Bulk viscosity-driven suppression of shear viscosity effects on the flow harmonics at RHIC

Bottom line: the part of this worth your time is the dynamical observation, not the headline. Tables III–V show that turning on bulk viscosity suppresses the mean and variance of pi00 through the evolution, and that the effect grows the longer the system expands. That is a clean, model-independent statement about the fluid, and it is the piece I would cite. The abstract's claim that bulk viscosity pushes v_n back toward ideal is true but only demonstrated where the freeze-out prescription is least trustworthy, and the authors mostly know it. Start with the tables, not Fig. 3. What the paper does well: the code is validated against two external benchmarks, the TECHQM shear test and the Gubser semi-analytical Israel-Stewart solution, in Appendix A. That is real and not universal in this literature. The event-by-event setup is standard and the bookkeeping in Tables II–V is careful. Where it is soft: at the paper's own default zeta/s from Eq. (9), the shear+bulk squares in Fig. 3 sit barely above the shear-only circles, and the authors say so. The visually striking recovery toward ideal appears at 10 zeta/s, which is exactly the case where they report the pion spectrum going negative and truncate the pT integral at 0.8 GeV, and where they drop the curve from the differential plots entirely. The second demonstration, Section IV C, uses zeta/s = eta/s = 0.007, an order of magnitude below the eta/s the rest of the paper uses. So the regime where the claim would matter is the regime where delta-f-Bulk is outside its domain. That is a sharper version of the MOM-versus-MH disagreement the paper already flags: the two prescriptions differ qualitatively for v2(pT), and neither is validated where the effect is large. The authors state both limitations in the conclusions, which is to their credit, but the abstract does not carry the qualification. Minor: pions only, no decays, no data comparison, all acknowledged. This deserves a serious referee. The dynamical suppression result stands on its own and the code validation is genuine; the flow-harmonic claim needs to be narrowed to what the parameter space supports.

The abstract (abstract:s1) states generically that bulk viscosity reduces the shear-induced suppression of v_n. The size of that reduction at the paper's own default ζ/s from Eq. (9) is small: in Fig. 3 the shear+bulk squares sit only slightly above the shear-only circles, and the authors themselves note that 'our initial bulk viscosity is so small that even after multiplying by a factor of 10, its peak is still not as large as the minimum of the shear viscosity.' The visually striking compensation — v_n pushed close to the ideal curve — appears only in the 10ζ/s diamonds of Fig. 3. But that is exactly the case where the authors report that the viscous correction to the distribution function drives the spectrum negative, forcing them to truncate the pT integration at 0.8 GeV (Section IV A) and to omit the 10ζ/s curve entirely from the differential plots (Section IV B). So the headline effect is demonstrated in a regime where δf is outside its domain of validity. The second demonstration, Section IV C, sets ζ/s = η/s = 0.007, roughly an order of magnitude below the η/s ≈ 0.08–0.2 used everywhere else and below the value the paper itself cites as the near-perfect-fluid benchmark. The regime where the claim would actually matter — ζ/s comparable to a realistic η/s — is precisely the regime the paper does not show, because δf^Bulk breaks down there. This is a sharper version of the reader's δf-model-dependence point: it is not only that MOM and MH disagree, but that the only parameter points where the effect is large are points where neither prescription is trustworthy.

The reader's weakest_assumption correctly identifies the freeze-out δf^Bulk as the soft spot, and the paper does acknowledge the MOM/MH disagreement in Section II D and the conclusions. I agree that this is the load-bearing issue, so my read does not move the verdict. Where I would sharpen the reader's framing: the problem is not merely that two prescriptions disagree, but that the parameter points where the effect is large are exactly the points where the prescriptions are unreliable. Fig. 3's dramatic 10ζ/s curve is computed with a truncated pT integral because the spectrum goes negative, and Section IV C's equal-viscosity demonstration uses ζ/s = η/s = 0.007, well below the η/s the paper otherwise adopts. The dynamical backbone — Tables III–V showing bulk viscosity suppressing π00 and π12 — is genuinely robust and independent of the freeze-out prescription, and the code validation in Appendix A against TECHQM and the Gubser solution is real, external support. So the paper's physics core stands; what is not established is the quantitative magnitude of the flow-harmonic compensation at realistic viscosities. That is a scope limitation the authors largely acknowledge, and it is the right basis for CONDITIONAL rather than a stronger verdict. I would not adjust the reader's verdict, but I would note that the abstract's unqualified phrasing slightly oversells what the two demonstration regimes actually support, and a sentence in the abstract tying the size of the effect to the ζ/s value used would be more accurate without weakening the paper's contribution. No concern about internal inconsistency or circularity arises; the issue is regime of validity, which routes through correctness risk as the reader already classified it. The forward-looking η/s re-evaluation remark is appropriately hedged with 'may' and is not presented as a result, so I do not treat it as a separate concern. The absence of a data comparison is acknowledged in the conclusions and is a scope choice, not an error, for a methods paper introducing v-USPhydro with bulk viscosity at freeze-out. Overall: a careful, well-validated simulation study whose central qualitative claim is supported, with a quantitative caveat the authors themselves flag and which the reader's CONDITIONAL verdict already captures. I would keep the verdict as is and, if anything, ask only for tighter abstract wording rather than new physics.

