Chapter 24 - Cowlings

Why build a Cowling

Piled high with blue foam I'm making my own cowling because I don't need the "cheeks" and I want all the airflow to the prop I can get. I also want to make one of those weird looking boat tail things while I'm at it. My plan is to make a plug and use it to lay up a temporpary cowl. Once I've flown the plane and settled on the cooling I'll pull a mold off the temporary cowling and make a new cowl, maybe with carbon fiber to save some weight. I'm doing the lower cowl while the plane is upside down. I've been looking forward to this job. Step one is to get the wings on.

Inverted flight

Putting the wings on took about an hour. I used the come-along to jack the plane up until the wings fitted with winglets just clear of the ground, then put lots of blue foam under the fuselage to support it on the trestles. Once everything was secure I took a minute to inspect the wings. Wow. The contrast between the finished and gloss painted fuselage & strakes and the wings was dramatic. The wings are in primer wrere supposed to be "ready to paint". Not. Given the experience gained while gloss painting the bottom, my criterea for quality of finish has changed. There had also been a bit of shrinkage here and there. (I'd read that micro can shrink a little, so you should wait a while before painting. I left my wings in the sun for a few weeks to make this happen now rather than later. In between making the cowl I began the task of bringing the bottom of the wings up to the same standard as the fuselage.

Another Cozy visit

Jeff & Alison Andre dropped in during their Florida vacation and took Char & I out to lunch. They plan to build either a Cozy or an AeroCanard together while also learning to fly. I wish them well and look forward to seeing them at Cozy dinners. Visitors to the project are becoming quite frequent. Almost one a week theses days. I don't know about the visitors, but it's great incentive for me, and it gets the patio cleaned up regularly.

Getting the shape

I had a vague plan for how I was going to make the cowl, but this was experimentation of the best kind. No danger, other than lost time, and lots of fun. I have no idea how to do this. I'll find out as I go along.

I didn't have the proper bulkhead for my flow guide yet, so I made one out of a bit of 1/4 foam. I used this to position the flow guide on the prop hub. Next I duct taped all around the edges and made some plywood supports both to support everything, and rough out the shape. I covered all this with a sheet of plastic and got out the pour foam. A cup of pour foam didnt go far. My rough guess is that I'll need a couple of gallons of the stuff to do the job. Hmmmm. I dashed over to Home Depot and got three cans of "Great Stuff" expanding foam. The first can did more than the pour foam, but didn't really make a dent in the shape I needed. I decided to try bits of blue foam to fill the voids. I placed chunks of blue foam all over until it seemed to cover the basic shape, then squirted "Great Stuff" in between the blue foam to stick it all together. It looks ugly as hell, but I think I'll be able to carve [read hack] it into shape. I need more "Great Stuff". Another 4 or 5 cans should do it.

The home depot guy recommended the Latex expanding foam, so I got a couple of cans as well as the "great Stuff". I squirted the latex foam into a bunch of voids and went inside to wait for it to set. Two hours later, I found that it was still gooey. I read the instructions which say it takes 24 hrs to cure. The instructions also mention that it's water soluble, so I power washed it out of the way and carried on with the Great Stuff (which sets in about an hour). After another six cans of Great Stuff and a lot of carving I had an approximate shape. Occasionally I'd uncover a void while carving and have to fill it with foam, wait for cure, then carry on. I think this would have been easier with blocks of urathane foam, but I didnt have any and didnt want to pay for all the foam I'd need. Actually, urathane would probably have cost less than 8 cans of Great Stuff at 4 something a can.

