Massive-Star Evolution Toward Collapse

Scrollable formula. Use Left and Right arrow keys to reveal hidden content.

Scrollable table. Use Left and Right arrow keys to reveal additional columns.

Scrollable figure. Use Left and Right arrow keys to inspect hidden content.

A massive star does not carry its birth mass intact to collapse. Successive burning stages build a stratified core while convection, overshoot, winds, rotation, and internal torques redistribute mass and angular momentum. The central lesson is therefore structural: the density and entropy profiles at the end of stellar evolution predict collapse behavior more directly than the zero-age main-sequence mass alone. This page follows that causal chain and explains what a one-dimensional stellar model can—and cannot—establish.

Helpful background. Black-Hole Geometry and Mechanics fixes the distinction between a physical mass and the length scale $GM$ ; Relativistic Matter and Compact Objects supplies stellar equilibrium and microphysical closure; Black-Hole Astrophysics supplies the observational role of a remnant model; and Black-Hole Dynamics and Collapse explains how the pre-collapse profile sets the initial conditions for collapse and horizon formation.

The core, not merely the birth mass

Hydrogen burning leaves a helium-rich core; helium burning sets the carbon-to-oxygen ratio; carbon, neon, oxygen, and silicon burning then proceed on successively shorter timescales. Neutrino cooling becomes increasingly important, so the last stages change the core faster than the envelope can thermally readjust. At the onset of iron-core collapse, a stellar-evolution calculation should deliver radial profiles such as

\rho(m),\quad T(m),\quad s(m),\quad Y_e(m),\quad X_i(m), \quad\text{and}\quad j(m),

Here $m$ is enclosed baryonic mass, $\rho$ is density, $T$ is temperature, $s$ is entropy per baryon, $Y_e$ is the number of electrons per baryon, $X_i$ are isotope mass fractions, and $j$ is specific angular momentum. These profiles, together with the binding energy outside the iron core, are the physical output needed by a collapse calculation. A single label such as “ $25M_\odot$ progenitor” discards most of the information that controls shock accretion and the available explosion energy.

The familiar initial-mass fate diagram is still useful as orientation, but its boundaries inherit the adopted wind, mixing, and explosion prescriptions. Heger and collaborators explicitly organize the outcomes by metallicity, mass loss, and helium-core structure in their classic fate map Heger et al. 2003, §§ 2–3 and Fig. 1. It should be read as a conditional model map, not a periodic table of immutable stellar fates.

Burning shells and mixing boundaries

Late burning creates composition interfaces. Their locations matter because the density gradient encountered by the stalled supernova shock controls the accretion rate onto the proto-neutron star. Convective boundary mixing changes the amount of fresh fuel available to a burning core and can shift the carbon-shell history; shell mergers can alter both composition and compactness. In a one-dimensional code these processes are represented by closures—mixing-length convection, overshoot, semiconvection, and sometimes diffusion coefficients—not resolved turbulent flows.

This explains a seemingly surprising result: neighboring initial masses can have quite different pre-collapse cores. Sukhbold and collaborators found nonmonotonic structure and interleaved successful and failed outcomes in a densely sampled one-dimensional progenitor grid Sukhbold et al. 2016, § 2.3 and §§ 4–5. The fine alternation is informative about shell physics, but its precise initial-mass locations are sensitive to numerical controls and stellar inputs. Robust interpretation compares profiles and integral diagnostics across grids rather than matching only one initial mass.

Winds, rotation, and internal torques

Winds remove the envelope and expose hotter layers. Because line driving depends on metal opacity, a lower-metallicity star generally retains more mass under otherwise comparable assumptions. Binary stripping can accomplish a similar removal through a different physical route, so a bare helium star need not have experienced a strong wind. This distinction becomes important when one tries to infer a birth environment from a remnant mass.

Rotation adds centrifugal support and composition mixing. It can enlarge a helium core by feeding hydrogen into the burning region, while internal angular-momentum transport can simultaneously drain the core’s spin. Hydrodynamic instabilities, magnetic torques, tides, and binary mass transfer may compete. Consequently, “rotation produces a rapidly spinning black hole” is not a valid general implication: the final dimensionless spin depends on the retained profile $j(m)$ and on which mass actually joins the remnant,

a_* = \frac{J_{\rm BH}}{G M_{\rm BH}^2}, \qquad J_{\rm core}(m_b)=\int_0^{m_b} j(m)\,dm.

Here $m_b$ is a baryonic mass coordinate in the stellar model, whereas $M_{\rm BH}$ is the final gravitational mass. The two angular momenta agree only if that material joins the hole without angular-momentum loss or additional supply. The formula uses the site convention $c=1$ ; restoring ordinary units gives $a_*=cJ_{\rm BH}/(GM_{\rm BH}^2)$ . Removing high- $j$ envelope material can reduce $J$ much more strongly than it reduces the core mass.

Structural diagnostics and their limits

A compact summary of the inner density profile is the compactness parameter

\xi_M = \left. \frac{M/M_\odot}{R(m=M)/(1000\ {\rm km})} \right|_{t_*},

where the evaluation time $t_*$ must be stated—often the onset of collapse—and $M$ denotes an enclosed baryonic mass coordinate, not the final black-hole mass. Large $\xi_M$ means that the chosen mass is packed within a smaller radius, usually implying a larger post-bounce accretion rate. This stellar diagnostic is distinct from relativistic compactness $GM/(Rc^2)$ : the normalization by $1000\ {\rm km}$ is conventional, and $\xi_M\gt1$ does not mean that a horizon has formed.

