Tips on Part Fixturing Inspections

Canadian Metalworking's Metrology Matters blogger Bill Reilly talks about setup tips and techniques for part fixturing inspection.

When it comes to fixturing parts for inspection, I think I’ve seen it all. OK, maybe that’s a stretch, but I’ve seen and personally experimented with a wide variety of techniques.

At my lab, Quality Inspection Technologies, the parts we inspect run the gamut, from small plastic injection moulded parts to stampings, castings, machined parts, forgings, plaster casts, fixtures, the list goes on. The list of fixturing techniques is just as long so how we go about holding the parts we inspect is integral in obtaining valid data, and the method employed varies from part to part depending on the type of part and what the measurement requirements are.

For instance, how would one go about restraining a small plastic injection moulded component about the size of a thumbnail? The method that works best for me is…crazy glue! A small spot of crazy glue on the end of a small pin gauge holds the part rock solid with no risk of distorting the part.

I wrestled with different methods before hearing about this simple technique from a fellow CMM technologist, and crazy glue has been a staple in my lab ever since. Remember, crazy glue is your friend, but use sparingly or you’ll be making shadow puppets for the rest of the day (stick your thumb and index finger together…you’ll see what I mean.) Hot glue also works well on a larger scale. The glue sets up fairly rigid and usually comes off clean. Take it from me though, steer clear of the dollar store glue guns, I had one go off like a flash bomb!

When deciding how you’ll fixture any given part, it’s essential to know the quantity of parts you’ll be inspecting. For me, anything up to 6 parts or so, I’ll stick with rudimentary set-ups that serve the purpose but don’t take a great deal of time to create. No point going all Rube Goldberg on it if you’ve only got a few parts to inspect.

The other end of the spectrum is ongoing inspection in a production environment, where you’ll be inspecting anything from several parts each shift to 100 per cent inspection for the life of the project. Here’s where you want to have a custom holding fixture built where you can easily drop parts into the fixture and run your program without having to take any manual hits to align each part since the fixture repeatably positions and orientates the part. In my lab we don’t perform any ongoing production inspection although our sister company, Elite Tool & Gauge, designs and builds holding fixtures which can be part of a full turnkey CMM package consisting of a holding fixture and CMM program. Somewhere in between the two extremes is where I’ll construct a sort of temporary jig which also positions and orientates each part allowing me to just load parts and run automatic. I’ll use components like 1-2-3 blocks, V-blocks, jack stands, magnets, various sizes and styles of vices, etc. which serve the purpose just fine for the short term and then can be disassembled and used for another inspection project.

Time is money, so for me and for my customers, the time spent on fixturing needs to make sense (and cents). There is another option in modular CMM fixture kits which are available from such manufacturers as Rayco, TE-CO and Witte, and can be used for a wide variety of parts and consist of dozens of standard components such as spring clamps, magnets, stand-offs, spring buttons, etc. These kits allow you to build even the most complex fixtures if you’re creative and mechanically inclined.

No matter how you go about fixturing the parts you inspect, it’s absolutely critical that you perform some sort of repeatability test in the set-up you’ve created. I recommend running the same part at least a few times, even on just a few key dimensions, to get a sense of how well the part is being restrained. If the results vary too much, it’s back to the drawing board.

There’s nothing worse than running 50 parts, only to find out the set-up was not repeatable and the data is unreliable. It may sound time-consuming but it’s just one of the costs of reliable data.

If you have questions for Bill or comments you would like to provide, please reply here.