WorksPhysics and Astronomy
Heavy ion collisions from of 62.4 GeV down to 4 GeV in the EPOS4 framework
Abstract
The EPOS4 project is an attempt to construct a realistic model for describing relativistic collisions of different systems, from proton-proton () to nucleus-nucleus (), at energies from several TeV per nucleon down to several GeV. We argue that a parallel scattering formalism (as in EPOS4) is relevant for primary scatterings in AA collisions above 4 GeV, whereas sequential scattering (cascade) is appropriate below. We present briefly the basic elements of EPOS4, and then investigate heavy ion collisions from 62.4 GeV down to 4 GeV, to understand how physics changes with energy, studying in particular the disappearance of the fluid component at low energies.
Introduction
Scatterings of two nuclei are fundamentally different at high energies (several TeV per nucleon) compared to low energies (several GeV). At high energies, one can clearly separate “primary scatterings”, happening instantaneously (at ) and “secondary scatterings” of the primary particles (for ), which may lead to thermalized matter (quark gluon plasma, QGP) and hadronic rescattering after hadronization. At low energies, one cannot separate these two stages anymore, and at very low energies ( GeV), heavy ion collisions can be treated by purely sequential hadronic scatterings, also referred to as hadronic cascade [1, 2, 3, 4, 5, 6, 7, 8]. So the physics picture changes drastically when going from several TeV down to a few GeV, and we try to understand these changes. We will start at high energies and we will see how (and when) the “high energy features” disappear when lowering the collision energies.
Most important for the discussion of primary scatterings at very high energies (several TeV per nucleon) is the observation that nucleon-nucleon scatterings must happen in parallel, and not sequentially, based on very elementary considerations concerning time scales. To take this “parallel scattering scenario” into account, EPOS4 brings together ancient knowledge about S-matrix theory (to deal with parallel scatterings) and modern concepts of perturbative QCD and saturation, going much beyond the usual factorization approach; see Refs. [9, 10, 11, 12].
Let us look more in detail at the relevant time scales (see [11] for details). The particle (hadron) formation time has to be compared to the collision time (the duration of an collision) and the interaction time (time between two nucleon-nucleon interactions). We define the high energy threshold (in the sense of energy per nucleon-nucleon pair) by the identity
and the low energy threshold by the identity
Considering central rapidity hadrons (), a formation time , and a big nucleus with , we get [11]
Beyond particle production starts only after the two nuclei have passed through each other, which means all the nucleon-nucleon collisions should happen in parallel, instantaneously, there is no time sequence. Below a hadronic cascade is appropriate. Between the two thresholds, one needs some “partially parallel scattering scenario”, which is not yet implemented in EPOS4. We will employ the full "parallel scattering scenario" down to lowest energies, and we will investigate where precisely and how it breaks down. Since PbPb at 5.02 ATeV and AuAu at 200 GeV have already been discussed in [11], we will focus in this paper on energies below 200 GeV.
In the overview [9] and in detail in [10, 11] it is shown how such a “parallel scattering scheme” for primary scatterings can be constructed, based on S-matrix theory, which we will sketch very briefly in the following.
An early realization is the Gribov-Regge (GR) approach [13, 14, 15, 16] for and scatterings. This S-matrix approach has a modular structure, it is based on so-called “cut Pomerons”, representing elementary parton-parton scatterings, with the associated mathematical object , with being the Fourier transform (with respect to the momentum transfer) of the corresponding -matrix, divided by , with referring to the Mandelstam variable. This so-called “impact parameter representation” with with an impact parameter makes formulas simpler. Although the GR approach is an excellent tool to deal with parallel scatterings, a serious drawback is the fact that the energy-momentum sharing between the multiple scatterings is not taken care of. And obviously GR has no answer to the question of how to connect the cut Pomeron expression and the corresponding QCD expression for parton-parton scattering.
In [17], a possible solution has been proposed, by taking into account energy-momentum sharing (let us call this approach “GR+”) and based on the hypothesis “ is equal to ”, where the latter is essentially a cut parton ladder based on DGLAP parton evolutions [15, 18, 19]. A detailed discussion about the calculation of can be found in [10]. Unfortunately, it turned out that implementing energy-momentum sharing has a very negative side effect: it ruins seriously elementary geometric properties such as binary scaling in scattering. In EPOS4, the first step towards a solution of the problem is a detailed understanding of what causes the problem and that it is fundamental, and not just a wrong parameter choice. Using
