a cura di Paolo Botton
11 febbraio 2013
On february 2014, after four months of data processing and simulations, I finished the task I was in charge to
carry out, based on a preliminary model whose goal was to graphically show the interaction processes
inside a tiny globular cluster.
In the beginning the program was designed and intended to be the well known simulation of the N-bodies interactions but the equations I was given to manage, once properly transcripted in machine language (6500 rows in Python) and fed with the preliminary data (boundaries conditions and preliminary values at time zero), instead of plotting the movement of the bodies, early starded to show a distribution that well suited the shape of a globular cluster.
As the stars started growing in number, the bodies moved slowly and collapsed to the center of the cluster,
following a spherical simmetry, without colliding each others. This, I guessed, was one of the expected behaviours
as far as I had to understand from the specifications, but the math model was written to simulate interactions
and not something that - to me - sounds to behave like a spooky accretion mechanisms.
As matter of fact each star entered in the model (read "born"), after a brief interaction, found a stable configuration inside the cluster and in the end I got a spherical shape that was very, very closed to a globular cluster.
Once warned my contact person, I was asked to keep going with the computation, without changing anything, to see
if the module should show any computability issue or become unstable.
Nothing bad happened until the end of the whole process... As I am just a computer science specialist, only skilled in maths and physic, I am not in the position, neither I want, to jump to easy or funny conclusions.
I just had the permission to write this short summary on my pages and put both graphs and the small resolution animation, in order to give the evidence of my job to the audience.
Now data have to be collected and deep analysis should be made by specialists, in order to proof if the model they designed can be used to give the evidence of a real globular cluster accretion model.
The equations I transcribed into the program produced 85000 binary files of 15MB each, containing all the data used or generated by the model during computation, as well as all the boundaries conditions, those that in my, and also general opinion, was in part responsible for the unexpected behaviour of the model.
|© Paolo B.|