Model Forum / General / Rockets / March 2007

Hi, John.

Actually, temperature correction is not peculiar
to RDAS. Most altimeters require it. It has always
been an issue with the FAA, and is more so now
that some planes use GPS and others use altimeters.

(I didn't read your email carefully, I've been at work
forever. You gave me a reference. Was this it?)

1)http://www.faa.gov/regulations_policies/rulemaking/historical_documents/1999/med
ia/ac91-xx.pdf

I did not think it news to anyone that temperature

correction is important. I was surprised that, in the

samples I tested, the base temperature appeared to

be 0 degrees C, and not 15 degrees. The most

controversial part of the whole thing is my value

of 128 for the displacement. It seems to work

better than other values I tried, and it is a round

number in binary.

Let me summarize a couple of other points:

1)Now that I think about it, I cannot characterize

all RDAS altimeters in exactly this way. Three

of my samples are from 2003. These were taken

at ~34 degrees C. I have another data set, the

vintage of which I cannot vouch for that is taken

at -15 degrees C. That dataset is not in raw form,

but is in intrpreted form. That is, the altitudes are

already computed. The R^2 statistic between

untransformed inertial altitude and barometric

altitude goes up from .94 to .9996 with the

conversion. The same sort of thing happened

with the other data sets, which were taken at

very different temperatures.

Perhaps later versions of the instrument do things

differently.

2) Check out the formula I suppled in my reply

to Dave Schultz. Note that since pressure is a

ratio, it may be represented in only relative

units. Most units, including RDAS, allow one

to download raw data. One normally adds/subtracts

a displacement from that (in my samples, it

seemed to be 128) and takes a ratio of each

individual reading to a ground reading. (I averaged the

negative time values.) Substitute that into the formula

using your actual base altitude temperature and you get

a calibrated altitude value. You don't need the value that

RDAS gives you, which is likely a (very very good)

polynomial approximation

Do a linear regression on the nominal values and you

can estimate the base temperature used by RDAS. Indeed,

the intercept is an estimate of the reference altitude.

(Problem is the regressor variable, whichever you choose,

has noise, so there is some bias. The intercept tends to be

bigger than it should be.)

Some units claim to be temperature corrected. The data have

seen with RDAS are not consistent with temperature correction,

IMHO - or temperature correction means something other

than the issue detailed here.

See my next (short) post.

Regards,

-Larry

Hi, Dave.

Yup. I have actually measured this on a takeoff
and landing of a jet liner. Standard rate is 6.5 Kelvins
per kilometer. I measured 6.38 over Beijing and 5.78
over Newark. Turns out such lapse rate differences
make for trivial errors in the formula.

FWIW, in the lower stratosphere, the lapse rate is zero.
It eventually reverses higher up, as ozone absorbs light
energy.

Altimeter formulas are based on two principles:
the first law of thermodynamics and hydrostatic
equilibrium.

Hydrostatic equilibrium says that at any altitude,
the the pressure on a given horizontal area of air
is balanced by the weight of the column of air
above it. That weight is affected by the temperature
of the air as the temperature lapses. It works out to

Const HC = 0.03418155
Const LapseRate = 0.0065

BaseTemp = BaseTempC + ZeroC

Ratio = PressRead / MeanGroundPressure

Altitude = (BaseTemp / LapseRate) * (1 - (Ratio ^ (LapseRate/HC))

Note that the whole thing is proportional to base temperature.

Well, this stuff isn't exactly controversial. :-)

I'm a little behind (OK. Some would say I'm a big behind...)
on this thread as I have production issues at work and have
to go in today, Saturday. Will reply to John D. and fill in more there.
The issue is involved.

Good to hear from you.
-Larry Curcio

and summarize to