Esoteric Aerospace Engineering Research [J.E.G.] ...
James Ervin Glover, Ph.D.
Mathematical and Statistical Scientist/Entrepreneur at Exec-u-Net Services ====> <exec-u-net.net>
==========================================
==========================================
ESOTERIC AEROSPACE ENGINEERING RESEARCH?
????????????????????????????James Ervin Glover, Ph.D.
?.???ENGINEERING CORE PROGRAM MIGRATIONS TO PHILLIPS CRAY-2
?????Converted a version of the Time History Interpolation
?????and Manipulation (THINMAN) Program to correct format
?????for execution on the Phillips Laboratory CRAY-2
?????supercomputer.?The source code was ported to the
?????CRAY-2 and was successfully compiled, linked, and
?????executed.?It was necessary to write a procedure for
?????execution on the CRAY-X/MP, which utilized the MATNN
?????Routine to port the required input datasets for
?????THINMAN to the CRAY-2.?Appropriate supporting MATCOMP
?????routines, as required by MATNN on the CRAY-X/MP and
?????the CRAY-2, were also written.
?.??INTEGRATED DYNAMICS ANALYSIS SYSTEM (IDAS)
????Installed and tested a Time History Interpolation and
????Manipulation (THINMAN) Program in a classified
????computing environment and provided the object code for
????use in Dynamics Analysis Network Computing Environment
????panels.?Successfully revised and ported THINMAN from
????the CRAY-2 to an IEEE classified workstation
????environment (HP-9000).?Ported THINMAN to the Los
????Alamos National Laboratory (LANL) CRAY Y-MP super-
????computer.?Generated MATRIX files for testing and
????comparing translation times for IEEE to CRAY and CRAY
????to IEEE binary data conversion.
?.??ENGINEERING CORE DATA GENERATION, CODE CONVERSION
????Converted a CRAY-2 version of NASPAT (a code which
????translates model entities from a NASTRAN data deck
????into a PATRAN neutral file) for successful execution
????on the MVE operating system.?Converted all RELEASE,
????ACCESS, SAVE, and ASSIGN commands to their equivalent
????forms in the MVE environment, using SCLCMD, a system
????command interface program. Used indirect methods to
????generate a suitable binary input file to be accessed
????on MVE.?Modified the OUT2MAT source code, which
????re-formats NASTRAN Output2 table data, to generate
????all output parameters as an ASCII data file from a
????namelist and a binary input file. Ported the ASCII
????file from the CRAY-2 to the MVE system and reconverted
????it to binary format, using the binary file as input to
????the NASPAT source code.
?.??STRUCTURAL DYNAMICS ENGINEERING DATA/MAGNETIC TAPE
????CONVERSIONS
????Contractors periodically transmit data which may be
????stored in a variety of formats within disk files,
????magnetic tape files, or electronic mail files.
????Irrespective of origin, the files were systematically
????transformed into a conformable format for existing
????numerical analysis programs.?Codes were written and
????executed to successfully translate and convert scores
????of contractor tapes into required binary output formats
????for the MATRIX routine.?As input files contained
????extraneous characters or were unusually huge files,
????some jobs required as many as 20 modular executions.
????Some contractors served were Martin Marietta Space
????Launch Systems (Titan IV Liftoff Forcing Function Data,
????Booster Vehicle Model Data, and Titan IV PLI Loads
????Data) and Lockheed Missiles and Space Company
????(Satellite Vehicle Max Propellant Slosh Model Data).
?.???STRUCTURAL MECHANICS MICROCOMPUTER SUPPORT
?????Installed and verified software on personal computing
?????equipment.?After determining hardware and software
领英推荐
?????requirements for installation of NCSA Telnet,
?????installed e-mail software, provided operating
?????procedures, established and tested electronic
?????networking capabilities.?Served as a member of a
?????Computing Quality Process Action Team.
?.???CRAY BINARY DATA CONVERSION
?????The Integrated Dynamics Analysis System was designed
?????so that a workstation user will be able to submit
?????jobs to the CRAY computer or a workstation and
?????process returned data in such a way that the
?????interfaces are transparent.?As all structural loads
?????analysis programs initially resided on the CRAY, the
?????process of rehosting programs to a workstation
?????required modifications to the code and the conversion
?????of CRAY binary data to workstation binary.?Hence,
?????a sequence of bit manipulation routines were developed
?????and coded which read various CRAY floating-point and
?????character type words and output them in binary, after
?????bit shifting, using the concept of masking.?A larger
?????conversion program was written and the routines were
?????successfully tested and used in this program to
?????convert arbitrary 64-bit CRAY floating point numbers
?????into the equivalent IEEE single-precision format on
?????the SUN workstation. Cases involving double-precision
?????were also handled.
?.???IBM BINARY DATA CONVERSION
?????Information on telemetry tapes is digitized and
?????stored on the IBM system in binary form.?A series of
?????routines was developed to convert IBM double-precision
?????binary data to standard IEEE double-precision format
?????in a way analagous to that used for CRAY data.?It was
?????also necessary to determine exponential biases which
?????differentiate the two formats.?Since the IBM operates
?????in hexadecimal, certain nuances were observed which
?????were not encountered in the case of the CRAY
?????conversions.?For example, it was necessary to
?????overcome the conflicts in various argument types
?????allowable for bit-shifting functions on the IBM,
?????as compared to the CRAY.?It was also necessary to
?????develop code to accomodate shifts across IBM word
?????boundaries, as well as, to utilize rounding techniques
?????to preserve bit values affected by shifts.?The sign,
?????characteristic, and mantissa of a given number were
?????generated in separate subsections of code and
?????appropriately masked to generate the correct CRAY
?????word.?Since various machines or platforms have
?????different bit configurations, a major consideration
?????in this conversion effort was the limitation on the
?????size of a number, induced by varying exponential bit
?????stream lengths, in both single- and double-precision
?????numbers.
===================================================
Power Point Presentation <=======
Slide Share <=======
==========================================
==========================================
???????
Published by
Mathematical and Statistical Scientist/Entrepreneur at Exec-u-Net Services ====> <exec-u-net.net>