Black Magic-Another “nominal” minimum diameter N5800 flight

[xenon]

Team Last Minute Rocketry travelled to Tripoli Las Cruces the weekend of December 29th for a final attempt at CTI’s N5800 Challenge. As our group name implies, we push things to the last minute, and this project was no exception. Construction started December 21 and was mostly finished by the 24th. A last minute 15 hour session on December 28-29 finished up the rocket in time for a 10:15 am launch at FLARE’s West Mesa, Las Cruces launch site.

I’m going to briefly cover the construction and flight in the same thread, with a few pictures.

Details on the rocket are as follows:
Pad weight: 43.2 lb
Dry weight: 10.5 lb
Max external diameter: 4.055”
Length: 84.5”
Expected Altitude: 51,700’
Expected Max Velocity: 3,795 ft/s (Mach 3.55)
Burnout Altitude: 7,800’
Launch system: 2x 1010 buttons, 10’ rail
Subsonic (launch) stability margin: 1.91 calibers
Burnout stability margin: 2.43 calibers

The rocket was basically all composite and mostly carbon. From the top down:
The nose cone was a Performance Rocketry 4” FWFG nose cone with red anodized aluminum tip. The bulkplate holding the #10 screw that holds the tip in was glued in place with 5 minute epoxy. The area above this bulkhead was filled with ProLine 4500, and the tip was screwed on. Excess resin that squeezed out was cleaned up with a rag before it cured. The provided coupler tube/shoulder was glassed internally with 2” wide strips of 10oz glass. The coupler was then glued into the nosecone with 5 minute epoxy then the coupler-nose joint was glassed in a similar manner. The forward part of the nose was filled with Evercoat 2 part foam. A bulkhead with a piece of ⅜” threaded rod bolted to it and a large number of holes to allow foam flow was glued into the nosecone with 5 minute epoxy about 10” forward of the base of the nosecone-coupler joint. More foam was poured on top of this until the fiberglass from the nosecone-coupler joint was reached.

The electronics were mounted on a pair of ¼” plywood boards that were attached to a ~3.8” bulkhead. These would slide onto the threaded rod in the nose cone. The goal was to keep all of the electronics forward of the nose-tube joint; however, they ended up extending about 1” below that. Electronics consisted of the following:
-Beeline 70cm High Power GPS with Ublox 6 chip configured to transmit every 5 seconds on the 2 second time slot and store every second. A featherweight magnetic switch was used to power this.
-Beeline 70cm Normal GPS with Ublox 6 chip configured to transmit every 5 seconds on the 0 second time slot on the same frequency as the above one, and store every second. A featherweight magnetic switch was used to power this as well.
-Raven 3 configured to deploy at normal barometric apogee (C02), baro apogee + 3 (BP charge), 15k AGL (BP in drogue bay) and 2,000’ (main).
-Stratologger configured for baro apogee and 700’ main.
The avionics bay was vented aft into the main parachute bay to avoid drilling holes into the nose or near the nose-tube transition (more for structural than venting purposes). All of the avionics were powered by magnetic switches, also to avoid holes.

The upper tube consisted of a 30” section of 4” LOC tubing stripped down to a minimal amount of layers. Two layers of Soller Composites sleeve using Aeropoxy was applied to the tube. A spare motor case was used as a mandrel to keep the cardboard from deforming during layup. A 3oz layer of fiberglass was also applied as a veil layer. A bulkhead was glued in such that it rested on top of the eye bolt in the top of the motor. A Rouse-Tech CD3 system was installed in this bulkhead with a 16 gram cartridge.

The lower tube was constructed similarly, except with a single layer of Soller sleeve and a single layer of 14oz or so carbon wrapped around it with no veil layer. It was also 30” long.

Fins were ⅛” G10 with a 12” root, 4.75” span, 8” sweep and 3” tip cord. They were laminated on each side with 3 layers of 14oz carbon using Aeropoxy, bringing their total thickness to approximately .21”. They were tacked to the tube with 5 minute epoxy, filleted with ProLine 4500 and had 4 layers of tip to tip 14oz carbon of increasing length using West 206. Final measurements showed a 1.25” L.E. airfoil length, 1.0” on the T.E, a L.E radius of 0.035 and an average thickness of 0.30”. The leading edges back to behind the outermost tip to tip ply were painted with Quick Weld (fast JB weld).

