Custom Search

Home About Me Projects Composite Material Info Project Submission Contacts

Subaru 2.5rs Carbon Fiber Strut Bar

This page is dedicated to the design and construction of a carbon fiber front strut bar.


The basic idea behind a strut bar is to improve the rigidity of the frame, which in turn reduces body roll and improves handling. The front struts mount to the frame of the car in front of the firewall. Typically a strut bar is attached in the engine bay and extends from one strut mount to the other. The result of this is keeping each strut parallel.


If you were to go look at the majority of strut bars in production, you will notice that they are made of three pieces: Two mounting plates, and a center section that spans the engine bay. These three parts are generally pinned together. This is poor design because pin joints allow for a certain amount of flex. If one strut mount were to flex in such a way that is was raised about the other, the only added constraint from the strut bar would be that of distance. Therefore I decided to design and build a single piece strut bar. This will provide maximum rigidity.

I decided to base the design off of the strut bar that I already had installed in my car. The two "Y" joints will make the construction more difficult but it also offers additional stability. Unlike the circular strut bar shown here, I have decided to use a T shape. The logistics of trying to mold a tube with two "Y" joints are just ridiculous for what I am trying to accomplish. By going with a T shape I can use stock materials to build the plug.

After a long search looking for the correct size T stock, I decided on ordering two 8' 1.5" x .75" x 3/16" pieces. This ran me about $60 with shipping. My plan is to build the spanning section with the T-stock and then use it as a plug.

Once the T stock arrived and I began machining the plug. I was able to use RIT's machine shop for the majority of my machining work, so if you do not have access to a mill, do not stress too much. A compound mitre saw can achieve comparable cuts if you use a carbide blade. Most compound mitre saws are built for cutting wood but they should be able to cut though this thin aluminum stock without much trouble.

The reasoning behind needing a mill or a mitre saw is that a compound cut is required. The T stock needs to bend up and around the air box. For this particular situation, the T stock must bend 6 degrees up and 6 degrees towards the firewall. Because I do not have the capability to physically bend the T stock as one solid piece I will make 2 angled cuts and reweld the T stock into the shape I need. In addition to this, each side will have a support to improve torsional rigidity.

The three pieces you see in these pictures line up perfectly because I cut each piece with a 3 degree cut. Had I left one side flat and cut the other at a 6 degree angle, the side with the 6 degree angle would overlap because of an increased surface area. The pieces were tacked together using a Tig welder. Remember before welding aluminum you have to brush the surface to remove the aluminum oxide surface coating that gives the material its corrosion resistant qualities.

Now to line up the pieces correctly, they were tacked in two places while laying flat on our welding table. This allowed me line up the first angle of the cut. After welding up the first tacks, I then bent the pieces up into place using a 6 degree angle block. Below you can see the finished product after I added the two supports.

I will now stray away from the body of the strut bar and focus on the base plates. The base plates are the sections of the strut bar that physically attach to the body of the car. Because of the geometry of the parts, there are many different forms of manufacturing at my disposal that I could use. One of the options I considered was to lay up a sheet of carbon fiber and then machine the carbon to my desired design. Unfortunately this is a very slow manufacturing method and if I needed to create multiple parts I must use a mold. For the plug I considered using a CNC mill to cut the shape I need out of aluminum, hand machine it, or have the aluminum laser cut. The CNC and laser cutting would be much to expensive for my budget and hand machining would not provide a part with a high enough accuracy. Therefore I opted to use this:

This is a sheet of high density polyethylene aka HDPE. I selected this material because of the low cost of the material, easy machinablity, and fiberglass and epoxy resins do not stick to it. The one issue that I will have with this material is keeping a high dimensional tolerance. Therefore I designed both base plates so that the dimensions are more of a guideline.

I designed the entire strut bar including the base plates using Solidworks. The 3D CAD construction allows me to ensure the nominal dimensions are correct, and to determine critical dimensions and constraints. After building the base plates in Soildworks I created a 1:1 scale detailed drawing. I printed out the detailed drawings and cut out the parts which effectively gave me perfect stencils. I then used a band saw to roughly cut out each base plate. They were both cut out slightly large; to ensure the edges have a smooth finish so the plug will easily be removed from the plug, I filed down any rough surfaces and created a slight draft on the edges. Take a look at the pictures below for some visual representation. 

