Computer simulations of complex, three-dimensional behavior.  We see it on television every night as an integral and intriguing part of the weather forecast.  You know, those color animated triple-Doppler displays.  And it seems to me that the accuracy of these weather predictions (if we discount anchor hype) has progressed a lot in the last few years and is now good enough to actually change our life, or at least our weekend plans.

If I might be cynical for moment – not usually my style, but here goes… Then why hasn’t 3D mechanical simulation really changed the way we design mechanical products?  OK, I know you’re going to say that your company has used finite element analysis and other simulation technologies to study all sorts of stress, vibration, and thermal problems in your products.  But does simulation really have the impact that it could and should have?

God knows I’m a believer in this stuff.  I vividly remember as a young engineer sitting in a restaurant in Little Rock, Arkansas (long before Bill was in charge) trying to convince an engineering manager that structural analysis was going to change the world.  Within a few years, I pontificated, there wouldn’t be a significant item in the restaurant – table, ceiling beams, fireplace grate – that would be released to manufacturing without first being thoroughly tested on the computer.  How idealistic I was!

Recently I visited a user site, almost thirty years after Little Rock, and my idealism was absolutely shattered.  This company designs and produces rail cars with a very competent staff and a reasonably well-run implementation of 3D CAD.  Now get this – out of the 40 full-time users at the site, they have one person using the simulation capabilities of the system.  When I found this out, I asked if they outsource a lot of their analyses.  Nope.  That one analysis specialist is pretty good, but I’d estimate that at best he has time to study say 1 out of every 20 parts and subassemblies that get released to manufacturing.  I have this nagging concern that the other 95% of the rail car components just might carry some load, might be subjected to dynamic excitations, or might have to dissipate some heat.  Oh, where is the mechanical engineering equivalent of the triple-Doppler that will finally convert the masses?!

The best and probably only way to permanently ingrain simulation in the product development process is to formally redefine the definition of "release".  At most companies there exists some manual that defines the prerequisites for "release".  Usually this boils down to creating enough data (i.e.  geometry on a drawing or, hopefully, in a 3D computer model) so that the part or subassembly can be manufactured.  The manual also requires certain checks on the data, some quite meaningful (inspection information) and some a bit archaic (oh, the lengths of those arrowheads must be correct!).

OK, here’s where we make an impact.  Rip up that old manual!  Take it upon yourself to work with some trusted colleagues to prepare a new definition of release, with the mindset of "we release when a design is ‘right’, and we use simulation to get the product right the first time".  Some checkboxes for your new release package might be… a complete self-documented 3D model, checked for packaging/assemble-ability/maintainability, checked for tolerance stackup and sensitivity, carries the loads without being over- or under-stressed, doesn’t vibrate, won’t overheat, etc, etc.  Management not only needs to approve the new handbook but also needs to instill the discipline that designs don’t get released until they are indeed "right".

And this mission isn’t complete until most of your designs are simulated and therefore, most of your design engineers use the simulation tools.  Every weather report relies on computer simulations.  Shouldn’t every design review?  It’s in your hands.  Let’s get serious – finally, my friends – about simulation.

A 35-year veteran of the computer-aided design, engineering and product lifecycle marketplace, Wayne McClelland is well-known as an energetic pioneer in the development and implementation of technologies for structural dynamics, finite element analysis, 3D solid modeling, and workgroup data management.  Wayne can be reached at waynemcc@wamware.com or www.wamware.com.