leads unavoidably to contradictions. In a second step, a solution could be presented. Let us look at the arguments of and : Both depend on (not written explicitly) and on the lightcone momenta of the external legs and . But most importantly, also depends on some low virtuality cutoff for the parton evolution, and this cutoff is now considered to not anymore be simply a constant, but a dynamical variable, named saturation scale . So we have . The fundamental relation between and is now
with and with (being crucial)
with the so-called connection number counting the number of Pomerons being connected to the same projectile and target nucleon as the given Pomeron. The quantity is the deformation of the distribution of (the Pomeron’s energy squared) in case of compared to the case . This deformation (a consequence of several Pomerons competing for energy sharing) destroys factorization and binary scaling unless one uses eqs. (5,6). Finally, is a normalization constant.
The primary scattering will produce “parton-ladders”. The link between primary and secondary scatterings is the “core-corona procedure”: The parton ladders are treated as classical relativistic (kinky) strings. So in general, we have a large number of (partly overlapping) strings. Based on the momenta and the density of string segments (referred to as prehadrons in the following), one separates at some early proper time the core (going to be treated as fluid) from the corona (escaping hadrons, including jet hadrons).
We will first discuss in Sec. 2 about particle production and the role of core, corona, and remnants for different energies, down to 4 GeV.
Then in Secs. 3, 4, and 5, we will show very detailed results from 62.4 GeV down to 7.7 GeV, considering the transverse momentum dependencies of yields and flow harmonics
for all energies,
for all possible centrality choices, and
for all hadron species where the corresponding data exist, namely pions, kaons, (anti)protons, but also hyperons (, , ) and mesons.
This is the most complete collection of model/data comparisons concerning gold-gold collisions in this energy domain, in particular concerning differential yields ( dependencies).
Over the last decade, a few other models involving a fluid dynamic picture have been developed for the energy range GeV, where the UrQMD cascade [20], a decelerating string picture [21, 22] and a 3D extension to Monte Carlo Glauber [23] have been employed for the initial state. These models succeeded in reproducing the collision energy dependence of the shape and the magnitude of the rapidity density of hadrons, and (except [21, 22]) of -integrated yields and elliptic flow coefficients. In Ref. [20], transverse momentum spectra of several hadron species at selected centralities and energies were shown.
The role of core, corona, and remnants for different energies
Multiple scattering diagrams (representing the primary scatterings) are the origin of particle production, which means first the production of so-called prehadrons, having the quantum numbers of hadrons, but not necessarily being final. Prehadrons originate from Pomerons (via kinky strings) or from remnants. A detailed description can be found in section 4 of [10], where we also discuss the different types of Pomerons and their relation with pQCD, including technical details for the pQCD computations.
The pQCD part (parton ladders) is a crucial element of a Pomeron at LHC energies. But with decreasing energy, it becomes more and more likely that these Pomerons are replaced by purely soft ones, see Sec. 3 of [12]. At 200 GeV, the relative weight of “normal” compared to soft Pomerons is roughly 1:1, at even lower energies, the soft dominates. Also the Pomerons get less energetic, producing fewer particles.
Based on these abovementioned prehadrons, we employ a so-called core-corona procedure (see [24], for a more recent discussion in the EPOS4 framework see [12]), to distinguish core from corona particles, at some given (early) proper-time . The core prehadrons constitute “bulk matter” and will be treated via hydrodynamics. The corona prehadrons become simply hadrons and propagate with reduced energy (due to some energy loss). Corona particles are either very energetic (then they move out even from the center), or they are close to the surface, or we have a combination of both.
We will try to understand the relative importance of the core part and of the fraction coming from remnant decay. In Figure 1, we show results for central (0-5%) PbPb collisions at 5.02 TeV, and central (0-5%) AuAu collisions from 200 down to 4 GeV (per nucleon). We plot four different curves: all prehadrons (red full), all core prehadrons (red dotted), prehadrons from remnant decay (blue thin full), and core prehadrons from remnant decay (blue thin dashed). Looking at the results for PbPb at 5.02 TeV, we observe that almost all prehadrons are core prehadrons, so the core dominates. Prehadrons from remnants are preferentially produced at large rapidities, but also here almost all are core prehadrons. Going down in energy, starting already at 200 GeV, we see that the core still dominates around , but at lage values of , the core contribution drops dramatically. We also observe that the remnant contributions become more and more important, and very dominant below 20 GeV. Nevertheless, close to , the core remains dominant - down to 7.7 GeV. There is actually little change from 39 down to 7.7 GeV, but things change from 7.7 to 4 GeV: the core part decreases strongly.

