Upward Propensity: My Level 2 rocket

[CarVac]

This thread will document the thought process, and eventually the construction of, my Level 2 rocket.

This rocket started with the name: I thought up the name after I came up with Mach-ingbird, and I decided to use the name for my Level 2 rocket.

Then, after making my Giga Drill Breaker, I had a 25-inch section of 4" PML phenolic, and my club had a piece of Kevlar that was EXACTLY large enough to wrap it twice, so I laminated it with that for durability. I decided, to give myself an engineering challenge, to not add any airframe to that at all (though of course I could if I chose to).

I decided I wanted the most obnoxiously large fins imaginable, so I made three 12" by 12" flow-cloth-cored, .1" thick CF panels, using 2 weaves and one bi-unidirectional (45-45 degree) layer of CF on each side. The fins I planned were 11.5" root chord, 2.25" tip chord, and 9" span, with 8" sweep.

I was working out the internal configuration, starting out with a 38mm motor mount and a baffle system that would have made it basically the big daddy of my level 1 rocket (It would fit the smallest motors of the cert level, and use baffles, and be very fat and short with huge fins). But then, I realized that with the space available inside, I could make it a dual-deploy and 54mm, and I increased the fin span to 10.5".

However, I found that the motor mount would fit the Aerotech 54/1706 case, which meant that I am almost compelled to level 2 on the Dark Matter K456, but the problem was that the simulations were poking Mach 1, causing a risk of flutter on the enormous, enormous fins. Flying anything other than a moonburner in that case size would make the problem even worse, clearly punching through Mach. Adding weight wasn't slowing it down much, and only sending it higher.

The fins are all ridiculously stiff, but I was still apprehensive of flying it supersonic, until yesterday, when I came up with a countermeasure to fin flutter. I was reading the Wikipedia article on the X-29 forward-swept wing experiment, and it mentioned that they got around the severe problem of flutter by laying up the composite fins so that bending the wing would intrinsically twist it to create a restoring aerodynamic force. I have already laid up the fins, but there is an alternative: counterweights.

If I rigidly attach a small mass (on the order of the mass of the fin) to the fin tip suspended a few inches in front, the mass will lag the motion of the fin tip, causing the tip to twist in such a way as to send it back towards the unbent position. In control systems terms, I took the potentially unstable aeroelastic system of the fin tip and the airstream, and coupled it to an overstable mass-fin tip system (which has a CP waaay behind its CG), rendering the whole system stable.

The way I am implementing this is by not using pure trapezoidal fins, but by having them extend out forward at the wingtips and epoxying a 1/32" thick and 1" wide strip of CF to the edge, where a winglet would be, to stiffen the extension in the plane perpendicular to the fin.

An OpenRocket file is attached. I chose to recess the motor mount because I wanted to give flames a chance to expand a little before emerging from the bottom of the rocket (aesthetics, basically), and for CG reasons.

What do you guys think about the design? About my solution to flutter? I'll detail the planned deployment setup in my next post.

EDIT: I'm not doing anything fancy to resist flutter, so I removed that from the title.

View attachment Upward_Propensity.ork

 

[CarVac]

For my recovery system, I had three priorities: reliability, no airframe breaks (I only wanted to use that one piece of body tube), not exposing any shock cord to black powder charges, and the ability to do motor ejection backup.

In order to do that, I used a method which was initially inspired by Adrian A's chute cannons, but ended up more like the Giant Leap Mariah, as far as I can tell. But in any case, it has significant differences from both, due to the choices I made for reliability, and the fact that this rocket is far from minimum diameter.

In front of the 54mm motor tube, there are a few extra inches of space not needed to fit motors. A coupler tube will be inserted and shear-pinned about 3 inches into the motor mount tube, and about one inch glued into a 54mm chute cannon.

The coupler will contain all of the avionics for the rocket: a Raven and a Telemini. There will be vent holes drilled through the glued portion of the coupler and the chute tube, and the tracking antenna will likewise protrude from there.

