Failure Analysis of Hyperion

Interesting. I was using the 1S based off of testing I've done in the past, using a 1S.

Flyfalcons said:

Interesting, as I'm planning on running the same combo as you and Matt. Perhaps it would be a good idea to run the RRC2+ as primary for the drogue, and program the -3 for a longer delay after apogee than the 2+'s one-second option.

Click to expand...

Possible, I guess. From now on, whenever I'm going to go with a dual redundant dual deploy system, I'm using 2 of the same computer regardless of brand. I think that's the only way to be sure that you're getting the same computation.

ECayemberg said:

Matt,

Nice to meet you and cool boost, man! Sorry the big bird is missing; I'm confident it will turn up over time. I'll begin wading through photos today to see if I got a shot of the liftoff. If it's the flight I'm thinking it was, I remember watching for an apogee event, seeing a puff of powder smoke, noting it was apart, then the main blossoming a few seconds later....it appeared that the main "shook loose" following the apogee event. IE: the main charge did not deploy the main at apogee.

Based upon weather conditions, etc., I'd bet a nickel that the three #4-40 let go at apogee. Most rockets were headed upwind at a pretty good clip due to the winds...I too suspect that upper winds were 50+mph. For reference, my 7.5" Patriot that flew earlier in the day (calmer) with a very reefed drogue drifted 1.13 miles downwind after arcing heavily upwind and deploying as intended from 8175' up. It was definitely notably windy up there! Forces involved with a speedy horizontal deployment are significantly greater than those of a calm day where the bird really slows down before arcing over.

Note, I have #6-32 and even some mega #8-32 pins for certain big honkin' cones. (But I did use *only* four #4-40 nylon screws on the 7.5" Pat on Saturday and it worked well).

-Eric-

Click to expand...

This was my instinct as well (that horizontal velocity caused a stressful drogue deployment), but see my post earlier - his nose cone was only 4lbs, so assuming the shear pins weren't damaged and were all engaged (vs some having a looser hole), the rocket would have had to be moving at 300fps or more horizontally to shear the nose cone pins with a drogue that lowers the rocket at 50fps, and Matt was flying drogueless so it doesn't seem to be explainable by horizontal velocity.

Perhaps the drogue shock cord tangled? That would cause a big yank, though his pins should've survived a nearly 40G event (150lb shear force and 4lb nose cone)

Some of the analysis is around the fact the main deployed. However, I haven't read here if the drogue deployed. Did you have all chutes out, or just the main at apogee?

The rocket was fully deployed (apogee shock cord, main shock cord, main parachute).

Your Eggfinder battery died, I think that that is actually a pretty established fact by now. Future advice: look at your Receiver until the countdown starts, and look for new packets being received and slight fluctuations in coordinates/altitude. The coordinates would be stagnant if the battery died while on the pad and the receiver would only report the last known coordinate. You could have stopped the launch had you been looking at it and saw stagnant coordinates.

Assumption 1: Your Eggfinder battery was dead before you launched the rocket.

Regarding your deployment, I see a couple potential points of clarification:

1. Are you positive that you put the shear pins in? It seems like you did a lot right before going to the pad (open main bay, hook up Eggfinder, seal Eggfinder, tape to cord, put the cone on, put rocket together, take pics, BS with people, etc. etc.) and it's feasible you forgot to put the pins in. Or, if they were loose, one or more could have fallen out.

2. How did you wrap your shock cord? If you z-folded and used masking tape you're probably fine, but there's a chance that a tangled and snagged cord could be the culprit. I've experienced a 25' cord getting knotted and behaving like an 8' cord, which in turn caused the main to come out early on a proven rocket design. I even have the video to prove it!

[YOUTUBE]tL1kdGE1O1g&t=39s[/YOUTUBE]

The key here is that, to Jim Hendrickson's point, I had an RDF beacon in it. Found it with very little difficulty thanks in large part to Dave McCann taking a series of photos as it descended into the trees. People don't think to do it, but TAKE PHOTOS AS A ROCKET LANDS!

And so...

Assumption 2: Your main came out because of inadequate or non-existent shear pins, and/or tangled drogue cord.

I definitely did have the shear pins in the nose cone. See below.

Agree 100% about pictures of the landing. Saved me twice this weekend. However, Matt's rocket was probably 2 - 3 miles away when it landed, and a smart phone camera isn't going to cut it. Without a really long lens, it won't be much good.

I wish I had been there to snap pics. I was out recovering when Matt launched. We saw it drift over, but did not see where it went.

blackbrandt said:

I definitely did have the shear pins in the nose cone. See below.

Click to expand...

Good deal! Now that still leaves whether or not they were adequate or if the drogue cord snagged. Unfortunately, I don't think you'll get answers to those questions.

Bat-mite said:

Agree 100% about pictures of the landing. Saved me twice this weekend. However, Matt's rocket was probably 2 - 3 miles away when it landed, and a smart phone camera isn't going to cut it. Without a really long lens, it won't be much good.

