Dry Sump Status
 

Here are some pics of our (marc and tom) dry sump project pan roughed out...a bit more machining on the pickups and then some finishing touches and ready for install. But still a ton more work!

It is designed to allow individuals to mount different baffle configurations by welding.

The plate started "solid" and is now mostly chips!

We are making 4 more after this one.

Timing...soon :)

Marc

Codename: Machupichu :)

Looks killer.

Looks interesting,

any pictures from other angles to share?
What will the cost be?

/Peter

It's not clear why you would choose such an expensive method to make this part -- why not weld it up out of plate?

CNC rules. You design it, then you cut it to tolerances that welders cannot fanthom.

this falls under one of those, "if you have to ask" catagories.

Cost will be about $1500 per pan. Once we test the first one, the remaining set will go really fast!!

Welding is not precise enough and the pan will end up warped and is suspect to cracking. The cost to machine, cut and weld up a pan is not much different in cost. Also, the oil pickup needs to be very precise and our method will ensure a leak free system, just like stock!!

Cheers,
Marc

Sorry, I don't buy that. When you cut that much material off of a block of aluminum, it moves all over the place as you relieve stress in the metal. I would be very surprised if the gasket surface on that part is any less warped than it would be if you welded it. In either case, if the warpage is beyond what you can take up by tightening the pan bolts, a cleanup pass will have to be made to flatten the gasket surface.

:roflmao: :roflmao: :roflmao: :roflmao:

You should check out a machining shop sometime to get up to speed on how it's done. Aluminum is soft stuff that does not have hardly any stress cast into it in billet form. When cutting it, a continous stream of coolant (which cools and lubricates) is sprayed directly at the cutting tool as it smoothly converts block to chips. These chips actually take the brunt of the heat and they are flung away at high speeds, usually landing on coworker's nearby toolboxes if the CNC machine lacks an enclosed shroud. At the correct cutter RPM and CNC feedrate, the heat soak to the workpiece is minimal and warpage non-existent.

I used to do this 8 to 14 hours a day, making checking fixtures within .050mm (0.002 inches) - with a cheap little off-brand CNC. When the program was done, I picked up the blocks with bare hands - warm but not hot. The only time I warped something was when I was cutting a thin plate and my coolant flow got choked by chips in the resevoir tank while I was eating lunch.

I'm 110% sure that Marc's beautiful dry sump is not warped even on a molecular level.

Yes, a machined block of aluminum is the best way to make the pan. The pan on a 928 is very shallow and not much more than a flat plate is required for a dry sump. If it were very deep than it would be impractical to machine a massive block and then welding sheets would be practical. Aluminum is very stable after being machined and will be far flatter than a weldment. I'm planning on doing the same thing in my shop, one of the things I look forward to with the method is knowing that the mounting surface will be the flattest possible.

Well, it wouldn't be the first time I was wrong. :D But, during the 1986-1995 part of my career where I was responsible for having many different tooling plates and fixtures made up for various purposes, this issue had been raised over and over again by machinists. One specific example I recall was a 1/2" cast aluminum plate, ~12" x ~20" had to be taken down to 10mm thickness. I was told by the machinist (later confirmed by our in-house mechanical engineer) that the plate had to be flipped repeatedly and light cuts made to keep the flatness within the .002" spec. Were these guys full of it or has machining technology/metallurgy changed that much in 10 years?

If that plate were stainless steel you would have significant warping from the machining forces and taking material away in steps like that would be helpful, I do that all the time with stainless steels. Most aluminum used today is extruded bar, usually 6061 alloy. I find it the most stable metal I work with as far as changing shape from machining, other aluminum alloys like high strength 7075 seem to be very stable as well. There must have been something funky with that cast plate in your experience and the whole idea that a particular peice of material can be funky is one of those things that keep us machinists from getting bored ;-)

Edit: more machining trivia: hmm. .002 over 12x20 10mm thick is tricky. The hardest part is holding on to the part without the clamping warping things. Two sided tape can come in handy for that and it can also be ground on a surface grinder for the last passes, a corse open wheel with some tap magic for aluminum will keep the wheel from loading up and the part will look nice.

