To simulate or fabricate?

I answered an email recently concerning my belief that actually making models is preferable to using simulation software and more likely to end in success, and I think I should enlarge on it here.

When I said in an earlier blog, that this type of experimentation couldn't be done with any kind of modeling program, what I meant was that, yes, you can test an existing design with simulation software, but in the design process, you can't really rely on it to the exclusion of hands-on design. You may miss some simple alternative design or a small modification to the existing one that you can see in front of you when you have the actual components in your hands. When you can physically move a mechanism by hand and study its range of movement you may find that it becomes necessary to alter something to enable it to comply with your design. You may start the design on paper or in paint on the computer but at some stage it is better if you make the mechanism and see it in action.

Having said that, if I had the expertise and a sufficiently powerful computer to use a simulation programs, I'm sure I might decide to test out a particular design and see if it worked. But I don't so I must build it to see if it works and of course if I found that it did work in simulation then I'd have to build it then anyway. But I would still prefer to build it and study it in action.

One of the things that testing an actual physical mechanism shows is what I call 'tight spots' where at some point in its range of movement, usually at an extremity, the mechanism stiffens and becomes bound. This usually requires some loosening but that can have a negative effect in the part of the range where it isn't tight This looseness can cause lateral sway which may cause overlapping parts of the mechanism to interfere with the full range of movement, but in my experience this can be reduced to an acceptable level with the inclusion of spring washers or other springs. This kind of problem will not show up in simulations and yet it is quite likely to occur.

I suspect that it was this kind of use Bessler was referring to when he implied that he might use springs but not in the way people might think.

These kinds of problems and solutions do not show up in simulation software and for that reason I think it is better to make the parts from the beginning.

JC