Fiberglass Frame

Investigating As to make a FG frame with had tools. I need a frame for my drive train tests and as I always wanted to do some FG stuff - well I am doing an FG frame.

What I decided on is an octagon as it gives me more room for pickup and other mounts it will have the corners of a rectangle cut off with a diagonal turning it into an octagon The FG will be Uchannel with holes every 2 in or thereabouts (not always 2 inches at corner. I decided on compression mold after some tests as its stronger and lighter (less resin). I made tests atm with both expoxy and Polyethylene and for now use polyethylene mostly. The epoxy is stiffer the Polyethylene takes more of a beating (boat resin hence that is why they use it in boats) . The polypropylene smells more but that is what I got PPE for and good ventilation in my new shed.

So here is my 2nd try (first one did not work out so fine)

The discoloration is from me making my marks on where to cut etc on the Glass - I got to do that differently probably make a template

Wall thickness is 2 mm and consists out of 3 layers of 1700 coaxial mat and on then on each side 8 once cloth. Then the bolts through the mold squeezes it down to 2mm.

This is basically the idea - I make them thicker now as those even with backing did not stay straight enough. For Mold material I have not landed on a choice yet and trying ABS, PETG and Nylon . Nylon is probably best with Epoxy but not that good with Polypropylene as even with release agent (wax and PVA) it sometimes sticks too much as you get it through the holes. (still something to work out).

The 7 in piece in the pic suspended between the ends will carry my weight (300 lb) and can be hit full force with a hammer. (not very scientific test I know) . The weight projection for a whole frame is going to be somewhere between 6-9 lb depending on configuration (excluding wheels motors, gear boxes etc of course.

Well wish me luck and a happy 4th to all of you.


Just as a relevent point: I think there was a team or two that used structural Fiberglass tube for frames in the 2011 era.

Perhaps someone else has more details. Here is the best relevent thread I could find from a quick search:


842, in the same general area, also did. @s_forbes was that before your time or during?


Thanks. I played with the idea of off the shelf stuff but feel that might be good for certain things Making custom molds will allow for all kinds of custom stuff.

Are they still doing it? How well did it work?

Hard to do a frame when Covid seems to have ended the team. That said, their 2013 frame was Kit, so seems like they stopped some time before then?

1 Like

Too bad - too many teams bit the dust during COVID. Just looking for any info that could be helpful

From a fundamentals standpoint, using fiberglass to replicate geometry that could be otherwise made with standard aluminum tubing and some machine work almost always requires an amount of effort that is not reflected in improvement (lower mass, greater stiffness) of the end product. I figure you’ve probably experienced that a few times now.

Modifying preformed composite tubes/etc if you can find them is fantastic, but if you’re doing all this legwork to start from scratch, it’s more beneficial to design a structural system that heavily benefits from the more complex geometry that can be created. Stressed skin panels and such. 842 was usually pretty good at balancing prefabbed parts and custom layups, but there were definitely a few robots that went too far in avoiding traditional materials.

I’m surprised there aren’t any CADathon teams (afaik) that have done something crazy like a fully integrated base and superstructure from a single fiberglass mold and plate reinforcements.



1 Like

I worked with pultruded fiberglass drivetrains on 1726 from 2006-2008, and on 842 from 2011-2012. The best implementation IMO was the NERDS 2008 drivetrain, which used cantilevered dead axles and just 4 pieces of c-channel for the frame. It worked because the fiberglass frame members were so stiff.

The implementation that we used is different than the custom layup method in the OP of this thread, which I don’t have much experience with. I’ve seen custom layup fiberglass parts from other teams in the past, but their numbers are eluding me.

These days I prefer aluminum, and sometimes plywood. Novel construction methods are great though, 2767 is a good example of doing something different that works well.