G10 Prepped

Wax paper on the carbon for tracing the profiles

Vacuum Bagged

Fresh out of the bag

Cleaned up and tapered

Fillets

Tip to Tip

Recovery was drogueless at apogee with a 58” bright orange topflight parachute at 2,000’. Lower tube rested (tightly) on the motor case, and the upper tube used the top 20” or so of the motor case as a coupler. This joint was to be separated at apogee, with the nose (using 2x 2-56 shear pins) coming off at 2,000’. A 1010 rail button was installed in the upper tube, along with in a waterjetted and sanded aluminum ring that was slid on the aft end of the case. The tailcone that we initially intended to fly was not used due to manufacturing issues (ending with the TSA taking our foam and glue, but not beginning there), and the fact that we were satisfied with the simulated altitude without it. Finally, a 70cm GPS beacon was taped and wired to the forward closure of the motor.

Final on the pad weight was 42.3lb with a stability margin of 1.91 (based off fully loaded CG measurements).

Launch occurred during continuity checks at 10:14:38 Mountain Time, so unfortunately we don’t have any pictures or video to show. Data shows a reasonably nominal flight until 2.43 seconds in. An increasing lateral acceleration that peaked at about 1.5g’s (23” forward of the CG) began to occur. At 2.67 seconds in going M3.15 (3,400fps) and 4,400’ or so (based off simulated data), the nosecone liberated itself from the rest of the rocket by some means. The nosecone decelerated rapidly, pegging both the axial and lateral accelerometers on the Raven simultaneously. The baro sensor reports that it reached an apogee of about 5,400’. Total flight time on the nosecone was 62 seconds.

Other notes:
-The fins came off too. We didn’t find them or get any pictures of them in flight
-The power wire on the HP GPS was cut on takeoff due to one of the boards shifting
-The GPS receiver antenna on the LP GPS was damaged on takeoff as well
-The transmit antenna on the LP GPS was damaged and stopped transmitting once it hit the ground
-The avionics bulkhead was cracked in half, and somehow managed to take the stratologger with it
-The only pieces we found were as follows:
Lower tube, sans fins, 1750’, 323 degrees from the launch site
Lower 20” of upper tube
Nose cone, 1350’ due north of the launch site
Nose cone coupler, 2” further out of the nosecone than it had been previously
Motor case
Part of the electrical tape that was holding the Beeline beacon to the motor.
The main parachute was seen drifting off

Front of the motor, missing the eye bolt

Other end of the motor

Used to be a fin there

Front again

Nose as found

Internals, post flight


So what happened?

It broke.

A couple theories:
-Stratologger blew the drogue charge on the way up (see baro trace on Raven). This is unlikely due to the fact that the upper tube was still on the motor case when we got it back. The small dip in pressure just before breakup occurs while the slope of the baro pressure is approximately 2,300ft/sec. This implies that even if the stratologger had a clue what was going on and detected launch (who knows if it did or not), that it would have locked out due to the velocity. Also, the CO2 charge isn’t much of a match against the 300lb of drag on the vehicle at that point
-A fin came off. Sure. There’s not much evidence for or against this other than it did happen at some point.
-The upper tube broke. See previous bullet point.
-Wind shear. Data from El Paso that morning supports a 10 knot shift in speed and 150 degree direction shift at the altitude where the breakup occurred. Observations at the launch site support stronger winds by the time launch occured. Rocksim simulations using 0 mph ground winds and similar speed (12mph) upper winds could result in oscillations that would create a lateral acceleration of similar size to that seen in the accelerometer data. This oscillation would nominally damp itself out, but for whatever reason (structural failure?) it didn’t.
-Aliens. Wouldn’t surprise me. We were in New Mexico after all.


Thanks to the New Mexico and Texas folks for helping with a place to finish up constrep (new term for construction/prep) and a great launch site and lunch.

View attachment Black Magic.FIPa

 

[JRL303]

Great try guys! Loved the write up and pictures. You've got moxy for sure!
Jim

 

[bobkrech]

No shear pins?

Bob

 

[xenon]

No shear pins?

Bob

On the nose, yes. On the drogue, no, but that joint never separated.

 

[stealth6]

Just two 2-56s apparantly.

hmmmmmmm.

s6

 

[CarVac]

I'm inclined to think that stress concentration could have done it: the forward edge of the motor case ends abruptly and bam there's a lot of load there.

