[starbox] Much of my design work involves “math modeling” gearsets to determine the lowest-cost components which will meet the requirements. Over the years, a designer develops his or her own set of guidelines for what geometry is acceptable and what is objectionable. A group of gear experts may agree on eighty percent of a design and argue for hours over the remaining twenty percent. This is particularly so in the area of “microgeometry” — lead and involute modifications that help improve performance at the extremes of loading. Software packages include sub-routines to suggest values for these modifications. Unfortunately, microgeometry requires a very strong understanding of what loading conditions are. Many times we do not have adequate data to support the assumptions we have to make. Pushing buttons on software input is easy; modern grinders make it possible to produce whatever shape the program asks for. Sadly, this mechanization can compound the problem. We once got FOUR different revisions to a lead chart during an eight-hour period. The machine’s operator could be forgiven for being slow to implement the second and third changes. The lesson we took away was to spend more time reviewing assumptions, and agreeing on them, before pushing the buttons for modification simulation. Sometimes the charts just “don’t look right” to the experienced eye. If this happens on your project, don’t get angry. Instead take the time to review with the commentator why the assumed load conditions would result in the shape shown on the chart. If you can’t convince them, you might want to take another look at those assumptions. The beauty of computer simulation is that you can consider many possible configurations in a short period of time without endangering the precious physical parts. The ugly side of computer simulation is that it is so easy to input faulty assumptions. The old adage — “garbage in, garbage out” — still applies. Spend the time needed to understand and verify load conditions before pushing those buttons.