Insanity Plea: A sub minimum diameter M2245 project.

I've been thinking about that. The fins had ~9.5 square inches of mounting surface. The epoxy was rated for 4500psi. 9.5 x 4,500 = 42,750 lbs of total holding force. The fins weren't that heavy. They weighed ~3.5 oz or .218 pounds each. 42,750 / .218 = 195,428 gees. I'm pretty sure it wasn't spinning enough to produce those kind of G forces.

ChrisAttebery said:

I've been thinking about that. The fins had ~9.5 square inches of mounting surface. The epoxy was rated for 4500psi. 9.5 x 4,500 = 42,750 lbs of total holding force. The fins weren't that heavy. They weighed ~3.5 oz or .218 pounds each. 42,750 / .218 = 195,428 gees. I'm pretty sure it wasn't spinning enough to produce those kind of G forces.

Click to expand...

There are different ways things are measured/quantified. You say rated at 4500 psi- what measurement? This is essentially a lap joint so I think lap shear would be what we need to look at.

With this spin rate I ask a couple questions ( I hate to ask these questions as I really am not trying to be rude). First, were the fins on straight? Second, were the fin bevels the same on each fin? Finally, were the fins strong/thick enough to not flex?

4500psi was the shear strength.

The fins were machined to be conformal to the motor casing. I used a wrap/mask to etch the case and it was put on very straight. The overlap along the 14" of the mask was even. The fins were CNC machined to be symmetric. The fins were as straight as I could machine them. I block sanded them to remove machining marks, but that shouldn't have caused a major difference in the shape of the fins.

For the record my Blackhawk 75 also had a spin. The fins on both rockets were the same shape and machined from the same model.



I have a couple theories about what happened. I think that the NC coupler failed. The only parts of if it that I recovered are the little bit that was still epoxied to the nose cone adapter and a wad of CF tow that was tangled up in the drogue. The sequence of events could have played out like this. The coupler or the airframe adapter failed. The nose cone turned and the drogue deployed. The NC hit the motor and at least two of the fins as it slowed down. Fins popped off, drogue deployed and turned the motor popping the remaining fins off. The shock cords snapped and the motor tumbled away.

The fins were probably too thin. They were ~.030" at the leading and trailing edges. The thickest point at the tip was .125" and the base was .1875". One of the recovered fins has 80% of its leading edge curled over about 3/16". That may have been the cause of the break up or it may have been a result of it. There's really no way to know.

Chris, I looked up the epoxy data sheet and saw the lap strength was as you said. What I am wondering is that was measured at 73 f. The temperature of the motor case just sitting in the sun would be higher than that by a fair amount. Throw is the motor burning and rapidly heating the case. I could not find the Tg of the epoxy. I also do not see info as to the change in lap shear strength with temperature increase.

Either way, if you lost the nose cone or other parts all bets are off.

Either way, I think the idea of this rocket is great. I like the machining, and in no way meant to be rude/insulting with my questions.

It is a shame that welding aluminum is a little messy. That might be the best way to attach fins like this, but that would likely cause issue with the case, much less make it a single rocket sort of motor.

Neglecting oblique shockwaves at nosecone for supersonic flight due to lack of data. The total temperature at Mach 3.5 is 1,390.91 Fahrenheit using advanced fluids compressible gas dynamics total temperature formula. If you have a nosecone angle I can recompute this for oblique shock which should lower estimate slightly. Also I assumed room temperature start point which is incorrect for Black rock conditions. If you have a actual launch date temperature that may help. That’s the tip of the theory world of what your facing.

What about Cte issues? The coefficient of thermal expansion will be fairly similar for both parts I assume, since they are both more or less are aluminium. Motor casing heats up during flight, probably expanding a significant amount. The fins are in the airflow getting a good dose of convective cooling off the not insignificant airflow. They are also in the lew of the shockwaves largely, so a high total temperature of a normal shock wave probably is significantly cooler in the post-shock area where most of the fin surface area is. So that is two things pointing to significant temperature differences, and change rates, during the flight.