Also: railbutton wave drag could have caused the rocket to fly at a not-insignificant angle of attack, thus snapping the nose off.
Did you move the rocket between the first two aftermath photos? The first makes me think that the tube peeled in the upward (in that photo) direction, while the railbutton was shown as being on the same side in the second aftermath photo.
Assuming none of the torn tube at the forward end of the motor got disturbed after the initial failure (a very poor assumption, I know) that leads me to think that the railbuttons might have at least contributed to the failure.

 

[Aksrockets]

I have a question
Sorry if this comes off rude at all. It's hard to convey a tone of voice on a forum.

Why did you think this would turn out different then any other N5800 rocket out there? There were some entries made of much stronger materials (no cardboard or fiberglass) and they still came apart. The only one that survived was one with a super tough aluminum fin can.

Alex

 

[CarVac]

(no cardboard or fiberglass)

I'm pretty sure every attempt had fiberglass. You can't get RF transparency, and thus GPS lock and telemetry, without it (or basically something other than carbon or metal). No GPS lock, no Tripoli record.

That said, it could be that they flew it for the same reason people play the lottery, except the odds are somewhat better and the payout much lower. AKA, for fun.

 

[stealth6]

Launch occurred during continuity checks at .......

This little (very casually mentioned) bit is what jumped out at me. Yikes!

s6

 

[Aksrockets]

That said, it could be that they flew it for the same reason people play the lottery, except the odds are somewhat better and the payout much lower. AKA, for fun.

Those are some expensive lottery tickets.

What I meant with the was most teams only used fiberglass where they needed it. Other then that, the strongest materials avalible were used.

I'm impressed it was built in under a week. I usually can't build an altimeter bay in that time.

Alex

 

[BLKKROW]

I have a question
Sorry if this comes off rude at all. It's hard to convey a tone of voice on a forum.

Why did you think this would turn out different then any other N5800 rocket out there? There were some entries made of much stronger materials (no cardboard or fiberglass) and they still came apart. The only one that survived was one with a super tough aluminum fin can.

Alex

I am one of the weird ones that believe a composite rocket can handle the N5800. But that is not what this thread is about.

Xenon thanks for the thread, and I learned a lot and love reading about these attempts. I hope you learned something from the launch that is the best thing you can get from a failed launch besides a good story

 

[edwinshap1]

I'm pretty sure every attempt had fiberglass. You can't get RF transparency, and thus GPS lock and telemetry, without it (or basically something other than carbon or metal). No GPS lock, no Tripoli record.

That said, it could be that they flew it for the same reason people play the lottery, except the odds are somewhat better and the payout much lower. AKA, for fun.

Yup, FG nosecone on most 5800 rockets I've seen.

 

[cjl]

I am one of the weird ones that believe a composite rocket can handle the N5800. But that is not what this thread is about.

I still believe that as well, even though my first composite attempt at the 5800 failed. Hopefully, I can give it another try next year - I have some ideas that may improve my chances in the future, but that's a subject for another thread.

As for the subject of this thread, that sounds (and looks) very much like the way mine failed, and I'm sorry to hear that it didn't work out. I know with reasonable certainty that mine failed due to upper tube failure, possibly exacerbated by a stress concentration right at the top of the motor case, and that would be my best guess for yours as well (though fin failure would also be a strong possibility). Do you have any plans to try again at some point?

 

[CarVac]

I happen to also think that a composite rocket can survive, but an aluminum fincan can be made to be more aerodynamic.

For the redesign of Bare Necessities, we are (I just pointed it out yesterday to CCotner) taking measures to minimize stress concentration, as well as making our tube super beefy, because those seem to cause 90% of failures.

 

[EMiR]

I commend you on your attempt, and love the balls displayed in giving it a shot with such a quick build.

Interesting results in your pics, I'm surprised that the single tip to tip layer stayed connected on the body tube part and tore clean at the fillet edges.

easy pick, would be fins came off, you got a change in angle of attack, leverage would have found the weakest part, and pressto.

Yer an Ali fin can and body tubes will do the job, by that's just the easy way out. (cheating really)

I think a real 5800 flight will be have to be done by a composite built rocket, And I too, think it's possible. Just have to get a little creative with the load distribution and design

 

[cjl]

I happen to also think that a composite rocket can survive, but an aluminum fincan can be made to be more aerodynamic.

