Prevent body drag coeffs from being negative

[AlexWKinley]

fixes #205

I think this should be a sufficient change to address the issue of getting negative drag coefficients in certain orientations.

The new body test fails without this change because its z drag coefficient is negative, causing the body to fly off the infinity. With the fix the test passes normally. Once others are happy with this change, I'll fix the merge conflict.

[RyanDavies19]

@AlexWKinley I have talked with Matt about this and I think there is a better approach. The translational and rotational drag coefficients should each be converted to a 3x3 matrix, and then multiplied by the orientation matrix and it's transform. This should yield a 3x3 drag matrix that then is multiplied by the velocity to give our drag force. We would do this for both the translational and rotational drag. coefficients, giving us an end result of a 6 DOF drag force vector:

$$F \\\ = 0.5 * \rho_w * \\\ OrMat\\\ \begin{bmatrix} CdA_x & 0 & 0 \\\ 0 & CdA_y & 0 \\\ 0 & 0 & CdA_z \\\ \end{bmatrix} OrMat^T *\\\ (\vec v_i * \\\ |\vec v_i |)$$

I'll put together a PR for this and have it up soon.