Cooling, conduction, compact objects: Gravothermal evolution of dissipative self-interacting dark matter halos

Technical Summary: Gravothermal Evolution of Dissipative Self-Interacting Dark Matter Halos

Problem Statement
Self-interacting dark matter (SIDM) models often predict radiative processes that dissipate energy, such as bremsstrahlung or decay of collisionally excited bound states. While elastic SIDM is known to drive gravothermal evolution (core formation followed by collapse), the specific impact of energy dissipation on this process remains under-explored in numerical simulations. Existing N-body methods for frequent small-angle scatterings (fSIDM) typically assume elastic interactions, while dissipative models often rely on fluid approximations or isotropic cross-section assumptions that may not capture the physics of light mediators (which are often forward-peaked). Furthermore, recent observations of a compact object in the strong lens system JVAS B1938+666 suggest a collapsing SIDM halo, but elastic models require unrealistically large cross-sections to explain the collapse by the observed redshift ($z=0.881$). This work investigates whether dissipative self-interactions can alter the gravothermal evolution of isolated halos to explain such observations with more moderate cross-sections.

Methodology
The authors present the first extension of the fSIDM N-body formalism (Fischer et al. 2021) to include effective dissipation.

• Numerical Implementation: The method treats frequent small-angle scatterings as a drag force and transverse momentum diffusion. The authors derive a dissipative extension where the drag force is enhanced by a dissipation parameter $r_{\text{diss}}$, and the stochastic velocity kick is modified to ensure the system loses energy consistent with the radiated energy $k_0$ in the underlying particle physics process. The time step is scaled down by $r_{\text{diss}}$ to maintain stability.
• Simulation Setup: They simulate isolated Navarro-Frenk-White (NFW) halos ($M \sim 10^{10} M_\odot$) using the OpenGadget3 code. Two series of simulations are performed:
  1. Varying the dissipation rate while keeping the heat conduction rate constant.
  2. Varying the heat conduction rate (via the transfer cross-section $\sigma_T/m_\chi$) while keeping the cooling rate constant.
• Comparisons: Results are compared against:
  • A gravothermal fluid model (augmented with a cooling term) to assess validity.
  • Isotropic "rarely self-interacting" (rSIDM) simulations to test the influence of angular dependence in the cross-section.
• Observational Application: The authors project the simulation results to derive surface density profiles and enclosed masses, comparing them to the mass distribution inferred for the JVAS B1938+666 perturber by Vegetti et al. (2026).

Key Results

1. Qualitative Change in Evolution: Dissipation does not merely accelerate collapse; it qualitatively alters the evolution. In the elastic case, heat conduction eventually reverses direction (outward) as the core becomes isothermal, driving collapse. In the dissipative case, sufficiently strong central cooling prevents the formation of an isothermal core. The velocity dispersion gradient remains positive at the center, keeping heat conduction directed inward throughout the evolution.
2. Mass Infall vs. Outflow: In elastic halos, the outer halo is heated by conduction, leading to mass outflow. In dissipative halos, outer regions cool efficiently (dominated by dissipation over outward conduction), leading to mass infall. This results in a larger region of mass infall and a less pronounced indentation between the core and outer halo in the final density profile.
3. Dependence on Parameters:
   • Dissipation: Increasing the cooling rate (controlled by $r_{\text{diss}}$) significantly accelerates collapse (by nearly two orders of magnitude for the parameters tested) and produces denser, less extended cores.
   • Conduction: The effect of the transfer cross-section is non-monotonic. At low cross-sections (long mean free path), increasing conduction accelerates collapse. At high cross-sections (short mean free path), increasing conduction suppresses the heat transport efficiency, delaying collapse.
   • Angular Dependence: The authors find that the evolution of isolated halos is largely insensitive to the angular dependence of the cross-section (forward-peaked vs. isotropic), provided the cooling and conduction rates are matched.
4. JVAS B1938+666 Application: The authors demonstrate that weakly dissipative SIDM halos can reproduce the inferred mass profile of the JVAS B1938+666 perturber. Compared to the elastic case, dissipative models can explain the observation with:
   • Significantly shorter evolution times (allowing collapse by $z=0.881$ with moderate cross-sections).
   • Or, equivalently, smaller self-interaction cross-sections for the same evolution time.
   • The required cross-sections ($\sigma_T/m_\chi \sim 10^2 \text{ cm}^2 \text{ g}^{-1}$) are consistent with constraints for halos of mass $M_{200c} \lesssim 10^7 M_\odot$.

Significance and Claims
The paper claims to open a "new route to connecting halo structure and recently reported compact objects to dark-sector microphysics." By extending the fSIDM formalism to include dissipation, the authors provide a framework to test models featuring light mediators and bound states. The primary significance lies in showing that dissipation can resolve the tension between the observed compact object in JVAS B1938+666 and the large cross-sections required by elastic models. The authors modestly note that their conclusions rely on specific initial conditions (isolated NFW halos) and that cosmological simulations and simulations extending deeper into the collapse phase (beyond numerical breakdown) are necessary to fully constrain dissipative properties. They also acknowledge that the angular dependence of the cross-section may become more relevant for satellites and mergers, a regime not covered by this isolated halo study.