University of Minnesota
School of Physics & Astronomy


Understanding Massive Stars

A snapshot of one of Kashi's simulations of a star undergoing a giant eruption. The figure shows a 2D slice of the density (in g/cm^3; logarithmic scale), 70 days after the onset of the eruption. Units of distance are in solar radii. Velocity vectors are superimposed on the figure, showing the stellar wind in the outer regions, and turbulent motion in inner layers.

Amit Kashi is a postdoc working on very massive stars with Kris Davidson and Roberta Humphreys. Kashi is working on stars that are typically a hundred times more massive than the Sun. Kashi says that these very massive stars are fundamentally different from others in that they are very unstable due to their size. The radiation from these objects, coming from the nuclear burning core, is so intense that it almost overcomes the gravity that is holding them together.

They can reach a stage in evolution in which they become sensitive to internal instabilities or external disruptions, such as a binary companion moving in close proximity. “They are barely held together. There is a delicate balance between radiation force, centrifugal force and gravity. An instability in such a star can unbind a part of it and cause a giant eruption.” These eruptions are so large that they are often mistaken for supernova by observers because the luminosity is similar. As such, they earned the name “SN Impostors”. The difference is that after a giant eruption the star survives and continues to function as a star, unlike a star that undergoes a supernova which leaves behind a neutron star or a black hole.

Kashi studies these so called “SN Impostor” events using computer simulations to try understand their underlying physics. “First I pretend the eruption has happened, without being concerned about the cause. I artificially create a giant eruption to see how the star recovers, and returns to be a star in hydrostatic equilibrium. This process occurs in a few centuries. This is a very short time in a life of a star, even massive one, but major changes occur during that period.” Recently, he was able to discoverdiscovered that the mechanism byin which the very massive stars eject winds after an eruption is related to internalcaused by pulsation in the star. The pulsation is that is induced by radiation attempting to escape from the center while interacting with the gas inside the star. His simulations also successfully reproduced observations of Eta Carinae from the last 15 years that showed that the mass flux from the star has dropped.

Kashi is also trying to understand what causes giant eruptions. “We don’t even know if the trigger of the giant eruptions is in the envelope of the star or in the core. I am trying different models to find that out.” Kashi says the entire topic has not been explored by researchers because they have so few examples to study. He is looking at the handful of SN Impostors in our gGalaxy, and only a few in other galaxies. Though the observation quality is relatively poor, it has been enough that some theoretical models have been proposed. One idea is that a binary companion in a very eccentric orbit passed very close to the star and tore out a piece of it. “Eta Carinae has such a companion, and I think that this companion when it passes very close to the more massive star has exerted tidal forces on the envelope of the star.”

Kashi says that it is difficult to extrapolate this model onto other stars because they are in distant galaxies and haven’t been as well observed. One example used is SN 2009ip, a famous supernova imposter, which erupted in 2009. It was thought to be a supernova, but in 2011-2012 it erupted again so it obviously could not have been one. By studying the light curve (velocity as a function of time) he has shown that the data could fit very nicely into the binary induced eruption model. “We still don’t know whether the events in 2012 were terminal or not. Material from the eruption is blocking the view. In one of my papers I have a model that suggests that those 2012 eruptions were non-terminal. We will have to wait a few more years for the stellar environment to clear in order to find out.”