Figure 1. The prehadron yield as a function of space-time rapidity (), with being the time and the longitudinal coordinate), for central collisions at different energies. The curves refer to all prehadrons (red full), all core prehadrons (red dotted), prehadrons from remnant decay (blue thin full), and core prehadrons from remnant decay (blue thin dashed).
In EPOS, is defined as the time corresponding to maximal overlap of the colliding nuclei. Fluidization takes place at a fixed proper time , which we later
call initial proper time. We set for all collision energies except for 7.7 GeV, where is used. Such setting ensures that the fluid stage starts after all the primary nucleon-nucleon scatterings have taken place. Having identified core pre-hadrons, we compute the corresponding energy-momentum tensor and the flavor flow vector at some position at initial proper time as
and
with being the net flavor content and the four-momentum of prehadron . The function is some Gaussian smoothing kernel (see [12]). The Lorentz transformation into the comoving frame provides the energy density and the flow velocity components , which will be used as the initial condition for a hydrodynamical evo lution [25]. This is based on the hypothesis that equilibration happens rapidly and affects essentially the space components of the energy-momentum tensor. In Figure 2, we plot the energy density at the initial proper-time as

Figure 2. Energy density at the initial proper-time as a function of the transverse coordinate for central collisions at different energies. The blue dashed line is the freeze-out energy density.
a function of the transverse coordinate in collisions at different energies. We also indicate as a blue dashed line the freeze-out energy density . At the highest energies, the energy density is way above the critical value , but with decreasing beam energy, the energy density approaches this value. At 7.7 GeV, the energy density is slightly above , and finally at 4 GeV, there is no fluid created.
As a next step, a viscous hydrodynamic expansion follows. Starting from the initial proper time , the core part of the system evolves according to the equations of relativistic viscous hydrodynamics [25, 26], where we use presently . The “core-matter” hadronizes on some hyper-surface defined by a constant energy density (presently 0.57 GeV/fm). In earlier versions [27], we used a so-called Cooper-Frye procedure. This is problematic in particular for small systems: not only energy and flavor conservation become important, but we also encounter problems due to the fact that we get small “droplets” with huge baryon chemical potential, with strange results for heavy baryons. In EPOS4, we will systematically use microcanonical hadronization, with major technical improvements compared to earlier versions (see [12]).
In the following, we want to study core and corona contributions to hadron production. We will distinguish:
(A) The “core+corona” contribution: primary interactions (S-matrix approach for parallel scatterings), plus core-corona separation, hydrodynamic evolution and microcanonical hadronization of the core, but without hadronic rescattering.
(B) The “core” contribution: as (A), but considering only core particles.
(C) The “corona” contribution: as (A), but considering only corona particles.
(D) The “full” EPOS4 scheme: as (A), but in addition hadronic rescattering.
In cases (A), (B), and (C) , we need to exclude the hadronic afterburner, because the latter affects both core and corona particles, so in the full approach, the core and corona contributions are not visible anymore. In the following, we will focus on energies below 200 GeV, since the corresponding plots for PbPb at 5.02 ATeV and AuAu at 200 GeV have already been discussed in [11].
In Figure 3, Figure 4, and Figure 5, we show ratios / core+corona versus , with being the corona contribution (blue), the core (green), and the full contribution (red), for AuAu collisions at different energies. In Figure 3, we show results at 62.4 GeV, for (from top to bottom) pions (), kaons (), protons ( and ), and lambdas ( and ). The four columns represent four different centrality classes, namely 0-5%, 20-40%, 60-80%, 80-100%. Looking at the green (core) and blue (corona) curves, we observe that the core contribution increases from peripheral to central collisions. Concerning the dependence, we observe a maximum of the green core curves around 1-2 GeV/c, at very low the core contribution goes down, so even at very small values the corona contributes. At higher energies (see [11]), one observes a crossing of the green core and the blue corona curves (core = corona) between around 2 GeV/c (mesons, peripheral) and 5 GeV/c (baryons, central), and for larger values the corona contribution dominates clearly. Here, the crossing is absent. There are two reasons. The spectra are softer compared to higher energies, and around 3-6 GeV/c, the core and corona curves are almost parallel. And since high becomes rare, we cannot reach very high in the Monte Carlo simulations.
Figure 3. The X / core+coron ratio, with X being the corona contribution (blue), the core (green), and the full contribution (red), for 4 centrality classes and four different particle species, for AuAu at 62.4 GeV.

