New standard for 3D Model Files
Design engineers have long thought it would be convenient if all the data describing a part were packed in one standard electronic 3D file. Recently, this concept has finally become practical thanks to the definition of a new military standard known as Mil-Std-31000A. As with many concepts introduced by the Dept. of Defense and NASA, this technology will likely filter down to revolutionize product development in all manufacturing industries.
The overarching concept of Mil-Std-31000A is called the Model Based Enterprise. Through MBE, for example, one all-inclusive electronic 3D model file would eliminate conventional hard-copies — that historically have included 2D drawings, instruction manuals, manufacturing data, and inspection documents — defining Army weapon systems. Mil-Std-31000A also holds the promise of allowing the translation of proprietary CAD files into PDF files that would include all descriptive information in one electronic package. This is a tremendous step forward in that PDF files are free to deploy and free to be read by the public using common HTML5 browsers.
The scope of the standard says it, “Provides requirements for the deliverable data products associated with a Technical Data Package (TDP) and its related TDP Data Management Products.” The first 25-pages include a lot of traditional procurement legalese. The 140 pages that follow include Appendices A, B, and C that furnish extremely useful information for designers.
Of particular interest is the scope of Appendix A, which says it “Provides guidance for Procuring Activity personnel on the acquisition of the various types of technical data and the completion of the TDP Option Selection Worksheet.” This worksheet acts like a pilot’s checklist to guide design managers through the conceptual design, development, product drawings, tooling and ordering phases of the TDP.
Appendix B covers what’s called organizational scheme. It’s designed to comply with ASME Y14.41. This is the ASME standard for digital product-definition data practices with a focus on(geometric dimensioning and tolerancing(GD&T) and product geometry. It is designed to work with any CAD system. “The goal of this schema is to define a common practice to improve design productivity and to deliver consistent data to the customer,” the document says.
Appendix B meets this goal by illustrating a sampling of requirements for title-block design, annotations, text, and detail views. It follows with instructions on methods of handling product data in the 3D model and continues by addressing naming conventions, combination views, annotations, and metadata. Metadata is described as, “Data that supports the definition, administrative, or supplemental data package. Metadata includes all relations, parameters, and system information used in the model. This data resides at the model and feature level.”
Appendix B also recommends that models contain elements (really, standard names) to accommodate MBE and includes a 77-item list of them. For example, element P1 stands for ALT_MATERIAL and is described as “Use as needed to define alternative materials”; P77 stands for WEP_SYS_NAME and is described as “The weapon system name that the drawing is applicable to.” I see this Critical Data Element List as an attempt to promote standardized and universal names for these elements. I expect these element numbers and definitions will be useful for troubleshooting inconsistencies between 3D-modeling software within organizations and between suppliers. But the average design engineer will likely have little need to use this list.
Appendix C is a key to understanding the new standard and is a must-read for designers-in-the-trenches. That’s because it goes into the various subtle defects in 3D models that can cause issues in follow-on applications such as numerically controlled manufacturing, finite-element analysis, and inspection with coordinate-measuring devices. It goes on to identify and classify the causes of defects in 3D digital models, indicates which defects are likely to cause difficulties for various types of computer applications that use the models, and recommends tolerances and acceptance criteria for models.
Appendix C also divides validation into three classes: geometry, annotation, and attributes. Geometry validation identifies defects in geometric entities such as lines, curves, and surfaces, and errors in relationships among entities, such as gaps or overlaps. Geometry validation also encompasses model topology including faces, edges and vertices. Annotations consist of dimensions, geometric tolerances, notes, and other symbols attached to the model. Validation examines the locations, orientations, values, layers, context, and syntax of annotations. Model attributes consist of views (model orientation and zoom state), layers, numerical values, notes, material descriptions, or combinations of them necessary to complete the product definition but not visible when viewing the model. Attributes must be evaluated for their existence, type (integer, floating point, text), and values.
Appendix C contains definitions, requirements for native 3D model data and detailed requirements for acceptance criteria such as: data checking, model healing, conversion of 3D models, remastering of 3D model data, and long-term archiving. Each of these acceptance criteria is well illustrated with examples of potential problems and recommended corrections.
There is also an exhibit within Appendix C that furnishes guidance for geometric integrity. It covers topics such as geometry, surface criteria, edge criteria, edge-loop criteria, shell criteria, solid-body criteria, assembly-structure quality criteria, and annotation-quality criteria. It contains over 100 illustrated examples of 3D drafting problem areas with a recommended correction for each problem.
One example is that of nontangent edges. These are tangent angle changes at the common endpoint of adjacent edges. As the exhibit explains, nontangent curves and line segments don’t necessarily indicate poor data quality. A designer might want nontangent curves, for example, on square corners with chamfered and beveled edges. If curves are intended to be tangent but aren’t, then a discontinuity (dubbed a G1 discontinuity in the standard) indicates poor quality. Nontangent conditions are usually visible to the unaided eye and can be felt by finger tips on physical models.
Several exhibits in Appendix C cover validation. The need for validation arises mostly because of the many different 3D-modeling formats that come from industry giants such as Siemens AG, Austin, Dassault Systèmes, Waltham, Mass., Autodesk Inc., San Rafael, Calif., and others. There is no universal language in 3D modeling. As a result Mil-Std-31000A must deal with errors stemming from translating back and forth between native formats of CAD platforms. Similarly, errors crop up from incompatibility between derivative formats and variations between outdated revisions of native formats. One exhibit defines 37 error conditions such as different solid volume, unsewn solid, unmatched part, merged assembly, and so forth, and lists their effects.
One notable practice in the standard is its referencing of commercial companies. It is extraordinary to find a military standard mentioning vendors. This practice shows that the private sector holds the key to the future of 3D modeling.
For example, another exhibit acknowledges commercial geometry-checking software such as CADdoctor from Elysium Inc., Southfield, Mich., and CADIQ from International TechneGroup Inc., Milford, Ohio. These companies can also provide geometry-healing software which can correct errors if 3D model data is to be translated to another format. Also covered is semiautomatic remastering which involves translating 2D sketches and parameters to 3D. There are several illustrations of a part remastered from Siemens NX to Dassault Systèmes Catia V5 using CADfeature software from Elysium. It continues with derivative model validation using CADIQ.
A number of commercial software houses have created programs able to transform native parts and assemblies with 3D GD&T, and other 3D annotations into high-fidelity 3D PDF and 3D HTML MBE documents. This will make it possible for hardware part suppliers to read 3D files from customers and for 3D files to be archived and easily read in the future. More importantly, a soldier in a foreign country can find part lists and assembly instructions that apply to a weapon system by simply calling up a PDF file on his or her laptop.
Those involved in solid modeling recognize they depend heavily on the major software suppliers such as Siemens, Dassault, and Autodesk, which dominate 3D modeling worldwide. These suppliers are updating their software to accommodate Mil-Std-31000A. Designers would be well advised to educate themselves about how their CAD software vendor is handling the needs of the new standard.
Resources: Explanation of Model Based Enterprise
Downloadable transcription of Mil-Std-31000A in PDF
Anark Corp., Denver, has taken the lead in providing tools to translate 3D models from proprietary CAD formats into PDFs