Dry Sump Designs

So, I mocked up a VR sump gasket which ended up looking like this:

[You must be registered and logged in to see this image.]


And then used it as a base for a *loose* model of the OEM pan:

[You must be registered and logged in to see this image.]


And then I started playing with ideas:

[You must be registered and logged in to see this image.]


(Sorry they're both upside down. It's the way I drew them )

The roundness of the pan helps channel the oil back to the scavenging points (on the flat face, not yet cut into model) and the flat edge should stop oil moving around too much whilst it is waiting to be drawn through the separator and then pure oil pumped back to the tank. I'm looking at approximately 2cc/rev coming off the crank in "drips" which as you can work out is about 10litres a minute at 5000rpm. So the pump willl be pushing the oil around way above that (a redline of 8000rpm would obviously be 16litres per min oil flow)


I've still got some more designs to come. A plain triangular one, triangle with a ditch so the oil can't run back out, W shape one, just for the sake of having a few designs really. I'm using the gasket template for the top edge so I can see fitment would be like. They all start from there.

I was working on the pump design but spur gears and calculation of involute curves was cooking my brain, so I've shifted to the pan for now. I will also be doing the air/oil separator and a basic oil tank which vents to atmosphere.

All thoughts, comments, questions welcome

Oh, also I need to add in baffles or walls between cylinders to reduce pumping losses in the crankcase, but for now it's just conceptual designs.

New design #2:

[You must be registered and logged in to see this image.]

Simpler than the round one, but perhaps less efficient.

any thoughts anyone?

So, does the pump have to be driven from the crank? And if not do you have to still vary the speed of the pump with respect to the engine speed? Just thinking pump run on when the engine stops might be useful, especially if running a turbot.

mrbeige wrote:So, does the pump have to be driven from the crank? And if not do you have to still vary the speed of the pump with respect to the engine speed? Just thinking pump run on when the engine stops might be useful, especially if running a turbot.

The current pump is gives a variable quantity of oil based on engine speed but as they are spur gear pumps they provide high flow with low pressure, its only restrictions in the oil ways that give you the pressure.

Looking at the VR self study guide the oil high pressure warning is at 1.4b with the PRV in the oil pump operating at 5.5b.

You could use the existing oil pump and add an extra electric oil pump in the dry sump system that is triggered at the same time as the after fan run on - although this may give you cavitation issues.

Sorry for the delay :/

Yeah, my design is based plainly on a pump running off the crank pulley and mounted just under where you'd find an alternator (except perhaps on yan's G60 valver!)

I'm basing it on a cartridge system so you have the main pump section, and then next to it a separate cartridge for each scavenge pump line, and then at the end of those the air-oil separator before it's all pumped back through the giant header tank/catch can and nice pure oil is ready for the pump again. the separator has a cartridge of its own too. See this pic for ideas:

[You must be registered and logged in to see this image.]


I've run into major problems trying to get the involute curves for the gear teeth profile done - about 3 weeks behind schedule :/

the separator is coming on ok, but it's just so hard to kinda understand how it uses centrifugal/centripetal forces to separate the fluids (so to speak). I always think some air must sneak through, my lecturer (Geoff Goddard of Cosworth fame) assures me the oil will be 0.5% air max after a good separation process. I guess this is what breathers and good oil can design are for too, to complete the system.

One last thing - I don't really see how I'm supposed to supply the main engine gallery with oil from an external oil pump. Do I use a modified pickup which would enter through the side of the dry sump and connect to the original gallery line that the OEM pump would use? I will check with my lecturer also.

Yan, any chance of a copy of that guide? I can swap you a technical paper on G laders from the SAE

Cheers,

Jon.

I've mocked up a pump now, in the same style as the one above. One gearshaft driven off pulley (via crank pulley) and the second just driven by first gearshaft. It's looking good...!

[You must be registered and logged in to see this image.]

Still need to do bearings, circlips, the 4 bolts that hold all the cartridges/sections together and the O-rings between sections also. I have modelled the gears and shafts out of steel, and the rest out of aluminium, with the main flow sections being blue anodised. The picture is actually 90deg off, as the in and out ports should be on each side, not the top and bottom. I keep drawing stuff on the wrong planes in CAD! :/ but it's cool, I can rotate the assembly around. I'll do an exploded one now so you can see what the innards look like (similar to above).

Also thinking about making the ports a large diameter. IIRC they're 16mm now.

Jon

boost panda wrote:One last thing - I don't really see how I'm supposed to supply the main engine gallery with oil from an external oil pump. Do I use a modified pickup which would enter through the side of the dry sump and connect to the original gallery line that the OEM pump would use? I will check with my lecturer also.

