This is really outside the box, but if you could build some really LIGHT 18mm to 13mm adapters, you could put A10-0T and A3-4T in mid stage and sustainer respectively. Not sure how much mass the adapter would cost you, but might do two things if it shaves off a few grams
https://www.rocketreviews.com/compare-estes-a8-to-estes-a10t.html
1. Drops your lift off mass.
2. Because the mass dropped is in the tail, should move CG FORWARD. on the other hand that may be bad as it will make the rocket MORE stable and more likely to weathercock. Hmmmm.
Interesting idea!
Assuming a theoretical zero-mass 13mm to 18mm adaptor, here's what we get with a configuration of B6-0; A10T-0; A3T-4:
Launch speed: 13.8 m/s. Apogee: 729 ft. Deployment velocity: 6 m/s. Max V: 63 m/s. Total mass with motors: 94.4g. Stability: 5.31 cal -> 6.08 cal -> 5.37 cal (per stage).
If I spec two 13-to-18mm adaptors at 3g each, I still get 13.4 m/s launch and about 720 ft apogee.
This looks promising. It saves considerable mass (14-20g) over the next-best option, the B6-0; B6-0; A8-5, and has much better launch speed. Stability is high, but it's already above 4.0 cal for every possible engine choice. This kit seems to have been designed with very high stability. I think the designer must have been worried about change in trajectory when staging.
As a bonus, the apogee is lower than any other choice I have (excepting the awful A8-0; A8-0; A8-5), which is really a bonus for me, I don't like losing things into trees. The deployment velocity is also lower than any other motor option, meaning less shock on cord and chute.
One interesting thing (see the plots below) is that the A10T-0 is an oddball motor. According to the
thrust curve, the engine burns almost entirely in 0.2 seconds, with a sharp spike up to 12 N; and then it burns for another 0.6 seconds at a tiny 1 N of thrust! The result is an average of barely over 2 N for the entire 0.8 second burn time. In practice this causes the upward speed on this rocket to noticeably decrease during the middle stage.
If I can build myself some 13-to-18 adaptors this is a real option.
Here are the plots for each of the stages on the above configuration:
B6-0 stage (bottom):
A10T-0 stage (middle):
(Open rocket has issues trying to model the tumbling of this stage)
A34T-0 (sustainer):