A numerical approach to particle creation in accelerating toy models

1. Problem Statement

The paper addresses the challenge of numerically simulating particle creation (specifically Hawking radiation) in dynamical gravitational spacetimes.

• Theoretical Context: While particle creation in cosmological backgrounds (like inflation) is well-understood analytically, calculating the full spectrum of particles created during gravitational collapse (e.g., star to black hole) or compact object mergers is technically difficult.
• Limitations of Current Methods:
  • Analytical solutions for black hole formation are typically restricted to "sufficiently late times" (after the black hole has settled), where the spectrum depends only on the final mass and angular momentum, ignoring the details of the collapse.
  • Standard numerical relativity uses Cauchy foliations (spacelike slices). These slices cannot reach null infinity ($\mathscr{I}^+$), where radiation is unambiguously defined. Extracting radiation from Cauchy slices requires post-processing extrapolation, introducing systematic errors.
  • Simulating the full process (from past null infinity $\mathscr{I}^-$ to future null infinity $\mathscr{I}^+$) requires a framework that can handle asymptotic boundaries directly.

2. Methodology

The authors propose a numerical framework based on hyperboloidal slicing to solve the scattering problem for massless scalar fields in asymptotically flat spacetimes.

A. Theoretical Framework

• Model: Instead of solving the full Einstein Field Equations coupled with the Klein-Gordon (KG) equation, the authors use a toy model in Minkowski spacetime.
• Effective Potential: They introduce a time-dependent effective potential $V(t, r)$ in the KG equation to mimic the gravitational dynamics of a collapsing star.
  • The potential acts as a barrier that scatters the field modes.
  • Time-dependence in the potential (amplitude or position oscillation) mimics the dynamical spacetime required to excite particles from the vacuum.
• Quantization: Particle creation is quantified via Bogoliubov coefficients ($\alpha_{\omega\omega'}$ and $\beta_{\omega\omega'}$). A non-zero $\beta$ coefficient indicates that the "in" vacuum state (at $\mathscr{I}^-$) contains particles when viewed from the "out" region (at $\mathscr{I}^+$).

B. Numerical Implementation

1. Hyperboloidal Foliation:
   • The Minkowski spacetime is foliated using hyperboloidal slices that asymptotically reach both past ($\mathscr{I}^-$) and future ($\mathscr{I}^+$) null infinity.
   • Compactification: A conformal factor $\Omega$ is used to map the infinite radial coordinate to a finite computational grid ($r \in [0, 1]$), where $r=1$ represents null infinity.
   • Coordinate Transformation: The authors transform from standard Minkowski coordinates $(\tilde{t}, \tilde{r})$ to hyperboloidal coordinates $(t, r)$ using a height function $h(r)$ and conformal factor $\Omega(r)$.
2. Evolution Scheme:
   • The KG equation is rewritten as a first-order in time, second-order in space (FT2S) system.
   • Method of Lines: Spatial derivatives are discretized using 4th-order centered Finite Differences; time integration uses 4th-order Runge-Kutta.
   • Boundary Conditions: Parity conditions at the origin ($r=0$) and outflow conditions at null infinity ($r=1$).
3. Data Translation (The "Scattering" Step):
   • The simulation is split into two phases:
     1. Ingoing Evolution: Propagate the initial signal from $\mathscr{I}^-$ using ingoing hyperboloidal slices.
     2. Translation: Interpolate data from the ingoing slices to a specific outgoing hyperboloidal slice ($t_p = t_{match}$). This step bridges the two foliations.
     3. Outgoing Evolution: Propagate the signal from the matching slice to $\mathscr{I}^+$.
   • Interpolation: B-splines (4th order) are used to reconstruct the signal on the new grid, with artificial decay added to the tails to handle regions not covered by the intersection of slices.
4. Bogoliubov Coefficient Computation:
   • The evolved signal at $\mathscr{I}^+$ is projected onto a basis of "out" modes (windowed plane waves) using the Klein-Gordon inner product.
   • The coefficients are calculated via integrals over $\mathscr{I}^+$.
   • Normalization: A specific rescaling factor is applied to the KG product to ensure the unitarity condition $\int (|\alpha|^2 - |\beta|^2) d\omega' = 1$ holds numerically, compensating for the finite width of the windowed modes.

3. Key Contributions

• Novel Numerical Framework: Successfully implemented a hyperboloidal slicing method to solve scattering problems from $\mathscr{I}^-$ to $\mathscr{I}^+$ in a single computational grid, avoiding the need for Cauchy-characteristic extraction.
• Toy Model Validation: Demonstrated that time-dependent effective potentials in Minkowski space can successfully reproduce the qualitative features of particle creation (non-zero $\beta$ coefficients) without solving full General Relativity.
• Data Translation Technique: Developed a robust procedure to translate data between ingoing and outgoing hyperboloidal foliations, a critical step for scattering problems in this framework.
• Normalization Protocol: Established a method to normalize windowed modes to satisfy unitarity constraints, addressing the sensitivity of the KG product to the width of the signal envelope.

4. Numerical Results

The authors tested four scenarios:

1. Vanishing Potential ($V=0$):
   • Result: The signal propagates unperturbed. $\beta_{\omega\omega'} \approx 0$ (two orders of magnitude smaller than $\alpha$).
   • Significance: Validates the code's ability to preserve the vacuum state in a static background.
2. Static Potential Barrier:
   • Result: The signal scatters (partial reflection/transmission) but remains stationary. $\beta_{\omega\omega'} \approx 0$.
   • Significance: Confirms that static backgrounds do not create particles (no frequency mixing).
3. Pulsating Potential Barrier (Amplitude Oscillation):
   • Result: The potential amplitude oscillates. The $\beta$ spectrum shows non-negligible peaks comparable to the $\alpha$ spectrum.
   • Significance: Demonstrates particle creation due to time-dependent amplitude. The total particle number is finite (convergence of $\beta$ at high frequencies).
4. Shaking Potential Barrier (Position Oscillation):
   • Result: The potential center oscillates radially. Similar to the pulsating case, significant $\beta$ coefficients are observed.
   • Significance: Confirms that spatial oscillation of the barrier also induces particle creation.

Convergence Tests:

• Pointwise convergence tests showed good agreement between medium, high, and highest resolutions ($N=1000$ to $4000$).
• Minor convergence losses were observed near the potential region and during the data translation step, but the overall results were deemed reliable.
• The unitarity condition $\int (|\alpha|^2 - |\beta|^2) d\omega' = 1$ was satisfied within numerical error (e.g., $1.000491$ for the pulsating case).

5. Significance and Future Prospects

• Bridging the Gap: This work provides a rigorous numerical tool to study particle creation during the entire dynamical process of collapse, not just the late-time equilibrium.
• Astrophysical Relevance: The framework is a stepping stone toward simulating more complex scenarios, such as:
  • Binary black hole mergers.
  • Dynamical spacetimes with oscillating black hole masses.
  • Full General Relativistic simulations where the background metric evolves dynamically.
• Future Work: The authors plan to apply this framework to fixed Schwarzschild backgrounds and eventually to dynamical spacetimes generated by numerical relativity simulations of gravitational collapse. They also aim to improve the slice-translation algorithm to reduce interpolation errors.

In summary, the paper successfully demonstrates that hyperboloidal slicing combined with effective time-dependent potentials is a viable and powerful method for numerically investigating quantum particle creation in dynamical spacetimes, overcoming the limitations of traditional Cauchy-based approaches.