Through-the-hood cooling scheme!

[aston@ultrasw.c]

I recently installed a 95 993 motor in my 86 with a shortened G50/01 trans.

Conversion went pretty well: no major surprises.

This is the story of the cooling scheme. It will be in a few installments as time allows.

So I took the new beast motor for its' first track day and oil temps quickly climbed to 250F. Not good. Ambient was around 100, but that's no excuse, it's always 'kin hot in the desert.

The old 3.2 never got above 210F no matter what. The cooling system had changed of course, the engine mounted cooler had gone and more power means more heat. I was relying on a front mounted Setrab 172 and the Carrera fender cooler.



So I purchased a Setrab 660 which has about 50% more cooling capacity.

First I just stuck the new cooler out in the airflow to test the concept. Temps stayed below 215, on the track on a very hot day. Now all I had to do was integrate it into the body shell.



I had always hankered after a through the hood cooling duct, so this seemed like a good time. I studied the Elephant Racing hood cooling design, for the tenth time, and started cutting.

First I had to remove the under the car duct and make a shroud to support the cooler.

The general idea:



Coming up: cut, weld, repeat!

[jittsl]

Cool!!!!

[RWebb]

will be more cool if you have a place for all the air to go - there's a thread here on cutting ducts thru the trunk area

could be Chad Plavan??

[wildcat077]

Subscribed to this thread

[aston@ultrasw.c]

Originally Posted by RWebb

will be more cool if you have a place for all the air to go - there's a thread here on cutting ducts thru the trunk area

That's the plan

[ShakinJoe]

Gonna love this!

[jittsl]

Originally Posted by RWebb

will be more cool if you have a place for all the air to go - ?

I already assumed that's what the blue tape on the hood was indicating.

[aston@ultrasw.c]

In case you haven't seen it, here's some very useful background from Elephant Racing

ELEPHANT RACING Hood Vented Oil Cooler Project

[DaddyGlenn]

This is going to be fun.

[$ Short]

Chris
Do you ever rest? Keep working and we look forward to see your car in November
Kurt

[Walt Fricke]

If you are DE only, or race in a class with no rules restricting what you do with the bodywork, through the hood is a good way to go.

It is worth giving some thought to ducting. The opening in the front valance should be smaller than the area of the cooler. You want to slow down the air as it passes through the cooler. The opening in the hood should be slightly larger than the opening in the valance. This is to speed the air back up as it exits, so it more smoothly blends with the air flowing over the hood, like entering a freeway. Slightly larger accounts for the fact that the air will have slowed some passing through the cooler, so it needs more help.

Carroll Smith's Tune to Win is required reading for those designing their own ducting systems. He suggests that the front opening be 60% of the cooler area, but qualifies that with a suggestion that you make the test bed adjustable, etc. Most of us aren't going to do that. And I didn't find the part about the size of the exit opening, so maybe I have that wrong - backward as to whether the exit should be a bit larger or smaller than the entrance. He does note that applying a Gurney lip on the top of the frontmost part of the hood opening will help get the air out. You see this on some cars configured this way. On mine, I have more than enough cooling as it is, so haven't tried that.

I also noticed that the structure of the 911 hood lent itself to making a kind of NACA duct (in reverse) to help the air out, so I made it narrower.

Smith stresses a smooth flow of air. However, with our 911s we are stuck with some changes of direction larger than what might be optimum. We've got a fuel tank in there, for instance. A stock tank may be less of a problem than the rectangular 15 gallon fuel cell I have, which is 10 inches high. But we can still insure that the ducts are as smooth as possible.




Be sure to run the hot oil first through your big aftermarket cooler before it runs through the stock cooler. If both coolers were equally effective, you'd want to run them in parallel. But they aren't, so in a serial configuration you want the best cooler first. The greater the temperature differential, the more efficient the cooler, so you want your best cooler to have the hottest oil.

But maybe you are an aeronautical engineer with a strong thermodynamics background, and know all this stuff and a lot more. Only thing so far I wonder about is why make two hood openings when you have a single cooler? A single opening ought to be easier to fabricate, and present a smoother flow to the exiting air.

[Flieger]

Originally Posted by Walt Fricke

The opening in the front valance should be smaller than the area of the cooler. You want to slow down the air as it passes through the cooler. The opening in the hood should be slightly larger than the opening in the valance. This is to speed the air back up as it exits, so it more smoothly blends with the air flowing over the hood, like entering a freeway. Slightly larger accounts for the fact that the air will have slowed some passing through the cooler, so it needs more help.