You can't sand Great Stuff, but you can carve it fairly accurately. My plan was to make a smooth plug, cover it with something for release and glass it. If I'd used Urathane the shape would have been easier to get, but then I'd still have to figure a way to release the glass. I tried covering the foam with sheetrock, but decided the cure time wasnt good enough and you couldnt fill large holes with it. While at Home Depot on one of my Great Stuff runs I came across plaster of paris. $7 for a 25 lb bag. I bought two bags. You get 30 minutes to apply it and then it sets up like a rock. I covered all the foam with a skim coat, waited for it to set, then sanded it. It doesnt sand very well. It just clogs up the paper. I mentioned the problem to Char. She had worked with plaster of paris before and told me that you can wet sand it. The water soaks into the top layer and it starts to "unset". Cool. I tried this and, sure enough, I was able to wet sand the plug. I'll need some more plaster to fill some dips, but I think this can be done. It's starting to look like a cowling.

Working in Plaster of Paris

I'd cleaned the patio for Jeff Andre's visit. Doing the cowl was extreme fun of the childish kind, and the patio got severely messed up again. Plaster of Paris (P of P) is a whole new medium to work in. Normally you mix it up and pour it into a mold. Making something with it is a whole different story. It gets unworkable after about 10 minutes and its hard to put on thin. I tried putting it on with a brush, a putty knife and a wet cloth. Once set I tried wet sanding it with various grits. Talk about experimental. One day my freind Mark came over to visit. Isn't that going to be a bit heavy, he said!

Sheetrock is the stuff

After two bags I gave up on P of P. It's just too hard to get any accuracy. I asked on the list and a few builders suggested joint compound. I'd tried this, but the stuff I had was the 24 hr variety. At Home Depot I found bags of fast setting easy sand joint compound in 20, 45 or 90 minute options. I got two bags of the 20. While there I also picked up a gallon can of returned mismix paint for $3. It's green, but who cares. I don't plan to paint the real cowl with it. But... then again...

I slapped on a skim of this stuff and, sure enough it's hard in 20 minutes. Still a little fast, but its much easier to spread. After a few hours it was still too wet to sand, but I found that I could "scrape" it easily. After a few experiments with various tools I settled on a sharp, clean 3 inch putty knife. I found I could scrape the surface smooth very easily this way, pulling off anything from just a sliver to a 1/2 inch lump as needed. In a couple of spots I got down to the P of P and had to hack it out and replace it with joint compound. I needed to shape the cowl around the bump of the NACA scoop inlet, so I just added an inch or so of sheetrock filler and, after cure, carved it into a "pleasing shape". Don't try this with micro! So far I've worked on the cowl for about 24 hrs total, and I have a fairly smooth plug.

After about five or six sessions with the sheetrock compound, with wet sanding in between, I had a 320 grit plug the shape I wanted. Next day I waited for it to dry out, then dry sanded. Once satisfied with the plug I sprayed it with the latex paint. This dried in a few minutes but seemed a bit soft and was difficult to shine. It'll probably peel off with the epoxy. I had a couple of cans of enamel spray paint I'd been using for guide coats when sanding. They were dark blue, but what the heck. Still not enough paint. Hmmm. I found a large can of epoxy appliance enamel. This did the trick. I was happy with the shine, so I polished the plug with turtle wax and taped the edges with packing tape for a good release. I replaced the duct tape around the edges and epoxied some peel ply onto the bare plug ready for the reinforcements later. This was my first layup for months. I was a bit out of practice. After 4 plys of BID and 5 hours work I was back in practice. I had a bit of a battle with some of the curves. You know, when you squeege the layup into one curve and it pops out of another. I battled the last ply for about half an hour then decided "screw it" and made a cut through the ply about 5 inches long in one of the curves. Now, finally, it would lay down flat. I cut a piece of bit to repair the cut and overlap an inch all around, then covered the whole lot with plastic and squeeged it again. No one will ever know.

