You have a couple of dimensional problems here.
1. The argument to the ln function must be dimensionless. You have it as an inverse length.
2. The argument to the LambertW function must be dimensionless. Even if you make the ln argument dimensionless the LambertW argument is currently inverse length squared.
If you correct these then you can evaluate the LambertW function by using the symbolic ->float,20. i.e. you could put Dp -> float,20 after the specification of Dp.
(Doesn't have to be 20, but don't be tempted to use too small a number with float).
Alan.