It is not a universal one-number explosion criterion. The two-parameter construction of Ertl et al. 2016, §§ 3.1–3.3, especially eq. (6) uses $M_4$ , the enclosed mass in solar-mass units at entropy $s=4k_{\rm B}$ per baryon, and $\mu_4$ , a dimensionless mass gradient just outside that location. Their practical gradient uses a finite mass interval of $0.3M_\odot$ and radial units of $1000\ {\rm km}$ . The pair retains information about both the accreting core and the surrounding density gradient.

Both diagnostics are compression maps: different profiles can share the same value. A responsible comparison therefore also inspects the density gradient, interface locations, envelope binding energy, and the calculation’s stopping condition. Detailed neutrino-radiation transport and gravitational-wave population likelihoods remain outside this page; the first belongs to the dedicated collapse treatment, while the second requires a detector selection model after the remnant population has been constructed.

A reproducible teaching check

A production calculation would pin a stellar-evolution code commit, reaction network, opacity tables, abundance mixture, wind switches, convection and overshoot controls, rotation and torque prescriptions, mesh and timestep tolerances, and collapse stopping condition. It would publish the inlist or equivalent controls and compare helium-core masses and full profiles at more than one metallicity. No such production grid was rerun for this page.

The following is instead a teaching reconstruction, not a claimed scientific rerun. It integrates the analytic profile $\rho(r)\propto1-(r/R)^2$ , locates the $2.5M_\odot$ mass coordinate, and compares $\xi_{2.5}$ with the analytic integral. The pinned environment used for the displayed output is Python 3.9.6 with NumPy==2.0.2. Save the following teaching script as core_profile_check.py.

import numpy as np

M_TARGET = 2.5
M_TOTAL = 3.0
R_KM = 3000.0

def enclosed_exact(r):
    x = r / R_KM
    return M_TOTAL * (2.5 * x**3 - 1.5 * x**5)

def exact_radius():
    lo, hi = 0.0, R_KM
    for _ in range(100):
        mid = 0.5 * (lo + hi)
        if enclosed_exact(mid) < M_TARGET:
            lo = mid
        else:
            hi = mid
    return 0.5 * (lo + hi)

def estimate(n):
    r = np.linspace(0.0, R_KM, n + 1)
    shape = 1.0 - (r / R_KM)**2
    shell = 4.0 * np.pi * r**2 * shape
    mass = np.zeros_like(r)
    mass[1:] = np.cumsum(0.5 * (shell[1:] + shell[:-1]) * np.diff(r))
    mass *= M_TOTAL / mass[-1]
    radius = np.interp(M_TARGET, mass, r)
    return radius, M_TARGET / (radius / 1000.0)

r_ref = exact_radius()
xi_ref = M_TARGET / (r_ref / 1000.0)
for n in (128, 256, 512):
    radius, xi = estimate(n)
    print(n, f"{radius:.6f}", f"{xi:.8f}", f"{abs(xi-xi_ref):.3e}")
print("exact", f"{r_ref:.6f}", f"{xi_ref:.8f}")

The program must produce the grid size, inferred radius, compactness, and absolute error. A clean save-and-run recipe is:

python3 -m venv .venv-core
. .venv-core/bin/activate
python3 -m pip install numpy==2.0.2
python3 core_profile_check.py

Grid intervals $N$	Radius result (km)	Compactness result	Absolute error
128	2480.157892	1.00800034	$1.525\times10^{-5}$
256	2480.191397	1.00798672	$1.629\times10^{-6}$
512	2480.194023	1.00798566	$5.619\times10^{-7}$

The analytic integral provides the independent exact-solution benchmark: its reference value is $R(2.5M_\odot)=2480.195406$ km and $\xi_{2.5}=1.00798509$ . Refinement reduces the interpolation and quadrature error. A real stellar profile can be harder: a principal failure mode is under-resolving sharp burning interfaces, and a smoothly converged integral can still be physically wrong if mixing or wind switches are inappropriate.

Common pitfalls

Reading initial mass as destiny. Initial mass indexes a calculation; it does not uniquely determine the core. Metallicity, mixing, mass loss, rotation, and binary history must accompany it.

Calling a closure a measurement. A one-dimensional diffusion coefficient can be calibrated and tested, but it is still a model of multidimensional transport. Agreement in final mass alone does not validate the internal profile.

Exercise

A pre-collapse model places the $2.5M_\odot$ mass coordinate at $1500$ km. Calculate $\xi_{2.5}$ , and explain why that value alone does not determine whether the star explodes.

Solution

Direct substitution gives

\xi_{2.5}=\frac{2.5}{1500/1000}=1.67.

The number says that this particular enclosed mass is concentrated within a relatively small radius at the declared evaluation time. It does not encode the density gradient outside that coordinate, composition-interface arrival times, envelope binding energy, proto-neutron-star neutrino response, or the explosion model. Those missing data can change the outcome even for models with similar compactness.

References

Ertl, T., Janka, H.-T., Woosley, S. E., Sukhbold, T., and Ugliano, M. (2016). “A Two-parameter Criterion for Classifying the Explodability of Massive Stars by the Neutrino-driven Mechanism.” The Astrophysical Journal 818, 124. DOI.
Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., and Hartmann, D. H. (2003). “How Massive Single Stars End Their Life.” The Astrophysical Journal 591, 288–300. DOI.
Sukhbold, T., Ertl, T., Woosley, S. E., Brown, J. M., and Janka, H.-T. (2016). “Core-collapse Supernovae from 9 to 120 Solar Masses Based on Neutrino-powered Explosions.” The Astrophysical Journal 821, 38. DOI; model archive.