Oil issues: Dry Sump, Accusump for road racing?

[walt555]

This was precisely the point of my question. There are 2 holes at the oil filter location on the bottom of the block. In stock form, the large, threaded hole in the middle is where the oil returns to the engine from the oil filter. The small, unthreaded hole to the side is where pressurized oil leaves the block to go through the oil filter.

In a dry sump system, you remove the oil through the scavenge plate. You could return it through the large threaded hole -- but you have to do something to block the small hole.

It sounds like biknman is using the large threaded hole as his oil feed to the engine, which was the point of my original question -- what is he doing to plug the other hole?

To clarify further, I'm not using the stock oil pump, but a 3-stage pump (2 scavenge, one pressure), as is biknman.

Could you use a canton remote oil filter adapter part numbe 22-597 and just plug the output side?

[Scargo]

Yes... Almost any number of solutions. Here's one.
Foto0093.jpg
If you retain the use of the stock oil pump you could also simply modify internally an adapter to return the oil, thus avoiding using a filter for that purpose.

If you wanted to go to extremes... I was looking at a block and the area where the filter mounts. I think you could tap the outlet hole and then plug both holes. Next you would redirect the oil supply right where the oil goes from the pump into the block by creating a hole up into the return passage. This new hole or passage would connect the existing passages going to and coming from the filter mount. It would require some drilling, tapping and plugging and possibly some work with a burr to smooth the transition of the new passage, but I think it could be done without welding. I assume welding on a case would be a bad idea.
All this, on the basis that there is a filter elsewhere in the DS system.

[walt555]

Doesn't the current DS pans ARE OR Aviaid in a 3 stage system, feed oil thru the stock location and gut the oil pump then either use stock oil filter or canton type filter adapter with filter and cooler. Are there draw backs going this route verses adapting to the oil filter location?

[FFRSpec72]

So what hose have people been using as I was looking at Earl's Pro-lite 350 hose and Aeroquip AQP Socketless hose

[RetroRacing]

Earl's Pro-lite 350 hose

[Scargo]

Aeroquip Startlite Race Hose

[Sgt.Gator]

So what hose have people been using as I was looking at Earl's Pro-lite 350 hose and Aeroquip AQP Socketless hose

I asked ANPlumbing.com that question, they said "Auto Flex, Perform o Flex, Ultra Flex SS , HS79, or Ultra Pro SS work fine".

[FFRSpec72]

I asked ANPlumbing.com that question, they said "Auto Flex, Perform o Flex, Ultra Flex SS , HS79, or Ultra Pro SS work fine".

Not heard of most of those before.

[C.Plavan]

Summit AN hose. Been using it forever (Porsche's and 818's). My Porsche oil hose is going on 11 years, no issues. Cheap and good.

[Scargo]

Many ways to skin a cat. RCR cover brings oil in from the external pump on the bottom. Integrated, undermount design similar to Daley. You still have to deal with the filter mount area.
RCM2668.jpgRCM2668.jpg

[Sgt.Gator]

In my post above about hoses I should have been more clear, I asked ANPlumbing for their recommendation for the scavenge hoses that see vaccum, not the pressure or return hoses.

[Zach34]

Is that RCM kit still available? That's the first I've seen of it.

[Sgt.Gator]

I've run into two issues installing the Cosworth DS.

1) The 720 mm HPD belt is about 10 mm too short. The instructions are very adamant that there should be NO tension on the belt. Heck I can't even get it over the pump pulley at all, much less no tension. I checked into a slightly larger belt but the next size up from three manufacturers is 800 mm which will be way too long. The possibility to use an 800 would be to run the tensioner on the cogged side (inside) of the belt instead of the smooth outside. I suppose that since the tensioner is just a smooth idler pulley exactly like the timing belt idlers and there's really no tension, it just takes up slack, it might be ok, but IDK.

For now I'm trying to stretch the 720 mm belt a half inch using a floor jack with a couple of 3" pipes on each side to not bend the belt over too sharp a corner. It's helped a tiny bit, but not enough yet. See pic of my Colonel Red Racing belt stretcher!