The chute cannon will contain, in order of increasing distance from the avionics bay, a piston, a 48" Spherachute, and shock cord. The piston will be attached by really thin cord to the bottom of the parachute.

The chute's shock cord will be anchored on the outside of the chute cannon, but the end of the cannon will be in contact with the bulkhead in the nosecone, so it will have scallops cut out of the chute cannon so as to not crush or abrade the cord.

All of the other shock cords will also be anchored on the outside of the chute tube: on one side will be the cords going to the nose cone and the booster, and the other side of the cannon will have the drogue chute and the main chute.

Deployment will go as follows:
1. A BP charge will go off aft of the AV-bay, pushing it forward and out, and also pushing the nosecone off. If I choose, I could have motor ejection backup, since it goes into the same space. If I'm worried about the volume being too large, such as if I use a shorter motor, I could put in a temporary bulkhead of some sort.
2. The drogue, which was stored alongside the chute cannon, would come out. The nosecone and the booster would hang from one side of the chute cannon, which would be suspended from the drogue on the other side, so the cannon would be oriented horizontally.
3. The main parachute, which up until now has been retained in the cannon with lots of blue tape, is fired sideways by a black-powder charge. That should provide plenty of clearance between it and the drogue to prevent tangles. The piston, which protects the chute, hangs shortly below. The whole assembly then descends to landing.

This meets all of my design criteria. It is reliable because there can be multiple apogee and main charges, with motor backup for apogee, and the chute cannon's charge can be sized to be capable of pushing the entire nosecone off. Tangles will be unlikely due to the horizontal chute cannon. It has only one airframe break, thanks to hiding that internally. Finally, it exposes no shock cord to black powder charges, ensuring a long lifetime without needing to replace components.

 

[CarVac]

Scared off by the walls of text, huh?

It seems like many fewer people have downloaded the .ork file than have looked at the thread, so I'll attach a screenshot of it in OpenRocket to help people visualize.

 

[T-Rex]

Interesting!! I read through the text, but I need to log on with a different computer to make use of the .ork file

 

[cbrarick]

think of each weight as a failure point for your level 2. If you like risk, go for it....if not do something more standard for your level 2 and then try this later...

JMHO

 

[CarVac]

What sort of failure mode do you anticipate for the weights? I didn't think there will be in-flight failure, but rather landing failure. Should I beef up the structure some more?

Or were you worried about tangles? I forgot about that. I could return the fins to normal: they are really stiff and not slender, so they may be fine WRT flutter...

 

[UhClem]

If I rigidly attach a small mass (on the order of the mass of the fin) to the fin tip suspended a few inches in front, the mass will lag the motion of the fin tip, causing the tip to twist in such a way as to send it back towards the unbent position. In control systems terms, I took the potentially unstable aeroelastic system of the fin tip and the airstream, and coupled it to an overstable mass-fin tip system (which has a CP waaay behind its CG), rendering the whole system stable.

I don't think so.

Stability in the CP/CG sense has little to do with this. It is a problem of the airflow exciting natural frequencies of the system. Adding mass tends to lower the natural frequency of a system and in fact one of the NACA reports on the subject said:

"The matter of leading edge counterweights has been investigated, in particular on wing 2C. Figure 23 shows the effect of moving a counterweight along the span. The weight has a rather surprising negative effect near the tip...", NACA-R-685, pg. 129


Adding weight at the end of a spar introduces new vibration modes. Depending on the frequencies of the new modes I suspect that there will be one that makes fin flutter worse even if the added mass doesn't lower the flutter velocity. If you are really lucky you will stay below that frequency. If you are merely lucky the spars will break off. If Murphy pays a visit you will lose one or more fins.

One under appreciated aspect of this is that the problem isn't so much one of too much speed but of speed near the velocity that excites the fins natural frequencies. A long burn motor that keeps velocity near the flutter point is worse than a fast burning motor that quickly accelerates past the flutter point.

 

[CarVac]

That's interesting... I hadn't thought about that fact.

You're right: simulating the fin bending in my head, I can imagine that the slowest vibrational mode could be the fin flapping in phase with the counterweight, which gives positive aerodynamic feedback.

