Research (updated 25/Jan/2020)
Wind Collision and Accretion Simulations of the Massive Binary System HD 166734
Published in MNRAS arxiv.org/abs/2001.08007
https://academic.oup.com/mnras/article/492/4/5261/5729029?guestAccessKey=8d399301-e4f3-4efa-8acd-adfdad418726
We run hydrodynamic simulations which follow the colliding winds structure of the massive binary system HD 166734 along its binary orbit, and show that close to periastron passage the secondary wind is suppressed and the secondary accretes mass from the primary wind. The system consists two blue supergiants with masses of M1≈39.5 M⊙ and M2≈30.5 M⊙, on a P≃34.538 days orbit with eccentricity of e≈0.618. This close O-O binary with high eccentricity is observed through its orbit in the X-rays, where it shows an unusual long minimum close to periastron passage. We use advanced simulations with wind acceleration and prescription treatment of accretion and simulate the entire orbit at high resolution that captures the instabilities in the winds. We find that the colliding wind structure is unstable even at apastron. As the stars approach periastron passage the secondary wind is quenched by the primary wind and the accretion onto the secondary begins. The accretion phase lasts for ≃12 days, and the amount of accreted mass per cycle we obtain is Macc≃1.3⋅10−8 M⊙. The accretion phase can account for the observed decline in X-ray emission from the system.
Published in MNRAS arxiv.org/abs/2001.08007
https://academic.oup.com/mnras/article/492/4/5261/5729029?guestAccessKey=8d399301-e4f3-4efa-8acd-adfdad418726
We run hydrodynamic simulations which follow the colliding winds structure of the massive binary system HD 166734 along its binary orbit, and show that close to periastron passage the secondary wind is suppressed and the secondary accretes mass from the primary wind. The system consists two blue supergiants with masses of M1≈39.5 M⊙ and M2≈30.5 M⊙, on a P≃34.538 days orbit with eccentricity of e≈0.618. This close O-O binary with high eccentricity is observed through its orbit in the X-rays, where it shows an unusual long minimum close to periastron passage. We use advanced simulations with wind acceleration and prescription treatment of accretion and simulate the entire orbit at high resolution that captures the instabilities in the winds. We find that the colliding wind structure is unstable even at apastron. As the stars approach periastron passage the secondary wind is quenched by the primary wind and the accretion onto the secondary begins. The accretion phase lasts for ≃12 days, and the amount of accreted mass per cycle we obtain is Macc≃1.3⋅10−8 M⊙. The accretion phase can account for the observed decline in X-ray emission from the system.

The YSO Transient ASASSN-13db
Amit Kashi, Amir M. Michaelis, Leon Feigin
The low mass star ASASSN-13db experienced an EXor outburst in 2013, which identified it as a Young Stellar Object (YSO). Then, from 2014 to 2017 it had another outburst, longer and more luminous than the earlier. We analyze the observations of the second outburst, and compare it to eruptions of Intermediate Luminosity Optical Transients (ILOTs). We show that the decline of the light curve is almost identical to that of the V838 Mon, a prototype of a type of ILOT known as Luminous Red Nova (LRN). This similarity becomes conspicuous when oscillations that are associated with rotation are filtered out from the light curve of ASASSN-13db. We suggest that the eruption was the result of accretion of a proto-planet of a few Earth masses. The proto-planet was shredded by tidal forces before it was accreted onto the YSO, releasing gravitational energy that powered the outburst for ≈800 days, and ended in a ≈55 days decline phase. When the accretion material started depleting the accretion rate lowered and the eruption light curve declined for almost two months. Then it exhausted completely, creating a sharp break in the light curve. Another possibility is that the mass was a result of an instability in the proto-planetary disk that lead to a large episode of accretion from an inner viscous disk. We find that the variation of the temperature of the outburst is consistent with the surface temperature expected from a depleted viscous accretion disk. The 2014-2017 outburst of ASASSN-13db may be the least energetic ILOT to have been discovered to date, with an energy budget of only ≈1e42 erg.
