Combining Additive and Subtractive Manufacturing

Anyone who has used a 3D printer has faced the reality that it is not an exact science. Parts can turn out misshapen, or fail to print all together. It was this frustration that has inspired my partner and me to spend the last 4 years developing and experimenting with how we can make better parts, bigger parts, or make them faster. Our first attempt focused on “better” parts, the idea that by using traditional subtractive technologies we could refine parts to have high tolerances and pristine surface finishes. This is how we set out to accomplish that vision.

We identified early on the need for a custom tool head, one that could switch from additive to subtractive. We wanted this to be automatic so we could switch multiple times in the same part. Before we could develop our tool head, we needed a platform to test it on. Luckily we had a chance to acquire a ShopBot Buddy. While the mechanical systems were fantastic, using the ShopBot software got in the way rather quickly so using the 37 pin connector and a breakout board like so http://www.ebay.com/itm/DB37-D-SUB-Male-Adapter-to-37Pin-Terminal-Two-Row-Screw-Breakout-Board-New-/302207741096?hash=item465cfc2ca8:g:HTYAAOSwUKxYi6U~, we were able to install a RAMPS board running Marlin firmware and print our first test cube. https://www.youtube.com/watch?v=v6G_m1X2JEw Now that we had a working platform, we needed to develop our head. The key here was iteration. Through multiple tests of various slide mechanisms and spindle motors we continued to refine our system and technology. It was during this time that we began to dream of a desktop printer that could seamlessly switch between additive and subtractive. So as our head began to perform acceptably we looked to create our first full prototype.

The first prototype was rough. We were lucky enough to find a awesome local waterjeting house to create some amazing parts for us out of .25 in aluminum. We still were running the ramps board from the ShopBot, but our motion system was much lighter (eliminating a lot of the table rocking seen in the shopbot video). This was accomplished using a wonderful product called OpenBuilds. This product is an awesome tool for anyone looking to make anything with linear movement. http://openbuildspartstore.com/ Essentially a linear rail built into aluminum 20x20 style extrusion, it allows for the combining of structural members and linear rails saving weight and money. We learned a lot on our R1 prototype, and we even took it to our first Makerfaire here in Portland, OR. With the feedback we got there we found a list of features of things that weren’t quite right.

No enclosure meant bits of plastic went everywhere Our tool changing still needed reliability Our bed wasn’t strong enough to deal with machining forces Our bed was underpowered and couldn't keep temperature above 70C. We took this insight and went back to the drawing board to recreate the next prototype.

The goal for our second model was what people call a “Looks Like, Feels Like” prototype. This meant that while the internals were still in flux, the goal was a printer that looked like the one we intended to bring to market.

The first thing we built was the new printbed. We started with MIC 6 aluminum. MIC 6 is an awesome product. It is ground to be incredibly flat, and because it is cast metal it has far fewer internal stresses making it less likely to warp under thermal stress. We then used a silicone heater that was powered off of 110V and switched with a relay. This allowed for far faster heat up times. These modifications left us with a highly reliable print surface.

The enclosure was CNC routed out of polycarbonate and then bent using a sheet metal brake. The front cover was mounted using inset magnets to allow for complete removal and easy access to the build environment.

To increase too change reliability we implemented our first “V-Head” this used two sets of linear ways to allow for the head to return to the same place when raised and lowered.

The final product created some rather remarkable parts. Overall this project was canceled due to a lack of interest from the overall 3D printer community. However, the learning from this project has fueled development of other new technologies by our team, and we hope that you can learn something from the process as well. All the CAD files are attached, feel free to poke around and use them as you see fit.