High mass accretion rates onto evolved stripped-envelope massive stars by jet-induced mass removal

TL;DR

The paper investigates whether evolved stripped-envelope stars, modeled as Wolf-Rayet–like objects, can sustain high-rate mass accretion without undergoing excessive expansion by invoking jet-induced mass removal of the outer envelope. Using 1D MESA simulations with a pulsed accretion scheme, the authors mimic jets that remove high-entropy outer layers, quantified by the removal fraction $\alpha = \dot M_{rm}/\dot M_{add}$, and show that increasing $\alpha$ substantially mitigates stellar expansion. They find that, for realistic pulses, the star can maintain a deep gravitational potential while accreting, allowing substantial accretion energy release, with estimates like $\dot E_{acc} \simeq \frac{1}{2} \frac{G M_* M_{acc}}{R_*}$ achieving values orders of magnitude above the star’s luminosity during brief episodes. These results bolster models of intermediate-luminosity optical transients (ILOTs), such as luminous red novae, powered by jets from accreting non-degenerate stars, and reinforce the role of jet feedback in binary-interaction–driven transients. The study demonstrates the viability of 1D pulsed-accretion modeling to capture essential physics of jet-induced mass removal, while acknowledging the need for 3D simulations to fully resolve angular-momentum transport and jet-star coupling.

Abstract

Simulating one-dimensional stellar evolution models with MESA, we show that removing the outer inflated envelope of a mass-accreting evolved stripped-envelope star, like a Wolf-Rayet (WR) star, substantially moderates the stellar expansion during accretion at high-mass accretion rates. We study the accretion onto a star via an accretion disk, which launches jets that remove the high-entropy outer layers of the inflated envelope. This is the `jetted mass removal accretion scenario.' By manually removing the entire hydrogen-rich envelope from a red supergiant, we build a hydrogen-deficient WR stellar model with a mass of 6.03Mo and a radius of 0.67Ro. We then accrete mass onto it at a high rate. We mimic the real process of simultaneous mass addition near the equatorial plane and jet-induced mass removal from the outer envelope by dividing the accretion period into hundreds of pulses: in the first half of each pulse, we add mass, and in the second, we remove a fraction of this mass. The removal of tens of percent from the added mass decreases the stellar expansion by a factor of about 2-5. Our results show that WR stars can maintain a deep potential well and not expand much while accreting mass at high rates. This allows the formation of an accretion disk and the liberation of large amounts of gravitational energy. Our results strengthen models of intermediate-luminosity optical transients, such as luminous red novae, in which a non-degenerate star accretes at high rates and launches jets that power the transient event.

Figures (6)

• Figure 1: The stellar mass, $M$, and masses of four isotopes, $M_{\rm i}$ as indicated, as a function of time. The upper two panels correspond to Stage A1, and the lower two panels correspond to Stage A2. The mass loss in Stage A1 is the regular stellar wind. In contrast, in Stage A2, we manually set a much higher mass-loss rate to mimic the external effects of a binary companion, e.g., common-envelope evolution. In the upper panels, the time is the stellar age in Myr; in the lower panels, it is the stellar age minus 9 Myr. Stage A2 ends in a stripped-envelope star (a WR star); see Table \ref{['Tab:Table1']}.
• Figure 2: Total abundances of four isotopes (see legend) as a function of radius at the end of the three stages (two different simulations of stage A3; see Table \ref{['Tab:Table2']}). The radius at the end of each phase is to the right of each panel. The pink area in the upper panel marks zones where the convective velocity is $v_{\rm conv}\geq 1 {~\rm km} {~\rm s}^{-1}$. The inset in the upper panel zooms in on the core. Note that we accrete hydrogen-rich gas onto the stripped-envelope star (Stage A3: here B4 and B5), which has no hydrogen at the end of Stage A2.
• Figure 3: Entropy profiles as a function of stellar radius at three times during one pulse of mass accretion in simulation B5. The three profiles overlap in the inner zones. The blue line shows the entropy profile at the end of Stage A2, just before we start the accretion process; only its right-hand, which extends to a radius of $0.67R_\odot$, is visible in this plot. The dotted-orange line shows the entropy profile after mass addition at the end of the first half of the first pulse, and the dashed-green line shows it after mass removal of $80\%$ of the added mass at the end of the pulse.
• Figure 4: The stellar radius as a function of time measured from the beginning of the accretion process (Stage A3) for simulation group B. In all simulations here the net accretion rate is $\dot M_{\rm acc}=0.015 M_\odot {~\rm yr}^{-1}$. The upper, smooth blue line shows the expansion as mass is continuously added. The other lines show accretion in pulses: half the pulse time is spent on mass addition, followed by mass removal. $\alpha$ is the ratio of addition to removal rates, i.e., $1-\alpha$ is the fraction of added mass that the star retains. The smaller the fraction of retained mass, the smaller the expansion.
• Figure 5: Similar to Figure \ref{['fig:alpha']}, but for cases D, E, and F, which have different mass accretion rates and different durations (see Table \ref{['Tab:Table2']}). Due to the large number of pulses (500) in simulations F1 and F2, the line width smears the zigzag. In the inset of the lower panel, we expand a short time interval to demonstrate that the zigzag exists.