Computer Flow Simulation and Verification for Turbine Blade Channel Formed by the C-90-22 A Profile
Round 1
Reviewer 1 Report
· This paper has obvious outcomes which are already published in previous literature. There is no novelty in work.
· Keywords: Ansys can’t be a keyword. It is an organisation/company. Similarly, verification seems like a non-technical keyword.
· Line-70-71: Authors say that “the choice of a computational grid that provides the number of control volumes within the range of 500,000 to 1,500,000 elements”. It is only true for a particular case. It may vary from case to case.
· The governing equations for fluid flow and for SST-k-omega turbulence model are missing.
· The equation of “friction energy loss coefficient” and the equation of error is missing, they will be used for comparison between experimental and simulated values.
· Authors have concluded that “it is recommended to take values of y+ close to 1 since such values provide an acceptable error for all considered configurations and modes.” This is obvious. The recommended value for y+ is 1 for SST k-omega model.
Author Response
Point 1. Keywords: Ansys can’t be a keyword. It is an organisation/company. Similarly, verification seems like a non-technical keyword.
Response 1. We removed this keyword from new version of , thank you for comment.
Point 2. Line-70-71: Authors say that “the choice of a computational grid that provides the number of control volumes within the range of 500,000 to 1,500,000 elements”. It is only true for a particular case. It may vary from case to case.
Response 2. We removed this lines from new version of varticle, thank you for comment.
Point 3. The governing equations for fluid flow and for SST-k-omega turbulence model are missing.
Response 3. We added this equations to new version of article, thank you for comment.
Point 4. The equation of “friction energy loss coefficient” and the equation of error is missing, they will be used for comparison between experimental and simulated values.
Response 4. We added this equations to new version of article, thank you for comment.
Point 5. Authors have concluded that “it is recommended to take values of y+ close to 1 since such values provide an acceptable error for all considered configurations and modes.” This is obvious. The recommended value for y+ is 1 for SST k-omega model.
Response 5. In new version of article we extended the conclusions, the main conclusion lies in various recomendation for y+ with the minimum error for various Re.
Point 5. This paper has obvious outcomes which are already published in previous literature. There is no novelty in work.
Response 5. In this article was clarified recommendations for choice of y+ for SST model with regard to blade channel. In literature there are only general recommendations without specific limits of applicability for varios values of parameters.
Reviewer 2 Report
Benchmark solutions are important tools for CFD analysis to assess the accuracy of the numerical method that used, and to validate with the practical implementation. The capability of the numerical solution should be examined in a clear way, including boundary conditions and initial conditions in each numerical condition. The accuracy of implementation of boundary and initial conditions should be fully examined. Formal order of accuracy should be given by assessing (i) grid/mesh convergence, (ii) code independence, and (iii) capability of capturing the solution. All of these important information to provide the accuracy of numerical results should be available before validating with the established correlation and/or the experimental results. Moreover, because of in this study focus on the simulation and verification, it is necessary to plot the error on y axis-versus delta X (a determined parameter) on x-axis, then find the order of convergence. Adding the CPU time to the results is also useful.
Author Response
Point 1. Formal order of accuracy should be given by assessing (i) grid/mesh convergence.
Response 1. In this study maximal order of convergence was established in 10E-4, so it is accetable level for convergence to engineering cases with significant turbulence.
Point 2. (ii) code independence
Response 2. Considering turbulence models are the same for other CFD codes, Fluent or Comsol for example.
Point 3. (iii) capability of capturing the solution
Response 3. Capability of capturing the solution considering in study with considering dependence of solver settings to numerical solution error in comparsion with experimental data.
Point 4. Moreover, because of in this study focus on the simulation and verification, it is necessary to plot the error on y axis-versus delta X (a determined parameter) on x-axis, then find the order of convergence
Response 4. We added variant of this plot with sensitivity analysis in new version of article, thank you for comment.
Round 2
Reviewer 1 Report
Can be published after following minor corrections
- Remove keyword "verification"
- Give the equation of percentage error used in Figure 5
Author Response
Point 1. Remove keyword "verification"
Response 1. We removed this keyword in the new version of article. Thank you for comment.
Point 2. Give the equation of percentage error used in Figure 5
Response 2. For the percentage error in Figure 5 and Figures 6-7 the equation 7 on lines 214-215 is used. On lines 220-222 we noted this separately in the new version of article. Thank you for comment.
Reviewer 2 Report
Improved.
It could be accepted.
Author Response
Dear reviewer
Thank you for comment
