Rocket Recovery Failure (see Lee's shit blow up)

[RiggerLee 61]

Armadillo has put up a little video from the last launch. It loaded really slow for me. I had to just leave it for a while and let it down load. What you're seeing is a composet of three cameras. The cool one from my perspective is the nose which is actually in side the tube. The side and GPS mast are in the way but It's still the best view of my shit shreading. I'll see if I can get better vid od it and stills or something of the seam failures along the burble fence on the ballute.

http://www.armadilloaerospace.com/n.x/Armadillo/Home/News?news_id=377

Lee

Lee
lee@velocitysportswear.com
www.velocitysportswear.com

[pchapman 279]

Fun to see some parachuting stuff beyond the usual sport skydiving.

Guess you must have a copy of Knacke in addition to the usual Poynter's sitting on the shelf. Where do you find other design guidelines? AIAA Decelerator Systems conference papers might help a bit, but really one would need contacts with others who have done such high speed work.

The ballute sounds solid, 5.5 oz Nomex, lots of Kevlar reinforcements, even if it didn't work in the end. I'm curious how the ballute is deployed, and sequenced relative to whatever the rocket is doing near the apogee.

What was being used as the descent canopy? Custom built, or an old tandem main, something like that?

[monkycndo 0]

That nose cone banging around looks like it could do damage or get tangled in ballute's lines.

50 donations so far. Give it a try.

You know you want to spank it
Jump an Infinity

[wolfriverjoe 1,523]

Quote

Armadillo has put up a little video from the last launch. It loaded really slow for me. I had to just leave it for a while and let it down load. What you're seeing is a composet of three cameras. The cool one from my perspective is the nose which is actually in side the tube. The side and GPS mast are in the way but It's still the best view of my shit shreading. I'll see if I can get better vid od it and stills or something of the seam failures along the burble fence on the ballute.

http://www.armadilloaerospace.com/n.x/Armadillo/Home/News?news_id=377

Lee

Nice. 140,000feet. Wow!

"There are NO situations which do not call for a French Maid outfit." Lucky McSwervy

"~ya don't GET old by being weak & stupid!" - Airtwardo

[DougH 270]

Is that another rocket under parachute @ 10:11?

So why a square for the recovery chute instead of a round?

"The restraining order says you're only allowed to touch me in freefall"
=P

[pchapman 279]

Quote

So why a square for the recovery chute instead of a round?

GPS & computer & servos added, so it can fly back, instead of having to chase your rocket across miles of desert...

[Nutz 0]

I'm thinking you should go to a bigger round parachute first then transition to the square after you have slowed down enough. But WTF do I know?

"Don't! Get! Eliminated!"

[jonathan.newman 1]

Could GPS and servos also steer a round? What about using a reefing line and cutter?

Have you taught the computer how to flare a square?

[Beatnik 2]

Quote

Could GPS and servos also steer a round?

Yes you can but a round will not get the glide or maneuverability that a square will. Also, you can use a smaller square parachute. Pack volume could be a variable in the design.

Teaching the computer to flare a square is something that has been taught. Look at the many JPADS' out there that use squares. I have videos of JPADS doing some pretty cool things. This is a similar setup I imagine.

[dorbie 0]

Quote

Quote

So why a square for the recovery chute instead of a round?

GPS & computer & servos added, so it can fly back, instead of having to chase your rocket across miles of desert...

Yes, but they seem to trivialize the complexity and risk of such a system. A round is far more likely to deploy successfully in an automated system. Fetching your spaceship is a lot easier than rebuilding it. I hope they learned a lesson from this. It's another poor design decision from these guys, like building a top-heavy spaceship with a narrow base that just toppled over on landing, makes me scratch my head in wonder.

