---
title: Collective-Flow Nuclear Shape Imaging
url: https://www.emergentmind.com/topics/collective-flow-assisted-nuclear-shape-imaging-method
type: topic
---

# Collective-Flow Nuclear Shape Imaging

Searching arXiv for the core paper and closely related work to support the article.
Searching for the 2025 nuclear-structure critique and the 2024/2025 STAR imaging papers.
Collective-flow-assisted nuclear shape imaging is a program in which ultrarelativistic nuclear collisions are used to constrain nuclear deformation parameters by exploiting the mapping from initial-state geometry to final-state collective flow. In this framework, deformation of the colliding nuclei modifies the initial transverse energy density of the quark–gluon plasma, and the subsequent hydrodynamic expansion converts these spatial anisotropies into measurable momentum anisotropies \(v_n\), mean-transverse-momentum fluctuations, and mixed cumulants. The method has been formulated as an “imaging-by-smashing” strategy for extracting quantities such as the quadrupole deformation \(\beta_2\), triaxiality \(\gamma\), and, in extensions, octupole \(\beta_3\) and hexadecapole \(\beta_4\) contributions [2401.06625]. Its status is, however, intrinsically dual: collider flow is a well-founded probe of initial geometry, yet claims of direct laboratory-frame “precision imaging” of even-even \(J^\pi=0^+\) ground states have been criticized as conceptually overstated, with robust inference requiring explicit attention to higher-body correlations and to pre-existing low-energy nuclear-structure constraints [2507.05208].

## 1. Historical emergence and domain of application

The method emerged from a convergence of relativistic heavy-ion phenomenology and nuclear-structure modeling. Event-shape engineering established that event-by-event flow fluctuations can be used to select collision ensembles with targeted initial anisotropies through the reduced flow vector \(q_n\), thereby providing a practical route to geometry-conditioned measurements [1208.4563]. In parallel, studies of small systems argued that collective flow in \(p+A\), \(d+A\), and light-ion collisions retains sensitivity to projectile geometry, including deuteron deformation, three-body triangularity, and \(\alpha\)-cluster substructure [1407.6478].

A more explicit precursor of nuclear shape imaging was the proposal to use ultrarelativistic \(^{12}\mathrm{C}\)-\(^{208}\mathrm{Pb}\) collisions to detect triangular \(\alpha\)-cluster correlations through the transmutation of initial triangularity into \(v_3\) [1407.8495]. Subsequent AMPT studies of \(^{12}\mathrm{C}+^{197}\mathrm{Au}\) argued that the ratio \(v_3/v_2\) can distinguish chain, triangular, and Woods–Saxon-like \(^{12}\mathrm{C}\) configurations, especially in the most-central events [1702.02507]. At lower energies, EQMD simulations of \(^{16}\mathrm{O}+^{16}\mathrm{O}\) around \(40\)–\(60\) MeV/nucleon found that directed flow \(v_1\) of free protons is more sensitive than elliptic flow to four-\(\alpha\) cluster geometries, establishing an analogous shape-imaging logic outside the QGP regime [1803.07978].

The modern formulation was articulated for ultrarelativistic heavy-ion collisions at RHIC and the LHC, with benchmark applications to \(^{238}\mathrm{U}+^{238}\mathrm{U}\) versus \(^{197}\mathrm{Au}+^{197}\mathrm{Au}\), to the Ru/Zr isobars, to Xe+Xe, and to proposed light-ion systems such as Ne on Pb in fixed-target configurations [2409.09599]. In this expanded usage, the targeted shape descriptors are primarily \(\beta_2\) and \(\gamma\), with increasing attention to \(\beta_3\), \(\beta_4\), neutron-skin effects, and cluster-driven non-ellipsoidal structure [2507.05208].

