arXiv subjects

D. John Hillier

Publications and source records attributed to D. John Hillier.

At least 91 records · Page 5Linked to original sources

arXiv2014

Critical ingredients of supernova Ia radiative-transfer modeling

We explore the physics of SN Ia light curves and spectra using the 1-D non-LTE time-dependent radiative-transfer code CMFGEN. Rather than adjusting ejecta properties to match observations, we select as input one "standard" 1-D Chandrasekhar-mass delayed-detonation hydrodynamical model, and then explore the sensitivity of radiation and gas properties on radiative-transfer modeling assumptions. The correct computation of SN Ia radiation is not exclusively a solution to an "opacity problem", characterized by the treatment of a large number of lines. It is also key to treat important atomic processes consistently. Besides handling line blanketing in non-LTE, we show that including forbidden line transitions of metals is increasingly important for the temperature and ionization of the gas beyond maximum light. Non-thermal ionization and excitation are also critical since they affect the color evolution and the Delta-M15 of our model. While impacting little the bolometric luminosity, a more complete treatment of decay routes leads to enhanced line blanketing, e.g., associated with 48Ti in the U and B bands. Overall, we find that SN Ia radiation properties are influenced in a complicated way by the atomic data we employ, so that obtaining converged results is a challenge. We nonetheless obtain a good match to the golden standard type Ia SN 2005cf in the optical and near-IR, from 5 to 60d after explosion, suggesting that assuming spherical symmetry is not detrimental to SN Ia radiative-transfer modeling at these times. Multi-D effects no doubt matter, but they are perhaps less important than accurately treating non-LTE processes [abridged].

astro-ph.SR↗

arXiv2014

Constraints on the explosion mechanism and progenitors of type Ia supernovae

Observations of SN 2011fe at early times reveal an evolution analogous to a fireball model of constant color. In contrast, our unmixed delayed detonations of Chandrasekhar-mass white dwarfs (DDC series) exhibit a faster brightening concomitant with a shift in color to the blue. In this paper, we study the origin of these discrepancies. We find that strong chemical mixing largely resolves the photometric mismatch at early times, but it leads to an enhanced line broadening that contrasts, for example, with the markedly narrow SiII6355A line of SN 2011fe. We also explore an alternative configuration with pulsational-delayed detonations (PDDEL model series). Because of the pulsation, PDDEL models retain more unburnt carbon, have little mass at high velocity, and have a much hotter outer ejecta after the explosion. The pulsation does not influence the inner ejecta, so PDDEL and DDC models exhibit similar radiative properties beyond maximum. However, at early times, PDDEL models show bluer optical colors and a higher luminosity, even for weak mixing. Their early-time radiation is derived primarily from the initial shock-deposited energy in the outer ejecta rather than radioactive decay heating. Furthermore, PDDEL models show short-lived CII lines, reminiscent of SN 2013dy. They typically exhibit lines that are weaker, narrower, and of near-constant width, reminiscent of SN 2011fe. In addition to multi-dimensional effects, varying configurations for such ``pulsations" offer a source of spectral diversity amongst SNe Ia. PDDEL and DDC models also provide one explanation for low- and high-velocity gradient SNe Ia.

astro-ph.SR↗

arXiv2014

[CoIII] versus NaID in type Ia supernova spectra

The high metal content and fast expansion of supernova (SN) Ia ejecta lead to considerable line overlap in their optical spectra. Uncertainties in composition and ionization further complicate the process of line identification. In this paper, we focus on the 5900A emission feature seen in SN Ia spectra after bolometric maximum, a line which in the last two decades has been associated with [CoIII]5888A or NaID. Using non-LTE time-dependent radiative-transfer calculations based on Chandrasekhar-mass delayed-detonation models, we find that NaID line emission is extremely weak at all post-maximum epochs. Instead, we predict the presence of [CoIII]5888A after maximum in all our SN Ia models, which cover a range from 0.12 to 0.87Msun of 56Ni. We also find that the [CoIII]5888A forbidden line is present within days of bolometric maximum, and strengthens steadily for weeks thereafter. Both predictions are confirmed by observations. Rather than trivial taxonomy, these findings confirm that it is necessary to include forbidden-line transitions in radiative-transfer simulations of SNe Ia, both to obtain the correct ejecta cooling rate and to match observed optical spectra.

astro-ph.SR↗

arXiv2013

Type II-Plateau supernova radiation: dependencies on progenitor and explosion properties

