Oil issues: Dry Sump, Accusump for road racing?

[Santiago]

Mulry, can you get a rough price quote for a full ARE system? For some reason I keep thinking that the last run down I saw was some sort of "shop price" (not for retail direct to customers). Anyway, have fun! I was supposed to be there but the wife decided to have a birthday this weekend...

Best,
-j

[Mulry]

The "full ARE system" is a pretty ambiguous thing since it's typically customized for every car. The price list that ARE gives out with "shop prices" or "racer prices" or whatever it says is the direct retail to customer price that I paid for the parts that I bought from them. PM me your email address and I'd be glad to share the price list that I got earlier this year with you.

[Scargo]

Chris of "Killer B" has been adamant that the Subaru motor needs four scavenging pumps. One for each head and two in the sump. Ask them what they think about this. I'm guessing they are using an external oil pump and don't use the stock pump, like Cosworth does. I know adjustments are simpler if it is external or by using an external pressure control valve. Perhaps you could get some feedback about these things.
I like the Daily setup (underneath), but beyond that I am fairly ignorant. I need to get off the fence soon. Thanks!

[Mulry]

I like the Daily setup too, and he'll be there too. I just didn't want to have to re-do my entire exhaust system yet, which would be a requirement for the Daily setup. I'll try to ask ARE and/or Daily about those topics. It makes zero sense to me that you would need a scavenge stage for the heads since those will be draining to the dry sump plate via the suction action of the scavenge pumps attached there. I'm interested in that answer too. The other problem with that issue is that there is no room for a 5-stage pump on the EJ engine, other than underneath (necessitating the exhaust rebuild) or WAY up high above it.

[Scargo]

Agreed. The idea is that at high G's that where the oil is and where it needs to be removed from. Not sure if I buy the logic since there is a lot of oil in a DS reservoir. not sure how it would be accomplished. I am fine with a custom exhaust where the turbo is low and in front.

[Mulry]

Yeah, I get the theory, but where I have some cognitive dissonance is in the idea that there is enough slosh in an actively scavenged motor to pool in the heads.

[C.Plavan]

Porsche 911's have been Non head scavenged since 1965-1998.

[Mulry]

Porsche 911's have been Non head scavenged since 1965-1998.

Exactly. I'm not sure that we really even need to loop the head breathers to the PVC port on the crankcase, I suspect we could just plug all 3 and have fewer avenues for air to leak into the scavenge system.

[Scargo]

Yeah, I get the theory, but where I have some cognitive dissonance is in the idea that there is enough slosh in an actively scavenged motor to pool in the heads.

DOH! Palm to the forehead! That makes a lot of sense...

[Sgt.Gator]

I'm repeating myself so please excuse me...but not only does the Element Tuning "kit" not use a pressure pump or scavenge ports in the heads, the Cosworth Engineering system doesn't either. You may not fully trust Element Tuning, but I trust the Cosworth Engineers to know what they are doing. After all, they build Formula 1 parts too.

[Scargo]

I have yet to use a dry sump system so I hope that keeps me from getting my head chopped off.
As I understand it, with a DS, an external pressure pump allows for fast and simple adjustments of the oil pressure. Otherwise it is a bit of a chore to get to the plug in the stock pump location and change the spring and/or shims and try it again. Then, you must change out the stock located pump to affect volume.
Oil pressure seems the single most critical element you can look at in an oiling system. Others are: temperature, volume, aeration and (?).
Perhaps there is not that much variation in oil pressure for most conditions. Needs also vary by all internal clearances and the overall quality of the system. Stock galleys and passages in the castings may not be optimal. I wonder if any additional cross-drilling is done on 2.5L cranks like is done on some 2L cranks? Lots of work is done by various race engine builders and not much info is shared. Hoses or fittings might be too small.

The other issue is cost. I did research all these systems to some degree. Cosworth doesn't give you much for the huge outlay. Since there seems to be some of variables in the Element Tuning system there is doubt in my mind about what you get for your money and why you need ET to do their pieced-together system.
I am a DIY who has no fear of ordering hoses, tanks and other pieces. I have a reliable non-DS system in my ST2 level STi track car. Yet, I know of many STi engines that "blow up" on the track when they are basically left stock.

