Assessment of RNG k-ε, SST K- ω and Reynolds Stress Models For Numerical Simulation of DLR Scramjet Engine

Abstract

In recent years, many researchers have used numerical simulation to investigate scramjet engine because, setting up of testing facilities for scramjet engine is very expensive. Completion of simulation work in less time is costeffective but, it is also important to achieve good outcome from the simulation work. In this context, the use of appropriate turbulent model for the simulation for DLR scramjet engine is a vital aspect. From the available literature, it is seen that many researchers have used Reynolds stress, SST kω and RNG k-ε models in the field of supersonic flow and combustion. Thus, in this study, the primary objective was to find the better turbulence model in between Reynolds stress, SST kω and RNG k-ε models for numerical simulation of DLR scramjet engine. Finally, from the study, it is found that using SST kω model for numerical simulation of DLR scramjet engine results in more accuracy at a lesser computational time.Impact Factor (JCC): 7.6197 SCOPUS Indexed Journal NAAS Rating: 3.11The simulation in this study is conducted on a workstation equipped with Intel Xenon processor of 8 GB memory.The computational time spend for cold flow simulation using RNG k-ε and SST k-ω models were approximately 16 hours and that for using Reynolds Stress model was 31 hours. Whereas, the computational time spends for reactive flow simulation using RNG k-ε and SST k-ω models were approximately 36.5 hours and that for using Reynolds Stress model was approximately 74 hours. CONCLUSIONSIn this study, the primary objective was to find the better turbulence model in between Reynolds stress, SST kω and RNG k-ε models for numerical simulation of DLR scramjet engine. Base on the results found from the simulations, the following conclusions are drawn.• The results predicted by SSTk-ω model is found to be more accurate than the results predicted by RNG k-ε and Reynolds Stress models in comparison with experimental results.• In reference to the results obtained by using SSTk-ω model, it is seen that the average deviation in mixing and combustion efficiency predicted by RNG k-ε model is less compared with Reynolds Stress model.• The computational time required for simulations using RNG k-ε and SST k-ω models are substantially less in comparison with Reynolds Stress model.• Thus, from this study, it can be finally establish that the use of SST kω model for numerical simulation of DLR scramjet engine will provide more accurate result at a lesser computational time.

A Preliminary Investigation of a Hydrogen Injection Wedge Strut for a Scramject Combustor Application

“…A modification of the k-omega model, known as Shear Stress Transport k-omega (SST k-𝜔) can combine the abilities of both the standard k-omega and k-epsilon models to predict turbulent effects within and far from viscous layers [20]. Additionally, a study conducted by Debnath et al [25] compared RNG k-epsilon and SST k-omega models applied to the DLR combustor geometry and initial conditions and concluded the SST k-omega model provided better accuracy with a lower computational time than the RNG k-epsilon model. The evaluation metrics between models of mixing efficiency, combustion efficiency, showed most alignment with the original DLR results, while contours of static temperature and Mach number were similar for both models.…”

“…A blending function multiplies and adds the k-w and k-e models which activates the respective model depending on the flow location, i.e., the k-w activates near a wall, inversely zeroing the ke model. This makes the SST k-w model more reliable than the standard model [25] for flows with shock-induced adverse pressure gradients and was derived by F.R. Menter for this purpose.…”