Link to the paper
Amit Kashi, Amir M. Michaelis, Leon Feigin
The low mass star ASASSN-13db experienced an EXor outburst in 2013, which identified it as a Young Stellar Object (YSO). Then, from 2014 to 2017 it had another outburst, longer and more luminous than the earlier. We analyze the observations of the second outburst, and compare it to eruptions of Intermediate Luminosity Optical Transients (ILOTs). We show that the decline of the light curve is almost identical to that of the V838 Mon, a prototype of a type of ILOT known as Luminous Red Nova (LRN). This similarity becomes conspicuous when oscillations that are associated with rotation are filtered out from the light curve of ASASSN-13db. We suggest that the eruption was the result of accretion of a proto-planet of a few Earth masses. The proto-planet was shredded by tidal forces before it was accreted onto the YSO, releasing gravitational energy that powered the outburst for ≈800 days, and ended in a ≈55 days decline phase. When the accretion material started depleting the accretion rate lowered and the eruption light curve declined for almost two months. Then it exhausted completely, creating a sharp break in the light curve. Another possibility is that the mass was a result of an instability in the proto-planetary disk that lead to a large episode of accretion from an inner viscous disk. We find that the variation of the temperature of the outburst is consistent with the surface temperature expected from a depleted viscous accretion disk. The 2014-2017 outburst of ASASSN-13db may be the least energetic ILOT to have been discovered to date, with an energy budget of only ≈1e42 erg.
Link to the paper
Simulating the response of the secondary star of Eta Carinae to mass accretion at periastron passage
We use high resolution 3D hydrodynamical simulations to quantify the amount of mass accreted onto the secondary star of the binary system Eta Carinae, exploring two sets of stellar masses which had been proposed for the system, the conventional mass model (M 1 = 120 M ⊙ and M 2 = 30 M ⊙ ) and the high mass model (M 1 = 170 M ⊙ and
M 2 = 80 M ⊙ ). The system consists of two very massive stars in a highly eccentric orbit. Every cycle close to periastron passage the system experiences a spectroscopic event during which many lines change their appearance, accompanied by a decline in x-ray emission associated with the destruction wind collision structure and accretion of the primary wind onto the secondary. We take four different numerical approaches to simulate the response of the secondary wind to accretion, each affects the mass loss rate of the secondary ifferently, and in turn determines the amount of accreted mass. The high mass model gives for most approaches much more accreted gas and longer accretion phase. We find that the effective temperature of the secondary can be significantly reduced due to accretion. We also test different eccentricity values and a higher primary mass loss rate and find their effect on the duration of the spectroscopic event. We conclude that the high mass model is better compatible with the amount of accreted mass, ≈ 3 × 10 −6 M ⊙ , required for explaining the reduction in secondary ionization photons during the spectroscopic event and compatible with its observed duration. LINK TO THE PAPER Download IAU346 Conference Proceedings |
Videos of density profiles on the orbital plane:
|
|

Type II intermediate-luminosity optical transients (ILOTs)
Amit Kashi and Noam Soker
We propose that in a small fraction of intermediate luminosity optical transients (ILOTs) powered by a strongly interacting binary system, the ejected mass in the equatorial plane can block the central source from our line of sight. We can therefore observe only radiation that is reprocessed by polar outflow, much as in type~II active galactic nuclei (AGN). An ejection of Mej,e=1e-4 M⊙ (1 M⊙) at 30 degrees from the equatorial plane and at a velocity of ve=100 km s−1 will block the central source in the NIR for about 5 years (500 years). During that period of time the object might disappear in the visible band, and be detected only in the IR band due to polar dust. We raise the possibility that the recently observed disappearance of a red giant in the visible, designated N6946-BH1, is a type~II ILOT rather than a failed supernova. For this case we estimate that the ejected mass in the polar direction was Mej,p≈1e−3 M⊙. Our scenario predicts that this event should reinstate its visible emission in several decades.
Read the paper: https://arxiv.org/abs/1611.05855
Amit Kashi and Noam Soker
We propose that in a small fraction of intermediate luminosity optical transients (ILOTs) powered by a strongly interacting binary system, the ejected mass in the equatorial plane can block the central source from our line of sight. We can therefore observe only radiation that is reprocessed by polar outflow, much as in type~II active galactic nuclei (AGN). An ejection of Mej,e=1e-4 M⊙ (1 M⊙) at 30 degrees from the equatorial plane and at a velocity of ve=100 km s−1 will block the central source in the NIR for about 5 years (500 years). During that period of time the object might disappear in the visible band, and be detected only in the IR band due to polar dust. We raise the possibility that the recently observed disappearance of a red giant in the visible, designated N6946-BH1, is a type~II ILOT rather than a failed supernova. For this case we estimate that the ejected mass in the polar direction was Mej,p≈1e−3 M⊙. Our scenario predicts that this event should reinstate its visible emission in several decades.