[RiggerLee 61]

Starting at the begenning:

There is some infomation in Recovery Systems but this is kind of a weird thing that's out side the norm. I did find a few papers that were relevent. the best one was a report of a drogue deployment test behind a fighter from 1.6-.75 starting at 50,000 ft. It had the best real world no shit numbers I could find. I found a coulpe of designs with good dementions for the design of the ballute it self. It's a 42 deg cone or 84 deg if you prefer. at the base. It gave ratios for distance to equator. It's a simple spine curve bringing the side up from the base to vertical at the equator. The section above is an elips. It also gave locations and diamiters for the burble fence which is a torus. It gave distance below the equator for the vents to be located and an angle for the inlet. All of the designs that I could find dementions for used a snorkal like inlet. With a metal ring in the mouth and a network of lines to suport it. It looked like just the kind of thing an engineer who had never seen a parachute would design. The guy obveously wanted to put a metal scoop on there, think super charger like on Mad Max's car, and tryed to make the same thing out of fabric. It's as fucked up as it sounds. I did find pictures of ones that had actually gone into production and were used on bombs and retarded submunitions that used a strong tandom style vent. So that's how I built it. Small ones worked great. Then we built a 5 foot for the first stig rocket. We didn't have good video on board and as far as we could see from the ground it seemed stable. As it happions we had an of nominal flight. Early burn out. The original spec on the job called for a 300-350 pound load deployed at 110 mph at 8,000. One of the harder things they asked for was a low decent rate. We were looking around for larger round canopies that would meet there needs when we talked to strong. They were playing with a replacement for the T-10. Basically a 35 ft canopy with a much lower decent rate. It was about the only thing we could find off the shelf between a personel parachute and a g-12 cargo chute. And Ted was by the way the only guy we could find that seemed interested in playing ball with us. The numbers on it looked very promessing. It should have given them about a 14 ft per sec decent rate.

As it turns out by the time it flew the empty weight was up to about 550 lbs. And with an early burn out it wound up around 900 lbs on opening. And the opening happend a bit high. The housing came lose on there pin puller and the cable wiggled out under vibration. Deployment wound up being at about 200 mph at... about 16 0r 17 thousand feet MSL I don't recall the numbers. Ouch.

Tore out a section of there bulkhead that they were attached to and broke the riser assimbally at the release. We'd talked about this. That they were getting too heavy, basically that they had out grown there shit before they had even flew, and that a high speed over weight abbort was not survivable. I still felt like a tool when I watched the mushroom cloud rise into the air.

The second vehical was kind of a side project. It was one of their short fat MOD rockets. We built a bigger nilon ballute for it. Eight foot. We did a drop test with it but unfortinatly the relese wasn't quite ready. and Phil wound up basically dirrect bagging it. So the opening was almost instentanious. Didn't really get to see how the ballute really behaved. It was washing around and didn't seem fully inflated but we put that down to it just not having time. This was our first use of the Set 500. We wanted to move towards a square and we were going to land the MOD on it's side. They eventually want to fly passengers and a soft landing under a big square looks a lot more attractive for there capsule.

I'll make the story short. It was another early burn out with an early low speed deployment. Bad news was the thing was tumbling hard. The ballute deployed but the vehical was still tumbling and wound the bridle around itself. If it had been given time to unwind it might have stabalized the vehical before the main was deployed. Unfortinently after the last flight they jumped the gun and fired the release manually. There were some issues with the door that covered the canopy compartment but the biggest problem was that the canopy was pulled out and wrapped around the fuselage. The airframe had been partially torn appart when it tumbled. There was jagged alluminum every where. Several lines were cut on the airframe before it unwound durring the opening shock. Spiral in crunch. The airframe was already toast it didn't hit supper hard but they lost part of their computer which was a bummer. The INS unit was however recerted and flew again. Another big L on my head.

Stig A The one in the wright up. You bassically have the story there. Oddly this was actually the softest of our landings. No deal killer dammage. Same vehical scedguled to fly late Jan.