We explore the properties of Type II-Plateau (II-P) supernovae (SNe) together with their red-supergiant (RSG) star progenitors. Using MESA STAR, we modulate the parameters (e.g., mixing length, overshoot, rotation, metallicity) that control the evolution of a 15Msun main-sequence star to produce a variety of physical pre-SN models and SN II-P ejecta. We extend previous modeling of SN II-P radiation to include photospheric and nebular phases, as well as multi-band light curves and spectra. Our treatment does not assume local thermodynamic equilibrium, is time dependent, treats explicitly the effects of line blanketing, and incorporates non-thermal processes. We find that the color properties of SNe II-P require large model atoms for FeI and FeII, much larger than adopted in Dessart & Hillier (2011). The color properties also imply RSG progenitors of limited extent (~500Rsun) --- larger progenitor stars produce a SN II-P radiation that remains too blue for too long. This finding calls for a reduction of RSG radii, perhaps through a strengthening of convective energy transport in RSG envelopes. Increased overshoot and rotation reduce the ratio of ejecta to helium-core mass, similarly to an increase in main-sequence mass, and thus complicate the inference ofprogenitor masses. In contrast to the great sensitivity on progenitor radius, SN II-P color evolution appears insensitive to variations in explosion energy. Finally, we document the numerous SN II-P signatures that vary with progenitor metallicity, revealing their potential for metallicity determinations in the nearby and distant Universe.

astro-ph.SR↗

arXiv2013

Massive stars at low metallicity: Evolution and surface abundances of O dwarfs in the SMC

We study the evolution, rotation, and surface abundances of O-type dwarfs in the Small Magellanic Cloud. We analyzed the UV and optical spectra of twenty-three objects and derived photospheric and wind properties. The observed binary fraction of the sample is ~ 26%, which is compatible with more systematic studies, if one considers that the actual binary fraction is potentially larger owing to low-luminosity companions and that the sample excluded obvious spectroscopic binaries. The location of the fastest rotators in the H-R diagram indicates that these could be several Myr old. The offset in the position of these fast rotators compared with the other stars confirms the predictions of evolutionary models that fast-rotating stars tend to evolve more vertically in the H-R diagram. Only one star of luminosity-class Vz, expected to best characterize extreme youth, is located on the ZAMS, the other two stars are more evolved. The distribution of nitrogen abundance of O and B stars suggests that the mechanisms responsible for the chemical enrichment of slowly rotating massive stars depends only weakly on the star's mass. We confirm that the group of slowly rotating N-rich stars is not reproduced by the evolutionary tracks. Our results call for stronger mixing in the models to explain the range of observed N abundances. All stars have an N/C ratio as a function of stellar luminosity that matches the predictions of the stellar evolution models well. More massive stars have a higher N/C ratio than the less massive stars. Faster rotators show on average a higher N/C ratio than slower rotators. The N/O versus N/C ratios agree qualitatively well with those of stellar evolution models. The only discrepant behavior is observed for the youngest two stars of the sample, which both show very strong signs of mixing, which is unexpected for their evolutionary status.

astro-ph.SR↗

arXiv2013

A Bake-Off Between CMFGEN and FASTWIND: Modeling the Physical Properties of SMC and LMC O-type Stars

The model atmosphere programs FASTWIND and CMFGEN are both elegantly designed to perform non-LTE analyses of the spectra of hot massive stars, and include sphericity and mass-loss. The two codes differ primarily in their approach towards line blanketing, with CMFGEN treating all of the lines in the co-moving frame and FASTWIND taking an approximate approach which speeds up execution times considerably. Although both have been extensively used to model the spectra of O-type stars, no studies have used the codes to independently model the same spectra of the same stars and compare the derived physical properties. We perform this task on ten O-type stars in the Magellanic Clouds. For the late-type O supergiants, both CMFGEN and FASTWIND have trouble fitting some of the He I lines, and we discuss causes and cures. We find that there is no difference in the average effective temperatures found by the two codes for the stars in our sample, although the dispersion is large, due primarily to the various difficulties each code has with He I. The surface gravities determined using FASTWIND are systematically lower by 0.12 dex compared to CMFGEN, a result we attribute to the better treatment of electron scattering by CMFGEN. This has implications for the interpretation of the origin of the so-called mass discrepancy, as the masses derived by FASTWIND are on average lower than inferred from stellar evolutionary models, while those found by CMFGEN are in better agreement.

