3D Printing and Prototyping

3D Printing and prototyping. It seems that these 2 things go hand in hand. Especially in the business world, additive technology is seen as a way to create physical prototype models faster than any other methods. Additive technology is also beginning to be used more in actual production parts. The housing and the breech of the Nimble are made from nylon printed using selective laser sintering (SLS), not unlike Boeing's use of SLS in the 787 Dreamliner.

Prototyping using 3D printing in our case was almost self referencing, using 3D printing to make parts for 3D printing to improve 3D printing as a whole. Nice!

How did we do it? It was a combination of printing in house and having parts printed. During the initial development we printed in house using our delta printers. But we quickly noticed that we were up against tolerances and wall thicknesses that were tricky to get right consistently, using FDM. So we turned to SLS printing and that allowed us to make a jump in quality and a size reduction. At the same time, we still printed other parts on the delta's, like adapters and effectors. It was a pretty intense development period.

The actual process went something like this:

• Initial idea to improve, we discuss it, think of ways to implement it and then model it in CAD.
• Take a good look, change it around till we think it is indeed a relevant improvement.
• The model then goes to Shapeways who print it and send the parts back to us.
• The parts are assembled and tested.
• Quite often to destruction. (it is surprisingly hard to wreck this nylon!) This cycle time is about a week.

Sometimes the developments are so fast that while we are waiting for the parts to arrive, we have already made changes and improvements to the CAD models and once the parts arrive, we can validate if the changes are good and order the new parts the same day.

Additive manufacturing and influence on development process

We started to really understand the strength of the material, both physical and in how to use it in the design process. A lot of time went into making the Nimble shape so that it could be injection moulded. Once we decided not to go that route, we felt liberated and could design shapes that were simply not possible to manufacture any other way. It was bliss!

Here is an example of a part that could have been injection moulded, but it would have been expensive. It would have required 3 pulls and very tricky seam management to be able to get the surfaces right. Using the SLS process we we able to just print it.

SLS, and the speed in which the parts arrived, allowed us to iterate faster and get a better understanding of what worked and didn't work. We were not afraid to fail, because within a week we would have an improved model. This meant that we reached optimal (minimal) size fairly quickly. As Colin Chapman said: Reduce complexity and add lightness. That we did.

One of the biggest lessons learned is that not all SLS processes are the same. After we had frozen the design and found a suitable partner to print in bulk (Shapeways did not want to do bulk work) we were amazed at the differences between the two suppliers.

These parts seem to be identical, but they are not. The right one is to spec, the right tolerances and bearing fit. The left one is 0.2 mm too wide, the bearing surfaces do not hold the bearing properly and the centre to centre distance is off by 0.1 mm. 

Parts that were perfect from one, were way oversized from the other, tolerances were out and we had a load of trouble to dig through. This caused a delay in shipping the first products. So now we do the final prototype using the production supplier.

Overall, prototyping with SLS allowed use to develop faster, and try more different shapes quickly. On top of that, it made us understand the material much better and allowed us to model weird shapes that could not be made any other way. All of these things resulted in a better product. The Nimble.