Mitigating Forced Shock-Wave Oscillation with Two-Dimensional Wavy Surface
Round 1
Reviewer 1 Report
The present manuscript shows CFD results of the control of shock wave oscillations in a supersonic diffuser that has a forced pressure perturbation at its exit. The control method was to apply a dimple onto the upper and lower wall surfaces of the diffuser. The present study was well organized, with good written English, but needs to fully reconsider the explanations and discussions on the results obtained from the CFD, based on the comments given in:
1) Time step applied should be very carefully verified to check if it is reasonably appropriate to capture major flow behaviors of unsteady shock boundary layer interaction in the supersonic diffuser. Fig.4 (b) shows a remarkable discrepancy between CFD and Exp data, which must be improved with more careful implementations of CFD work. Moreover, the authors should check whether the boundary conditions in the present study are physically reasonable.
2) The authors should not be using some unclear terms in manuscript : secondary shock, secondary flow, outflow pressure, etc. These may cause readers to be much confused with general fluid dynamics terminologies.
3) The time dependent flow field given in Fig.6 and 12 should be more carefully validated. In particular, the local flow separation far downstream of shock waves should be in detail explained to show if it is due to real flow physics.
4) The generation mechanism of the second and third shock waves has been well known as the post-shock expansion. The authors should not infer or guess this kind of flow feature, simply based on the flow confinement effects, but they should give detailed quantitative data from the present CFD.
5) The flow distortion parameter should not be used as the aerodynamics performance parameter and the total pressure recovery parameter may also be used as the total pressure loss or the ratio of total pressures.
6) How could the authors calculate the shock excursion speed v in Eq.(5)? Did they obtain this value from the CFD results? Then the authors should check if it is correct from the time step and two-equation turbulence model(time average tech). Also it is quite interesting that the authors explain the alleviation of shock oscillations based on the thermodynamics law. The reviewer has studied hard to understand what the authors try to explain using the thermodynamics law but it was not very easy to understand that. This may be the same to a lot of potential readers. Thus, the authors should provide more straightforward and comprehensive explanations on the mitigation effects. In general, shock location, regardless of time-mean or time-dependent flow, is determined by the pressure ratio upstream and downstream, local flow state including the boundary layer characteristics and local flow passage geometry. Based on these parameters, shock wave oscillations should be analyzed in terms of their stability, but not in the work balance explained by the authors.
7) With regard to 6) above, the dimple applied in the present study is too massive and big for the purpose of controlling the shock boundary layer interaction in supersonic diffuser. It significantly modifies the entire geometry of flow passage and is not a good option to practically implement this kind of control strategy in engineering applications. Thus, the shock oscillations with completely different flow fields are being compared as the control effectiveness.
Comments for author File: 
Comments.docx
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The manuscript describes a numerical simulation based work on the use of a variable 2D “dimple” just aft of the nozzle throat to mitigate shock wave oscillations. The results presented are compelling and demonstrate an interesting way of controlling or perhaps eventually getting rid of the significant problems caused by these oscillating shock waves. There are a few comments that nevertheless should be addressed before the manuscript can be accepted:
-
As far as the reviewer can tell, simulations were conducted on 2D “dimples” where only depth and length can be varied. However, realistically a “dimple” should be a 3D depression along the surface, where its width varies in the streamwise direction. In this case, rather than a dimple, what we have is actually a 2D wave imposed on the surface. Hence, the authors should actually rename the “dimple” to a “2D wavy surface” or something equivalent, unless the flow behaviour can be demonstrated to be similar, which is unlikely. 3D dimples not only have streamwise recirculatory flows but spanwise recirculatory flow components as well that could spiral out of the dimple. This will not be produced by a 2D wave along a surface.
-
The optimal “dimple” is deemed to be the one that produces the smallest shock wave amplitude but how about the frequency of the shock wave oscillation? Is that not an important consideration as well?
-
Why is the “dimple” only imposed along the flat bottom wall? Fig. 6 results seem to suggest that a similar approach along the curved top wall could help as well, particularly with the significant recirculatory flow. Also, is the Mach number indicated in the legend magnitude? It could be worthwhile to plot the streamwise component, so that the extent of recirculatory flow can be better appreciated.
-
Can the authors comment if a flexible surface that is able to go and down is able to emulate the present wavy surface’s functions? It seems like it may work, judging by earlier studies that the authors cited and present results.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Earlier comments have been addressed well enough.