astro-ph.SR↗

arXiv2012

One-dimensional delayed-detonation models of Type Ia supernovae: Confrontation to observations at bolometric maximum

The delayed-detonation explosion mechanism applied to a Chandrasekhar-mass white dwarf offers a very attractive model to explain the inferred characteristics of Type Ia supernovae (SNe Ia). The resulting ejecta are chemically stratified, have the same mass and roughly the same asymptotic kinetic energy, but exhibit a range in 56Ni mass. We investigate the contemporaneous photometric and spectroscopic properties of a sequence of delayed-detonation models, characterized by 56Ni masses between 0.18 and 0.81 Msun. Starting at 1d after explosion, we perform the full non-LTE, time-dependent radiative transfer with the code CMFGEN, with an accurate treatment of line blanketing, and compare our results to SNe Ia at bolometric maximum. Despite the 1D treatment, our approach delivers an excellent agreement to observations. We recover the range of SN Ia luminosities, colours, and spectral characteristics from the near-UV to 1 micron, for standard as well as low-luminosity 91bg-like SNe Ia. Our models predict an increase in rise time to peak with increasing 56Ni mass, from ~15 to ~21d, yield peak bolometric luminosities that match Arnett's rule to within 10%, and reproduce the much smaller scatter in near-IR magnitudes compared to the optical. We reproduce the morphology of individual spectral features, the stiff dependence of the R(Si) spectroscopic ratio on 56Ni mass, and the onset of blanketing from TiII/ScII in low-luminosity SNe Ia with a 56Ni mass <0.3 Msun. We find that ionization effects, which often dominate over abundance variations, can produce high-velocity features in CaII lines, even in 1D. Distinguishing between different SN Ia explosion mechanisms is a considerable challenge but the results presented here provide additional support to the viability of the delayed-detonation model.

astro-ph.SR↗

arXiv2012

Radiative Properties of Pair-instability Supernova Explosions

We present non-LTE time-dependent radiative-transfer simulations of pair-instability supernovae (PISNe) stemming from red-supergiant (RSG), blue-supergiant (BSG) and Wolf-Rayet (WR) star rotation-free progenitors born in the mass range 160-230Msun, at 10^-4 Zsun. Although subject to uncertainties in convection and stellar mass-loss rates, our initial conditions come from physically-consistent models that treat evolution from the main-sequence, the onset of the pair-production instability, and the explosion phase. With our set of input models characterized by large 56Ni and ejecta masses, and large kinetic energies, we recover qualitatively the Type II-Plateau, II-peculiar, and Ib/c light-curve morphologies, although they have larger peak bolometric luminosities (~10^9 to 10^10 Lsun) and a longer duration (~200d). We discuss the spectral properties for each model during the photospheric and nebular phases, including Balmer lines in II-P and II-pec at early times, the dominance of lines from intermediate-mass-elements (IMEs) near the bolometric maximum, and the strengthening of metal line blanketing thereafter. Having similar He-core properties, all models exhibit similar post-peak spectra that are strongly blanketed by FeII and FeI lines, characterized by red colors, and that arise from photospheres/ejecta with a temperature of <4000K. Combined with the modest line widths after bolometric peak, these properties contrast with those of known super-luminous SNe suggesting that PISNe are yet to be discovered. Being reddish, PISNe will be difficult to observe at high redshift except when they stem from RSG explosions, in which case they could be used as metallicity probes and distance indicators.

astro-ph.SR↗

arXiv2012

Super-luminous supernovae: 56Ni power versus magnetar radiation

Much uncertainty surrounds the origin of super-luminous supernovae (SNe). Motivated by the discovery of the Type Ic SN2007bi, we study its proposed association with a pair-instability SN (PISN). We compute stellar-evolution models for primordial ~200Msun stars, simulating the implosion/explosion due to the pair-production instability, and use them as inputs for detailed non-LTE time-dependent radiative-transfer simulations that include non-local energy deposition and non-thermal processes. We retrieve the basic morphology of PISN light curves from red-supergiant, blue-supergiant, and Wolf-Rayet (WR) star progenitors. Although we confirm that a progenitor 100Msun helium core (PISN model He100) fits well the SN2007bi light curve, the low ratios of its kinetic energy and 56Ni mass to the ejecta mass, similar to standard core-collapse SNe, conspire to produce cool photospheres, red spectra subject to strong line blanketing, and narrow line profiles, all conflicting with SN2007bi observations. He-core models of increasing 56Ni-to-ejecta mass ratio have bluer spectra, but still too red to match SN2007bi, even for model He125 -- the effect of 56Ni heating is offset by the associated increase in blanketing. In contrast, the delayed injection of energy by a magnetar represents a more attractive alternative to reproduce the blue, weakly-blanketed, and broad-lined spectra of super-luminous SNe. The extra heat source is free of blanketing and is not explicitly tied to the ejecta. Experimenting with a ~9Msun WR-star progenitor, initially exploded to yield a ~1.6B SN Ib/c ejecta but later influenced by tunable magnetar-like radiation, we produce a diversity of blue spectral morphologies reminiscent of SN2007bi, the peculiar Type Ib SN2005bf, and super-luminous SN2005ap-like events.

