Optimization of steel frames using two-phase approach
Mercader Ardevol, Anna (2019)
Mercader Ardevol, Anna
2019
Rakennustekniikka
Rakennetun ympäristön tiedekunta - Faculty of Built Environment
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2019-05-03
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201905061518
https://urn.fi/URN:NBN:fi:tty-201905061518
Tiivistelmä
The goal of this thesis was to develop a program able to solve steel frame optimization problems using a two-phase approach. This optimization problem consists of minimizing the weight of a steel frame ensuring that the solution found satisfies all the strength and stability criteria establish in Eurocode 3.
The two-phase approach consists of first finding a continuous solution and after it, using this first solution to select the closest standardized profiles and find among them the optimum discrete solution. This optimization method was implemented using Python and checked with Ftool that all the structural calculations were correct.
Two different formulations had been written following the same structure but with a big difference, the first one has only one design variable (h) and the second has 5 (h,b,t_f,t_w,r). As the first formulation has only one design variable, the correlation between h and all the other design variables needed to be found. To be able to approximate all the other variables it is necessary to know which profile family is each member of the frame, for this reason in the first formulation all columns are HEA profiles and beams IPE. However, in the second problem, all members can be from any of the I families.
The two-phase approach consists of first finding a continuous solution and after it, using this first solution to select the closest standardized profiles and find among them the optimum discrete solution. This optimization method was implemented using Python and checked with Ftool that all the structural calculations were correct.
Two different formulations had been written following the same structure but with a big difference, the first one has only one design variable (h) and the second has 5 (h,b,t_f,t_w,r). As the first formulation has only one design variable, the correlation between h and all the other design variables needed to be found. To be able to approximate all the other variables it is necessary to know which profile family is each member of the frame, for this reason in the first formulation all columns are HEA profiles and beams IPE. However, in the second problem, all members can be from any of the I families.