Velocity Mapping of an H₂ − O₂ Exhaust Jet in Air by Means of Schlieren Image Velocimetry (SIV)

Round 1

Reviewer 1 Report (Previous Reviewer 1)

The manuscript presented both the experimental and CFD studies of a H2-O2 jet. SIV was applied to obtain the jet velocity data that are compared with CFD simulation. Due to the inherent nature of schlieren imaging, which integrates a 3D jet along the optical path to from a 2D image, it's difficult to obtain accurate data because the loss of spatial resolution along the optical path. More discussion and justification on the comparing of the experimental results with the CFD simulation should be provided. There are also mistakes and lack of clearance in the manuscript. For example, the discussion of Fig. 10 does not seem to match the presented figure. The font size is too small in many of the figures. There is an error in generating the right figure in line 160.

On the content side, some 2D maps of the SIV and CFD would improve the general understanding of the flow structure and the similarity between the experimental results and the CFD simulation.

 

It's generally readable but it can be further improved.

Author Response

Hi, 

Thank you for your suggestions. Please find attached a word document with more explanation of the modifications made.

Author Response File: Author Response.docx

Reviewer 2 Report (Previous Reviewer 3)

I thanks the authors for taking the time to respond to my original comments, though it would have been easier to assess your response if the particular revisions had been clearly stated in your response, rather than simply referring back to the manuscript. The introduction has been greatly improved and many of my queries around the methodology have been answered. I find it rather odd that the results from the Lagrangian tracking methods appear to have been completely removed from the paper, yet the method is still discussed in detail and the in the results section your mention the accuracy of this approach but do not provide any data to back this up. Without providing this data, I do not see how you can discuss these results. These either have to be included or this section has to be completely removed from the paper, however without this I do not see sufficient content to warrant publication. There are also a number of comments made in the paper that require rewording and or clarification, because at present I cannot follow their intent. These comments can be found in the attached file. I still have fundamental questions about the validity of this method that have not been addressed here. A paper that demonstrates that these approach appear to have 20% error is not particularly surprising, but even that does not address if this is a fundamental error in the experiment or a failure of the turbulence model used in the CFD to capture the true physics. 

Comments for author File: Comments.pdf

Author Response

Hello! Thank you for all the meaningful suggestions.

Please find attached a document with the response from the comments. 

Most suggestions (I hope all of them) have been included in the article's text.

Author Response File: Author Response.docx

Reviewer 3 Report (New Reviewer)

The authors have presented a methodology to extract velocity using two methods, i.e., the Lagrangian tracking method and direct image correlation using a kymograph for an H2-O2 micro thruster exhaust jet. Finally, results from SIV (schlieren image velocimetry) are compared to results from CFD simulations to check the algorithm's accuracy. The work is very interesting. Still, in my opinion, the authors have to explain the extract/image manipulation using MATLAB, and ImageJ should be described a little bit in detail. 

1. for example, details of the first 5 steps, which are usual for any image processing researcher, will help others. This can be given in appendix or present snippet code or given part of in github so that others can use it. This would help the author to enhance visibility of the publication. (section 2)

2. Similarly, details of kymograms extraction using imageJ has to be detailed out. 

3. Will the standard k-e model predict the phenomenon well? In my opinion, authors should try a better turbulence model.  The description of the computational domain is not clear. Also, did the authors check grid independence?

 

There are several places statements and words used are not grammatically correct. I request the authors to get the manuscript checked thoroughly. I am stating the mistakes in the abstract.

1. in line 10, 'scientifical' should be 'scientific' 

2. line 13, 'The test case of this study centers consists out of' should 'The test case for the present study consists of'

3. line 15 'firstly' should be 'initially'

Author Response

Hello!

Please find attached the answer to your previous comments.

Thank you!

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report (Previous Reviewer 1)

The manuscript has been improved but only marginally.

Author Response

Dear Sir/Madame,

1. A 2D map of the method accuracy has been added as requested by Reviewer 2.
2. The CFD simulation has been further explained.
3. The results of the lagrangian method have been explained on the new uploaded graph.