astro-ph.SR↗

arXiv2012

The influence of rotation on optical emission profiles of O stars

We study the formation of photospheric emission lines in O stars and show that the rectangular profiles, sometimes double peaked, that are observed for some stars are a direct consequence of rotation, and it is unnecessary to invoke an enhanced density structure in the equatorial regions. Emission lines, such as N IV 4058 and the N III 4634-4640-4642 multiplet, exhibit non-standard "limb darkening" laws. The lines can be in absorption for rays striking the center of the star and in emission for rays near the limb. Weak features in the flux spectrum do not necessarily indicate an intrinsically weak feature -- instead the feature can be weak because of cancellation between absorption in "core" rays and emission from rays near the limb. Rotation also modifies line profiles of wind diagnostics such as He II 4686 and Halpha and should not be neglected when inferring the actual stratification, level and nature of wind structures.

astro-ph.SR↗

arXiv2012

Non-thermal excitation and ionization in supernovae

We incorporate non-thermal excitation and ionization processes arising from non-thermal electrons that result from γ-ray energy deposition, into our radiative transfer code CMFGEN. The non-thermal electron distribution is obtained by solving the Spencer-Fano equation using the procedure of Kozma & Fransson (1992). We applied the non-thermal calculations to the blue supergiant explosion model whose early evolution was studied in Dessart & Hillier (2010). Non-thermal processes generally increase excitation and ionization and decrease the temperature of the ejecta. We confirm that non-thermal processes are crucial for modeling the nebular spectra. Both optical HI and HeI lines are significantly strengthened. While optical HeI lines are not easily discerned in observational spectra due to severe blending with other lines, HeI 2.058 μm provides an excellent opportunity to infer the influence of non-thermal processes. We also discuss the processes controlling the formation of the HeI lines during the nebular epoch. Most lines of other species are only slightly affected. We also show that the inclusion of FeI has substantial line-blanketing effects on the optical spectra. Our model spectra and synthetic light curves are compared to the observations of SN 1987A. The spectral evolution shows broad agreement with the observations, especially Hα. The uncertainties of the non-thermal solver are studied, and are expected to be small. With this new addition of non-thermal effects in CMFGEN, we now treat all known important processes controlling the radiative transfer of a supernova ejecta, whatever the type and the epoch.

astro-ph.SR↗

arXiv2012

On the nature of supernovae Ib and Ic

Using non-LTE time-dependent radiative-transfer calculations, we study the impact of mixing and non-thermal processes associated with radioactive decay on SN IIb/Ib/Ic light curves (LCs) and spectra. Starting with short-period binary models of \leq5Msun He-rich stars (18-25Msun on the main-sequence), we produce 1.2B ejecta which we artificially mix to alter the chemical stratification. While the total 56Ni mass influences the LC peak, the spatial distribution of 56Ni, controlled by mixing processes, impacts both the multi-band LCs and spectra. With enhanced mixing, our synthetic LCs start their post-breakout re-brightening phase earlier, follow a more gradual rise to peak, appear redder, and fade faster after peak due to enhanced gamma-ray escape. Non-thermal electrons, crucial for the production of HeI lines, deposit a dominant fraction of their energy as heat. Because energy deposition is generally local well after the LC peak, the broad HeI lines characteristic of maximum-light SN IIb/Ib spectra require mixing that places 56Ni and helium nuclei to within a gamma-ray mean-free-path. This requirement indicates that SNe IIb and Ib most likely arise from the explosion of stripped-envelope massive stars (main-sequence masses \leq25Msun) that have evolved through mass-transfer in a binary system, rather than from more massive single WR stars. In contrast, the lack of HeI lines in SNe Ic may result from a variety of causes: A genuine helium deficiency; strongly-asymmetric mixing; weak mixing; or a more massive, perhaps single, progenitor characterized by a larger oxygen-rich core. Our models, subject to different mixing magnitudes, can produce a variety of SN types, including IIb, IIc, Ib, and Ic. As it is poorly constrained by explosion models, mixing challenges our ability to infer the progenitor and explosion properties of SNe IIb/Ib/Ic.

