Model Forum / General / Rockets / November 2006

  Be realistic, Bob.  Do you REALLY think that they're going to put
that kind of communications in WRITING???  :-)

  Wouldn't help much, I'd think.  The flame-front would propagate
upwards AND downwards, so the rest of the propellant would light soon
after ignition.

  If they *really* wanted to accurately measure the burn-rate, they
would have used a strand-burner chamber... not this 'Keystone-Cops'
setup they outlined.  :-)

Um, but that doesn't make it an end-burner since the flame can travel up the
grain slot.

Oh, and if he is not familiar enough with rockets to figure out the burn
rate, then he is working outside his area of expertise, an action that
is also highly frowned upon by the licensing boards.

GC

No.  The government's own standards are quite different.  For instance,
the FAA requires structural material properties for aircraft analysis in
which 98% of samples exceed the values with 95% confidence.  This is
documented in 14 CFR 25.613.  The FAA's guidelines for developing these
properties typically involve at least hundreds, and sometimes thousands,
of samples from multiple manufacturers and multiple heat treat lots.  If
you tried to certify an aircraft with properties based on 3 samples, the
FAA would laugh you out of the room.  We know for a fact that there is
significant scatter in the burn rates and impulse of commercial APCP
motors, and this would only serve to increase the number of samples
required.

Eric Albrecht
Aircraft Landing Gear Structural Analysis

Sure, you can compute an unbiased standard deviation from just two
samples. Understand that the average and standard deviation of a set of
measurements are the average and standard deviation of just those
numbers. They are estimates of the mean and variance of the underlying
population from where those numbers came.

Once you have a standard deviation you can discuss whether a particular
value (e.g. the manufacturer's published burn duration as in the ATF
doc) falls within some number of standard deviations of the average.
Three standard deviations gives a 99.7% "confidence interval". That's an
unfortunate choice of words because it implies more confidence than it
strictly means. Strictly speaking, "99.7% confidence range" means
there's a 99.7% chance that one more measurement (taken using an
identical testing protocol) will be within 3 standard deviations of the
average. It says nothing about how well the average and standard
deviation of those three measurements relate to the mean and variance of
the underlying population.

... which is what you're asking, Bob: how well does the average & std
dev of just three numbers estimate the mean and variance of the
population. Student's t-test for the null hypothesis is a common way to
answer this question. I haven't read all the ATF papers, but I suspect
this sort of thing wasn't done. :-)

However, taking only three measurements is hardly an indictable offense.
The statistical analysis done in the ATF tests is not the issue.

What struck me was the inconsistent test procedure (unconfined/confined)
for blackpowder and smokeless propellant vs. APCP.  ATFE will probably
claim that the APCP was tested confined in order to determine its burn
rate 'when used as intended'.  However, this should have applied to the
IMR4189 as well since to my knowledge its only intended use is in
cartridges for firearms where it is totally confined upon ignition, and
where its burn rate is vastly greater than when lit in an open trough.
Confined 'as intended' in a cartridge, it would be up with blackpowder
at the top of the list on page 141 of the report.  But by employing this
apples-and-oranges procedure, APCP is neatly bracketed on the burn rate
continuum by blackpowder and smokeless propellant, both of which ATFE
often sees employed in homebrew explosive devices.

--

It is interesting that they included the Orange Book.  Here is an agency
of the U.S. Government submitting into evidence a document that is to
support their view of their agency's regulatory expertise...

...and the document is completely out-dated.  References to 5400.4 forms
used to sell explosives to in-state residents who don't have LEUPs...
If this is their evidence to support their position, they could have at
least submitted up-to-date documentation.

Or in a little more organized fashion here:
http://www.tripoli.org/documents/batfe/Oct2006Filing/2006Oct30-ATF.shtml

<snip>

Jeezuz.  As a science educator, I see so much wrong in the "lab"
test that I am just speechless.

A PhD and an Air Force officer signed a test result that I would
have failed were it a high school science project. This is not
stupidity, IMHO, its much too incompetent to be simple stupidity.
It seems to me that, somehow, BATFE has got a research lab to
become their patsy for producing bad data. If the data doesn't
hold up, BATFE won't take the heat, the lab will.

Its obvious, especially reading the Final Rule, that the BATFE is
intent upon regulating rockets and not APCP rocket motors, per
se.  When cornered, their rationalizations invariably revert to a
"large rockets which could deliver a destructive warhead" type
diatribe because they have no data or statistics to back up their
original claims. This lab result is just a pathetic attempt to
justify some supporting data, no matter what the cost.

What gets me is the lengths they seem willing, even anxious, to
go to in order to support their "claims". BATFE is not this
incompetent. A PhD at a research lab is not this incompetent. The
conclusion I draw from this "test" is that BATFE has been "told"
to regulate rockets, period.

We're "fighting" the wrong group, people. If the government is
willing to go this far to regulate APCP sport motors, we might
consider what may happen if the court rules against the BATFE on
this count. I think its bigger than a hobbyist squabble with a
regulatory agency.

If its not, then the Air Force needs to do some serious firing
and retiring at their labs. I have some spec ed high school
students which can do a better job and they are looking for work.

Well, I never would have believed it if I had not read the report.  I
find it incredible that the individuals producing this AFRL report did
not even have the basic fundamentals of how a solid rocket motor works.
Dividing the burn time into the propellant length instead of the web
thickness to get burn rates is an enormous error.   The scale of the
error is sufficient to state that the individuals producing the AFRL
report are technically incompetent with respect to solid rocket
propellant and solid rocket motors.  The fact that the ATFE did not
catch this indicates that all ATFE individuals reviewing the data are
also technically incompetent on solid rocket propellant and solid
rocket motors.  It is a farce to assume that technically incompetent
individuals could correct classify APCP solid rocket propellant as an
explosive or not an explosive.  After reading their report to the
court, I would seriously question whether the ATFE has the technical
competency to even correctly generate an Explosives List.  What is
striking to me, beyond the burn rate calculations, is that the ATFE
went to Tyndall AFB for the tests rather than Edwards AFB.  Edwards is
an expert in solid rocket propellant and motors.   The fact that the
ATFE went to the wrong facility in AFRL for the tests shows another
whole level of technical incompetency.   Not even having the basic
knowledge of who are the experts and who are not.   I think a GAO
investigation of the technical competency of the ATFE with respect to
explosives is called for when the new Congress convenes in January.

It is clear that NAR/TRA will fix the incorrect burn rate data for the
court, but I don't think that will be the end of the story or lead to
victory.   The ATFE essentially argues that any oxidizer/fuel mixture
is an explosive.  That needs to be addressed by showing a variety of
oxidizer/fuel mixtures are classified as flammable solids by DOT.
There are even AP/fuel mixtures classified as flammable solids.
Second, it needs to be acknowledged that there are APCPs that are
detonable and properly classified as explosives.   However, the APCP
used in hobby motors do not fall into this category.  The explanation
as to why needs to be made in layman's terms to the court.  I'm not
sure this has been adequately done.

John Wickman