There has been a lot of discussion recently about black powder not functioning at altitude and a lot of speculation as to why it behaves in that way. Bill's monograph explains the mechanism behind the failure.  What concerns most amateurs and experimenters is how generate enough gas to blow a nosecone or separate an airframe at altitude so that a recovery system can reliably deploy.  See more after the paper below.

William Colburn

In 1962 I tested initiators for Thiokol at our lab in Sunnyvale, CA. In firing these initiators at altitude in a vacuum chamber, the first unit expelled the 2A pellets (B/KNO3/Polyester Resin) without igniting them. This was a serious problem indicating that the second stage Pershing would not ignite! The Thiokol engineer, Sam Zeman, happened to be on site. Confronted with the problem he offered that the test conditions were unfair and that the initiator should be fired into a closed volume, as it would be on the missile. We added a plastic cap to the threaded initiator output and they al fired successfully. Intrigued by the problem, I fired similar units with black powder output charges. They scattered black powder into the chamber without ignition.

The phenomenon is generally referred to as a ìdP/dt Failureî. This refers to the fact that a sharp pressure drop occurs, either at ignition or even during burning. Why does that cause the material to quench?

One theory of solid propellant burning is the Summerfield Theory. In that theory several layers are proposed which exist in burning propellant. The first layer is the unburned propellant, next is the heated material, then the products of vaporization or sublimation, then the combustion of the vapors. The last layer is luminous (at least in the infrared) and is pumping heat back into the other lower layers. 

During an abrupt lowering of pressure, the vapor layer expands suddenly lifting the luminous combusting layer to a greater standoff from the lower layers. This stops heat transfer or minimizes it and the propellant ceases  to burn as vapor stops forming. 

This phenomenon was useful in early rocket motor research as grains could be ejected from the motor by failing the nozzle attachment. The ejected grain would quench due to the dP/dt change and the partially burned grain could be examined.

We built a detaching nozzle for UTC. On the first test I stood by at one of the blockhouse windows. About 2 seconds into the burn the nozzle blew, being fired by a timer, the grain ejected into a pile of wet hay. It was still violently combusting. This is to say that not all propellants will quench when undergoing a drop in pressure. This particular propellant was highly aluminized. It is possible that the increased radiance from the luminous zone was enough to keep up combustion. 

What to do about the problem when using igntion material which are sensitive to dP/dt? Confining the propellant until all burnt (in the case of an ignition charge) would be one solution to the problem. Slowing the pressure drop another. Using materials which are not sensitive to pressure drops would be the best solution. 

Unsubstantiated by any testing, I would recommend materials which have high metals content. Mag/Teflon would be my first choice. T-1, Copper Oxide/Titanium would be another. There is some evidence that indicates that ignitercord or thermalite is not sensitive to pressure drops. 

A good field for testing. Anyone out there willing to do the experiments?

4-20-02

Springs have been used in the past with good success.  The trick is developing a reliable method of deploying the spring at the proper time.  Mechanical systems can be tricky and unless tested to death are prone to failure.

Compressed gases are also somewhat popular.  Rapidly evacuating a vessel pressurized with CO2, Nitrogen, or Helium will expel the contents of an airframe.  The key is getting the gas out of the bottle in a short time period and often a Burst Diaphragm is necessary to accomplish this.

More to follow.