I'm not convinced that's true, given proper composite construction. I'm not sure you could ever achieve the required (aerospace-grade) construction techniques to do it, but at least in theory, the substantially higher modulus and ultimate strength of CFRP (as compared to high-grade aluminum alloy) should allow for thinner, more aerodynamic CF fins to survive than would be possible in aluminum. In practice, the aluminum is a bit easier to work with however, so you may be right at least as far as achievable construction techniques are concerned.

 

[CarVac]

I'm not convinced that's true, given proper composite construction. I'm not sure you could ever achieve the required (aerospace-grade) construction techniques to do it, but at least in theory, the substantially higher modulus and ultimate strength of CFRP (as compared to high-grade aluminum alloy) should allow for thinner, more aerodynamic CF fins to survive than would be possible in aluminum. In practice, the aluminum is a bit easier to work with however, so you may be right at least as far as achievable construction techniques are concerned.

Yes, CFRP has very high properties, but it can only have said high properties in two directions. It would require ridiculous 3d weaving to get the required strength for a fincan, which is to say you have to keep it from delaminating.

I'm actually not worried about the fin strength itself, but the fincan, which has a more complex stress state. Perhaps a metal fincan and tightly fitting carbon fins could be optimal...

 

[xenon]

I'm inclined to think that stress concentration could have done it: the forward edge of the motor case ends abruptly and bam there's a lot of load there.

Also: railbutton wave drag could have caused the rocket to fly at a not-insignificant angle of attack, thus snapping the nose off.
Did you move the rocket between the first two aftermath photos? The first makes me think that the tube peeled in the upward (in that photo) direction, while the railbutton was shown as being on the same side in the second aftermath photo.
Assuming none of the torn tube at the forward end of the motor got disturbed after the initial failure (a very poor assumption, I know) that leads me to think that the railbuttons might have at least contributed to the failure.

I'm interested in investigating the rail button thing a bit more. As far as I know, we're the first ones to use them on something like this. Also agreed on the stress concentration on the forward edge of the motor. Post flight discussions at the field point to the Soller sleeves having some drawbacks. Maybe making the upper tube out of normal weave cloth would've been a better idea.

I have a question
Sorry if this comes off rude at all. It's hard to convey a tone of voice on a forum.

Why did you think this would turn out different then any other N5800 rocket out there? There were some entries made of much stronger materials (no cardboard or fiberglass) and they still came apart. The only one that survived was one with a super tough aluminum fin can.

Alex

I don't recall saying I thought that

By the time Balls rolled around, there were something like a half dozen build/destruction analysis threads. We took a bunch of different things from them, this list not being anywhere near inclusive:
-PR FWFG noses probably need to be reinforced
-FWFG upper tubes are going to snap
-Keep your stability margins non-zero during the entire burn
-Building as short as possible is just going to be a pain when it comes to stuffing recovery in
-Going for 100k isn't necessarily a bad thing, but if you only need 50k for the record, it'll make the flight easier in a lot of aspects
-etc

Yes, you could consider it an expensive lottery ticket. Or it could be considered a good excuse to go hang out in the desert for a few days with some good friends and launch a rocket as well.

As for the subject of this thread, that sounds (and looks) very much like the way mine failed, and I'm sorry to hear that it didn't work out. I know with reasonable certainty that mine failed due to upper tube failure, possibly exacerbated by a stress concentration right at the top of the motor case, and that would be my best guess for yours as well (though fin failure would also be a strong possibility). Do you have any plans to try again at some point?

Agreed on the failure mode. As for trying again, I don't really see a reason other than the CTI challenge to fly an N5800 like this. I would like to fly a min-diameter N5800 again, but not with the (now expired) CTI challenge goals in mind. Their new 03400 also has a burn profile that seems much more suited for a high altitude flight than the N5800.

easy pick, would be fins came off, you got a change in angle of attack, leverage would have found the weakest part, and pressto.

This statement is preceded by an "if" for a reason, but if Rocksim is to be believed, the change in AOA by the wind shear was ~0.5 degree, which, if my analysis was correct, would result in less than 50lb of lift on a single fin. This was another one of those fin cans I wouldn't have a problem standing on, so there were probably other factors involved.

