5 Stamp collection of collision strengths

We present some of our collision strengths graphically in the manner of a "stamp collection'', each collision strength being plotted against the final electron energy within a small rectangle resembling a postage stamp. In this way we are able to put together 12 transitions per journal page. The present collision strengths contain structure in a narrow energy range which typically extends for 2 or 3 electron volts starting at threshold energy. Since we believe that these pseudo resonances (a) have no physical significance being an artefact of the CCC method and (b) only affect the thermally averaged collision strengths $\Upsilon$ at relatively low temperatures, we have not attempted to delineate them in great detail for most transitions.

An interesting result emerges from our calculations at low energies, namely that for some transitions the collision strengths appear to tend to finite non zero values at threshold. However this behaviour may simply arise because of pseudo resonances since there is no apparent physical explanation for it. When carrying out the thermal averaging we impose the condition that all our collision strengths be zero at threshold.

Figures 2, 4, 6, 8, 10 show collision strengths for transitions $\rm 1 \to 2$, ... $\rm 2 \to 29$ with the final electron energy varying from zero to 0.2 rydbergs (2.72 eV). Figures 3, 5, 7, 9, 11 are for the same 55 transitions but with the final collision energy now going from 0.2 Ry to 1 Ry (13.6058 eV). It should be noted that the vertical scale varies from one "stamp'' to another and is chosen in order to optimise the legibility. Pseudo resonances can be seen in most of the low energy "stamps'', i.e. when the final electron energy is between 0 and 0.2 Ry. The occasional irregularities occurring above 0.2 Ry are presumably also produced by pseudo resonances.

Collision strengths for electron excitation of neutral atoms do not contain the enormous amount of resonance structure associated with positive ions: see, for example, Hummer et al. (1993). There is some structure, but we have considered it to be of secondary importance in so far as the thermally averaged collision strengths ${\sl\Upsilon}(i-j)$ are concerned. It would be too costly in computer time to attempt to delineate the structure associated with each transition. Also this structure tends to occur over a limited energy range just above threshold energy.