Rear Engine Top Cowl

This is going to be a bit of a longer post, so please bear with me. I am going to try to get the blog up to date as of today, then keep it current going forward…. WHAT??? It could happen.

As I mentioned in a previous blog post, the cowls don’t fit. Neither front nor rear. But that is part of the beauty of composites, everything can be fixed. I also had to build my own cowl for Betty, so I wasn’t in the least bit intimidated by this revelation. So hopefully I can take and apply to cowls number 2 and 3 what I learned from Betty as well as some advice from some composite expert friends. I am sure I can do a good job on both cowls.

The front is going to be a simple fix. The spinner and prop just needs clearance. I have a plan and don’t think it will take long to fix. It is going to be a simple amputation of the front 2-3 inches of the cowl, and re-build with proper prop clearance. The contours are already well defined by the shape of the cowl, inlets and spinner, so not much magic there.

Aft is a bit more work and head scratching, so naturally I started there. The aft cowl is about 5 inches too narrow to fit the engine. Most of that 5 inches is in the top cowl as the seam between top and bottom cowl runs about the bottom of the rocker covers.

As you would suspect, I started by cutting away what doesn’t belong. This allowed me to put the cowl on and assess the situation. I held straight edges and splines up in every direction until I couldn’t stand it. I was not sure how to make Brian’s original shape, albeit larger, work. About then, for a fleeting moment, I did think of going a la Cub and leave the cylinders out in the breeze.

Ok, back to reality. I spent a lot of time discussing with multiple people and researching. I had 3 priorities for fixing this problem.

1. Stay true to Brian’s DDA design and the intent of the project.
2. A low drag cowl shape.
3. An aesthetically pleasing cowl.

Number One was my big hang up. Luckily, it was alleviated by talking with Brian’s wife Arleen. I had asked her if there was a solution Brian was already planning. She was unaware of any plans he had for the rear cowl. She further told me to do what I think is right.

I had a huge sense of relief when she said that. I want to be a good steward of Brian’s DDA. I don’t want to change more than I feel is necessary to get it flying. I didn’t think I was going to be able to keep his original cowl shape and get a workable solution, without huge drag and aesthetic penalties. Arleen giving me the permission to do what I needed to do gave me great comfort, and creative license.

Two and Three are a bit easier. These I can R&D (rip off and duplicate) and apply basic “rules of thumb,” and TLAR (That looks about right) to get it done. After all, if it looks right, It will fly right.

Since the Defiant is a Widebody Canard, I needed to look more at the Cozy and Velocity for inspiration. I looked at tons of stock photos of both models and then started thinking about Canard airplanes that I personally knew of. There are a few that stand out as having tons of ramp appeal, nicely modified, or being extremely fine examples of stock airplanes. I focused specifically on Scot Carter’s Limo Ez, Chris Esselstyn’s Cozy 540RG, Rene Dugas’ Velocity, and Bob Bittner’s Cozy MK IV. I also couldn’t forget my roots in the Long Ez world, and looked over my favorite top cowl, Wayne Blackler’s. There were bits and pieces of each cowl I liked and wanted to incorporate.

Being that I had CAD drawings of Brian’s DDA, I figured it would be simple to use those drawings to modify the shape. I would use CAD to then loft a larger cowl with features from my favorite widebody canards. This would give me templates to graft onto the existing cowl. But it wasn’t quite that simple. The existing CAD drawings for Brian’s cowl needed modification. Even once they were modified and converted for modern CAD, it was not a simple task to loft a nice looking cowl.

I spent quite a few hours for a few days trying to get this to work. Given enough time, I am sure we could have gotten something to work out how I wanted. But in the middle of the CAD learning curve, I was at the hangar and on a whim decided to give the old method a shot. So I started adding pour foam onto the cowl. Then more, and more.. well, you get the idea. I spent an entire day adding pour foam and sanding. What I came up with was abysmal. Awful. Looked horrible and likely would have been draggy. It certainly wasn’t befitting of Brian’s DDA.

It was so bad I had to stop working for the day and go home. I texted my brain trust of builders for their input and lamenting my woes. While I was in the pool contemplating my dastardly looking cowl graft, I remembered reading an article about how a Cozy builder built their cowl. The basic premise was to build splines out of pink Home Depot foam, jig them on the airplane, contour, apply a filler, and then make a mold. That is what we did.

Saturday, Kevin and I spent the day at the airport. We started by cutting away the right half of the cowl to start fresh. Then we cut a few splines out of a 4×8 sheet of foam. We then jigged these splines onto the cowl in key areas which I will call control splines. These Key areas were the cowl seam between upper and lower cowl, the lower corner of the rocker box, the upper and outboard most point of the engine assembly which was actually part of the plenum, the centerline of the top cowl, and the step feature that allowed the cowl to integrate with the spinner shape.

