Oops. I just realized I forgot to convert from Newtons to lbs in the last wind shear calculations, so the 375 N converts to 86 lbs, which is right about what Chris calculated the 2nd time. I think we're converging.

It's kind of a shame that you can't just graph or predict dynamic pressure in rocksim though.

But you can graph density and velocity, so it's pretty easy from there.

Great discussion....for the fin planform, there is the option to go with short fins with minimal sweep and as sharp a LE as you can make to reduce pressure wave. Think of an F-104 Starfighter wing or a Nike rocket fin. This is a more efficient lift/drag shape, but has to contend with effective supersonic flow. Option two is to sweep back to 71-75 degrees, have a long root edge, go a bit thicker if needed, and deal with the extra parasitic drag. Does the math just discussed favor one fin planform over the other? Do materials and construction?

I note that the New Zealand commercial sounding rocket goes with option #1, whereas designs like the Shadow Aero rockets go with option #2.

Chris, did the calculations you do figure in the lower density of air and the change in Reynolds numbers?

Anyone know how to calculate skin temperature during flight or to what degree heat would soak into airframe? I seem to recall the SR-71 Blackbird getting up to above 400 degrees F throughout, with the vertical fins at about 1000!

Neat image of rocket at M2.5.

Great discussion....for the fin planform, there is the option to go with short fins with minimal sweep and as sharp a LE as you can make to reduce pressure wave. Think of an F-104 Starfighter wing or a Nike rocket fin. This is a more efficient lift/drag shape, but has to contend with effective supersonic flow. Option two is to sweep back to 71-75 degrees, have a long root edge, go a bit thicker if needed, and deal with the extra parasitic drag. Does the math just discussed favor one fin planform over the other? Do materials and construction?

I note that the New Zealand commercial sounding rocket goes with option #1, whereas designs like the Shadow Aero rockets go with option #2.

Chris, did the calculations you do figure in the lower density of air and the change in Reynolds numbers?

Anyone know how to calculate skin temperature during flight or to what degree heat would soak into airframe? I seem to recall the SR-71 Blackbird getting up to above 400 degrees F throughout, with the vertical fins at about 1000!

I did account for the density, although I didn't bother with the reynolds number. I was going for a rough approximation, just to get a feel for whether the flutter or the shear would be a larger problem. It seems that the flutter is the bigger problem, and I would lean towards the highly swept delta rather than the stumpy straight fin in that respect (I'm not sure how they compare in drag though). For reference, the fin I've been simulating around has been a 12" root, 2" tip, and 3" span trapezoid with a 9.5" sweep (72.75 degree sweep angle, which is good up to about M3.3).

Another question: is it better to mount the fins flush with the rear of the rocket, for maximum CP benefit, or is it better to mount them somewhat forward of the rear end. I know the ShadowAero rockets mount them something like a full diameter forward of the rear of the rocket because it is supposed to reduce drag, but that's the only place that I've heard that from. Anyone know whether it actually does reduce drag to mount the fins forward from the rear end?

Nice pictures, Chad. That really shows why someone might want to keep the fins swept back enough to stay within the Mach cone angle. I agree with Chris that sweeping the fins to 74 degrees would be a lower-risk structural design, both because it's hard to get a long, short-span fin to twist, and also because sweeping the fin back so much makes the fin root create the pressure wave, rather than the fin tip.

John, your description of the fin failures has convinced me to be conservative in the structural design of the fins. It would be great if we can find out some more details so that we might guess as to what went wrong with those failures. You've noted that the problems you have seen are for Mach 2 and higher. Do you think the fins in those cases were swept back 60 or more degrees for Mach 2, or 74 degrees for Mach 3?

I think the reasoning behind mounting the fins some distance from the aft end is the same that drove some jets to taper the fuselage where the fins are thickest. Supersonic drag is minimized when there aren't sharp changes in the cross-sectional area, so you don't want to have the end of the fin coincide with the end of the body tube. When the fillets are relatively large and the fins thin, you can use the fillet taper in front of the fin and behind to to do some smoothing of the cross-sectional area. I was doing that for Violent Agreement, but then the tube aft of the fins got crushed when I had to use a sledgehammer to get it off of the mandrel after my tip-to-tip layup.

There was a discussion on Rocketry Planet in which someone provided a formula for the skin heating. I'd like to find that again. It was comparable to a hot heat gun for the Mach numbers I was interested in at the time, but I don't remember if that was M2 or higher than that.

I'd also really like to know what the motor case temperature profile is. It's possible that the motor case would act as thermal mass to keep the structure cool early in the flight, and then by the time the motor heat soaks through, the peak skin friction heating is already done. The HighCarbYen (good article about it in Rocketry Planet) is designed with an oversize tube and FG centering rings to keep the hot motor away from the fin can. That rocket got to M2.3 and sustained no apparent damage, similar to Violent Agreement at M2.0.

