Solve the form ahead of time.
It is easier to solve general forms when one uses rational values instead of floating point such as -47.3 . But if you substitute 473/100 and solve then the reasons for various constants that show up in the solution become obscured. So for the moment substitute a variable, V, for the -47.3, declare all the names involved as syms, and solve. simplify() and expand() and simplify() again. What you can get out will look like,
fb_ = 1000*10^(-(1/40) * V + (1/40) * deltaTL + 1/8) * (m1+m2)^(1/2) / (m2^(1/2) * d^(1/2) * m1^(1/2))
Once you have that formula you can put it into the source code, eliminating the solve() call.
