Revised maximum altitudes

1 view
Skip to first unread message

Eric

unread,
Jul 31, 2008, 2:02:12 PM7/31/08
to HALE TEAMS
After going over the GPS logs yesterday, we have the final maximum
altitudes

Balloon #2 (first one launched with FLL team 90, LEGO Mindstorms,
Gypsy, Little Joe) went to 99,712 feet

Balloon #1 (second one launched with LUXPAK, Brix-catcher, Peeps, and
Reel-E) went to 99,570 feet

Little Joe was released at 82,338 feet

I've posted a graph of all 3 altitude profiles on the mobileme
galllery.

I'm going to get the Excel files in a presentable format and then post
them on the HALE website so that you can also have the GPS locations
(lat and lon). I'm still working on getting a google earth file - new
to this so it's taking a while.

It looks as though the two balloons passed by each other at 8:16:41
PST at an altitude of 55,500 feet. They were further apart than I
thought - 11 miles, which will make it difficult to see in a photo.
Balloon #2 was on it's way down and balloon #1 was on its way up -
exactly due East from balloon #2 - which means the photos will have
the sun it in just to make things more difficult. We can always
hope...

Eric

Eric

unread,
Jul 31, 2008, 2:15:45 PM7/31/08
to HALE TEAMS
I just calculated that on it's way down, Little Joe passed Balloon #1
at 8:01:04 PST at an altitude of 33,000 feet and at a distance of 15
miles. I don't think there is much chance of Brix-catcher seeing
Little Joe in a video at that distance. But Eugene may want to look at
the video VERY closely around that time.

Eric

Brian Davis

unread,
Jul 31, 2008, 4:30:51 PM7/31/08
to HALE TEAMS
[I started out with a short description. Honest I did.]

On Jul 31, 2:02 pm, Eric <lego.profes...@gmail.com> wrote:

> After going over the GPS logs yesterday, we have the final maximum
> altitudes
>
> Balloon #2 (first one launched with FLL team 90, LEGO Mindstorms,
> Gypsy, Little Joe) went to 99,712 feet

I've been digging through the data from Gypsy & Lil' Joe as well. The
pressure log from Gypsy looks great - lowest pressure was 1.4 kPa
(right at about 0.0138 Atm). In trying to convert these to equivalent
altitudes I've two problems: knowing the elevation of the launch and
landing sites (as calibration points) and trying to convert pressure
to altitude accurately. Using a constant scale height doesn't cut it
(overestimates the peak altitude 20,000'), but I've not tried a
"standard atmosphere" look-up table or similar yet. It does however
show some wonderful detail, and was logged about every 3 seconds.

Gypsy recorded 4,396 individual timestamped records in six different
formats:

2,835 "Ambient" records, of pressure, sensor temperature, and light
levels (CDS sensor). Light levels did increase slightly with altitude,
as I'd expect due to less atmosphere. The data has a distinctly
bimodal distribution - almost certainly due to the sensor rotating
first toward then away from the sun which would have been at low angle
(near horizon) during the flight. Once I correlate these with the
pitch angle, I'll be able to find light levels for "up" and "down" as
well (as the platform was pitched to those orientations). Between MET
(Mission Elapsed Time) 6.55 and 13.67 minutes, you can see the peak
CDS readings drop distinctly - I suspect this was a cloud deck the
balloon went through, which a rough estimate for from the pressure
data would be between 14,000 and 17,500 feet (based on a launch
elevation of 4,700', and a lot of guessing and approximations).
Landing was at MET 114.4 minutes according to Gypsy's pressure sensor
& internal clock. The balloon burst at about MET 78 minutes (all times
sort of rough, as I'm reading these off of graphs, not the data right
now).

651 "Averaged" records, of minimum net acceleration, average sound
level, and peak sound level during the previous 2 second interval. It
turns out most of the mission is silent (about 5.5 out of 100), as
you'd expect, but with "creaks", "pops", or "groans" all over the
place - the peak sound levels during the intervals were all over the
place. At cut-down, however, there *was* a significant amount of sound
- the average level registered jumped to about 22 out of 100, which is
remarkable considering that the ambient pressure was so low (below 0.1
Atm) that "sound" transmission should be very weak (and was, see
below). This continued through the entire descent, with the sound
level falling to 4.7 after landing. also after landing, the peak sound
level was much lower (average around 18 out of 100) and with very
little scatter - those "creaks", "pops", & "groans" were gone after
landing. The "acceleration" data i messed up somehow - I think I had
it set to record the minimum, instead of the maximum, acceleration
during the 2 second interval. But it's still very interesting. The
balloon was clearly swinging & swaying during the ascent with steadily
decreasing amplitude up to about MET 40 minutes. After that there is
very little motion until the balloon breaks at MET 76 minutes, when
the minimum accelerations are equal to those early in the ascent
phase. On the ground, the minimum acceleration is very similar to (if
not slightly noiser than) what the balloon experienced aloft... in
other words, again, the flight of the balloon high in the atmosphere
is amazingly calm, quiet, and undisturbed. until the balloon goes
away, of course.

