And misshapes along the way.
fin airfoils vs other shapes
- Thread starter lr64
- Start date
Help Support The Rocketry Forum:
And misshapes along the way.
Admittedly I've only done it on LPR balsa fins, but when I've done root to tip thickness tapers (for a scale WAC Corporal / Tiny Tim), I stole a tip from the C/L stunt community, who taper elevator stock in this manner.
I used music wire affixed to my workbench. One piece of a diameter to match the fin thickness at the root, then two more pieces; one the diameter of the desired thickness at the tip and a second length that 'splits the difference' between those two. For example, 1/8" diameter/thickness for the root, 1/16" diameter for the tip, and a piece at 3/32" diameter.
Butt the root of the fin against the 1/8", and set the 3/32" diameter rod at the tip. Use those as guides with a sanding block to 'plane' the fins down. Remove the 3/32" rod and replace with 1/16", flip the fin over, and do the other side...
I would not want to attempt the same with aircraft plywood fins, for those I am quite content to do leading and trailing edge bevels.
Coming back to this, the point isn't to shame anyone about how they make fins. Make 'em however you want, they'll work of course. The idea is that with a little bit of effort you can reduce the drag on your rocket substantially, even more so with more effort. So if that's something you'd like, what's the right approach for you?
In terms of typical subsonic model rockets, the main thing I'd like to bring awareness to is that rounding the fin leading edge makes everything much better aerodynamically, definitely for drag in terms of keeping flow attached, and I think also for lift—and therefore stability with minimum drag—as mentioned by others above. If you are OK with not having the edges square, if you do nothing else, rounding the leading edge is the most important thing.
Here's a rough idea of what the airflow might look like, comparing a square-edged fin to one with a rounded leading edge as well as to a fully airfoiled fin, all at a slight angle of attack (flow going from left to right of course). Rounding the leading edge gets rid of the flow separation bubbles (local stalling) seen in the case of the square-edged fin. I don't have any specific wind tunnel or simulation data for this, it's just going on what I've seen before.
So as a practical matter, here are some different levels of effort versus drag payoff, numbered from least to most, with an example balsa wood fin:
One time I got so zealous about sanding a shallow trailing edge taper that I didn't notice I was also sanding the top surface of my thumbnail. I stopped when it was paper-thin in the middle, fortunately before it was sanded through. Felt weird. I don't recommend it.
In terms of typical subsonic model rockets, the main thing I'd like to bring awareness to is that rounding the fin leading edge makes everything much better aerodynamically, definitely for drag in terms of keeping flow attached, and I think also for lift—and therefore stability with minimum drag—as mentioned by others above. If you are OK with not having the edges square, if you do nothing else, rounding the leading edge is the most important thing.
Here's a rough idea of what the airflow might look like, comparing a square-edged fin to one with a rounded leading edge as well as to a fully airfoiled fin, all at a slight angle of attack (flow going from left to right of course). Rounding the leading edge gets rid of the flow separation bubbles (local stalling) seen in the case of the square-edged fin. I don't have any specific wind tunnel or simulation data for this, it's just going on what I've seen before.
So as a practical matter, here are some different levels of effort versus drag payoff, numbered from least to most, with an example balsa wood fin:
- Square-edged fin: For ease of build, looks, etc.
- Round off just the very edges: Helps with drag a bit. On the trailing edge it doesn't do much because the airflow is going to separate back there anyway, so it's mostly for matching looks. Also it makes the edges less prone to denting damage than if they were sharp square.
- Fully round edges: On the leading edge, helps with drag. Again rounding the trailing edge doesn't really help, just for matching looks, so it's optional.
- Elliptical leading edge, tapered and blended trailing edge: Very good for drag. In this example the trailing edge is shown with a blunt face for durability; it could instead be sharp for even less drag. Blending the trailing edge taper into the flat part of the fin (rather than making a flat angled trailing edge facet, like on a supersonic airfoil) helps keep the flow smoothly attached all the way. Note that the tip airfoil is proportionately thicker than the root airfoil, so the tip is probably draggier than it could be.
