| Data Driven Design Optimization Methodology (DDDOM), as an application of Dynamic Data Driven Application System (DDDAS) in the engineering design field, was proposed by D. Knight et al, [18]. DDDOM combines experiments, simulations and design optimization into one automated loop and tends to achieve better designs in less time with less effort than traditional methods. The present research focused on the development of the DDDOM software system and its application. Six elements of the DDDOM system were developed: User Interface, DDDOM Controller, Experiment Design, Simulation, Surrogate Models and Optimization. The software development involved system programming, code wrappers and a graphical user interface. Perl and Perl/Tk are the principal programming languages used. Part of the code was also written in C. The DDDOM software system provides an integrated design development environment for users in the engineering field. The application of the DDDOM was the design of a cooling system for electronic components. Thermal management in cooling of electronic equipments is the main technical problem in obtaining faster, smaller and more reliable electronic systems. In our study, two dimensional mixed convection heat transfer over two isothermal protruding heating elements, simulating electronic components, located at the bottom surface of a horizontal channel was considered. Three cases were investigated within the DDDOM software system: (1) single objective optimization problem, minimizing pressure drop at Re = 600; (2) multi-objective optimization problem, minimizing pressure drop and maximizing heat transfer (Stanton number) at Re = 900; (3) multi-objective optimization problem, minimizing pressure drop and maximizing heat transfer (Stanton number) at Re = 2100. In all these cases, the design variables are the locations of two heat sources. The results showed that the DDDOM yielded a better design in a more efficient way than the traditional sequential design method. |