Also consider the epoxy between the surfaces as an insulator. There will be only a low heat flux through that and only for a short time. Sounds like a potential differential expansion issue, with some other not insignificant forces at play.

Let me know if I got any of that aerodynamics stuff wrong. I am an electronic engineer

Did you have any idea how the high loads were on the NC coupler? Any chance they exceeded the strength of the design? From what you posted the parts looked pretty thin...

Great work on this build , unfortunate end to the project.
MK2 will be an absolute BALLS tearer I suspect.

ChrisAttebery said:

I've been thinking about that. The fins had ~9.5 square inches of mounting surface. The epoxy was rated for 4500psi. 9.5 x 4,500 = 42,750 lbs of total holding force. The fins weren't that heavy. They weighed ~3.5 oz or .218 pounds each. 42,750 / .218 = 195,428 gees. I'm pretty sure it wasn't spinning enough to produce those kind of G forces.

Click to expand...

Another thing you could try is friction stir welding to attach the fins:

https://www.researchgate.net/public...ance_of_a_Homemade_Friction_Stir_Welding_Tool

This is often used in aerospace because it doesn't melt the metal, so retains much of its strength at the weld.

Bob Clark

RGClark said:

Another thing you could try is friction stir welding to attach the fins:

https://www.researchgate.net/public...ance_of_a_Homemade_Friction_Stir_Welding_Tool

This is often used in aerospace because it doesn't melt the metal, so retains much of its strength at the weld.

Bob Clark

Click to expand...

Welding by its very nature melts two materials together, friction stir welding just does it without overheating the metal too much, yes it is better than arc type welding for maintaining material properties.

Don't mean to be a jerk but welding does not "melt" two materials togerther. It is a material joining process. In some cases material does change state to liquid and then resolidifes, hence melting would be correct description in that case. In the case of friction stir welding the materials do not change state. Friction stir welding is a solid state welding process, meaning the materials never liquify or "melt", which is why the material properties of the parts being joined are not affected as much.

There are some aluminium brazing rods that work really well. I used them when making some sidesteps for my 4x4. Unlike welding, they tend to melt and work a few hundred degC below the melting point of the Al, so that might be an option. Not sure how important that temp region is to the material properties of the case alloy. Ask a metallurgist . I know they have been used on a fincan that I think flew on an N5800 successfully.

OverTheTop said:

I know they have been used on a fincan that I think flew on an N5800 successfully.

Click to expand...

Correct. That was Mike Passaretti's "Don't Debate This" N5800 flight from BALLS 21 which was the first successful N5800 MD flight.



Build Thread.
https://forum.ausrocketry.com/viewtopic.php?f=10&t=3593

That's the one. I was sweating on that flight being a success as I had suggested the braising .

OverTheTop said:

There are some aluminium brazing rods that work really well. I used them when making some sidesteps for my 4x4. Unlike welding, they tend to melt and work a few hundred degC below the melting point of the Al, so that might be an option. Not sure how important that temp region is to the material properties of the case alloy. Ask a metallurgist . I know they have been used on a fincan that I think flew on an N5800 successfully.

Click to expand...

You guys need Gerald or GT back here to explain this stuff better. The temperature will affect material properties of aluminum significantly and can alter tempering of the metal you ordered. It’s not funny when the material property tensile strengths drop by gee 30-40%. I’m referring Gerald because he has experience with it. Temper matters a lot. Screwing up temper on Metals is bad. Very bad. He pulls all these thermal temperature to strength charts out for different grades of aluminum and well it’s scary. Brazing should be fine. Welding sh*t gets real.

https://www.rocketryforum.com/threads/high-strength-aluminum-alloys-for-motor-casings.146437/

In Gerald’s posts he references 2024T-3 over some common aerospace alloys and thermal reasons why. 7075T6 is not arc weldable that I know of it microcracks.

OverTheTop said:

What about Cte issues? The coefficient of thermal expansion will be fairly similar for both parts I assume, since they are both more or less are aluminium. Motor casing heats up during flight, probably expanding a significant amount. The fins are in the airflow getting a good dose of convective cooling off the not insignificant airflow. They are also in the lew of the shockwaves largely, so a high total temperature of a normal shock wave probably is significantly cooler in the post-shock area where most of the fin surface area is. So that is two things pointing to significant temperature differences, and change rates, during the flight.

