Rubies computational method

Computer methods callec HEMP, TIGER RUBY, used to calculate deton vel and other parameters of solid liq expls are described... 

Detonation pressures estimated by the simpler method are compared in Table VI with the corresponding ruby values for twenty-eight materials. Since we are at this point concerned only with reproducing ruby results, it is unimportant to the present discussion whether input heats of formation in the ruby computations are accurate although the AH/ for picric acid in Ref,... 

A first trial by the method of least squares involved fitting detonation pressures predicted by the ruby computer code8 to an expression of the form... 

Computer codes are used for the calculational procedures which provide highly detailed data, eg, the Ruby code (70). Rapid, short-form methods yielding very good first approximations, such as the Kamlet equations, are also available (71—74). Both modeling approaches show good agreement with experimental data obtained ia measures of performance. A comparison of calculated and experimental explosive detonation velocities is shown ia Table 5. 

Because it was considered that the intrinsic inexactness of the computer s input information led to uncertainties of at least 5% in ruby s predictions of P, we originally felt that differences of about the same magnitude between arbitrary N, M, and Q and ruby s values could be tolerated in the present study. For this reason the analyses of Tables IV and V led us to set Po= 1.4 g/cc as a tentative lower limit of applicability of the present calculational method. The results below show this restriction to be unnecessary. 

The above observations regarding the sensitivity of N, M, and Q and the relative insensitivity of to exact compositions of products from individual explosives lead to a variety of interesting observations. These involve (a) the present, calculational method (b) the ruby code and similar computer-based methods of calculation (c) effects of equilibria on actual detonation properties and, eventually, on damage effects and (d) methods which are widely used to intercompare the predicted performance of explosives. 

In discussing Eq. (1) in the earlier papers of this series, it was emphasized several times that despite some serious questions which had been raised regarding the accuracy or appropriateness of the computer codes input information, our temporary intent was to devise a simple calculational method which might best reproduce ruby results. It was also mentioned that the Kistiakowsky-Wilson equation of state7,8 parameters used in ruby6 were chosen by Mader5 so as to best accommodate five experimental measurements the detonation pressure of RDX at 1.80 g/cc, the detonation velocities of RDX at 1.00 and 1.80 g/cc, and the detonation velocities of TNT at 1.00 and 1.64 g/cc. 

The non-relativistic version of DVME method was developed in 1998 and was applied to the analysis of multiplet spectra of ruby [6-8]. This method was later applied to the analysis of a variety of TM-doped solid-state-laser materials [9-11]. The relativistic version of DVME method was developed in 2000. However, at that time, it was still difficult to calculate multiplet spectra of RE ions due to the limited performance of available computers. On the other hand, the relativistic... 

The multiplet structures of ruby calculated by the DV-ME method with and without the correlation correction are shown in Fig. 4. The experimental values are also shown together. In this case, the multiplet structures are calculated directly using the molecular orbitals of the impurity states obtained by the cluster calculation. In the calculated results, each level is broadened by the presence of the trigonal crystal field. Although the split of each peak seems to be somewhat overestimated due to the computational errors, it can be improved by increasing the number of sampling points (2). 

---