README.md 4.44 KB
Newer Older
Jake Read's avatar
Jake Read committed
1 2 3 4
# Roller Milling

Spent some time working this out on the whiteboard today, and threw together some rhino sketches.

Jake Read's avatar
Jake Read committed
5 6
![mill](/images/rct-machine-00.jpg)

Jake Read's avatar
Jake Read committed
7 8
![wb](/images/wb-01.jpg)

Jake Read's avatar
Jake Read committed
9
I'm going for simple, strong-enough, and easy to build. Using rollers and pulleys, or, just as easily, rack and pinion or leadscrew-type actuation.
Jake Read's avatar
Jake Read committed
10 11 12

Next up, getting into fusion to get through the details, then fabbing one axis... designing a simple 3axis machine, and fabbing that.

Jake Read's avatar
Jake Read committed
13 14
I pulled this off with a simple set of flat-milled parts - I was trying to work with the constraints present in the set of FREP and Milling tools in MODS, to see if we could finish end-to-end gantry generation and milling.

Jake Read's avatar
Jake Read committed
15
The result is OK, it'll certainly do - there's a line to toe between being kinematically wicked and loosing complexity. This is a bit of a midpoint - all of the axis of constraint have 'tuneability' in that their preloads can be adjusted, but all are over-constrained. There's enough (lots) flex in HDPE that this will be OK. There's a version which (not actually existing yet) is dead simple and totally constrained, and a version which is *perfect* motion-wise.
Jake Read's avatar
Jake Read committed
16 17 18

Overall, it's a bit bulky. There's some detailing left out before it is complete. But I think in terms of an all-arounder axis, this is pretty OK, and certainly OK for being millable from one sheet of HDPE (or alu, whatever).

Neil Gershenfeld's avatar
Neil Gershenfeld committed
19
![one-mill](/images/rob_rct_one.jpg)
Jake Read's avatar
Jake Read committed
20

Jake Read's avatar
Jake Read committed
21 22
## CAD

Jake Read's avatar
Jake Read committed
23
Available Here: https://a360.co/2qQf6wf
Jake Read's avatar
Jake Read committed
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

password is *rollercoaster*

The gantry is built with a set of 'kinematic elements' - a Motor / Pulley Unit, Lateral (into the gantry) and Cross (across the gantry) supports:

![bits](/images/little-bits.png)

These can be jimmied around to make a full set of constraints:

![more bits](/images/multiple-bits.png)

And combined into a complete gantry:

![comb](/images/combined-bits.png)

Here's the back of that:

![allup](/images/backside.png)

Jake Read's avatar
Jake Read committed
43 44 45 46 47 48 49 50 51 52
## Kinematic Bits

Currently the set assumes 3/8" HDPE, 8x22x7mm bearings, NEMA23 Motors and 9mm Wide GT2 Belts

Version for 1/4" HDPE, 5x11x4mm bearings, NEMA17 Motors and 6mm Wide GT2 Belts.

See [cad/assembly/elements/](/cad/assembly/elements) for kinematic elements, and [cad/assembly](/cad/assembly) for one-machine made with these elements.

Motor and XY Constraint:

Jake Read's avatar
typos  
Jake Read committed
53
![motor](/images/kunit-motor.png)
Jake Read's avatar
Jake Read committed
54 55 56

XY Constraints, with preload adjusting and without:

Jake Read's avatar
typos  
Jake Read committed
57
![xy-adjust](/images/kunit-xy-adjust.png)
Jake Read's avatar
Jake Read committed
58

Jake Read's avatar
typos  
Jake Read committed
59
![xy](/images/kunit-xy.png)
Jake Read's avatar
Jake Read committed
60 61 62

YZ Constraints ('out of plane'):

Jake Read's avatar
typos  
Jake Read committed
63
![yz](/images/kunit-zy.png)
Jake Read's avatar
Jake Read committed
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

**Assembling in Rhino**

![basically](/images/kunits-basic.png)

or

![units xz](/images/kunits-before-form.png)

filling in

![one](/images/kunits-one.png)

![two](/images/kunits-two.png)

![three](/images/kunits-three.png)

![four](/images/kunits-four.png)

Jake Read's avatar
Jake Read committed
83 84
## Hardware Notes

Jake Read's avatar
Jake Read committed
85 86 87 88 89
- needs more clearance on belt-side
- belt tensioning?
 - when rounding corner, teeth must at some point disengage b/c different lengths
 - same size M3 tensioning

Jake Read's avatar
Jake Read committed
90 91
Motor Idlers / Topside Way Bearings mount on M5 SHCS with washers on *both sides* and washers used as bearing shims. Use 3DP bushing (or purchased?) in axel. Important for carrying some moment to plate.

Jake Read's avatar
Jake Read committed
92 93 94 95 96 97
Clearances here for pulley to be mounted / tensioned! Consider order of assembly, tensioning.

M5 Nuts 93625A200
M5 SHCS (length not selected) 91292A128
M5 Washers 93475A240

98 99
also uses long M3's for tuning, nuts, washers. Length to be selected.

Jake Read's avatar
Jake Read committed
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
### Egregious Errors
 - bottom rail extremely unhappy, should be smooth and sideways ways should be clear

### Minor Errors
 - missed a few holes in CAM, double check! 
 - y block internal boxing has co-dependent assy 
 - bearing bushings not satisfactory
 - on double rollers, accomodate washer in between 
 - motor collar coule be +0.05mm in diameter 
 - need long M5 for motor -> bearing 
 - need bigger wall on captive nut for tuning 
 - need channel in side of motor collet for screw turning 
 - XZ situation has lots of unaddressed assembly issues 
 - add clamping features, hole access for final beam -> gantry assembly
 - have to watch where bolt heads, bolts are - should include those in kinematic units modules 
 - y gantries should include two rollers in xy space (machine space) constraints 

117 118 119 120
## Some Simulation

It occured to me in class today (Rob Hart fabricated the below component!) that the two-bar flexure I had intended was acting much more like a one-link flexure. Whoops, all well. Still does its job. Here's an image of that in any case:

Jake Read's avatar
Jake Read committed
121
![onebar](/images/preload-sim.png)
Jake Read's avatar
Jake Read committed
122 123 124 125

## ooooh

![milling](/images/milling.jpg)