Smoothing Length Coefficient

edited June 2013 in Old versions
Where did the value of 0.866025 (sqrt(3)/2) come from for the smoothing length coefficient? I'm running some 2D wave tank tests, and the waves are dissipating very quickly as a result of this low coefficient... What I saw in my simulations was that the particles on the surface of the wave were more than a smoothing length away from any other particle, so the forces were not being transmitted between particles on the surface of the water... A coefficient of 1.41421356 (sqrt(2)) seems to work better in this case, but has anyone found a systematic way to find an optimal coefficient?

Comments

  • You only have to assume a minimum number of neighbours for each particles, thus values between 0.7 and 2.0 are ok.

    High values (1.2,1.5,1.7) minimises the wave decay simulating propagating waves. However the friction with boundaries becomes more important.

    Low values reduce the friction with boundaries and leads to faster simulation. But wave decay appears soon.
Sign In or Register to comment.