PyTOpt: A Nonlinear Topology Optimisation Program in Python
Typ
Examensarbete för masterexamen
Program
Applied mechanics (MPAME), MSc
Publicerad
2021
Författare
Pettersson, Daniel
Säterskog, Erik
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Ever since the finite element method was introduced in the 1940’s, the behaviour of structures
could more easily be predicted. Combining this with an optimisation method made it possible
to minimise a structure’s weight while keeping the strength. Several topology optimisation
programs have been created for this specific purpose since then. In this thesis, a topology
optimisation program is implemented in Python with the method of moving asymptotes and
the optimality criteria method as optimisation algorithms. By utilising previous works from
universities, such as Lund University and the Technical University of Denmark, and combining
the best parts a new topology optimisation program was created. The program has new
functionalities not previously implemented in similar topology programs. I.e. functionalities
such as the ability to utilise a variety of material models, being able to use an arbitrary design
domain in 2D, body forces and asymmetrical yielding. This program also fits well in educational
purposes due to its versatility. The results shows that optimality criteria method gives a
better and faster result than the method of moving asymptotes. The results are also compared
with commercial topology optimisation in ANSYS. The both programs gives indistinguishable
results but ANSYS outperforms the thesis’s program in regards to computational time.
Beskrivning
Ämne/nyckelord
Topology , Optimisation , MMA , OC , FEM , Python , Body forces , Compliance , Density , SIMP