Week 16: Wildcard (Composites)¶
Goals for Composites¶
- Build a useful composite part for my boat.
- Finally, wrap my mind around composites and their production techniques**
- ~ Compare a simple layup part to a vacuum bagged part in terms of strength to weight ~
~ are stretch goals ~
**I can already hear you Dr. Fagan, “They didn’t teach you that at State??” They did, and I’ve understood them in theory for a while, but I don’t think it will fully click until I make a few myself.
Project selection¶
I’ve gone back and forth a lot for Wildcard week. My friend wants me to make him a convertible top cover for his 70’s Chevelle. We would make a composite and then he would have it upholstered. As cool as that would be, the size of the project (~ 60”) was daunting for my first foray into composites. I also considered vacuum forming a plastic part for him, but I was unable to find a vacuum former this week, let alone one that large. Having not fully moved on from my initial exploration for molding/casting week, I decided I would build a part for my sailboat afterall. Specifically, the Aquadyne Sailbird has issues with waves. Calling it a “wet” boat is putting it nicely. See an extreme example here of 20 knot winds in Ocean conditions.
Others sailbird owners have purchased Hobie parts to this end. Here is Jack Young’s solution for instance.
I decided this would be a major improvement to the boat if I could build a strong composite here.
Procuring Materials¶
Even still, I decided the only way I was going to try and tackle composites, would be if I had a partner in crime. When I opened it up to the group, none other than my CNC buddy Aarush stepped up to the plate. Brave man. So he and I set about creating a wishlist of materials.
All else being equal, we did our best to choose biobased materials that would be friendly to the environment and our lungs. That meant no fiberglass, and paying up for biobased epoxy. I am unsure how this epoxy will perform with extended UV exposure, but I think this may be able to be alleviated with a top-coat.
Design¶
So I set about designing it in Fusion.
This came together quite quickly as I tried to keep it very simple. I plan to make the left and right dodger as one part and them cut them in two with a saw. Additionally, I don’t have great dimensions on the hull so the plan is to sort of oversize it and then cut it back as to line up with the hull.
Aarush and I determined that the deepest we could machine with the 1/2” flat end mill would be ~ 2.25”. The layers of foam we will be gluing together are each 0.75” thick, so this conveniently works out to 3 layers. As such, I have designed the mold to fall entirely within 3 layers of foam, with a 4th on the bottom for extra flat area around the edges. This is good practice for vacuum bagging according to this resource here, which I read thoroughly.
I will need to update this design according to what size ball-end mill we have available. Filleted corner features will be more machinable and aid in the vacuum bagging process.
Back to the drawing board¶
I had put the above design down for some time before returning to this work. In the meantime, I had a chance to take some more measurements and reference images of the boat that would be used to make the part work better. I also took the opportunity to make a reference model out of posterboard. This model allowed me to capture the dimensions of the part transfer them into CAD later on. Here are the reference images and snapshots of the modeling process using them.
Finally, here is the most simplified design I could come up with. I finally settled on the part wrapping underneath the metal bar of the boat (ockas) rather than overtop. This allows the entire mold to be nearly at the same height and machined out of just a few pieces of foam.
You may notice the two wings share a central channel. This has the benefit of reducing the size of the mold and machining time, with the drawback that I will only be able to mold one wing at a time. In hindsight, with all the setup and mess involved in laying the composites, it would have been ideal to be able to mold both wings at once. Even still, with time running out, I will be happy to produce one good part now and its counterpart down the road.
Another drawback of this design is that the surface that will be primarily visible to boaters is the top-surface, which will be facing the bag rather than the mold. After reading several Fab academy pages from former students, it seems clear that this side will be more wrinkly than the bottom-side that faces the mold. In the future, I might choose to make such a part with a negative mold rather than a positive.
Milling the mold¶
Then came the milling process. I was lucky to have Aarush by my side for this process. We exported our models out of Fusion as .STL files and followed a tutorial on how to import them into aspire, lay them out, and begin creating our toolpaths. Luckily, this was pretty straight forward. One trick I learned were to make sure to import the STL as mm rather than inches in order to get the scale right. This is despite my Fusion design being made in inches and the shopbot operating on inches. Another important note is to make sure the model is laid flat on the virtual build-plate. The settings to set this up correctly in aspire were reasonably intuitive. The software was then able to determine that my model was 1.25” high with 0.25” of stock above it (starting stock is 1.5”).
To create our 1.5” stock, we glued two 0.75” sheets of insulation foam together. To create our toolpaths, we used the 3D roughing (0.5” FEM) and 3D finishing (0.5” BEM) options. These toolpaths did not have a lot of options but we did discover the difference between a 3D raster and a working off of Z-levels. For Aarush’s batmobile, which had many complex curves, the former seemed more appropriate whereas the latter seemed better for my model.
