I want to make a diffuser/engine pan for my car, and last night I started gluing foam sheets together to make a mold. My plan is to form the foam into the shape I want and then seal it with body filler, and then primer and polish. I have been reading up on making molds for composite work and most seem to talk about making a copmosite piece for the mold rather than what I am planning. If I make my mold as described will my composite part rip the mold apart when I seperate the two? (assuming I use a PVA mold release film).

Also, if I make the part with carbon fiber, how do I finish the edges of the piece when I'm done?

I would like to vacuum bag this part because the part will be visible from the top and the bottom... Can I get away not bagging it and get a decent finish? (I'm not super concerned that it has a perfect surface really).

-Tony

Finished part will look something like this...

With the long straight edge meeting up against the firewall and the two tunnels exiting right under the bumper.

First of all, SWEET! I'm an ME student and aero-junky , so if you pull this off, I want one too!

As for the composite work, I dont know much about making it, just that it can be made in any shape the engineer (me) wants it.

A couple of questions, do the exhaust headers drop down each side or run inside the diffuser then out the back and where will the engine cooling air exit from? If you check Patrick's web site, he has a 914 vintage racer that is "canned" in the rear to smooth out the air flow underneath and I remember him telling me it was done as an experiment and the results were not worth all the work involved. Since the firewall is about the lowest part of the car I would want the diffuser to start there?

Second of all:

WHY, To catch the oil?

BahnBrenner: If my mold lasts I can make another copy.

John: My exhaust sits up almost against the floor of the rear trunk. I did this knowing that I would eventually put in a set of tunnels. The exhaust will exit either through my bumper (a pair of holes, one on each side of the licenes plate) or through the gap between the lower edge of the bumper and this diffuser plate. (which will sit 6-7 inches below the bumper, with the top of the tunnels almost touching the bumper ) So my exhaust headers turn up, right after the engine, which is at the point of maximum clearance from the tunnel roof.

My design has the enitre panel attach to the firewall/floorpan point. So there will be no gap between the diffuser and the floor pan.

Paul: Subaru engines don't leak oil At least not much.

There are a few reasons for doing this. First is an expired reason. I had planned on using this as part of my mid-mounted radiator cooling system, as a way to exhaust the air better. Since I moved the radiataor up front, that is no longer necessary.

Another reason is that without a rear valence, the car looks uncomplete. I don't really want to put a rear valence on and wanted to do something different. It will actually serve to keep a little road grime out of the engien bay area too..

And lastly, I have always wanted to make something out of composites, so this being a relatively simple shape, seemed like a neat first project.

I don't actually expect it will make much difference in terms of driving. First of all I don't drive fast enough, and if I did I would be worried about the front end lifting.

Well I thought you had a stock-type teener, so there's no way I could use what you have.

I still love the idea though I'll be very interested to see what kind of aero you get with it installed. Whether it reduces/increases drag, gives you some downforce, or just sucks itself right off the bottom of the car at 100 mph!

Here's kind of what I anticipate the fitmet to be like, although probably more clearance between the diffuser and the headers.

-Tony

Tony,

Appears you are actually making a part with composite (sandwiched) molds for a single layer CF part. The foam should work fine, I've even seen it done with cheap house insulation blue foam. The guy used cheap wall plaster "mud" and sealed it since the resin would melt it.

I would carve and form the male side, finish it and then make dams to pour liquid urethane to make the female side. You would treat it with a mold release like PVA. Finish that side and be ready to press the resin and CF cloth. The finish will depend on the amount of mold finishing. You can also lay in mylar if the surfaces aren't compound.

I'm not sure but cutting the ends should be like cutting fiberglass. Kevlar is the material that's difficult to cut.

Felix: Not sure about what you mean in terms of what I'm making. The mold that I'm working on should end up looking like this:

Basically its a foam piece (with sealer and primer, etc. on the surface) and then I will layer the cloth on top of that. I will probably use some kind of mat between CF layers to reduce cost, but this large foam piece won't be part of the finished product. The whole foam piece will be glued to a 1/2 inch MDF board. Which in turn will be supoprted by 2x4s. I figure that should help during the vacuum bagging, to prevent mold warpage.

-Tony

I might be easier to make it out of aluminum or stainless sheet metal. I just took a couple of pieces of paper and folded one up quite easily. A sheet metal fabricator could
bend one up I assume very reasonably.
Something to think about.

Chappy's right....it would be easier to make it out of sheet metal as their are no complex curves/hidden angles. Once made out of metal you could use it as the master to make a glass one if you so desired.