Ideas?

2) I thought all the connections were -12AN. They are not, the hose from the pump > cooler> tank is -16AN. Of course my tank has all -12AN fittings and the cooler as well. Somewhere either right at the pump or probably at the cooler I have to go down to -12AN. Since this is for my STI and the cooler +tank are in the trunk I think it's best to neck down right at the cooler so the long line is still -16AN. However it would be a lot easier and cheaper to run -12AN all the way from the pump. The Element Tuning systems many of use run a -12An but the lines to the tank are pretty short. And in Element's Tie Attack car the tank is in the passenger foot/door area which is a pretty short run. Having had the tank in the same place in my Legacy, there's no way I'm putting the tank in the cockpit for enduro races, the heat is unbearable.

And a couple of pics of the engine build. 2.7L Stroker.

Belt Stetcher.jpgInlet Mock Up.jpgCosworth DS Mock Up.jpgExhaust-Turbo Mock Up.jpgCosworth DS Pump Belt Issue.jpg

[biknman]

image.jpeg Just sold my setup ( thanks to Doowop) so took couple quick pics of the tap 1/2 -16 BSF I used to make custom length crank bolt so you can install an HTD pulley on a mandrel and Align to the pump

these are of a factory crank bolt image.jpegimage.jpeg

[biknman]

Doesn't the current DS pans ARE OR Aviaid in a 3 stage system, feed oil thru the stock location and gut the oil pump then either use stock oil filter or canton type filter adapter with filter and cooler. Are there draw backs going this route verses adapting to the oil filter location?

Yes they do. I did it both ways. No real drawback just a different way of doing it. Both routes the stock oil pump gears are gutted. Via the stock oil pan pick up point though a dry sump pan you're pushing oil up through the stock guted oil pump then through the galley to the oil filter thru the filter into the galley to feed the bearings. Which allows you to use the stock pressure bypass spring in the stock oil pump and the stock oil filter pluses. Or you can bypass all that and feed the oil from the pressure stage of the dry sump straight into the bearings via the stock oil filter location center hole (12mmx1.5). This requires having a pressure spring bypass on the dry sump's pressure stage which allows you to externally set oil pressure via bypass spring a plus and then you'll need to filter the oil after the pump before the oil galley, I did this with an Peterson 400 series filter.

Hope this helps :-)

[biknman]

I've run into two issues installing the Cosworth DS.

1) The 720 mm HPD belt is about 10 mm too short. The instructions are very adamant that there should be NO tension on the belt. Heck I can't even get it over the pump pulley at all, much less no tension. I checked into a slightly larger belt but the next size up from three manufacturers is 800 mm which will be way too long. The possibility to use an 800 would be to run the tensioner on the cogged side (inside) of the belt instead of the smooth outside. I suppose that since the tensioner is just a smooth idler pulley exactly like the timing belt idlers and there's really no tension, it just takes up slack, it might be ok, but IDK.

For now I'm trying to stretch the 720 mm belt a half inch using a floor jack with a couple of 3" pipes on each side to not bend the belt over too sharp a corner. It's helped a tiny bit, but not enough yet. See pic of my Colonel Red Racing belt stretcher!

Ideas?

2) I thought all the connections were -12AN. They are not, the hose from the pump > cooler> tank is -16AN. Of course my tank has all -12AN fittings and the cooler as well. Somewhere either right at the pump or probably at the cooler I have to go down to -12AN. Since this is for my STI and the cooler +tank are in the trunk I think it's best to neck down right at the cooler so the long line is still -16AN. However it would be a lot easier and cheaper to run -12AN all the way from the pump. The Element Tuning systems many of use run a -12An but the lines to the tank are pretty short. And in Element's Tie Attack car the tank is in the passenger foot/door area which is a pretty short run. Having had the tank in the same place in my Legacy, there's no way I'm putting the tank in the cockpit for enduro races, the heat is unbearable.

And a couple of pics of the engine build. 2.7L Stroker.

