Week 07:
Computer-Controlled Machining

Objectives

Make something big on a CNC milling machine. We have a ShopBot.

Software

Machines

Practices

 Methodology for how to make a CNC milled product

 Research

This is basically a running blog of my thoughts:

First inspired by Christoph Niemann's lockers for BluePrint Co-working space in Hong Kong.

However, attracted, finally, to a more efficient use of material. I like this design b/c it looks even better when there's more of it:

Now, let's start with the idea that one of the shelves could be a standing desk surface.

JOINTS

And look at this! A lovely CNC milled shelf system

And the Kalo chair, that highlights materiality; a milled pattern that is gorgeously irresolute; and some biomimicry:

And,

I add another constraint into the final project: wabi sabi. See this example of raku kintsugi, a Japanese ceramic technique. Can I get good enough at 3-D modeling while in Fab Acad to model something like this?

 Drawing by hand



 SolidWorks: parametrics

NOTE: Control 8 is critical for resetting the view for us non-Windows users who keep skewing the picture plane. Because of the skewing, I had plenty of lines become angled that were supposed to be horizontal / vertical.

See below some global variables set in SolidWorks in an attempt to help myself out with parametrics. I am using the variables, converted to a text file and then linked to the next SW part file, to define the parameters of the part files.

"vertical leg width"= 150mm
"tab width"= 55mm
"angle offset from vertical"= 10deg
"material width"= 15mm


I'll add one file here, which is a quick test for a type of joint in Rhino.

joint .3dm

 Tolerance test on the ShopBot

The best fit for our 15mm OSB is, surprisingly, tabs of 15mm! No offset seems to work best for a press-fit joint.

Starting a daily account of the shelf-making ...

16 Mar 16. Wed.

As my fluency in SolidWorks increases, I succumb and resist and succumb to making a more and more graceful design. I do hold myself accountable for not having actually milled a single part yet; I have fallen prey to my design and the pleasure of getting some sense of competency with a 3-D program, for which I'd been starving in the Fab Acad context. My struggle today was with working with the SolidWorks assembly. The various parts files go together, or "mate," in the assembly file. I have some tricky notches that I was fiddling and fiddling with to get to match. The challenge is because the sides of my shelf are diagonal, not vertical, and they are mating with horizontal shelves, so I'm finessing the joints so there's no obscene gap, but instead a modest join.

Explanation of the block of pics below, starting from top left and going clockwise:
(1) Santi said this solution was nonsense and suggested I be more discreet. I agreed, and made
(2), which, as you can see from the pic, did not match up perfectly. After a lot of prodding (like a blind piezoelectric Roland scanning sensor), I discovered I was working with different depths for the shelf in the part files! This is easy to do when the measurements are kind of arbitrary in the first place, and you just change them 'cause you feel like it (this is not a very sophisticated way of designing, and this lack of sophistication installs itself in your files if you don't get up and fabbercise regularly).
(3) You can see all the small measurements on the shelves to make the notchings fit.
The bottom left drawing is of the "simple" finished shelf joint.

CLOCKWISE:

How it turned out:

17 Mar 16. Thurs.

/// struggles in parametric design

Having been so proud of using equations in SolidWorks, I'm now spending a long time resolving errors that arise because of them. It is holy work, of course, having watertight numbers ~ but I need to slip whole days of hours into the nights to get it done.

This slot-plus-pocket took me hours and hours to figure out the right sequence of commands so as to not to get errors and accomplish the form.

Inspired to see a beautiful pattern in strange cancer of arrows that popped up:

18 Mar 16. Fri.

Going to lasercut, running into problems to get the file to cut. The computer is not communicating with the Spirit Laser Pro, which I have not used before. I spent a half hour watching the attempted resolution, then abandoned my hawkish and birdbrained observation in order to recoup some time finessing the design on SolidWorks.

Gori reminded me to ALSO scale the thickness of the material: it needs to be 3mm. I have a secured a 3mm piece of scrap and fit the parts onto the footprint of the scrap piece.

1:5 scale of parts to lasercut

/// lasercutting: the MultiCam 2000

Ingi did a tutorial on the MultiCam 2000. It's overkill for my project, but the Epilog was occupied with a multi-hour project and the Spirit pleaded the Fifth and refused to talk.

* Check out a great tutorial on the Multi-Cam! *

operating MultiCam 2000

/// chamfering a(t the) bit

And another pot of functions in SolidWorks: the chamfer / fillet command. I played around with these.