709 "Engineering" records, each of two more internal temperature
measurements (one on the NXT batteries, one near the camera
batteries), the NXT battery level (wow do those Energizers last!), the
heater circuit status (on/off) and the position of the motor
controlling the platform pitch. Curiously, this last seems to show
that at least at one point the platform "drifted" while the motor was
floated (a problem to solve for the future). The thermal data shows
that the heater circuit was turned on at MET 8.8 minutes when the
sensor near the camera battery dropped to 282 K (9 C or 48 F), and
remained on until MET 171 minutes, nearly an hour after landing (MET
114 minutes). The heaters evidently helped, with the temperature of
the NXT leveling off near 284 K (11 C or 52 F) and the camera (near
the front of the payload, with the hole cut for the lens) dropping to
a chilly 269 K (-4 C or 25 F). But when the balloon popped... rapidly
all three payload temperatures start a downward trend - clearly the
thin air at this altitude was swirling through the payload from the
lens hole, chilling things at roughly 1.5 K/minute (the rearmost
sensor, protected in the pressure sensor housing, dropped somewhat
slower than this). By MET 98.5 minutes the temperature trends reversed
sharply (reaching a low of 265 K (-8 C or 23 F) for the NXT and 254 K
(-19 C or -2 F) for near the camera), climbing again rapidly until at
MET 119 minutes they were all at essentially the same temperature (278
K or just above freezing), and then they continue to warm
significantly slower until the data runs out (peak temperature of 302
K (29 C or 84 F), and I've not doubt it got warmer before the team
finally picked it up). Curiously the payload internals, very near
freezing when they hit the ground, warmed far *slower* after landing
than during the fall, indicating that it was the air motion and front
opening that was drastically changing the payload temperature, not
conduction through the payload shell. The interesting thing about the
battery level data is how "boring" it is - after an initial drop to
9.02V, the battery voltage slowly drops to a minimum of 8.77V right
about the time of landing, then continues and almost linear climb
after that to 9.18V. Clearly the batteries cooled down slower than the
sensor, but also bottomed out and warmed up slower... yet none of this
remotely approached critical power levels.

153 "Imaging" records: yep, the program was reading the script files,
and executing the sequence perfectly, even if for some <bleeped!>
reason the camera has no pictures. I'm thinking William Shatner may
have seen something sitting on the payload during this flight...

26 "Pitch" records, recording when the platform was commanded to
various angles. It also recorded what the sound levels where while the
motor was running, while nothing was running, and while the NXT was
playing a loud "Beep!". The sound level drops remarkably as the
pressure drops but does not go to zero, probably due to solid
conduction through the structure (I could test this at home because my
cheap "vacuum chamber" isn't nearly big enough to hold the payload in
flight configuration). I have to study these a little more, but
there's something odd about these records, and it may tell me
something about what was happening to the platform orientation during
descent.

22 "Video" records, indicating a command to either turn the camera on/
off (this happened about 10 times) or to start/stop a video. which, as
I've mentioned, didn't work for unknown reasons. Nothing exciting, but
at least this way I know exactly where the program was in the various
script files during the mission.

> Little Joe was released at 82,338 feet

...and fell like a fairly high-velocity big blue rock. The individual
moment-by-moment velocity from the GPS logging data shows "spikes" as
high as 500+ mph, but a more careful analysis suggests Lil' Joe hit
more than 100 mph during the fall... and had a rather rough ride.
While hanging below the ballon was literally as calm as sitting on a
table (the accelerometer data does not change during the first two
seconds of the record), within 1.5 seconds there's evidence of
rotation or a "wobble" (remember, it's falling tail-first initially
with the heavy end "up", so it has to do a flip somewhere during the
descent), with it perhaps flipping over around 8-9 seconds after cut-
down. I have high-resolution (about 30 Hz) 3-axis data for the next
110 seconds, during which the acceleration exceed the 2.5 G limit of
the sensor along at least two of the three directions. Net
accelerations probably topped out below 4.5 G's, but there's a lot of
scatter in the data. Parchute deployment was at 61 seconds after cut-
down... but remarkably does not show up in the acceleration record at
all. Either the parachute was torn out of the payload earlier (and the
resulting deceleration lost in the acceleration "noise") or it's
deployment was so ineffectual it isn't visible in the data. Given the
way it was falling, I've no way to know which is the case, but it
seems unlikely that the deployment was really early... and if the
parachute had somehow been puled out during ascent (*very* unlikely),
the acceleration signal of the payload with the parachute pre-deployed
I think would have been obvious.

After 120 seconds of high time resolution data, Lil' Joe switched to a
"log 2 seconds, wait 58 seconds" pattern for the rest of the flight
down. This allows me to pinpoint the landing time to within a minute
(originally Lil' Joe was not going to have a GPS logger... thanks for
that, Eric & Co.!). The 2 second "windows" of data during the descent
show that the payload was swinging, but I'm not sure how to interpret
it yet... the "downward acceleration" remained around 0.5 G's, not the
1 G i'd expect even if it was swinging at terminal velocity below a
parachute. I've still got a lot that can be teased out of this data,
but it will be "interesting".

