Finally got back to this.

This PR changes the tube fin stability calculations to use calculated stability of tube fins instead of the "three fin hack" previously in use. It completes the rewrite of tubeFinCalc.java.

The stability calculated with this code is close, but not identical, to that using the code from 15.03. In the case of the tube fin example rocket, the CP location in centimeters from the nose of the rocket is:

	15.03	new
Mach 0.1	54.1	53.4
Mach 0.9	54.5	52.3

My expectation is that the new code is more accurate, but we won't really know until we can find a wind tunnel.

A bit of background: originally, neither RockSim nor OpenRocket was able to simulate tube fins. Bruce Levinson developed a technique for simulating tube fins using flat fins: he used three times as many rectangular fins as there were tubes, with a total area equal to the tube fin area (frustratingly, I am unable find his original article. He mentions it in an article he wrote in Apogee's Peak of Flight newsletter #27, "Simulation of ring fin designs using RockSim 4.0", but the link in that article is dead and google isn't helping me). This is known as the "three fin hack".

OpenRocket, as of version 15.03 (and no doubt earlier, but I didn't go looking) had the three fin hack embedded in the code. The tube fin calculation took pretty much all of the code from FinSetCalc.java, and set the geometry according to the three fin hack. This PR removes all that and replaces it with calculations from a paper cited in the code.

Fixes #1258