We were able to do a heavy drop test with the final system right before we flew. The was some flutter in the drouge. We knew we had a problem but not how bad it was going to be. We didn't have a lot of options at that point and decided to fly the system as it was. The Ballute it self has quite the gridwork of kevlar tape on it. I was actually very confedent that it would hold togather. I expected the main load to be in the skin of the ballute. I did not expect to find the burble fence to be the failure point. And I certinly did not expect to have it fail at the top edge of the fence. That was a surprise to me. Right now we're playing with vent designs to try and make them more stable. With stable inlets I think we'll have a solid system.

Back to questions. The nose cone is held on with nylon sqews. There is a heliom tank in the nose cone. It blows off the nose cone by shearing the screws. Bang. The nose cone is attached to the bridal by a Screemer. That tears in elongation under load ripping multiple rows of zigzag. The screws were cold and we blew off a little harder then expected ripping the screamer to full extension. It absorbed a lot of KE but it still rebounded a little hitting the fusolage. Not bad.

The computer has GPS and INS. It fires the nose cone at appogy.

The nose cone was attached high on the bridle below a swivel. The decision was made to go with a reletively short bridal to keep the ballute from reaching the tail fins. The nose cone was placed high on th bridal to keep it away from the fusolage. It was expected that it would hang down. As it turns out we were not makeing as much drag as we had expected with the ballute and it tended to float up. There was a lot of ossolation. The nose cone would touch the canopy and get bounced out to the side. Once it was sideways it would swing up and hit the ballute. It also stored energy pulling on the base of the ballute. We've been doing some drag test and we are seing a corilation between the flutter of the heavy bridal and a rippalling collaps mode in the canopy. Weight and natural frequency of the bridal are playing a part in the instability. We have vid of the collaps running in a circle around and around the ballute. Fasinating. Next time I think we're going with a much longer bridal. Swival down low with the nose none attached much lower. and a thin kevlar rope going up to the ballute.

The ballute dosn't have lines as such. There are a few small holes from where the nose cone touched it. The nose cone did not tear the ballute by "hitting" it. More the ballute was fluttering so fast that it wore holes in it self where the nose cone brushed against it.

We have talked about larg rounds, with reefing lines. We've even talked about useing a timed cutter to hold the slider up in case of another high speed opening. If we fix the ballute problem we should be looking at a 110 mph deployment which is well withen the canopys envolope and I hate to complicate the universe.

Tired of writeing.

Lee

Lee
lee@velocitysportswear.com
www.velocitysportswear.com

[dorbie 0]

Quote

Quote

Could GPS and servos also steer a round?

Yes you can but a round will not get the glide or maneuverability that a square will. Also, you can use a smaller square parachute. Pack volume could be a variable in the design.

Teaching the computer to flare a square is something that has been taught. Look at the many JPADS' out there that use squares. I have videos of JPADS doing some pretty cool things. This is a similar setup I imagine.

JPADS has a bit more R&D testing behind it and a more conventional operational envelope.

It would be cool to develop this as part of a civilian project, but not at the same time as you're developing a launch vehicle and relying on the reliability of one for the preservation of the other.

If they could get a license to stick a real JPADS on this thing that might be wiser. Then they're working on a drogue deployment to replace a direct bag rather than the whole enchilada.

[RiggerLee 61]

Acually the GPS recovery system is an off the shelf proven tecnology that is in use under contract by the military today. Waymor has performed flawlessly every time the system has flown in drop test for us. That has not been where our problems have layed.