Wikipedia gives the stagnation temperature, for air, as Tstag/Tamb = 1 + .2M^2. So for Mach 3, the ratio of absolute temperature would be 1+ .2*9, or 2.8. So for 300K ambient temperature, the stagnation temperature would be 840 Kelvin, 567 C, or 1052 F. I think there are some factors that make the actual temperature lower than the stagnation temperature. Here is another quote from Wikipedia: "For example, the SR-71 Blackbird jet could fly continuously at Mach 3.1 while some parts were above 315°C (600°F)"

Cotronics has epoxy that has a service temperature of 600F, and has reported 2 kpsi strength at 400 F. It cures with 4 hours of 250F.

I've read that the trailing edge of the fins should be one fin span in front of the truncated end of the fuselage. The idea was to have a clean airflow before the end was encountered and keep the drag bubble attached (bubble separation causes drag to increase with each event). Clearly this is only an issue after boost, as the motor blows away the drag bubble.

In subsonic flight when the fin has a high angle of attack, it is beneficial to have a smooth blend of the fin to the body, kinda like a curved rearward strake. Many kit planes and gliders have this, but I think this may actually hurt in supersonic flight. If it were me, I'd build a small one and buy me some extra margin of stability on the initial boost- there is a lot that the Barrowman equation isn't telling us, and little things like good fillets and smooth transitions can improve the lift coefficient and thus the stability.

John, your description of the fin failures has convinced me to be conservative in the structural design of the fins. It would be great if we can find out some more details so that we might guess as to what went wrong with those failures. You've noted that the problems you have seen are for Mach 2 and higher. Do you think the fins in those cases were swept back 60 or more degrees for Mach 2, or 74 degrees for Mach 3?

The can that I saw shred was pretty similar to an Acme fin can in general shape, though not construction. The shredded can also had a "wedge" shape like you see on Nike, etc. - in other words, viewed from the tip looking at the airframe, it was an elongated diamond. The fins were a number of layers of carbon and kevlar, all vacuum bagged. The fins also had a shorter span than a traditional Acme can.

I couldn't guess at the velocity, but I can tell you that it was a three inch "N" motor that was from a terrific builder. I'd say the shred occured at about 3,000' but that is a total guess on my part.

Not to keep dragging gees into this, but Adrian - your transmitter failed on your last "I" boost. Do you think that was due to gees or other issues?

Thanks for the info.

Maybe the lack of fin sweep was a contributing factor then. The shock wave would definitely be along the whole leading edge, not just the root. Worse, the rectangular-ish shape gives the tip a lot more leverage to twist the fin than if it had a shorter tip length and more sweep. And multiple materials didn't help prevent fin flutter, if that's what it was. Kevlar has more damping than CF, but structural damping is probably a small factor when you have a strong forcing function.

The tracker and altimeter did just what they were supposed to do last flight, despite the high Gs. I just set up the altimeter to turn on the tracker when the main deployment charge went off, and since the apogee deployment pulled the main out early, the main deployment altitude didn't happen until about 15 minutes into the flight. Then I got a nice strong signal until the bird went behind the bluffs.

In february I also flew the same av-bay on an h999 and recorded over 120 gs.

If you're worried about the fuel grains, you can test the effect of g loading by just pressing on the stack. The force on the lowest grain will just be the the weight of the whole stack of grains times the number of gs.

After thinking about it some more, I realized that just compressing the fuel grains wouldn't be a very representative test, considering the higher temperatures, internal pressure, and reduction in mass and volume of the fuel during the burn. It'd be pretty hard to test without flying it.

In my sims so far, I was somewhat surprised to see that this rocket doesn't benefit from any additional mass beyond the minimum airframe + motor. The Cd vs. velocity for Rocksim's nosecones are pretty hokey though, so when I get a chance I'll sim it with my own Cd vs. velocity curve that I measured with my I flight to see the effect of mass. If we use the same nosecone shape, I think we should be able to get about the same overall Cd that I got on Violent Agreement; it will just be scaled up to a larger base diameter.

Disclaimer: From a person with almost zero MD experience, who has absolutely no formal training, and horrible at math.

Going backward in the thread a bit, it doesnt seem like this passes a sanity check...

If I ignore the center layers of the layup, and assume that the middle 6" of the fin is taking all the stress, then with a 0.01" thick layer of uni carbon on each side of the core, and 0.04" between the midpoints of the layers, (for a total thickness of 0.050") then I get 62500 psi stress at the root of the fins. The published strength data on the aircraft spruce & specialty website for a mostly-uni layer comes out to about 90 ksi. In other words, you could put a 430 lb weight 3" from the root a 6" long fin, and the fin would only have to be 60 mils thick, with 10 mil outer layers of uni-carbon, to withstand that force without breaking.

