Al
NFPA 1125 specifies the allowable tolerances for consumer rocket motor performance, and these are the parameters that NAR and Tripoi measure and certify in their motor testing.
Model Rocket Motors
Total Impulse: absolute maximum variation +/- 20% of certified total impulse value; lot maximum standard deviation +/- 6.67%
Average Thrust: absolute maximum variation +/- 1 N or +/- 20% whichever is greater
Delay Time: maximum variation +/- 1 second or +/-20 % whichever is greater but not to exceed +/- 3 seconds
High Power Rocket Motors
Total Impulse: absolute maximum variation +/- 10% of certified total impulse value; lot maximum standard deviation +/- 6.67%
Average Thrust: absolute maximum variation +/- 10 N or +/- 20% whichever is greater
Delay Time: maximum variation +/- 1 second or +/-20 % whichever is greater but not to exceed +/- 3 seconds
The motor burn time and the ejection delay time variation specifications are largely decoupled.
In practice for most high power motors, the allowable variations in total impulse and average thrust would result in a nominal burn time within +/- 1 second of the certified burn time under the worst conditions. The minimum allowed ejection delay variation is +/- 1 second for delays of 5 seconds or shorter, and +/- 20% for delays between 5 and 15 seconds, and +/3 seconds for delays longer than 15 seconds (which don't currently exist).
In practical terms, when you consider the overall boost and coast vertical flight to apogee, I think the allowable motor variations yield a difference of only +/- 50 ft to +/-100 ft altitude if you normalize the flight to the ground level density altitude. While this is probably not a conventional normalization, it is very relevant for aerodynamic deployment stress which is proportional to air density and and to velocity squared.
Near apogee, gravity determines the velocity of the rocket. The variation in velocity is 32 ft/second/second near apogee, so if you're ejection charge fires within 3 seconds of apogee, the rocket is moving at less than 100 ft/second. The air density determines the maximum deployment shock load. A good HPR should have a recovery system that is rugged enough to withstand this the forces of deployment at this velocity, and most will, so if the delay burn is within +/- 3 seconds of apogee, your rocket will recover without separating.
The stock delay choices selected by Aerotech for each motor maintain these margins. You can and should buy the correct ones and not modify the ones you have if they have the wrong delay. It's not that is unsafe, it's just against the rules.
A comment for Fred.
The burn rate of a propellant is a function of the chamber pressure. The fuel grain(s) geometry and the nozzle throat area determine the chamber pressure. The delay column is also propellant, but its burning surface area, drilled or drilled, is so small compared to the fuel grain(s) that its recession rate is independent of its surface area and is determined only by the chamber pressure temporal profile. As I stated earlier, the end that you drill from won't make a significant difference in the delay time, but if you drill from the propellant side, the greater burning area should give a denser tracking smoke.
I believe that the forward ejection of burning delay column particles is primarily an issue with short delays when residual fuel grain combustion is still providing enough mass to pressurize the motor casing. If the pressure in the casing is still above ambient, buring delay charge could be spitting out on the shock cord for several seconds after the BP has gone off.
Bob Krech