Marcus said:

>The only explanation I can give is that the banking is
>done to increase ballistic coefficient ...

No. Ballistic coefficient is not a function of bank angle. Ballistic coefficient can be modified by changing the angle-of-attack which changes the drag coefficient. As a first cut when designing a trajectory, we assume a constant ballistic coefficient and constant L/D ratio for the entire trajectory (a very rough approximation). A better approximation is then used where we assume Modified Newtonian aerodynamics (impact theory). Modified Newtonian aerodynamics works great for drag and lift for blunt vehicles at Mach numbers greater than 6. Modified Newtonian aerodynamics is also susceptable to closed form solutions which makes it ideal for real time simulation. Unfortunately Modified Newtonian begins to crap out below Mach 6 and is no longer believable below Mach 2. For high fidelity simulation, we use CFD (wind tunnel data is vastly better but really expensive). Modified Newtonian provides a ballistic coefficient that is a function of free stream density and velocity. However the L/D ratio under Modified Newtonian is not sensitive to free stream density and velocity (that's a limitation of the approximation). Modified Newtonian is pretty rough for calculating moment coefficient and trim angle-of-attack (well executed wind tunnel data is good for that).

Gary