Yeah I think those are interesting ideas. May need a new header to install a deeper pan, but that is a possible route to go down. Similar idea to a slight overfill. Doesn’t solve the surging/sloshing, but may reduce the impact.

I have read about some modifications to pickups to remove bends (I believe this was in a forced induction FA20). They argued that the bend was causing cavitation and, thus, starvation. So in general I’m hesitant about bends unless they have significant testing, which seems lacking even in the factory FA24.

In theory this pressure drop could be a result of cavitation in the irregularly shaped FA24 pickup rather than lack of oil, but it’s hard to say without more research. At this point, I’m hoping that the prototype baffle works and I keep can tracking this car for years to come.

I’m hoping someone finds a solution soon!

I think “non-track” people would love to see this same test done in a canyon or mountain road… or a comparable ratio of right hand turn g measure to equal net of pressure loss which is no real constant. I think If the street only people know they are carrying atleast 40+ psi of oil pressure this may calm some people down.

Or you will see people selling brz’s based off how many total right turns they made at what g… ha ha

[OldBiker]

The idea of cavitation in the pickup appeals to me... especially since it appears they have expanded the mid-section (screen at the top of the tube) to provide MORE screen area because of the RTV problems. If that is the case, it would take more than oil pan baffles to fix it... and would explain why the older engines seem to do better in this regard. Either way, a scary problem as even relatively low-g right handers seem to cause pressure drops. Kudos to 900BRZ for the great work.

[Frost]

Originally Posted by 900BRZ

Thank you! Yeah I think those are interesting ideas. May need a new header to install a deeper pan, but that is a possible route to go down. Similar idea to a slight overfill. Doesn’t solve the surging/sloshing, but may reduce the impact.

I have read about some modifications to pickups to remove bends (I believe this was in a forced induction FA20). They argued that the bend was causing cavitation and, thus, starvation. So in general I’m hesitant about bends unless they have significant testing, which seems lacking even in the factory FA24.

In theory this pressure drop could be a result of cavitation in the irregularly shaped FA24 pickup rather than lack of oil, but it’s hard to say without more research. At this point, I’m hoping that the prototype baffle works and I keep can tracking this car for years to come.

As someone who has spent 13+ years in the pump industry (specifically chilled water loops, heating and fire pumps), I would caution the implication that cavitation is the cause of the pressure drop. Given the way the oil system has to wait for oil to drain back to the pan, the oil pickup isn't pressurized. Correct me if I'm wrong, but in this case the oil pump actually has to suck oil UP from the pan to pump it through the engine.

EDIT: Please note, I am writing this to add constructive feedback to the the diagnosis, I LOVE your critical breakdown and data. I mean no disrespect. You have provided by FAR the most data driven analysis of this problem. Many thanks.

From having seen internal documents within Subaru, they said they adjusted the shape of the pan to reduce frothing at the pickup. I assume this is more to ensure that consistent oil flow can be had at the pickup. This leads me to my point:

I think the issue is that the pickup is not reliably being presented with oil to suck up. The pump will cavitate as a result but the cause isn't the cavitation but a byproduct. If frothing was an issue, the easiest way to have fixed that was either to make the pickup go deeper into the pan where the highest amount of G doesn't affect it BUT you also have to consider the oil return coming back is actually hitting the spot where the pickup is intaking oil. If the G is high enough, maybe oil isn't flowing back to that spot. Short of seeing the oil paths, I am baffled (pun intended) as to why right turns don't allow for oil to return in enough volume back to the intake.

[Frost]

Also, I'd like to add, the video does make mention of first gens not getting as severe and as long pressure drops on the same turns. Any chance we can get full lap of the first gen followed by a full lap of the second on the same track?

Thanks!

[Teseo]

Hope Toyota's dealerships doesn't see this video...otherwise they could start decline warranties

[Frost]

A little explainer on cavitation:

https://dienerprecisionpumps.com/pum...w-to-avoid-it/

I believe our oil pumps are rotary positive displacement pumps. Again, it will crop up IF you don't have adequate oil supply and thus make the pump have to suck harder to draw up oil.

Remember that in order for cavitation to occur, the fluid in question has to flash vaporize due to the pressure it is being exposed to in effect lowers its boiling point. Thus it flashes into vapor creating a pocket of gas. Then you expose it back to high pressure and your force the bubble to collapse thus creating a shock wave. If we were having cavitation, you'd be wrecking the oil pump pretty damn quick.

I'm not saying cavitation isn't happening, I am saying we're looking at a symptom and not the cause.

The answer is lack of oil accessible to the pickup based on the data we've been shown so far.

[OldBiker]

There might be some cavitation at the impeller, but I doubt it. The cavitation we are talking about (imprecise choice of words) is happening in the pickup tube. The theory being it sucks some air/froth as a result of braking/turning in. But the shape (and according to me the bulbous mid-section) prevents it from quickly re-establishing good flow to the pump.


