All,
A while back I wanted to model a solar sail so I used the special
relativy equation and integrated over a distance to get the
approximate velocity vs time. Basicly, I did a force ballance to
calculated the accelleration and then integrated the accelleration to
get the velocity and the distance (two integrations). I think this
is how I did it.

p = gamma*m*u

dp/dt = d gamma/dt * (m*U) + d(m*u)/dt * gamma
dpt/dt = d gamma
d gamma/dt = (u/c^2)*gamma^3* du/dt
after some manipulation (I love when they do this in books )
dp/dt = ma*[gamma^3 * (u/c^2) + gamma]

I obtained a final equation for the accelleration:

a = (dEphotons/dt)/ m*c[gamma^3 * (u/c)^2 + gamma]

m = mass
u = accelleration
c = speed of light
E = energy of the light
p = momentum
(Note: all of these values are in the stationary reference frame)

I integrated (numerically) this eqation with respect to time to get
the velocity and again to get the distance.

Now, I am an engineer and not a physicist so I used Euclidean space to
integrate over - I'm not concerned about small changes in cordinates
due to small variations in the the coordinated system due to gravity.

Does my rational have any inherent flaws? The results seemed
reasonable.

Eric