Hybrid Machine Design

[nnfuller]

Combined Additive and Subtractive Machine

Additive manufacturing has come a long way in the past decade, specifically in the areas of FDM (Fused Deposition Modeling). We have seen astounding growth in methods, precision, materials, and speed. Still, certain instances occur where the desired combination of traits for a specific application aren't attainable in a single step. Here post processing such as vapor finishing, sanding, or even post machining have been used to help create parts that more fully satisfy the demands of industry.  Post machining shows perhaps the most promise for creating engineering grade parts, due to the high level of controlled precision compared to the previously mentioned methods. It is also the only process which can increase the precision of a 3D print. Up till now post machining required a second highly sophisticated machine, with additional work holding and tooling. This is the issue which this project will attempt to address. Below are my comments on Pros and Cons of Hybrid Manufacturing.

Pros

Increased precision

Because the printed object isn't repositioned, there is no uncertainty as to its location.  The standard practice of touching off reference points can cause problems when using a 3d printed part is uncertain dimensions.

Increased Throughput by Decreased Print resolution

Parts designed to be post machined can be printed faster at lower resolution, then have critical areas machined to tolerance. These parts could be thought of much like a casting, where non critical areas remain rough while other precision areas are honed within the desired tolerance.

Machined Internal Cavities

At a high level of functionality, this machine could be capable of more than a 2 step process, machining geometry after any number of layers had been printed, accessing areas that would later be covered by additional material,

Cons

Increased Machine Cost, Weight, and Complexity

Because the system must have the all the traits needed of both a 3D printer as well as a mill, there will be certain trade offs required which will likely result in greater weight and cost, while a head that can alternate between additive and subtractive processes will be more complex than a single purpose head.

It is my opinion that the benefits mentioned above outweigh the potential detriments and obstacles presented by combining these two processes into one machine.

[nnfuller]

Current State of Project

In order to pursue this project, I first needed a reliable set of axes that promised to be both rigid enough for machining, and quick enough for printing. The ShopBot Desktop was the perfect tool, and Autodesk was able to supply one that could  be hacked in order to function as both a machine tool and a 3D printer. Below is the shopbot with the bed removed for maintenance after shipping.

The next step was to create a CAD model of the system to allow the additive head to be retracted to allow for machining or lowered to print. This was all done in Autodesk Fusion 360. Preliminary renderings can be seen below.

These parts have been printed and I am waiting on hardware and Aluminum plate before the parts can be mounted. Example of printed hotend carriage can be seen below.

[daniel_lyall]

the only problem I can see with it you are going to lose cut size, print size there are ways around it but not on that machine the x axis will need extended to over come that problem over run on each side of the machine

[nnfuller]

You are totally right. For now I will only be using a small portion of the 24" x 18" bed area. My hope is to use this Shopbot to prove the technology and then design a custom machine for this purpose traits like an extended X axis, and more Z cut height for taller parts.

Another huge obstacle is workholding. We are still working on how to secure the printed part to the bed well enough that it can be machined. Would love to hear people's ideas and solutions!

[daniel_lyall]

a vacuum table that can be easily be turned of an on.

to do it in a small form just use a moving table design with over hangs on the x axis to accommodate the increasing Z axis size

[MichaelAubry]

I love this project. DIY combined with progressing the future of manufacturing. I'm really excited to see how you progress. 

[nnfuller]

Made some progress this week. Here is the printed and laser cut extruder head:

I also created some concept art for apply for a grant to develop a comercial version of a Hybrid Machine: 

[MichaelAubry]

How are you sizing your parts to fit?  Are you downloading custom content or are you using calipers?

[daniel_lyall]

your printer head looks good 

but your frame is a but off it would be fine for a 3D printer but not a router it would be ok if you went slow but if you run it at a fast feed and speed it would move around just look how the shop bot is built its that way for a reasion. the shopbot desine is not the best ever it has some problems with staying straight racking of the X axis.

also if you wont to have diffrent heads on a machine you need to acount for the size of the Z axis if you keep the Z axis inside the uprights you will have a lot of dead space on the machine you will need to have the x axis connected to the frount of the uprights and have it so the head can go past the uprights.

also add a laser engraver to it as well you will easyley be able to sell a machine if it can cut, engrave, print 

[nnfuller]

That's some great advice! I like the idea of having the heads able to extend past the uprights. The laser engraver option is definitely a cool idea, and will require some research into how powerful a laser we can get while still being affordable and compact.

[nnfuller]

We currently design use either engineering drawings and model of the shelf items ourselves, or download parts from sources like McMaster Carr. The real trick with design for 3DP is tolerancing. We have spent a great deal of time learning the proper offsets for different elements. If you look at the extruder model for example, the bearing are press fit into the part for simplified assembly.

[daniel_lyall]

what program do you plan to run this machine off, a 445 2w laser with control will only add a couple of 00 to the price.

if you wont I can send you a desine for the rebuild of my small router I am changeing the X axis to give more travel inside the uprights

[nnfuller]

Weekly Update:

Firstly we have managed to drive the Shopbot using the RAMPS 1.4 control board. This is a huge step because it gives our team complete control over the firmware. However as of right now we do not have endstops, so cannot move much for safety concerns. We are looking to use limit swithches on the X and Y, but we still need a good way of sensing Z. If anyone has a cost effective example of how modern CNC tools would do this please let us know. Our team has also finished the driving portion of our bowden extruder pictured below:

I will try to post video of the shopbot moving once endstops are in place. 

Cheers,

Nathan

[daniel_lyall]

in what way do you wont  sensing of Z

[nnfuller]

As a team we have talked about implementing either sensing system that detects when a drill bit contacts a corner of the bed completing as circuit, or using a touch plate. Other ideas would be welcome!

[nnfuller]

Progress has been made! After one of the Fused Machine team memebers milled out the almuminum mounting plate we were able to mount the extruder on the shopbot and reattach the bed now that all the new wiring has been routed through the machine. See the pictures below:

Speaking of wiring, also got a bunch of help from the Fused Machines software guru Andy, and the Shopbot is now wired into the RAMPS 1.4 board. Andy is still tweaking some firmware but we hope to have this thing moving around here in the near future. Picture of the Frankenstein wiing job below(the stepper motor was for diagnostics):

[nnfuller]

Our extruder is work! So you may have seen in an earlier post that there was an extruder design posted, but now with the addition of a E3D bowden head, we have finally gotten everything working. Here are some pictures and a quick video of our setup being bench tested before we take it to be mounted on the test machine (ShopBot).

[carloquinonez]

Hi,

I was wondering what software tool chain you're using to generate the toolpaths and run the shopbot?

Regards,

Carlo

[nnfuller]

We are still wrapping up the machine design phase of this project. However, we are expecting to use Slic3r to generate the 3D printing Gcode, and Fusion 360 to generate the CNC Gcode. These two codes will then be spliced together, first by hand and then by an program written by the software team at Fused Machines. I hope this answered your question. 

Nathan

[carloquinonez]

Thanks for the info Nathan. So the makerbot is running a RAMPS board, right? Did you take a look at the TinyG? https://www.synthetos.com/project/tinyg/

And if you had to do it all over, would you stick with the shopbot or switch to a different gantry?

Thanks,
Carlo