## 2. Geometric and hydrodynamic framework

Most implementations start from a deformed Woods–Saxon or Fermi density in which the nuclear surface is expanded in spherical harmonics. In its axial form, the radius is written
\[
R(\theta)=R_0\left[1+\sum_{\lambda\ge 2}\beta_\lambda Y_{\lambda 0}(\theta)\right],
\]
with density
\[
\rho(r,\theta)=\frac{\rho_0}{1+\exp((r-R(\theta))/a)}.
\]
For quadrupole triaxiality one uses
\[
R(\theta,\phi)=R_0\left[1+\beta_2\left(\cos\gamma\,Y_{20}(\theta)+\sin\gamma\,\frac{Y_{22}(\theta,\phi)+Y_{2,-2}(\theta,\phi)}{\sqrt{2}}\right)+\ldots\right].
\]
A common low-energy comparison is the mapping
\[
Q_0 \approx \frac{3}{\sqrt{5\pi}}\,Z e R_0^2 \beta_2,
\qquad
R_0 \approx r_0 A^{1/3},
\]
and typical magnitudes quoted for well-deformed nuclei are \(\beta_2 \sim 0.2\)–\(0.3\), \(\beta_3\) up to \(\sim 0.1\) in pear-shaped regions, and non-negligible \(\beta_4\) in actinides [2507.05208].

Event by event, initial anisotropies are quantified by participant eccentricities,
\[
\epsilon_n e^{i n \Phi_n} = - \frac{\langle r^n e^{i n \phi}\rangle}{\langle r^n\rangle},
\]
with density-weighted averages in the transverse plane. Static quadrupole structure contributes primarily to \(\epsilon_2\), whereas \(\epsilon_3\) is often fluctuation-dominated at high multiplicity. Hydrodynamics then supplies the leading-order response
\[
v_n \simeq \kappa_n \epsilon_n,
\]
with \(\kappa_n\) determined by system size, multiplicity, viscosities, and freeze-out conditions. Higher harmonics receive nonlinear contributions such as
\[
v_4 \simeq \kappa_4 \epsilon_4 + \chi_{422} (\epsilon_2)^2,
\qquad
v_5 \simeq \kappa_5 \epsilon_5 + \chi_{523}\epsilon_2\epsilon_3.
\]
This structure makes quadrupole imaging comparatively direct through \(v_2\), whereas triaxiality, octupole, and hexadecapole inference is entangled with nonlinear mode coupling and model dependence [2507.05208].

In the ultra-central uranium program, STAR further emphasized mixed size–shape observables. In that context, hydrodynamic simulations yielded the parametric forms
\[
\langle v_2^2\rangle = a_1 + b_1 \beta_2^2,\qquad
\langle (\delta p_T)^2\rangle = a_2 + b_2 \beta_2^2,\qquad
\langle v_2^2 \delta p_T\rangle = a_3 - b_3 \beta_2^3 \cos(3\gamma),
\]
which underlie the extraction of both \(\beta_2\) and \(\gamma\) from central \(^{238}\mathrm{U}+^{238}\mathrm{U}\) data [2409.09599].

## 3. Observables and inference strategy

The basic observables are integrated and differential flow harmonics \(v_n\), their multiparticle cumulants \(v_n\{2\}\), \(v_n\{4\}\), \(v_n\{6\}\), symmetric cumulants
\[
SC(n,m)=\langle v_n^2 v_m^2\rangle-\langle v_n^2\rangle\langle v_m^2\rangle,
\]
event-plane correlations, and event-by-event mean-transverse-momentum fluctuations. In the STAR implementation, the shape-sensitive set for uranium imaging included \(\langle v_2^2\rangle\), \(\langle(\delta p_T)^2\rangle\), and \(\langle v_2^2\delta p_T\rangle\), while the isobar program used ratios of \(v_2\) and \(v_3\) versus multiplicity to constrain \(\beta_2\), \(\beta_3\), \(R_0\), and \(a\) [2409.09599].