astro-ph.SR↗

arXiv2012

On the influence of the companion star in Eta Carinae: 2D radiative transfer modeling of the ultraviolet and optical spectra

We present 2D radiative transfer modeling of the Eta Carinae binary system accounting for the presence of a wind-wind collision (WWC) cavity carved in the optically-thick wind of the primary star. By comparing synthetic line profiles with HST/STIS spectra obtained near apastron, we show that the WWC cavity has a strong influence on multi-wavelength diagnostics. This influence is regulated by the modification of the optical depth in the continuum and spectral lines. We find that H-alpha, H-beta, and Fe II lines are the most affected by the WWC cavity, since they form over a large volume of the primary wind. These spectral lines depend on latitude and azimuth since, according to the orientation of the cavity, different velocity regions of a spectral line are affected. For 2D models with orientation corresponding to orbital inclination angle 110deg < i < 140deg and longitude of periastron 210deg < omega < 330deg, the blueshifted and zero-velocity regions of the line profiles are the most affected. These orbital orientations are required to simultaneously fit the UV and optical spectrum of Eta Car, for a half-opening angle of the cavity in the range 50-70deg. We find that the excess P-Cygni absorption seen in H-alpha, H-beta and optical Fe II lines in spherical models becomes much weaker or absent in the 2D models, in agreement with the observations. The observed UV spectrum of Eta Car, dominated by Fe II absorption lines, is superbly reproduced by our 2D cavity models. Small discrepancies still remain, as H-gamma and H-delta absorptions are overestimated by our models. We suggest that photoionization of the wind of the primary by the hot companion star is responsible for the weak absorption seen in these lines. Our CMFGEN models indicate that the primary star has a mass-loss rate of 8.5x10e-4 Msun/yr and wind terminal velocity of 420 km/s around the 2000 apastron.

astro-ph.SR↗

arXiv2012

Time Dependent Radiative Transfer Calculations for Supernovae

In previous papers we discussed results from fully time-dependent radiative transfer models for core-collapse supernova (SN) ejecta, including the Type II-peculiar SN 1987A, the more "generic" SN II-Plateau, and more recently Type IIb/Ib/Ic SNe. Here we describe the modifications to our radiative modeling code, CMFGEN, which allowed those studies to be undertaken. The changes allow for time-dependent radiative transfer of SN ejecta in homologous expansion. In the modeling we treat the entire SN ejecta, from the innermost layer that does not fall back on the compact remnant out to the progenitor surface layers. From our non-LTE time-dependent line-blanketed synthetic spectra, we compute the bolometric and multi-band light curves: light curves and spectra are thus calculated simultaneously using the same physical processes and numerics. These upgrades, in conjunction with our previous modifications which allow the solution of the time dependent rate equations, will improve the modeling of SN spectra and light curves, and hence facilitate new insights into SN ejecta properties, the SN progenitors and the explosion mechanism(s). CMFGEN can now be applied to the modeling of all SN types

astro-ph.SR↗

arXiv2011

Multi-Wavelength Implications of the Companion Star in Eta Carinae

Eta Carinae is considered to be a massive colliding wind binary system with a highly eccentric (e \sim 0.9), 5.54-yr orbit. However, the companion star continues to evade direct detection as the primary dwarfs its emission at most wavelengths. Using three-dimensional (3-D) SPH simulations of Eta Car's colliding winds and radiative transfer codes, we are able to compute synthetic observables across multiple wavebands for comparison to the observations. The models show that the presence of a companion star has a profound influence on the observed HST/STIS UV spectrum and H-alpha line profiles, as well as the ground-based photometric monitoring. Here, we focus on the Bore Hole effect, wherein the fast wind from the hot secondary star carves a cavity in the dense primary wind, allowing increased escape of radiation from the hotter/deeper layers of the primary's extended wind photosphere. The results have important implications for interpretations of Eta Car's observables at multiple wavelengths.