I guess that controlling flutter this way would be best done by doing a custom fin layup that is extremely stiff spanwise but happy to twist, making the lowest frequency mode self-stabilizing. My fin layup is fairly isotropic, so I can't really trust that to be the case unless I make the wing fairly slender and long, not my goal.

So I guess I'll stick to a conventional (though gigantic) trapezoidal fin and hope that its stiffness will survive. When I tap on the 12x12 panels, their resonant frequencies are very high, and can only get higher as I cut some off. I can see how it survives at constant speed on the CTI 54-2G and 3G Classics, and the 4G Green3 if I borrow a case, since they burn long enough at constant thrust to do that. (If only the AT J99 were still certified...)

But I'm glad that it's got a good chance of surviving a high-thrust motor; I was thinking of trying a K1103X after the Dark Matter.

 

[CarVac]

AAAAAAAAAAnd, more than half a year later, I have resumed work on this.

I cut the motor mount long enough for a 54/1706 with extended forward closure, and cut centering rings.

I am almost certainly going to revise the deployment method. If I can figure out a way to have an av-bay that doesn't get pressurized by the ejection charge, I'll do dual deploy with a cable cutter.

 

[CarVac]

Obnoxiously large fins, indeed.

I cut the fin slots today: I clamped the tube to my school's table saw, started the saw, raised the saw blade into the tube, and then lowered the blade. Then I lengthened the forward part of the slot by a half an inch with the saw on my pocketknife. The table saw blade is the same width as the fin thickness, and the length of the cut is just barely shy of the fin tab length. A happy coincidence.

CCotner tells me that my rockets (save for Disappearing Act, which can vanish away (ha ha) into any nook or cranny) tend to defy storage solutions. My L1 scratch squat rocket Chubby has to stand on the floor on its fins. Giga Drill Breaker just takes up room and in certain configurations can roll around, and this one is... has a simply excessively large fin span.

 

[Coop]

I completely want to see this fly.



Later!

--Coop

 

[troj]

CCotner tells me that my rockets (save for Disappearing Act, which can vanish away (ha ha) into any nook or cranny) tend to defy storage solutions. My L1 scratch squat rocket Chubby has to stand on the floor on its fins. Giga Drill Breaker just takes up room and in certain configurations can roll around, and this one is... has a simply excessively large fin span.

But hey, if they make you happy.....

Are those fins nomex honeycomb?

-Kevin

 

[CarVac]

But hey, if they make you happy.....

Are those fins nomex honeycomb?

-Kevin

They are made of this: https://www.acpsales.com/Aero-Mat.html with three layers of carbon fiber on each side. It has great crush strength and the edges have decent integrity, so it should be more durable (though heavier) than Nomex honeycomb. Plus, nomex honeycomb wouldn't end up this thin.

I'm probably just going to seal the edges of the fins with Aeropoxy ES6209, and then gloss-coat the entire fin surface. The nosecone is a 4" plastic PML cone painted bright red (though the months and months in storage have caused a few chips), and the rest of the rocket is going to be natural material color.

 

[CarVac]

Yesterday, I started attaching the fins.

They have a tremendous leverage, given that the fins have 10 inch semi-span and the roots are only 1 inch in, so I am going for maximum strength. To this end, I am first epoxying them on the outside of the body tube, then removing the motor mount, filleting them on the inside of the tube, then adding the motor mount back in, filleting the fins to the motor mount tube, and then gluing the last centering ring back in.

Because of the potentially extreme stresses, I probably will sand down and then fiberglass a portion of the motor mount in order to prevent shear failure in the layers of phenolic paper. I'm much more concerned about internal delamination than a fin snapping, due to the raw stiffness of this foam-cored carbon fiber.