Just to answer my own question, I think I would probably go in through the oil filter plate, and replace the oil filter. It's a possible idea. I just don't see how the oil gets into the main gallery. I know an OEM pump uses its own pickup tube and pushes oil up through the bottom of the block,but that will be covered with dry sump goodness...

[You must be registered and logged in to see this image.]

As the pump is in sections, I obviously can't use a shaft which has gears fixed to it. I'm thinking about using little keys here (cutting a little groove into the interior of the gear hole, and into the exterior of the shaft and then fixing together with a little piece of metal in a very tight fit (transition perhaps).

Just mocked up the O-ring seal gap on the blue casing (both sides) using a swept cut of 1mm diameter circle, and created the O-ring by doing the opposite; a swept boss following the same path of 1mm diameter circle.

[You must be registered and logged in to see this image.]


And a closeup of the cross-section vertically through the side of the casing:

[You must be registered and logged in to see this image.]

I guess bearings next. But that will take a bit more work and I will also have to find correct bearings for the loads etc and speeds. Will ask lecturer tomorrow.

Any questions?

Anyone interested at all? trying to keep OU alive here...!

LOL, I was just thinking what a lead balloon this thread is

I must admit, dry sumping isn't something that's sparked my interest before as it's not something you'd really bother with on a road car, but it's a nice little project for you to get stuck into!

So the main advantage of dry sumping is to get the engine lower for better C of G and cure oil surging problems on race tracks?

Yeah it is a bit. I'd expect that on CF as people are clueless, but not on the mighty OU!

Well it's my Dissertation Kev. So I thought I would share it with everyone, as it's pretty relevant. I mean change the pan layout and everything else is interchangable.

Yeah, the benefits are:

- lower mounting of engine so better C of G height,
- obsoletes oil surge problems,
- regains up to 10% engine power from crank spinning through air (based on 70hp increase on an F1 engine and 25hp increase on BTCC engine).
- much easier to do oil and filter changes! lol

It really is a bolt on mod too, just swap pans and connect the rest of it up!

Wonder how many horses ae dissipated in your turbo Kev, spinning through all that oil? Depends on viscosity too of course.

The only snag with the use of an F1 engine compared to a road car is the engine revs, at 18.5k the drag of air is making a impact let alone oil

So aside from maybe a bit of a crisper engine response and a couple of ponies as Kev said for a road car it's not hugely useful. You can't lower the engine on a C without a feck load of work and I only managed to get the oil buzzer and light to come on after going round a roundabout 3 times quickly.

So despite the fact that I like reading this sort of stuff the chances of applying it a car are slim.

Now if you came up with a solid reliable separate oil feed for the charger complete with some good science on a more suitable oil to use and an integrated oil cooler I would be very interested

Well maybe we should all think of it as a nice piece of engineering material, whilst i will keep evolving the design. I mean, this is an engineering forum after all.

It's all good guys. Just sitting here silently reading to myself.

Carry on....

Yandards wrote:Now if you came up with a solid reliable separate oil feed for the charger complete with some good science on a more suitable oil to use and an integrated oil cooler I would be very interested

What do you have in mind? My CAD is always hungry I'm always on the lookout for new mini-projects to build up my portfolio, especially as I'm not using the industry standard CAD package. I told Prodigal I would redesign his boost piping. I think you could possibly reroute them to be more efficient. Bends = pressure loss.

I think perhaps a chemist would be better suited to answer the oil question.

But for bespoke / standalone G oiling, what about the company on the same estate as G Werks?

boost panda wrote:Yeah it is a bit. I'd expect that on CF as people are clueless, but not on the mighty OU!

Well it's my Dissertation Kev. So I thought I would share it with everyone, as it's pretty relevant. I mean change the pan layout and everything else is interchangable.

Yeah, the benefits are:

- lower mounting of engine so better C of G height,
- obsoletes oil surge problems,
- regains up to 10% engine power from crank spinning through air (based on 70hp increase on an F1 engine and 25hp increase on BTCC engine).
- much easier to do oil and filter changes! lol

It really is a bolt on mod too, just swap pans and connect the rest of it up!

Wonder how many horses ae dissipated in your turbo Kev, spinning through all that oil? Depends on viscosity too of course.

The Mighty OU Yes I despair at some of things I read on the CF sometimes, I really do!

Crank spinning through air, eh? I thought it was the crank sloshing through oil that sapped power, or do you mean by removing the sump, there's no air space for the crank to spin in?

