دانلود رایگان مقاله تاثیر مدلهای اغتشاش در دقت تحلیل CFD از حلقه حرارت مکانیزم رفت و برگشتی

Abstract

A bellows-type Reciprocating-Mechanism Driven Heat Loops (RMDHL) is a novel heat transfer device that could attain a high heat transfer rate through a reciprocating flow of the working fluid inside the heat transfer device. Although the device has been tested and validated experimentally, analytical or numerical study has not been undertaken to understand its working mechanism and provide guidance for the device design. In an effort to improve earlier numerical models of the RMDHL, different turbulence models for the RMDHL design have been studied and compared with prior experimental results to select the most suitable turbulence modeling techniques. The governing equations have been numerically solved using a CFD solver. For the three-dimensional fluid flow, several turbulence models have been studied for the RMDHL, including Standard, RNG, and Realizable k-ɛ Models, Standard and SST k-ω Models, Transition k - kL-ω Model and the Transition SST Model. The results of the simulations have been analyzed and ranked using numerical model calibration template. It was found that the standard k-ω Models provided the least accurate results while the RNG-k-ɛ Model provided the most accurate predictions. It is expected that the results will help improve the accuracy of the work on the RMDHL modeling.

6. Conclusion

In order to improve the accuracy of CFD model for the RMDHL, extensive numerical simulations with different turbulence models have been carried out to study fully developed turbulent, low velocity and high heat reciprocating flow. The flow statistics obtained from the numerical results of seven selected turbulent prediction models and available measurement data have been compared, and in general a good agreement between the numerical simulation and experiment is obtained. The accuracy of the numerical models against the experimental data was also quantitatively analyzed. To identify the most accurate turbulence prediction model, template of Moriasi et al. [42], adopted. The results indicate that for a close prediction of the average cold plate temperature, the standard k- ɛ Model models provides the best accuracy while the RNG k- ɛ Model provides more accurate information with respect to variations in temperature across the cold plate surface and will generally provide the best qualitative and quantitative basis for additional modeling consideration for future RMDHL study.