A bit of experimental data.
The 17 years of the ARLISS (A Rocket Launch for International Student
Satellites - www.arliss.org <http://www.arliss.org/>) program has some
interesting data to contribute to this discussion. This is the first and
oldest CanSat competition - begun by Professor Bob Twiggs (CubeSat co-inventor)
and continued by AeroPac in partnership with Unisec Global (unisec-global.org
<http://unisec-global.org/>). Each year we fly about 50 flights of 15 kg
airframes with 1 kg university student payloads to about 3k meters on M motors
… where the payloads are deployed at apogee for autonomous return … each
payload is an autonomous robot. Most land to crawl home across the Black Rock
playa floor, but some fly in various ways. Each payload has a rather complex
DIY electronics payload as well as the flight avionics … subject to boost
acceleration of up to 8 Gs plus deployment shock (from either black powder or
CO2). We now have about 800 flights on M motors with this flight profile
during the program - on a fleet of very similar airframes, on essentially the
same motor - the Aerotech M1419.
In 2013, we decided to investigate more deeply about the flight dynamics of our
student payloads and designed a custom flight recorder with 3x accelerometer +
3x gyro with a recording rate of ~1.3 kHz to examine these dynamics. A team
of our members, led by Bob Feretich, designed and implemented the custom
datalogger and post processing software
(http://www.rafresearch.com/rocketdatalogger/index.html
<http://www.rafresearch.com/rocketdatalogger/index.html>) to implement this
data collection and analysis. A portion of the database of these flights from
2014 can be found here
(http://www.rafresearch.com/arlissdatalogger/flightdata/ARLISS2014/index.xml
<http://www.rafresearch.com/arlissdatalogger/flightdata/ARLISS2014/index.xml>)
including spectral analysis of the experienced by the payloads during boost and
deployment.
One of the surprising results of the analysis is that while there are some very
large magnitude shocks (up to 50g! - particularly from black powder deployment)
- there is little correlation with payload electronics failure. Remember
these are student payloads … so build quality is highly … variable. We
speculate that while the absolute magnitude of the shock can be high, the
duration is very short … so the total energy in the shock is modest.
We think this will correlate well to the shock profile for ARLISS Extreme
suborbital flights as well. We will fly one of these recorders this coming
season to confirm - but the motors and deployment methods are not dissimilar.
A paper describing an overview of this recorder can be found here
(https://www.dropbox.com/s/1kpkt3dg6038pbx/ARLISS%20Data%20Logger%20Project%20Part1.pdf?dl=0).
Bob Feretich has these recorders for sale as well.
Here is an overview of the ARLISS program.
https://www.dropbox.com/s/5f6sl7dsr5lrdl8/Sport%20Rocketry%20ARLISS%201.2014.pdf?dl=0
A copy of Dnepr user manual is enclosed … this has some standards for testing
payloads on a big solid to LEO - note pages 54-58 for specs on typical payload
acceleration and deployment shock..
https://www.dropbox.com/s/ncnk03oly4eodxu/Dnepr_User_Guide.pdf?dl=0
<https://www.dropbox.com/s/ncnk03oly4eodxu/Dnepr_User_Guide.pdf?dl=0>. To my
eye, the boost acceleration magnitude and the deployment shocks are not
dissimilar between Dnepr and ARLISS and perhaps might be extrapolated to other
solid fueled systems.
We see very few flight avionics failures (three out of 800 flights - and we
ascribe those to pilot error) and most of the student failures seem to be from
connector failure on landing shock.
Ken
On Dec 24, 2015, at 3:00 PM, Henry Spencer <hspencer@xxxxxxxxxxxxx> wrote:
On Thu, 24 Dec 2015, Henry Vanderbilt wrote:
Playing catchup here... I can't prove it, but my suspicion is that amateurs
could get most of the benefit here at a fraction of the cost. If your
electronics can't stand being bolted in the bed of a pickup truck for three
fast miles down a dirt road...
One of the folks involved in the Oersted magnetometry satellite (Denmark's
first scientific satellite, launched 1999) once commented: "the launcher
people suggested that a very good approximation of launch conditions were to
strap the prototype PCBs directly to the engineblock of a two-stroke car and
drive around with it for a couple of hours".
That may have been a rather extreme worst case -- I think Oersted went up as
a secondary on a Delta II, which is notorious for giving its payloads a rough
ride. (Our lab director has been heard to say that when the very last Delta
II flies and there is thus no chance that anything of ours will ever be
launched on one, we're going to ceremonially burn the Delta II
secondary-payload manual.)
Wholly scientific, no. But then adding some sort of decent-bandwidth
vibration/shock sensor - a small weight suspended on three axes with strain
gauges on each axis? - is there such a thing off-the-shelf?
There are off-the-shelf accelerometers for vibration measurement -- the
giveaway is that the DC response (to steady acceleration) is lousy or
nonexistent, while the response to oscillating loads is specified up into the
kilohertz range -- and I expect you could get complete three-axis packages,
but I've never gone looking.
Henry