Assume that Chris' data seems to match up somehat.

Also assume that the revised 76lb number is correct force that you have to deal with.

Now, i can buy the above if say, said material is a plane, and we have 6 inch wide fin, 6" of which is hanging off of an infinite slab with perfect material creation and an infinitely strong joint. You place 76lb weight 3" from the root. Assuming CF layup has zero deformation before its breaking point, you then have even load distribution over that 3 inch wide area; 60 mils still seems mighty thin, but lets give CF the benefit of the doubt.

Now do the exercise again, except where at the root, the layup starts curving away at 60 degrees over 1/2". You have just created a lever with a very short load arm and very long effort arm. CF wont deform to spread out the load.

To put it a different way:
It seems like you are imparting the strength of the material in a planar configuration to the airframe-fin joint. Thats a moment arm, and it seems to me that the load distribution is all within a few millimeters of the joint.

60 mils of CF in that configuration with a 6 inch joint is going to hold 76lb standing 3 inches from the root? Go look at something of similar construction and ask yourself what happens if you sit on it (two fins, 120 degrees apart, each bearing half of a 150-200lb load). Then think about how much less of a lever you have created if you have a four fin configuration. Not saying that four fins is a good idea, just trying to illustrate that moment arm.

Disclaimer: read disclaimer at top of post 🙂

Keep in mind that Adrian was assuming that CF could withstand 62,500 pounds per square inch without trouble. That would require near perfect CF work - probably with a high pressure, high temperature cure and top grade epoxy. Realistically, most amateur carbon work could not stand up to anywhere near that (which is why that strength seems so incredible). I'm pretty sure the math works out though. Of course, you'd want them thicker than that for flutter reasons anyways.

Oh, and a fillet should reduce the stress. I don't really follow your statement about the 60 degree curve - as long as the tube is fairly buckle-resistant (which it should be, given that the motor casing would be supporting it at that region), there shouldn't be any problem with the strength at all. Oh, and wouldn't the curve be 90 degrees, regardless of the fin count?

Good question. I'm glad you asked. The quickie analysis I gave was for the flat part of the fin, where the two face sheets are held apart by 0.04" of something, where that something could be G10 or CF or even balsa. It just has to be strong enough to keep the face sheets parallel with each other and keep them from buckling. That allows the face sheets to do all the work of resisting the bending. In both cases the main stress in the material is due to bending, but the flat part of the fin next to the root has has the worst of it, because its lever arm is only 0.04" from center to center.

As you pointed out, the fin root is also in bending. And for a tip-to-tip layup, it resists bending the same way. Starting with the flat part of the fin closest to the tube, as you get closer to tube, the distance between the face sheets gets farther and farther apart, so the face sheets get more and more leverage to resist the bending moment. Now instead of just the 0.04" core, you get into the fin root fillet material, and as long as it can keep the facesheets from buckling, the joint will get stronger as you get closer to the tube. On the tension side of the fillet, the fillet material has to keep the curved facesheet from straightening and pulling away from the fillet. On the compression side, the fillet material has to resist getting pushed inward by the facesheet. It is possible that if the fillet material is too wimpy, it will let the facesheets buckle or bend, and you can get a failure there. But the stress inward and outward is just a small fraction of what the facesheets see, so it's usually not a problem.

In your example, assuming a relatively low friction surface, and a 160 lb load, then each fin would get 80 lbs at a 30 degree angle from pure bending. It would be equivalent to 40 lbs in the plane of the fin, and 70 lbs of bending. So you chose a good example, that's pretty similar to the situation I was looking at. In that case, I would be confident that the fins would hold if I stood on the tube.

I know all the discussion here has been on composite fincans but I wanted to give input about metal ones. I built one for my L3 - it was 2.5" diameter with a 7" root and 4" span. I brazed the aluminum fins to an aluminum cylinder. It survived great except the part about landing in a ravine, one fin being wedged between two rocks while a 10' long rocket fell and created a huge moment arm to bend the fin. It didn't break, just bent.

I made v2 of my fincan tonight out of 1/16" tube and 3/32 aluminum fins. I made two of them and destroyed one. The aluminum fin is the weak part because it becomes very malleable - but only at very high stresses. It took about 175 pounds at the very tip to start the fin bending and I had to get to 325 pounds to have it break the brazing off of the tube. I was impressed.

I have 3 hours into the fabrication of both fincans so they are very easy and not that time consuming to make.

Edward