Sitting in the hot tub that night we admired the new dimension to the plane. I'm very pleased with the shape. Somehow the cowl just looks "right". Next day I lifted the cowling gingerly. It came off, complete with all the paint and quite a lot of the sheetrock compound. Hmmm. I thought that latex hadn't stuck well. I pulled on my release tapes and peel ply around the edges and the paint came away from the epoxy cleanly. To get the paint and sheet rock out of the middle I resorted to the power washer. It took a while, but eventually it came clean. The inside of my cowl is very smooth. It will make an excellent mold if I decide to make another cowl. So far in my experiments I've proven that neither Turtle wax nor Pledge work for mold release. I think I'll get a can of the proper stuff for next time.

Making the upper cowl

ready to make the upper cowl I think this is going to be interesting. With the engine and lower cowling in place I bolted on all the extra bits of engine I'd been accumulating. Water pump, smog pump, turbo, alternator. After talking to a few people on email, I decided to make the cowling the "right" shape, then make the engine live within that shape rather than the other way around. There was a pipe coming out of the side of the turbo that looked like it was going to get in the way, so I removed it. I also removed the water pump pressure cap housing. This was definately sticking up too high. I didn't bother with the oil cooler, intercooler or radiator because they are going to have to fit in the space available. I taped all around the wings and fuselage, firewall etc., then wrapped the whole engine in cling film. Piled high with blue foam I'd been repairing the swimming pool skimmer and had some 2 inch PVC pipe lying around. This'll do for an exhaust pipe! I stuck a piece into the turbo and aligned it so it was sticking out the back of the cowl. I collected a bunch of blue foam bits and stuck them in place. Rather like making a 3D jigsaw where all the pieces are blue and the shape you want isnt really fixed yet.

Once I had the rough shape of the cowl outlined in blue foam I used 3 cans of triple expansion Great Stuff to stick it all together. Char was doing some night gardening so I got her to pitch in. Next day I began carving. An hour or two of carving got me to a reasonably even shape.

Carving the foam

Char doing some night gardening Char applying the Great Stuff A really rough cowl shape Carving the foam A foam cowling The final shape of the cowl was still to be determined, but the foam got me to a point where I could start to see what I wanted. Two and a half 18lb bags of sheetrock (about 50lb) later my cowl was looking a bit better. Contrary to my own advice to keep the foam an inch below the surface, I'd carved the foam fairly close to the actual shape, so there were quite a few bits of foam sticking through the sheetrock. A few cuts with the carving knife and a bit more sheetrock cured this problem. As the shape took form I added large gobs of sheetrock to the left side to make it match the right. Two bulges sort of emerged from the mess and aligned themselves with the fairings along the top of the strakes. I tried to cut these lumps down, but somehow they kept coming back. I think they want to be there.

I tried both dry and wet sanding. Both techniques have their pro's and cons. An unexpected advantage of wet sanding, if you're small like me, is that the wet sheetrock builds up on the soles of you're sneakers. If you start with the sides you gradually gain height. By the time you're ready to do the top, you can reach it! Dry sanding 80 grit or so is better for fine work, but you have to wait a day or so for the sheetrock to dry. As I got closer to the final shape I removed the PVC exhaust and plugged the hole.

A sheetrock cowling - first cut A sheetrock cowling - ready to glass A sheetrock cowling - ready to glass I spent a couple of days filling and sanding sheetrock until I'd got a symetrical and pleasing shape. The cheeks are there because the turbo needs some room and I felt that I needed some kind of structure, otherwise the cowl would be weak and would look like a "lump on the back". The outside edges of the cheeks are parallel with the fairings on the fuselage. So airflow and drag wise, they don't stick out at all. One issue I'm not sure about yet is the fact that the upper and lower cowlings dont match each other. When taken as a whole, they don't merge. This is purely an aesthetic issue. Maybe I'll extend the lower cheeks to match the upper ones later. Maybe not.

