This article is from the Model Rockets FAQ, by Wolfram von Kiparski with numerous contributions by others.
From: Leviathan@nighthawk.medtechnet.com (Leviathan)
So would everyone else... but there's always that chance that something
may fail when staging and/or clustering... and probably double the chance
when staging & clustering. Therefore, my best advice to anyone attempting
(large?) projects requiring staging/clustering is to invest in some sort
of recovery system such as the Adept altimeters with deployment. In the
case of staging... if the upper motor fails to ignite the altimeter will
still deploy the upper stage chute(s) SAFELY with NO damage to the rocket.
In a case were you're clustering and a motor (or 2, or 3....) fails to
ignite in the cluster, and the rocket WILL fail to reach a safe altitude -
or more precisely now WILL have a delay time that is TOO LONG - again the
altimeter will SAFELY deploy the chute(s). IMO it's a small price to pay to
protect a much larger investment of the rocket itself. As a matter of fact
in my current project - a 1/4 scale 3 stage Argo D4/Javelin - each stage
will carry it's own altimeter with the 3rd stage carrying the Adept OBC2
recording altimeter. Not only will this provide for dual deployment of
each stage's recovery system, it should provide protection against failure
of the 2nd and/or 3rd stage ignition. I also plan on carrying a Pratt
system aboard to provide redundant back up.
From: John Dunbar (firstname.lastname@example.org)
I really recommend the teflon sheathing method of thermalite.
Now you can try using a flash pan to get that first stage bohemian going.
Its just a nice circular pan, with fine black powder in it. Just order a
can from your local gun shop, they can ship US postal to your front door
without a single eyelash batting on the face of a BATF agent. You stick
those engine down into the power with a thermalite whick protruding from
the engine, and WHOOSH...
The more parts you have to worry about, the greater the likelihood that
something is going go terribly wrong. Now if you don't care, and just
want to do it for fun, GO FOR IT, otherwise think of ONE BIG MOTOR for the
first stage and one smaller, yet BIG MOTOR, for the second. Do not rely
on mercury switches for high power ... thats a NO NO. Instead, use timers
in a way that causes the second stage to start its ignition while the
first is still under power. Now you can drag separate, and that's fine to,
but make sure your bird is flying straight and true, or it will be doodoo!
Technical reports on this subject are available from NARTS and California
Rocketry (see part02 of this FAQ):
California Rocketry report AIR-3
North Coast report NCTRA1 (from NARTS)