 

[patelldp]

This post is not meant to be snarky, but more to focus on what your mindset was going into this attempt. I will accomplish this with a series of questions you are free to answer or ignore.

1. Did you happen across a cheap/free N5800 reload and case at the last minute, leading to a hurried build?
2. Did you fully consider other projects and their modes of failure when deciding on a tube made out of only 2 layers of CF sleeve?
3. Was there research done on Proline 4500 epoxy to see if it would be a sufficient epoxy for this? Did you learn anything that may be of use to others surrounding this epoxy with "mysterious properties"? My guess is no, as it appears that the epoxy itself was liberated almost entirely from the surface of your tube.
4. Was there precedent surrounding your altimeter choices? This seems like a hostile environment for a barometric-only altimeter. The extreme flight profile that this rocket experienced is right up the alley of what Jim Amos is looking to learn about to make his RRC3 altimeter suitable for such flights. I doubt this was done on the SL100.

So, let this be another warning to those trying to utilize hobby techniques on an amateur rocket. Most people (have to leave that uncertainty in there) do not have the means to make an optimized tube in their basement/garage that can withstand a MD N5800 attempt. Heck, even those with filament winders have proven that they aren't likely to be able to do it. This is not likely to be done with standard practice. We should all look at Mike Passaretti's attempt as a baseline and look for ways to improve on his techniques/designs.

 

[xenon]

1. Did you happen across a cheap/free N5800 reload and case at the last minute, leading to a hurried build?

Cheap, yes, free, no. We already had a case as well. There seems to be a lot of focus on the last minute aspect of this. I do not think that that was a contributing factor to the failure, and it's also the way we do things. Flights to El Paso were booked by mid-November and we did a test flight with a very similar (stand in fins and lower ply tubes) rocket that was built the night before MDRA's December launch, so this by no means wasn't planned. We knew it would be a last minute build, as usual. This build time is pretty normal for us- I built and flew my L3 in 6 days, 3 of which were spent painting it. We've done over a half dozen various USLI rockets, none of which were started more than a week out, some were started less than 24 hours out, all with layups.

2. Did you fully consider other projects and their modes of failure when deciding on a tube made out of only 2 layers of CF sleeve?

2 layers of sleeve was all we could fit between the OD of the peeled down LOC tube (ie, peeled down such that it was nearly falling apart) and the OD of the nose as the sleeve is pretty thick. Vacuum bagging sleeves has proven difficult for us in the past, so we didn't try it. I've seen too many snapped pieces of PR FW carbon and glass to try to use that for something like this. We did not get a chance to do FEA analysis of the tube.

3. Was there research done on Proline 4500 epoxy to see if it would be a sufficient epoxy for this? Did you learn anything that may be of use to others surrounding this epoxy with "mysterious properties"? My guess is no, as it appears that the epoxy itself was liberated almost entirely from the surface of your tube.

ProLine 4500 is still a mystery. The only thing we learned was that the Boston TSA folks don't think it's flammable and the LAX ones think it's flammable. The lack of epoxy remaining on the tube isn't really evidence of whether or not there was research done.

4. Was there precedent surrounding your altimeter choices? This seems like a hostile environment for a barometric-only altimeter. The extreme flight profile that this rocket experienced is right up the alley of what Jim Amos is looking to learn about to make his RRC3 altimeter suitable for such flights. I doubt this was done on the SL100.

We had the altimeters, they fit and were advertised to work at the altitudes expected. I see no reason that a baro-only altimeter wouldn't work with this flight profile. If there is such evidence that baro only altimeters won't work in this application, I'd like to hear more.

 

[patelldp]

We had the altimeters, they fit and were advertised to work at the altitudes expected. I see no reason that a baro-only altimeter wouldn't work with this flight profile. If there is such evidence that baro only altimeters won't work in this application, I'd like to hear more.

https://www.rocketryforum.com/showt...ks-product-in-the-works-!&p=455173#post455173

Reach out to Jim Amos.

 

[patelldp]

This little (very casually mentioned) bit is what jumped out at me. Yikes!

s6

Agreed. That's the biggest failure of the project. That's scary, in my opinion.

 

[bobkrech]

Yes, CFRP has very high properties, but it can only have said high properties in two directions. It would require ridiculous 3d weaving to get the required strength for a fincan, which is to say you have to keep it from delaminating.

