The technique that I employ is to calculate the tributary width for each beam on which the loading area spans and multiply the same with the applied pressure. Then I have on my hand a Uniformly Distributed Load applied on the members that support the load area. I am not sure why STAAD has set limitations on the applicability of area loads on closed panels alone. In the end, I am certain the program does the same thing that I do. determine the influence width of beams, multiply it with the applied pressure.
The corner co-ordinates of the rectangle specified essentially demarcates the boundary with in which the load acts. The geometry of loaded area is therefore very clear. Please note that I am applying neither on a inclined surface nor on a skewed geometry. I am essentially applying my loads on regular rectangles.
As I understand it, utilising the One Way load distribution is the more practical way of simulating the situation.
The manual on this very clear, I will read it again and again if need be. What I do not understand is why should the specified rectangle be within a closed panel.
If the process involved is determining the tributary width of the beams on which the area spans, multiplying it with the applied pressure and applying the associated UDL on the beams, what is the factor that necessitates the use of closed panels?
I am by no means stubborn and if there is a way for me to work around this issue, I will. I wanted to implement sass3k's suggestion. The only thing I am worried about is that it will seriously screw up my model if I introduce dummy beams now. (I have beams that are on top of each other and hence specified with definite member offsets). Still, it's better than calculating the member loads manually.
I will post a pic early tomorrow showing how I calculate the tributary width.