Investigation of the Starting-Up Axial Hydraulic Force and Structure Characteristics of Pump Turbine in Pump Mode
Round 1
Reviewer 1 Report
Manuscript ID: jmse-1090216
Title: Investigation of the Starting-Up Axial hydraulic force and Structure Characteristics of Pump-Turbine in Pump Mode
Authors: Zhongyu Mao , Ran Tao , Funan Chen , Huili Bi , Jingwei Cao , Yongyao Luo , Honggang Fan , Zhengwei Wang *
In this manuscript, using CFD commercial software CFX and the Reynolds averaged Navier-Stokes (RANS), the authors presented a numerical study in pump mode.
My comments are as follows:
- In commercial code (e.g., CFX) or open-source code (E.G., Code_Saturne), the Reynolds averaged Navier-Stokes (RANS) model is well defined. Also, in computational fluid dynamics, the k–omega (k–ω) turbulence model is common. Then, what is the new knowledge gap or originality in this paper?
- Several similar investigations in the literature for the starting up axial hydraulic force and structure characteristics of pump-turbine in pump mode are available. Then, what is a new contribution to this manuscript?
- The authors presented a literature review, then on page 3, suddenly jumped in 3D turbulent flow simulation. It looks like continuity with the previous section is missing.
- Please explain clearly the initial condition and boundary condition.
- Please explain the justification of using the elastic model in Table 2, while the authors used the von Mises stress.
- In Figure 8, the authors compared the experimental results with the numerical results. But, details of experiments are not presented.
Author Response
We sincerely thank the comments that helped to improve the manuscript. Details are given below.
1) In commercial code (e.g., CFX) or open-source code (E.G., Code_Saturne), the Reynolds averaged Navier-Stokes (RANS) model is well defined. Also, in computational fluid dynamics, the k–omega (k–ω) turbulence model is common. Then, what is the new knowledge gap or originality in this paper?
Reply: The originality of this paper including two terms of fluid and structure. In terms of fluid study, the transient process of pumping-mode starting-up is proceeded based on the 1D transient simulation of pipe and the 3D CFD simulation of pump-turbine. As the wave transmission in pipe, the in-out boundary conditions are dynamic. Based on the 1D transient simulation results of pipe, the dynamic boundary conditions can be applied to the 3D CFD simulation of pump-turbine. The theories of 1D transient simulation and 3D CFD simulation are both based on the N-S equations.
We have elaborated that on Line 92-93.
2) Several similar investigations in the literature for the starting up axial hydraulic force and structure characteristics of pump-turbine in pump mode are available. Then, what is a new contribution to this manuscript?
Reply: There are three contribution improving the exist investigations: First, the dynamic boundary conditions are considered in the transient process of starting-up, as mentioned above; Secondly, the hydraulic force applied on the structure of runner is based on the one-way FSI method; Thirdly, the structure model is based on the system of support bracket, shaft and runner of pump-turbine, while only the structure of runner were focused before.
3) The authors presented a literature review, then on page 3, suddenly jumped in 3D turbulent flow simulation. It looks like continuity with the previous section is missing.
Reply: Thanks for your comment. Our literature review is done in the 1st section. As mentioned in the last paragraph of introduction, the 1DHT and 3D turbulent flow are proved important based on literature review. Therefore, it is common to introduce the governing equations of 3D simulation. This why we put it after introduction. Also, this is common in preparing the manuscript of CFD/FEM works. Please accept this arrangement. Thank you.
4) Please explain clearly the initial condition and boundary condition.
Reply: The initial condition and boundary condition are given in section 3 with added explanation after review. See Line 150, 171.
5) Please explain the justification of using the elastic model in Table 2, while the authors used the von Mises stress.
Reply: The elastic model in Table 2 is refer to the material properties of prototype pump-turbine. Table 2 is to introduce the material properties including density, Young’s modulus and Poisson’s ratio which are required in solving equation 7~9. By solving equation 7~9, the 1st~3rd stress can be predicted with calculating the von Mises stress. Hence, Table 2 is proper here.
Elaborate on Line 200.
6) In Figure 8, the authors compared the experimental results with the numerical results. But, details of experiments are not presented.
Reply: The details of experiments have been added to section 4.1. See Line 226 to 234.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Overall, the manuscript can be accepted after revision. I attach my review report.
Comments for author File: 
Comments.pdf
Author Response
We sincerely thank the comments that helped to improve the manuscript. Details are given below.
1) In this manuscript, the numerical simulation for flow part is based on CFD commercial software, CFX. The structure component is based on ANSYS. The authors may need to clearly state their new contribution to the implementation of the existing software. Otherwise, there will be no new contribution to the numerical technique, and the study will become a numerical exercise of the existing software.
Reply: There are three contribution to the implementation of the existing software: First, the dynamic boundary conditions are considered in the transient process of starting-up based on the 1D transient simulation of pipe; Secondly, the hydraulic force applied on the structure of runner is based on the one-way FSI method; Thirdly, the structure model is based on the system of support bracket, shaft and runner of pump-turbine, while only the structure of runner were focused before. In summary, this study integrates the flow field of 1D pipeline and 3D pump-turbine and the structure filed of pump-turbine system based on the software CFX and ANSYS.
We have elaborated that on Line 92-93.
2) In Figure 8, larger errors exist in the initial of starting up. Is there any way to avoid it or it is natural occurred?
Reply: At the initial of starting-up, the guide vane opening and mass flow rate are very small. As discussed in section 5.1.2, there are high speed jet flow between two guide vane blades and complicated vortex in runner and guide vane. The simulation at this status require extremely small mesh size and computational accuracy which may charge massive computing resources. In the future work, the further study will focused on the initial of starting-up and find out the feature of axial hydraulic force.
3) Figure 9 shows the comparison between 1D & 3D models. There is a fluctuation of both 3D and 1D models for the time between 0 and 5 sec. Any explanation?
Reply: There is the water hammer effect in pipeline with the guide vane open, causing the outlet pressure fluctuate.
Elaborate on Line 183-184.
4) In Section 4.2, “Figure 9 shows the mesh of structural field...”, should it be Figure 10?
Reply: Sorry for the mistake. Corrected.
5) Also, in the same paragraph, Figure 10 should be Figure 11.
Reply: Corrected.
6) Line 405: "In order understand" should be "In order to understand".
Reply: Thanks to point it out. Corrected.
7) Why does Dmax always occur at D1?
Reply: The reason is that the downward hydraulic force on the D1 side is larger than that on the D2 side, which caused by the asymmetric flow pattern in guide vane and volute.
Elaborate on Line 424.
8) Lines 450-454, the authors mentioned "the stress of runner is not only influenced by the hydraulic forces on crown and band but also affected by the blade force". I agree with this statement. Although this manuscript focuses on the axial hydraulic, it could be better to know how many percentage of contribution from axial hydraulic force.
Reply: That’s a good piece of advice. The axial hydraulic forces and blade forces work together as a joint force on runner. Actually, the proportion can be decomposed by order-reduction. However, it requires a large amount of time and effort in algorithm but it is not the main concerns of this study. Thanks for your advice, it will be a good individual topic to decompose and understand the contribution of force components. Please allow us to conduct analyses and discussions in the future works.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Accepted in the present form.
