The tables below were calculated by Roy Bossingham using Magboltz 2, V3.1 (Biagi, 2000) for typical values of interest to STAR: T=23.9C, P=760 Torr and B=5.0, 2.5 and 0.0 kG, with additional values to help with interpolation. E values were most concentrated around E=148 V/cm, and were parallel to B in all cases.

The drift velocity and diffusion constant are functions of the ratio of E over density; to find the drift velocity and single-electron diffusion at fixed B, general values of the drift length L and small changes from the input values of (cm)E (V/cm), P (Torr) and T (K):

• Use (E/P) * (T / 297.04 K) for the "E/P" column.

• Look up the appropriate value (note that, for parallel E and B in Magboltz, Sigma_T is a function of B, but drift velocity and Sigma_L are not).

• Scale Sigma_L and Sigma_T by sqrt[L * (760 Torr / P) * (T / 297.04 K)] to get the diffusion in um.

Magboltz 2 uses Monte Carlo methods, with the statistics determined by the collisions parameter. Statistics were used corresponding to collisions=120, or higher, giving a precision of at least 0.04% in drift velocity and 1.6% in diffusion.

Drift velocity in P10 vs E for P=760 Torr and E parallel to B, from a rather global viewpoint.
The STAR TPC normally operates just beyond the peak in the right-hand plot.

Diffusion in P10 vs E for P=760 Torr and E parallel to B.

Drift velocity and longitudinal diffusion in P10 vs E for P=760 Torr and E parallel to B.
The plotted parameters are close to the operating conditions for the STAR TPC drift volume.

Transverse diffusion in P10 vs E for P=760 Torr and E parallel to B. The plotted
parameters are close to the operating conditions for the STAR TPC drift volume.

Blum and Rolandi define the variable D, which is proportional to the transverse diffusion squared and a function of E and B; they plot it for the special case of P9 (relevant to Aleph) and E=110 V/cm. D(0)/D(B) becomes a straight line at sufficiently high B (higher than in STAR, however). As the lefthand plot below shows, D(0)/D(B) varies smoothly and slowly with B^2 and E, making it useful for interpolation. However, as the righthand plot shows, Blum and Rolandi's P9 plot does not apply especially well to STAR's operating conditions.

Transverse diffusion in P10 vs E for P=760 Torr and E parallel to B. The plotted
parameters are close to the operating conditions for the STAR TPC drift volume.

Calculated drift velocity [cm/us] and longitudinal diffusion [um/sqrt(cm)] in P10 for parallel
E and B at P=760 Torr and T=23.9C  . Values are nearly independent of the magnitude of B.

Transverse diffusion [um/sqrt(cm)] in P10 (calculated)
for parallel E and B at P=760 Torr and T=23.9C  .

Page created by Roy Bossingham; maintained by   .
Last modified on June 29, 2000   .