Belt Stetcher.jpgInlet Mock Up.jpgCosworth DS Mock Up.jpgExhaust-Turbo Mock Up.jpgCosworth DS Pump Belt Issue.jpg

Here's my two cents you do not want and HTD or a glimmer belt to be too tight you will wear out your bearings surfaces in your pump super quick :-0. You should be able to twisted the belt halfway. So it's considered kind of rather lose compared to other type accessory drive belts. Belts are cheap and can be bought lots of places as they are used in lots of other applications other than motorsports check out McMaster Carr. Heck I even have one on my bicycle replacing the old chain and cogs ;-P!

Yeah them dog-gone tanks ;-)! For the tank I've had it located in the front right corner of the car (not 818) behind the engine where the stock turbo location would be after moving my turbo to the front of the engine, both of which the tank got beat up by gravel from the tires, bump drafting, and armco / Jersey wall smacks hee hee :-0! The tank gets really dirty if in the engine bay or external. For some reason those oil tanks they get so hot oil just kindof leeches out or whatever through the aluminum (for lack of better explanation) and the whole oil tank is covered in oil which attracts dirt and makes a big old mess. So the best location to keep the thing clean is in the cab but that adds an immense amount of heat to the cockpit which has to be managed :-(. But this all was on a Suby chassis not an 818 chassis.

[DanielsDM]

Thanks for the info biknman!
We are putting together our own dry sump setup and it is good to have details on how it's been done on the EJ engine.

[Scargo]

Today I have been looking at Reynolds values, ID's of AN fittings and flow rates for 10, 12 and 16AN fittings. For instance a 10AN hose is .56" nominal ID and the fittings are .484" nominal ID. 12AN is .69" ID and the fitting ID is .609". 10AN is .184” squared. 12AN is .291” squared. That’s 63% more area!
Good AN fitting-hose-tube sizing info from Pegasus website. There is where I got this chart.
Attachment 60201
I have been curious about the fact that I have nothing that is even 12AN and only fittings on my oil cooler that are 12AN. I admit, I did not study fluid dynamics and I still have some concepts that are fuzzy in my head. The fittings restrict flow more than anything else. As an orifice, are they completely dominant when it comes to the overall flow rate? A 16AN hose may improve efficiencies and offset reduced flow from bends in fittings and hoses and internal friction for that portion of the run (especially the longer the runs) but any smaller fitting in the run would determine the final flow rate, wouldn’t it? Is the 16An there to allow a slowdown of the velocity of the oil so air separates out? Doesn’t the vacuuming volume from the sump need to exceed the output of the pressure pump or otherwise you would have oil accumulating in the bottom of the engine? This infers that a fair amount of air is drawn in with the oil.
With my wet sump system I run only 10AN. I have two hoses in my external oil filter and oil cooler route that are 4 to 5 feet long. I ran an 11mm pump for 70+ track hours without any oiling issues and no bearing failures. This may have no bearing on a DS system however. I welcome any clarification or correction.

[DanielsDM]

Today I have been looking at Reynolds values, ID's of AN fittings and flow rates for 10, 12 and 16AN fittings. For instance a 10AN hose is .56" nominal ID and the fittings are .484" nominal ID. 12AN is .69" ID and the fitting ID is .609". 10AN is .184” squared. 12AN is .291” squared. That’s 63% more area!
Good AN fitting-hose-tube sizing info from Pegasus website. There is where I got this chart.
Attachment 60201
I have been curious about the fact that I have nothing that is even 12AN and only fittings on my oil cooler that are 12AN. I admit, I did not study fluid dynamics and I still have some concepts that are fuzzy in my head. The fittings restrict flow more than anything else. As an orifice, are they completely dominant when it comes to the overall flow rate? A 16AN hose may improve efficiencies and offset reduced flow from bends in fittings and hoses and internal friction for that portion of the run (especially the longer the runs) but any smaller fitting in the run would determine the final flow rate, wouldn’t it? Is the 16An there to allow a slowdown of the velocity of the oil so air separates out? Doesn’t the vacuuming volume from the sump need to exceed the output of the pressure pump or otherwise you would have oil accumulating in the bottom of the engine? This infers that a fair amount of air is drawn in with the oil.
With my wet sump system I run only 10AN. I have two hoses in my external oil filter and oil cooler route that are 4 to 5 feet long. I ran an 11mm pump for 70+ track hours without any oiling issues and no bearing failures. This may have no bearing on a DS system however. I welcome any clarification or correction.