To create a manageable mold I will need to place the polyethylene cut outs on flat surface. For this I decided to use more polyethylene because of its resistance to sticking to fiberglass. I experimented with a number of adhesives and bonding techniques to attach the HDPE parts to the HDPE sheet and my search did not end up well. I even tried melting the pieces together using a hand held blowtorch. (Test pieces of course) This warped the material too far out of spec so it would not be a preferable means of attaching the two pieces. The best way I have determined to attach the two pieces is screws. I will cover the top of the screws with clay so that the gel won't stick to them.


The sheet of HDPE is very flexable. If you use this method to lay up a mold be sure to place the HDPE on a flat surface and do not move it while the fiberglass cures. Doing so may result in a warped mold. After the fiberglass and gel coat has cured, attach some thin sheets of plywood or foam to the mold to ensure that there will be minimal flex. After the supports have cured you can seperate the plug and the mold.

My Base Plate Lay Up

This lay up was extreemly uneventful, which is great. I used some of the mold clay I had lying around to build a small barrier around the edge of HDPE so the gelcoat does not run off  the edge. After laying up four layers of fiberglass I added two small pieces of 1" x 3" wood for support. I left the layup to cure overnight and then next day it was ready to be seperated from the plug.

I had some difficulties removing the plug from the mold. The gel coat did not stick to the HDPE sheet but it did manage to get between the base plates and the sheet in such a way that the base plates could not be pulled free from the mold. In addition to this problem the heads of the screws were positioned towards the gel coat so the HDPE sheet could not be unscrewed. I was forced to use a hole saw to cut around the screws. This removed the HDPE sheet, and then used a box cutter to cut though enough of the base plate to pull it out of the mold.

There were some air bubbles that formed which created some small holes in the mold. I used Bondo to fill in the holes and shape portions of the mold to ensure the mold has enough of a draft to easily create multiple parts. The nice thing about Bondo it that you have control over how fast it hardens. The more hardener you add, the faster it cures. The Bondo will progressively get harder. Before it is fully cured I recommend using a sharp razor blade to remove any excess Bondo. It will have to be sanded either way, but Bondo is relatively difficult to sand so you can save yourself some time. You may have to add Bondo to the mold multiple times. This is normal, sometimes you miss holes that need to be filled it and you have to go back and fix them. Take your time and be thorough!!! If you are not meticulous during the preparations, then your parts will reflect the poor quality.

Baseplate Mold

This is the finished mold. The orange color you see is the orange tooling gelcoat. The white is where I used bondo to fill in cracks and adjust the draft. The darket areas on the mold are where the gelcoat is thinner. They appear streaky because I used a brush to apply the gelcoat. Had I sprayed the gelcoat on, the color would be more consistant. This however does not effect the strength of the mold. Tooling gelcoat is mainly used because it is easy to spot flaws in the mold.

Baseplate reverse 

This is the reverse side of the mold. As you can see I placed a 2" x 4" along each side for support. These pieces provide excellent rigidity and double as work stands for the mold. I used a single layer of fiberglass along each edge of the wood while laying up the final layer of fiberglass.

The edges appear smooth in these last two pictures because I used a band saw to clean up the sides. Before I did that, I had fiberglass strands sticking out all over the place. I lay fiberglass slightly over the edge to ensure that the area I wanted to cover received the full four layers. I also rounded off the corners using a file for safety sake. Don't kid yourself, this stuff is sharp and will easily cut you.

I want you to also notice the lack of air bubbles on the reverse side of the mold. Compare this to the mold that I did for the 2.5rs Hood. Because I did the layup inside away from UV light and in a slightly cooler area, the resin did not cure as quickly which gave me a longer working time. If you have to work outside and/or in a very warm area, reduce the amount of harder you add to your resin. The less you use, the greater amount of working time you have. The recommended amount is 1% hardener for standard applications.

Mold Preparations

For this small mold I added 5 coats of wax and a single layer of PVA.

Carbon Fiber Layup

There are two ways I could go about laying up the carbon fiber in the mold. The first is by cutting out pieces of cloth the approximate shape of the final part. The pieces of cloth that are cut to shape fit better into the mold and will result in a final part that is dimensionally close to the part the mold was made from. The layup method I will be using is oversized pieces of carbon fiber fabric. This will not conform as well to the mold but can I cut them out of the roll much quicker than a custom shape. At the thickness of the cloth I am using, 15 pieces are required for a thickness of 1/4". This first layup is only a prototype so this first layup is a proof of concept effort.