A central methodological device is the use of matched-system ratios, such as U/Au or Ru/Zr, to suppress common QGP-response uncertainties. In the STAR isobar analysis,
\[
R_{\mathcal O}\equiv \mathcal O_{\mathrm{Ru}}/\mathcal O_{\mathrm{Zr}}
\approx 1 + c_1 \Delta\beta_2^2 + c_2 \Delta\beta_3^2 + c_3 \Delta R_0 + c_4 \Delta a,
\]
while in the uranium case the ratio of U+U to Au+Au observables isolates deformation-driven changes in central collisions [2409.09599]. A later STAR analysis extended the same logic to \(v_3\), \(v_4\), and their covariances with \([p_T]\), arguing that \(v_3\)-based ratios provide sensitivity to a possible octupole component of \(^{238}\mathrm{U}\) [2506.17785].

Nonflow suppression is handled primarily through multiparticle cumulants, \(\eta\) gaps, and subevent methods. Event-shape engineering provides a second layer of selection: by binning on the reduced flow vector
\[
q_n = Q_n/\sqrt{M},
\qquad
Q_n = \sqrt{Q_{n,x}^2+Q_{n,y}^2},
\]
one enriches classes with larger or smaller initial anisotropy, thereby increasing the resolving power of geometry-conditioned measurements [1208.4563]. The critique literature argues that, for credible deformation inference, this observable set should be embedded in a global, multi-observable, multi-model fit with priors anchored to low-energy data, uncertainty propagation across initial conditions, hydrodynamics, and hadronic afterburners, and explicit reporting of model discrepancy [2507.05208].

## 4. Benchmark systems and reported results

The method has been applied or proposed across a broad set of systems. The table summarizes the principal benchmarks.

| System or comparison | Main observables | Targeted structure |
|---|---|---|
| \(^{238}\mathrm{U}+^{238}\mathrm{U}\) vs \(^{197}\mathrm{Au}+^{197}\mathrm{Au}\) | \(\langle v_2^2\rangle\), \(\langle(\delta p_T)^2\rangle\), \(\langle v_2^2\delta p_T\rangle\) | \(\beta_2\), \(\gamma\), with extensions to \(\beta_3\), \(\beta_4\) |
| \(^{96}\mathrm{Ru}+^{96}\mathrm{Ru}\) vs \(^{96}\mathrm{Zr}+^{96}\mathrm{Zr}\) | \(v_2\), \(v_3\), multiplicity dependence, ratios | \(\beta_2\), \(\beta_3\), neutron skin and radius differences |
| Xe+Xe | flow systematics | triaxiality claims |
| \(^{12}\mathrm{C}\)+heavy target | \(v_3/v_2\), multiplicity trends | triangular \(\alpha\)-cluster correlations |
| Pb+Ne vs Pb+Ar or Pb+O | \(v_2\), HBT anisotropies | light-nucleus deformation and clustering |
| \(^{48}\mathrm{Ca}+^{48}\mathrm{Ca}\) vs \(^{40}\mathrm{Ca}+^{40}\mathrm{Ca}\) | \(v_3\), \(v_4\), \([p_T]\) cumulants | neutron-skin thickness |

For \(^{238}\mathrm{U}\), STAR’s proceedings reported, from IP-Glasma+MUSIC+UrQMD, \(\beta_{2,U}=0.297\pm0.015\) and \(\gamma_U=8.5^\circ\pm4.8^\circ\), and after combination with Trajectum, \(\beta_{2,U}=0.286\pm0.025\) and \(\gamma_U=8.7^\circ\pm4.5^\circ\) [2409.09599]. A later STAR analysis based on anisotropic and radial flow reported \(\beta_{2,U}=0.300\pm0.016\) and \(\gamma_U=8.3^\circ\pm4.7^\circ\) from the joint use of \(R_{(\delta p_T)^2}\) and \(R_{v_2^2\delta p_T}\), while \(R_{v_2^2}\) alone yielded a lower bound \(\beta_{2,U}=0.240\pm0.018\) [2506.17785]. Related RHIC proceedings quoted a combined result \(\beta_{2,U}=0.286\pm0.025\), \(\gamma_U=8.7^\circ\pm4.5^\circ\), emphasizing a large deformation with a small but nonzero departure from axial symmetry [2510.02689].