Imagine sucking fluid through a straight straw... if the straw loses contact with the fluid, you will suck air, but flow will resume immediately as the straw is immersed again. But consider a straw with a big bulb in the middle... it takes longer to re-establish flow because the bulb is full of air now and has to be drained--its pressure must fall sufficiently below ambient to suck up the oil... and a larger volume of gas takes longer... especially since the pump impeller is designed for low-volume hi-pressure operation in fluid, not sucking on air. Just a theory...

[cmiovino]

Originally Posted by vindiesel

I think “non-track” people would love to see this same test done in a canyon or mountain road… or a comparable ratio of right hand turn g measure to equal net of pressure loss which is no real constant. I think If the street only people know they are carrying atleast 40+ psi of oil pressure this may calm some people down.

I'd also like to see this. I'm a heavy autocrosser with a 1st gen, but when you tell people cornering at even autocross on 200TW tires in stock form isn't (or shouldn't) be anywhere close to what you do on a public road, people don't get it. People shouldn't be concerned at all with heavy street driving if you're going 8 or 9 tenths and certainly if you don't have 200TW tires. It's not the same.

Track driving in 3rd or 4th gear around some corners... totally different realm than even 50mph in a wide autocross sweeper.

IMO, it's really track driving these oiling issues crop up at. Also factor in track driving has repeated high RPM use, braking, rinse and repeat. It's a lot on a car.

[Frost]

Originally Posted by OldBiker

There might be some cavitation at the impeller, but I doubt it. The cavitation we are talking about (imprecise choice of words) is happening in the pickup tube. The theory being it sucks some air/froth as a result of braking/turning in. But the shape (and according to me the bulbous mid-section) prevents it from quickly re-establishing good flow to the pump.


Imagine sucking fluid through a straight straw... if the straw loses contact with the fluid, you will suck air, but flow will resume immediately as the straw is immersed again. But consider a straw with a big bulb in the middle... it takes longer to re-establish flow because the bulb is full of air now and has to be drained--its pressure must fall sufficiently below ambient to suck up the oil... and a larger volume of gas takes longer... especially since the pump impeller is designed for low-volume hi-pressure operation in fluid, not sucking on air. Just a theory...

What you're referring to is loss of prime on the pump suction. However, for a positive displacement pump, this isn't an issue because guess what? When your car is turned off and you turn it on, the pump will start and suck up oil from the pan. Unlike centrifugal pumps, they can self prime.

What pushes the fluid up the intake is the atmospheric pressure within the system. Liquid doesn't move to a void, it gets pushed from an area of high pressure towards the void.

The issue here is that you don't want to lose suction flow mid corner at high load.

[OldBiker]

They self prime, but it takes a few seconds... I think that is the problem. I suspect, without proof, that the priming function is slowed down because of the now bulbous shape of the tube... forcing it to suck more air. There is probably the same momentary loss of suction as the older engines, but it now takes longer to re-establish prime and flow.


Actually it is the difference between atmospheric pressure and the pressure at the top of the inlet tube that pushes up the fluid... the rotating impeller simply reduces the pressure a little and up comes the fluid. But all of this takes time.

[gnarjunkie]

Originally Posted by cmiovino

I'd also like to see this. I'm a heavy autocrosser with a 1st gen, but when you tell people cornering at even autocross on 200TW tires in stock form isn't (or shouldn't) be anywhere close to what you do on a public road, people don't get it. People shouldn't be concerned at all with heavy street driving if you're going 8 or 9 tenths and certainly if you don't have 200TW tires. It's not the same.

Track driving in 3rd or 4th gear around some corners... totally different realm than even 50mph in a wide autocross sweeper.

IMO, it's really track driving these oiling issues crop up at. Also factor in track driving has repeated high RPM use, braking, rinse and repeat. It's a lot on a car.

100%

Driving even super hard on the street doesn't put anywhere CLOSE to the stress on a car that track driving does.

Even in the case where someone has a fast curvy road, that they know extremely well, that is 100% clear of any traffic, on 200tw tires, where they are willing to push every corner up to the point of potentially going off the road... there's still the fact that you probably can't do lap after lap after lap of that section for 20 minutes straight without interruption.

[Espi]

Youtube recommended me the video last night THEN I came here as I just knew this thread would exist! LOL

Anyways, good luck everyone and hope a solution is found.

[drift86]

Thanks for making this video @900BRZ, love your work!

I'll like to see the Greddy baffle tested. It has a plate that sticks up that is designed to prevent oil from going into the front chain cover.

The FA20/FA24 engine has a structure in which a large amount of engine oil accumulated in the oil pan flows into the front chain cover due to intense braking G when driving on a circuit. The oil pressure will drop, and in the worst case, the engine may blow.

[Blighty]

Originally Posted by KillerBMotorsport

The engine’s Timing Cover is attached to the front of the engine and upper sump. It is a large piece with a lot of open area. There are some PCV provisions, and there are no baffles or other features to control the oil. There is a large opening from upper sump into the timing cover. The argument could be made that the timing cover is also part of the ‘sump’ because any amount of oil above the minimum level dipstick dot, is also held in the timing cover.

Interestingly just noticing that the design in the BRZ technical changes document from Subaru shows a different layout of the chain cover than what you are showing.

Very strange - wondering what's going on here.