astro-ph.SR↗

arXiv2011

A Spectropolarimetric Comparison of the Type II-Plateau Supernovae SN 2008bk and SN 2004dj

The Type II-Plateau supernova (SN II-P) SN 2004dj was the first SN II-P for which spectropolarimetry data were obtained with fine temporal sampling before, during, and after the fall off of the photometric plateau -- the point that marks the transition from the photospheric to the nebular phase in SNe II-P. Unpolarized during the plateau, SN 2004dj showed a dramatic spike in polarization during the descent off of the plateau, and then exhibited a smooth polarization decline over the next two hundred days. This behavior was interpreted by Leonard et al. (2006) as evidence for a strongly non-spherical explosion mechanism that had imprinted asphericity only in the innermost ejecta. In this brief report, we compare nine similarly well-sampled epochs of spectropolarimetry of the Type II-P SN 2008bk to those of SN 2004dj. In contrast to SN 2004dj, SN 2008bk became polarized well before the end of the plateau and also retained a nearly constant level of polarization through the early nebular phase. Curiously, although the onset and persistence of polarization differ between the two objects, the detailed spectropolarimetric characteristics at the epochs of recorded maximum polarization for the two objects are extremely similar, feature by feature. We briefly interpret the data in light of non-Local-Thermodynamic Equilibrium, time-dependent radiative-transfer simulations specifically crafted for SN II-P ejecta.

astro-ph.SR↗

arXiv2011

A Method for the Study of Accretion Disk Emission in Cataclysmic Variables I: The Model

We have developed a spectrum synthesis method for modeling the UV emission from the accretion disk from cataclysmic variables (CVs). The disk is separated into concentric rings, with an internal structure from the Wade & Hubeny disk-atmosphere models. For each ring, a wind atmosphere is calculated in the co-moving frame with a vertical velocity structure obtained from a solution of the Euler equation. Using simple assumptions, regarding rotation and the wind streamlines, these 1D models are combined into a single 2.5D model for which we compute synthetic spectra. We find that the resulting line and continuum behavior as a function of the orbital inclination is consistent with the observations, and verify that the accretion rate affects the wind temperature, leading to corresponding trends in the intensity of UV lines. In general, we also find that the primary mass has a strong effect on the P-Cygni absorption profiles, the synthetic emission line profiles are strongly sensitive to the wind temperature structure, and an increase in the mass loss rate enhances the resonance line intensities. Synthetic spectra were compared with UV data for two high orbital inclination nova-like CVs - RW Tri and V347 Pup. We needed to include disk regions with arbitrary enhanced mass loss to reproduce reasonably well widths and line profiles. This fact and a lack of flux in some high ionization lines may be the signature of the presence of density enhanced regions in the wind, or alternatively, may result from inadequacies in some of our simplifying assumptions.

astro-ph.SR↗

arXiv2011

On the nature of the prototype LBV AG Carinae II. Witnessing a massive star evolving close to the Eddington and bistability limits

We show that the significantly different effective temperatures (Teff) achieved by the luminous blue variable AG Carinae during the consecutive visual minima of 1985-1990 (Teff= 22,800 K) and 2000-2001 (Teff=17,000 K) place the star on different sides of the bistability limit, which occurs in line-driven stellar winds around Teff ~ 21,000 K. Decisive evidence is provided by huge changes in the optical depth of the Lyman continuum in the inner wind as Teff changes during the S Dor cycle. These changes cause different Fe ionization structures in the inner wind. The bistability mechanism is also related to the different wind parameters during visual minima: the wind terminal velocity was 2-3 times higher and the mass-loss rate roughly two times smaller in 1985-1990 than in 2000-2003. We obtain a projected rotational velocity of 220 +- 50 km/s during 1985-1990 which, combined with the high luminosity (L=1.5x10^6 Lsun), puts AG Car extremely close to the Eddington limit modified by rotation (Omega-Gamma limit): for an inclination angle of 90 deg, Gamma_Omega > 1.0 for M< 60 Msun. Based on evolutionary models and mass budget, we obtain an initial mass of ~ 100 Msun and a current mass of 60-70 Msun for AG Car. Therefore, AG Car is close to, if not at, the Omega-Gamma limit during visual minimum. Assuming M=70 Msun, we find that Gamma_Omega decreases from 0.93 to 0.72 as AG Car expands toward visual maximum, suggesting that the star is not above the Eddington limit during maximum phases.

astro-ph.SR↗