Hi Mr. Michael Wood,
I am not sure if the exercise, is for aesthetic or engineering purposes?
If the former is your concern, developing Mr. Hackney concept might be the solution.
If the latter is your goal, we start talking about modelling a catenary surface .
In such case the eyeball approach would be probably unacceptable, and some number crunching would be necessary.
There are software on the market which would do the job easily, like ones based on Finite Element Methods or Boundary Element Methods (more appropriate ones).
Also, as the case is academically simple, there might be analytical solutions for it (note two symmetries). If you find the equations get back to me, I will try to build the F360 model for you (free of charge).
Finally, I am not an expert on F360 generative design but possibly you could also try this heavy handed approach deriving into some outcome ( or hundreds of them).
Regards
MichaelT
MichaelT