Here is my journey. I started with FRC oh over 14 years ago. And in the beginning it was all either kit frames like Andymark - usually heavily modified or aluminum extrusion. Both usually had certain problems. Namely they started to look like swiss cheese cause we had a tough time weight wise. And most of the time they needed mr sledge hammer to make all 4 wheels touch the ground at the same time. We were a hand drill/sawsall kinda team an a 90 deg angle was a challenge. True we made regionals more than half the time and qualified for nationals a couple of times. And being a poor team we always were struggling $$ wise. I know teams who do not have those struggles. So then we got (and we is either 1989 or other teams after they “Died”) into 3dp reinforced with 1/2 in square al. tubing. Advantage was we got much lighter (no more swiss cheese) and its quite easy to make a 90 or 30 or 22 deg angle etc with 3dp and then slide the 1/2 in tube into it. It works and is feasible but it takes a lot of time to print and as you cant reuse parts from offseason or prior seasons you better have a good number of printers and the ability to run them 24/7. The cost is lower than an andymark frame (of course not counting the kids labor but usually they have fun) So the idea was/is if you can do some stuff off season. Like making molds - they are not part of the robot they are not COTS or custom parts or components they kinda fall under tools. And to my knowledge there is no rule that you are not allowed to use a tool purchased/made in the off or prior season.

So Figerglass was/is an idea. It weighs much less than aluminum and in some cases is stronger especially if you compression mold it. Now buying extrusions and then putting them together like Al extrusion is an option you get the advantage of FG (weight, strength etc) but you again get the accuracy problem if you work with hand tools. So the idea is 3dp a mold and use that. You need to be done filling the mold or doing a hand layup during the workable time of the resin which is between 10 to 20 min usually depending on resin . Then there is the cure time but that does not require labor just leave it alone. Then demolding and maybe a bit of sanding/post processing if necessary. If you don’t count the curing time where no work is involve you can maybe (there was a discussion) fall into the “doing a part in less than 30 min by hand” off season exception to make an off/prior season part “season legal”.

I tried shapeways (I think) to make 3dp “season legal” But found that for a whole skateboard frame aka Andymark frame that is impossible as according to their prices the whole frame would about to > $6k so would violate the $$ limit most seasons. Then you could say (if it was cheaper) I can use last seasons parts as anyone can buy them and I am making a functional analog (you can print a whole frame from petg for $100-150 in filament depending on what bulk you buy it in and from who.

Now prefabrication ex/pulltrusion of FG if you figure in shipping and cutting as you cant reasonably ship 20 foot sticks you wind up > $10/foot. If I am successful In figuring this out You can bring it down to about $1/foot plus the “free” “child labor” lol.

So my self assigned task (I am retired and this beets binge watching Netflix) is to see if I can find a way to compression mold FG in a setting that could be replicated in an FRC team setting.

In PPE the only thing that should be added besides gloves and goggles is a mask KN95 or better. I personally use a full face respirator. Mainly as the filter cartridges are cheaper in the long run than throw away single use masks, are better and full face - well I always struggle with goggles keeping from fogging up. I can work with my full face respirator with 60921 or 60923 filter cartridges safely for hours without fogging up and a better field of vision.

In resin so far I have tried both polyethylene (boat) resin and epoxy. For a school setting probably epoxy is better as it smells less. For me boat resin is probably the better choice its less expensive and way more impact resistent - hence that is why they use it for boats which take a pounding in the waves and robots take a pounding too. Mold making is a work in progress. And I just destroyed one cause the bolts going through it also made the holes in the part for bolt holes and the resin escaped through the threads and made demolding impossible without destroying part of the mold (so back to the drawing board/Computer cad)

If it works I will - like always share all my working stuff on grabcad. The frame I am making will serve as a test platform and will be able to take pretty much anything in drive train - tank, meccanum, swerve, x-drive and if I figure out a couple more challenges a ball drive

4130 and 3DP, LET’S GO.

1 Like

Chromolly or team 4130
Didnt use 4130 just regular steel
2020 1989 robot never competed due to covid and later was scrapped for parts the square holes are filled with either aluminum or steel square 1/2 in tubes. This was not the latest version as we felt we were too light so we put a 3/16 solid aluminum plate 30x30 in with the necessare openings for some motors on it to bring the weight up to a more reasonable 110is lb. This robot had about 15 kg of filament on it. So the FG is to do it with less during build season printing

There is a YouTube channel called Too Fast Matt. [Superfastmatt ] He builds/modifies automotive type stuff in his garage using FRC level tooling. He his currently making a fiber/resin body for his salt flat speed car. Well worth a watch.


Just a bit off on the name, pretty funny too in my opinion.

Make sure to put a ham sandwich in your robots fiberglass!


Is that the guy who epoxied a ham sandwitch into his rear stabilizer or am i thinking of someone else?


You are correct besides the name:

I seen that dont intend to do that lol I rather eat it

Among many other very necessary things