 |
| Hey -- where's the wheel coupling? |
Perhaps you've been in the same position: you just bought something in a great deal -- but when you get it home you find something important is missing.... This happened to me recently when we got an excellent stainless-steel table in an "all-sales-are-final" auction -- was so excited to find it was actually brand new and had great wheels, but when we got it to the office and opened up the boxes the wheel mounting couplings were missing. Without them the wheels could not be attached to the table, which made the table almost worthless.
But I'm a Maker, and I can fix this! I hope you enjoy this story of how Maker tools -- 3D design and printing in this case -- were applied to a practical problem.
For several reasons (perhaps to be discussed another time) I decided to use PVC pipe as a supporting sleeve around a new fitted 3D-printed coupling. There would be some fun design in this, as some parts were round and others hexagonal, and all parts were of different sizes. There was also a little engineering involved to devise a way to deal with potential lateral stress since FDM 3D prints (FDM = fused deposition modeling, which is the most common technology used in current 3D printers) are susceptible to this, as lateral stress tends to break the weakest part of FDM prints (the inter-layer bonding).
 |
The red component is the tighly-fitted
3D printed coupling. I will be supported
on the outside by a PVC sleeve. |
 |
A 2-part PVC sleeve was placed over
the 3D printed coupling for more
strength and stability. Note: there's
another fitted 3D printed ring under
the larger section of PVC. |
For the sleeves to be strong with minimal susceptibility to lateral stress, the fit of the 3D printed pieces to the existing mechanical parts had to be precise. To achieve this I used my basic rule for 3D design against existing parts: measure twice and print thrice. Truly one of the great advantages of the whole desktop manufacturing revolution-- with 3D printing as the current crown jewel -- is rapid prototyping. By designing some test prints I was able to quickly arrive at the exact dimensions for all the critical parts, fitting both the internal and external surfaces of the existing mechanical parts. But why did it take up to 3 test prints to achieve this? Because the goal was to achieve a fit with a tolerance of less than 0.25 mm over imperfect surfaces so there would be no mechanical wiggle or rocking of the joint, and I wanted to avoid the use of glue. 3D printers do a variety of things that slightly alter the radius of curves (a good topic for another article), so when you need an exact fit some test prints are highly recommended. When fitting to existing curved surfaces, it's also critical to use a 3D printer that can print near-perfect circles (curves) -- so I used that one that from experience gives me the best circles. Because of the test prints, the final 3D printed coupling fit perfectly on the first full test.
 |
| The final result -- a full-usable rolling table! |
The final 3D prints fit snugly -- so well they needed a little "encouragement" from a rubber mallet to slide all the way on the existing pieces. Normally I'd be a little worried about putting that much stress on 3D prints because cracks could eventually appear, but since everything was to be held together by an equally-tight PVC sleeve I'm confident no stretching or expansion (and therefore no cracks) will happen. Here's what the final result looks like -- it's perfectly stable, level and rolls well. The table has been supporting more than 100 lbs of load for a while now, with no problems. 3D printing wins again!