I just want to know whose DS system has a solid track record. Literally. I don't feel like I can call up Cosworth and ask for advice. Their system seems discontinued.* With Dailey, I feel that they are available. I am unsure about ET. The thing is, I won't be calling any of them up and asking if their system is superior. One TT weekend per engine rebuild is not convincing enough for me. Perhaps that's short-sighted? I want users telling me that their DS system works. Ones that can get a season out of the engine. There just aren't that many Subie DS users out there and data seems hard to find.
Another small benefit from an external pressure pump is the ability to prime a new engine with oil before cranking it over or starting it. Normally you would have to find a place to tap into the system, create a special adapter, change that back afterwards and, depending on your setup these spots can be difficult to get at. I have had a tap on the left AVCS banjo bolt for this purpose and to mount a pressure sensor. That location seems the most accessible.
*Correction: It seems that only Rallispec has discontinued selling the Cosworth DS system. I see that it is available elsewhere.

[Mulry]

I have some definitive answers from a conversation I just had with Gary Armstrong. Too much to type on a phone but will supplement later. Stay tuned

[Zach34]

FFR has been running Element's setup, if I'm not mistaken? A call to them would probably get you a good testimonial.

[DanielsDM]

When I visited FFR in August Tony told me they were running a Cosworth dry sump on their grey 818R now. I don't know what they had before.

[C.Plavan]

When I visited FFR in August Tony told me they were running a Cosworth dry sump on their grey 818R now. I don't know what they had before.

Jim at SEMA said the Element Tuning Drysump.

[Mulry]

I have some definitive answers from a conversation I just had with Gary Armstrong. Too much to type on a phone but will supplement later. Stay tuned

Ok, sorry for the delay. I wanted to get to a keyboard for this and all I had was my phone and tablet while I was in Indy but am home now.

On Thursday morning at PRI, I had a longer conversation with Gary Armstrong of ARE about some various dry sump configuration questions. With regard to Sprintric location, he said that it's ideal to locate the Spintric closer to the dry sump oil tank rather than further from it. Keeping it closer permits sending only one line from the scavenge side of the oil pump to the oil tank location. This is especially useful (from a cost and complication perspective) if you are going to keep the oil tank far from the engine, like they do in NASCAR, where the oil tank typically resides behind the driver.

With regard to the issue of needing a separate stage to scavenge the heads, he does not believe that it is necessary because the oil pump will scavenge all the oil from the heads, but even if there is limited pooling in the heads in a long sweeper-type turn, there is enough excess capacity in the oil tank that you won't lose lubrication. We should loop the head breathers to the crankcase vent (where the PCV valve is located, but we don't need the PVC valve) and then tie those three locations back to the top of the oil tank so that there is ventilation through the engine. Without providing that ventilation, the oil sump can't suck -- it will create too much suction pressure and fail. He used the analogy of if you put a straw in a drink and cap the straw with your finger and remove the straw from the drink -- the suction prevents the fluid from leaving the straw. Same thing if we don't provide ventilation to the crankcase and heads. Those lines shouldn't see very high pressure, so the solution that Sgt Gator came up with (I think?) should be an effective solution.

That's all I can remember on that topic right now. It was a long (and incredible) day and a half and my head is filled with info about various topics from it, so I might be forgetting something. We did have a longer conversation, so if you have any other questions just ask and I might have already asked them of Gary. Cheers.

[Scargo]

This part, "then tie those three locations back to the top of the oil tank so that there is ventilation through the engine. Without providing that ventilation, the oil sump can't suck -- it will create too much suction pressure and fail.", is the part of the comments I can't resolve.
I read that a major benefit of DS is the negative crankcase pressure it develops and how it assists in sealing the rings and boosts horsepower. As with the system Cosworth offered, a regulator would be in the system to control crankcase vacuum and pressure and the system would be otherwise sealed. If the pump is having issues then it should be upsized. I see ARE offers a pressure regulating module that ties in with the scavenge pump cluster, but they're using the stock Subaru pump, correct? What size do they recommend? I would guess an 11 or 12 that might be unregulated...

[C.Plavan]

This part, "then tie those three locations back to the top of the oil tank so that there is ventilation through the engine. Without providing that ventilation, the oil sump can't suck -- it will create too much suction pressure and fail.", is the part of the comments I can't resolve.
I read that a major benefit of DS is the negative crankcase pressure it develops and how it assists in sealing the rings and boosts horsepower. As with the system Cosworth offered, a regulator would be in the system to control crankcase vacuum and pressure and the system would be otherwise sealed. If the pump is having issues then it should be upsized. I see ARE offers a pressure regulating module that ties in with the scavenge pump cluster, but they're using the stock Subaru pump, correct? What size do they recommend? I would guess an 11 or 12 that might be unregulated...

FYI- I'm going back to venting to the tank.