Interesting experience! 0.002" flatness over a 10mm thin 12" x 20" plate is actually pretty tight tolerances. The 1/2" cast stock was probably not anywhere NEAR .002" flat from the vendor, doubly so if the flatness has to be on both sides. So that's why they had to flip it and make light cuts. Perhaps they didn't have very cutting-edge cutting tools either, modern coatings have made a world of difference.

Mike and JEC are right, most aluminum can be relied upon to stay flat IF you cut it right... we make a lot of large flat plates, like bulkheads, with many pockets in them, and floor thickness as little as .030". and they stay flat within a .002- .005". As long as the cutter is very sharp, little stress is left behind. But, a slightly dull cutter rubbing across the surface will turn the part into a potato chip...
Marc's pan has relatively large section thickness, so warpage should not be a problem.

I wonder if I could design that in Solidworks. Or even Emachine shop. Making ten or more of them should be cheaper then 1500.

I don't think you are allowed to use the word cheap in the same sentence as dry sump.

I could have used "more inexpensive"

OK, thanks for your input guys... a have received more info now from a source that I trust, and I see now that I was mistaken. Advances in cutter geometries & coatings, combined with other advances in CNC controls, spindle rgidity, etc. etc. have changed things. Aside from the ridiculous "molecular level" comment, you guys were right and I was wrong. Learn something new every day! :cheers:

Oh, and yes the 0.002" was very tight for such a large, thin plate. Fortunately the spec also called for holes for 6 countersunk screws on either side so the plate could be mounted either side up on the robot where it was to be used; these were used to hold the workpiece during machining.

:biggulp: :biggulp: :biggulp: :biggulp: Ok I have come to my senses. I got a bit sideways there, these things happen on Friday. :typing:

I love this forum for sharing interesting technical stuff.

Can't wait for the acid test of the Devek dry-sump system!

Pretty!
 

Precision of the sealing surface will be determined by how heavy the final cuts are...

If you cut 20 or 30 thou then deflection will be sever... but it your last cuts are only taking 1 or 2 thou then it should be great...

I imagine that Marc is making the prototypes as perfect as possible as the extra time to get to the final tolerances is negligable with CNC...

The time required in precision machining is all about setup... 90% of the billet can be blasted away in the first 20% of the cuttng process...

The cheapest approach for a final product will be to cast a rough blank that has some handling poimts (reference datum) and then machine all of the prceision features onto it... I think that cast aluminum is nothing like cast iron when it comes to machinabililty... hehe

Looks Great Marc. Maybe someday I will have to get one!

Loren

This is a great project! Thanks!

Harvey

Pan update
 

I just came from the shop where the dry sump pans are being machined. I watched the final machining being done to the rough outside shown in the pic Marc posted. The cutter stepped horizontally 0.01" to start each new pass as it ran up and down ramps and rounded corners, producing a very nice finish. We have discussed the possibility that the gasket flange may not end up as flat as we'd like, and if that's the case, we'll make certain it is flat before it leaves the shop!

The pan is just the beginning of the $$$$ outlay! You'll need a scavenge pump (at least 3-stage), a remote tank, some hoses and AN fittings, mounts for the pump and tank as well as drive sprockets and belt. I must have forgotten a thing or two!

In anticipation of building a dry sump system, I bought a 10 qt. tank on eBay a few years ago. This is a large tank--probably 10" in dia by 16" tall. Too large. The only reasonable place for it is in the rear seat area. That makes for overly long hoses and I will probably look for something smaller that will fit somewhere up front.

This first pan will probably appear on my car in about 3 mos. I wish it would just appear!

Cheers,

Tom Cloutier

What I got for Christmas
 

I picked up the first finished pan on Sat. and it is simply BEAUTIFUL! Much too nice to hide away under an engine where it will be subjected to the insults of sandblasting, oil, grime, etc. I couldn't talk my normally cooperative wife into displaying it as sculpture in the living room, so I guess I have little choice but to mount it on an engine!

