Friday, January 24, 2014

Building 3D Printer: Part II

Like I said in my previous post, I've wanted to design my own printer since I built my Clonedel. I kept putting it off due to a variety of things, but that's neither here nor there.

In summer of 2013, I got an email from the MSE Lab Manager, Tuesday, asking if I would help them build a 3D printer for the MSE Junior labs. I didn't have anything to do at the time, so I agreed.

I had started compiling a mental list of printer kits, and various options, all based on the department's budget, when I attended a meeting to discuss the printer with Tuesday and some other students (Ryan and Jacob). Tuesday explained that we had a cool 1000$ to play with. I was pretty happy with that amount -- you can get a darned good kit for that price (And the FlashForge is sometimes on sale for that: dual extruder, fully assembled, calibrated!).

Tuesday then explained that she wanted us to design the printer ourselves. I was a bit thrown by this, but ended up agreeing... After all, this would motivate me to design that printer I had always planned on!

(In retrospect, I feel that asking 3 MSE students to design a 3D Printer from scratch is asking for failure... CAD design is not required for MSE, and there is very little chance that they have had any sort of mechanical design training. Fortunately, we all had some CAD experience, and I had been around printers enough to know how not to make them)

Regardless, we sallied forth. We put together a timeline

6/27 - 7/4 ==> Model printer
7/4 - 7/11 ==> Order parts/print parts
7/11 - 7/18 => Assembly 
7/18 - 7/25 => Assembly
7/25 - 8/1 ==> Calibration
8/1 - 8/8  ==> Calibration
8/8 - 8/15 ==> Write Manual
8/15 - 8/22 => Write Manual

Anyone who's had any design experience can tell you what is wrong with that timeline...

ONE WEEK TO DESIGN A 3D PRINTER FROM SCRATCH!?!?!

This was the first big mistake our group made. We actually accomplished a huge amount in just a week, and ended up with something like this:

The main frame was to be composed of aluminum angles. This goes back to what I said before about how not to make printers. I wanted a rigid frame to support my prints. The Grawmet has no frame, and is a flimsy little printer. The Mendel's not much better -- its frame, while strong, gets in the way of the printer, and relies on a million and one nuts to be held together. Another WOOF officer, Mark, had what was essentially a maraca after bringing his Clonedel on the buss.

I wanted something that looked good, used a small amount of fasteners, and was practical. Angles fit this profile pretty well, and are pretty cheap, so I went with them.

You'll notice that the top corners are all just lined together and attached with screws (Screws not shown). I had the gumption to make the assumption that my aluminum angles would be 90 degrees...

*CUE THE FIRST DISASTER*

A screenshot of the "Aluminum Angles" page on onlinemetals.com
If you don't know 6061 and 6063 are different alloys of aluminum. I went with 6061, since I've used it before. The angles we got had the fillet on the interior, as seen above. This, of course, ruined everything. After a few days of confused searching, I finally found the above page, showing that to get a right-angle geometry, I had to order a specific alloy. It was rather confusing, but the "fun" didn't end there.

Oh no. This was only the beginning.