Please find the correction added in the new version of the manuscript.

Reviewer 2 Report (Previous Reviewer 3)

Thank you for your revisions. I think the way the results from the Lagrangian tracking are present is still not entirely clear. Principally it is not clear how the data in Figure 9 has been used to come up with an accuracy for this approach. The text mentions that traceable elements were only found from 34 mm downstream of the nozzle, but data is present much closer to the nozzle, at least for the v component, but how can you get a v component without a u component? Also you state that error should be small away for x > 120 mm, but I don't see how you can make this claim without presenting data from that region. Is this figure of all tracers at a given x location, or only those on the centre line? Importantly, given the scatter in this data, how do you arrive at an accuracy of 62.5%? Is this based on the statistical distribution of these samples and accuracy of the mean? Does this represent a global accuracy from 0 to 120 mm? The scatter plot shows does not enable the reader so see the the average velocity measured at say x = 60 mm, particular given the tracer size and the scale (-400 to 500) used. The presentation of this data has to be improved. For example the probability density function of measured the velocity could be provided. As stated in previous reviews, CFD data can not be called a theoretical profile. 

The manuscript and particularly the most recent changes, need to be checked again and corrected.

Author Response

Dear Sir,

Please find the answer to your inquiries in the attached file. 

Author Response File: Author Response.docx

Reviewer 3 Report (New Reviewer)

The manuscript can accepted in present form. Annexure should be published.

The language of the manuscript is good.

Author Response

Thank you!

Round 3

Reviewer 2 Report (Previous Reviewer 3)

I thanks the authors for taking the time to address my comments. I am now happy that the lagrangian tracking results have been presented in a way that readers can better appreciate the authors conclusions.

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

The manuscript reported velocity mapping for a round turbulent ?? − ?? exhaust jet by schlieren image velocimetry. The technique is not new and there are many other research studies available but the authors only had a very brief review and limited discussion on the state of the art in this field. A more comprehensive literature should be carried out. A schlieren imaging is an integration along the optical path and great care is required to interpret the obtained velocity based on one 2D PIV principle. More uncertainty and error analysis has to be carried out to justify the comparison with the CFD data.

Reviewer 2 Report

1. Presents a limited synthesis of the literature cite. The last of reference should be number.

2. Demonstrates an inadequate understanding of the critical aspect of a literature review and briefly to critique sources (gap analysis).

 

Reviewer 3 Report

The authors present a manuscript focused on assessing the extent to which the temporal variation in schlieren images can be used to estimate the velocity. Given that schlieren images represent integrated line-of-sight refractive index gradients, it is not surprising that this approach yields a poor estimate of velocity. Even if this could faithfully track particular eddies, this approach would always be biased toward particular gradients due to the knife-edge cut-off and it is known that individual eddies convect at velocity that differ from the local bulk velocity. Based on the literature that has been discussed in the manuscript, it is not clear what motivates this particular work and what unresolved questions this work has set out to address. The discussion of previous work mentioned what has been done, but offers little in terms of critic. Importantly some of the discussion in the introduction incorrectly represent some of the previous work such as [2], which does not appear to consider velocity determination from BOS data at all. Overall this manuscript lack explanation behind some of the parameters, in particular those used for the PIV type analysis, which uses unusually small windows for a single pass approach. Similarly it is not clear why it is necessary to perform an edge detection. Applied algorithms require reference to the originating works, more than the name of a MATLAB function. It is important that readers know what the function is actually doing. Axis label on multiple plots are either missing or incorrect, and in some cases lines like those in Figure 6 are added without an explanation for how this points were selected or why this is representative of the mean. Critically the basis for validation of the results is a CFD simulation, but from what is presented, it is not possible for the reviewer to access how well this CFD present the experiment. It is mentioned that the validation was based on pressure and thrust comparisons but there are not presented, nor is a reference to this provided.

In addition to this, I have included an annotated PDF that includes additional points for consideration and highlights some of the issue raised above.

Comments for author File: Comments.pdf