- Tapered fin, elliptical leading edge, tapered and blended trailing edge: Very good for drag. Keeps the same proportionality in the airfoil from root to tip. It's a lot more work to sand that taper along the span though.
- Fully tapered and airfoiled: Excellent for drag, with a consistently-proportioned airfoil throughout. The maximum thickness shown is actually slightly greater than on the other fins, since you can get away with that and still have less drag overall.
One time I got so zealous about sanding a shallow trailing edge taper that I didn't notice I was also sanding the top surface of my thumbnail. I stopped when it was paper-thin in the middle, fortunately before it was sanded through. Felt weird. I don't recommend it.
Last edited:
lr64
Well-Known Member
6 could be much thicker with very little or no different drag penalty. If you lengthened the ellipse and lengthened the taper until it was tangent to the ellipse, number 5 would be a legitimate airfoil with reasonably low drag. It's probably not hard to make a constant chord mold in this shape using a very large milling bit* and inclining the workpiece the right way. Inclined another way, you could get the flat taper. I wonder if a drill press would work well enough for this. Of course, if you've got NC, there's no need to be this clever. That's how they made a 7 percent thick mold I designed for some RC sailing guys for their "fins", which is what they call a fin keel. It would probably be very good for rocket fins. Maybe they'd let me borrow it.
*Or a hole saw, or a fly cutter.
______
Thin fingernails can be reinforced with very light glass and thin CA. Be sure to use very little CA and spread it around with polyethylene plastic. If you use too much, or if you use accelerator, you might burn yourself. 3/4 oz is probably about right. Maybe 2 layers, but not all at once. I remember sanding down my nails by accident when making a Jetco glider when I was a kid.
*Or a hole saw, or a fly cutter.
______
Thin fingernails can be reinforced with very light glass and thin CA. Be sure to use very little CA and spread it around with polyethylene plastic. If you use too much, or if you use accelerator, you might burn yourself. 3/4 oz is probably about right. Maybe 2 layers, but not all at once. I remember sanding down my nails by accident when making a Jetco glider when I was a kid.
Nearly forty years ago, I met a classical music guitarist (i.e., a picker, not a strummer), who superglued all of the nails of his picking hand. He probably wasn't a rocketeer, or he might have known about fiberglass.
lr64
Well-Known Member
Come to think of it, someone probably makes wood blade blanks for RC helicopters, maybe even ones that aren't too thick. I think 12 percent is common, which is maybe slightly thicker than we'd like. Or maybe the blanks would be thinner, to allow for a wrap. Some heli blades use symmetrical foils. Although they're really too thick for the low Reynolds numbers involved, some RC sailboat guys use them for fins. Glass or carbon ones might be too heavy at the thickness we're talking about. Completed blades might be ballasted.
Unless one is going for a specific airfoil, I don't think one has to be perfect in shaping the leading edge into a shallow ellipse and making a blended taper on the trailing edge to get most of the effect. I always just eyeballed it while sanding it by hand.
lr64
Well-Known Member
I don't know the relative drag, but a full airfoil can be thicker, and therefore stronger, stiffer, and maybe even lighter if you design for it. I suspect that someone with a good eye can probably make a decent airfoil that way, too, though probably not as good as with the method I linked to earlier.
By now, I'm sure someone, somewhere, has some data on stuff like this.
Try pricing 25 pounds of APCP, or finding it in today's shortage; plus shipping. Next price the liquids and metals for fuel.
lr64
Well-Known Member
In fact, I implied APCP was expensive. But it's not the only way to make a rocket go. Maybe it is if you want to go Mach 2.
Very well done!
