Cosmological galaxy formation simulation software | Hacker News Reader

Cosmological galaxy formation simulation software

tng-project.org

61 pointskrpl8 comments

I like seeing how familiar structures appear at cosmological scale. Long ago I created a webgl visualization of the Millennium Run, an early large-scale cosmological simulation: https://www.asterank.com/galaxies/

It was a nice way to learn about three.js/webgl and how to make many particles performant. There are probably better visualizations out there nowadays.

Would the empty regions be like that end of cause and effect which is predicted for an expanding universe?

No, voids are normal in these types of simulations and are observed in real world observations as well. Some dark energy alternatives actually looked at the effect of voids on our observations, but could not sufficiently explain the observed acceleration.

Maybe it's naive for me. Every time we talk about a simulation, we got an accurate model and we have some model validation to run our model. The next big question is, okay, what can we do with those models? If we can do CFD simulations, we can use that tools to build fast car, to build efficient SpaceX rocket to reduce the cost, but in terms of this cosmology galaxy formation, what can we do with this model?

One simple idea for me is to, okay, is there any significant difference from a large-scale model with at the galax scale and the model we have, let's say, simulating a von Karman vertices because both of them are showing very similar structure. Yes, that's an idea of comparison.

The answer to your first question is more philosophical, and I think outside the scope of what I am able to reply to here.

But for the second, I think while there are similarities, the details are awfully important for galaxy formation simulations. That being said, there have been comparisons between large scale structure and things like slime molds [1], but beyond techniques we do not expect identicality to the level of precision that is the current state of the art.

[1] https://iopscience.iop.org/article/10.3847/2041-8213/ab700c

With anything cosmological, what we can do is make predictions, and find interesting things the universe does when our predictions fail. Those failures often cascade back into practical outcomes here on earth.

Think of generations of scientists observing orbits, painstakingly recording their measurements by hand, building books of tables, until finally someone was able to put it all together into a comprehensive theory that gave us the ability to predict the effects of gravity.

If we build what's supposed to be a valid model of a galaxy, but we look out and a galaxy we see violates what we expect, we have an interesting place to explore.

The model in question is physics laws and values for parameters therein, including evolving parameters. The purpose of running the simulation is to produce a state that matches cosmological observations: e.g. are stars indeed forming galaxies, are those galaxies distributed as we see (like the films in a foam)? The simulation thus (in)validates basic assumptions of the model. The model in turn can generate features that can be clues to what to look for in cosmological observations: are planets for more proliferated than we've seen so far and should we really spend money on the MostBiggestTelescope to observe them? Even better: can we derive the parameters for the telescope that can definitively confirm or rule out their existence? If we found good agreement on parameters, what does the early universe look like in the model? The future universe?

Tons of basic science, in short.

I'm not sure I understand your question. Cosmological models are used to study cosmology, no more no less. They modelise the physics of the universe as much as we can understand it, and are used with observations.

Reply on news.ycombinator.com