[Sgt.Gator]

To clarify, the analogy would only be correct if the entire system were sealed, but it's not. The DS tank is vented to open air thru a catch can. So the oil coming from the DS tank to the stock oil pump is never working against a closed suction. The OEM pump will not fail because of it. In fact since the DS tank is higher than the OEM pump, the pump is getting oil fed with a slight positive pressure due to gravity.
You did say the oil sump suction pump would fail, and yes there the system is under suction, if you don't count the gas blow by of combustion past the piston rings. Which of course is the purpose for venting a crankcase, otherwise positive pressure builds up and blows the seals outwards and other bad things.
I'm surprised he has that concern since most dry sump system builders work hard to seal up the engine and create negative pressure, up to a point of diminishing returns. I find it a little humorous that most builders like Killer Bee, say you must have multiple scavenge pumps to be able to handle all the positive blow by pressure because there is too much positive pressure for the Element Tuning and Cosworth pumps to keep up. And now ARE is saying to vent the engine block because you will create too MUCH negative pressure!
Obviously dry sumps on Subarus is still an Art and not a Science!

[C.Plavan]

I just was having some weird oil pressure readings after sealing the motor up. I never had them before when it was vented to the tank. I'm not risking it.

[Sgt.Gator]

I just was having some weird oil pressure readings after sealing the motor up. I never had them before when it was vented to the tank. I'm not risking it.

Further proof your engine is possessed by a gremlin.

[C.Plavan]

Further proof your engine is possessed by a gremlin.

True!

[Scargo]

Sarge, I agree... it's black art in many a mind. Certainly lots of variables. I do suspect you could copy or mimic what Porsche has done and have success, though I can't find details about their system and in one place it said the oil reservoir was built-in to one of their engines! Perhaps someone has some specs on the system and performance specs, like oil pressure, vacuum levels and controls.
I have been reading for days now and I'm not sure I'm a lot smarter than when I started. I hope I am not repeating things already discussed here, but I'm going to include everything I feel is relevant anyway. Grab a beer and find a quiet place. I got a little carried away...
A boxer motor may have some unique problems and solutions. I question the efficacy of sticking with the stock Subaru oil pump for DS. Can they really be relied on, whether 10, 11, or 12mm, as a "one size fits all" solution? Perhaps within 7,000 RPMs, but what about my desire to turn 9,000? Already we see pulleys being changed to slow scavenging rates and some oil pumps may be constantly bypassing at high RPMs from too much volume and subsequent high pressure.
As to having or not having a closed system: With all case venting closed off, head vents interconnected and valve covers sealed properly (IE, no leaks) the system is essentially closed. The pressure pump feeding the engine is pulling nothing but oil from the reservoir (with atmospheric pressure help). The pump is creating internal pressure which works against what the scavenging pumps are doing. As long as you can adjust each pump then you can control the vacuum that the scavenging pumps are able to generate. The blowby is where it becomes tricky in that a turbo can have a significant effect. Not enough scavenging and your oil seals and/or gaskets might fail or even blow out. Too little suction from the scavenging pumps could have a detrimental affect on oil pressure in a sealed system. *This is where the regulating is a critical element of the system.* A vacuum regulator keeps you from exceeding the perhaps optimal 8-10, or even possibly 14 inHG of crankcase vacuum. It seems that around 14" or higher you need squirters to keep the pistons and pins lubed.
From this Engine Labs article: "In road racing applications, and some drag racing instances where there is frequent throttle fluctuation from changing conditions or “pedaling” the throttle, there may also be a need for a pop-off valve to help relieve built up pressure. Under these circumstances, the engine can actually go from a vacuum situation to a positive pressure situation. Generally, these pop-off valves have one-way operation and are typically placed on the valve cover. They open at a specified pressure and allow crankcase pressure to vent to the atmosphere."
So, what I conclude is, too little suction in a closed system is the biggest concern. *Increased vacuum will lower the oil pressure so it must be compensated for.
I'll buy into the fact that aeration is, or can be a big problem. A good DS storage tank with air/oil separation capabilities is absolutely essential. I have a hunch that with enough oil, a well-designed tank and a properly set-up scavenging system, you are likely to get enough entrained air released to not require something like the Spintric air-oil separator. It's also important not to oversize the scavenging return hoses. From what I've read, a single, -10 return line would be the maximum needed and only a -12 for the feed line.
What about optimum pressures? Some suggest 65 PSI@2K to 85 PSI max(?) is the sweet spot for our motors. I don't know what kind of bearing clearances were being used but the Aussies that were suggesting this were building top performing, high HP, TT cars.