Figure 4. Same as Figure 3, but for 39 GeV, and we consider , , , and hadrons.
Figure 5. Same as Figure 3, but for 19.6 GeV, and we consider , , , and hadrons.
The red curve, full over core+corona, represents the effect of the hadronic cascade in the full case. The pions are not much affected, but for kaons and even more for protons and lambdas, rescattering makes the spectra harder. We should keep in mind that rescattering involves particles from fluid hadronization, but also corona particles from hard processes. Concerning the baryons, rescattering reduces (considerably) low yields, due to baryon-antibaryon annihilation.
In Figure 4 and Figure 5, we show results for AuAu collisions at 39 and 19.6 GeV. The high particles are getting rare, we hardly get (for simulations with reasonable CPU times) beyond 5 GeV/c. Another “low energy effect”: the difference between particles and antiparticles becomes more and more important, this is why we consider separately , , , and .
The biggest effect can be seen when comparing the effect of hadronic rescattering (red curves) for protons and antiprotons: with decreasing energy, the proton curve goes slightly up, and the antiproton curve goes significantly down (they are rare, and most of them are annihilated). Interesting observation: also at lower energies, the core ratios (green curves) get down at low , but stay close to unity at intermediate .
Corresponding plots for AuAu collisions at 27 GeV, 11.5 GeV, and 7.7 GeV can be found in the appendix.
When discussing these results at low energies (7.7–19.6 GeV), we should keep in mind our discussion in § 1, where we estimated that our parallel scattering scenario should be valid beyond 24 GeV per nucleon, and below we have to take into account the fact that particle production starts before the two nuclei have passed through each other (not yet done in EPOS4).
But anyway, in the following sections, we will explore to what extent the model “works” (and can explain data), and where it fails, and what we can learn from that.
Results concerning spectra
In this section, we show simulation results compared to data. We will not add too many comments to each curve, the main purpose is to check if the concepts discussed in the previous sections give a coherent picture (and reproduce the data) or not.
The number of plots is huge, and it would be tempting to show only a selection of results and provide a simple message. And it would be tempting to optimize parameters to have “nice” plots. But the EPOS4 philosophy is different: It is an attempt to
have a single unique approach, with version number (here 4.0.0), and
make a maximum of tests, as complete as possible, for all kinds of observables and at various energies.
Modeling heavy ion collisions is complex, consisting of several stages, with considerable uncertainties at each step. And there is nothing like the simple “smoking gun” to prove or disprove certain concepts. Flow harmonics for example are popular observables showing the effect of flow asymmetries, but the flow affects also the spectra, more or less strongly depending on the mass of the particle. So rather than looking at few selected curves and expecting precision results, the idea is here more to get a global view and to see to what extent a given theoretical scenario can provide an overall good description of a very large set of data.
We will see later, that spectra and elliptical flow , although both are affected by flow, they are at a different degree sensitive to particular details of the model, so we get complementary information, and therefore both observables should be studied.
This “global view” strategy requires a huge amount of tests, and therefore the“tuning” of parameters and options is far from being optimized, But we expect reasonable agreement in the energy range (beyond 24 GeV) where the parallel scattering approach is applicable.
In the following we discuss spectra. We will start with the highest energy, and then move down till 7.7 GeV per nucleon. To go this way is convenient in the sense that from the theory point of view, the high energy case contains in principle everything, we do not need to add “features” at low energies, simply certain phenomena “die out”.
We will focus on energies from 39 GeV down to 7.7 GeV, the corresponding plots for PbPb at 5.02 ATeV and AuAu at 200 GeV have already been discussed in [11].
Results for 39 GeV
In Figure 6 and Figure 7, we show transverse momentum distributions of , , , , , in AuAu collisions at 39 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
Figure 6. Transverse momentum distributions of , in AuAu collisions at 39 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .

