Sensitivity of H-alpha Routine

The H-alpha routine in the code determines the amount of H-alpha emission (neutral hydrogen de-excitation line from level 3 to level 2). It is dependent on electron number density, temperature, ionization fraction, and hydrogen density. There are 2 main components to this emission: a recombination term and a collisional excitation term.

Collisional excitation is what most people usually think about when talking about H-alpha emission: electrons collide with neutral hydrogen and the excited hydrogen atom goes through the H-alpha transition. The recombination term comes into play when you have ionized hydrogen that wants to recombine. When it recombines, there is a nonzero probability that it goes through the H-alpha transition. So, if you have a lot of ionized hydrogen at a low temperature, there should be a lot of recombination and the recombination term can dominate the H-alpha emission.

Below are two plots showing the H-alpha emission as a function of temperature for two different ionization fractions. The hydrogen number density is the same in both plots, and for simplicity the gas was assumed to be all hydrogen. nH = 5560 was chosen because it is the average of the post-shock densities of two simulations which I will talk about.

Notice how in the x=0.99 plot, the recombination term is much stronger and dominates the H-alpha at low temperatures longer. This is because 99% ionization is unrealistically high for these temperatures, so the gas should rapidly recombine and thus go through many more H-alpha transitions.


3D Mach stem simulations

For reference, see ehansen02242015.

For the 3D Mach stem simulations, we noticed a significant decrease in H-alpha emission from the M = 10 runs to the M = 5 runs. The best way to explain the difference is by comparing the temperatures in a couple of these runs. Left is M = 5, right is M = 10, both at separation distances of 3 rclump.

I measured the temperatures in the bow shock intersection regions and got about 6900 K for M = 5, and 19,700 K for M = 10. This is approximately a factor of 3 which seems reasonable given that postshock temperatures are approximately proportional to vshock2

These simulations have relatively low ionization fractions throughout, and at the bow shock intersection points, the ionization remains near 0.01 for both runs. So we should look at the x = 0.01 H-alpha plot.

It would appear that these simulations have hit the sweet spot. There is a sharp increase in the H-alpha going from 6900 K to 19,700 K. This is where the collisional excitation term takes over and dominates. I estimate a jump of about 2 or 3 orders of magnitude which is consistent with the emission maps.

Also, look at the maximum temperatures in the simulations, and then look at the x = 0.01 H-alpha plot. This explains why the M = 5 model does not show a wide range of H-alpha intensities like the M = 10 model does.

In conclusion, the H-alpha routine is working just fine. The parameters I chose for the simulations just happen to illustrate the temperature sensitivity in our code.

Attachments (3)

Download all attachments as: .zip

Comments

No comments.