You are correct that the fittings restrict the flow most (on a system where the pipe length is short in comparison to the number of fittings). However having a couple fittings that are a smaller size won't restrict the flow down to where it would be if the entire system was the smaller size. Everytime there is an abrupt change in diameter or direction there is pressure drop. So use as few fittings as possible, make bends as large as possible.
There are some online flow calculators that may be useful. http://www.pressure-drop.com/Online-Calculator/ This one is free, a little clunky UI but it's free.
http://www.pipeflowcalculations.com/ This one is better but not free.

[Scargo]

...There are some online flow calculators that may be useful. http://www.pressure-drop.com/Online-Calculator/ ...

Funny, that it is exactly what I have been using. I get a Reynold's number of 7-8K for .56 ID (10AN). It does seem to be in the range of being turbulent but that's as much as I can make of it. Commonly used 12AN is still at 5.9-6.7K.
I wonder if turbulence (from too constricted a system) is something I/we should worry about in a closed system? It's not like there is an opportunity for air to be introduced. The pump appears to be capable of dealing with and overcoming the 8+ PSI (or so) drop in the system I have roughly calculated, if accurate. A 12AN size hose drops pressure loss to around 3 PSI.
Again, I'm getting into an area where I lack the educational background to feel very comfortable in spouting off... so what works for me is purely anecdotal.

[DanielsDM]

Every piping system I have worked on operated in the turbulent region. It is not economical to size components to operate in the laminar flow region (in most cases). You just determine how much pressure drop you can live with and design accordingly.

[Sgt.Gator]

Aeroquip has a page of technical documents. One download is Flow Capacities of Hose Assemblies http://aeroquipperformance.com/file/...Assemblies.pdf

Using the Aeroquip chart and the largest Subaru oil pump flow of 3.5 gallons/min, a .750 -12AN hose is more than large enough, unless I'm missing something.

The rest of the documents are here: http://aeroquipperformance.com/rt-28...echnology.html

[Zach34]

I've run into two issues installing the Cosworth DS.

1) The 720 mm HPD belt is about 10 mm too short. The instructions are very adamant that there should be NO tension on the belt. Heck I can't even get it over the pump pulley at all, much less no tension. I checked into a slightly larger belt but the next size up from three manufacturers is 800 mm which will be way too long. The possibility to use an 800 would be to run the tensioner on the cogged side (inside) of the belt instead of the smooth outside. I suppose that since the tensioner is just a smooth idler pulley exactly like the timing belt idlers and there's really no tension, it just takes up slack, it might be ok, but IDK.

For now I'm trying to stretch the 720 mm belt a half inch using a floor jack with a couple of 3" pipes on each side to not bend the belt over too sharp a corner. It's helped a tiny bit, but not enough yet. See pic of my Colonel Red Racing belt stretcher!

I don't think stretching that belt is a good idea. If you stretch it, the distance between the teeth will increase, but the distance between the grooves on the pulleys does not - which means you will wear out the belt fast and possibly cause it to fail. Seeing it rigged up on your floor jack like that......... I would just throw that belt away now.

I agree with biknman in that I don't think you want to run those belts with much tension on them, either. The only experience I have with them is on Ducatis, and they have non-spring-loaded tensioners that are set to a surprisingly loose tension (the ballpark estimate is when you can twist the belt about 45 degrees with two fingers). Much different than a serpentine belt.

[Sgt.Gator]

I don't think stretching that belt is a good idea. If you stretch it, the distance between the teeth will increase, but the distance between the grooves on the pulleys does not - which means you will wear out the belt fast and possibly cause it to fail. Seeing it rigged up on your floor jack like that......... I would just throw that belt away now.