Well this is the thing with the Turbot, it has an abundance of horses to begin with, so a few escaping the paddock won't really be missed But yeah, I'm using 10W/60 which is pretty thick old stuff, but it's what you need with sloppy tolerance aftermarket pistons....

boost panda wrote:

Yandards wrote:Now if you came up with a solid reliable separate oil feed for the charger complete with some good science on a more suitable oil to use and an integrated oil cooler I would be very interested

What do you have in mind? My CAD is always hungry I'm always on the lookout for new mini-projects to build up my portfolio, especially as I'm not using the industry standard CAD package. I told Prodigal I would redesign his boost piping. I think you could possibly reroute them to be more efficient. Bends = pressure loss.

Actually based on my personal experience with the Nugget the biggest drain on boost on a G60 is the bloody plumbing. When I swapped over to the BBM boost tubes (constant diameter throughout) and coupled with the chargecooler I am making exactly stock boost of 0.7b from my stock charger on a stock pulley. I did not have a boost gauge fitted when I fitted the boost tubes prior to the chargecooler but I did an MFA test and was getting the 0.7b numbers again. As my boost gauge is a T piece from the MFA vac line it is the same source so the data should be reliable.

Crazy D did comment that my G needed higher fueling than he expected for a mostly stock car, I suspect this is down to more CFM and less pressure, after all pressure is simply a factor of restriction..

As for the charger oil feed, a compact pump that uses the existing supply and return lines to keep things simple, this will also regulate the pressure and flow to standard G60 specs, with around a 3b delivery probably electrically driven with a self contained, baffled 1l tank.

kevhaywire wrote: Crank spinning through air, eh? I thought it was the crank sloshing through oil that sapped power, or do you mean by removing the sump, there's no air space for the crank to spin in?

Yeah, it's a parasitic loss as the crank uses energy to shear the boundary layer of the oil every time it revolves, so spinning in air removes that loss.

I'm confused, did i say somewhere that cranks spinning in air are a bad thing? Ahhh, I see now.

The crank spinning through oil on any given engine has been known to use up to 10% of the engine power to spin through the oil, and overcome the shearing forces there. Letting the crank spin through air, by removing the wet sump (whilst still suppling oil to all galleries and bearings) can regain this lost energy. Does that make sense?

There are some gains with emissions also as the crankcase (area in which the crankshaft is located, typically the changing space underneath the pistons) needs to be vented to stop blowby gasses diffusing into the engine oil, reducing its life. Obviously, this also causes the engine oil to last longer (the blowby gas thing and the fact there isn't 10kg of hot steel churning it up all the time!)

hope that helps. I'm not an expert on the matter, but it's good to learn together.....group hug?

Yandards wrote:

Actually based on my personal experience with the Nugget the biggest drain on boost on a G60 is the bloody plumbing. When I swapped over to the BBM boost tubes (constant diameter throughout) and coupled with the chargecooler I am making exactly stock boost of 0.7b from my stock charger on a stock pulley. I did not have a boost gauge fitted when I fitted the boost tubes prior to the chargecooler but I did an MFA test and was getting the 0.7b numbers again. As my boost gauge is a T piece from the MFA vac line it is the same source so the data should be reliable.

Crazy D did comment that my G needed higher fueling than he expected for a mostly stock car, I suspect this is down to more CFM and less pressure, after all pressure is simply a factor of restriction..

As for the charger oil feed, a compact pump that uses the existing supply and return lines to keep things simple, this will also regulate the pressure and flow to standard G60 specs, with around a 3b delivery probably electrically driven with a self contained, baffled 1l tank.

Ah ok, interesting stuff with the G plumbing. What CC do you run on the nugget, and where is it located? How about the 16v?

Yeah I would think that more air flow would be the reason behind your injectors working longer. I imagine the Digifant monitors all that and adjusts fuelling accordingly? I am not too au fait with EFI, even old setups! Mass flow rate of air is much more important than air pressure as you know

The Design spec for the standalone oil system sounds interesting. Electric pump would be a simple affair too. Depending on where you mount the 1l tank, you could make it tall, therefore eliminating any oil surge as the pickup will always be under 30mm of oil for example. It's a crude but effective design, and means the tank can be pretty much just a cylinder with little or no internal modifications like baffle plates.

Despite almost killing myself, I managed to persevere with the evils of involute curves and all the maths involved with calculating gear profiles and I have re-engineered my pump cartridge to be much smaller and flow more oil so I'm really chuffed about that.

It's now a 7-tooth instead of a 10-tooth. The OEM pumps are 7 as they're much more efficient and don't leech power like a 10t one does. I've run a few iterations of motion studies on the gears meshing on the computer and they interact really well and smoothly I have some other equations to look at which allow me to calculate flow rates so I will calculate all of that at some point.

I've been stuck on gear profiles for a few weeks now so yesterday was a major breakthrough Pics to follow for those interested.

Why are the 7-tooth gears more efficient? Is it down to how much tooth is in contact with another at any one time?