The new pan weighs about 12lbs as compared to 19 for the stock unit, but scavenge pump, tank, etc will make for a slight weight increase. Marc T. has said that as much as 7hp per liter can be reclaimed with a good dry sump lubrication system, not to mention the near total elimination of oil pick up problems during spirited driving. I'd say these advantages far outweigh a little additional weight.

Also the new pan is much stiffer than the original as the sealing flange is wider and about 1/4" thicker and it's made from billet 6061 aluminum rather than cast material. Early concerns about the machining process leading to warping at the sealing surface did not materialize; the flange is very flat.

What's next? Testing, testing, testing. Yahoo! I've been a dry sump fan since the purchase in 1970 of a brand spanking new Norton Commando 750. It's due for some rehab work and may sprout fuel injectors, but I digress and that project is a long way off.

Happy New Year!

Tom Cloutier

Looks nice. Good luck with the testing.

Looks incredible, Tom! Good luck, and keep us posted! :cheers:

Tom Cloutier said "In anticipation of building a dry sump system, I bought a 10 qt. tank on eBay a few years ago. This is a large tank--probably 10" in dia by 16" tall. Too large. The only reasonable place for it is in the rear seat area. That makes for overly long hoses and I will probably look for something smaller that will fit somewhere up front."

Tom, a while ago now I had the fortune of doing some laps at Nurburgring (as a passenger) in a genuine 924 Carrrera. It had dry sump oil reservoir mounted in the rear in the spare wheel well. If that's worked well for a factory 924 Carrera that has done 30,000 quick kms around Nurburgring, wouldn't a similar location also work in a 928? Locating the oil reservoir atop the battery box (in which case it'd have to be removable to access the battery when required) would seem to provide for enough vertical height for such an oil reservoir. Another option might be to relocate the battery into the spare wheel space proper, and use the additional depth of the battery box to provide for the oil reservoir. (though the battery location may be critical to overcoming central shaft harmonic problems, so more investigation about that aspect might well be warranted). Just my comments, for whab they're worth. Good luck with this.

Dave O
Dubai

Beautiful!

Are there enough threads for the fittings?

When will mass production of a dry sump kit for the racer start?
Looks great :thumbup:

/Peter

:thumbup:

thats properly nice. what have you done with the oil pump feed? i can see the hole in the inside of the pan and im just guessing it will be a press fit to whatever you do with the existing oil pump supply..

Tom:

'70 Norton Commando 750! Ah, yes, I had a '71 for many years. It was a torque beast. My '75 850, owned since new, needs to get out more.

The pan looks stellar. :thumbup:

The oil pump feed is well designed to be leak free. Many designes were reviewed, but the requirement for leak free and ease of assy drove our decisions.

We will most likely only sell the pan and the pickup, as people have pump and other part preferences.

Location of the actual sump is up to the designer of the car.....polar moment vs weight distribution is very debatable topic.

Cheers,
Marc

think you can keep the stock accessories and make room for the pump/maybe kinda like the Centrifical blown cars do?? ever make anymore of these??

Back from the dead - did this go anywhere past prototype?

uh oh - did it start snowing again up there in the cold country/looking for a winter project already?

Funny you should say that. My neighborhood is full of cottonwood trees.
All weekend it looked like it was snowing, usually lasts a couple of days. During this time work in the garage is out since having the garage door open any longer than necessary will fill the garage with the white crap from the trees.

Bump

Why spend the money on a CNC'd pan? So many dry sump pans are welded aluminum to a standard flange. Just have a water jet cut mounting flange and weld the pan together from that. A bit more dirty labor, but much cheaper.

Plenty of PACE drysump kits out and about now guys

Do PACE sell them directly? (Couldn't see any 928 info on their website)
Or are these the systems that Gantspeed sells?

They will sell direct to you, i got all my parts from them bar one part

I think its neil you need to speak to at pace

I seriously considered this one for my track 928...ended up going with a accusump....

Pace products make the dry sump kit or bits for gnatspeed, if you were to buy from pace they will allow you to buy bits at a time gnat will not.

Great, thanks for the info.... :thumbup: