Peter Fearey

3D Printing and Scanning

Assignment

Design and 3D print an object (small, few cm) that could not be made subtractively. Also, 3D scan an object (extra credit: print the object and make the scanner).

Ideas and Approach

I have been fortunate to have been able to make a few items this year with our school's Replicator 2X and Digitizer.  That being the case, I decided to work on three different challenges this week...
1) Create a part for a new "exploratorium" wall at the school where PK/K students will build marble wall for constructing a cool sculpture.
2) Scan, cleanup print/generate a simple 3D object (a Dunkin Donuts cup), and
3) Take a 123D Catch scan of myself and, after cleaning up the model, printing it. 

Files

• Final "Marble Redirector" STL file
• Scan of me after cleanup (Meshmixer file)
• Scan of Dunkin Donuts Coffee Cut (Meshmixer file)
  

Steps Taken

Marble Parts

For the first part of the assignment I decided to help a fellow teacher who wanted to create a "marble wall" for students. 

	Here's a creative "marble wall" that our primary school librarian is building.  Her goal is to make a wall where students can create a contraption that will move the marble down the wall.  Her hope is to make this a way to explore everything from angles to probabilities to simple physics.  Given this week's assignment, I thought it would be cool to make something that would catch the marble and redirect it.  It needs to have a hole for the marble to fall through and then a path that will redirect the marble back or straight down.
	Here's a printout of my first draft.  It has a few issues... 1) The prongs are long because there are two sheets of pegboard (they put one behind the first to make the pegs more secure).  The thing is, they aren't perfectly aligned, which I didn't realize until I tried to mount it.  FIX: In the next round I am going to shorten the bottom ones so they only need to go through the first piece of pegboard. 2) I felt like a total idiot, but I didn't realize the bottom hold pushes the marble out "the front" rather than through the side.  FIX: Rotate the hold to the left or the right. 3) The hole is the right size, but the marble flies right over it.  FIX: I need to add a backstop that acts like a funnel.
	Here's another picture of the first draft mounted on the board.  Some of the issues mentioned above are more easily seen here.
	Here's the STL file / version 2 with all of the changes proposed above...

When all was said and done (and after four different revisions of the part, here's what we we finally ended up with (with the entire Marble Mover) pictures.  Checkout the green starter part in the top :->...

Dunkin Donuts Cup Scan

Using the school's MakerBot Digitizer.  I scanned a Dunkin Donuts coffee cup.  I had two goals with this:  1) show different ways of outputting the model once it was done , and 2) learn to do simple editing of a mesh. 

	Here's what the coffee cup looked like right after it was scanned in using the MarkerBot Digitizer.  Notice the strange growth on the left side of the cup.  That's what I will eventually try to get rid of.
	Here are three different kinds of output that I generated once I had the STL file of the coffee cup (from left to right): 1) 2Bot Subtractive Cutaway:  Using the 2Bot milling machine, I cut three pieces of 2" think foam insulation out into the three pieces of the coffee cup.  The cup is just under 12".  2) 123D Make Cutaway: Using 123D Make, I was able to get the laser cut plans that would turn the coffee cup STL into linkable parts.   The cardboard behind the different coffee cups shows how the parts were laid out on a piece of cardbard after they were cut (I didn't use enough power so this cut didn't go through). 3) That's the original 4) That's the shrunk MakerBot version of the coffee cup.
	Using MeshMixer, I was finally able to clean up the lid (a bit).  While it wasn't perfect, I did get it much better so there isn't as big a growth on the far side.

Scanning using 123D Catch

Once I had the simple coffee cup cleaned up a bit (and I was starting to understand the basics of MeshMixer), I decided to try something more difficult.  So using the iPad (and the 123D Catch iPad app), I tried to make a funny scan of myself.  Then once I had that scan I tried to fix many of the problems. 

	OK, I admit it...I was messing around to see how it would handle the overhangs.  I also thought my brothers would really enjoy my attempting to flex.
	While my first reaction was that 123D Catch did a wonderful job of generating the model, I later started to see many flaws that I realized I was going to need to fix.  Some examples: 1) My right hand is merged with my head. 2)The background is included and mixed in and I need to get rid of it. 3) My body is hollow on the inside 4) I need a flat cutoff on my torso so it will rest on a table 5) There are many extra growths in my model.  For example, my left arm has some extra While there were many extraneous parts on the perimeter, I later found out that these are easy to delete in tools like MeshMixer.
	Wow...here's the model after it was imported into MeshMixer.  What a hot mess!  You can see how small I am when all the walls and surrounding mesh are included.  Fortunately, this was quite easy to get rid of.
	After 90 minutes or so of trying to learn MeshMixer, I ended up with a halfway decent model of my silly pose.  Here's how it looked in MeshMixer.
	...and here's how it showed up on the MakerBot print bed.

Learnings

Here are my key learnings from this lesson:

• Label Output: Label the parts as you take them out of the laser cutter.  You'll notice that the cutouts aren't in the perfect order (and we are missing one at the top).  This was because I was speeding through all this and I thought I could figure out which one went where...I was wrong and I ended up breaking a piece because the fit was so tight it ripped. 
• MeshMixer (and other mesh editors) will take some getting used to.  I really struggled with learning some of the basic concepts.  For example, there were so many ways to smooth items and none of them did exactly what I wanted.  I found the quick videos interesting, but a bit useless because they were all based on an old version of MeshMixer.
  
• Planning Pays Off: I made some dumb mistakes with the marble project.  I was rushing and didn't bother drawing this on a piece of paper first.  That was silly so I wasted a bunch of time.  It really pays to draw the pictures on a blank piece of paper first so you don't waste time. 
  
• 123D Apps are good, but platform matters.  Many of the 123D apps work across all the platforms (Mac, PC, iDevice).  The thing is, some of them are pretty buggy.  The 123D apps seemed to crash a bunch on the PC.  I did have good experiences on the iPad. 
  
• Save Often: Many of the programs I used (123D and MeshMixer) crashed a bunch.  It was a bit of a bummer.  You need to save often.