The whole round square debate is ongoing. When they first asked me about this my responce was a big round. It raises questions about bulk and weight. As I said the problem started out really easy and we started out with a dyper deployed round. We talked about going to a more robust round with a cutter, a very proven tecnology, but there were other future issues that they were looking towards. Contracts for MOD style rockets that would require them to land withen certin distances on take off. Remember these guys started out focased on powered flight and powered landing. Most of the work deals with guidence and control systems. The high speed high altitude thing is a new direction for them. and what you're seeing is not really a final design but more an engion test platform. As an example one path they are looking at is decent under canopy followed by relighting and landing under power. To make that mile stone the have to be able to land withen a certin distance. They need a square for that. They can't carry enough fuel for that kind of cross wind range. That's just one example but the point is you're only seeing half the story behind the design decisions. As to a square for this project. Well it's still under debate. It's a direction that they want to go for other reasons and I'm not sure it's a bad decision for this. The thing will fall over. It will do that with a round as well. It could have a lower vertical decent rate during the flare then you could acheave with a resonable sized round and we're thinking the horizontal landing will kick the base back under neath it laying it down softer on it's side then a round dropping it on it's tail and just letting it fall over like a tree. Then again it could be a hell of a pole vault. We will see, hopefully in about a month.

Now I really am done typing.

Lee

Lee
lee@velocitysportswear.com
www.velocitysportswear.com

[pchapman 279]

Plenty of companies have developed GPS guidance systems for parachutes. But finding something off the shelf for rocket recovery might be an issue. I agree, dorbie, that it sucks to develop the recovery system at the same time as the rocket. It's like an untested engine on an untested airplane. But sometimes, there isn't much in the way of good options.

Thanks for the writeup, Lee.

[dorbie 0]

@pchapman, On the issue of off the shelf systems, you could drogue it down to whatever envelope its designed to operate in, so the drogue is where any novel work would be (perhaps oversimplifying things but there you go).

@RiggerLee

thanks for the awesome detailed responses, very generous of you to share this information.

I wasn't trying to be overly critical of the amazing work on a tight budget, just the complexity added by the fundamental question of round vs. square and you've made it clear what the initial recommendation was and hinted at the hidden design decisions that led you to this solution. Thanks again.

[rrmtopo 1]

Very cool video Lee. Have you looked at what was done with the X38 (CRV)? Obviously they started with the lifting body vehicle, but a PC to round to square deployment sequence seemed to work when performing test drops.

http://www.youtube.com/watch?v=0OkiRt5Etbw&feature=related (skip to 2:00)

[RiggerLee 61]

We have in fact talked about another intermedeat stage but no one wants to add that kind of complexity. I was board and started playing around with reentery models. To be honest I'm rather sceptical about it's acceracy. Let me refrase that. I know it's wrong. I'm just not sure by how much. It's a diffrent world at those renalds numbers. I use that term losely. It starts out in what for all intents and purpouses is vacume. It falls building up speed and then whacks into the atmosphear like a wall. The density increases faster then it can slow down and it sees ablout 5 G's. There's another spike at the trans sonic. The odd thing that I see in my model, keep in mind that this is in the area where I'm pretty sure it's breaking down, is that there seems to be two humps in that first big load. It's like a slap when it first hits the outer atmosphear and then another as the density rises faster then it decelerates. Keep in mind that this spreadsheet came straight out of my ass. But there seems to be a sweet spot in the diamiter of the drouge where this kind of evens out and the load is distributed over a longer time. Too big and the inital spike becomes domminant. To small and you don't really slow down early enough. If we went to a little smaller drouge we could cut down the peak G loading by about 27%. This is again based on the "shaky" end of my model. It says we're a little bigger the optimal. But we would like to keep it simple for now. We're trying had to go straight from drouge to direct bagged main. The curent 8 ft diamiter should give us about a 105 mph deployment at 15,000 msl. The snatch force is rather high but we're working that problem. It seems to be in hand.

Should be more drag test tomarrow then I want to mod one of our old drouges and see if we can do a heavy torture test to try to simulate full 5 g loading. It wont really be the same, no good way to reproduice the same mach numbers or true air speeds. But seeing those seams hold togather under load would give me a warm and fuzzy.

Lee

Lee
lee@velocitysportswear.com
www.velocitysportswear.com

[rrmtopo 1]

Sounds like a really interesting project Lee, after looking at the X38 video again I was not sure if their deployment sequence was optimized just for the drop test or an operational deployment. Your regime is obviously a lot more complex than what was experienced there. Good luck and look forward to hearing how further tests go.