Also consider the epoxy between the surfaces as an insulator. There will be only a low heat flux through that and only for a short time. Sounds like a potential differential expansion issue, with some other not insignificant forces at play.

Let me know if I got any of that aerodynamics stuff wrong. I am an electronic engineer

Click to expand...

You can design the metal thickness to account for thermal expansion this tactic is called interference fit and use coefficient of thermal expansion to make the parts thermally lock into place when heated. The thermal expansion is known. Think SR-71 fuel pipes or l3 rocket named don’t debate this has interference fit fins. There is not convective cooling. The fins are being heated by supersonic flow. This is an aero heating problem. Air friction is heating the entire airframe. Yes in places it varies. Tips of nose and leading edges of fins hotter than middle sections. See above total temperature post on how hot this thing gets theory wise at Mach 3.5. The fins act as a heatsink from casing briefly until the rocket is going high enough Mach that the air friction becomes convective heating. The temper of the aluminum itself may change due to extreme heating if prolonged. There is not a normal but shock but weak oblique shocks at fins and nosecone. My normal shock calculation is higher than what really happens without more data from OP such as a radius and length of nosecone to actually account for actual oblique weak shocks. Normal shock is for windtunnel test pipes etc ie not rockets which are sharp objects by compressible gas dynamics theories. The rest of your posting seems okAy. For an EE guy you have a good idea that thermally this thing is screwy.

Predicting thermal loading and heat transfer through curved or complex geometries requires a computer grinding away. At undergrad level we had thermal sims crash our crummy desktops. Even a simple block with hole took an hour plus. Lol. You can model heat transfer through a nodal analysis approach by hand and differential elements but it only works for flat surfaces by hand. Then you need computers for more complex models. I’m still undergrad mech.

Andrew_ASC said:

My normal shock calculation is higher than what really happens without more data from OP such as a radius and length of nosecone to actually account for actual oblique weak shocks. Normal shock is for windtunnel test pipes etc ie not rockets which are sharp objects by compressible gas dynamics theories. The rest of your posting seems okAy. For an EE guy you have a good idea that thermally this thing is screwy.

Click to expand...

Thanks for the thumbs-up. I think I implied that the temps would be a bit lower than the total temp for a normal shock wave on the bird in question, I know I meant to anyway. I was tapping that out on the train on the way home from a Royal Aeronautical Society lecture. Not the most ideal environment for thinking.

Isn't there a massive drop in enthalpy across the shock boundary? That boundary gets set up by the leading edge of the fillet and fin (or the nose at the front). Behind it has to deal with a lot of heating by any disturbed areas of the boundary layer (convective transfer) caused by impinging and interacting shock waves. Areas where the boundary layer is laminar should be relatively tame in heating terms I feel. I don't think radiative heating at that velocity would be very big. I don't have much of a feel for the amount of drop across the boundary and the amount of coupling in practical terms unfortunately. Not my bailiwick.

Yeah. Finite element analysis takes a lot of time. I did some thermal simulations (hand-crafted programs!) back in 1980 for my year 12 Computer Science subject. The computer was a DEC PDP-8.

Sorry I haven't replied to this thread in a couple weeks. I've been a busy boy.

I've been looking over the parts and thought I'd post some pictures. First off is the nose cone:

This is the fin that the leading edge curled up. I'm sure it got hot and was too thin. This was the lightest and thinnest fin on the rocket. The leading edge curled up 270 degrees towards the front.

I just got this one back today. It clearly got hit by the nose cone. The dent matches the radius of the FG where the NC tip should have been.

Thanks for sharing those photos. Definitely shoes the extreme forces the flight was enduring! Do you have plans to redesign or attempt it again?

The nosecone is interesting. Do you think the damage is thermal? Obviously the tip is gone. I wonder two things here. First, would the tip be better served being part of the cone versus a mechanically attached piece? Maybe a tip cast in place made from a high temp ceramic. Secondly, do you think an ablative coating on the nosecone might help?

