On Apr 18, 1:11*pm, "Joe Taicoon" wrote:
I am starting a project, university computer science, in which I want to
model the formation of a solar system from an initial cloud of gass/dust.
In doing so I plan on using smoothed particle hydrodynamics and consider
forces from mutual attraction and friction between colliding clouds of
particles.

I have been unable to find any good documents describing how I shoud go
about calculating the friction force between two such clouds, so I post here
hoping for some good references.

Seeing that I come at this from a computer science background rather than
from astronomy, I might as well write a little extra about my thoughts and
let you tell me if I am missing some important point entirely.

I have made simple tests already with a thousand particles which all attract
each other through gravity. The gravity is not calculated as GMm/r^2 which
goes to infinity as r goes to zero. Instead I use the smoothed version
representing gravitational pull between two objects which are not points but
rather clouds: GMmr/(r^2+epsilon^2)^(1.5). Here epsilon0 is a softening
factor which ensures that the force inscreses as r shrinks, but only to a
certain point after which the force shrinks and reaches zero at the same
time as r does. Two clouds of particles do not pull each others centers when
they are exactly on top of eachother.

This lets the simulation run fine without having particles ejectes when they
collide.

First extra question is if it, considering it is compressible gass, is
correct to let a number of particles become one in that they have same
velocity vector and position. They have grown into one larger cloud.
It seems correct to me. Calculating for example the density of the space
occupied by two such collided clouds will show it to be double of one cloud
and the gravitational pull is also double. Any thoughts on that?

Second extra question is if it will be correct to model a particles friction
against other gass by calculating the average velocity vector of the other
gass in the continium and its density and then simply calculate friction
based on delte velocity and density and decelerate the paricle based on
this. After changing velocity the change in energy could then, based on mass
and density, be converted into heat and thereby updating the heat in the
simulation.

Any references, pointers, hints ect. will be greatly appresiated since I
have had a hard time finding good info on this subject. Though SPH started
out doing exactly what I am trying, it seems 99% of the texts dealing with
it now are all about simulating water.

As far as I know, the velocity of a supernova shockwave that's
migrating through a sufficiently dense molecular cloud of mostly
hydrogen, usually doesn't exceed 0.1c, and I don't believe that
singular event alone is sufficient to start the stellar creation
process.

What public funded supercomputer do you have access to?

Computer simulations should more than do the trick, although don't
expect much constructive help from within Usenet/newsgroups (aka
Google Groups).

~ BG