Next Contents Previous

B.2. Radial grid

Essentially two types of radial grids have been used : one kind for the uniform (r0) density distribution and another for the r-1 case. This was necessary because, the code CSDUST3 has convergence problems, if the distance between successive radial grid points is such that the optical depth approx 1 at any of the frequencies of the grid. This problem is most severe at the two boundaries of the cloud, where some UV radiation (stellar or ISRF) is incident, so that a finer grid is required there. Hence for the uniform density distribution, the entire radial range has been divided into two parts, with 71 points in a logarithmically increasing grid spacing in the first part ( i.e., from the inner boundary to mid point of the cloud), and an exact mirror image on the second part. This gives a total of 141 radial grid points with a finer grid at both the boundaries. It may be noted, that the necessity for having a fine grid towards the outer boundary of the cloud, exists only for uniform density distribution, since in this case alone there remains enough dust grains to produce high optical depth at the outer boundary. For the case of r-1 density distribution, the dust density falls off so rapidly, that towards the outer edge, the optical depth between successive radial grid points is never large enough. As a result, a smooth convergence is achieved even with a logarithmically increasing grid spacing right from the inner to the outer boundary of the cloud.