Its not to hard to do Tony. If your mould is smooth and blemish free, assuming the foam part you are using is the mould and not a plug, it will be easy to get a good finish. What I do is lay up just resin or epoxy like three times for a good build up to sand later to a glass finish. It should get some acrilic clear on it to keep them UV rays off so it will get a good grit sanding before that also. For me, the hardest thing is to get the CF as thin as possible and as strong as it needs to be. The best time to trim anything is when the stuff gets geled and not yet hardened since you can cut it with a razor knife but you can use a grinder to get the shape you want when its completely hard.

The two part mold would be similar to a die essentially replacing bagging which helps drive out all the air pockets and forms two finished sides. You can accomplish the same with laying the layers on your one piece mold and then a piece of mylar and squeeging out the air although you wouldn't control the thickness.

Structurally you can probably use one layer to take advantage of the CF strength.

Looks like a fun project...

Get ahold of the pan/diffuser for a 964 or 993 and make it as close to that as possible. They only really work at highway speeds (60mph+).

From the reading I've done, it seems that 10-12 degrees for the tunnel is optimum. For a 6 inch high tunnel exit that means 31 inches long. It will probably make a difference at high speed, but it will be far from optimum..


Here's a nice diffuser... (~$1100)

My experience is vacuum bagging 60" composite model sailplane wings panels.
We use mylar for the surface finish which can come out beautiful when doing paint design transfers.(no pinholes)
My suggestion is to bag ALL your individual precut pieces flat, and lay in your spars and predetermined attatchment points within your lay-up.
Then you can assemble the pieces. You may have a hard time getting the bag to conform to your concave 90's and get some flooding. (not the worst place though).
Blue or pink extruded foams work well with the epoxy finishing resins, and there is no sealing requirement, but it won't deal with direct heat well.

Kevin

I've used plaster of paris before, casts smooth, and it can be as massive as you want.

Kwn

First of all, thanks for all the tips! I'm going to put this stuff in a progress thread soon.

So about heat... I am goig to be curing this inside my garage, which means I can't really take the temeprature up above say 100 degrees for maybe an hour before my propane heater runs out. Its my understanding that there are room temperature curing resins. Are they far inferior to heat cured resins or is it a matter of time?

As far as a vacuum pump goes, what are the cheap options? I've seen some people using refigerator compressors.. Can I use my air compressor inlet? How much of a vacuum really needs to be pulled anyway?

-Tony

If you are careful (NO FIRE OR FLAMES) , a cardboard box can be heated well above 100 degrees F. That and more is quite attainable with a hot air gun venting into an enclosed box.

I've been kicking around bilding one for my race car and have collected several dozen pictures. here are a few

mo

llll

just remember that if you make it, and it's cool, everyone will want one!

On some of them the vertical fins are bolt on not molded in. In case of damaged on track and easier to make.

These things are way cool. You have a great idea here, I know Patrick Motor Sports has fashioned underbellies but not one of these .
I could have my buddy bend me up one so you could see how it works and looks, then you can take it from there.
He builds drag bikes and does all my tig welding for me. He also builds drag snowmobile sleds. He is a master with the sheet metal brake.

Some sort of canard fashioned into a spoiler would set it off front to back. Stay tuned...

Paul: Nice pics! There's a web site (unfortunately I don't have the URL handy) called Mulsanne's Corner and has a lot of detailed pictures and aerodynamic discussion of LeMans prorotype cars. There seems to be considerable variation among the number of strakes, the width and number of tunnels etc. The one seemingly constant thing though is the angle of the tunnels and from other reading I have found, 10-12 degrees is the magic number. Beyond that, determinung how many strakes to have etc., is probably something learned from wind tunnel testing or CFD simulations, neither of which I have access to. Without a truly flat bottom and the approriate amount of air getting under the car, its hard to say how much downforce this will produce.


tds: If everyone wants one then I might be able to recoup the mold costs At this point if I go with 3 layers of carbon fiber, its going to be somewhere in the north of $350 (very rough estimate) range just for materials per copy.

-Tony

Yup I want one for lapping days and commuting back and forth to events. Difussers are not legal for my Autocross class so it need to be removable.

Unless you are using a very specialized resin, you are not going to need alot of heat when working with epoxies. They can be temp sensitive, and do have curing ranges. As far as vacuum, I do composite laminates for a living (prosthetics)and in GENERAL 20-25 inches Vacuum gets the job done, and you'll need to sustain it for 16-20 depending on your resin. You can buy small cheap pumps these days.
Check out www.fibreglast.com for your start for supplies/info.

