No abstract

Expertise concerning fault tree methodology was extended primarily by T. H. Smith, W. K. Winegardner, and P. J. Pelto. Their concern and direction were the principal motivating factors behind the development of this method to fault tree analysis. Effective utilization of the hashing algorithms and the NAMELIST input package was accomplished through the helpful suggestions of K. B. Stewart and B. H. Duane.

No abstract

No abstract

A computer code, GSSLRN-II, is presented which is written in generalized form so as to allow analysis of any data containing peak shapes which can be adequately represented by an analytical function form. GSSLRN-II utilizes a second order Newton-Raphson least-squaresiterative scheme in its analysis of peak shapes. Effect~ve use has been made of the measurement variance and the least-squares deviations to justify automatic recycling attempts of least-squares fits on groups of peaks.The code will locate peaks, group them in appropriate sets, optimize the interval-of-fit, least-squares fit the set of peaks, recycle for another fit if necessary, and upon ootion will generate plots of thp fitted results. ~1any other options are available which allm'l the user a high degree of versatility \'lith respect to input and/or output options.• . -• ".'• . . I. I ntrodu cti on BNWL-1579 GSSLRN-I I A LEAST-SQUARES COMPUTER CODE FOR THE ANALYSIS OF PEAK SHAPES WHICH CAN BE DEFINED ANALYTICALLYA computer code, GSSLRN-II, is presented which is written in generalized form so as to allow analysis of any data containing peak shapes which can be adequately represented by an analytical function form.GSSLRN-II is a much more versatile code than GSSLRN-I (1) because it is not limited to the analyses of fission-product data using a fixed analytical function. In addition, GSSLRN-II contains a major improvement with respect to the use of measurement variance obtained from the least-squares fits for the purpose of justifying recycle (multiple fitting) attempts.The code makes use of the least-squares package, LEARN,(2,3) which utilizes a second order Taylors' expansion in its least-squares iterative scheme.The code contains a wide variety of options which are designed to allow the analyst many override features with respect to the input and/or output options.The net result is a code which is adaptable to many types of data with a minimal amount of necessary modification. [. Highlights of the Code Some of the features of GSSLRN-II which allow a wide variety of application are:• Ability to rewrite the analytical function which describes a continuum under the peak shapes without destroying the integrity of the code• Use of nonlinear least-squares techniques accompanied by a substantial regression analysis• Judicious use of the measurement variance in order to quantify the performance of a fit• An. automatic recycling logic which enables the code to build in hidden or unaccounted for peaks• Optional use of a peak searching routine(4) to locate peaks from data spectra collected on large multichannel analyzers• Optional input and/or output specifications including automatic on-line or off-line plotting features• CALCOMP plotting-options which display the measured s~ectr~m, fitted curves and their root-mean-square deviation envelope, fitted parameters, curves of each peak as they are resolved from a multf~let, or a three-dimensional plot of multiple spectra• No prior knowledge required to locate and successfully fit hidden or overlapping spectral c...