The foam here comes in 1″ thickness. I chose to cross cut the foam into 1/2″ strips. This gave me a rigid enough piece of foam that it acted as a spline, but flexible enough I could bend pretty significant curves into each one. Side by side, these essentially made a contour map of what the cowl should look like. The splines were held into place mostly with pour foam. This pour foam is close enough in density and sanding properties to the pink foam. This means that when you are contour sanding they sand at nearly the same rate making a smoother surface. We also used some hot glue and foam core to jig the pink foam splines in place. We were careful to not allow hot glue anywhere near where we were going to be sanding.

By the end of Saturday, Kevin and I had half a cowl that we were proud of and that fit our vision of R&D-ing our favorite canard airplanes. I could certainly see how the cowl was going to finish out. So Sunday, I amputated the dastardly first attempt at the cowl graft. A cross cut saw and all the ugliness was gone.

We then worked on adding the same control splines on the left side as we started with on the right. Then it was a simple matter of filing in the area between the control splines. By referring to the right side, I could put splines in on the left side in much the same order and orientation as I did on the right. This was important. The same starting point on both sides makes symmetry easier.

Monday morning I started contouring the foam, and since I had already done it on the right side, it went very quickly. A couple hours and I had a visually similar shape.

Next step was to make templates to check that symmetry. The resolution you can get with foam is not as fine as you can get with laminates or fillers, but you should try to get the symmetry as close as possible to make subsequent steps easier. When I checked the symmetry, I started with what I labeled as station 28. It was 28″ aft of the cowl to fuselage seam. On the left side, station 28 was the largest conflict with the old cowl. So I used the left side shape of station 28 to define the shape I wanted and matched the right side to that template. From there I worked aft, and both sides were actually almost identical. Then I worked forward from station 28 to the cowl seam. I would chose which side had a nicer contour, template that side, and match the other. This sounds tedious, but with all the prep, making the control splines the same, and the care in spline placement, this only took a couple hours.

Once I was happy with the shape and symmetry, I then used quick setting drywall compound over the foam. I used the 5 minute variety. It was a hustle to get it spread in a productive way. As usual with fillers, it was sand, fill, sand, fill, sand fill, sand fill, etc. I probably did 6-8 hours of that until I was happy-ish. You never get to the point where there aren’t flaws. You more so get to the point where you are done.

I have since learned that I could have improved my technique to get a better drywall compound surface. I was dry sanding since the compound is both activated by water and water soluble. Turns out, if you use a moist sponge, you can fix a lot of things after it has dried. This would have allowed me to remove craters, gouges, and other such low spot flaws. Knowing that would have both allowed me to save a lot of time as well as have a much better surface that was nearly mold ready.

As it was, I decided I was done with drywall compound. For each application I made, I was making as many problems as I was fixing flaws. The foam isn’t structural, neither is the drywall compound. So if I got really into an area trying to fix contour, I would either get to the point where I would crack the compound or I would get enough flex in the foam that it wasn’t sanding but moving out of the sanding block’s way. I was chasing my tail.

It was time to make a mold or a part. So I broke out the wax, and started waxing and buffing my plug. I ended up with 4 or 5 coats of wax. Then I sprayed it with 3 coats of PVA mold release. It was ready for a layup.

After much discussion with Richard, one of my friends who is a composite expert, we decided that I should make a tool cowl. This would be a layup over my plug that would end up being a “positive” of the cowl. Of course, this means I will be off contour by the thickness of the laminate. But we can fix that. We decided that I should do a tool cowl because we know that the plug surface isn’t perfect. Therefore there is more filler and surfacing to do. If we make a tool cowl, we can leave all that weight in the tool, not the flying part.

That brings us to yesterday morning. I was able to get a 5 ply layup done in a hour and a half. The layup schedule was one each BID +/-45, one BID 0/90, and 2 plys 2 oz deck cloth. The deck cloth would mostly serve as sacrificial thickness to contour sand.

With the 120 degree hangar temperatures, it only took a few hours for the tool cowl to cure. Once it was cured, I started contour sanding it. It is amazing how flaws came right to the surface.

It is important to know when you have sanded all the high spots off. Once this happens, you have to stop sanding. You can’t fix a low spot by sanding down to it. You have to fill the lows. Sometimes it isn’t easy to tell when this happens. So I had some black sandable primer, and shot the tool cowl with that. Unfortunately, I ran out.

That is the current status of the aft cowl until I get to the hangar today. A mostly contour sanded tool cowl, half in primer. I know I have 2 or 3 areas I need to give a bit of attention to with filler. Once we have contour corrected, we can remove the tool cowl from the plug. This will allow us to thin out the flange areas from the inside. By thinning these flange areas, I can correct the thickness mismatch and have the cowl tangent to the fuselage. Then we can make a mold, and finally vacuum bag a carbon fiber part.

If you have read all the way to here, either you are mom, a builder, or have serious time on your hands. Any which way, thanks for following along on this project. I hope you learned something. It sounds like a lot of fussy work, and it is. But I think the end result will be worth it. I could certainly have made this tool cowl the actual flying cowl, but it would be heavier than it needs to be, and that would have annoyed me for years to come. I know that I will be proud of this part when it is done.