I agree with biknman in that I don't think you want to run those belts with much tension on them, either. The only experience I have with them is on Ducatis, and they have non-spring-loaded tensioners that are set to a surprisingly loose tension (the ballpark estimate is when you can twist the belt about 45 degrees with two fingers). Much different than a serpentine belt.

I tossed it. It was an interesting experiment though, even using as much pressure as I could it wouldn't stretch.

I did solve the problem though. The Cosworth pulley is 36 teeth. Luckily the pump uses a standard shaft and keyway. So a Peterson 35 tooth HTD gear fit very well. The new belt/pulley is not as loose as I would like, there is still no need for the A/C idler pulley to take up any slack, but I think it will be fine. A 34 tooth pulley would probably be perfect.

[Desertrunner]

Removed
as usual you guys have all the answers.

[flynntuna]

The same reason that the transmission fluid level on the dip stick is off too I suppose.

[Scargo]

Don't know if you guys are aware that the Subaru motor in a 818 sits flatter then in a Subaru car and what it does is give you a false reading on the dip stick. It will show the sump is full when in fact its not.
I tried to find a image of the motor and gearbox in a car so you could see what I mean.

It would be nice if you would post this elsewhere since it does not pertain to the thread.

[Bob_n_Cincy]

One of the main reasons I went for the micro v belt over the toothed belt had to do with the clearance between the harmonic balancer and the firewall. Adding a additional pulley on the front was to tight for me.
Bob
belt.jpg

[Rob T]

I did a partial oil replacement on my dry sump system yesterday. This was accomplished by draining and refilling the separator tank which holds just shy of 5 quarts. I am estimating that I got about half the oil in the system. As I plan to do frequent oil changes, this will probably work fine. I was hoping the forum could offer some additional advice. If I were to disconnect the AN fitting/hose that returns to the tank, could I keep the separator tank full and allow the engine to pull the oil from the separator tank and the scavenge pump to push the oil through all the hoses and coolers, thereby purging all the old oil and replacing it with new? I haven't been able to find out how fast the oil is pumping because if it is too fast, this would be a really bad idea. It may be a really bad idea anyway, but I think the concept works. This was not one of those things I was just going to "give it a go". Any thoughts, data, or other ideas would be welcome. If this is the dumbest idea you've ever heard, be kind, I am learning. Thanks

[Bob_n_Cincy]

Hey Rob,
I think your way is good to get all the used oil flushed out. If your changing to oil often, I don't think it is necessary to change every drop.
Here is a link to the corvette factory dry sump oil change procedure.

http://www.101corvetteprojects.com/2...e-z06-and-zr1/

basically; drain your tank, drain the small amount in the pan, change the filter and refill.
I always fill the filter with oil before screwing it on.
Bob

[FFRSpec72]

I thought that I would document my dry sump install parts, cost was around $3100

1 -Earl's Hose, Pro-Lite 350, BraidedNylon, -12 AN, 20 ft. Length
1 - Moroso Breather Tank, Dry Sump,Round, Aluminum, Natural, 1 qt. Capacity
6 - Earl's Fitting, Hose End,Swivel-Seal, 90 Degree, -12 AN Hose to Female -12 AN, Aluminum, Red/Blue
1-Earl's Fitting, Hose End,Swivel-Seal, Straight, -12 AN Hose to Female -12 AN, Aluminum, Red/Blue
2 - Earl's Fitting, Hose End,Swivel-Seal, 45 Degree, -12 AN Hose to Female -12 AN, Aluminum, Red/Blue
1-Earl's Fitting, Hose End,Swivel-Seal, 30 Degree, -12 AN Hose to Female -12 AN, Aluminum, Red/Blue
1-Element Tuning Dry Sump Kit (no hoses) with 7.5" diameter tank
1-Stock Subaru A/C mount and Idler adjuster
1-GatesK050345 Multi V-Groove Belt (using stock crank pulley)
1-tube Genuine Toyota Fluid 00295-00103 Formed-in-Place Oil Pan Gasket - 3 oz. Tube

[Scargo]

Thanks, Tony! Are you happy with the Earl's fittings and are there any brands you won't use? I've shopped around a bit and had a couple that leaked or broke but, the problem is I don't know their brand...