It depends on whether you assume air is incompressible or compressible. The constant-density assumption is good for up to about 1/3 of the speed of sound or about 250mph at room temperature. As you increase the temperature the speed of sound increases. Decrease the temp (like at altitude) and it decreases. If you have compressible flow (trans sonic, sonic, or supersonic) then air speeds up as the cross sectional area of the duct expands. Hence why rocket engine nozzles are bigger at the back but water hose nozzles are smaller at the end.

This is worth a read:

http://www.homebuiltairplanes.com/forums/aircraft-design-aerodynamics-new-technology/15614-debunking-meredith-effect.html

So, no real useful drag reduction at your speeds. Just make sure to keep the bends/expansions/contractions, etc. to a minimum but maybe add some vortex generators at the front of the car to add turbulence and enhance heat transfer. What you are going for is cooling without adding the lift that would come from exhausting the air under the car.

[Flieger]

Pressure must be less at the back than at the front for the flow to go out the back, but the thrust (or drag reduction) comes from the increased speed (momentum change) of the air coming out the back compared to the same mass flow rate coming in the front.

Mr. Meridith's Mysterious Effect

[boba]

Similar configuration, 3.2ss MFI slide valve twin plug with no engine cooler (993 filter console). I went with a B&B front cooler, vented under the car. Shrouded to force air flow through the cooler.

[aston@ultrasw.c]

Without a wind tunnel and CFD analysis it was always going to be a bit of a swag.

I followed the smaller inlet, slightly bigger at the cooler and much bigger at the exit, guidelines.

Also the hottest oil goes to the bottom of the best cooler.

[aston@ultrasw.c]

OK folks here's E2/S1

Mounting is very important as oil coolers are easily cracked with the shock and vibrations of hard driving.

The standard mounting brackets restrain the cooler at 4 discreet points plus they consume precious space above and below the cooler matrix. This is not ideal, so I cut them off, and thought about how to provide uniform support around the entire cooler.

I also knew that I needed a shroud to prevent high pressure air from escaping around the sides of the matrix, so I combined the shroud with the mounting scheme.

It would have been easy (er) to simply weld the shroud on all 4 sides but how would I get the cooler out if it needed? So I welded 3 sides and fabricated a bolt-in closure for the forth side. The shroud has a channel added to actually retain the cooler matrix lined with a soft rubber molding.

The lower shroud section was made of heavier steel sheet the resist the inevitable off-track excursion.

BTW adding the splitter to the original cooler setup dropped temperatures significantly. It helped direct the high pressure “bow wave” through the opening in the valance instead of allowing it to escape underneath the car.

[aston@ultrasw.c]

As the more observant have already noticed I had an eye on nostril hood vents after the GT40.

Initially, I looked around for a FG nostril panel that I could graft onto the 911 hood.

A fellow even brought a 40 to one of our track days....



aaand I saw a few at the Silverstone historics



...but I was no closer to finding a panel.

Then I came across VRaptor Speedworks

Gen I Hood Louvers

- who make all kinds of blingy parts for mega cars. So for $130 I thought I would try to install aluminum louvers on my steel hood.




When the louvers arrived, I have to say I was very impressed.

They are water jet or CNC nibbled from a single sheet of 090 alum sheet. The louvers just tilt back and they are locked in place by a couple of stringers underneath.

Actually fitting the panels to a 911 hood was another matter, but that's a story for another day.

[Quicksilver]

Walter as usual brings good engineering info. To add to the confusing fray...

The outlet of a duct is generally more important and more difficult to get right then the inlet. The inlet has the advantage of being on a forward facing surface which is a naturally high pressure area.
An "on the hood" outlet has the inherent issue that it is variable pressure location. (Under the car behind the airdam is a low pressure location) The farther back you go the higher the pressure. Remember the L-88 hood scoop that faced the windshield?

If I was hacking a car apart to build a major duct I would build a multiple tube pressure gauge with a bunch of clear tubing loops filled with liquid so you can measure a bunch of locations at once. 8 to 10 sections leading from tubes taped at different points up the hood would give you a serious quantity of data to make a real informed decision with.
(aero is one of the most common places where "That looks good" is just plain wrong.)

[Walt Fricke]

A couple of more pictures.

[Flieger]

997 RSR puts it far forward and the air flows at a much steeper angle upward.



BMW puts them further back and a shallower angle


Ferrari goes sort of all the way down the hood