For balsa fins with the grain nearly parallel to trailing edge, you should not sand the trailing edge too thin, unless you intend to cover them with tissue. For the fin platform shown, if the grain is parallel with the leading edge, there is enough angle to the trailing edge grain to support a thinner trailing edge. I like to use aircraft plywood where you can make the trailing edge razor sharp and thin, without sacrificing durability. In fact, I often salvage and reuse airfoiled plywood fins.
lr64
Well-Known Member
C-grain, aka quarter sawn is best for fins. Stiffer across the grain and less likely to warp.
So aerodynamics for the trailing edge are pretty much the same for rounded or square? Rounding it is just a waste of time unless you want it rounded for looks?
Based on some simulations I did a while back, I'm compelled to suggest that before people get too enthusiastic about low-drag profiles, they ask themselves just how much fabrication accuracy is required to achieve the low drag, and to compare the result of an inaccurate profile with that of a flat fin with rounded edges. After all, we don't fly theoretical airfoils, we fly fins made more or less accurately with the time, tools, and materials we have at hand.
My sims used xfoil to compare a 64a010 (exact) with the same profile "damaged" with a .005" flat spot of the sort that might be caused by inaccurate sanding (on a 2" chord). What I found was that the damaged fin had very close to the same drag as a flat fin with the same thickness ratio and rounded leading/trailing edges. Speaking as an amateur machinist, I am pretty sure no one is going to "eyeball" a fin to that level of accuracy, and, even if one could, how many fins are you willing to scrap before you get 3 (or 4) that are all accurate (on both sides, for the full span of the fin)?
Then, too, one has to ask whether the contribution of fin drag to the total drag is large enough that reducing the fin drag (with an accurately shaped fin) is going to make a noticeable difference. That's easy enough to test with RASAero and it turns out, motor variations (or wind) probably swamp the difference, at least for the flights I simmed.
I had been quite keen to try a bi-convex fin, but after making a cutter (using a swing arm and my mill), using the cutter to make a mold, molding some CF fin material and seeing the result, I was motivated to do the sims. That's when I decided flat was good enough.
--Steve
My sims used xfoil to compare a 64a010 (exact) with the same profile "damaged" with a .005" flat spot of the sort that might be caused by inaccurate sanding (on a 2" chord). What I found was that the damaged fin had very close to the same drag as a flat fin with the same thickness ratio and rounded leading/trailing edges. Speaking as an amateur machinist, I am pretty sure no one is going to "eyeball" a fin to that level of accuracy, and, even if one could, how many fins are you willing to scrap before you get 3 (or 4) that are all accurate (on both sides, for the full span of the fin)?
Then, too, one has to ask whether the contribution of fin drag to the total drag is large enough that reducing the fin drag (with an accurately shaped fin) is going to make a noticeable difference. That's easy enough to test with RASAero and it turns out, motor variations (or wind) probably swamp the difference, at least for the flights I simmed.
I had been quite keen to try a bi-convex fin, but after making a cutter (using a swing arm and my mill), using the cutter to make a mold, molding some CF fin material and seeing the result, I was motivated to do the sims. That's when I decided flat was good enough.
--Steve
lr64
Well-Known Member
I think the smoothness of the shape is far more important than just how close it comes to some ideal shape. I mean how smooth the cross section would look if drawn out. I've never been able to get Xfoil to handle anything like à flat plate. I'll have to look it up again, but in a publication called Soartech 8, which you can find on line, wind tunnel tests with hand made, imperfect models showed a pretty significant difference between a thin flat plate with a rounded leading edge and tapered trailing edge vs a normal airfoil, though I forget 2hich one. Unfortunately, the Reynolds numbers were only about 300k, which corresponds to a fairly small and slow rocket. What Reynolds and Mach numbers were you using? A surface with a 2 foot chord going hundreds of feet per second will have a much higher Reynolds number, higher than what we"d see with most amateur rockets. Drag coefficients would be lower. Such a rocket would probably also have much more weight for its surface area, making aerodynamic drag of lesser relative importance.
