I am assuming that half a K2050 has about 0.7430 lbm ( 1/2 of 674 grams) mass of propellant. Using a total impulse from the graph in the previous post for the Aerospike test, I get 130 lbf-sec/.7430 lbm = 174.98 sec or a specific impulse of 174.98 seconds. This compares with my calculations for a full K2050 with a convergent/divergent nozzle of 209.51 seconds. So, although the aerospike is not as good as the regular K2050 with a standard convergent/divergent nozzle, it is functioning as an aerospike for the supersonic gas. Off the top of my head I am thinking that a nozzle working only as a throat is going to have a specific impulse about roughly 58 percent of a K2050 or about 121 seconds or a little over 1/2 the K2050 specific impulse. (I am thinking that the critical pressure ratio is 0.58, but that depends on what someone uses for the specific heat ratio. I am thinking intuitively just simply multiply the fully expanded nozzle performance by the critical pressure ratio to get a throat-only performance. This is intuitive and needs more justification. In any case as I recall just a throat alone has a coefficient of thrust of 1.0)