I wish I had been there to snap pics. I was out recovering when Matt launched. We saw it drift over, but did not see where it went.

Click to expand...

Dave has a nice camera and camera skills (said in Napoleon Dynamite's voice). You don't need a long range lens, just a series of pictures where you can make out the rocket. A decent SLR with short lens at full zoom will probably capture a rocket even that far out.

patelldp said:

Good deal! Now that still leaves whether or not they were adequate or if the drogue cord snagged. Unfortunately, I don't think you'll get answers to those questions.

Click to expand...

And that's something that will only be determined if I can recover the rocket.

Honestly, I don't care as much about the lost rocket from the monetary standpoint. I can mow more lawns and work more hours to pay that off. It would be nice to not have to build another giant rocket, but hey, it's the hobby.

I mostly want it back to figure out what went wrong so I can analyze it for future use.

patelldp said:

Dave has a nice camera and camera skills (said in Napoleon Dynamite's voice). You don't need a long range lens, just a series of pictures where you can make out the rocket. A decent SLR with short lens at full zoom will probably capture a rocket even that far out.

Click to expand...

PS...you can also (per one of Dave's other posts here or in one of the other threads about this rocket) do some math and count pixels to figure out how far out it was. It's not a perfect method, but it's better than nothing. Go to a Google Map of the site, mark the photographer's position, the sight line based on the last picture, estimate distance from the photo...start looking!

Shear pin holes were snug right? No abiilty of the nose cone to move before engaging them and all pins would have engaged at the same time right?

This thread is making me think I should confirm with a real-life test my calculation that on my L3 build my pins won't shear until ~85lbs.

rocketsam2016 said:

Shear pin holes were snug right? No abiilty of the nose cone to move before engaging them and all pins would have engaged at the same time right?

This thread is making me think I should confirm with a real-life test my calculation that on my L3 build my pins won't shear until ~85lbs.

Click to expand...

The shear pin holes were slightly on the loose side. However, we did have a piece of tape over the heads to ensure that they stayed in, as well as twisting the nose cone slightly to lock everything in place.

Each 4-40 should hold 50-76 lbs. My plan calls for 3. I do have cutting plates in place as well. I wonder how much they reduce the strength of the pins?

patelldp said:

PS...you can also (per one of Dave's other posts here or in one of the other threads about this rocket) do some math and count pixels to figure out how far out it was.

Click to expand...

The visual distance to the horizon from an observer on the ground is about 3.2 miles. So in this case the count pixel method would estimate the distance to the horizon, which is already a known quantity. Fortunately with the strong and steady upper level winds the direction of the wind was pretty stable at about 320-330 degrees. Get the exact value and you know the line. Start at 3 miles and draw the line to 5 miles. In between is where your rocket is. (more or less).

Matt, best I can do to help, which isn't much. Here is about what I saw. Anybody's guess. The large circle is a GUESS, not based on anything other than the direction of travel.

I really enjoyed your build thread and sorry to hear the result. My first L3 attempt wasn't successful either. I hope to see a new build thread for Hyperion II soon!


Sent from my iPad using Tapatalk

Here is my guess.

Wind data suggests wind direction was 320 - 330. I drew the line at 325 degrees from the away cell. Length is 4 miles. If you have better data on actual upper level wind direction let me know and I will redraw the line.

Did you have shear pins on booster tube to electronics coupler?

I have to agree with mikec. The eggfinder TX should not be operated under any circumstance from a 1S Li battery pack. Even with the best LDO on the market, it would still be questionable. Because the eggfinder has seperate LDOs for the locator TX and the flight computer, that would explain why you lost lock but still deployed. The drop across the LDO increases as you draw more current, so the TX side could shutdown even though the flight computer side was still running.

mccordmw said:

Each 4-40 should hold 50-76 lbs. My plan calls for 3. I do have cutting plates in place as well. I wonder how much they reduce the strength of the pins?

Click to expand...

I would guess a decent amount? I have a knife that can cut an orange in two with no downward force from me other than the weight of the knife. A plastic knife, even if weighted to the same weight, not so much I believe (but someone else I'm sure can speak more authoratively on this!) that the shear pin strength calculations assume a true shear vs a cut with a blade.

I'll certainly be confirming the actual yield strength of my shear pin configurations going forward to make sure things like looseness, asymmetric loading, and sharp edges haven't broken my assumptions.

jderimig said:

Here is my guess.

Wind data suggests wind direction was 320 - 330. I drew the line at 325 degrees from the away cell. Length is 4 miles. If you have better data on actual upper level wind direction let me know and I will redraw the line.

View attachment 317023

Click to expand...

Man, I wish I could get back out there and look. I live 75 minutes away, so it's no casual drive. And we don't return to Higgs until after crop season, usually Nov. or Dec.

Banzai88 said:

I am beginning to question this premise.

