I’m far from being a composites expert—I primarily build metal airplanes. But inevitably, all aluminum rivet-bangers will have to do some glass work (just as all glass builders will have to build aluminum engine baffles), so we learn what we can and do what’s necessary. External fairings and wingtips require lots of fine finishing work, while internal things like ducts can be a little rougher—but it’s fun to see what you can do with composites, like making parts that have no straight or parallel lines!
I needed to build an air duct from the baffles to the oil cooler on our F1 Rocket project a few weeks ago. I knew where it had to start and where it had to end, so I cut out some scrap pieces of particle board and tacked them in position with double-sticky tape. Then I grabbed some hardware-store pink foam blocks I had handy and started filling in the space between—spray adhesive and a band saw make this easy. Once I had built up all the space between my two “ends,” I took it off and started sanding away all the stuff I didn’t need, ending up with a shape you see in the photo above.
The next step was to wrap my male mold in “release tape”—otherwise known as electrical tape. Fiberglass won’t stick to it, and it stretches to curves nicely (packing tape doesn’t do that). I mounted my “duct blank” to another piece of wood so I could mold in mounting flanges as well.
Finally, out came the messy stuff—glass and resin. This is always a hand-to-hand combat job for a non-composite guy like me, but I ended up with a nice shell of a shape that matched the mold. The pink foam dissolves out with acetone, and the electrical tape pulls out of the resulting gooey mess. Now I can take this duct, add some more layers to make it pretty, and voilà—I have my cooling duct!
And there’s no way I’d want to make this sort of shape out of aluminum.
Great article Paul!
I never knew about the electrical tape trick!
Keep up the great work educating us. We really appreciate it.
Brett Hahn
But Paul, the engine shakes and the cooler is fixed at the firewall. How will account for this in your design?
This is just the raw duct – not a finished assembly. There will be a slip joint and seals at the engine end before it is finished and ready for flight. Of interest, however, is that the engine really only moves significantly in “roll”, which – where the duct contacts the back of the baffle – translates mostly into up/down motion. This makes the interface simpler. We know of a number of old-timers that have been using this kind of setup for many decades (and thousands of hours) without a problem. What you DON’T want to do is bolt the front of the duct to the engine (as you point out) – that will fail immediately.
How do you intend to connect the independently vibrating cylinder baffle face to the fixed-end of the duct? Some sort of flexible coupling?
In the production aircraft world, composite ducting used to be laid up on “wash out” mandrels. These were mandrels cast to the “inner mold line” of the duct out of a water soluble salt based material. You could wrap the mandrel with wet lay up or pre-preg high temp composite, vacuum bag it and toss it in the oven. After the laminate was fully cured, you de-bagged it and washed the mandrel out with running water. Nice, light composite duct work with service temperatures up to 400F, depending on the resin system being used.