Numerical Evaluation of Rotordynamic Coefficients for Compliant Foil Gas Seal
Round 1
Reviewer 1 Report
The article is interesting and describes current issues concerning rotordynamic coefficients for foil gas seal. Nevertheless, it contains many faults that need to be corrected.
1) Lots of editing errors when formatting text (see instructions for authors).
2) Point 3.1, the number of nodes and elements for FEM is not given. This has a major impact on the accuracy of numerical calculations.
3) How was the dimensional rotordynamic coefficients determined for dimensionless pressure?
4) The calculation methods used are not sufficiently described.
5) Not all drawings show correct unit designations.
Author Response
1) Lots of editing errors when formatting text (see instructions for authors).
I changed the edit errors according to the instruction of Applied Science.
2) Point 3.1, the number of nodes and elements for FEM is not given. This has a major impact on the accuracy of numerical calculations.
Thank you for your suggestion. The number of nodes and elements for FEM I used is based on the number nodes Lu’s paper and around 16000. I will test the grid independence and accuracy by myself in future.
3) How was the dimensional rotordynamic coefficients determined for dimensionless pressure?
The outlet pressure is common value in this paper. But the inlet pressure changes. The dimensionless pressure will change with the inlet pressure changes, so the rotordynamic coefficients changes with the inlet pressure changes in fact.
4) The calculation methods used are not sufficiently described.
I added the central difference method of FDM to solve the rotordynamic coefficients of the T groove of compliant foil gas seal in this paper.
5) Not all drawings show correct unit designations.
I changed some drawings errors, including Fig. 2 (b) and (c).
Author Response File: 
Author Response.docx
Reviewer 2 Report
The review is in attachments.
Comments for author File: 
Comments.pdf
Author Response
The changed part of this paper was marked with track revise
Review report of reviewer2 with response
- The authors should explain more, why the stationary model (eq. (2.1.1))
is satisfactory for the description of the problem. Is the problem without
dynamic effects at all?
I think the eq.2.1.1 is an equilibrium equation. It can be satisfactory for the description problem, including stationary and dynamic. It’s one of equations for solving the analytical solutions.
- The authors should (at least briefly) specify the boundary conditions of
(2.1.1).
Yes, I lost them. Thank you for your tips. I added them in the part 3.1 of the paper.
- Eq. (2.1.1) is nonlinear. The authors should explain how the nonlinear problem is solved. It would be nice, if the authors could also briefly mention which linear underlying solver is used.
The equation is nonlinear. But I replaced the every part linearly by citing the Taylor expansion and FDM and the value is approximation.
- The reference on Cheng HS should be supplemented by citation.
I add the reference in this paper.
5.The Comparison with Cheng HS and for different data is not a perfect verification of the reliability. A possible way for improvement is an application of a posteriori error estimates. Is it possible to implement some kind of a posteriori estimates to the authors’ code?
Thank you for your suggestion. You are right and the comparison is not a perfect verification. But there is not test rig for it and it’s just in the theory. I think I can only get the same trend as this paper use the different groove type and parameters. The number of nodes and elements for FEM I used is based on the number nodes Lu’s paper in this paper and around 16000. I will prepare a posteriori estimates by myself in future.
- Fig 2. (c) the curves contains quite large oscillations. The authors should give some explanation to this phenomena.
The oscillations may be due to the T groove. I had explained it briefly in the paper. Because the parts of large oscillations include the structure of T groove. The grooves lead to the pressure increase abruptly. That’s why we designed the structure.
- Table 1.: How L and h are chosen?
The L and h I chose is according to the reference[9] and [10].
- In the sentence ’... they can be guaranteed to study ...’, there should be the word ’guaranteed’ (probably) exchanged by ’used’.
Yes, I changed them. Thank you for your suggestions.
- Is Q leakage?
Yes. Q represents leakage. I added the explanations.
- The argument how (3.2.1), (3.2.2), (3.2.3), (3.2.4), (3.2.5) are derived should be more detailed.
- The references on equation (3.2.2), (3.2.3), (3.2.4), (3.2.5) seems to be wrong (they are shifted).
These equations are just explained for discussing the results. I discussed and analysed the trend by these equations. I added the related reference in later.
- Fig 4.: How the fitted curves are obtained?
- The number of points for fitting curves seems to be low in Fig 4. and the resulting information could be wrong (or very inaccurate). The conclusion mentions this topic as well. (similarly Fig 6. and Fig 8.)
Yes, this depend on the coefficients of equation I design. The meaning of "the fitting curve" is that the curve fitting should be plotted with more points and more parameters. The curve in the paper was plotted by solving the coefficients of equation, which just include important parameters I assume at the beginning. If more points, there would be errors in the solver, so I hope I can improve the fitting curve. It is benefit for improving the accuracy of prediction.
- Fig 7. There should be some (at least short) explanation, why the character of leakage changes around the speed 9000 rpm.
Thank you for your tips. The character of leakage changes around 9000rpm may be that the amplitude of the axial direction of speed increase differently though the blocking flow is still increase. I will focus on the question in future.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Thank you for your answer.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
This manuscript shows the numerically calculated rotordynamic coefficients of a gas seal. Several points are missing in the manuscript, especially the comparison between the present result and results available in literature. At this state, the manuscript is not recommended for publication.
The parameters introduced in eq. 2.1.1. are not explained (R, L, ...). The nomenclature at the end could be deleted because the reader will not find it when reading the manuscript from the beginning. I suggest of integrating the explanations throughout the text, as it is done already at some locations in the manuscript.
The parameters in Table 1 have units but these are not given.
C is commonly labelled clearance between the rotating and the non-rotating part and not mean gas thickness. Please check the wording carefully.
Eq. 2.1.2 uses C for clearance while in the introduction C is the damping.
A reference to the geometry of the chosen gas seal in Fig. 2 is missing.
Eq. 3.2.1 is identical to eq. 3.2.2. One equation should be deleted.
A reference to the short bearing approximation is missing.
A comparison with values available in the literature for the bearing coefficients is missing. For example, Fig. 3 can be found in literature for specific sealings. Does the Matlab code used here provide the same results?
Units are not italic (typeface). This is inconsistent throughout the manuscript.
Conclusion (4) is not clear: What is the meaning of "optimize the fitting curve"?
Looking at Fig. 6a, it seems that the fitting of the calculated points leads to a dramatic undershooting. I strongly suggest including more points (10 instead of 6) before doing a curve fitting.
Page 6, line 166: The value mentioned y = -10^8 is outside the region of Fig. 8? The unit of y is N*m while Fig. 8 shows N/m.
Reviewer 2 Report
The graphical presentation of the results needs to be improved.
The axis markings are too small.
In figure 2 it is non-dimensional P - should be p - according to the nomenclature.
And a few more editing errors "no spaces".
