Dissecting altimeter data

Recently I had some anomalies on a launch of my 54mm MD Tomach to 15,000+’ AGL @ a velocity of Mach 1.8ish. The accent was nominal however the decent appears to had some issues, which I would appreciate some insight.

There were two apparent anomalies. One was a slow decent rate from apogee. The other, my primary SL100 altimeter stopped reporting data @ 1,889’ AGL or 543.55 seconds after launch detect. The rocket was recovered without issue 2.4 miles away.

There are two scenarios I was considering for the slow decent. One being high winds, which was confirmed, the other being my main deployed @ apogee, for whatever reason (electronics or failed shear pins). I am leaning towards the premature main deployment, based on the decent rate and consistency of it recorded on the altimeters.

As for the lack of altimeter data record, I have two working theories. First off I am not 100% sure if my altimeter was off when I recovered the rocket, I am pretty sure it was off when I think back. However as it was recovered 2.4 miles away close to an hour later, at the end of the day, I was more concerned about finding the rocket and just put it in the back then drove away and turned off the electronics when we got to the bar J so I am not 100%. The first theory is that there is an issue with the altimeter; the second, that there was an interruption in power. My gut tells me power, as all tests on the altimeter have come back nominally, however the wiring also checks out…so it’s a bit of a mystery.

A bit about the rocket and altimeters: This was the rocket’s fourth flight and the AV bay electronics has been put through the mock deployment bench/ground tests, at least couple dozen times now. Besides working out the BP charge (1.5 g), I have never had an anomaly. The AV bay is literally keyed, so it is impossible to assemble upside down, (drogue instead of main or vice versa). The wiring is hard wired to the drogue bulkhead and keyed to a removable forward, main bulkhead. There is adequate slack in the wiring harness, in case of any movement. The battery (1S Li-Po) was fully charged @ 4.2, and just shy of 4.2 volts two weeks later. The altimeter was tuned back on post flight, same battery two weeks later, to grab the data and run tests; it ran for about 1.5 hours still reporting about 4.1 volts. The altimeters are both StratoLoggers, main (which had the anomaly) an SL100, the backup (which performed nominally) the new SLCF. Wiring is all stranded 20 gauge, with all connectors securely crimped and finished with heat shrink tubing. The terminals on the featherweight magnetic switch are soldered to the ring terminals after they are first crimped to the wires, followed by an inspection under a LED magnification lens then tested with a bench top multi-meter, which is sensitive to 1/100^th of an ohm. All wiring and connections were inspected and tested post flight, and all nominal. The rocket was about 11lbs on the pad and 7-8 something on decent. I use a 12” Fruity chute classic elliptical drogue, which normally provides a 75-90’/s decent rate, depending on AGL. The main is normally 25’/s. The drogue is deployed when the altimeter bay, payload tube and nosecone separate from the booster via dual 1.5g charges, 1 second apart shearing 3 2/56 nylon shear pins. Main is deployed after dual charges 200' apart using 1.5g charges to separate the nosecone from the payload tube and altimeter, after shearing 3 2/56 nylon shear pins.

Attached are the two altimeter log files, “Tomach SL100 Loki K830 SF Aug 29 2015.xlsx” is from the main altimeter which had the anomaly. “Tomach SLCF Loki K830SF Aug 29 2015.xlsx” is from the backup altimeter.

View attachment 271980
View attachment 271981

Based on the descent rate it seems obvious to me that you deployed your main at apogee. Your descent rate is in the low to mid 20's from just after apogee to touchdown at 270 feet. If your main would have deployed at 700 feet the rate would have dropped off shortly after that. Plus, you say that normal descent rate under drogue is in the 75 - 90 feet per second range. Whether the main deployed due to failed shear pins or a wonky altimeter is difficult to say.

I looked up the memory limitations on the two altimeters:
SLCF has a memory capacity of 18 minutes (approximately 1,080 seconds) per flight.
SL100 has a memory capacity of 9 minutes (approximately 540 seconds) per flight. So this explains the shutdown of your main altimeter.

Main deployed at apogee, looks pretty clear from the data. You're lucky it only ended up 2.4 miles out.

Zeus-cat said:

I looked up the memory limitations on the two altimeters:
SLCF has a memory capacity of 18 minutes (approximately 1,080 seconds) per flight.
SL100 has a memory capacity of 9 minutes (approximately 540 seconds) per flight. So this explains the shutdown of your main altimeter.

Click to expand...

That certainly explains the truncated logs, but I am a bit surprised with the memory limitation. My interpretation of the memory limits was for historical data and assumed it would roll-over, not shut down in-flight.

This makes me believe that the manufacturer's claim of good to 100,000’ AGL is not a very grounded claim…at least I wouldn’t choose it.

I enjoy the challenges of higher altitude launches and this is just another reminder of how little I know. Going forward my altimeter choices will be based on a worst-case formula, something like a decent rate of 15’/s or even ultra conservative 10’/s. Regardless of the formula I believe it to be prudent based on the limitations of the technologies.

Now that the truncated data log mystery is solved I need to focus on why my main deployed at apogee. I have already discounted a miswired altimeter or altimeter failure based on inspecting and testing the altimeter post flight. Also the logs clearly show deployment charges where fired not main charges, at apogee.

Some considerations:

My BP charge of 1.5g is 50% higher than the basic amount required, that being 1g. If both charges 3g went off close enough together I wonder if the force would have been enough to cause the nosecone shear pins to shear when the 20’ Kevlar harness maxed out. Kevlar does not have much elasticity.

Both logs show that apogee happened around 29.4 seconds and both logs show very high velocities just after apogee quickly followed by very abrupt -v numbers…thoughts?