I'm actually not worried about the fin strength itself, but the fincan, which has a more complex stress state. Perhaps a metal fincan and tightly fitting carbon fins could be optimal...

Nonsense.

There are 2 simple mechanical solution that allows you to obtain any ratio of longitudinal to hoop strength in a composite tube without difficulty or cost.

1. Filament winding. A filament winding machine can be programmed to make a tube with a wide ratio of known longitudinal to hoop strengths simply by changing the angle of the windings. This is how the pros do it for composite pipeline pipes, and motor casings.

2. Uni-cloth hand layups. Where a fixed longitudinal to hoop strength ratio is desired, you simply make a series of uni-wraps in the following sequence:

1 layer wrap at some angle, +a, a second layer wrap at an angle -a, and the 2 layer wraps at 90 degrees. The angle is adjusted to make the tube of equal strength in both longitudinal and hoop. If you want to make the tube much stiffer to prevent bending and column bucking failure, you can put one layer wrap at 0 degrees to inhibit column buckling. You can repeat this 4 or 5 wrap process as many times as required for the desire strength. I have built 8 layer CF motor tube using this techniques.

The pros use uni for obtaining the required directional strength with the minimum weight.

Bob

 

[CarVac]

I did not mention hoop strength but rather peel strength, which is a property fixed once you choose a resin-fiber combination.

How does using uni prevent delamination of the outer layer as in the Project60k failure? It is not a bulk material strength issue; wouldn't adding more uni mean that the tube is merely thicker once the fin peels off?

So far as I know, composites are ultimately limited in peel strength and that is what is tested in the can part of fincans.

 

[MasonH]

This was a very good attempt! I enjoyed reading it, and I do not say that often. Building something this complex in a week takes dedication!!

 

[CCotner]

I think CarVac was talking about fincans, not tubes. It is possible to achieve a sufficiently high-strength (longitudinal and hoop) with conventional CFRP techniques (not even filament winding), as you described. A fincan has significantly more complex loading and I think CarVac is correct that without 3-D fiber weaving and subsequent autoclaving, it's probably not possible to outdo aluminum, and aluminum is easy if you have shop access. This is why we chose aluminum.

CarVac, the idea of aluminum with thin CF fins is...interesting. I'll investigate how much weight and aerodynamic savings we might get on the way home today; I'm thinking a 1/16" G10 core laminated in CF out to .1875" thickness, widening to .25 at the base (so it fits in our existing aluminum fin can) would be plenty strong enough. The challenge would be preventing de-lamination; maybe Mudd II style aluminum leading edge inserts? We should talk about it more.

Also, J-B Kwik, as this group discovered, is NOT an acceptable substitute for J-B Weld. The instructions plainly state that it has a significantly lower strength than the classic formula, and more importantly, 300F instead of 550F temperature resistance.

 

[xenon]

Also, J-B Kwik, as this group discovered, is NOT an acceptable substitute for J-B Weld. The instructions plainly state that it has a significantly lower strength than the classic formula, and more importantly, 300F instead of 550F temperature resistance.

Very true. It was used as an ablative layer on the fin leading edges, so those differences were unlikely to matter too much. If we had a fin back, I'd report back on how it did.

 

[bandman444]

I won't add very much to this discussion but have to paraphrase the words of Ken Biba at XPRS. "See the demographic of the ones attempting The Challenge? nuff said."

As being one of those young and dumb teams that tried it I can say without a doubt that we learned something, learning something that could only be had in the real world.

If a N5800 could rip the layers of filament wound carbon from the tube, and 8 layers of tip-to-tip from a mongoose, than 2 wraps of carbon on a cardboard tube and a couple layers of tip-to-tip will only last less than a second.

Of course I am most upset that my team had a small error in the fabrication (completely my fault) because I strongly believe that we could have made it through the burn. And that is what is pushing me to investigate further. I would love to fly an N5800 again, but not before flying it on other 6XL-like motors to prove itself.

Great write-up.

Could you elaborate on the premature launch portion of your story?

 

[bobkrech]

I did not mention hoop strength but rather peel strength, which is a property fixed once you choose a resin-fiber combination.

How does using uni prevent delamination of the outer layer as in the Project60k failure? It is not a bulk material strength issue; wouldn't adding more uni mean that the tube is merely thicker once the fin peels off?

So far as I know, composites are ultimately limited in peel strength and that is what is tested in the can part of fincans.