Is there a consensus of the order of the DS components and what is needed? Some say you go from the pumps scavenging the hot oil straight to the tank. I think you would want to go through the Spintric or other de-aeration device first, if equipped. I agree with the argument that air entrainment is best resolved when the oil is the hottest and thinnest. From the tank, which also must slow the oil flow and separate out air, you then go to the pressure pump (with adjustable pressure relief valve), to the main filter, to the cooler and finally into the engine. Some use large particle caching, in-line screen filters, like an Oberg, or, in the case of a Dailey or Zen system, screening in the pan to catch big particles as the oil is scavenged out, but before they can reach the scavenging pumps. A thermostatic bypass valve can be inserted just before the oil cooler. Then there's the breather tank off the main tank and the main tank may have a heater element installed.
I'm not sure if any check valves for oil are required. That might depend on the tank placement. Could you potentially see the tank drain its contents into the engine? Oil filters typically have an anti-syphoning check-valve but that would not prevent oil from moving towards the engine.

So, back to the "black art" part where, because of variables in builds and use, you need someone who knows what they are doing to spec the system for you or you may be doing a lot of R&D and spending a lot of money to get it right.
One of my next searches will be to see if there is a better piston ring combo for DS. I'm going to start with calling Total Seal. Some research has indicated that that's the case.
I'm certainly hoping for discussion and debate, especially from those with experience, and stand to be corrected, as 98% of this is "from the internet".

[Mulry]

Here's how I've designed my setup. I haven't run the engine yet, but I did enough research before I started buying (and follow up after) that I'm confident that lubrication won't be a problem:

Dry sump pan -> 2 scavenge stages on DS oil pump -> -16 long pipes to the front of the car -> Spintric -> oil tank (with pre-heater and Olberg tattletale filter on the inputs) -> oil filter -> oil cooler -> -16 long pipe return to the engine -> pressure stage of DS oil pump -> oil input @ former oil filter location on block.

I've gutted the gears out of the stock oil pump so that the pulley is still there for timing belt purposes, but won't develop any pressure.

I don't think that DS oil systems are really any great mystery, nor is the EJ motor, nor is the effect and use of DS systems on boxer engines. Some of the specifics are different, but Porsche basically solved this problem in about 1965 and the fundamentals haven't changed since then.

[RetroRacing]

-16? Wow! At 100psi at 10,000 rpm, we only use a -10 oil line with straight 40 racing oil in our rotary's

Is the oil system really that different on these engines?

[Mulry]

-16? Wow! At 100psi at 10,000 rpm, we only use a -10 oil line with straight 40 racing oil in our rotary's

Is the oil system really that different on these engines?

I don't think that it's that different on these engines, that's what Gary Armstrong recommended. But my oil cooler outputs on a -12 line, so I may use that for the feed. He strongly suggested a -16 for the output due to the air entrained in the oil.

[Scargo]

Output means what? A large hose from the output from the scavenging pumps goes against several sources I found. I have a -12 on one side of my Setrab 12x12x2 oil cooler: 450-550 HP #50-640-7612 12.99"X12.20", 45,000-67,000 btu/hr range. I assumed that the -12 fitting was the input to the cooler. I only have -10 going in and out of it.

[Sgt.Gator]

Mulry how are you going to force oil thru the filter and oil cooler if they are after the tank and before the DS pump? The oil coming out of the tank is only under the pressure of gravity.

[Scargo]

I just want to say that having read a few more things I came across this, which seems to answer what I asked about the possibility of the oil tank draining into the engine: This usually won't happen right away. When the pump is brand new it will be tight enough to hold oil against gravity. (The tank will invariably be higher than the pump.) Then when it gets a little wear on it, the pump will let the oil drain back through. You can fix this by a) draining the oil out of the tank between events, which is good practice anyway, or b) installing a shut valve or check valve in the line.
Thanks for the suggestion about the Oberg filters. I think these ARE fittings with screens are a great idea if the scavenge pumps allow for them, but still wonder how to deal with a Dailey type system when and if you end up with chunks of metal entering the pumps. Perhaps Dailey has a solution?

[Sgt.Gator]

I've read about guys putting an electric drill to the scavenge pump and spinning it up for a minute to move excess oil out of the DS sump and back to the tank. If the pump also includes an engine pressure stage it pre-oils the engine before you start up. That is relatively easy for pumps mounted on top of the engine, more difficult for the pumps mounted low. Probably both are near impossible in a mid engine location like an 818.

[Mulry]

The issue of the oil tank draining into the engine is interesting, but the possibility largely rests on the location of the oil tank versus the location of the oil pump. If using a Dailey pump integrated into the dry sump, then you could have the pump lower than the oil tank output. However, if you have a 3-stage pump that's mounting above the driver's side cylinder head, then it's unlikely that your oil tank will be higher than your pressure stage. In that case, you might want a check valve to prevent oil from backflowing out of the pressure stage when the engine is not running, although there is a fair amount of debate on whether or not that obstruction in the lubrication path is a good thing.

