Optimal Design Of The Heat Transfer Unit Fully Based On The CFD Model

Since energy consumption gradually increases with the development of society,it becomes increasingly important to improve the energy conversion efficiency and reduce the energy transmission loss.As the basis of heat transfer equipment,heat exchangers have been widely applied in the industry,where kinds of heat transfer units are basic components.Therefore,the performance enhancement of heat transfer unit can effectively improve the performance of the heat exchangers,which will further increase the heat transfer efficiency and have a significant effect on economic development and social progress.In the optimization of heat transfer units,analytical model,surrogate model,and CFD model are three common performance calculation models.However,simplifications in the derivation process of the analytical model will bring accuracy problems,while the sample training and performance prediction of the surrogate model will also cause errors.On the other hand,the optimization fully based on the CFD model requires amounts of computing resources.Nevertheless,it gradually becomes possible since the computation performance keeps increasing in the recent years.Therefore,this dissertation further improves the optimization approach coupling CFD software and optimization algorithms,and subsequently optimizes four typical heat transfer units based on different key performance indicators(generalized thermal resistance R_h,power consumption P_f,etc.).In this dissertation,the shape of a tube under laminar flow is optimized by the simplified conjugate gradient method and the single-objective genetic algorithm.In the process of optimization,seven polar radii are selected as the design variables,and then optimized to achieve the minimum of the objective function composed by the generalized thermal resistance and the power consumption.Results show that the optimal drop-shaped tube has a good comprehensive performance,and performance improvement is always better than that of the circular tube from 8.12%to 24.10%even in the flow fluctuation of±25%.Furthermore,optimal tubes under different inlet flow rates and different weighting coefficients are determined to meet different requirements.Subsequently,since the weighting coefficient needs to be adjusted to meet different requirements and kinds of performance evaluation criteria exists in the literature,this dissertation applies the multi-objective optimization algorithm to optimize the following heat transfer units,including tube bundle,porous ring,and minichannel heat sink.This approach is more general because researchers could not only select the final solution according to the corresponding performance evaluation criterion,but also select the best compromise solution according to the decision maker technique.In this dissertation,we applied a decision maker technique,TOPSIS,to select the best compromise one from Pareto solutions,and then the decision effect is also compared with other classic performance evaluation criteria.Results show that TOPSIS can effectively balance two conflicting objectives function,while the solution selected by PEC will be the same with the one selected by maximum Nu when optimizing the porous ring.Moreover,after the optimization and selection,the overall performance of different heat transfer units has been significantly improved.For the shape optimization of a single tube,we have found that although R_h is 17.13%higher than that of the circular tube,but the P_f is decreased by 61.93%.For the arrangement optimization of tube bundles,it is found that the JF factor of the best compromise solution determined by TOPSIS is about 1.04.In the optimization of a porous ring inserted in the tube,PEC of the best compromise solution determined by TOPSIS is approximately 1.46(without the constraint)and 1.18(with the constraint of f<50).For the minichannel heatsink,the optimal solution by the geometrical optimization could reduce 35.82%in thermal resistance and 52.55%in pumping power,respectively.On the other hand,the optimal solution by the cross-sectional shape optimization could reduce 7.13%in thermal resistance,while pumping power increases slightly by 2.42%.Finally,the core flow principle is validated in the process of optimizing the porous ring.For a series of non-dominated solution,we found that the position of the porous ring is always in the core flow area of the tube.This result indicates that inserting the porous ring in the core flow area will have a better performance,while those solutions deviated from the core flow area have been eliminated by the evolutionary algorithm.The above results are just consistent with principles of enhanced heat transfer in the core flow.