Read the paper: https://arxiv.org/abs/1611.05855
Accretion at the periastron passage of Eta Carinae
A. Kashi
We present high resolution numerical simulations of the colliding wind system η Carinae, showing accretion onto the secondary star close to periastron passage. Our hydro-dynamical simulations include self gravity and radiative cooling. The smooth stellar winds collide and develop instabilities, mainly the non-linear thin shell instability, and form filaments and clumps. We find that a few days before periastron passage the dense filaments and clumps flow towards the secondary as a result of its gravitational attraction, and reach the zone where we inject the secondary wind. We run our simu lations for the conventional stellar masses, M1 = 120 Msun and M2 = 30 Msun, and for a high mass model, M1 = 170 Msun and M2 = 80 Msun , that was proposed to better fit the history of giant eruptions. As expected, the simulations results show that theaccretion processes is more pronounced for a more massive secondary star.
Download the paper (8MB, PDF):
A. Kashi
We present high resolution numerical simulations of the colliding wind system η Carinae, showing accretion onto the secondary star close to periastron passage. Our hydro-dynamical simulations include self gravity and radiative cooling. The smooth stellar winds collide and develop instabilities, mainly the non-linear thin shell instability, and form filaments and clumps. We find that a few days before periastron passage the dense filaments and clumps flow towards the secondary as a result of its gravitational attraction, and reach the zone where we inject the secondary wind. We run our simu lations for the conventional stellar masses, M1 = 120 Msun and M2 = 30 Msun, and for a high mass model, M1 = 170 Msun and M2 = 80 Msun , that was proposed to better fit the history of giant eruptions. As expected, the simulations results show that theaccretion processes is more pronounced for a more massive secondary star.
Download the paper (8MB, PDF):

etacaracc2.pdf | |
File Size: | 8038 kb |
File Type: |
Simulating the Onset of Grazing Envelope Evolution of Binary Stars
S. Shiber, A.Kashi and N. Soker
We present the first three-dimensional gas-dynamical simulations of the grazing envelope evolution (GEE) of stars, with the goal of exploring the basic flow properties and the role of jets at the onset of the GEE. In the simulated runs, a secondary main-sequence star grazes the envelope of the primary asymptotic giant branch (AGB) star. The orbit is circular at the radius of the AGB primary star on its equator. We inject two opposite jets perpendicular to the equatorial plane from the location of the secondary star, and follow the evolution for several orbital periods. We explore the flow pattern by which the jets eject the outskirts of the AGB envelope. After one orbit the jets start to interact with gas ejected in previous orbits and inflate hot low-density bubbles.

mnras-2016-shiber-mnrasl-slw208.pdf | |
File Size: | 1215 kb |
File Type: |
The ILOT Club
I am maintaining a webpage devoted to Intermediate Luminosity Optical Transient (ILOTS). Most broadly, these are astrophysical stars or stellar systems that experience outbursts with a peak luminosity somewhere between the peak luminosity of novae and supernovae.
See the page at http://phsites.technion.ac.il/soker/ilot-club/
Below is the Energy-Time Diagram used to classify ILOTs.
See the page at http://phsites.technion.ac.il/soker/ilot-club/
Below is the Energy-Time Diagram used to classify ILOTs.
Numerical Simulation of Giant Non-Terminal Eruptions and Recovery in Very Massive Stars
Amit Kashi, Kris Davidson and Roberta Humphreys
We use a hydro-and-radiative-transfer code to explore the behaviour of a very massive star after a giant eruption, i.e., following a supernova impostor event. Beginning with a reasonable model for an evolved VMS, we simulate the change of state caused by a giant eruption via two methods that explicitly conserve total energy: 1. Synthetically removing outer layers of mass while reducing the energy of the inner layers. 2. Synthetically transferring energy from the core to the outer layers, an operation that automatically causes mass ejection. Our focus is on the aftermath, not the poorly understood eruption itself. Then, using a radiation-hydrodynamic code in 1D with realistic opacities and convection, interior disequilibrium state is followed for about 300 years. Typically the star develops a 400 km/s wind with a mass loss rate that begins around 0.1 Msun/yr and gradually decreases. This outflow is driven by -mechanism radial pulsations. In some cases a plateau in the mass loss rate may persist about 200 years, while other cases are more like Car's known history.