Figure 7. Same as Figure 6, but for , , , .
In Figure 8, we show transverse momentum distributions of , , , , , , , in AuAu collisions at 39 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Figure 8. Transverse momentum distributions of , , , , , , , in AuAu collisions at 39 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
In general, the simulation results are close to the data, concerning identified particles as pions, kaons, and protons, and as well as hyperons, although for kaons and antiprotons at low the simulations are slightly above the data. Concerning the meson, the simulations are somewhat above the data at low
Results for 27 GeV
In Figure 9 and 10, we show transverse momentum distributions of , , , , , in AuAu collisions at 27 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
Figure 9. Transverse momentum distributions of , in AuAu collisions at 27 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
In Figure 11, we show transverse momentum distributions of , , , , , , , in AuAu collisions at 27 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Figure 11. Transverse momentum distributions of , , , , , , , in AuAu collisions at 27 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
In general, the simulation results are relatively close to the data, concerning identified particles as pions, kaons, and protons, and as well hyperons. But compared to 39 GeV, for kaons, antiprotons, and the meson, the deviation (simulation compared to data) at low gets bigger.

Figure 10. Same as Figure 9, but for , , , .
Results for 19.6 GeV
In Figure 12 and Figure 13, we show transverse momentum distributions of , , , , , in AuAu collisions at 19.6 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
Figure 12. Transverse momentum distributions of , in AuAu collisions at 19.6 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .

Figure 13. Same as Figure 12, but for , , , .
In Figure 14, we show transverse momentum distributions of , , , , , , , in AuAu collisions at 19.6 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Figure 14. Transverse momentum distributions of , , , , , , , in AuAu collisions at 19.6 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
In general, the simulation results are relatively close to the data, concerning identified particles as pions, kaons, and protons, and as well hyperons. Similar to what we have already seen at 27 GeV, for kaons, antiprotons, and the meson, the simulation is somewhat above the data.
Results for 11.5 GeV
In Figure 15 and Figure 16, we show transverse momentum distributions of , , , , , in AuAu collisions at 11.5 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
Figure 15. Transverse momentum distributions of , in AuAu collisions at 11.5 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .

Figure 16. Same as Figure 15, but for , , , .
In Figure 17, we show transverse momentum distributions of , , , , , , , in AuAu collisions at 11.5 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].

Figure 17. Transverse momentum distributions of , , , , , , , in AuAu collisions at 11.5 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Here we see for the first time (compared to higher energies) significant deviations between simulation and data. The most striking is a large proton excess at low . And (as already seen earlier) a excess. Surprisingly, the hyperons are doing well.
Results for 7.7 GeV
In Figs. 18 and 19, we show transverse momentum distributions of , , , , , in AuAu collisions at 7.7 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .
In Fig. 20, we show transverse momentum distributions of , , , , , , , in AuAu collisions at 19.6 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Here, at 7.7 GeV, essentially all spectra from the simulation are too soft, the yields at low too high, with the biggest excess observed for protons and mesons. So here the model does not work.
Figure 19. Same as Figure 18, but for , , , .
Figure 18. Transverse momentum distributions of , in AuAu collisions at 7.7 GeV for different centrality classes. EPOS4 simulation (lines) are compared to data from STAR [28]. From top to bottom, we multiply the curves by , .