I have two rockets configured this way, and I keep having issues with the main (RRC3) and the back up (RRC2) firing at about the same time (with the RRC2 set with the delay). I have confirmed that it's NOT because the charges are interacting with each other and ripple firing. This is in 2 different rockets with 2 different sets of computers. I've had it happen twice on each rocket.

The first time it happened at about 3500 feet on each. After re-checking my math and again doing ground tests, I flew them again on motors to go about 1800 feet to make it easier to see. Sure enough, same darn thing, both ripple fired.

Don't rule it out just because it didn't happen on the ground (you did ground test your main charges with the back ups in place to rule out a ripple fire potential?) that it won't happen in flight. Mine NEVER failed on the ground in every ground test that I did, but BOTH rockets failed in flight BOTH times.

Since moving to only one altimeter in each rocket, they've flown 100%.

I'm thoroughly convinced that the sense of the RRC2 and RRC3 are so close that the delay set up for back up just hasn't worked for me. If I were to NEED to guarantee the back up, I would use 2 of the very same computer.

Click to expand...

Ok, So lessons learned are backup RDF tracker though if the Eggfinder was powered by a 2S/7.4V pack AND the power supply wasn't ripped from the board, the flight could have been tracked.
With all that space and small size of RDF trackers just tack one on the apogee harness.

If the Missileworks electronics have the potential to behave like Tom clearly states, could have had the Main shearpins give way by an exuberant, simultaneous apogee charge firings. Easy enough to remedy by using a
different device.

I hope the rocket is found so the failure scenario can ultimately be determined. I do think the suggestions made by many, if heeded would lead to a successful flight next time. Kurt

Bat-mite said:

Matt, best I can do to help, which isn't much. Here is about what I saw. Anybody's guess. The large circle is a GUESS, not based on anything other than the direction of travel.

View attachment 317021

Click to expand...

I was with John when we saw it. I wish I would have stayed with it. We were a little further north in the field that the me circle in the pic. Based on what I saw the arrow in the picture should be shifted up to the 313 marker right about it. That was consistent with the area I pointed out to you Friday afternoon. At that point we turned away and it was still between 2,000 and 3,000 ft AGL. Seeing the grid and going off what I saw I would have pegged the search radius right around mission branch road along that line. You didn't go back that far did you?

Regarding the Lipos, for reference my 2S lipo running the TRS was at 8.3V at launch. I had a 3db antenna on both the TRS and receiver and had packet updates all the way down to touchdown.

Here are my suggestions based on your experience and mine.
- Get a team of spotters in place. One to look at GPS and/or RDF and 2-3 spotters for visual track (you being on of them). I wish you would have waited for John and I to get back.
- Get a recovery vehicle idling and ready to go. You/your spotters could have probably gotten under it/close to it if you would have been ready to jet out. This was more important than I realized on a windy day after seeing no less than 5 big rockets including my own hit the ground and get dragged across the sod with each wind gust.
- The RDF tracker Gus had in John's F200 came in handy. I would definitely couple one with a GPS tracker in the future.

NateLowrie said:

I was with John when we saw it. I wish I would have stayed with it. We were a little further north in the field that the me circle in the pic. Based on what I saw the arrow in the picture should be shifted up to the 313 marker right about it. That was consistent with the area I pointed out to you Friday afternoon. At that point we turned away and it was still between 2,000 and 3,000 ft AGL. Seeing the grid and going off what I saw I would have pegged the search radius right around mission branch road along that line. You didn't go back that far did you?

Click to expand...

Mason Branch is one of those creeks, just FYI.

Bat-mite said:

Mason Branch is one of those creeks, just FYI.

Click to expand...

My bad. Shows how much I know about the launch site.

I suspect that the main deployed at apogee due to insufficient shear pin retention on the nose. On our big rocket (payload 36"x12.625") we used six 4-40 pins that were all tightly threaded in. We also used about half (4 grams of FFFF) the ejection charge. Even a little slop (it sounds like there was some) makes it very much easier to break the pins as well.

RocketFeller said:

I suspect that the main deployed at apogee due to insufficient shear pin retention on the nose. On our big rocket (payload 36"x12.625") we used six 4-40 pins that were all tightly threaded in. We also used about half (4 grams of FFFF) the ejection charge. Even a little slop (it sounds like there was some) makes it very much easier to break the pins as well.

Click to expand...

I've think it's the opposite. If there is slop in the shear pin hole the pin has a good chance of entering an elastic deformation state. If it deforms it can take 30-40% more energy to finally shear. Having the hole as tight as possible (i.e. tapped for 4-40) pretty much ensures a clean break.

I do think you needed at least 5 in the nose.

NateLowrie said:

If there is slop in the shear pin hole the pin has a good chance of entering an elastic deformation state. If it deforms it can take 30-40% more energy to finally shear

Click to expand...

I'm not a mechanical engineer, but couldn't you have the case where the shear pin holes deform unequally (i.e. the nose cone is separating in a lopsided way).. then instead of breaking 3 shear pins, you have a force that's breaking one shear pin at a time? Thus requiring less overall force?