Glassing the cowling

Glassing the cowling Getting the paint off the inside of the lower cowl had been a lot of work. This time I decided to do without the paint. Instead I sealed the sheetrock with some wax and dived in. This was to be my last big layup. You'd think I'd be well prepared, me being an experienced epoxy guy an all. I epoxied 2 inch peel ply around the edges. The first ply of BID went well. No problems. When I layed the glass on for the second ply the BID fell off the end of the roll. Damn. I had enough for the second ply, but this was to be a 4 ply layup. I searched around and found some large off cuts, but nowhere near enough for 2 plys. I've read that some use UNI instead of BID for one or two plys. Hmmm. I have a fair bit of UNI left. I did one ply in UNI, then pieced the last ply out of BID cut offs, overlapping an inch on each seam. Finally the plastic ply and I was done. Phew. It had a long 3 hour layup on a hot July day. I dived in the pool to cool off, then collapsed on the couch.

I was determined to leave the cowl to cure overnight, but later that evening I went out to inspect the layup. I decided that it would be a good idea to do some finishing work while the cowl was still supported on the plug. I mixed up some dry micro and started in. I put a skim (may 1/16th) over the relatively flat parts, but when I started on the rounded cheeks I came across the old problem of how to apply dry micro to a curved surface. I came up with a new (to me) idea.

Upper cowl at the bottom of the pool How did the waxed sheetrock plug work out?, you ask. Well, it came away with a very thin layer of a new filler I've invented. Instead of micro, mix epoxy with sheetrock putty. When cured the result is hard, heavy and quite difficult to get off. It's saving grace is that its water soluble. I dumped the cowl in the pool overnight. Next day the stuff scraped off without too much effort.

Summary of cowl building experience

The lower cowl took me about 5 days (say 40 hrs) and cost about $50. The upper cowl took a little less time and cost about $25. Cowl making was probably more fun than any other major task on the plane. You get to carve a lot of foam 'to a pleasing shape', wet sand sheetrock, and make a huge mess. My cowl is the shape I wanted, and its a perfect fit. If I were to do it again, even if I were building with a plans engine installation, I might consider making my own cowl.

Things I learned about making plugs:
1. The (wood) base you use for you're foam plug needs to be strong and well anchored so that the foam can't move when you press on the plug later. 2. You need to build the initial foam plug at least 1/2 inch smaller than the final shape to be sure the foam doesn't poke through anywhere, so the accuracy of the foam carving isn't really important. Given this fact, you can make the initial plug with just about anything. Lumps of blue foam stuck together with great stuff work fine. Carve the foam fairly close to the shape and dig it out if it comes close to the surface anywhere. 3. Sheetrock compound isnt very strong when thin. I think my plaster of paris coat might have helped because it gave the compound a good solid base to rest on. The only problem with this was that the plaster of paris base was too close to the final shape and it showed through here and there making it difficult to sand the mixed surfaces. On the upper cowl I didnt bother with P of P and the job was much easier. 4. Use a good mold release wax, or paint the plug with something that will peel off the epoxy easily (like latex). Or, don't paint the plug at all. Sheetrock is easy to get off the cured epoxy. 5. I should have done the cowling before painting the fuselage bottom. My duct tape masking pulled off some paint when I removed it. 6. NEVER try to remove paint from epoxy with paint stripper. Thankfully I did a test first. The paint stripper removed the epoxy too. 7. Making plugs is fun!

Modifications to the cowl

The cowling after fitting Once all the cowling lips, reinforcements and nutplates were installed (See: engine) I decided to make a few modifications to the shape. I'd been expecting to have to do this, and I'd received quite a few helpful comments about the shape of the boat tail. The general consensus seemed to be that it had too much curvature at the back. Also, I researched the "hershey kiss" spinner and found that I needed a small gap around the flow guide so air could be accelerated past the flow guide and onto the spinner. I decided, fairly arbirarily, on a 1/2 gap around the spinner. I wasn't happy with the way the cowlings joined at the back. The easiest thing to do when I made them was come straight across from the aft wing root to the spinner. I wanted this curved, so I picked an aesthetic curve that not only looked right, but would allow me to get further forward for preflight inspection purposes. (Note: I havent mentioned this yet, but I'm planning to put doors on either side of the upper cowl.)