Figure 20. Transverse momentum distributions of , , , , , , , in AuAu collisions at 19.6 GeV at central rapidity for different centralities. EPOS4 simulation (lines) are compared to data from STAR [29].
Results concerning integrated yields
In this section, we show the integrated yields for every particle species as a function of , at mid-rapidity in the most central AuAu collisions. We summarise thus the discussion from the previous section, by displaying how particle production changes with collision energy for every hadronic species, compared to STAR data from [28, 29].
We compare in Figure 21 the yields from EPOS4 (lines)

Figure 21. Integrated yields of , , , , , and at mid-rapidity () in the most central AuAu collision class (0-5%), as a function of collision energy . EPOS4 simulation (lines) are compared to data from STAR [28].
of , , , , and , measured within , with STAR data from [28]. We do the same in Figure 22 for , , , , , and , measured within , compared with STAR data from [29]. For all hadronic species, the events considered are from the 0-5% centrality class, except for the mesons for which we consider 0-10% centrality events due to the available data. We only omit here and , as the data published by STAR in [29] is not using the same centrality classes for all energies, making it impossible to display meaningful results about the evolution of integrated yields with collision energy.

Figure 22. Integrated yields of , , , , , at mid-rapidity () in the most central AuAu collision class (0-5%), as a function of collision energy . EPOS4 simulation (lines) are compared to data from STAR [28].
When looking at the whole ensemble of results, we see that particle production in EPOS4 simulations above 19.6 GeV is, overall, in good qualitative and quantitative agreement with STAR data. As expected, it reflects the conclusions from the last section. At lower energies, we observe a clear overproduction of protons, as well as and baryons, unlike their respective anti-baryon counterparts. The yield of mesons is also largely overestimated at low energies, and displays notably almost no dependence in energy.
Finally, the yield of shown in Figure 22 is overestimated for the whole range of collision energies displayed here, compatible with the spectra at low . However, its energy dependence follows qualitatively well the data.
On the contrary, the meson is the only species which is underproduced in collision with energies 19.6 GeV and above. But this is in contradiction with the spectra shown earlier, where we see an overproduction.
Results concerning
In Figure 23, Figure 24, Figure 25, Figure 26, Figure 27, and Figure 28, the transverse momentum dependence of of identified particles within a pseudorapidity range of in AuAu collisions at energies from 62.4 GeV down to 7.7 GeV are shown. Different centrality classes are considered: 0-10%, 10-40%, and 40-80%. We show full simulations (thick red lines) and those

Figure 23. Transverse momentum dependence of of identified particles in AuAu collisions at 62.4 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).

Figure 24. Transverse momentum dependence of of identified particles in AuAu collisions at 39.4 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).

Figure 25. Transverse momentum dependence of of identified particles in AuAu collisions at 27 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).

Figure 26. Transverse momentum dependence of of identified particles in AuAu collisions at 19.6 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).

Figure 27. Transverse momentum dependence of of identified particles in AuAu collisions at 11.5 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).

