Review Reports - Numerical Study on Flow Stability Improvement of an Automotive HVAC Compressor’s Expansion Chamber Using Parametric Mufflers

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study only reports pressure pulsation and mass-flow fluctuation; no sound pressure level, transmission loss, or acoustic analogy result is provided.

Experimental validation is important. The CFD model should be validated against measured pressure/mass-flow data for at least the baseline and optimal cases.

The double-muffler optimization is incomplete. The influence of the clearance was not thoroughly considered.. The conflicting optima for pressure and mass flux are attributed to “simulation uncertainties” without quantitative error analysis.

Is pressure pulsation peak-to-peak? Is mass flux fluctuation RMS, standard deviation, or peak-to-peak?

CFD model setup is insufficient. Provide refrigerant properties, compressibility/phase-change treatment, turbulence-model justification, time-step independence, and the source of the inlet pressure curve.

Manuscript quality needs improvement, including inconsistent symbols, grammar/typos, and unclear figures/units.

Additionally, the references should focus on the research conducted in the past five years.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Majot finding in this paper is - the optimized single muffler reduces outlet pressure pulsation from 3700 Pa to 2470 Pa, while the optimized double-muffler configurations reach 1880–2020 Pa. The authors also report reductions in the reported mass-flow fluctuation.

most important issue - The paper repeatedly frames the work around flow-induced noise, vibration and NVH, and introduces Lighthill's acoustic analogy. However, the actual simulations appear to calculate only - pressure pulsation, mass-flow, velocity. There is no acoustic propagation calculation, acoustic pressure, sound pressure level, transmission loss, acoustic power, FW-H calculation, FEM/BEM acoustic model, or experimental noise measurement. The authors themselves acknowledge that pressure pulsation is being used as an acoustic excitation indicator. Therefore, statements such as reduction in pressure pulsation leads to suppression of flow-induced noise should be presented as a implication, not as a demonstrated noise reduction. I strongly recommend revising the title, abstract, introduction and conclusions accordingly.
The manuscript says that density, viscosity, thermal conductivity, enthalpy and temperature are defined according to the physical characteristics of the working fluid, but the actual refrigerant is not clearly specified, nor are its thermophysical properties or property correlations given.
The theoretical section presents the conventional Navier–Stokes equations, but the paper does not adequately establish whether the simulations are compressible or not
The manuscript defines Sections A–D and states that Section D is an extension used for boundary conditions. It then specifies the condenser-side pressure at the end of Section D, while the realistic outlet is stated to be at Section C. clearly explain Where exactly is the inlet, outlet, Why is Section D required, Is the pressure boundary applied at C, D, or both, etc
The inlet pressure Pin is taken from previous simulation results of the compression chamber. The actual pressure-time waveform should be provided
I could not find sufficient information about the actual compressor operating condition.
A time-step independence study is needed, just as a mesh independence study is presented.
total cell numbers, mesh quality parameters, inflation/boundary-layer treatment, is not reported.
The study uses the standard k−ϵ model. the authors should justify why standard k−ϵ is adequate.
The manuscript calls the second objective “mass flux fluctuation”, but the reported unit is kg/s. kg/s represents mass flow rate.
If the numerical uncertainty is comparable to the difference between candidate designs, the manuscript cannot confidently distinguish those designs. The authors should quantify uncertainty
The introduction cites several studies where CFD predictions are compared against experiments, and emphasizes the importance of numerical/experimental validation. However, this manuscript itself appears to contain no experimental validation. This is particularly problematic because the paper is based on a real mass-produced compressor geometry. the authors should discuss this limitation explicitly..
convergence criteria are set to 0.1 for both flow and heat models........ What does 0.1 represent?
Table 4 contains a unit error
Grammar and technical English require editing....comfortability, “flow filed, “the famous Navier-Stokes equationsetc.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript presents a transient CFD study of the discharge-side expansion chamber of a mass-produced automotive HVAC compressor, in which a short muffling segment is introduced into the discharge pipe and the offset between the pipe centrelines on either side of it is varied. Forty-three geometries, in single- and double-muffler configurations, are compared on two criteria — outlet pressure pulsation and outlet mass flow fluctuation — with reported reductions of up to 49 % and 62 % relative to the original chamber.

