Table 1 - uploaded by Leifur Leifsson
Content may be subject to copyright.
Source publication
Simulation-based optimization has become an important design tool in microwave engineering. Yet, employing electromagnetic (EM) solvers in the design process is a challenging task, primarily due to a high-computational cost of an accurate EM simulation. This paper is focused on efficient EM-driven design optimization techniques that utilize physica...
Context in source publication
Context 1
... optimized surrogate model design in Fig. 4(a) is not as good as that in Fig. 4(b) because the constant output SM correction terms causes a significant distortion of the model response while moving away from x (0) . Table 1 shows the results of optimizing the bandstop filter in Fig. 1 using the surrogate model based on output SM and ARC. The latter technique is able to yield a better result using only three fine model evaluations, whereas the optimization cost with output SM is eight fine model evaluations. ...
Citations
... Typically, these models are set up in the entire design space [15]. More efficient methods involve surrogate-based optimization with physical surrogates, e.g., suitably corrected lower-fidelity simulation models [16]. Utilization of physics-based surrogates is not widespread in the context of CFD modeling of compact heat exchangers of complex geometries. ...
Two-dimensional numerical investigations of the fluid flow and heat transfer have been carried out for the laminar flow of the louvered fin-plate heat exchanger, designed to work as an air-source heat pump evaporator. The transferred heat and the pressure drop predicted by simulation have been compared with the corresponding experimental data taken from the literature. Two dimensional analyses of the louvered fins with varying geometry have been conducted. Simulations have been performed for different geometries with varying louver pitch, louver angle and different louver blade number. Constant inlet air temperature and varying velocity ranging from 2 to 8 m/s was assumed in the numerical experiments. The air-side performance is evaluated by calculating the temperature and the pressure drop ratio. Efficiency curves are obtained that can be used to select optimum louver geometry for the selected inlet parameters. A total of 363 different cases of various fin geometry for 7 different air velocities were investigated. The maximum heat transfer improvement interpreted in terms of the maximum efficiency has been obtained for the louver angle of 16 ° and the louver pitch of 1.35 mm. The presented results indicate that varying louver geometry might be a convenient way of enhancing performance of heat exchangers.
... Metamodels are used in many fields and a large amount of works was found related to this method. Applications for structural mechanics [17] ,Computed Fluid dynamics [18], electromagnetics [19], [20], or forming process [21], [22] can be mentioned. Furthermore, metamodels are used to faster fulfil FE models objectives faster, as model approximation, design space exploration, sensitivity analysis [23] and optimisation [7]. ...
... Metamodels are used in many fields and a large amount of works was found related to this method. Applications for structural mechanics [17] ,Computed Fluid dynamics [18], electromagnetics [19], [20], or forming process [21], [22] can be mentioned. Furthermore, metamodels are used to faster fulfil FE models objectives faster, as model approximation, design space exploration, sensitivity analysis [23] and optimisation [7]. ...
Recent evolutions of computer-aided product development and massive integration of numerical simulations to the design process require new methodologies to manage the continuously increasing flow of data and decrease the computational costs of numerical design of experiments. This paper presents a literature review of Simulation Data Management strategy and adaptive design of experiments methodology to detect possible links between these two fields and identify potential improvements for simulation process shortening. Adaptive design of experiments is based on several methods implying a profusion of different technics. Re-using best practices may help designers to choose relevant methods to reduce computational cost and simulation process duration
