| Spec box: Working Model 3D 4.0 Working Model 3D is a 3D dynamics simulation program for Windows 95 and NT 4.0. Recommended PC configuration is a Pentium processor, 32 Mbytes of RAM, and up to 125 Mbytes of hard disk space.
List Price: $5,995
Working Model Inc., 66 Bovet Rd., Suite 200, San Mateo, CA 94402; Tel: (800) 766-6615; Fax: (415) 574-7541; www.workingmodel.comProduct Code 4461
A similar product: DADS--CADSI, 2651 Crosspark Rd., Coralville, IA 52241; Tel: (319) 626-6700. Product Code 4462
To speak with a company representative call 1-800-828-6344 x011 and enter the product code
Working Model 3D (WM3D) 4.0 adds enhanced integration with CAD software, improved collision detection, additional constraint navigation tools, function-driven motion, joint limits, tracking cameras, an added generic constraint, and a units-smart formula editor.
I typically try to experiment with most new features of a product--however, Version 4 found me at a point of real need in three designs, so my use was intensive and problem-focused. I'll discuss the features which I found most useful in these applications, and mention a few older features which were of great value. All of these models were difficult analytically, and included indeterminate collisions of complex geometry. Two problems involved moderately high part counts, and two involved unusual combinations of high mass ratios and high externally applied forces.
I imported a majority of the model geometry for all problems from Mechanical Desktop. As the analyses progressed, I made changes in the geometry in Desktop, then re-imported the models into WM3D. I found that, with a little forethought, excellent associativity can be maintained--although, a review of constraints and custom mass parameters eliminated surprises. CAD associativity is a very productive feature.
My first problem made good use of the new generic constraint. This constraint permits general control of six degrees of freedom (DOF) with a single constraint. My problem included several 3D bending beams which I originally modeled in groups of three beam elements, each joined with a single DOF revolute spring/damper. Changing to a generic constraint which combined all three DOFs reduced the number of elements by a factor of 3, significantly simplifying the model. I can see a wide variety of applications for the generic constraint.
Throughout my work, I made constant use of the object navigation tools, and found them indispensable. For the first time, I made extensive use of transparency, allowing me to visualize, and show others, complex interactions among hard-to-see parts. The value of WM3D's ability to continually replay a simulation from varying vantage points, and with varying visibility characteristics, can't be overemphasized. WM3D's ability to aid communications is outstanding.
The new formula editor parses expressions for unit consistency, clarifying an issue which was confusing in earlier versions. The editor is usually able to highlight inconsistent units, make correct assumptions, and generally keep the user out of troubles which can be serious in complex analyses. Although I found it possible to confuse the units parser, particularly when parentheses were deeply nested, it usually complained, and I have the feeling that some of the problem (perhaps all?) was mine. I'd ask, though, for a bit more complete documentation of this feature, and the generic constraint as well.
The new collision engine appears to make intelligent decisions about regions likely to collide, ignoring those which cannot collide near the current time, thus saving computation time. This was of particular value in one problem which included a complex structure, with only small areas in which collisions could occur. I used the entire CAD model, rather than rebuilding and adding small portions as separate objects. I experimented with some local modeling to see if I could improve performance, and found that I couldn't. WM3D's collision engine seems well optimized.
In general, I found that the time to get to a preliminary, functioning simulation was very short. This is a tribute to CAD associativity. The majority of my time was spent optimizing the simulation to reduce run time, and this is the area where attention to detail and an appreciation for the problems associated with dynamic simulation prove valuable. My highest priorities for improvements to WM3D are features that reduce computation time associated with complex collisions, and additional examples and instructions for optimizing such difficult problems.