In SolidWorks, I used the measuring tool to determine how to determine a distance-to-distance-defined chamfer:

I tried a fillet on the main shelf with a radius of 15mm. How is the top edge of this fillet going to look, I wonder? Will it be a smooth line to the main surface, or a slightly recessed edge? We'll see ...

Go wild clean angry Multi-Cam, go!

For some reason, multiple paths cut. "Sel Dup" in Rhino to eliminate unnecessary overlays of lines. What is the equivalent in SolidWorks?
On further reflection, I think the machine seemed to cut multiple times in the same place because of me and my fancy fillets.

/// cut extrude

A major challenge I had with SolidWorks is realizing that, in order to make a pocket cut or a track, I need to make a "cut extrude" that exists IN A DIFFERENT SKETCH from the part on which I'm working. The sketches do a boolean between each other. Lord, I spent hours fussing with that one. Consequences of the bad judgment of a tired person. Did you know we actually lose IQ points when we are underslept? It's the same as coming to work drunk. It might be more refreshing to stop work and actually get a bit drunk instead.

/// lessons from lasercutting a prototype

Maybe I need another part. The unbalanced design I created is nifty and sleek, however, considering the uncertain circumstances and multiple users at IaaC, it seems foolhardy not to add a part that wil stabilize the shelf if someone pulls it away from a wall.

I also learned that, in order to prototype, it might be better NOT to use designs with pockets, because the laser cutter cannot imitate a half-width cut. Something to consider.

/// prepping a 3-D SolidWorks file to lasercut:

Here's the process for getting a file from .sldprt to .dxf:

lasercut .dxf

19 Mar 16. Sat.

IaaC is sweet and calm today.

/// patterns

Working on design adjustments based on the model I lasercut last night. I thought to make some more triangles. And fiddling with chamfer / fillet adjustments. Getting the hang of cuts.

I suspect my top shelf is in the wrong position. I am impatient now, though, and simply want to mill a product! And make adjustments based on that afterwards. So, finished enough for now:

/// laying out the cuts

I decided not to worry about ensuring all my pieces fit on one board. As a result, I need two OSB sheets! Keep this in mind for next time: some foresight at the beginning can prevent this superfluous expense. The layout had to be re-designed for two sheets, and the chamfers had to be taken away: I over-chamfered! The machine needs special milling bits to deal with these, and I think it needs some different g-code, too. No time to deal with this level of detail before I even have a first milled piece. Onward!

20 Mar 16. Sunday.

Frustjustments.
I know the machine is going to have trouble with the acute angles, and the design will not be improved by a splatter of dogbones. Damn it. I want to mill in the morning. So, a little more design adjustment.

22 Mar 16. Tuesday.

/// PartWorks

THE SHELF CONTINUES. From SolidWorks, export each part as a .dxf to PartWorks and lay out the pieces on the correct size sheet. Join all the vectors. In order to make pockets, the shape of the pocket needs to be a closed shape, so in the .dxf export from SolidWorks, re-draw the lines to make sure each cutting zone is a totally enclosed shape. Verify this by selecting and seeing what shapes appear as hot pink units.

See the pink outline defining a pocket on the right side of the image:

Verify with zooming to the utter maximum that the lines that appear to line up actually line up.

From a distance it seems to be aligned ... but it is not.

When everything is lined up and has its own clean outline, you can add your fillets from the menu on the left.

Note this problem I had ~ there was a profile toolpath error that stymied me: PartWorks insisted there was an open path. Couldn't see it. Ferdi set "join vectors" to a tolerance of 100mm and, whatever it was, the program corrrected the problem. Then the profile toolpath could be brought to completion.

I expect the milling to take a few hours.

SETTING UP THE FILE TO MILL.

Specify the sheet size and the tool as you see in the following screenshots.

Here are the two boards laid out ~ the one on the right fits onto an offcut from someone else's project that I found:

Send each toolpath (drill holes, pockets, and through-cuts) as separate .dxf files.

toolpaths .sbp

23 Mar 16. Wednesday.

/// milling with the ShopBot

Follow these steps:

Vacuum. Place board. Collet for assembly. Zero "Z." Second hit is an error ~ it makes Santi shriek! Zero "X" and "Y" via the console. Check spindle speed reading. Fuses on the left should be up.
load part -> start sequence -> mill
Wrapping up, the underlay is more pocked, but clean again.