[RiggerLee 61]

Just an up date. Flew the same airframe last weekend. There were a few modifacations. The nosal was extended to 10 inches in diamiter. and the section with the roll vaines was turned 90 deg. The engion burned to compleation. I think it was a 51 sec burn. The nosal showed signs of being slightly over expanded at sea level. The heat band was about a inch short of the edge after the burn test at Caddo. Ground elevation at space port america in NM is about 4500 ft. So I doubt there was much loss. Hoever at higher altitudes it really shined. Much higher speeds. They melted all of their stickers off. The paint blistered off the nose cone. The very tip of the cone looks poorus like the end of a bone. Looks to me like they vaporized a little of there ressen at the very tip. might have to go some kind of higher temp resin? They lost telimitry part way up. They have a pretty good drag model that fit the gps data from the last flight very well. Putting the telimitry from sat into it, I think they lost it some time after burn out around 100,000 ft, it looks like they topped out at around 311,000 ft. So I don't think they can quite call it a space craft yet but they seem to be getting close. This shot wasn't expected to excead 250 so they are very pleased with the progress on the engion.

Now the bad news. I made another crater. We did recover one camera. It looks like that's the only chip that survived. It was an upwards faceing video mounted in the side. It's field of veiw went from almost vertical down to the horison. You can see the nose deploy. The ballute looks very good. You can see it in several places in the video. No sign of flutter. I think that problem is solved. but there was an uncomanded release of the ballute and early deployment of the main. The relese for the ballute broke 92 sec after the nose was deployed. That's at about 180,000 ft if you beleave my reentry model. That's also right about when the force is supposed to be peaking, best guess 1900 lbs. In the video it apears that the kevlar rope between the ballute and the swival is not fully extended. It was rubberbanded togather in three large stows. It was a 100 ft 12,000 lb kevlar line. Before you even start they used a 100 ft line in there testing and insisted on useing the same length on the flight. It looks like the line became intangled. YOu can see it in a bundle at the base of the ballute. Right before it failed the ballute seemed to extend further out like that friction knot let go and suddently you had a big snatch when it came taunt. Main tryed to deploy at about mach 2.5. snapped three PD reserve slinks and broke all the lines off the fourth riser. Rocket hit nose down like a lawn dart about 1000 ft from the trailor. compressed 28 feet worth of rocket into about seven feet. They haven't released and video or stills. I'll try to a least get the picture of the fuel tank. It is so cool. Looks like an acordion. You got to see it.

In any case we are now looking for release designs for the drogue. I'm back to looking at cutters. Waiting on some info from airborn sys. If any of you have heavy drop background I'd love to hear any thoughts about macanical or soft releases that we could actovate electronicaly or with a pin puller. We're looking at about 5 g's, Next airframe around 800 lbs. So we're predicting a load of about 4,000 lb. That's assuming all goes well. Note that we had a big spike on this last flight thanks to that line. I'll try to get back to you with more detailes.

Lee

Lee
lee@velocitysportswear.com
www.velocitysportswear.com

[pchapman 279]

Quote

Main tryed to deploy at about mach 2.5. snapped three PD reserve slinks

Conclusion:
PD Reserve Slinks are dangerous!!

.... at least at Mach 2.5 (& unknown dynamic pressure)


Thanks once again for the interesting report, even if it is all getting a little expensive to furnish us with these stories.

[popsjumper 2]

Quote

Quote

Main tryed to deploy at about mach 2.5. snapped three PD reserve slinks

Conclusion:
PD Reserve Slinks are dangerous!!

.... at least at Mach 2.5 (& unknown dynamic pressure)

Talk about a hard opening!

My reality and yours are quite different.
I think we're all Bozos on this bus.
Falcon5232, SCS8170, SCSA353, POPS9398, DS239