Figure 28. Transverse momentum dependence of of identified particles in AuAu collisions at 7.7 GeV at central rapidity for different centralities. EPOS4 simulation, full simulations (thick red lines) and without hadronic cascade (thin green lines), are compared to data from STAR [30] (dots).
without hadronic cascade (thin green lines), and they are compared to data from STAR [30] in the Beam Energy Scan at the Relativistic Heavy Ion Collider. We employ the same event plane method as described in the experimental paper. For the upper plots, one should note that the ranges on the ordinates change, depending on the centrality, since increases with decreasing centrality. In general, the simulation results describe the data reasonably well, also the centrality dependence. The biggest deviation is observed for protons for 40-80%. Also for the pions the agreement is not so great. But the hyperons ( and and their antiparticles and ) and as well the
and the are relatively close to the data.
Summary and Conclusions
We reviewed briefly the EPOS4 approach, a detailed discussion can be found in [9, 10, 11, 12]. Most important is the concept of parallel scattering of primary interactions, which is needed at high energies, to be more precise above 24 GeV [11]. We also reviewed briefly “secondary interactions” in EPOS4, composed of core-corona separation, hydrodynamic core evolution with subsequent decay (microcanonically), and final state hadron cascade
In the EPOS4 formalism, there is a smooth transition from high to low energy, certain features change (or disappear) gradually. The parton ladders become less frequent, they are replaced by soft Pomerons, and most importantly, the relative importance of particle production from remnant excitation and decay increases. We discussed how this affects the core-corona procedure. Below 30 GeV, even central rapidities are dominated by particle production from remnants, and essentially all prehadrons go into the core (for central collisions). We computed the energy densities of the core (the fluid initial condition) for the different systems, to observe that they drop from about 40 GeV/fm at 5.02 ATeV to roughly 5 GeV/fm at 11.5 GeV, and then drop dramatically. At 4 GeV, there is no fluid anymore.
With all the model details already being published elsewhere [9, 10, 11, 12], the main purpose of the paper is the presentation of a very detailed test, considering a very large set of experimental data in the RHIC energy domain, covering the spectra and also the dependence of the elliptical flow, for identified particles. Spectra and provide complementary information about the fluid expansion. A particular aim was the investigation of a possible breakdown of the model at low energies, expected at around 24 GeV from theoretical considerations.
We first showed comparisons of simulations with data concerning spectra of identified particles, from 39 GeV down to 7.7 GeV. For the higher energies, down to 19.6 GeV, the simulation results are relatively close to the data. At lower energies, we observe “problems”. At 11.5 GeV, we see for the first time (compared to higher energies) significant deviations between simulation and data. The most striking is a large proton excess at low . At 7.7 GeV, essentially all spectra from the simulation are too soft, the yields at low too high, with the biggest excess observed for protons and mesons. So here the model does not work. These results concerning low yields are also summarized, more globally, through the integrated yields of different hadronic species displayed as a function of energy in the most central collisions.
The situation is quite different concerning . Here the simulation results describe the data reasonably well, also the centrality dependence. There is no “significant deterioration” at very low energies, as in the case of transverse momentum spectra. Although the model is obviously wrong at low energies, it works for .
The success (at energies 19.6 GeV) and the failure (at energies below 19.6 GeV) of the model correspond (amazingly well) to our earlier estimate of where the model should work and where not: it should work above 24 GeV (roughly estimated). The main problem of the (full) parallel scenario at low energies is the fact that a given nucleon hits all target nucleons on its way, although in reality it does not exist anymore for the final scatterings. This explains the large proton excess. One of the future projects will be to take that into account.
Acknowledgements
IK acknowledges support by the grant GA22-25026S of the Czech Science Foundation. MS acknowledges support by the U.S. Department of Energy grant DE-SC0020651. J.J. acknowledges support by the MUSES collaboration under National Science Foundation (NSF) grant number OAC-2103680.
Appendix
Core and corona contributions for 27 GeV, 11.5 GeV, and 7.7 GeV
In Figs. 29, 30, and 31, we show ratios X / core+corona versus , with X being the corona contribution (blue), the core (green), and the full contribution (red), for different hadrons. The four columns represent four different centrality classes, namely 0-5%, 20-40%, 60-80%, 80-100%. We show results for AuAu collisions at 27 GeV, 11.5 GeV, and 7.7 GeV.

