Detection and Analysis of an Alternate Flow Pattern in a Radial Vaned Diffuser

Round 1

Reviewer 1 Report

See comments in the attached pdf: ijtpp-614073-review.pdf.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is reporting, based on unsteady pressure measurements, the observation of an alternate flow pattern in a radial vaned diffuser with a even number of vanes. The link between this phenomena and the transition to surge is then analysed.

The quality of the paper is very good and interesting. The results are original and the scientific approach is clear.

The reviewer is just proposing some minor remarks :

1/ The main problem is a problem of vocabulary : the authors refers several times, to qualify the different regimes of the flow, to an axisymmetric flow. But the flow is never axisymmetric (it can be seen for example, in figure 2 : the pressure is depending on the azimuthal coordinate).

The phenomenon described in the paper is not a loss of axisymmetry but a modification of periodicity.

2/ 'These' instead of 'those' (Page 1, Line 25 and page 5, Line 153).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

1: The authors did not provide a second case in order to make their results more robust. This is a limitation of the work.

2. Line 43: why is the variable “asymmetry” rate changed to “alternate” rate?

3. The authors should clarify the difference between (line 13) “alternate stall pattern” and (line 36) “alternate pattern”.

4: Line 40 makes a reference to “alternate pattern” in previous studies [4,5]. How does this fit with the abstract “a new variable, the alternate rate A”?

5. The authors were asked to define the test case so the reader may understand what the measurement is for. Despite the addition of Fig. 1 or Fig. 2, it is still unclear what the test case looks like.

6. There are details in Fig. 2 that needs to be explained.

7. Line 77-78: where is the location of the test rig inlet, perhaps it should be indicated in Fig. 2? The calibration of the diaphragm flowmeter should be indicated.

8. Multi figure graphics (Fig. 3) might need annotation a) b) and c).

9. Line 100: perhaps a “flush-mounted” sensor is a better word?

10. Line 108: the authors introduce a numerical steady simulation. The authors should clarify why a steady flow is applicable to rotating stall and surge.

11. Line 108-109: The abbreviations RANS and elsA should be clarified. The authors were asked to clarify the numerical simulations but say its “unnecessary information”. Please explain why this is “unnecessary information”. If the objective is to describe the “alternate pattern” feature it seems crucial that the numerical solution is presented in a comprehensive way and that the prediction is correct.

12. There seem to be errors in Fig. 4. The authors need to clarify why there is an aperiodic blade-to-blade distribution in the numerical steady simulation. Perhaps the simulation is not converged? There is a shock line indicated in Fig. 4b but the numerical solution does not seem to have a sharp gradient. There are unexplained features in Fig. 4. Does the subscript “abs” stand for absolute? If the radial diffuser is stationary then why is there a need for the subscript? The “min” “max” on the pressure colorbar needs numerical values. The authors were asked to clarify the location of stall and the location of the jet. Since Mach number is a scalar property in Fig. 4b, it is still unclear where the flow is separated and where it develops a jet. The authors might want to indicate the diffuser inlet (Line 114) and “leading edge of the diffuser” (Line 115) Fig. 4.

13. The authors need to explain why they distribute the sensors instead of direct measurement in the desired channel or in two adjacent channels.

14.  The authors were encouraged to formulate Eq. (1) in discrete form since the measurements were taken at discrete points. The rebuttal contains an interesting comparison with the uniformity index. It might be useful to add this discussion and shown under which circumstances that the “alternate” index reduces to the uniformity index. The authors may want to elaborate on when the “alternate” index is necessary and when the uniformity index fails to be sufficient.

15. The authors were asked to define the projections 1p and 2p (now Fig. 5) but the manuscript does not seem to be updated with this information.

16. The rebuttal states “2D pressure maps”. Thus, is the numerical simulation based on 2D and how is this applicable to a 3D flow?

17. Some figures have poor quality, e.g. Fig . 6 contains text annotations that are not readable.

18. “The limitation A<0.007 actually corresponds to the double of the noise error due to the difference in the discretization between the one channel and the two-channel projection (cf. Fig.2 (a)).” It's unclear what “double of the noise error” means. Fig. 2a does not seem to contain this information?  The explanation in the rebuttal does not seem to have been added to the manuscript? How does region A < 0.007 change with respect to the discretization? What happens if the resolution increases a factor two?

19. The authors were asked to clarify the interpolation procedure employed (now Fig. 7). It is still unclear why the result is not biased when only four speed-lines are used in the analysis. How come some speed-lines are classified and some others are not? It is customary to define all necessary operating conditions so that other research groups can reproduce the test setup and independently confirm the results.

20. Line 185:  it is not clear what “the flow is similar” implies. Does it imply that the velocity profiles blade-to-blade are equal between the passages?  

21. Line 189: please clarify “a slight alternation”. How exactly is the flow changed between the two channels?

22. The authors were asked to define mild and deep surge (Line 193-194). However, it is still unclear if there is any flow reversal in the investigated test case. (Line 195-221) a high-temperature reading at the rotor inlet might be necessary but may not be sufficient to determine if the flow is reversed. Suggest a rewrite.

23. The authors were asked to clarify how Fig. 6 was constructed. Is it an accurate representation of the test case? Where is the location of the blade-to-blade segment with respect to the diffuser channel? The rebuttal states “sketches are drawn to highlight the separated flow regions (colored areas) in the diffuser. This information has been added to the paper.” It is not clear if this has been added to Fig. 6?

24. The authors were encouraged to clarify stall patterns and stalled diffuser passages for the test case so it can be better connected with the different alternate rate regimes. However, the added figures still do not seem to contain this information.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

NA

Author Response

Nothing to modify according to the reviewer.