So this blog starts off with some bad news. Since the end of my last post on this project, I have been completely swamped and have not had time to get fuselage manufacture working. Some background: The last month of the semester is a nonstop grind for any classes with a final project, and needless to say I was working nonstop for the entirety of November and December. Now that I’m in Winter break/ IAP (independent activities period), I will have *so* much more time to work on my projects; and work I will.
Back to the fun.
Since my last post I hunkered down and worked more on manufacturing the wings, and have quite a lot to show for in that department.
What follows is the result of about 4 hours of work in the lab of carefully cutting a replacement wing for the one I bricked in the last post, coating all wings, and finishing them to have nice leading and trailing edges, as well as removing bubbles.
First Came cutting out the new wing. At this point I was running really low, dangerously low even, on that light load foam I had managed to scrap. Placing one of my current wings on the block confirmed my worst fear: the block wasnt long enough; chordwise that is, I had enough for 2 or 3 tries in the spanwise direction. However, I would need to do this not only without lead-in, but without a cooling loop.
What do those words even mean?
These terms come straight from the jank application we use on the departments foam cutting computers to cut our wings, and each has a special purpose. Well, sort of the same one, but…
As I write this I don’t currently have access to the computers, so instead I will use my inkscape skills to diagram this out. Below, we have an airfoil. Now to cut this out, our cnc foamcutter would guide the wire along this path, right? Ideally starting and stopping in one place?
In reality we want to offset from this path. Why? Well our foam wire cutter isn’t infinitely thin, moreover, it actually melts an area of foam all around it. It is very important we get the same shape out of our machine that we designed, and since our machine isn’t smart, it is up to us to find a way to compensate. So we put an offset of some sorts around our airfoil and use that path instead. One might suggest uniformly scaling the airfoil about the center, and using that path. So let’s try that.
Well, I hope you see the issue here: This doesn’t line up nicely at all! This comes down to a common misunderstanding people seem to have in my experience when it comes to airfoils for manufacturing: they aren’t really shaped nicely for scaling. Their combination of concavity and convexity, combined with their overall long and narrow nature means a uniform scale just stretches it out. We can also see issues like on the left, where the new path seems to actually go over a different part of the previous airfoil. This is no good. Because our scaling is uniform, It is proportional to the dimensions, so it gets shifted more in the x axis than the z axis. There has to be a better way… Perhaps some way to get points that are all the same distance away from the surface?
To some this may have all seemed obvious, but this is one of the many things that make me love vector mathematics. Essentially, We want a curve outside of our airfoil, that is tangential to it at all points, and an equally distance from the original airfoil at all points. To do this, we can use the derivative of our airfoil curve at each point to construct an orthogonal vector (perpendicular) to the surface at each point. Giving these the same magnitude will make them all mark a new point the same distance away. I made another diagram below to illustrate better.
And just like that we have a path to cut along. However, we have one issue: our foamcutter doesn’t move at a constant speed. Why? well around complex curves like near the leading edge, the machine slows down in order to accurately move around that high curvature area. A secondary issue is that when we eventually get back to our starting location, typically at the trailing edge of airfoil, we will stop right at a point, and since the wire is still live, we will melt the trailing edge. So to solve this, we need our trusty lead-in and cooling loops!
Okay now I will actually discuss lead-in and cooling loops I promise
Let’s lead in with lead-in. (I would put one of those drum sound effects to point out the joke but this blog is text format so that doesn’t really work) Okay, anyways, lead-in; what is it? Well as the name suggests, we lead in to our airfoil cut, preferably with a straight line, this way our hot wire doesn’t linger anywhere we don’t want it. The end result is this:
So on to the cooling loop. I don’t always use this when making larger wings, but it is nice for smaller wings, as the amount of time the hot wire spends near the leading edge goes up as the wing gets smaller, since we are increasing the curvature, so slower cutting around that edge. The leading edge is very critical for good airflow. It is basically the conditioner for the rest of the airflow. Bad flow on the leading edge is bad airflow everywhere. So We need to let the foam cool down for a bit while we cut the leading edge. The way the cutting software handles it in the lab is by adding a semicircle path after the first half of the cut, letting our leading edge cool down and prevent any warping.
I’ll admit that was a massive tangent, back to your regularly scheduled wing making shenanigans
So I got to cutting that first wing, and well, lessons were learned immediately: I will need to sacrfice a bit of the trailing edge to preserve the leading edge. It of course was learned the hard way when the first cut was cancelled after it didn’t even get to the leading edge before cutting out the other side of the block. Risky trial and error later, and I managed to get one perfect cut out of the remaining foam. Lucky me, that was the last chance I had too.