Overall, the work is relevant to the journal, however it requires major revisions:

Several abbreviations are used before being defined: OEM, NEV, ICE, BEV and IDV. Please define each at first use in the text. For NEV, a brief gloss of the category would help readers outside that market.
Could the authors state the length of Section D and show how the results change when it is doubled or halved.
Why didn’t the authors place the fixed-pressure outlet with a non-reflecting/impedance condition or extend the downstream pipe so that the sampling plane is several wavelengths from the boundary. In the current position, the pulsation is constrained by the boundary, not only by the chamber. And this also depends on the length of D.
Both boundaries are pressure boundaries, so the mass flow is an output of the model, not an input. Every modification adds flow resistance, and the model can respond by passing less gas — which by itself shrinks the absolute fluctuation. The 54.50 % and 62.49 % reductions may therefore be a throttled flow rather than a stabilised one. The added resistance also costs pressure drop, which is never reported although it determines compressor work and COP. Authors could report the mean mass flow (with the fluctuation normalised by it) and the mean pressure drop for the baseline and the three optimal designs. Both are available from the runs already performed.
The explanation given (longer path, longer dwell time) is asserted without any supporting quantity, so the design rule cannot be transferred to another compressor. Authors could provide the frequency content of the inlet and outlet pressure, quantify the claimed mechanism (residence time or an outlet uniformity index), and discuss the results against classical muffler theory.
There are several important criteria missing and they should be added
- a table of the refrigerant, its thermodynamic model, the inlet temperature, the wall thermal condition, and the solver settings.
- a dimensioned drawing, or at least a table of the main dimensions; if the geometry is proprietary, give normalised dimensions.
- the compressor type and speed and confirm the fundamental frequency in Figure 4 corresponds to it.
- Report cell counts and y⁺, and justify k-ε or repeat a case with SST k-ω.
The mesh study does not cover the geometries the conclusions rest on. The authors could
- Repeat the mesh study on at least one muffler geometry (Model 9 and the optimal double muffler) and for both metrics.
- Report cell counts and use a formal estimator such as the Grid Convergence Index
- Add a time-step refinement study (for example halving Δt).
- Restate the convergence claim within the real output precision.
Numerical settings are under-specified and partly contradictory
- In line 286, “The numeric scheme for velocity and pressure is upwind”. First-order upwind is strongly dissipative and will damp precisely the pulsation amplitudes being measured. If the scheme really is first-order, the reported attenuations have a numerical component.
- Line 287 to 288, “Second-order implicit backward Euler” is a contradiction in terms: backward Euler is first-order; the second-order member of that family is BDF2.
- Line 289 to 290, “The convergence criteria are set to 0.1” gives neither the variable nor whether this is an absolute or a normalised residual. As written it is a very loose criterion.
- In line 280, “The time step is set to 1800” confuses the number of time steps with the step size; Δt = 8.33×10⁻⁵ s is correct for 0.15 s / 1800.
- Line 321 to 324, “The run covers only about five pulsation cycles, and the peak-to-peak value is taken “across the full-time domain”, which includes the start-up transient. A single start-up spike can set the reported metric.
The title, abstract and Section 3 all use “optimization”, but the procedure varies one factor at a time with fixed values for the others.
The paper computes no acoustic quantity, yet the abstract, title and introduction are framed around noise and vibration.
For the single-muffler study δ is treated as a magnitude from 0 to 12 mm; for the double-muffler study a sign convention is introduced. The relation between the two definitions should be stated explicitly.
Line 234 to 241 argue that a CFD-based pulsation reduction “can be reasonably expected to translate into reduced radiated noise”, supported by references [9] and [10]. Those studies concern a hermetic refrigerator compressor and an industrial reciprocating suction system — different machines, different frequency ranges.
Figures 9 and 13 are compared qualitatively, but the colour scales differ (0–10 m/s versus 0–8 m/s), and it is not said whether the streamlines are instantaneous or time-averaged, nor at which instant.
“approximately 20 % of R&D expenditure … is spent on NVH issues” no citation supporting the quantitative claim
the compressor “compresses the refrigerant coming out of the evaporator into gas” The refrigerant leaving the evaporator is already vapour; the compressor raises its pressure and temperature.
Figure 1 shows a belt-driven compressor with an orifice tube and accumulator, i.e. a conventional combustion-engine system, which contradicts the BEV motivation of the introduction.
“computational fluid mechanics (CFD)” Computational fluid dynamics
In equation 2, there shouldn’t be ∇ in the term is ρ ∂u/∂t
From equations 1 and 2 “form the famous Navier–Stokes equations”, they become Navier–Stokes only with a Newtonian constitutive law for τ, which is not given. The energy equation and the turbulence equations are missing although both are solved.
“mass flux fluctuation” in kg/s. kg/s is a mass flow rate; mass flux is kg/(m²·s). Also, “fluctuation” is never defined — presumably peak-to-peak
Figure 5, “Coeffcient” is misspelt in the legend
δ denotes both the pipe offset and the Kronecker delta (δᵢⱼ). Different symbol for the offset could be used.
Grammar errors

"noise and vibrations … is" --- are

"manufactures" --- manufacturers

"[15].The" --- (missing space)

"flow filed" --- flow field

"lay the foundation" --- laid

"turns into" --- turn

"time step is set to 1800" --- 1800 time steps

"reducing abundant computational time" --- avoiding excessive computational time