During the cowling modifications, I took a day off to help Bulent install his spar.

Cutting the upper cowl All the above meant that the rear foot or so of my upper and lower cowling needed to be reshaped. With the engine, redrive and flow guide in place, and the cowls screwed on tight, I started in with the dremel. First I cut the curve from the aft wing root to the "cheeks" going through both cowls at the same time. This left me a gradually widening gap between the cowls from nothing at the wing roots, to about an inch where the curve met the cheek. I considered making a vertical join here, or even using this as an air outlet. After staring at it for a while and pacing around it, I decided this would look better if the two cowls met at the back and the trailing edge of the wing simply curved into the cheek. To make this happen I had to cut the join at the bottom of the cheek almost all the way forward to the firewall. Once this cut was made the "flat" part of the cowl could be lowered a little. The net result is that the curvature of the wing is simply extended all the way to the cheek. This was one of those times when, all of a sudden, everything just seemed to look right. I clamped the trailing edges together and bondoed sticks to hold the two parts in place when I remove the upper cowl later and bridge the gap on the inside with glass. Next I followed the curve of the cheek all the way to the spinner and cut the horizontal lips off. Finally I added more sheetrock on top of the existing cowl to give me the 1/2 gap around the spinner and the redesigned boat tail.

The revised boat tail and curved trailing edge I glassed over the new sheetrock with 4 ply BID and then cut away the old cowl from underneath. I added 3 ply BID lips along the base of the top cowl all the way to the spinner so I could attach it to the lower cowl and added a couple more nutplates along the lip.

Another Cozy Visit

Augusto & Javier (plans # 963) try a Cozy on for sizeAugusto & Javier (plans # 963) dropped in for a quick visit. They said I was doing wonderful work, but at this stage they're easily impressed. The revised boat tail The finsihed cowl with spinner and prop in place. Char wants to paint the spinner in black and yellow stripes

Cowl Door

I planned to put full size doors in both sides of the cowl, but didn't get around to it until the plane was down at the hangar. I taped where the doors would go and put the cowl in place to see how it looked and check that I'd got the edges straight. I left it sitting like that and went home for the day. Next day, Paul Mason came over as soon as I arrived in the hangar. "I want a word with you", he said. Uh oh. I wondered what I'd done to offend this vertically challenged Aussie gentleman and canardian of ill repute (see my introduction to Paul ans Al in chapter 27).

"Are you planning to put cowl doors where you have those tapes?", he asks. "Uh.. yes", says I. "Why?" was his one word reply. "Well, I've flown cessnas with their tiny little oil check door, and owned a Piper with full size cowl doors. I liked the Piper much better because I could have a good look around during pre-flight", I explained. "How fast did the Piper go?" asked Paul. When I said 110 with the wind behind it, he asked "Do you know how much pressure is on the upper cowl at 200mph?". It wasn't hard to see where he was going with this. Basically, with a series of leading questions, he advised me not to make large doors in my cowl. I took his advise, but I did expand the plans door to about 10 inches * 7 inches. I figured out how to install the Hartzell catch only after looking carefully at Al's catch on his LongEz. I had a spare strip of extruded hinge which was just about the right size. Rather than riveting it in place I installed three nutplates on each side. When the door was done I decided that there was a chance the airflow could get under the leading edge, so I added two nutplates for torx screws just to be on the safe side.

I was just finishing off the door installation when Tom (Glasair builder from across the way) dropped by. "You know, most people do this stuff before final paint", he says smuggly. "Yea, I know.", I replied. Tom's been a good buddy since I left the hangar, but I swear I'll drop him one of these days....

The rest of the cowling work comes under engine and Finishing so this chapter is now, finally, done. But - I'll add here for anyone thinking of making their own cowl - finishing the cowling was a LOT of work. The finished cowling before polishing

<<< Back | Index | Next >>>