Running a Calculation: Initial Routines

All calculation using the SPH code are run using the executable ``sphagr'', but you will need different input files to tell the code to do different things. We start with a description of routines used to start a new calculation from scratch. Note that this covers the case of using the output of a single-star run as an initial condition for a binary run. We discuss in Sec. 4.6 how to pick up a run exactly where it left off.

First, and most importantly, you must decide how many stars you want in the calculation. We have routines that can set up a single star, or two stars in an irrotational binary, a corotating binary, or a hyperbolic orbit. Once you know what you want, you can edit the ``sph.init'' and ``sph.input'' files accordingly, or better yet, run the ``testinput'' utility we have written. This program queries the user about all the option which go into the input files, creating all that need to be created.

Some input routines require other files as well, and we discuss this below. For more on file I/O, please see Sec. 5.

It is not difficult to write your own initialization subroutine, and we encourage you to do so if ours do not meet your needs. You have complete freedom to do so, but be aware you will need to include ALL of the following things:

• Read in the run parameters from sph.input, as well as any other input files you need.
• Set the positions, velocities, smoothing lengths, masses, and entropic variables of all the particles, these are the common block variables X, Y, Z, VX, VY, VZ, AM, HP, A.
• Set the number of particles N to the correct value, since it is needed for the next step.
• Call the subroutine GRAVQUANT, which sets the proper limits for parallel computation
• Call the subroutine LFSTART, which initializes the leapfrog algorithm (see Sec. 3.1.5).