[FFRSpec72]

Thanks, Tony! Are you happy with the Earl's fittings and are there any brands you won't use? I've shopped around a bit and had a couple that leaked or broke but, the problem is I don't know their brand...

Yes and the hose was great to work with and NO LEAKS !

[Mulry]

I'll have more on this topic sometime later this spring, but after putting my hands on them at PRI last December, we've moved to XRP's ProPlus Extreme hose for the lines that will see heavy suction, for a several reasons. It's a convoluted hose but the serrations are on the outside of the hose, so when you scatter a motor, the hoses are easier to clean out (versus those that are internally convoluted, it can be impossible to get all the metal out of those and I don't like throwing out expensive lines, if possible). And that means a smooth interior bore, so better flow rate. In addition, it has a stainless steel wire wound into it so it has the collapse resistance of a hose that you install a spring into, but it has better flexibility. They have reusable hose ends for these now, so you can build them yourself. And they're less expensive than the comparable Earl's product (which is internally convoluted, to boot). I bought mine through ANPlumbing.com, they don't always list these on their website but they were great to work with on the phone. There was a little misunderstanding on one of the sets of hose ends -- just a miscommunication -- and they fixed it ASAP, even sent me a UPS label to ship back the exchanged ends. Highly recommend them.

[FFRSpec72]

I'll have more on this topic sometime later this spring, but after putting my hands on them at PRI last December, we've moved to XRP's ProPlus Extreme hose for the lines that will see heavy suction, for a several reasons. It's a convoluted hose but the serrations are on the outside of the hose, so when you scatter a motor, the hoses are easier to clean out (versus those that are internally convoluted, it can be impossible to get all the metal out of those and I don't like throwing out expensive lines, if possible). And that means a smooth interior bore, so better flow rate. In addition, it has a stainless steel wire wound into it so it has the collapse resistance of a hose that you install a spring into, but it has better flexibility. They have reusable hose ends for these now, so you can build them yourself. And they're less expensive than the comparable Earl's product (which is internally convoluted, to boot). I bought mine through ANPlumbing.com, they don't always list these on their website but they were great to work with on the phone. There was a little misunderstanding on one of the sets of hose ends -- just a miscommunication -- and they fixed it ASAP, even sent me a UPS label to ship back the exchanged ends. Highly recommend them.

For those of us that are not running a vacuum we don't have to worry about the line collapse but we do have to worry about the ability to clean everything in case of a engine failure. So with a oil cooler, remote oil filter I have a lot of lines and parts to worry about, thanks for reminding me !

[Scargo]

I like the idea that XRP is coming out with a silicone coated version. Easy to keep clean and probably more heat resistant.
How do you know that a normal hose will collapse? I would think that a normal bend in SS braided hose couldn't collapse. I mean, it would have to be a hell of a vacuum, and an undesirably high one to do that, wouldn't it?
Just as an aside, I have successfully used -10AN Aeroquip Startlite Race Hose and another just like it (with a red tracer woven in) for several years without issues for remote filter, cooler and Accusump. The downside is it absorbs fluids into the weave (but so does SS covered hose).

[RetroRacing]

We don't run a ton of vacuum, so we have not seen an issue with hose collapse on the DS side of the system (3 dash 12 lines) using the Kevlar covered lines. Everything else is under pressure.

[Sgt.Gator]

The 2017 BRZ GT300. The most interesting thing is that I see a dry sump tank in the trunk. Apparently they have a solution to DS the FA20 engine.
https://youtu.be/vGyl8PERQco

And the 2017 Nurburgring 24 Hour STI is testing. Just race car goodness!
https://youtu.be/WCoaxhpp4cE

Wish I could understand Japanese !

[Sgt.Gator]

I was asked the distance between the face of the crank pulley and the face of the Cosworth drive mandrel. The answer is 1 7/8" .

[Bob_n_Cincy]

I was asked the distance between the face of the crank pulley and the face of the Cosworth drive mandrel. The answer is 1 7/8" .

Gator,
That would put a dry sump belt pretty close to the frame. I will measure stock pulley to frame.
Bob