Because my oil system will be largely lower than the pressure stage of the oil system, I'm more concerned about priming the system prior to starting (when cold/dry). Sgt Gator notes the problem of the motor location in the chassis, but it might be possible to route a 90-degree fitting through a firewall cover to permit access to the nut on the front of the oil pump drive. Downside is that you'd also need to remove the belt prior to spinning and that may be a PITA, but is still better than wear on the bearings...

Scargo, those screened dry sump fittings are pretty commonly found in the dry sump world. Dailey uses something similar if memory serves from looking at their gear at PRI, but they also may use a screen in the pan itself.

FWIW, Gary Armstrong recommended using a #30 screen in the Olberg filter. Fine enough to catch the chunky bits, coarse enough to not impede oil flow into the oil tank. Finer bits should be captured by the media-based oil filter as the oil returns to the engine.

[Sgt.Gator]

Mulry I'm still wondering about post 348....

And I recall a guy that paddocked near me with a NASCAR truck. He used the electric drill trick to oil prime his engine every morning before startup. I guess he had it down so it was no big deal to him.

[C.Plavan]

http://www.jegs.com/i/Peterson+Fluid...VP0aApBf8P8HAQ

[Mulry]

Sgt. Gator, 3 options available. Chad beat me to the answer on one of them. Either:

1. That Peterson remote filter mount;
2. By pulling the oil drive belt & using a drill on the nut on the dry sump oil pump drive shaft, if I can get to it and it's not too much of a PITA; or
3. Install an Accusump that I have sitting on the shelf. I don't really need the extra oil capacity but that would be the easiest way to prime & pre-lube the system, just flip a switch.

So probably #3 since I wouldn't have to buy anything new. Unless option 2 is easier than I'm anticipating.

[Sgt.Gator]

#2 not only primes the engine but also empties the dry sump back into the DS tank, which is what started this discussion.

I still don't understand your post #344 on the schematic of putting the filter and oil cooler between the DS tank and the pump. Are you planning on the pump suction to pull oil thru the filter and cooler?

[Mulry]

#2 not only primes the engine but also empties the dry sump back into the DS tank, which is what started this discussion.

I still don't understand your post #344 on the schematic of putting the filter and oil cooler between the DS tank and the pump. Are you planning on the pump suction to pull oil thru the filter and cooler?

That's a good catch. I misread the diagram. So that's not the plan anymore Thanks! Gonna have to reconfigure that part.

[Mitch Wright]

I have been talking to a gentleman from the Netherlands who is with Saker Race cars which use a EJ engine and are pulling lateral g of 1.7 m/s2 using a wet sump. We have been discussing the solutions that they have come up with and the issues that they have had. As I get more information I will pass it along.
What has been really positive is as I have waded through the BS the direction I have taken is pretty much inline with Saker has been successful doing.
http://sakersportscars.com/?lang=en

[FFRSpec72]

I have been talking to a gentleman from the Netherlands who is with Saker Race cars which us a EJ engine and are pulling lateral g of 1.7 m/s2 using a wet sump. We have been discussing the solutions that they have come up with and the issues that they have had. As I get more information I will pass it along.
What has been really positive is as I have waded through the BS the direction I have taken is pretty much inline with Saker has been successful doing.
http://sakersportscars.com/?lang=en

I'm still not convinced that dry sump is the way to go, my attempt will be to stay wet sump, killer-b pickup, killer-b baffles, killer-b oil pan (have STI pan now), killer-b mechanical control valve, air oil separator and be done with it. What I have seens so far are pseudo dry sump systems.

[Scargo]

I'll be interested in what you find out.
Tony, I don't know what you mean by pseudo dry sump. Do you not consider Element Tuning's Dry Sump Kit a real dry sump system and if not, why? Is it the use/reliance on the internal Subaru pump/not creating a vacuum on the case?

[FFRSpec72]

I'll be interested in what you find out.
Tony, I don't know what you mean by pseudo dry sump. Do you not consider Element Tuning's Dry Sump Kit a real dry sump system and if not, why? Is it the use/reliance on the internal Subaru pump/not creating a vacuum on the case?

One of the factors is as you point out the total reliance on the stock pump system, others being that most are 2 stage only, I have not seen anyone install a 5 stage system. I think that you may achieve the effect of keeing the heads oiled using a Element Tuning dry sump system but I think you can also achieve the same effect with much less money and uncertainty.