Most amateurs do not make composites anywhere near the strength of aerospace composites because they are nort using professional resins which require heat assisted curing to develop maximum tensile strength at high temperature. I get the opportunity to do a lot of failure analysis in my professional life so I'll try to explain what I think happened but since there are no videos of this flight, I can only go by what was stated by the principals. If I understand the sequence correctly, the airframe first separated at high mach number, and then the aft airframe went sideway and the fins were striped off. Assuming I got the sequence corrrect....

(1.) If the airframe was vented and pinned at the nose cone and coupler(s), then the initial airframe failure had to be due to column buckling of the airframe either at the forward end of the motor casing or at a coupler. Either is a result of the airframe or couple lacking sufficiet stiffness to prevent buckling. Buckling is a real possibility if the airframe was constructed from 1 or 2 layers of coaxial braid which is weak in both hoop and longitudinal directions. If the airframe was not vented and/or pinned, then any section not pinned most likely drag separated at the reduction of thrust near burnout. Regardless, at high mach number, once the airframe separates at high mach mumber, the rocket will be destroyed in a few tenths of a second by aerodynamic forces.

(2.) The motor casing prevents the aft airframe from bending, and the fins appear to be stiff, and the photographic evidence shows the fins separated due to a shear failure of the bond between the fins to the airframe. It appears that much faith was put into tip to tip laminating for attachment strength which is the probable reason why the fins came off, aided by a poor epoxy bond due to insufficient surface preparation on the airframe and fin root to enhance direct epoxy bonding.

Many sounding rockets have detachable fins that are bolted to L brackets which are bolted to the airframe. This method of attachment is stiff enough and strong enough for high mach flight. The composite analog is multiple L-shaped composite fillet layes attached to the airframe and fins. If the fillet joint is as stiff and strong as the fins and the airframe it will not fail.

This is my interpretation on what happened and why it happened. Your milage may vary.

Bob

 

[CarVac]

Many sounding rockets have detachable fins that are bolted to L brackets which are bolted to the airframe. This method of attachment is stiff enough and strong enough for high mach flight. The composite analog is multiple L-shaped composite fillet layes attached to the airframe and fins. If the fillet joint is as stiff and strong as the fins and the airframe it will not fail.

I do not dispute your failure analysis of this flight in particular, but your assertion (perhaps exaggerated by my own ego?) that under all situations, composites can be made so that they exceed the useful strength of a metal.

In your example, you give an "L-bracket" "bolted" to the "airframe", and "L-shaped composite fillet layers" "attached" to the "airframe".

An L bracket bolted through an airframe puts compression and shear loads on the airframe, which metal, or probably composite tubes are capable of surviving. However, what is your proposed "attachment" of the fillet to the fins and the airframe? I don't think surface mounting even with proper surface preparation is enough. My basis for this assertion is that there is precedent for the surface bond strength to exceed the intrinsic peel strength of a filament-wound composite tube.

Do you think that it would be better done with fasteners like the L-bracket fins on sounding rockets? That immediately steps out of the realm of all-composite fin attachment mechanisms, in my book.

On the other hand, you assert that there probably exists a professional aerospace resin that is significantly enough better than what is used in Performance Rocketry filament-wound carbon fiber tubes to increase the peel strength by a significant margin. I'd ask for 2x as strong, or at least 1.5x if I'm being very kind, based on what I saw of the Project60k remnants.

Perhaps I should test peel strength of convolute-wound CF tubes made with PTM&W 5712 high-temp resin post-cured according to manufacturer instructions against FWCF.

If anyone has multiple short samples of ~4" Performance Rocketry FWCF they'd donate for an experiment, I'd gladly make an extra tube or two when practicing layups for Bare Necessities to help clear up the debate.

If our tubes hold out better, then it's quite possibly the resin in PR tubes that limits their capabilities as substrates for fin mounting, or it's the fact that they're filament wound diagonally versus convolute. If PR tubes hold out better, then our techniques are probably insufficient. If they're about the same, then I suspect that further improvement in delamination resistance would require 3d-woven socks of some sort with fibers running back and forth between the surface and the core, of a significant thickness, with the outer surface having a peel ply finish before bonding.

Anyone up for this? I want to clear this issue up with some real experimentation, but I'm not committed enough to purchase a 4" PR carbon fiber tube to chop up.