Commentary: Elements of Analysis, November 2006

By Richard H. MacNeal, The MacNeal Group

Almost two decades ago British Aerospace captured the accounting dataused to create a now well-known relationship showing that commitmentsto costs associated with producing a specific design are made early inthe design process when actual expenditures to develop a specificdesign are low compared to the overall cost. The graph of thisrelationship also contains an implicit management message: 80% of costsare committed during the first 5% of project expenditures. If that istrue, why not front load design with simulation-based trade-off studiesto ensure that design changes won’t show up in the detailed designphase where they will most likely lead to additional tooling costs,scrapped parts, and slippage of the product delivery date?
The value proposition of functional-driven design is that it deliversthe holy grail of product development: We innovate, we develop on timeand within budget, and we design robustness into the final product sothat it can withstand the unknown unknowns in operation.

Recent forecasts by marketing organizations that track the MCADindustry indicate that management has gotten the message aboutinvesting in CAE. Their current forecasts predict CAE to be thefastest-growing segment of the PLM market with compound annual growthrates projected to be 10% to 12% through 2009. But it remainsquestionable as to whether management understands how to best channelthe new investments. Evidence suggests that money will continue to bespent on CAD-centric, form-driven design processes rather than makingthe shift to CAE-centric functional-driven design processes that trulydo move CAE into the front of the design process.

So how did we get to the point of CAD-centric form-driven design? Themarket for CAD exploded in the early 1980s and soon led to theestablishment of major CAD companies. On the other hand, the deploymentof FEA/CAE was not nearly so swift. The inhibitors in FEA deploymentwere its CPU-intensive nature, its lack of a man-machine interface, andthe fact that very few engineers understood the intersection ofengineering mechanics, finite element approximation theory, andcomputers.

CAD had no such problems. The major problem that CAD suppliers had wasgaining its acceptance by individual draftsmen. However, the promisedproductivity gains were so large that CAD was pushed down into theorganization by management because it was an easy sell in the boardroomwhere stratospheric ROI (return on investment) was promised by the CADvendors. CAD was an easy sell because management could “see” the costassociated with acres of drafting tables.

The current focus of CAE model generation on the established geometrycaptured in the CAD model reminds me of the old joke in which a drunklooks for his keys under the street light instead of the dark alleywhere he dropped them.

I wonder if that’s not where we are in the CAD-centric mind set. We areaware of the advantage of front-loaded design but the light is betterwhen looking at the CAD-centric design practice.

It is interesting that the MCAD industry forecasts suggest thatengineering managers expect to use more CAE in the next 10 years. Thissuggests that they buy into the value proposition for up-front design.But I wonder if they are still “looking under the streetlight.” Hasanyone really bothered to tell them where the keys to front-loadeddesign really are?

Up-front functional design is a disruptive technology that can providesubstantial competitive advantage. I firmly believe the CAE supplierswill soon have the products and technology required to deliver thevalue proposition in force. The real question is whether top-levelmanagement is willing to move away from the CAD-centric streetlight andshine the front-loaded CAE light into the alley to support a paradigmshift in the engineering design process.

On the other hand, what have they to lose?

Richard H. MacNeal is a co-founder and chairman of The MacNeal Group(tMG; Altadena, CA), a company established to continue the developmentand advancement of CAE technologies. He was also a co-founder of theMacNeal-Schwendler Corp. (now MSC Software Corp.) where he developedmany of the underlying technologies of the NASTRAN finite elementsoftware program. Dr. MacNeal received his Ph.D. in electricalengineering from the California Institute of Technology. Send yourcomments your feedback about this commentary through e-mail by clickinghere. Please reference “MacNeal Commenatry, November 2006” in yourmessage.