At 500 fps, if I didn't screw up the calculations, a difference in drag coefficients of .010 corresponds to a change in drag of 1.8 Newtons on 20 square inches of fin area. That's a pretty small rocket, which would be quite light. It might cause a g or two of deceleration.
At 500 fps, if I didn't screw up the calculations, a difference in drag coefficients of .010 corresponds to a change in drag of 1.8 Newtons on 20 square inches of fin area. That's a pretty small rocket, which would be quite light. It might cause a g or two of deceleration.
Last edited:
I am aware of the relevance of Reynolds numbers 
I don't have the exact numbers right at hand, but I tended to use ones that corresponded roughly to 200-500 mph at sea level (3e5, 5e5, 7.5e5). I did find, as you suggested, that xfoil tended to have problems with shapes that didn't really resemble true wings, though I sometimes got it to not blow up if I tried different Re and angles of attack. While it wasn't the ideal tool for doing the investigation, it seemed better than nothing.
I suspect you are correct about the smoothness of shape being more important than the actual curve. That was the point of my "dent" that I presented to xfoil--it wasn't about changing the entire curve, it was about changing just a part of it.
As far as the drag itself goes, keep in mind that the profile will affect only the pressure drag and not the friction drag [of the fin itself]. I doubt, therefore, that you'll see a change in Cd of 10 counts due to the airfoil change, but I'm willing to be wrong about it.
Even if the change in Cd is 0.010, I don't get a change in drag of 2N at 500 fps. I used a fuselage diameter of 2" and Cd of 0.3 and 0.31 and got drags of 8.9N and 9.2N.
My point remains, however, that before one goes to the trouble of trying to do something hard, it often pays to spend some time to discover whether the problem can actually be solved given what one has to work with; in the discussion so far, I wasn't seeing that this larger picture was being considered.
--Steve
I don't have the exact numbers right at hand, but I tended to use ones that corresponded roughly to 200-500 mph at sea level (3e5, 5e5, 7.5e5). I did find, as you suggested, that xfoil tended to have problems with shapes that didn't really resemble true wings, though I sometimes got it to not blow up if I tried different Re and angles of attack. While it wasn't the ideal tool for doing the investigation, it seemed better than nothing. I suspect you are correct about the smoothness of shape being more important than the actual curve. That was the point of my "dent" that I presented to xfoil--it wasn't about changing the entire curve, it was about changing just a part of it.
As far as the drag itself goes, keep in mind that the profile will affect only the pressure drag and not the friction drag [of the fin itself]. I doubt, therefore, that you'll see a change in Cd of 10 counts due to the airfoil change, but I'm willing to be wrong about it.
Even if the change in Cd is 0.010, I don't get a change in drag of 2N at 500 fps. I used a fuselage diameter of 2" and Cd of 0.3 and 0.31 and got drags of 8.9N and 9.2N.
My point remains, however, that before one goes to the trouble of trying to do something hard, it often pays to spend some time to discover whether the problem can actually be solved given what one has to work with; in the discussion so far, I wasn't seeing that this larger picture was being considered.
--Steve
If you hang around EX launches, not many are playing with Sugar, and there are reasons for that. Making your own BP is not an option unless you really want to risk your life.
Making your own Hybrids works also, if you can get low cost N2O. I have seen this coming around again, even in my area. You can print fuel grains. But you still need some AP pre-heaters unless you build a mouse trap of a GSE ignition system.
lr64
Well-Known Member
Other than performance?
Yes
lr64
Well-Known Member
Verboten to discuss even without mentioning composition?
Right, because when we're talking about aft-facing surfaces, even with a rounded trailing edge, the airflow near the surface can't negotiate that sudden of a turn when it gets there. So the boundary layer separates and forms a turbulent wake, not that different than if the trailing edge is square, like in the hand-drawn sketch above. That wake turbulence takes a toll as drag.
Contrast that with a gently-tapered trailing edge, which allows the airflow near the surface to follow along and not separate.