Of course, I'm using Aeropoxy 6209. It is super strong, bonds really well to composites and phenolic, and it cures fairly clear so it doesn't uglify the joint between the Kevlar and the carbon fiber. The cure properties dictate that I wait 12 hours between fillets:

Lay first fillet, and mount the fin guide to keep it perpendicular to the body tube.
Wait 12 hours. At this point, the first fillet is stiff but still susceptible to creep.
Turn to other side of first fin, remove fin guide, lay second fillet, and reinstall fin guide.
Wait 12 hours. At this point, the first fillet is fully cured, so...
In this same orientation, lay the first fillet on the second fin, and move the fin guide to the second fin.
Wait 12 hours.........

It'll take some time.


My one concern is regarding dual deploy: I'm planning on using an Archetype Rocketry cable cutter, with drogueless descent: nosecone pops off at apogee and the parachute stays bundled up.

However, I read fairly recently on some thread that rockets with very large fins and drogueless dual-deploy can glide around, covering long distances. This rocket has more fin area for its size than anything short of a glider, so if it should happen to any rocket, this one would suffer worst.

Any thoughts on this possibility? It's not a deal-breaker if it happens, and it's unlikely to go in a straight line when gliding what with the nosecone and burrito flopping around, so I might just see what happens...

 

[CarVac]

I laid fillet number 4 this morning.

Notice how recessed the motor mount is: I do this for stability and so that the flame will kinda just emerge from the whole bottom of the rocket. I'll cover it in aluminum tape for protection.

 

[bobkrech]

That's not design feature you'll see on professional rockets. Recessing a motor within the airframe causes recirculation of the hot exhaust gases and puts a lot of undesirable heat on the exposed airframe and the motor retainer. It also introduces significant base drag.

Bob

 

[CarVac]

That's not design feature you'll see on professional rockets. Recessing a motor within the airframe causes recirculation of the hot exhaust gases and puts a lot of undesirable heat on the exposed airframe and the motor retainer. It also introduces significant base drag.

Bob

Is there any research on this?

It's actually less dramatic than it looks like, especially on motors with protruding nozzles (not snap-ring cases or Medusa nozzles). Plus since the motor mount isn't glued in yet I still have the option to stick it out farther, at the expense of a tidbit of stability. Or would that not help (the recessed base plate itself causing recirculation regardless of where the motor is)?

Base drag... This is so far from a high-performance rocket that I don't care in the least. I imagine that simulation programs might overestimate altitude in this case, though.


It's going to be fairly tolerant of heat. The airframe is PML phenolic, so I'm not worrying about heat damage to it, and I can coat the inside with high-temperature epoxy and use aluminum foil tape to attempt to reflect heat (though it will get sooted up). It's not going to use motor ejection, so some heat soak on the retainer won't be the end of the world. Or could it?

 

[COrocket]

check out this explanation of krushnic effect https://www.rocketreviews.com/krushnic-effect.html and consider whether you still want to recess you motor. I believe this is documented at the Estes rocket level, where a model just sits on the pad after ignition and roasts on the pad. There is also an NAR report on the subject https://www.nar.org/pdf/TCR1.pdf Your rocket may or may not have issues, but I wouldn't risk it if it was my rocket.

 

[CarVac]

I somehow doubt that a 4" rocket with the bottom recessed only 1.75" and the K motor even less than that, could possibly experience the Krushnic effect, especially when it's not sitting flat on a blast deflector plate.

Heat, yes, but it will never get stuck to a pad.

 

[cwbullet]

What is the blue foam that you have against your fins?

 

[CCotner]

It's not foam, it's a 3D-printed fin guide courtesy of Layer by Layer (https://www.layerbylayer.com/).

 

[CarVac]

This is the one you see here, for 0.140" thick fins:





This one was used on Disappearing Act:

 

[CarVac]

The rocket, with myself for scale. Either the design just feels bigger in person than this photo depicts it, or I'm bigger than I imagine myself to be...

At this point, I have removed the motor mount and centering rings from the inside, and I prepared an internal-fillet-smoothing tool for use starting tomorrow.

 

[CarVac]

Fin glazing time! 1 of 3 fins glazed.

Cell phones don't do very well when taking pictures of black things in dark rooms...

I also laid up a fiberglass tube that couples PML 54mm airframe on the outside, for mounting the av-bay. I hope it turns out well.