The fins are certainly messed up. They clearly need to be thicker, or they need to be a sturdier material. How do you feel about machining titanium?

I’m sure the damage is thermal. I think the tip needs a long shoulder that can be epoxied in place. Yes, I think an ablative is needed.

I think the fins would work if they were a bit thicker. Titanium is out of the question. There’s no way I could machine it.

Wow! Amazing pics.

I love the way that fin leading edge seemed to have started peeling near the root and then worked along.

Do you think a thicker leading edge would help? The thin edge has almost no thermal capacity or conductivity so it would heat up quickly, thus losing any strength. I think the thin edge might also allow the normal shockwave to approach it more closely, increasing heatloading. Remember, when they were developing warhead reentry they had to go to blunt shapes to survive the atmosphere by keeping the shockwaves further away. A thicker edge would increase drag marginally but keep the normal shock a bit further away I think. A knife-edge is in theory optimal for drag, but practicalities get in the way sometimes. Any aerospace engineers out there please feel free to chime in on my comments and correct where necessary .

I love the ablative coating you used on the NC .

The Lockheed Martin Mi-15 and MA-25S ablatives while pricey are X-15 ablatives and were used on some areas of the space shuttle. It’s a spray on paint like speciality ablative. High dollar crapola. You can get the repair kit sizes for smaller projects. It would protect aluminum from severe heat loads. Single use. Thermal protection overkill. Missiles often use one or the other. Normally it’s a combined system for maximum performance. You may be able to get Mi-15 easier as it has aircraft repair applications. You might be able to compare spec sheets to a cheaper ablative.

Mi-15 data sheet.
https://www.lockheedmartin.com/content/dam/lockheed-martin/space/documents/thermal/MI-15 product sheet.pdf

MA-25S data sheet.
https://www.lockheedmartin.com/cont...ce/documents/thermal/MA-25S product sheet.pdf

The rolled leading edge is impressive and neat looking. Definitely a keeper when it comes to rocket related trinketry!

The Lockheed Martin Mi-15 and MA-25S ablatives while pricey are X-15 ablatives and were used on some areas of the space shuttle. It’s a spray on paint like speciality ablative. High dollar crapola...

Click to expand...

It would be a shame to put an ablative on nice aluminum fins when all that's needed is a dimensional adjustment. And you certainly don't need "X-15" ablative. That stuff is just pencil eraser in sprayable form. There's a few effective recipes for homebrew ablative rolling around here.

How uneven is the erosion on the nose cone? I’m thinking that the root cause is the curled up leading edge. This would greatly affect the fins ability to function causing a high angle of attack. As the rocket turned the nose would be ripped free as it is dragged sideways through the supersonic air and the coupler is simply not designed for that load.

I was wondering about the NC damage. Your hypothesis regarding non-zero AoA after fin loss is plausible.

Those are really interesting pictures that show how powerful aerodynamics can be.
I found this reply that I set aside to consider. I think parts of it may still be relevant. I really think the thickness and stiffness of the fins are key to the failure.

ChrisAttebery said:

I've been thinking about that. The fins had ~9.5 square inches of mounting surface. The epoxy was rated for 4500psi. 9.5 x 4,500 = 42,750 lbs of total holding force. The fins weren't that heavy. They weighed ~3.5 oz or .218 pounds each. 42,750 / .218 = 195,428 gees. I'm pretty sure it wasn't spinning enough to produce those kind of G forces.

Click to expand...

That might be what it’s rated at maximum, but I’d be really surprised if the glue joint each fin could withstand the weight of ten cars.
You had a whole bunch of stuff going on, flutter, which exercises a tiny portion of the bond at a time and caused parts to bend away from each other, causing peeling and stress risers; obviously spinning, which causes a lot of stress and vibrations in the airframe; and heat effects.
Something that I believe also happens during fin flutter is heating in the portion of the fin that experiences the most flex. Usually that’s near the root edge, right where the epoxy is.
I don’t know that any one of these is responsible, but in an extreme flight like yours, a combination would exist.