This section starts with me being overly enthusiastic about the possibility that my airplane would have blended winglets, and ends with me being disappointed in my execution and memory of my high school geometry. All is not lost on this page though. I used the same technique (except for the error pointed out below) for the final version of the winglets, which turned out great.
To back up, I decided to saw off my original, per plans, winglets because I was unhappy with my workmanship on the structural attachment, and I was hot to trot to have blended winglets. I found a builder website showing that the blended winglet installation was actually pretty easy. Additionally, I knew of a person in my EAA chapter who had installed blended winglets on his Cozy MK IV. So I got into contact with him, and he sent me all of his build photos for his blended winglet conversion. Armed with this information, and excitement, I was off and running.
My first step was to take the photos that I had received and write out my own plans in order to think through the entire process. The basic description of those plans are what follows in this section and the next section. NOTE: I had extensive help review these plans, both builders and an engineer. If you are attempting a major structural/aerodynamic mod like this, consultation with a structural and aero engineer is a must.
The first decision I had to make was what should the minimum blend radius be. After consulting with the engineer, he indicated that data points to the optimum blend radius being 1/4 the span of the winglet. For the Long Ez that is a 12″ blend radius. The previous builders that have attempted this mod have used smaller radii.
With that decision made and a set of my own plans, I was able to determine a list of materials required to complete the modification. Those materials were blue foam to shape the cores, materials for a new dipole com antenna, BID and UNI fabric, UNI spar cap tape, plywood, 1/8″ aluminum, and rod end bearings.
1. The very first step in actually starting the modification was a big one; Cut off the winglet and 12″ of the wing. That is right, cut off 12″ of wing. In order to maintain the same wingspan, this is necessary. But what about the lift lost from the remove wing section? The blend core lift vector projects a component of lift both in the vertical and horizontal directions so the blend acts to restore the vertical lift.
2. The structural attachment for the new winglet and blend is to extend the wing spar through the blend and into the winglet. Therefore, the next step was to prepare the wing spar for scarfing on a spar extension through secondary bonding. In order to do that, I had to create access to the spar beneath the wing skins. So I got the trusty sharpie out and marked the forward and aft edges of the spar cap on top and bottom. I also marked inboard from the new ‘wingtip’ stations where the spar extension plies will start.
Next was modifying the existing spar to prepare for the extension. I first sanded down through the wing skin over the spar cap. This sanding was done as to create a taper in the existing cap from close to zero thickness at the new wingtip, to existing thickness 12″ further inboard. This was done both top and bottom. Then I cut an inch of the wing out in front of the spar to allow access to the spar shear web. This shear web was prep-sanded as well.
3. I started this step by leveling the wings in both pitch and roll and jigging them to a table, which was also leveled in pitch and roll. This is important because it gives a method by which to ensure the winglets will end up properly positioned. Now on to fabrication. I started by building up a block of blue foam that is large enough for the blend core. How big is that? It is as wide as the minimum blend radius plus the max thickness of the wing, plus some margin. Sound complicated? It’s not, but to be sure, I made and measured a template.
I made a hotwire template for the wing max thickness station. Basically this template is a quarter of a circle with radius 12 inches. Then I measured the thickness of the wing, and made the template that thickness throughout the quarter circle. These dimensions and arcs were drawn on the plywood templates. I then add a ‘foot’ to the template. The purpose for this foot was to give a convenient way to accurately jig the template, and the foam core to the wing and table. Notice that the template for the blend is not 90 degrees of a circle. It is actually 91 degrees of the circle. This is because the plans winglet is canted 1 degree inboard.
Alright, so we have our block for the blend core measured and cut to raw size. Next I used my blend core templates to hotwire the shape you see here. Recall that since there are compound curves in every direction here, this shape is only accurate for one station, the max thickness of the wing. However, the leg of the foam core gives a convenient way to hold the blend core above the table accurately, and the area where the winglet will attach is canted 1 degree inboard accurately in preparation for the winglet core.