Remember that when working with carbon fiber, you should always use epoxy resin. I used a 3 to 1 medium cure time resin. The bottles say the resin has a work life of approx 20-25 minutes but I was in a cool room so I had a bit more time. Epoxy resin is very different then polyester resin in that epoxy resin uses a Part A mixture and a Part B mixture that chemically reacts together. The ratio must remain consistent in order for the reaction to occur properly.

I cut out a total of 30 squares, 15 for each base plate. For the epoxy, you generally want to use an amount that is equal 40-60% of the weight of the carbon. I find that 50% is a good midpoint. Only mix up an amount you can use before the resin starts to cure. I started with enough for one base plate. I wetted out each piece of fabric individually and then placed it on the mold.

Wetting Out Fabric

To wet out a piece of carbon fiber, place the fabric on a sheet of HDPE or a surface you do not mind ruining. Pour a small amount of resin on to the fabric and use a small squeegee to move the resin around the fabric. You will notice that the areas that are fully wetted out will no longer be shinny. After the first side is fully wetted, remove excess resin to the table top and flip the fabric over. Brush the back side of the fabric with the squeegee and add additional resin if necessary. Don't forget to ration your resin. If you have made enough for multiple pieces, be sure not to use too much on a single piece and leave others lacking.

After the first 15 pieces were completed and on the mold, I mixed up another batch of resin and then wetted out the next 15 pieces. After positioning them into the mold, the set up my vacuum pump and bagging system and left the parts to cure over night.

The next morning I removed the vacuum bagging set up and after a quick rinse under hot water to remove the PVA the parts popped right out.

I will use a bandsaw and grinder to remove the excess carbon from the part.

For these parts to accuratly fit on the car, they need to flat. The excess carbon created a large lip that needed to be removed. To do this I used a mill and simply ground away at the top until I had a flat surface. I also used the mill to cut out the center portion of the base plate. I should note here that it is a good idea to use a particulate mask when cutting or sanding carbon. The dust is very fine and can adversly effect some people. For the smaller strut bolt holes I used a drill press to cut through the carbon. At this point I encountered a problem where the drill bit would catch on the carbon and swing the part around. After some trial and error I found that three seperate clamps were needed to hold the parts in place so I could drill a clean hole. Most of the cuts I made were not very clean but a file cleared that up quite nicely.

Once the parts were in acceptable enough condition to place on the car, I was able to verify that the strut bar lines up with the tabs on the base plates. Some minor adjustment was needed and then I finished welding the sperate pieces together.

One of tricky things about pulling a mold off this shape is the lack of draft on the T stock. I decided to use Bondo to build a draft on surfaces. Bondo is great for plug modification but is relatively strong and therefore difficult to sand. I have found that using a razor blade to cut off excess Bondo while it is in the early stages of curing before it has become rock hard it much easier than trying to sand out the larger imperfections. Bondo should generally be wet sanded. Without water as a flushing agent, the particles of Bondo will quickly clock the sand paper.

When I was satisfied with the plug, I fabricated a base using a sheet of HDPE and clay. As you can see from the pictures I used the sheet of plastic as a base and used the clay to build a base around each end of the strut bar. The center of the bar is flat but each end is angled up at a 6 degrees. Besides being able to make shape I prefer out of the clay, gel coat and fiberglass do not stick to it.

Strut Bar Plug

In the picture above you can also see the spray bottle of PVA.  At the time I made the mold, the bottle was not functioning correctly and I was forced to put on the mold release with a brush. I believe this caused some of the problems I had getting the plug out of the mold, I'll talk more about that later on.

As with the other molds I have made, I am using orange tooling gel coat. I find that It makes imperfections in the mold easy to see. I have found that gel coat does not apply very well to sharp corners so if you are making a part with sharp corners, be prepared to do some repair work on your mold.

After the gel coat became tacky I applied 4 layers of fiberglass mat and a generous amount of resin. Unfortunately for me, the fiberglass did not want to conform to the part and I ended up with a large amount of air bubbles in the mold. A fix for this problem would have been vacuum bagging the entire part, which it what I would have done had my vacuum pump been at my shop and not in my apparent. Needless to say I have a lot of Bondo repair work on this particular mold. To ensure a strong mold, all the air bubbles behind the gel coat must be filled in. This means poking each part of the mold with a razor blade searching for the air bubbles, and then filling the holes in with Bondo.

Regretfully, the draft I created on the edges was not enough to allow me to remove the plug from the mold. I was force to cut the mold in half to remove the plug and then fiberglass it back together. After repairing all of the air bubbles, I will increase the draft to ensure that all parts being made will release easily.

To Be Continued....

Back To Projects

Copyright © 2008 Bar One Composites