Functional Design Revisited

Autodesk's philosophy supports design by function

The technical software world always seems to have some controversy or point of contention swirling around. I’ve witnessed this time after time during the years I’ve covered and participated in the mechanical/manufacturing CAD industry. Recently, though, I’d be hard-pressed to come up with a software feature, capability, claim -- whatever you want to call it -- that’s stirred up more attention than functional design in Autodesk Inventor has. I’ve gotten more inquiries about functional design from users, competitors and even other editors -- what it is, what it isn’t, is it really unique and so on -- than just about any other topic in recent memory.

Functional design has materialized in Inventor 11 and is not just a command or tool set, but rather a method or philosophy. This is an evolving Inventor usability issue and a different approach to 3D design. I know that I’ll get arguments from competitors, but with functional design, form doesn’t necessarily follow function; instead, function can drive form. Although still relatively young with a lot of room to grow, functional design is a knowledge content toolset that represents a movement from geometric descriptions to rudimentary knowledge capture. Functional design isn’t just a set of functions for creating 3D representations, it supports design by function.

In my mind, functional design really began with Autodesk’s acquisition of MechSoft and its integration into Inventor. MechSoft was a comprehensive mechanical design reference source, mechanical calculator and assembly parts manager -- in short, a digital Swiss army knife for MCAD. What made MechSoft so innovative was its digital version of the mechanical engineer’s bible, Machinery’s Handbook. It was an e-handbook for mechanical design – on steroids. Another aspect that was especially appealing was that you could explore different design iterations with individual component parts, plus entire assemblies, such as drive trains and other types of mechanisms.

To help clarify the issue, last week I spoke with Andrew Anagnost, senior director of CAD/CAE products in Autodesk’s manufacturing solutions division, to get his take on functional design and the implementation of it. Of course, his is a biased view, but one I thought would set some things straight. The discussion was not so much a technical discussion as it was a philosophical one.

When asked what constitutes functional design, Anagnost said, “Functional design requires that engineering knowledge must be encapsulated somewhere in the [CAD] application to drive geometry creation in some automated fashion. Some of our competitors in the past have explored this method through things such as KBE [knowledge-based engineering], but their efforts stalled and they quit investing in it. Others have invested in what Autodesk classifies as geometry wizards for geometry creation or repair, and not true functional design environments.”

He went on to say that the workflow for functional design combines many factors, including engineering requirements, analysis, simulation, automatic geometry creation, visualization and digital prototyping. The goal is to spend less time creating model geometry and, instead, provide for more true problem solving within the CAD system. Functional design is not intended to automate the design process, but to automate the geometry creation process. In other words, the intent is to enable users to spend more time designing and less time worrying about how to create geometry.

I asked Anagnost to compare parametric design with functional design. He said that parametric design is geometry-centric, whereas functional design is engineering-centric, and minimizes the need for geometric parameters. He said that although parametric design introduced the ability to change 3D models, it came at a price. It added constraint systems -- things that users have to figure out -- which increase complexity and time. Parametric design demands that users understand how their modeling choices or steps involved in creating a model impact their later use of the model. According to Anagnost, “Autodesk feels that the next generation of CAD must remove the geometry burden from users so they can be engineers first and CAD users second.”

Anagnost said that the initial target market for functional design is machinery design, but it will ultimately spread to all areas of mechanical design. He also said that functional design is figuring into all decisions that Autodesk is making with Inventor moving forward, and is the benefactor of significant amounts of R&D.

Anagnost stressed that Autodesk is not trying to replace engineers, but make them more effective by giving them the tools to use CAD more effectively. He said, “Functional design is not the end of CAD, but a redefinition of it. We want CAD to be viewed as an engineering tool, not just a geometry creation tool. We want to change the way CAD is applied in the engineering process.”

In a nutshell, that’s functional design. Is it entirely unique? Well, no, as some competing products do have bits and pieces of what it provides. However, to my knowledge, no one else has the breadth and depth of what functional design offers today at a reasonable price point.

Will functional design make or break a mechanical CAD sale? Probably not right now, but in the future, as its capabilities expand and users discover what it can do, it might become more influential in a buying decision. Even though it’s still early in the game, functional design could potentially help users feel more confident that their designs will, in fact, perform as intended. Obviously, there is a lot more to functional design than can be covered here, but it’s important and one of the aspects of Inventor that will continue to distinguish it down the road.

For more on functional design, read Cadalyst’s online article by Kenneth Wong, “Inventor’s Functional Design: Is it Really New?” where he finds out what competitors think of Autodesk’s new catch phrase.