Things that go bump in the night
Hello again, and welcome back to my PRACE SoHPC blog! In case you missed my introductory blog post, you can find it here. Last time around, I mentioned that my project this summer at the CINECA HPC facility in Bologna deals with the visualisation of supernova explosions in a magnetised inhomogeneous ambient environment. Over the last few weeks, I spent a considerable amount of time familiarising myself with the Galileo cluster at CINECA, learning how to edit and run models on the supercomputer. The supernova explosion models were run using The PLUTO Code, a freely-distributed software for simulating astrophysical fluid dynamics.
Starting with the basics
Once I was comfortable implementing sample models on GALILEO, I began running the simulation of an exploding supernova (SN) and the subsequent supernova remnant (SNR). I firstly considered the case of a spherically symmetric supernova blast wave expanding in a uniform circumstellar medium (CSM). The CSM contains the mass-loss from the progenitor star that was stripped away by the stellar wind in the years leading up to the explosion.
When the expanding SN ejecta interacts with the CSM, some of the material will continue to propagate outward in what is known as the forward shock. Conversely, some of the material will travel backwards into the freely expanding ejecta after colliding with the CSM. This is known as the reverse shock. The diagram below shows a 2-D slice of the density profile of the remnant at a time of 1,000 years since the supernova explosion, and each of the regions are easily distinguishable.
One of the input parameters that was defined in the model was the density of each of the regions. In the image above, the region with the highest density is the unshocked ejecta (blue) at the centre of the remnant. If we then move outwards, the density drops off as we go to green and then to red. The orange/yellow colour of the unshocked CSM represents low densities (as expected, as this region is much larger and has less mass than the remnant).
For a better sense of the evolution of the blast wave, scroll down and check out the short video below!
Making things more life-like
When I was happy with the shape and evolution of the symmetric blast wave, I began to introduce asymmetries into my model to make it more realistic in comparison to SNR’s that we have observed (such as Cassiopeia A). As a first step, I added a clump of material to one side of the initial remnant. I set the density of this clump to be 10 times greater than the rest of the SN ejecta, and the velocity of the clump was set to 5 times the velocity of the ejecta. The image below shows that the effect of this clump is to cause a protrusion of the outer remnant, distorting the remnant compared to the spherical case that we saw earlier.
The video below shows the comparison between the two models as they evolve over time.
For the remainder of the summer, I will continue to introduce more asymmetries in my models to better resemble real-life SNR’s. This can be achieved by adding more clumps, all of different sizes, velocities, and densities. Another possibility is to add asymmetries in the ambient environment, and this area is being investigated by my colleague, Cathal Maguire. Check out his latest blog here, where he attempts to include a torus (doughnut-shaped) feature within the CSM.
I will also spend more time analysing my models using the 3-D visualisation software Paraview. Finally, I plan to upload my models to SKETCHFAB, a platform for publicly sharing 3-D models. Exciting times lie ahead!