In the isobar system, STAR reported \(\beta_{2,\mathrm{Ru}}=0.16\pm0.02\), \(\beta_{3,\mathrm{Zr}}=0.20\pm0.02\), and neutron-distribution differences \(\Delta R_0\approx0.07\) fm and \(\Delta a\approx-0.06\) fm for Ru–Zr, inferred from \(v_2\) and \(v_3\) ratio systematics [2409.09599]. For octupole deformation of uranium, hydrodynamic calculations predicted that in \(0\)–\(2\%\) ultra-central events the ratio \(R_{v_3^2}\) should follow a linear dependence on \(\beta_3^2\), with \(R_{v_3^2}\approx1.05\) to \(1.12\) for \(\beta_{3,U}=0.078\)–\(0.10\), while \(R_{v_3^2\delta p_T}\) should show a characteristic suppression [2504.15245].

The method has also migrated to fixed-target LHC kinematics. In PbNe and PbAr collisions at \(\sqrt{s_{NN}}=70.9\) GeV, LHCb measured a significantly larger elliptic flow in central PbNe than in PbAr, with
\[
v_2(\mathrm{PbNe})/v_2(\mathrm{PbAr}) = 1.40 \pm 0.06
\]
in the \(0\)–\(3\%\) centrality bin, interpreting the result as qualitative confirmation of the distinctive bowling-pin shape of \(^{20}\mathrm{Ne}\) [2509.12399]. A complementary AMPT study of Pb+Ne and Pb+O found that the freeze-out eccentricity
\[
\epsilon_F = 2R^2_{\mathrm{side},2}/R^2_{\mathrm{side},0}
\]
is significantly larger for NLEFT-like \(^{20}\mathrm{Ne}\) than for Woods–Saxon neon or either oxygen configuration, indicating that azimuthally sensitive femtoscopy can function as a geometric discriminator alongside flow [2506.05849].

For clustered light nuclei, GLISSANDO and AMPT studies of \(^{12}\mathrm{C}\) with heavy targets showed that triangular configurations produce increasing \(\langle\epsilon_3\rangle\) and hence increasing \(v_3\) with multiplicity, while chain-like or smooth configurations favor larger \(v_2\); the discriminating observable is the ratio \(v_3/v_2\), which satisfies \(R_{\mathrm{Triangle}} > R_{\mathrm{WS}} > R_{\mathrm{Chain}}\) in central \(^{12}\mathrm{C}+^{197}\mathrm{Au}\) collisions [1702.02507].

## 5. Nuclear-structure interpretation and conceptual controversy

The central conceptual dispute concerns the meaning of “shape” in an even-even \(J^\pi=0^+\) ground state. The critique literature stresses that such a state is rotationally invariant in the laboratory frame and therefore spherical in that frame; it does not possess a fixed orientation or an “instantaneous shape” that can be directly photographed before impact [2507.05208]. On this view, deformed Woods–Saxon sampling of one-body densities is a practical phenomenological ansatz, not a literal representation of the laboratory-frame many-body wave function.

The same critique sharpens the information-theoretic content of the inference problem. According to that analysis, axial shape information in a \(0^+\) state resides in two-body densities, or conditional probabilities for finding one nucleon given another, whereas triaxiality requires three-body densities. Standard initial-condition generators, which sample nucleons independently from one-body deformed densities, do not encode this correlation structure. This is the basis for the claim that flow–geometry phenomenology is sound, but laboratory-frame “precision imaging” language is not [2507.05208].

The practical consequence is not that collider data are irrelevant, but that they must be interpreted as model-dependent constraints cross-validated against low-energy structure. Decades of spectroscopy, static quadrupole moments in odd-\(A\) nuclei, \(B(E2/E3/E4)\) values, charge radii, and electron scattering already define deformation systematics with high precision; collider analyses should therefore incorporate these as priors and benchmarks rather than attempt unconstrained extraction in isolation [2507.05208].