4. Next we need a winglet. So I used the plans method to hotwire winglet cores. This was done exactly per plans until I had a full winglet core. Now again, we have a 12 inch blend radius. So, we need to hack 12 inches off the bottom of the winglet in order for the winglet tip to be in the same location as the per plans winglet. Just keep the winglet cores in the blocks, measure up 12 inches from the root, and hotwire them off. Additionally, we need to cut 8 inches off the top of the winglet. This is to facilitate the spar construction, and will be bonded back on.
5. Now that we have both cores, it is time to locate the cores to each other and the wing. This was done by drawing a line representing the chord line of the winglet on the top of the blend core. This line was drawn so that the toe-in of the winglets were at the same angle as the plans winglet, and so that the max thickness station of the winglet was flush with the inboard edge of the blend core. Now we have the angles taken care of and the left/right location taken care of.
THIS IS WHERE I FOULED UP… BIG TIME. Now we need to determine fore and aft location of the winglet on the blend core. Now, from discussions with other people, in order for this to look and work right, I knew I needed to move the winglet aft to get a continuous leading edge. It will become apparent why in a little bit. From the discussions I had, and from the TLAR (That looks about right) method I decided I needed to move the winglet aft 4.5 inches. So that is what I did, and carried on.
My biggest mistake was to move forward with this 4.5 inch shift. I didn’t quite understand why I needed to move the winglet aft or how far, but deferred to people who had been there before me who told me I needed to move the winglet. I did so without a full understanding of the problem requiring this shift, much less the impacts of shifting the wrong amount. But lets continue forward, as I did.
The winglet is now located on the blend core. I took many measurements from the centerline of the airplane, on the instrument panel to the leading edge and trailing edge of the root of the winglet as well as the trailing edge tip to ensure they were the same both left and right. A stick is then bondoed in diagonally from the trailing edge of the winglet to the trailing edge of the wing to lock the winglet into place.
6. Next I decided to start carving the blend. I decided to start carving the blend before microing the blocks together so that I wouldn’t have the difficulty of carving over micro lines. This also turned out to be a mistake and complicated the carving more than aiding it. So I stopped carving, and microed the blocks together.
You can see from the next photos that I started the carving by making sort of a relief map. I measured the excess foam of the blend core and connected areas of equal excess on the wing side, to areas of equal excess on the winglet side. I then used a dremmel and a sanding drum to remove 80% of the excess foam in these areas. (Thanks TMANN for these ideas) These steps were then blended together by hand using 12 inch radius piece of the foam that was cut out from the blend core. I started this process on the inboard/top side.
Then I moved to the leading edge and worked the leading edge of the blend between the wing and winglet. This was done through the same process. I removed areas of major excess foam then hand sanded to blend the two airfoils together. It took a while, but eventually I ended up with a shape and it thought I was ‘there.’
Something looked not quite right to me, and I justified it by saying I am just not used to it yet. I was reassured by overly nice people who visited the shop to looked at it and said, “looks great.” That is until I posted pictures of my ‘pride and joy’ on the Canard Aviation Forums. Right away I got comments on things like: whats wrong with the leading edge, why are the leading edges not continuous, why does it look like a hump back whale? Then it hit me. The leading edge of the blend moves forward and changes angles as it moves outboard. It does look like a hump back whale. SHIT!
Remember when I told you about my major mistake, not knowing how much or why to shift the winglet aft, this is why. A member of the canard aviation forum reminded me of something that I should have remembered from high school geometry. Two lines in space define exactly one plane. That is to say the leading edge of the wing and the leading edge of the winglet define exactly one plane. Therefore they need to be positioned relative to each other so that both of the leading edges are in the same plane in order for the entire leading edge, wing through the blend to winglet, to be continuous.
To this point, I had probably two months of off again, on again work into the blended winglets. I was faced with a choice. Do I start over and do it again the right way, or do I live with what I have. I had people persuading me to go either way. But every time I looked at the winglets, I saw a hump back whale. I knew I wouldn’t be proud of those winglets when it was parked on the flight line at Oshkosh.
I called a mentor of mine, Dick Keyt, and shared my lament and dismay over my hump back leading edges. I was leaning toward just moving forward as is when he told me,“The answer is simple. If you are not proud of it, you do it again.”
So that is what I did. I got the handsaw out this time, and hacked the foam off. It was now time for the third set of winglets for the Long Ez.