Building A Carbon Fiber Splitter For A Mitsubishi Evo

One of the benefits of being a university professor is that I get to be involved in interesting projects outside the classroom all the time (at least when my hectic work schedule will allow for it). Many of the projects I get to dabble in come about because of my connection with students from class. So when one of my past students, Josh McGuckin (of Camera Courage fame), told me that his friend needed some help with his race car, my interest was immediately piqued. I have a hard time saying "no" to car projects--for better or for worse. In this particular instance, Josh's friend, Dave Kern, needed a new splitter for the front of his highly modified Mitsubishi Evo that he would be racing up Pike's Peak during the summer. His old splitter had been damaged when he hadn't been able to adequately negotiate a high-speed turn during a race the previous year, and he wanted something a bit thinner, lighter, and stronger to replace his old, mangled part. After looking at the way the original one was built, I figured a rebuild of this part wouldn't be a hard task to pull off; the old splitter was basically constructed as a low-cost/low-performance composite sandwich construction with a single skin of carbon composite over either side of a plywood and a open-celled polypropylene core. It was plenty heavy and lacked the stiffness that is really necessary to make an effective aerodynamic control surface. I knew I could make a splitter with significantly more stiffness than the original part, and even include all of the custom features he needed for the shaping and performance of his particular car. So, I said yes to the design challenge...and this blog entry will give the quick-n-dirty low-down of how I made the part. (Readers will note that some of the build details have been left out of this "how-to". The reason for this is that I am held under certain contractual non-compete constraints because of the books I have written on composites; I simply can't duplicate/publish those things I've previously written about for my publisher. For more info on the details specifically related to techniques shown in this build, though, check out my book Composites Fabrication Handbook #2.)

I first started out with a cardboard mockup of the part. I cut and taped some corrugated cardboard to make a representative piece that fit under the actual car as closely as possible. It included references for all the body lines and connection points to the frame, as well as the shape of the ducts that would direct airflow around brakes (as shown below). This mockup process took about 3 hours, but this included consultation and drive time to the shop where the car was, so the build of the mockup itself was actually very quick:

After examining the existing part and the bottom of Dave's car, I could tell that the trickiest part of creating this splitter would be fabricating the ducts, since the rest of the part was completely flat on the bottom. Flat parts are simple to fabricate in composites because flat sheet metal or plastic can be used to make a mold for the part with relative ease. More complexly shaped parts usually require much more planning in the mold-making stage. One of the easiest ways to fabricate a section of the mold for something like these ducts is to cut out wood and then simply cover it in formed sheet metal. These pieces can then be attached to a sheet of acrylic plastic which acts as a base for the mold. (For this mold, I opted for 3/16" thick acrylic that would be fastened over plywood because it is already very smooth and easy to polish and wax...plus I figured it would be easier to shape into other future mold components after using it for this project. Acrylic is not recommended for molds where polyester or vinylester resins will be used...the styrene in these resins will actually attack the acrylic!) I stacked up some scrap pieces of plywood and MDF (making sure to get just enough height in the stack for the needed dimensions of the duct) and used wood glue to join them all together. Next, I cut the stacked wood pieces on a table saw in successive passes, raising the height of the saw blade with each cut as needed until I was able to complete the contour needed for the top of the duct. I followed this cutting step with some sanding using a large spindle sander to smooth out the rough steps in the surface of the wood piece. I then cut the sides of the duct with a bandsaw and sanded it to a smooth final shape. Here is a side view of one of the pieces...

...and here is the top view of the same piece (shown below). It wasn't pretty, but it worked like a charm!

To create a quick, non-porous, mold-ready surface, I cut some 24 gauge steel sheet with some shears (dressing out the edges with a file) and shaped it to fit over the wood piece:

I then covered the edges of the sheet metal and wood with the type of aluminum tape that is made for HVAC repairs (as shown below). This is a quick and effective way to get the edges completely sealed prior to molding over the form.