A later microscopic development moved the discussion closer to the critique’s requirements by formulating shape information directly in terms of transverse two-body densities. For the reduced azimuthal correlator
\[
C(r_\perp,\Delta\phi)=\frac{\rho^{(2)}_\perp(r_\perp,\Delta\phi)}{\rho^{(1)}_\perp(r_\perp)^2}-1
= a_0(r_\perp)+\sum_{n\ge1} a_n(r_\perp)\cos(n\Delta\phi),
\]
the harmonic coefficients \(a_n\) furnish a pair-density fingerprint of intrinsic collectivity, and the associated moments define deformation measures \(\mathfrak B_2\simeq\beta_2\) and \(\mathfrak B_3\simeq1.06\,\beta_3\) in the small-deformation limit [2512.18926]. This suggests a microscopic route for reconciling collider observables with the requirement that axial and octupole information be formulated through correlated many-body densities rather than through one-body snapshots alone.

## 6. Systematics, best practice, and current directions

The method is highly sensitive to systematic uncertainty in both initial conditions and medium response. The main confounders include the choice of initial-condition model, such as MC-Glauber versus IP-Glasma; nucleon–nucleon correlations beyond independent sampling; subnucleonic fluctuations and finite nucleon size; neutron–proton density differences and surface diffuseness; multiplicity-based centrality calibration; residual nonflow from jets and resonance decays; transport coefficients \(\eta/s\) and \(\zeta/s\); pre-equilibrium dynamics; hadronic afterburners; and the random orientation of the colliding nuclei [2507.05208]. A separate hydrodynamic study of Au+Au, Cu+Au, and O+O emphasized that while \(v_2\{4\}/v_2\{2\}\) tracks \(\epsilon_2\{4\}/\epsilon_2\{2\}\) in large systems over suitable centralities, hadronization, nonlinear response, and hadronic rescattering become much more intrusive in small systems [2402.02348].

Best practice therefore consists of anchoring nuclear inputs to validated low-energy structure data, performing multi-observable global fits across centrality and collision species, scanning multiple initial-condition and hydrodynamic frameworks, and reporting credible intervals that include model discrepancy [2507.05208]. Multi-particle cumulants and matched-system ratios are preferred over absolute two-particle observables, and consistent centrality calibration across species is essential. For claims involving triaxiality, the critique literature further recommends explicit tests of sensitivity to three-body correlations, since current one-body initialization practices do not encode them [2507.05208].

Several current directions extend the method beyond quadrupole imaging. Quantum constraints on collective fluctuations have been incorporated in ultra-central Pb+Pb through an energy-weighted sum-rule restriction on quadrupole variance, which reduces \(\epsilon_2/\epsilon_3\) and resolves the long-standing \(v_2\)-to-\(v_3\) puzzle in central Pb collisions [2008.07304]. Femtoscopy has been proposed as an additional imaging channel, with azimuthally sensitive HBT radii supplying direct information on freeze-out source eccentricity [2506.05849]. Calcium-isotope studies within AMPT indicate that neutron-skin thickness in \(^{48}\mathrm{Ca}\) affects \(v_3\), \(v_4\), and especially the variance and skewness of \([p_T]\), suggesting an “imaging-by-smashing” route to \(\Delta r_{np}\) [2512.00114]. At the microscopic end, harmonic analysis of two-body correlations in ab initio calculations of \(^{20}\mathrm{Ne}\) and \(^{16}\mathrm{O}\) proposes a bridge from collider flow observables to the intrinsic many-body correlation structure itself [2512.18926].

Collective-flow-assisted nuclear shape imaging is thus best understood not as a literal camera for laboratory-frame ground-state shapes, but as a constrained inverse problem linking nuclear densities, many-body correlations, initial-state geometry, and final-state collective observables. In that more precise sense, it has become a distinct interface between nuclear structure and high-energy QCD phenomenology: strongest when used comparatively, statistically, and in conjunction with low-energy information, and most credible when its limitations are made as explicit as its geometric sensitivity.

Source: https://www.emergentmind.com/topics/collective-flow-assisted-nuclear-shape-imaging-method