When I first started molding stuff I bought a vacuum pump off Ebay for about $50, it was a refurnished medical unit. They run from that ($50) to very expensive.

So tonight I finished stripping the metallization off the foam sheets and got most of the mold blocked out. The foam I used is just polyisocyanurate home insulation that you can get at Home Depot. I used two or three full 1 inch sheets, and a single 2 inch thick sheet. All of that sits on top of a 1/2 inch MDF board. The next steip as mentioned previously is to strengthen that board with 2x4s for structural rigidity. For now, the foam blocks are partially glued together but they are not glued to the MDF.

-Tony

view from the other side.

Oh boy, don't burn that foam. No hot knife allowed. It will produce cyanide gas. We generally work with blue foam... which can be a little hard to come by in Southern California because it's used more up north. But I think Home Depot at least has like the 1" thick sheets of it.

Your MDF vacuum table sounds good. As far as pumps- an 'adequate' setup is more complicated then you would think. You do want to pull 15-25lbs per square inch... for pretty much a 24 hour period. Most pumps if hooked up directly will have a hard time maintaining a steady pressure and they may burn up before the end. The best home setup I've seen/worked with had a good pump, a propane tank vacuum reservoir- to ease the burden on the pump, and vacuum pressure regulator, and a pressure trigged relay to turn the pump on when vacuum was needed. A good system wont be cheap.

Don't get an old medical pump. You have no idea what it was used for... it probably hasn't been cleaned well enough, and it will spit out diseased exhaust all over your workshop. Definitely not cool.

Mylar will give you a great finish on the top. it comes in different thicknesses. We would use fairly thick stuff (1mm or 2mm?) doing wings... but we didn't go around any tight radiuses. I think a thinner piece that completely pulls down to the shape of the mold would yield good results.

The key to good composites (and the reason for using vacuum) is so that you can squeeze the epoxy out of the part... leaving the glass with the minimal amount of excess epoxy, ie un-wanted weight. Without vacuum bagging you just can't make really lite weight stuff.

As far as your mold plans- it looks pretty good. The mold, which sorta resembles what you would normally call the plug.. anyway... your bottom part that you're gonna lay the glass up on... the finish is everything. You're gonna want to get that surface so polished and glassy- so that you can actually seperate your big flat stuck part off of there. I don't put much faith in PVA. The mold release wax is where the money's at. Put some on. Then put some more on. Then put some more. Get the idea? At the end of the day you'll probably still have to use some plastic spatulas to pry the part off.

As far as fabricating your mold... I would probably fabricate it the same as the top of your part- with mylar. Shape the foam, lay up a few layers of 8 or 10 ounce glass, finish with a layer of 2 ounce, then vacuum bag it with mylar so you get a glassy surface.

Shoot... with how flat your part is.... why don't you just mylar both sides? Build your foam mold to shape... then skip the one week of work polishing and waxing... instead just lay the first piece of mylar over your shaped foam... wet out your clothe on top of that... slap the second piece of mylar on top... then into the vacuum bag. Make the mold... mylar... and glass extend about 2 inches past where you want the part to end.. then just jigsaw your final part to the correct dimensions.

Prepare for a 2-5 times learning curve. In other words don't waste your money on carbon fiber the first go around. Infact, where do you plan on getting carbon fiber in the first place? When the war started carbon fiber was basically rationed- and if you could even get it for a non military application it would be really really expensive. Maybe it's available again, I don't know. But your part doesn't really call for using carbon fiber anyway. You would be better off using S glass cut on the diagonal. Personally I would do one layer of 2 ounce, one layer of 4 ounce on the 45 degree bias, another layer of 4 ounce straight, a third layer of 4 ounce on the bias, and finish with 2 ounce again. You can even paint the mylar.

There are ways to make the part super lite weight... but then the finish on one side of the part would come out textured instead of glassy smooth. It would involve peel-ply and bleeder clothe. Definitely the most hi tech lite weight way to go.

I wouldn't worry too much about baking the part. Heat cured epoxies are another ball game entirely.. and they operate at like 250 degrees or something I think. Instead I would just throw a space heater in your garage over night. Epoxy curing in an 85 degree garage does really well since it was designed to work at 70 degrees room temp.

Anyway- we used to build model sailplanes as well. And some top secret stuff for the F-22. Now I have to kill you.

