[OP2] Data Structure Oddity and possibly [H3D] Data Processing
Hello~
I come from the old school days of Nastran before other companies apart from that single monopolizing one got a hold of the source code and developed on it or off of it. OP2 data structure (along with DMAP) was explained and well spelled out in what was called the 'Programmer's Manual'. Programs were developed off of the data structure and the newer kids on the block took the same formulation and ported to python and the like it seems. (Used to be well aligned with the archaic Fortran back in the days).
In any case, Optistruct and compatible hyperworks modules makes mention of OP2 (Output2) type of files, seemingly claiming it works with this file type and its data structure. I had a reason to query the ELSTRESSes of a rather large analysis results file (think 33GB h3d file), and wanted a binary file to postprocess my own personal way of handling a subset of the full data. Since (I don't know how to read an) H3D file, due to its esoteric (proprietary and undocumented) nature, I thought I was clever by outputting a seemingly understandable OP2 file to get at some of those data. Unfortunately, Altair's generated OP2 file seems like a **** child in the makeup of the binary data structure. There is an inkling of it being somewhat like a true OP2, and some data headers and tables were readable and discernible, including stress, but some binary protocols and structures, positions have been shifted where without some sort of a roadmap this is NOT USABLE.
My question is two-fold.
One. Which wise guy decided to change the OP2 file structure, and if so, where is the binary map to how the data is newly housed.
Two. Say I stick with the native H3D file. Is it possible to somehow read this Altair specific, proprietary file to get at fundamental data structures, ones asked for in the Case Control. If this is not how the meat of the binary file is arranged and requires an Altair medium to translate those result, that's fine. Is there some commandline, like using Hyperview itself somehow run in commandline mode like:
C:\> hyperview.exe -spitout OES1x (or OUG1) results.h3d subsetquery.pch (or .txt or .op2 or .bin) or something like zat?
or is there somehow a way to run hyperview in 'modeless' (or blind) mode, that takes a tcl file and performs on it, as if it is being run in GUI console directly in Hyperview, which then could possibly GetValueList of data wanting extraction and WriteData into a file in the operating system.
I am actually really interested in the latter approach as I want to adopt Altair ecosystem more and more, as I see merit in the software system for pre- and post processing. There are aspects of Hypermesh that I find invaluable for an analyst and I simply haven't seen such a thing in the 35 years I've been in the industry. Kudos to Altair!!!
Also, if you would when answering these types of questions, (as I've found in other posts) to speak directly to the problem at hand and not simply 'tossing' out a link to another 'similar' post, for the poor user to have to go decipher, when it doesn't read or explain very well. (Think Customer Oriented mindset) If someone does want to link to a good side reference, fine, but please make a concerted effort to adapt that explanation to the problem at hand. I've seen references to a batch mode commandline processing, that simply is too cryptic, and like much of Altair's documentation, simply state the interfacing information and not explaining the meat of the matter. Please answer directly here, and I'm most certain this post will be a great living document, very clear and practical, for myself and others in this community (posterity).
OP2 and/or H3D data handling or processing ('modeless' meaning not farting around with the GUI manually ^^)
Thank you~~~