Conclusions:

Cracks, resonance, tubular burning, delayed or non burning of some top and bottom surfaces, in itself, don’t seem to be able to explain the behavior o the BEM 3 test. So far only erosion, seems to make any chance. Erosion may however be triggered by any of the other possibilities when gas speed or mass rate is exceptionally increased.

The configuration of the BEM motor seems able to lead to significant erosion of the propellant segments. The fact that the BEM 1 test generated a relatively normal thrust diagram is most probably due to the fact that the burning rate was significantly lower than expected. But also in the BEM 1 test erosion probably occurred given the very long tail-off. 

Still very little is known about when and how erosion occurs in KNSB motors. Results from AP propellants don’t seem to fit well for KNSB. Much more reliable information on motors with progressively higher ratios of Ap/At are necessary to increase our knowledge.

The major point however is whether the SS2S motor will suffer from the same problem. In AP rocket motors is was found that extrapolation from small scale tests to large scale motors is difficult but in general erosion seem to decrease with size. In the following figure the SS2S is simulated under normal conditions with a throat diameter of 7.5 mm.

SS2S simulation with a nozzle throat of 7.5 mm

One can see that the gas speed is limited to about 160 m/s as compared to the BEM where roughly 190 m/s is reached. But also the pressure is low with a maximum of 45 bar as compared to almost 70 bar in a normal burning BEM. Hence these conditions are favorable (if gas speed is the dominating parameter) and erosion may be limited.

In the next simulation the throat diameter was made smaller (65 mm). Now the maximum pressure goes up to 75 bar. Gas speed is lower because with higher pressure the density of the gas is higher. The mass rate (not given in the figure) is however about 35% higher.

If erosion is rather a function of gas speed then a high initial pressure (strong diaphragm) and high burning pressure (small throat area) must be favored. In general when the pressure is very low gas speed close to 1 Mach may be obtained. This may be the case with a weak diaphragm. If on the other hand mass rate (expressed per unit cross area of the void) is the determinating parameter, the opposite will be true. But also the Reynolds number could be important as turbulence may play an important role.

SS2S simulation with a nozzle throat of 65 mm

The fact that the second static test of the NEAR SCAA9901, the largest KNSB motor known, performed very well is however a good sign.