The milling went off largely without a hitch. We upped the depth of cut to 0.42” with the 1/2” FEM. In hindsight, we could have likely upped the feeds/speeds as well because it naturally cut the foam like butter. After about 45 minutes of roughing and 10 minutes of finishing, you can see the final results above. You may notice the leading edge of the mold did not receive a finishing pass. This would have been ideal, but I missed it when programming the toolpaths. For the sake of time, I will sand these edges by hand before moving on to the molding process.
Composite laying and Vacuum bagging¶
Then came time for laying the composites. We read the epoxy safety datasheets and settled on wearing KN95 masks and Nitrile gloves. Additionally, we decided to lay the composites in our ventilated kiln room.
We then developed a todo list.
Adrian’s page and this resource from West Systems were especially useful to help us determine the order of the layers. see below.
We glued Aarush’s mold to the base of mine so that we could keep everything intact. We used the same spray adhesive glue mentioned above.
We wrapped the complete mold in clear plastic kitchen wrap. This actually became a source of error later on. More on that later.
To cut the linen to size, I created a sketch in Fusion, projected the correct geometry, exported as DXF, and used our 100W laser cutter to cut them out. The laser made quick work of the linen. I would go on to use 4 layers of linen to create my part.
We then cut the perforated film layer and breather/bleeder/blanket layer large enough to fit over the entire mold.
We then taped the vacuum bag material to the floor in a large enough area such that we could fold it back over the mold and tape it down.
Preparing the vacuum bag for the hose was not easy. Unfortunately, our kit did not come with any sort of shutoff valve, so the best we could do was cut a small slit for our hose and tape around the edges of the hose to seal it off.
We then mixed the epoxy in the suggested 2 to 1 ratio. With only a 21 minute pot time to work with, the clock was on. Here is a great timelapse Aarush captured that shows us laying the epoxy/laminates, covering the mold, sealing it off, and engaging the vacuum pump.
As you can see, the vacuum pressure helped suck the laminates down to the mold quite well.
Without the shutoff valve, we decided to leave the vacuum pump running overnight. Much to my suprise, I came back to a foggy room the next morning.
Upon further inspection, the vacuum pump spewing some sort of mist or fog. My first thought was fire and 2nd was epoxy fumes. I dismissed both because there was no major smell. Additionally, it wouldn’t make much sense to see that many fumes coming out 10 hours after the epoxy was left to set (4 hour set time). My only remaining theory is that this mist was coming from the vacuum pump itself. The vacuum pump was running hot, likely because it was left running all night. In hindsight, we should have purchased a shutoff valve and not left the pump running overnight.
Anyhow, the parts looked done.
After unwrapping the parts, they quite easily came off of the mold.
Then we did a little post-processing with scissors and palm sanders. As suggested in the safety data sheets, we wore dust masks while doing the sanding.
as you can see, the finished parts came out very well.
In hindsight, the only things I would change about the mold design would be the height of the walls. The 0.5” walls did not come out super well-defined. This problem was exacerbated by the way we wrapped the mold in clear plastic. You may notice that the parts not suck down completely to the mold and instead have a outward curve around the edges. We think this occurred because the clear plastic kitchen wrap did not have enough slack. This issue can be seen in the following two pictures.
In the future, we will need to find a better way to protect the mold without affecting the shape of the finish parts.
Finishing touches and Installing on boat¶
As finishing touches, I spray-painted the part blue and drilled some necessary holes for attaching.
The piece was then easily attached to the side of the hull and ocka with a small string.
Now all thats left to do is test it out on the water and make the opposing part… someday.
Former Students’ pages which proved helpful¶
Emergency Embroidery¶
Let the record show that my awesome team-mate Nidhie helped me crank out some embroidery on 6/28/22, the day before all this was due, just in case our composites failed or somehow we ran out of time. Luckily, this was not the case, but here is the piece we embroidered.
I will not bother documenting this any further since the composites worked.
Link to Group work¶
Aarush and I documented our group work for this week here.
- Composites are labor intensive and produce a lot of waste
- The result we achieved was very strong and flexible.
- Next time, we need to find a better way to protect the mold. We can either use the kitchen wrap and simply add extra slack or find another material that will more closely hug the mold.
- Next time, I would make the walls of my mold taller and more defined.
- Leaving the vacuum pump on all night is likely not a good idea. In the future, we should procure a shutoff valve.
- I had thought my two sided mold concept was clever in that it reduced my machining time, but in the end, not being able to mold both pieces at once was not worth the tradeoff.
Note: All design files and images can be accessed in my git repository found here.
All works shared on this site 'A Fab Academy Journey with Charlie Horvath' by Charles W. Horvath are licensed under Attribution-NonCommercial-ShareAlike 4.0 International