Figure 29. The X / core+coron ratio, with X being the corona contribution (blue), the core (green), and the full contribution (red), for 4 centrality classes and four different particle species, for 27 GeV.
Figure 31. The X / core+coron ratio, with X being the corona contribution (blue), the core (green), and the full contribution (red), for 4 centrality classes and four different particle species, for 7.7 GeV.
Figure 30. The X / core+coron ratio, with X being the corona contribution (blue), the core (green), and the full contribution (red), for 4 centrality classes and four different particle species, for 11.5 GeV.
References
- [1]S. A. Bass et al., Prog. Part. Nucl. Phys. 41, 225 (1998), nucl-th/9803035.
- [2]M. Bleicher et al., J. Phys. G25, 1859 (1999), hep-ph/9909407.
- [3]E. Bratkovskaya, W. Cassing, V. Konchakovski, and O. Linnyk, Nuclear Physics A 856, 162 (2011).
- [4]W. Cassing and E. Bratkovskaya, Physical Review C 78, 034919 (2008).
- [5]W. Cassing and E. Bratkovskaya, Nuclear Physics A 831, 215 (2009).
- [6]P. Moreau et al., Physical Review C 100, 014911 (2019).
- [7]T. Song et al., Physical Review C 92, 014910 (2015).
- [8]J. Weil et al., Phys. Rev. C 94, 054905 (2016), arXiv:1606.06642.DOI
- [9]K. Werner, Phys. Rev. C 108, 064903 (2023), arXiv:2301.12517.arxiv.org/abs/2301.12517
- [10]K. Werner and B. Guiot, Phys. Rev. C 108, 034904 (2023), arXiv:2306.02396.arxiv.org/abs/2306.02396
- [11]K. Werner, (2023), arXiv:2310.09380.arxiv.org/abs/2310.09380
- [12]K. Werner, (2023), arXiv:2306.10277.DOI
- [13]V. N. Gribov, Zh. Eksp. Teor. Fiz. 53, 654 (1967).
- [14]V. N. Gribov, Sov. Phys. JETP 29, 483 (1969).
- [15]V. N. Gribov and L. N. Lipatov, Sov. J. Nucl. Phys. 15, 438 (1972).
- [16]V. A. Abramovsky, V. N. Gribov, and O. V. Kancheli, Yad. Fiz. 18, 595 (1973).
- [17]H. J. Drescher, M. Hladik, S. Ostapchenko, T. Pierog, and K. Werner, Phys. Rep. 350, 93 (2001), hep-ph/0007198.
- [18]G. Altarelli and G. Parisi, Nuclear Physics B. 126, 298 (1977).
- [19]Y. L. Dokshitzer, Sov. Phys. JETP 46, 641 (1977).
- [20]I. A. Karpenko, P. Huovinen, H. Petersen and M. Bleicher, Phys. Rev. C 91, no.6, 064901 (2015) doi:10.1103/PhysRevC.91.064901, arXiv:1502.01978 [nucl-th].DOI
- [21]C. Shen and B. Schenke, Phys. Rev. C 97, no.2, 024907 (2018) doi:10.1103/PhysRevC.97.024907, arXiv:1710.00881 [nucl-th].DOI
- [22]C. Shen and B. Schenke, Phys. Rev. C 105, no.6, 064905 (2022) doi:10.1103/PhysRevC.105.064905, arXiv:2203.04685 [nucl-th].DOI
- [23]C. Shen and S. Alzhrani, Phys. Rev. C 102, no.1, 014909 (2020) doi:10.1103/PhysRevC.102.014909, arXiv:2003.05852 [nucl-th].DOI
- [24]K. Werner, Phys. Rev. Lett. 98, 152301 (2007), arXiv:0704.1270.DOI
- [25]K. Werner, B. Guiot, I. Karpenko, and T. Pierog, Phys. Rev. C 89, 064903 (2014), arXiv:1312.1233.DOI
- [26]I. Karpenko, P. Huovinen, and M. Bleicher, Computer Physics Communications 185, 3016 (2014), arXiv:1312.4160.DOI
- [27]K. Werner, I. Karpenko, T. Pierog, M. Bleicher, and K. Mikhailov, Physical Review C 83 (2011), arXiv:1010.0400.DOI
- [28]STAR collaboration, L. Adamczyk et al., Phys.Rev.C 96, 044904 (2017).
- [29]STAR collaboration, J. Adam et al., Phys.Rev.C 102, 034909 (2020).
- [30]STAR, L. Adamczyk et al., Phys. Rev. C 93, 014907 (2016), 1509.08397.












