LayOpt

Although the problem of finding the optimum structural layout of discrete members to equilibrate external loads, was studied over a century ago by Maxwell and Michell, as far as practicing structural engineers are concerned this is still an area of  academic rather than practical interest. Indeed in practice structural engineers still rely on their intuition to optimize their designs. Furthermore, interest in this academic research area appears to have faded in the period since the 1970s (when workers such as Chan, Cox, Hemp, Prager, Parkes, and others were active). Since the late 1980s (especially following the seminal paper of Bendsoe and Kukuchi), FE based continuum optimization approaches have became predominant in the field of structural layout / topology optimization research.

In late 1950s, Hemp derived solutions for several problems `to encourage research in this very important field', but this was clearly not sufficient. At the time closed form analytical solutions for specific problems were difficult to obtain, and limitations in computer power made it only possible to solve trivial sized problems using alternative computational approaches. Nowadays the situation has changed as immense computational power is readily and cheaply available, as are highly  efficient LP solvers. Therefore the reasons for ignoring this field need to be re-evaluated. Furthermore, the recent `adaptive member adding' technique proposed by Gilbert and Tyas makes it feasible to solve real-world scale problems on a standard PC, potentially giving structural optimization a second life.

In order to raise awareness of `adaptive member adding' technique, and encourage other researchers to use it, we followed the 99 MATLAB script by Sigmund which made SIMP method widely used, and implemented the `adaptive member adding' procedure in a very clear, concise MATLAB code, with an interactive free online tool.

For more details, and to try the free online layopt tool please visit http://www.layopt.com/

Comments