We were planning on using a BKNO3 igniter. We did not have an ATF Explosives User's Permit and were developing pellets using polyester resin as binder per Mil Spec 46994B. With a professional BKNO3 igniter as plugger and others have said the goal is "instant-on" - the igniter is to bring the motor to above the critical pressure (50-150 psi) for appx. 100 ms and apply a certain heat flux. This pressure goal can be aimed for without worrying about the dynamics of sealing the nozzle to burst at a certain pressure.
Since such an igniter is quite violent, we intended to do a static fire test of the M1378 with the robust BKNO3 igniter the week prior to our launch, but another team member mistakenly sent the liner for the static fire to New Mexico, along with the flight hardware. Thus we had no way to static fire the igniter, and due to fears of over pressurizing the motor, we instead flew with a thin aluminum burst disk (a few layers of aluminum tape) and sealed/bonded it to the nozzle exit diameter Silicone glue. The igniter used was a Quickburst Fat Boy, a long burning igniter that isn't able to pressurize the motor rapidly enough. The igniter fired but the motor chuffed out. I suspect the ambient pressure in the motor core had dropped to that of the atmospheric pressure at 35 kft, meaning the motor wasn't perfectly sealed.
For future flights, both for Princeton Rocketry Club, and Operation Space, Inc., an organization that I serve as design lead for, we plan on obtaining BKNO3 pellets with a burn time of approximately 50-100 ms, and sizing the igniter charge per the Bryan-Lawrence relation, which is another method (like the mass flux plugger mentioned) used in industry.
https://www.dtic.mil/dtic/tr/fulltext/u2/307914.pdf
FYI the quotes we are getting for the BKNO3 are over $5000 at this time for a MOQ, not including HAZMAT freight from GA.
The charge required seems quite large: I calculated 10g of BKNO3 from the equation, with a range of 5g-20g. This seems like a lot, but for reference the 2.75" FFAR rocket uses ~8g of BKNO3 pellets, with a total impulse of ~5000 Ns. Using this as a guideline, then one develops a testing apparatus of a steel pipe closed at both ends, with one end having a hole equal to the nozzle exit diameter. A pressure transducer is mounted to the apparatus to ensure that the pressure-time curve is appropriate, applying a pressure of ~50-100 psi for at least ~50 ms. The charge size can be modified at this time iteratively. We do not plan on sealing the nozzle since it was found to be unreliable last time.
Our plan for the igniter itself for Operation Space, Inc. is using the threaded bolt that goes into the forward closure of a Pro98 6GXL case, drilling a hole through the forward closure, and installing a much longer threaded bolt to serve as a solid mounting point for the threaded igniter assembly. Obviously the forward closure is sealed with epoxy. The igniter assembly is a 3D printed basket that is slightly smaller than the core diameter of the N5800 sustainer, and the igniter assembly holds the BKNO3 pellets. The 3D printed igniter basket is threaded and fits onto the rod, and sits inside the top grain core. 2 electric matches go inside the igniter basket for redundancy. For the N5800 motor I calculated a charge mass of ~30g.
For a bit more information about solid rocket motor ignition, here is a paper from the development of the Black Brant sounding rocket. They experienced upper stage ignition failures, and redesigned their igniter to be more reliable. They found that for their APCP propellant ignition was probable at 50 psi for a heat flux of 627 W/cm^2 for at least 40 msec. They also note that the core mass flux guideline was wildly inaccurate.
https://drive.google.com/file/d/1u8cSIIOlEvsLKY0CcCBbqWKnNdOCtdOB/view?usp=sharing
Since their final igniter charge mass was similar to that predicted by the Bryan-Lawrence equation, I suspect that we will be okay without measuring the heat flux, and that an igniter charge mass in the Bryan Lawrence range will produce sufficient heat flux. I do think that measuring the pressure output in a vented chamber will be a useful experiment. Upon verification of the pressure-time trace, we will static fire the rocket motor to make sure it doesn't over pressurize and CATO.
For more information about the theory of solid rocket motor ignition and igniters see the following literature:
1.
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19710020870.pdf
Be sure to read the igniters section, and the pelleted pyrotechnic igniters section. Pyrogen (propellant) igniters are typically for larger solid rocket motors like the Shuttle SRBs, from my understanding, and often have a pelleted pyrotechnic in the ignition train. Ignore the initiators section, as it is deceiving - this is only the initial spark that starts the ignition charge. For relatively small solid rocket motors like these we are using an ematch (squib) as an initiator. Read the ignition theory section, specifically about the critical pressure. Below the critical pressure (often equivalent to about 30,000 ft) ignition cannot be achieved regardless of heat flux applied. Since copper thermite igniters do not produce substantial gas, no matter how "hot" they are they will not light a sustainer at high altitude.
2.
"Igniter Material Considerations and Applications:"
A very useful, short, and simple paper detailing different igniters for solid rocket motors, from copper thermite to BKNO3, and how early igniters were a flash can of black powder, and how and why igniters moved to the precise wire cage BKNO3 pellets we see today (mostly to minimize shock from powder).
https://drive.google.com/file/d/1uzv4NAQa-V_2T_K86CjnDuqpjW5LL6KK/view?usp=sharing