If this were the case then possibly changing my backup delay to 2 seconds would be prudent. Both altimeters show a small change in alt at the 2-second mark after apogee. I also wonder what the net effect of 1.25g vs. 1.5g BP would be. And finally if I can squeeze a 25’ harness the extra 5 feet may help in bleeding off some force.

mikec said:

Main deployed at apogee, looks pretty clear from the data. You're lucky it only ended up 2.4 miles out.

Click to expand...

The decent on main averaged out to 26+'/s, slowing as it got closer to the ground, which helped. Plus the farmer who's filed I landed in was very accommodating and had to be the nicest farmer I have ever met...so lucky all around.

Do you do anything with the Kevlar cord? To help absorb some of the energy I normally fold the cord in about 1 foot loops multiple times and tape the cord to itself. I will do this process multiple times along the length. So the cord has to break the tape which absorbs some of the energy. The orange represents the Kevlar cord and the blue represents the tape. You can control the amount of energy by the number of tape lines, the number of wraps of tape, etc.

Zeus,,
Fantastic catch...
I've recently had mechanical switch problems,,
so that's where my mind was leaning till I read your post..

Michael,,
I love how technical you are with your flights...
That's why you enjoy pushing the envelope,,
because those flights are less forgiving / more challenging...

Very cool...

Teddy

mpitfield said:

That certainly explains the truncated logs, but I am a bit surprised with the memory limitation. My interpretation of the memory limits was for historical data and assumed it would roll-over, not shut down in-flight.

This makes me believe that the manufacturer's claim of good to 100,000’ AGL is not a very grounded claim…at least I wouldn’t choose it.

Click to expand...

I don't think it shut down, it just stopped recording data. I have an SL100 and it stops recording data after it sits on the ground for a brief period. I assume it looks at the change in altitude and if sees nothing for a certain number of samples it stops recording and goes into post flight beeping mode. That is not the same as shutting down. In the case of your flight I assume it would keep sensing altitude changes and stay in flight mode even if it was not recording the data. This is easy to figure out; did the SL100 fire the main charges? If it did, they most likely fired when they were supposed to. Maybe you can contact Perfectflite and find out if the altimeter stays in flight mode even after it stops recording so we know for sure.

My opinion is that your failure mode is that your drogue charges severed your shear pins and caused your main to deploy at apogee. Fix that and the rest of the issues probably go away.

Zues-Cat is correct on all counts.

1. Your main deployed at apogee. Both altimeters indicate that.

2. Original StratoLogger stores 32 flights with ~9 minutes of data at 20 Hz. (10,800 points) Longer times if you reduce the data rate. You ran out of memory before the main charge should have fired.

3. The new StratoLogger stores 16 flights with ~18 minutes of date at 20 Hz. (21,600 points) Longer times if you reduce the data rate. This altimeter would have deployed the main at 500' if it had not been deployed at apogee.

4. All both use the same barometric sensor which is accurate to 100,000'.

2., 3., and 4. are found in their respective manuals.

Bob

les said:

Do you do anything with the Kevlar cord? To help absorb some of the energy I normally fold the cord in about 1 foot loops multiple times and tape the cord to itself. I will do this process multiple times along the length. So the cord has to break the tape which absorbs some of the energy. The orange represents the Kevlar cord and the blue represents the tape. You can control the amount of energy by the number of tape lines, the number of wraps of tape, etc.

Click to expand...

Hi Les, yes I normally do tape the z-folds on my recovery gear but not on this rocket. The space is very restricted and it is quite challenging to fit everything in. Having said that I just had Gene at Fruity Chute make me some custom 12" Classic Elliptical compact chutes, so I may be able to move back to a 25' harness...Teddy will be getting another call.

I think my plan of action will be the following:

- move from a 20' to 25' cord
- adjust my redundant delay from 1 to 2 seconds
- back off on the over charge from 1.5g to 1.2 (1g was successfully ground tested)
- see if I can add some tapped off z-folded sections on the recovery harness
- add some self-clinching nuts to the back of the nosecone shear-pin holes, which is what I have on the booster section.

The last change allows me to easily screw the shear-pins in and out without adding any stress to them and they shear just as easily. I did remove the shear pins twice on the nosecone, pre-launch, and it very likely reduced their effectiveness in the process.

This should further mitigate the possibility of this re-occurring.

bobkrech said:

Zues-Cat is correct on all counts.

1. Your main deployed at apogee. Both altimeters indicate that.

2. Original StratoLogger stores 32 flights with ~9 minutes of data at 20 Hz. (10,800 points) Longer times if you reduce the data rate. You ran out of memory before the main charge should have fired.

3. The new StratoLogger stores 16 flights with ~18 minutes of date at 20 Hz. (21,600 points) Longer times if you reduce the data rate. This altimeter would have deployed the main at 500' if it had not been deployed at apogee.

4. All both use the same barometric sensor which is accurate to 100,000'.

2., 3., and 4. are found in their respective manuals.

Bob

Click to expand...

Thanks Bob and Zues-Cat.

I was aware of the specs on the number of recorded flights, as well as the accurate to 100K, but not in the detail you provided Bob. One thing I was not aware of was the logging limitation...good to know.

I confirmed the following:

- reviewed my post flight check and yes the main charge did go off
- confirmed with PerfectFlight, the implications the RAM limitation does not stop the altimeter's other functions.

So my number two anomaly is nothing more than a memory buffer filling up, and by design...this allows me to focus on the root cause, and continue to have faith in the StratoLogger products.

You can really control exactly how much of a buffer you give the separation,,
you're already "Z" folding..
A single wrap of masking tape on every "Z" folded bundle of 3 "Z" folds...
Do a low flight so you can watch the apogee separation happen...
Then you can adjust the buffer,, more or less according to what you saw..

Teddy