24 Mar 16. Thursday.

/// finishing, fitting, analysis

ASSEMBLING AND SANDING THE SHELVES.

sanding tools
noticing jags and chips due to the low density of OSB

first shelf!

ANALYSIS: The joints without lateral connections tend to splay. And for some reason I totally misjudged the cutout in the shelves needed to nestle in the back piece. I also moved a hole for the bottom shelf and forgot to adjust its corresponding slot! I'd say I skipped the final edit!!

bad joints
fair-to-middling joints

Some problems with the fit because of not tracking 3-D changes properly. Back to the drawing board for two of the pieces.

Also, I mysteriously shortened the whole thing! I will look and see which parts to alter and try to remill. I'd like to do a few more laser-cutter tests. Let's see if this gets done, as well as documenting these last two weeks, before we're at it again on Wednesday. We're technically on break now, but it's hard to pause when there's so much to do ...

The shelf is populated at last.

/// intimacy with objects

The irrational numbers that log the hours have led to some new and more invigorated thoughts on the Final Project. I would like to keep working with furniture ~ there is a lot of programming for behavior modification that would satisfy me personally re: the furniture I use. I would be happy to be in interaction with the objects that support and cradle my body for so many hours of the day; we should have a more humane, affective connection. So here's a little bit more about that here.

Theories, Ideas, Other

Neil Gershenfeld’s TED talk

I looked at Neil’s TED talk about what FabLabs mean and was inspired (whew, thank goodness). Here are some of the thoughts that tickled me:

Jargon-free good sense, powerful possibilities. Keep my teeth in the process; it’s genuine doer-based learning.

Simple Beauty (is rarely simple)

João was drawing on a Wacom beside me today and made me thirst to be spending some time drawing. Unfortunately I had to try to pay strict attention: he is already fluent with the CAM software we are learning; I know very little! After padding in a circle and shouting inside my own head for nine hours, when I came home I let myself just make something beautiful, something that included some drawing, even if it was on the computer.

I started on a possible crest or mascot: Jacob fighting the Angel. It’s from a Delacroix painting in Saint-Sulpice, where the human is battling with fierce anger and humiliation, and the angel is a wall, radiating a steady counterforce.

Another element of therapy: I’ve been in conversations with three people here who believe that blonde women are the perfect kind of woman. I thought coming to Barcelona instead of staying in New England, I could have a breather from pro-fair-skin chauvinism. In response, I’ve made the nose of the angel look more Eastern Mediterranean.

I thought this would be a quick project after I got the file ready! Troubleshooting. I didn’t consider time for that. There was trouble with the Roland vinyl cutter. We found out that the CUPS printer interface needed to be installed so the Raspberry Pi could talk to the vinyl cutter. While doing this with Ferdi, we authored a incipient tutorial on the process.

/// Tutorial on setting up Raspbery Pi ///

IAAC lecture: Urban Futures with the mayor of Somerville, MA

Joseph A. Curtatone (Somerville) and Christian Gärtner (Audi Urban Futures)

INTRO. Curtatone partnered w Harvard-Kennedy business school to think of new leadership strategies for Somerville, which not too long ago was called Slummerville, but currently has an average monthly apartment rent on par with New York City. It’s true ~ I know it as one of the most sought-after neighborhoods in Boston, reputed to have excellent walkability.

IDEAS

--> This talk was so jargon-filled with politician-speak that I didn’t stay. This American-politico-lingo of inflated optimism - salesmanship - and quoting old research - made me realize I care about finding out what I don’t already know, not seeing details about how what I already support works. Too much hubris, J: no need for proof if I’ve already figured it out.

IAAC lecture: Primavera de Filippi

Subject: blockchain.

Lecture was very similar to recitation we had with her in Week Two, on 08 Feb.

Handsome French-y inflection with some neat vocabulary that nevertheless all sounds a bit sadistic:

recitation: Silvia Lindtner on the culture of making in Shenzhen

Silvia's talking about the culture of making in China relates to Western ideas of a manufacturing renaissance. It is not a renaissance there; it is an ongoing phenomenon that is quite healthy.

Silvia has mentioned "cool and shiny" Western models, Silicon Valley ... Reinier de Graaf here at IAAC was talking about the credibility of charisma for architects ... what about the relation between charisma - design - looking cool - having credibility? Is this a notion that hinges on a Western model of the individual? Surely not ... however, I don't have insight into how "coolness" works in China and other non-Western cultures.