Amit Kashi, Kris Davidson and Roberta Humphreys
We use a hydro-and-radiative-transfer code to explore the behaviour of a very massive star after a giant eruption, i.e., following a supernova impostor event. Beginning with a reasonable model for an evolved VMS, we simulate the change of state caused by a giant eruption via two methods that explicitly conserve total energy: 1. Synthetically removing outer layers of mass while reducing the energy of the inner layers. 2. Synthetically transferring energy from the core to the outer layers, an operation that automatically causes mass ejection. Our focus is on the aftermath, not the poorly understood eruption itself. Then, using a radiation-hydrodynamic code in 1D with realistic opacities and convection, interior disequilibrium state is followed for about 300 years. Typically the star develops a 400 km/s wind with a mass loss rate that begins around 0.1 Msun/yr and gradually decreases. This outflow is driven by -mechanism radial pulsations. In some cases a plateau in the mass loss rate may persist about 200 years, while other cases are more like Car's known history.

The Effects on the Photosphere of the Secondary of Eta Car Due to Accretion Close to Periastron Passage
We perform an accretion calculation to estimate the radius and effective temperature of the secondary of Car close to periastron passage. The calculation is based on Bond-Hoyle-Lyttleton accretion with modifications and including other effects such as radiation pressure of the secondary. We find that for a narrow range of eccentricities and primary mass loss rates, an inflated envelope around the secondary can form, resulting in a lower eeffective temperature. We find that low eccentricity e < 0.85 requires very high primary mass loss rate that might be unrealistic. The lower secondary effective temperature causes a change in the spectra of the binary system, preventing from high energy UV photons from reaching the surrounding gas and thus creating the spectroscopic event. As the spectroscopic event does occur, it suggests that these rare conditions actually exist. We find that a variation in the mass loss rate from the primary will not necessarily change the accreted mass (and hence the duration of the event) in the same direction of the variation. However, below a certain primary mass loss rate there will be no accretion at all, and the event would either not happen or look very differently than we know.
The figure shows the photospheric radius, effective temperature, and the secondary's mass accretion rate according to our model. The eccentricity is taken to be e = 0.9 and the secondary mass is M2 = 50Msun. The time is measured from periastron passage. The accretion is limited from below by the radius of the star and from above by the size of the accretion radius. For a lower mass loss than the allowed range rate the secondary will be able to accrete all mass and would not form a thicker photosphere. For a higher mass loss rate than the allowed range the secondary will be able to accrete efficiently, and the gas would not enter the gravitational influence radius of the secondary.
Modeling of Emission Line Profiles and Transfer Functions for the Broad Line Region: Application to a Line-Driven Disk Wind Model
The physical picture of the broad line region (BLR) in active galactic nuclei (AGN) is often described as a photoionized wind launched from the AGN accretion disk. Reverberation mapping of the BLR in AGN is now beginning to be able to reconstruct two-dimensional transfer functions, which describe the distribution of broad-line reverberation time delay as a function of the projected velocity of the line. Theoretical transfer function models for a variety of possible BLR structures are needed in order to interpret these observational results. Here, we compute full two-dimensional transfer functions and synthetic line profiles for hydrodynamic simulation of a line-driven disk. We use several different prescriptions for the line emissivity in the wind. We find that the line shape strongly depends on the opacity of the gas, and in some cases depends weakly on the inclination angle. We find that the shape of the reverberation transfer function can depend strongly on the inclination angle to the observer's line of sight. Previously we showed that the wind in this simulation is virialized over a large distance from the disk. Therefore, observing line shapes similar to those that are presented here, may serve as an indication of a virialized wind in which the broad-line widths can be used to estimate the mass of the central black hole. To check if indeed such similarities exist, we compare the transfer functions and synthetic line profiles to observations and find good agreement between the observed profiles of the optical Balmer lines and the strong UV lines.
Did SN 2009ip really explode in 2012? The SN impostor SN 2009ip experienced a number of outbursts, most recently in September 2012 (outburst 2012b). The several outbursts in 2009–2011 were LBV major outbursts. However, it is not yet widely agreed what was the nature of the 2012a and 2012b outbursts. Some papers attribute the 2012a to a supernova and the 2012b to the collision of the SN ejecta with a previously ejected gas. However, it may well be that what was observed is an explosive ejection of the envelope of a massive progenitor star. In my paper we attribute all the outbursts to periatron passages of an eccentric binary system. We find our interpretation to be supported by a dominant time scale of ∼38 days in the light curve previous to the 2012b outburst. We consider two scenarios. In both scenarios the major 2012b outburst with total (radiated + kinetic) energy of ∼5e49 erg was powered by accretion of ∼2–5 solar masses onto the companion during a periastron passage (the first passage) of the binary system approximately 20 days before the observed maximum of the light curve. In the first scenario, the surviving companion scenario, the companion was not destructed and still exists in the system after the outburst. It ejected partial shells (or collimated outflows or clumps) for two consecutive periastron passages after the major one. The orbital period was reduced from ∼38 days to ∼25 days as a result of the mass transfer process that took place during the first periastron passage. In the second scenario, the merger scenario, some partial shells/clumps were ejected also in a second periastron passage that took place ∼20 days after the first one. After this second periastron passage the companion dived too deep into the LBV envelope to launch more outflows, and merged with the LBV. |

08/2014
Eta Carinae
The spectroscopic event is here!