If the effect holds at physically relevant values of zeta/s, it would matter: shear-only fits to flow data would systematically bias extracted eta/s, and the interplay between the two viscosities would need to be treated jointly. The code is validated against two independent benchmarks in Appendix A (the TECHQM shear-viscosity test in Fig. 11 and the Gubser SO(3) x SO(1,1) x Z2 semi-analytical Israel-Stewart solution in Fig. 12), which is a genuine strength and gives confidence in the numerical evolution. The dynamical observation that bulk viscosity suppresses the shear stress tensor (Tables III-V) is a robust, delta-f-independent result and is the most convincing part of the paper. The phenomenological reach of the flow-harmonic claim, however, is limited by the delta-f model dependence and by the parameter regimes in which the compensation is demonstrated.

Heavy-ion collisions at RHIC produce a droplet of quark-gluon plasma whose azimuthal flow harmonics $v_n$ are the standard handle on how fluid-like the medium is. Shear viscosity is known to damp those harmonics below their ideal-fluid values, and that damping is what previous work used to pin down $η/s$. This paper asks what happens when bulk viscosity is switched on at the same time, using an event-by-event 2+1 relativistic hydrodynamic code that carries both viscosities through the evolution and through freeze-out. The answer the authors argue for is that bulk viscosity softens shear's damping, pushing the harmonics back toward ideal-hydro values, and that the effect is visible both in the evolution of the shear stress tensor and in the final $v_n$. If that is right, fits that ignored bulk viscosity would have compensated by choosing too small an $η/s$.

What carries the argument. The vehicle is v-USPhydro, a boost-invariant 2+1 Lagrangian code that solves the Israel-Stewart equations for $Π$ and $π^{μν}$ with smoothed-particle hydrodynamics on Monte Carlo Glauber initial conditions, and then applies viscous corrections to the Cooper-Frye freeze-out distribution. The load-bearing mechanism is the indirect coupling between the two dissipative sectors through the fluid velocity: bulk viscosity resists radial expansion, which in turn reduces the diagonal components of the shear stress tensor, which in turn lessens the shear-driven dilution of $v_n$. The paper quantifies this with percentage changes in the mean and variance of $Π$, $π^{00}$, and $π^{12}$ across centrality classes and across early, late, and full freeze-out slices, and then translates it into ratios of viscous to ideal $v_n$.

Core claim. The paper's central claim is that bulk and shear viscosity do not act independently on the flow harmonics: adding bulk viscosity reduces the shear-viscosity-induced suppression of $v_n$ and brings the harmonics closer to their ideal-hydrodynamic values. The authors support this in two ways. First, in the hydrodynamic evolution itself, they show that switching on bulk viscosity lowers the mean value of the $π^{00}$ component of the shear stress tensor (by roughly 18 to 45 percent depending on centrality, and by less at early times) and shrinks its event-by-event variance by around 20 percent, while $π^{12}$ is only mildly affected. Second, in the final spectra, the viscous-to-ideal ratio of the integrated $v_n$ sits closer to unity for shear-plus-bulk than for shear alone, and the same ordering appears in the differential $v_2(p_T)$ through $v_5(p_T)$. The authors emphasize that the coupling is indirect: $Π$ and $π^{μν}$ do not appear in each other's equations of motion in their setup, so the effect runs through the bulk-viscosity-modified flow velocity, and it grows the longer the system expands. They also show a case with $ζ/s = η/s = 0.007$, where the integrated harmonics return almost exactly to the ideal values, with $v_5$ the exception.

The premise it rests on. The differential-flow half of the claim rests on how the bulk viscous correction to the freeze-out particle distribution is modeled: the paper's own two prescriptions, Monnai-Hirano and the moments method, give visibly different $v_2(p_T)$, with one nearly reproducing the ideal curve and the other suppressing it, so the claim holds only if at least one of them is close to the true correction.

Extractions of $η/s$ from hydrodynamic fits to RHIC flow data that omitted bulk viscosity would be biased low, since bulk viscosity supplies part of the flow that was attributed to a smaller shear viscosity.

The bias should grow with system lifetime, so the same effect should be stronger at LHC energies and at lower freeze-out temperatures, where the fluid spends longer in the plasma phase.

Higher-order harmonics are the least able to be fully restored by bulk viscosity, so $v_4$ and $v_5$ are the more discriminating observables for separating the two transport coefficients.

In central collisions a large $ζ/s$ compresses the gap between $v_2$ and $v_3$, which is a qualitative shape change in the harmonic spectrum rather than a uniform rescaling.

If bulk viscosity can mimic a smaller shear viscosity in the integrated $v_n$, then a joint $η/s$–$ζ/s$ extraction from the same data should show a valley of acceptable fits rather than a point, and the shape of that valley is itself a testable prediction.

The authors' observation that the effect strengthens at late times suggests a cheap diagnostic: comparing flow harmonics on freeze-out slices at different temperatures would reveal whether the compensation is dynamical or merely a freeze-out bookkeeping artifact.

Because the coupling here is indirect, the size of the effect is a direct measure of how much the bulk pressure distorts the velocity field; a calculation with the direct $Π$–$π^{μν}$ coupling terms switched on would tell us whether the effect found here is a floor or a ceiling.

The claim also implies that the commonly quoted near-perfect-fluid value of $η/s$ is not by itself evidence about the plasma, since a larger $η/s$ paired with a non-negligible $ζ/s$ could reproduce the same spectra.

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review