The last 10 seconds of GPS data show an average descent speed of about
30.4 ft/sec... it hit the ground going only 20.7 mph, much lower than
my previous estimate of 30 to 50 mph. Still faster than I like, but
I'm not sure how critical a parachute is anymore for a payload like
this. I'm beginning to suspect a suitably "fluffy" payload may not
need a parachute at all.

Remind me never to skydive from 82,000'. It really doesn't look like
fun. In fact, it looks fatal.

Eric, one thing I don't have (and dang it, I should have) is the
moments of inertia and center of mass of the payload, and where those
are relative to the sensor. I think i can get those, but... I need the
right mass distribution. If you are shipping me everything (parachute
included) for me to analyze (untangle) and ship back to you, please
include the webbing, SPOT unit with original batteries, etc. that was
I can hopefully tease more out of the acceleration data about what the
free-body motion of the payload was. It was complicated, that's for
sure. If you don't want to ship all that stuff back here, if you
could weigh the SPOT unit at least and estimate where it's CofM is, I
would really appreciate it. I honestly never thought this far ahead,
or thought there would be this much information in the data.

> I've posted a graph of all 3 altitude profiles on the mobileme
> gallery.

Very cool. I look forward to the other GPS datasets (with altitudes?).
On the Lil' Joe GPS set, in detail there are a number of "adjustments"
to the altitude, where the payload "suddenly" jumps up or down
somewhat irregularly in height. I'm not sure what to make of these, or
the times when the GPS cuts out altogether for periods of time... but
to describe the landing, whatever happened the GPS unit doesn't report
in for the next 7 minutes after "landing", half the length of the
entire descent. No accelerating data on the impact, as it happened
sometime during the "wait 58 seconds" portion of the terminal logging
pattern. I do know that the payload didn't move after that :) - no,
Eric, no record of the team picking it up off the desert, as the NXT
had completely filled it's memory prior to that point.

> I'm still working on getting a google earth file - new
> to this so it's taking a while.

That will be incredibly cool - thank you Eric!

> Balloon #2 was on it's way down and balloon #1 was on its way up -
> exactly due East from balloon #2 - which means the photos will have
> the sun it in  just to make things more difficult.

Well, we can always have fun looking... the SLR camera was on Balloon
#2, but the Brix-Catcher video and ride-along student payload (that
took photos?) were on Balloon #1, so there should be some pictures
from each towards the other...

--
Brian "been doing a little bit of analysis" Davis

Eric

unread,
Jul 31, 2008, 5:19:40 PM7/31/08
to HALE TEAMS
Brian,
I already untangled the mess because we were planning on shipping you
Little Joe sans parachute and SPOT. But I'm more than happy to put it
back together and send it all to you.

Interested in helping with Big Joe?

Eric
> 30.4 ft/sec... it hit the ground ...
>
> read more »

Claude

unread,
Aug 1, 2008, 7:34:26 AM8/1/08
to HALE TEAMS
Our mission report with data evaluation at first glance. Some data
need further reflection, other ones require confirmation and surely
discussion.

http://www.convict.lu/htm/rob/hale3.htm

Interesting anyway.

Brian, could you send us your air pressure profile?

Anyone else could he or she send us the temperature profile?

Thanks Eric for having patiently gone through the complicated upload
procedure.

Claude

Davis, Brian L.

unread,
Aug 1, 2008, 9:22:02 AM8/1/08
to hale-...@googlegroups.com
> I already untangled the mess because we were planning on shipping
> you Little Joe sans parachute and SPOT.

Ouch. That's "ouch, that must have been a fair amount of work", not "ouch, I needed the tangled mess" :). Honestly, for the accuracy I'm likely to get, I don't think I need the "flight hardware". But it turns out I still have your parachute bag and the SPOT box sitting around here somewhere, so if you want me to ship those back to you I'll be sending yet another DHL payload your way, and I can return the SPOT & parachute in those.

> Interested in helping with Big Joe?

<GRIN> You even need to ask? If you dig through that very long post, you'll see that parachute deployment happened after 61 seconds (didn't quite reach the 80 second limit), implying Lil' Joe "thought" it was at 69,900', 83 minutes after it was turned on. That's good, in that the program worked exactly as intended. But what a ride it had down... apparent sustained accelerations above 3 G's for about 20 seconds (flat spin?), and when the parachute "deployed" there is absolutely nothing visible in the acceleration profile. At that altitude, you might as well wave a wet towel over your head as use a parachute.

--
Brian Davis

Eric

unread,
Aug 1, 2008, 1:34:11 PM8/1/08
to HALE TEAMS
Brian,
We'll send the stuff out with Little Joe and you can send it all back
later. No rush, we won't do another launch for a few weeks and even
then we probably don't need the SPOT and parachute for that (we have
more).

Eric
Reply all
Reply to author
Forward
0 new messages