η Car is a massive binary system consisting of a massive LBV star and a companion star, having together a mass of 150Msun or more. Both stars blow strong winds and those collide creating a wind collision region with interesting physical phenomenons. The light curve of the system shows a 5.54 yr periodicity is observed in all wavelengths. According to most models, the periodicity follows the 5.54 years periodic change in the orbital separation in this highly eccentric, e≃0.9−0.95, binary system. Every period, close to periastron passage where the two stars are at their closest separation, the system undergoes a dramatic variability and many spectroscopic lines change their intensity or disappear.
More about it is explained in a presentation I gave in August 2009 in the International Astronomical Union general assembly in Rio [PDF].
The 2014 event started late July and is still going.
Here are a few interesting links for recent observations of the event:
Eta Carinae
The spectroscopic event is here!
η Car is a massive binary system consisting of a massive LBV star and a companion star, having together a mass of 150Msun or more. Both stars blow strong winds and those collide creating a wind collision region with interesting physical phenomenons. The light curve of the system shows a 5.54 yr periodicity is observed in all wavelengths. According to most models, the periodicity follows the 5.54 years periodic change in the orbital separation in this highly eccentric, e≃0.9−0.95, binary system. Every period, close to periastron passage where the two stars are at their closest separation, the system undergoes a dramatic variability and many spectroscopic lines change their intensity or disappear.
More about it is explained in a presentation I gave in August 2009 in the International Astronomical Union general assembly in Rio [PDF].
The 2014 event started late July and is still going.
Here are a few interesting links for recent observations of the event:
- Strange variability in the UV
- The X-ray light curve (thought to be created in the wind collision region that according to my models is accreted onto the companion during theX-ray minimum.)
More about my research interests:
Simulations of Accretion in Eta Carinae -- a video of my talk (scroll to 1h46m) from a the BRITE conference, Vienna, August 2019
Simulations of Giant Eruptions and Supernova Impostors (PDF of slides, Video of my talk, ESO, Garching, July 2017)
Intermediate luminosity optical transients (see also a video of my talk)
On the connection between Eta Carinae and Transients
Merger – Bursts of Brown-Dwarf and a Planet
Type Ia Supernovae: The Core-Degenerate Scenario
The Virialization of AGN winds
AGN Accretion and Feedback in Cosmological Simulations (see also My AAS poster )
Ph.D. Thesis:
The Periastron Passage of the Binary Star Eta Carinae. Supervisor: Prof. Noam Soker.
Popular level articles:
Universe Today article on my P-Cygni (the star itself, not the line profile) binary model
New Scientist article about my model for the transient SN 2009ip (also available in PDF: NewScientist_SN2009ip.pdf)
Ynet article on Eta Carinae (Hebrew)
An article on Eta Carinae in “Galileo” journal (Hebrew)
Simulations of Accretion in Eta Carinae -- a video of my talk (scroll to 1h46m) from a the BRITE conference, Vienna, August 2019
Simulations of Giant Eruptions and Supernova Impostors (PDF of slides, Video of my talk, ESO, Garching, July 2017)
Intermediate luminosity optical transients (see also a video of my talk)
On the connection between Eta Carinae and Transients
Merger – Bursts of Brown-Dwarf and a Planet
Type Ia Supernovae: The Core-Degenerate Scenario
The Virialization of AGN winds
AGN Accretion and Feedback in Cosmological Simulations (see also My AAS poster )
Ph.D. Thesis:
The Periastron Passage of the Binary Star Eta Carinae. Supervisor: Prof. Noam Soker.
Popular level articles:
Universe Today article on my P-Cygni (the star itself, not the line profile) binary model
New Scientist article about my model for the transient SN 2009ip (also available in PDF: NewScientist_SN2009ip.pdf)
Ynet article on Eta Carinae (Hebrew)
An article on Eta Carinae in “Galileo” journal (Hebrew)