Numerical Simulation of Breathing Mode Oscillation on Bubble Detachment

Round 1

Reviewer 1 Report

REVIEWER COMMENTS ON [Fluids] Manuscript ID_ fluids-795250

Numerical simulation of breathing mode oscillation on bubble detachment

 

summary - the authors present results of bubble nozzle pinch-off and subsequent oscillation on an intermediate temporal and spatial scale - experiment plus computation is performed

general judgement - One novel aspect is the attempt to compare the absolute pressure of experiment and computation, and this effort is only approximate for reasons detailed below.   The other novel aspect is the observation of an apparent low frequency modulation (or apparent subharmonic) in the post-pinch-off bubble oscillation as recorded by hydrophone and also predicted in VOF computation.  However the authors make no attempt to explain this observation. This paper would be much improved by addressing these two points of originality.  Otherwise, bubble pinch-off and subsequent oscillation has been well and thoroughly studied in previous literature, which is not very well cited by the authors

 

particular notes:

Already grammar errors in first two sentences of the abstract!!!!!

Eotvos number sp in Nomenclature

 

the writing is problematic throughout, including unfortunate grammar and misspellings (even in the last citation, the book author is “Clift”, not “Crift”

 

If f0 ~ 1kHz, then 40 kS/s is ok for analog voltage from hydrophone:  But, 1kfps for camera is not!!!!

 

fig. 1 too small, what are your lateral dimensions??  figure caption needs work

 

line 143 it is Eotvos not Etovos

 

fig 2 please give details such as fps, image scale (FOV, for example), nozzle diameter, and were all experiments air in water? continue throughout the manuscript

 

fig 2 b so you know the polarity of your hydrophone?  that is, do positive voltages correspond to compression or rarefaction?  Moreover, if you are truly trying to compare absolute pressure of measurement and computation, you must either include the hydrophone (a large hydrodynamic and acoustic object) in your computations, or make efforts to back out the true pressure in the absence of the hydrophone from the hydrophone measurement.  Absent these efforts, comparison of absolute pressure numbers from experiment to theory is approximate at best.  

 

you note the hydrophone signal is not a pure decaying sinusoid - given at least one of your tank dimensions is 0.5m, one would expect the first return of the emitted sound of the bubble to arrive at roughly 0.3 ms - meaning that the entire waveform is a superposition of emitted and (multiply) reflected waves - this has bearing on your observations of modulated sounds and the subharmonics in theFFT.  

 

Section 3 beginning line 174 and following, equations 4 - 6 - if you are trying to perform an absolute pressure analysis, you must incorporate an expression that describes how the pressure varies in space outside the bubble - you cannot take the apparent acoustic pressure from the hydrophone and assume that is the quasi static pressure at the surface of the bubble, which is your approach.  You must use the first integral of the Bernoulli equation and include both the r^-1 and the r^-4 term as well since the hydrophone is fairly close to the bubble.  This pressure is well documented in many bubble publications.  This would her a better approximation for comparison with VOF calculations than your r^-1 correction

 

So far as I can ascertain, the only original observation in this manuscript is found in the time series ( Figs2b, 6c, 8a&b, and. the spectra in Figure 11.  Both VOF simulation and experiment show this apparent lower frequency, with a ratio of roughly 1/3 the Minnaert frequency.  However, no analysis or physical mechanism evocation is attempted to explain the lower frequency peaks in the FFTs Figs. 11a and b or,  the apparent lower-frequency modulation in the time series Figs. 10 a&b.  the authors only say something about ‘coupling’ with the nozzle interface.     Why not work to explain this??  much previous work has been done in computing of shape and volume oscillations (one avenue), and much previous work has been done in 2-bubble coupling (another avenue) just to name two possible routes towards explanation not to mention trying to rule out the notion that you might be mode-locked to an eigenmode of your test chamber.

 

references are a bit slight for a topic so thoroughly investigated previously in the literature.

Author Response

Thank you for your valuable comments. We completed the revision. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript presents experimental and simulation-based results for a bubble detaching from a nozzle and the acoustics that result. It is an interesting problem and the results are generally quite good (figure 2 is particularly beautiful). These are coupled with detailed analysis via Rayleigh--Plesset-type equations and physical justification. The quality of the work is good, but I see room for improvement. Particularly, the disagreement between experiment and simulation is blamed on grid-resolution, though I don't find this particularly compelling. I believe the authors should perform axisymmetric simulations to allow for significantly higher grid resolution to justify their claims. Also: the writing is generally vague in many places. It would also benefit from grammatical improvement throughout; perhaps consulting with a native English speaker would be useful.

Some general comments are below:

• Section 4.1 generally needs to be improved. What justifies the modeling choices employed? 
• Line 225/226: How is $n$ computed? For an interface capturing scheme this is challenging and often involves advecting a color function for the materials. It is also often not robust. We need more reason to trust that the surface tension implementation is accurate and validated.
• Line 231: What is a MARS scheme and why is it chosen here?
• Line 232: Why is Simpson's rule used for time integration? I find this to be a strange choice and was appreciate some justification.
• Figure 6: Are these log or linear color bars? How is the interface (black line) computed?
• Figure 7: Insufficient information in capture and figure. What is plotted, what do the letters mean? What do the numbers mean? What is \phi? 
• Line 248: This sentence is vague and meaningless in its current form. Please improve or remove.
• Line 262: Awkward is not the correct word here.
• Figure 5(a) bottom: What cross section is this? Vertical or horizontal? Please make this obvious in the caption or picture.
• Figure 5(b): the \phi used here doesn't match the style of the \phi used in the text.
• Relating to Figure 5: Why is the domain so large? Is this necessary? 
• Relating to the mesh: Is a fully three-dimensional unstructured mesh used? If so, why? It appears that this problem can be solved by assuming axi-symmetry, especially after looking at the experimental results. This would save considerable computational time and allow for further grid refinement as is alluded to later in the paper and indeed in line 402. If such grid accuracy is important for this study, I recommend simulating much larger configurations using an axi-symmetric assumption.
• Line 298: comparing these disparate figures, several pages away from one another, is challenging. The authors should ease this by putting them side-by-side for similar time instances. 
• Line 375: What amplitude?
• Line 379: 'the solution will become excellent'. This seems speculative and I'm not sure why I should believe it.
• Line 382: 'essential subject in the future'. Seems speculative and I'm not sure why I should believe it. I don't think this kind of language is necessary or appropriate.

Author Response

Thank you for your valuable comments. We completed the revision. Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

I appreciate the responsiveness of the authors and recommend publication.