I should probably explain here that it was necessary to make a solid wood support under the sheet metal because the composite layup was to be vacuum-bagged over these mold components; even ambient air pressure would have crushed a hollow sheet metal construction like an empty pop can under a Mack truck, so support underneath this feature was critical. 

I used drywall screws to attach the acrylic sheet to a sheet of construction grade OSB (for added rigidity). When fastening acrylic to anything, always make sure to first create clearance-holes for the screws; acrylic cracks very easily under stress, so driving screws through an improperly sized hole in acrylic will quickly fracture it. I then adhered the duct forms to the acrylic base using double-sided mastic tape, and then sealed the edges of the forms to the base with modeling clay. Note that a couple areas of the modeling clay had been strategically removed at the edges of the duct forms to accommodate flat areas on the part where hardware would be used to mount the splitter to the car (as shown below). Fabrication of this mold for the splitter took only about 4 hours. 

After a nice, generous waxing (which took about 2 hours), I performed the layup: one ply of plain-weave 3K carbon fabric topped with a ply of +/- 45 stitched unidirectional (non-woven) cloth for enhanced torsional strength. These were both wetted out in succession using epoxy resin and a squeegee:

I added a layer of peel ply fabric and then applied a vacuum bag over the part and allowed the resin to fully cure (overnight):

After cure, I removed the bagging materials, inspected the part...

...and prepped the surface for the rest of the layup by carefully grinding down any resin wrinkles left on the surface of the part by the bagging materials:

Next, I used some craft paper to make patterns for the core materials that I planned to include in the second half of the layup. Since the part was symmetrical, I only needed to make half of the patterns and then double up on the cuts from the core material stock.

I marked and cut out the PVC foam core material based on the shapes of the patterns, paying close attention to  the placement of the patterns so I could use the core material as wisely as possible (because it ain't cheap):

I then "dry-fit" them over the lower half of the splitter that I'd previously formed. I beveled the outer edges of the core and scored the bottom side of the core pieces that would be placed over the curved faces of the ducts (since scoring the foam core helps it better conform to curved surfaces):

Next, I used epoxy filler to tack down and gap-fill the core to the bottom half of the splitter (as shown below). Baltic birch plywood worked well as an effective, lightweight compression resistant core in the areas where mounting hardware would be later used to bolt the splitter to the car:

I then laid up the second half of the composite sandwich (which was a reverse of the previous layup) and bagged the part like before (see the photo below). Layup of both sides of the part and prep of the PVC foam core took about 16 hours.

I then removed all the bagging materials from the part to reveal the nicely consolidated composite beneath:

Here is a view of the cured part, straight out of the vacuum bag:

I demolded the part to reveal its smooth, mold-facing bottom side...

...and then used a scraper to remove the modeling clay (as shown). A little naphtha and a rag cleans up any left over residue quite well. 

I then used the original cardboard pattern and an abrasive cut-off wheel on a rotary air tool to trim the edges of the part. (Note that the photo below shows the location of the mounting hardware, drawn onto the top of the part using a silver-ink Sharpie marker. These marks helped in aligning the part to the mounting locations on the frame.)

Here's a photo of the completely trimmed part, ready for final holes to be drilled for mounting to the car. Final demolding, part cleanup and trimming took about 4 hours.

The final part was significantly stiffer, thinner, and lighter than the original. After about 26 hours of work, the new splitter was ready for the race! Here's a little photo of the finished car peeling around the turns up Pike's Peak (notice the flames coming from the exhaust in the front!) with the new splitter mounted just below the front bumper:

With this little baby mounted to his car, Dave ended up getting second place in his class and fourth place overall. Not bad for any Pike's Peak racing machine! Maybe I can further help him get a first-place winning vehicle all dialed in for next year...if my day job doesn't kill me first.

As demonstrated here, high quality parts can still be fabricated in composites using relatively inexpensive mold materials and techniques. All that's needed is a little ingenuity and know-how (both of which come with practice). Now go build something for yourself!