Tony,

If I were you I would make this as easy as possible and just lay up some fiberglass mat and use some cheap poly resin. You're spending a lot of time on molds, material and vacuuming when you don't know if this is even going to work. Make the first one simple and cheap and once you get it to the shape you what than use that one for your mold and use the carbon.

Everyone here has given great advice but you're jumping in with both feet when maybe you should take smaller steps.

Just my $.02.

+1

I've built a number of one-off's for race vehicles with the cheap and in-expnsive method. Saves many hours of labor, and in the end it gets dinged-up anyway.

A lot depends on the complexity of the shape, but your's seems relatively simple. So, this is how I'd do your part. Construct the simple shape in MDF (positive or negative). Radius the outside corners, and use bondo for the inside radii (use PVC pipe or steel tube of the right size). Knock down the rough areas with sandpaper. If you can get Mylar to lay down flat to your surfaces, then simply do a wet layup with fiberglass/epoxy (cloth and mat combination). BTW, you can buy a two part laminating epoxy (I think they call it "Boat Epoxy") from Home Depot that is plenty adequate for one-off's (and no need to make special trips to a plastics supplier). Finish the top side using a short knap roller to both compress the layers of fiberglass layers together as well as push excess epoxy off the part and discard it. In the event that the Mylar cannot lay flat against your form, then spray several coats of grey primer, and wet sand with 400. Spray with PVA release agent (two coats minimum), then glass over that.

For more complex shapes like ducting, I shape the part from urethane foam and use the same process. Yes, it is un-sophisticated, but produces a part quickly and relatively in-expensively by comparrison. Your alternative of course, is to go with "highways" method.

Andys

Okay, so it sounds like there are a bunch of different ways to do this. First off, yes carbon fiber is expensive (~$60-$80 a yard depending on patern) and I won't start with that. If however my mold is only good for one shot then I might as well go with it, but thats a big risk....

Eventually I would like to use carbon because it looks cool and can be made thinner (lighter) than a glass counterpart. I am not going for the smooth glass finish like you see on vehicle interiors etc., but rather for the minimum epoxy lightweight construction.

So my initial plan was to make my mold as in figure D below. Basically shape the foam as I like, fill in the voids and make a smooth surface with body filler, and then sand some more. On top of that spray a primer, sand, polish and wax. Then coat with PVA and then start making a part. The vacuum bagging would probably not be done on the first article. Instead I would just lay up a few layers of glass, squeegy out extra resin and then let it cure. Then after that is done, pry the part from the mold and repeat, this time with vacuum bagging. And after I get that right, go for carbon.

So Paul's option as I see it is to cover the foam portion of the mold with a layer of fiberglass itself, with poly resin, and then use that part as my mold. I think if I ever want to vacuum bag it, that wouldn't be the preferred way to go, as my mold would now be a rather flexible large piece of fiberglass.

I think the method closest to D is what andy reccomends.

The whole deal with mylar is just that it has a very smooth surface? Last time I remember playing with mylar though, it has a tendency to wrinkle... Am I thinking of a different kind of mylar? Also as I said earlier, surface finish isn't my primary conern, and I really don't actually want a glassy layer of resin, rather I would like minimal weight, which means peel or perf-ply and an abosrbing mat layer during vacuum bagging.. right?

-Tony

http://www.fibreglast.com/

sells instructional DVDs.

Ed aka W9R1

I have also done some composite work on R/C pylon airplanes and had good luck with mylar as a face for the mold. The product I used was an adhesive backed mylar film that had a low temp adhesive. This film was applied with a covering iron. Any hobby shop that sells R/C hobby products will be able to give you more information than I can recall from 5+ years ago. I wish you luck in your project however.
Going this route will save you time with sanding, painting, and polishing your part to get that perfect surface.
I will advise that if you go this route, please practice on scrap first as I don't want to be the one to have caused a melt in your mold. There may be alot of help gained from a KNOWLEDGEABLE person at the shop. The material comes in 3 ft. wide rolls x 30? and the fact that the adhesive is heat activated will help avoid wrinkles as you iron and stratch the material over your plug. The heated adhesive will seal edges down so you could cover 1 section at a time.

Thanks for the tip! I've got to be careful with heating this mold. If I get it to the point that the foam melts it will release cyanide gas... Not something I wish to breathe too much

-Tony

if you do a search and find the sites of Coverite, Monocote, they may have spec sheets listing melt and shrink temps. I'd look for you but where I am our server locks out most sites.