4 Discussion

The theoretical Stark FWHM (w) dependence on the electron temperature together with the values of the other authors and our experimental results at the electron density of 1 $10^{23}\ \ \rm m^{-3}$ are presented graphically in Figs.  2 and  3 for the transitions 3s-3p and 3p-3d, respectively. Theoretical values calculated by Blagojevic et al. (1996), using the semiclassical-perturbation formalism (Sahal-Brechot 1969a,b), are presented with symbols $DSB_{\rm e}$ and $DSB_{\rm i}$ on the Figs. 2 and  3. $DSB_{\rm e}$ denote electron impact width only, while $DSB_{\rm i}$ denote the sum of electron + ion impact width calculated for our plasma compositions at relevant plasma parameters. Calculated values on the basis of the semiclassical (G) (after Griem 1974), simplified-semiclassical approximation (GM) (Eq. (526) in Griem 1974) and modified-semiempirical formula (DK) (after Dimitrijevic & Konjevic 1980) are presented by Dimitrijevic & Konjevic (1980, 1980a,b). Theoretical predictions by Hey (1976) (x), calculated with the semiempirical method, at plasma parameters obtained in Popovic et al. (1975) are also presented.

  

Figure 2: Theoretical Stark FWHM (w) dependence on the electron temperature scaled to the electron density of a 1 10$^{23}\ \rm m{-3}$ for the 3s-3p transition. $\bullet$, our experimental results and those of the other authors: $\diamondsuit$, Popovic et al. (1975); *, Källne et al. (1979); $\ifmmode\hbox{\rlap{$\sqcap$}$\sqcup$}\else{\unskip\nobreak\hfil \penalty50\hsk... ...box{\rlap{$\sqcap$}$\sqcup$} \parfillskip=0pt\finalhyphendemerits=0\endgraf}\fi$, Purcell & Barnard (1984); $\blacksquare$, Puric et al. (1987); $\triangle$, Glenzer et al. (1994a); $\circ$, Blagojevic et al. (1996). Theory: --, semiclassical electron impact widths ($DSB_{\rm e}$) and - - -, semiclassical electron + ion impact widths ($DSB_{\rm i}$) (evaluated only for the plasma parameters in our experiment A and B) using values calculated by Blagojevic et al. (1996) on the basis of the semiclassical-perturbation formalism; ...., semiclassical electron impact widths after Griem (G); - - -, simplified semiclassical approximation (Eq. (526) in Griem 1974) (GM); -- -- --, modified semiempirical formula (after Dimitrijevic & Konjevic 1980) (DK). The results of: (G), (GM) and (DK) calculations are presented by Dimitrijevic & Konjevic (1980, 1981a,b). x describe theoretical predictions by Hey (1976) calculated at plasma parameters obtained in Popovic et al. (1975). $\overline{\lambda}$ is the mean wavelength for the multiplet. The error bars include the uncertainties of the width and electron density measurements

  

Figure 3: Same as in Fig. 2 but for the 3p-3d transition

The theoretical Stark shift dependence on the electron temperature together with the values of the other authors and our experimental results at the electron density of 1 $10^{23}\ \ \rm m^{-3}$ are presented graphically in Fig.  4 for the transitions 3s-3p and 3p-3d. Theoretical values are calculated by Blagojevic et al. (1996) using the semiclassical- perturbation formalism and are presented with symbols $DSB_{\rm e}$ and $DSB_{\rm i}$. $DSB_{\rm e}$ denote electron impact shift only. The ion contribution to the shift is calculated only for our plasma conditions (parameters and compositions in experiments A and B). $DSB_{\rm i}$ denote the sum of electron + ion impact shifts.

  

Figure 4: Theoretical Stark shift (d) dependence on the electron temperature scaled to the electron density of a 1 $10^{23}\ \rm m{-3}$ for the 3s-3p and 3p-3d transitions. $\bullet$, our experimental results and those of the other authors: $\circ$, Puric et al. (1988); $\triangle$, Blagojevic et al. (1996). Theory: --, semiclassical electron impact shift ($DSB_{\rm e}$) calculated by Blagojevic et al. (1996) and - - -, semiclassical electron + ion impact shift ($DSB_{\rm i}$ ) evaluated only for the plasma conditions (parameters and compositions) in our experiment (A and B). $\overline{\lambda}$ is the mean wavelength for the multiplet. The error bars describe the uncertainties of the shift measurements

On the basis of our measured Stark parameters and existing experimental and theoretical values one can conclude:

1. For the lines of 3s-3p transition our experimental Stark FWHM data agree well with theoretical predictions calculated on the basis of the semiclassical theory (G) after Griem (1974). Same trend shows measured values from Glenzer et al. (1994a) at about 90 000 K electron temperature. In the range of the electron temperature: 30 000 K - 54 000 K, within the experimentally accuracy, our data agree, also, with predicted Stark FWHM values based on the semiclassical-perturbation formalism including the ion contribution ($DSB_{\rm i}$). Other experimental data lies under predicted (G), ($DSB_{\rm i}$) and our experimental values.

2. For the lines of 3p-3d transitions our experimental Stark FWHM data, about 35 000 K electron temperature, shows agreement with predictions based on the semiclassical-perturbation formalism including the ion impact contribution ($DSB_{\rm i}$) and, also, within the experimentally accuracy, with prediction based on the simplified semiclassical approximation (GM). Experimental values from Blagojevic et al. (1996) agree with our experimental data. The situation is, however, different at elelectron temperatures over 50 000 K. Namely, our Stark FWHM data at 54 000 K and those from Glenzer et al. (1994a), at about 90 000 K, agree with theoretical predictions calculated on the basis of the modified semiempirical approximation (DK).

3. In the case of the Stark shift one can conclude that our measured $d_{\rm m}$ data, which are equal to zero, within experimental uncertainties ($\pm$0.0015 nm) are not in contradiction with theoretical predictions (see Fig. 4). Namely, the only existing theoretical results of the Stark shifts, calculated on the basis of the semiclassical-perturbation formalism ($DSB_{\rm e}$) are very small and have negative sign. Inclusion of the ion contribution to the shift ($DSB_{\rm i}$), lead to their increase. Measured Stark shifts by Puric et al. (1988) and Blagojevic et al. (1996) have, also, definite negative value. It should be pointed out that the theoretical prediction of the Stark shift values is very sensitive to the number of the perturbing levels included in the calculation. Namely, the number of the perturbing levels has appreciable influence on the shift, including its sign. Omitting some of them, may lead to erroneous results. On the other hand, the use of available oscillator strengths, or those calculated from Coulomb approximation, can lead to results of opposite sign in Stark shift calculation (Djenize et al. 1993).