Bumps in the road: or fun facts about boundary layers
You may have noticed the carbon fiber shims I have inlayed on each wing. These take the relatively weak foam and kick up it’s loading capacity through the roof. There is of course, one issue. I didn’t make their channels deep enough everywhere, so each wing had an ever so slight bump in them. Now this might not seem terrible, but it can be pretty bad depending on the operating conditions. To explain, let’s talk boundary layers.
One of the fun aerodynamics facts I love dropping on people is the fact they are constantly surrounded by a bubble of air they can’t get rid of. Because of the no-slip condition in fluid mechanics, fluids will have zero tangential velocity on surfaces, meaning they stick to surfaces, and along the surface, particles of that fluid will be stationary. Depending on the viscosity, or how thick like syrup that fluid is, we will have differing amounts of distance from that surface before the air is “completely” free from that viscous effect. We call this the “boundary layer” as it creates a boundary around objects going through the air along which air can flow according to it’s potential, without interference from viscous forces. I spent a lot of time fawning over the boundary layer in my aerodynamics class. For example, did you know you can determine the drag on anything just by finding the “lost energy” in the wake of something? It’s crazy stuff, and I love it. OK anyways we will normally have laminar, or smooth flow at small scales such as these, unless the flow gets “tripped” to turbulence.
Laminar and Turbulent?
Anyone familiar with Aerodynamics can feel free to skip right past this section, I’ll just be doing a cursory overview of the two forms of flow.
Laminar is the favorite child in the world of aerodynamics. Less drag, better performance, etc. The issue is laminar flow doesn’t really happen that often in actual airflows. Once our Reynolds number (a non dimensional value characterizing our flow) gets to some point, turbulent flow is almost guaranteed. If you are flying on a commercial jet, essentially all of the flow over the wing is turbulent.
Now I don’t know why but a lot of people don’t seem to like turbulent flow. But I love it. The chaos, the unpredictability, the surprising beauty of it all. Oh yeah also it is *terrible* for performance. So much energy going to waste.
Now coming back to this plane; our Reynolds number should be below the threshold to automatically trip to turbulent, which leaves one potential bump in the road towards optimal performance. Bumps. Literally. Those slight ridges created by the lovely carbon shims are about to ruin my day. See if something protrudes outside of the thickness of the boundary layer (which, mind you, is often very thin), we could trip to turbulent, which will look something like this.
That new turbulence near the boundary layer thickens it, and if we remember my little aside earlier, we know thicker boundary layers go hand in hand with more drag, and that isn’t the greatest thing.
Time to actually build the wings!
Okay so with that in mind, I got to spend the next hour carefully sanding down just the parts of the carbon fiber that protruded from the foam wings. Sadly I didn’t take any pictures, but just imagine a pile of fine black powder atop a table, and me sitting there looking all tired and sad behind my respirator.
Here’s a note to all the makers just getting started. Write things down! About everything! I had to remeasure all of my wings to get the cuts just right on my new one for the flaps, as well as on getting the same angles for the sweeps of each wing.
After the new wing was all cut, and the new shim set in, I had the last, and most painful part. Monokoting the rest of the wings. The red monokote I was using was on the thicker side of things, which means keeping the temperature nice and high, while not keeping the iron in one place for too long. I found a god technique was doing small circles and working your way across the wing. Kind of like brushing your teeth.
I found it was best to do the wings with one piece on the top and bottom each, with a bit of overlap around the trailing edge. Once everything is to your liking using an xacto knife works nicely to slice off the excess plastic. Similar to when performing a two cloth fibberglass layup on the wing, we have the issue of a somewhat sharp leading edge forming due to overlapping pieces of plastic. Unlike fiberglass, we can’t just sand down that edge. Instead of sanding, I cut a bit extra off, and used some sealing tape around the leading edge of the wings, getting a nice clean leading edge.
I decided to try something new with the next set of wings, grabbing some of the black monokote we have lying around. This turned out to be a great learning opportunity on how applying different types of monokote works, and I will try to share all of my wisdom here:
Rule 1: Keep heat proportional to thickness. Thinner monokote can melt pretty easily and although it may seem obvious, the relationship between thickness and heat is pretty sharp, I was using max heat on the red, and min heat on the black.
Rule 2: Work along a path from one side to the other. Having to backtrack is a death sentence when it comes to avoiding bubbles. Don’t be afraid to peel back the monokote to try again.
Rule 3: Don’t keep the iron still! Once the monokote is hot enough for the adhesive to activate, it doesn’t need any more heat, and all that heat from the iron is going to go right into melting the foam.
Rule 4: Pressure is good, But don’t distribute it. I used just the tip of the iron to heat the monokote. Keeping good pressure on a small surface area helped to get the monokote adhered nice and quickly.
I will end off this post with a nice shot of the finished product. Hopefully in my next installment of this series I will have a nice name for it, and get some more CAD work done. ::)