Creating the Initial Conditions

Under construction
  1. Starlab v3.2 was used to create 16384 points distributed as a King model with W0 = 5: mkking -n 16384 -o -w 5 >& initial_snapshot (Beware! 1.1Mb)
  2. From this a list of 16384 phase space coordinates was created with  awk -f  make_phase_list.awk  initial_snapshot > initial_phase_list (Beware! 1Mb)
  3. The masses of the first 12288 were created according to the specification, and the remainder were split into binaries, again according to specification, using the f77 code make_ic.f.

Testing the Initial Conditions

1. make_ic produces the fundamental list of initial conditions (ic.dat) and the following data:
  1. Total mass 10174.2720189169 solar masses
  2. Galactocentric radius 8490.49 pc
  3. kT = 0.709278 solar masses*(km/s)**2
  4. median a: 8.61854E-07 pc
  5. median e: 0.586226
2. another code analyses ic.dat and prduces the following data:
  1. total mass, N: 10174.2720340000 20480
  2. cm position 1.9714438172285D-02 5.0410882363923D-02 4.3476247750797D-02
  3. cm velocity 1.6761459116220D-02 -1.1675330167132D-02 -1.7084032622201D-02
  4. median single mass 0.25310300290585
  5. median binary mass 0.58639198541641
  6. median radius 5.1955181628045
  7. median a, e 8.6185507453150D-07 0.58622576933932
and also the following plots of the distribution of
  1. cosine of inclination
  2. e**2
  3. log a
  4. longitude of node of binary orbit
  5. mass ratio q
(In each case the values are plotted in rank order. Thus these plots are plots of the cumulative distribution, with the axes exchanged: steep slope means low frequency.)


  1. The median single and binary masses, and the median radius, agree (within O(N**(-1/2)) fluctuations) with those expected from the corresponding distributions
  2. The median a and e agree with those of make_ic.f
  3. The distributions of cos i and node are as expected
  4. The distribution of e^2 is not flat; small values are relatively more probable, presumably because of the cutoff in small pericentre distances at 2(R1+R2)
  5. The distribution of log a is not flat. Very large values require large masses to satisfy the condition E>10kT, and large masses are rare; the steepening at very small values may be caused by a steepening in the radius-mass relationship below about 0.3 solar masses.
  6. The distribution of q is not flat. Low values are relatively improbable, presumably because of the limits on the individual stellar mass.