Mass Loss on an HR diagram

[Lizard]

TL;DR

Understanding drop in luminosity during mass loss in HR diagram.

I am trying to understand the red line in the below diagram. This is Case B mass transfer (on a thermal timescale, through Roche-Lobe overflow).
Mass transfer begins at point B. Because this is mass transfer on a thermal timescale, it must be dynamically stable, meaning that the radius of the donor star must initially shrink in response to mass loss. However, I don't quite understand what is happening on the thermal timescale.

The first thing I don't understand is why the luminosity drops from point B to point C. It seems like this must be due to expansion from a perturbation to thermal equilibrium, but I don't understand why expansion is the response to mass loss.

The second thing I don't understand is why luminosity increases from point C to point D. The star reaches its maximum mass-loss rate at point C. Why would a decrease in mass loss rate result in an increase in luminosity? Why don't the arguments that describe the decrease in luminosity from B to C apply from C to D?

 

[Ken G]

Your questions really require someone who is an expert on binary mass transfer, not to mention someone who understands the particular assumptions being made in the simulation you are citing the graph from (binary mass transfer cannot be simulated from first principles, it requires uncertain ad hoc assumptions, like that the mass transfer is "conservative", but there are many others as well). Didn't that graph come with a deeper explanation of what was happening? Is there a citation that could help us all understand where this comes from?

In partial answer to why the curves go to lower L and then back up again, note the importance of the dotted path. That is what the 10 solar mass star would have done without mass transfer, and for some reason, after the star has lost a bunch of its mass, it returns to that same path, though it then goes blue instead of red. Since point C is when mass loss is fastest, it is probably the place where the stars are closest. Conservation of angular momentum requires that for the "conservative" case considered here, the stars are at their closest when both masses are 9 solar masses, and the closest point should be the peak in mass transfer, so that must be point C. After that, additional mass transfer pushes the stars apart, and this causes the mass transfer rate to drop precipitously. So from C to D, there is a very slow mass transfer, and that curve should take much much longer than B to C. I don't know why the L drops from B to C, but remember this is happening so fast that the star does not have time to fully adjust to its lower mass, so this would be a star very far from a normal type of equilibrium. By the time it gets to point D, it seems to have returned to where it would have normally been in its evolution after it initiated core helium burning. The star has way less mass on the red curve at D than the dotted line at D, so I really have no idea why it has the same L and T, but you see the same thing again at point E. So that would be my question looking at this, the 10 solar mass star probably has a mass more like 1 solar mass by the time it gets to points D and E, so why do those points have any connection at all to the dotted curve for a 10 solar mass star that is also undergoing core helium burning?

I am also mystified by the same thing you are, which is why is the 9 solar mass star at point C so much larger than it was when it was a 10 solar mass star at point B. At point B it filled its Roche lobe, so its radius should have been something like 10 solar radii and the separation something like 20 solar radii. Then it should have pulled slightly closer at point C, so if anything the star should be slightly smaller, it certainly should not be able to be about 10 times larger like that H-R diagram suggests. It seems like it is completely impossible for it to have that L and T at point C.