Monday, June 27, 2016

[Updated] M2080 recovery failure analysis - Part 1

[Please see the notes in the entry in brackets, the addendum at the bottom of this entry, as well as part 2 of this report above this entry.]

This has been a rough entry to ponder but I need to commit to history and move on.  In memorium...  I had flown this Wildman Extreme DarkStar workhorse 10 times over the course of ~4.5 years before the April 2016 recovery failure at Holtville:
The recovery recipe I used for the first L585 flight remained consistent and served me well up through the M3100 flight.  The only difference in the setup for the most recent M2080 flight was the use of M-tek/Firewire electric matches (blue&white-striped leads) rather than J-teks (solid yellow or black leads). To my knowledge J-teks have never failed in any of my flights. I should note that at least one of the two Firewires used for apogee/CD3 deploy worked as the drogue was out but... Despite using separate, heterogeneous, redundant altimeters neither of the M-teks used in the main parachute's CD3 system fired as you'll see below.  For the record the Raven 3 used a freshly charged LiPo battery and the ARTS2, set for single-battery mode, used a fresh 9V battery.
     According to their website these Firewire Initiators should measure 1.0  +/- 0.2 Ohm:
FirewireInitiatorResistanceSpec
Here I'm disassembling the CD3 system for the first time after impact: MTek-RemovalFromCD3
After dissecting the nearly totaled rocket I tested the intact matches and both measured high for resistance; 3.5 and 4.1 Ohm, respectively:
Untitled Untitled
I also fired both matches with my ground ignition system after measuring their resistances so I remain thoroughly stumped.  Click to play these videos:
Video May 13, 5 32 33 PM 4.1 Ohm Firewire Initiator test
I had heard that there were issues with early batches of these matches but I ordered two boxes with 7 ft. leads (40 per box) directly from the vendor on 12/17/2015.  I see no lot code or batch date on either box [lot numbers are on the underside of the taped lid but I had only been accessing them from the sides of the boxes and I shared them with the vendor] so I have no idea how to identify these except for the order date.  That's all I think I know except for sadness.  After colliding with the sturdy concrete, nozzle-first, I was only able to salvage one Pro75 threaded closure, the ARTS2 (I think it still works?), the Raven 2, and my Beeline GPS tracker.  If I were to replace all that I lost it would run a hefty toll, $850, + shipping:
Painful.  I will contact the Firewire Initiator manufacturer tomorrow but I highly recommend that folks not use that product for recovery. [Instead just be sure to test the initiators for resistance and verify that they work with your electronics.] I will [likely] only use the remainder of those two boxes for ground ignition where failure is an option.  Behold the carnage:
Untitled
Direct impact on concrete with only a small drogue parachute.

Untitled Untitled Untitled
I had to saw the airframe off to try to get the bent casing out but no dice. After hacksawing off the forward end of the motor I discovered that the forward closure parts were dented and damaged beyond repair anyway:
Untitled Untitled Untitled Untitled
Similarly the CD3 for the main compartment bent on impact such that the red anodized aluminum cylinder was totaled.

Untitled
And this demonstrates that the 16g CO2 cartridge remained intact. The only elements that failed were both of the main Firewire Initiators. Did I mention that? Ugh.

Addendum:  After additional testing at the request of the initiator vendor and considering that the system changed post-impact I can no longer assert that the matches were definitely to blame with any certainty.  All I know for sure is that both initiators measured high and well out of spec for resistance after extracting and measuring them post-mortem (3.5 ad 4.1 Ohms, respectively).  See part 2 of this report above on 7/31/16 for details.

1 comment:

Anonymous said...
This comment has been removed by a blog administrator.