The leading edge is different: there, the air can negotiate a fairly sudden turn because it's flowing from high pressure to low pressure. (There is a limit to how sudden of a turn, hence why it matters to round the leading edge rather than leave it square.) By the time the airflow gets toward the trailing edge, it's "coasting" on inertia from low pressure into high pressure, so if it runs out of steam, especially if you try to get it to make too sudden of an inward turn, it'll peel off from the surface.
It's the same phenomenon that, when it happens on the upper surface of an airfoil, is called a stall.
lr64
Well-Known Member
Square te is better than rounded.
lr64
Well-Known Member
Those Reynolds numbers correspond to relatively small rockets. Somehow I missed that you wrote 2 inches, so my comment about big rockets was silly.I am aware of the relevance of Reynolds numbers
I suspect you are correct about the smoothness of shape being more important than the actual curve. That was the point of my "dent" that I presented to xfoil--it wasn't about changing the entire curve, it was about changing just a part of it.
As far as the drag itself goes, keep in mind that the profile will affect only the pressure drag and not the friction drag [of the fin itself]. I doubt, therefore, that you'll see a change in Cd of 10 counts due to the airfoil change, but I'm willing to be wrong about it.
Even if the change in Cd is 0.010, I don't get a change in drag of 2N at 500 fps. I used a fuselage diameter of 2" and Cd of 0.3 and 0.31 and got drags of 8.9N and 9.2N.
My point remains, however, that before one goes to the trouble of trying to do something hard, it often pays to spend some time to discover whether the problem can actually be solved given what one has to work with; in the discussion so far, I wasn't seeing that this larger picture was being considered.
--Steve
Friction drag can be lower if a substantial part of the flow is laminar. My number is based on multiplying the dynamic pressure times the area times the change in drag coefficients.
Sanding balsa into airfoils isn't very hard.
Edit: I calculated again and got 1.8 newtons again. What was the area of your fins?
Last edited:
- Joined
- Jun 6, 2011
- Messages
- 7,985
- Reaction score
- 2,126
Damn, buddy! I loved that place! I didn't get half the books I wanted and could spend hours in there, much to the Viking Princess' dismay. I also spent too much, but hey! It was like cocaine...it kept me quiet in the corner with a glazed look on my face.
lr64
Well-Known Member
You're assuming that it's hard to do. Big assumption. I suspect it's easier and less expensive than some other things people do for better performance.
I've found it handy to use a straight edge and a single light source to cast a shadow on the surface, revealing the cross section. Also to draw chord wise lines with a soft pencil. I used these tricks for a glassed foam replacement wing tip on a 3 meter glider. I used some of the fancier tricks found at charlesriverrc.org to shape a 36 inch span glider wing that came out very pretty. That wing had a curvy taper in both planform and thickness. I think it took me about 10 hours. I've been afraid to mess it up by continuing the build ever since. No doubt a straight taper, or no taper, would have saved a lot of time. Also, an airfoil that was elliptical in cross section, with tangent straight lines to the trailing edge, ought to be much less work. That ought to be fine for subsonic speeds. See Mark Drela's HT08 airfoil for something similar. Xfoil has a hard time telling the difference, at least at low Reynolds numbers. I think TLAR would probably yield fine results, though.
I could see making a dummy rocket, catapulted with a humongous piece of rubber or, better, a bow, as a crude method for testing fin cross sections without burning up expensive motors. The dummy could have an altimeter inside. No doubt there are other easy ways to get qualitative results.
Rschub
Well-Known Member
I'm gonna make all my fins like this:
- Joined
- Aug 6, 2022
- Messages
- 4,637
- Reaction score
- 4,768
Malemute still does it that way, if I'm remembering the right name with the right fins.
I recognize that airfoil !
It looks like the good ole X-15 Mach 5+ hypersonic wedge airfoil
How fast are you flying ?
-- kjh
Rschub
Well-Known Member
The beauty of this shape is that it doesn't matter.
Similar threads
