Here we are, at the final stage of our Design-Build-Test of our office chair arm replacement! If you haven’t seen the Design and Build phases yet, go check those out first, then come back to see how it all came together!
So after all the hard work, the time spent modeling, and the time dedicated to building this piece of replacement, surely it’s just a simple plug and play, right? Wrong! (at least in our case, anyway)
Rinse and repeat
At each iteration, we tested out the printed part on our chair to determine what dimensions we needed to adjust in order to get the right fit. Here are just some of the issues we ran into during our iteration tests:
1. Our first version was too small and actually couldn’t fit on the chair arm
2. The second version, while we could insert it over the arm, couldn’t slide all the way into
position because the channel was too narrow
3. The third version had printing issues so we couldn’t complete the piece period and had
to go back to the modeling board to make the piece more printable (our handy dandy
Spaghetti Detective could have been really useful here and saved us many a headache).
4. And so on and so forth…
We were able to correct the scaling and interference issues from the previous versions by altering the print orientation and utilizing Magigoo (our go-to bed adhesion solution), but still needed to go through additional iterations to get the final details correct.
What went wrong
These iterations and tests continued until we finally landed on a working chair arm with no fit or use problems. Now that we have the right part, it’s time to look at what issues we ran into during our tests so we could prevent some of those issues from arising in our future designs, because as much as we love rapid prototyping, the best way to cut down on lead times is to get the model as close to right the first time around.
1. Measure twice, cut once
When starting from scratch, dimensional errors can stack fast. The more assumptions or estimates you make can lead to more problems and worse assumptions. So be sure to carefully get your measurements and get them right. This will greatly reduce the chance of running into fit issues. For example, we made an initial assumption that the pictures we used as tracing reference would offer sufficient scale and accuracy. However, slight variations in the tracings came back to haunt us; these small variations were exacerbated by additionally assuming that we needed to scale up the model by 10% to allow for fitment variations. These two assumptions started stacking errors both in the model and in the print including impossible geometry, difficult-to-print sections, and unstable build mechanics.
2. Materials matter
Remember to take into account the properties of the original material and the material of the 3D printed part in order to account for their differences in your design. The original chair arm was made out of flexible material whereas our 3D printed part was made from our U-HIPS material which is rigid. To accommodate for this material difference, we added more support structure under the tongue of the chair arm. The original chair arm is meant to flex over the metal prong that anchors the chair arm to the char. In our case, this would not work with HIPS material, so we had to alter the structural support of the model itself to account for this.
3. Designing for use is not the same as designing for printability
A good example of this point is a simple extrusion print. A two-dimensional object that is then given some 3D attributes by extruding it into the Z-direction is very easy to print, but not terribly useful as a chair arm. The chair arm, a flexible part that interacts with metal fittings and fasteners, is a fairly simple idea; however to make something like it using desktop printing requires some compromise. In this case, we had to make the chair arm bulkier, thicker, and more angular (less curvy) to ensure that it could be printed with minimal problematic overhangs. So we had to sacrifice some of the aesthetic qualities of the original part in order to make the 3D printed part work.
4. Additional hardware comes with its own tolerances
Any design that requires other hardware like metal fasteners will also need to take into account the tolerances of those hardware pieces as well. This will help ensure that the fit is right for both the hardware and the part itself.
Our tests failed so many times - 35 times to be exact. But! This is precisely what makes 3D printing so revolutionary in the prototyping phase of any design. It may have taken many tries, but rapid prototyping allowed us to quickly iterate and test until we got it exactly right. And with any luck, we’ve learned from some of the mistakes we made this time around so that our next Design-Build-Test will be more efficient and more precise. And now we have an awesome chair armrest for our office!