[Users] Inquiry Regarding Parameter "N" in BBH Simulations

[Peter Diener]

Hello,

The parameter N defines the numerical resolution, i.e. the grid spacing, 
used in the simulation.

The setup of a Llama multipatch system is somewhat complicated, so in
GW150914.rpar, N is used to calculate all the necessary quantities
for setting the coordinate system up. A larger value of N means higher
resolution while a smaller value means lower resolution.

So when you decreased the value of N from 28 to 20, you effectively 
decreased the numerical resolution. This then resulted in, not only, each 
timestep taking less time (due to less grid points) but also the 
simulation was able to take larger timesteps.

With the decreased resolution, you now also have larger numerical errors. 
It is not straightforward to predict if the larger errors from less 
resolution will result in shorter or longer merger times. However, in this
case it looks like larger errors leads to an increase in the merger time.

Cheers,

   Peter

On Fri, 10 May 2024, Burgundy|?6?1'-'?6?1)?1?2?7?7 wrote:

> 
> Dear Einstein Toolkit Team,
> 
> I hope this email finds you well. I am writing to inquire about a specific
> parameter used in the simulation command for binary black hole (BBH) mergers
> in the Einstein Toolkit's Gallery.
> 
> Recently, I have been experimenting with running BBH simulations using the
> following command:
> 
> simfactory/bin/sim create-submit GW150914_28 --define N 28 --parfile
> par/GW150914/GW150914.rpar --procs 128 --walltime 24:00:00
> 
> In this command, I have noticed the use of the parameter "N" with a value of
> 28. However, when I changed the value of "N" to 20, I observed a significant
> reduction in the total runtime of the simulation, despite the fact that the
> merger time of the binary black holes increased from around 900M to 1000M.
> 
> Could you please provide some clarification on the purpose of the "N"
> parameter in this simulation context? Additionally, could you explain why
> changing the value of "N" affects the runtime of the simulation and the
> merger time of the black holes?
> 
> I would greatly appreciate any insights or explanations you could provide on
> this matter. Thank you very much for your time and assistance.
> 
> Sincerely, Xinyue Bian(an undergraduate who has a bit interest in NR)
> 
> 
> 
>