 

[CarVac]

My L1 rocket, Chubby, in the foreground with the (empty) Upward Propensity in the background which is waiting for its parts to finish gluing before getting installed.

I redid the simulations, and it seems like it will be adequately stable on any motor that fits, which is nice. Previous iterations of the simulation indicated that it would be less stable on 2-grain 54mm motors, but it's ~0.95 calibers stable.

Normal one: View attachment Upward_Propensity.ork

38mm adapted: View attachment Upward_Propensity_38adapter.ork

 

[CarVac]

I just drilled holes for a Raven in the av-bay, glued the 2" outer-coupler tube onto the motor tube, and made a motor retainer from aluminum.

The last thing to do is make one more centering ring, drill a few holes, mount rail guides, and glue everything together.

 

[CarVac]

Today I drilled a vent hole in the avionics bay , made a centering ring, mounted a forged eyebolt on it, and glued it to the avionics bay just forward of the vent hole. That way, ejection charges will not affect the air that the altimeter reads pressure from. (I still have to drill holes in the airframe to vent that space, though)
I also mounted the aft Acme rail guide. I sanded the tube, making the Kevlar fuzz up, which should provide an excellent bond. I also used 60 grit sandpaper on the aluminum surface itself.
That was 3:30 pm.

Then, at 11 PM, I glued the motor mount tube/avionics bay assembly into the airframe, and laid up some rudimentary fillets between the motor mount tube and the fin roots. I'll evaluate how they turned out once they're cured, and reinforce if necessary before adding the aft centering ring.

Plans for the future: finish av-bay (find a place for a switch, and wire it up). Furnish some way to attach shock cord to nose cone. Mount forward rail guide. Drill 3 vent holes for the altimeter. Study for L2 test (or maybe I don't really need to). Ground test.

 

[accpool]

Where do you go to school? your work is always so clean and neat.

 

[CarVac]

Where do you go to school? your work is always so clean and neat.

I go to Harvey Mudd College in Claremont, CA (east of Los Angeles).



Aside from that: The news from today.

I decided that I would grab the Apogee article on fin flutter and plug the numbers from my fins into the formula. Seeing the formula, I was surprised, because I expected the mass of the fin (and thus its resonant frequency) to have some sort of influence on the results, but there's nary a sign of it. Nor is there any mention of sweep. I guess it does have to do more with the characteristics of the forcing functions, but I don't know why wing sweep isn't taken into account for that.

The result varies depending on what the effective shear modulus of my foam-cored CF fins are. I wasn't sure, but this page (https://www.performance-composites.com/carbonfibre/mechanicalproperties_2.asp) gives two results for the in-plane shear modulus: 5 GPa and 33 GPa, depending on whether the fiber is oriented 90/0 degrees or 45/45, respectively, with respect to the direction of stress.

With these two values, I get ~.35 Mach and ~.9 Mach. The former is more likely, since my fins aren't completely CF, but my fins have a very large amount of 45/45 in the layup so it might be somewhere in between.


So. 1/3 of the speed of sound. That's an awfully limiting speed for a fairly lightweight 54mm rocket. Here I'll enumerate a few options I had.

1. Limit motor size and thrust. REJECTED
2. Add mass. REJECTED because I do want to be able to fly on smaller motors on occasion, and because it's largely assembled already.
3. Add drag. Unlikely to affect top speed too much unless I mount a saucer the size of the fins on the bottom of the rocket, so it's not going to work great.
4. Reinforce fins somehow.

So, I'm going to reinforce the fins with tip-to-tip fiberglass. Straight from tip to tip. In fact, I'm going to literally run G10 from one fin tip to the next to prevent the fins from twisting. It's going to be a beam section in the shape of a V with a wooden dowel glued in between to prevent them from fluttering on their own.

Fun side benefit: more drag, so it'll stay lower on the big, long-burning motors I intend to fly this on.

 

[CarVac]

This is what I'm going to use for tip-to-tip reinforcement. Real tip-to-tip. With all of the drag that entails.

I kinda used up a 12x24" board of G10 for this...will have to buy more for the club.