New Kinematics Options, PlyPAD Progress, and Software Updates: The Aug 30th Update

Hey Everyone!

 

Since this is our first every other week update instead of every week update we’ve got a lot of ground to cover and it was getting pretty text heavy so I made a video version of the update in addition to the text. 

 

The first thing I want to talk about this week is something we’ve alluded too in the last few updates which is the idea of moving the the point where the chains meet on the sled to the center of the router bit.

Locating the intersection of the chains at the router bit has a couple of big advantages. It significantly simplifies the math because now the system makes a triangle where the lengths of all three sides are known. This also makes the calibration process much easier because there is just one variable which matters, the distance between the motors and we already have a good system for measuring that automatically.

There are other advantages too like that if the chains are anchored at the router bit, if the sled twists it won’t affect the cut because the sled twists around the router bit.

Of course we can’t actually join the chains in the center of the router bit because the router is there so we have find a mechanical system which has the same properties. This can be achieved by attaching the chains to something which pivots around the center of the router without actually being there, like a large bearing.

Since even before the kickstarter, the community has been coming up with innovative ways to solve this problem. I was initially not enthusiastic about them because I underestimated how difficult mathematics and calibration process for the current system would be. I also had reservations because most of the early solutions relied on either low cost lazy-susan type bearings which I was worried would fill with sawdust and stop working, or large expensive sealed bearings which were out of our budget.

In the update a couple weeks ago I mentioned that I was playing around with a system where the chains anchor to wheeled trolleys which run along a large metal ring such that the chains always point to the center, and the community response was amazing. Here's a {link: http://forums.maslowcnc.com/t/throwing-my-hat-in-the-sled-modification-ring} to the entire conversation, but the gist of it is that the community did research, talked through ideas, sourced parts, and generally made it happen in the most amazing way possible.

My experimental setup with a steel ring and the chains attached to rollers

My experimental setup with a steel ring and the chains attached to rollers

Within a matter of days the community came up with a system of linkages to achieve the same effect at a potentially even lower cost, and that’s what I want to share with you guys this week. To be totally clear, this system of linkages is the community’s idea pioneered mostly by Logan Needham, and tested by Kevin Keegan.

The general principle is that a system of straight linkages can simulate the same motion as a larger bearing at a fraction the cost. They’ve already come up with the idea, built it, and are testing it. Here’s some photos and footage from the testing process:
 

The design as conceived and tested by Logan Needham:

Demo Build 2.jpeg
Demo Built 1.jpeg
A beautiful diagram of possible linkage systems from Logan Needham

A beautiful diagram of possible linkage systems from Logan Needham

As built and tested by Kevin Keegan

As built and tested by Kevin Keegan

A cat cut by Kevin Keegan using his new rig

A cat cut by Kevin Keegan using his new rig

 

Footage from Kevin Keegan’s testing:

 

 

If you would like to get in on the experimentation, this type of system is now fully supported by Ground Control and the firmware, you just need to go into advanced settings and change the kinematics type to “triangular”.

 

PlyPAD:

I also want to talk a bit about PlyPAD this week. We’ve been working on building this house for quite a while now and haven’t really gone into any details. That’s mostly because we’re figuring it out as we go along, and we’ll do a final recap about what worked and what didn’t at the end. Until then, I want to share a couple of the technical details about how we’re building it.

 

Use of templates:

The first idea I want to talk about this week is the idea of using Maslow to cut templates and then using the templates to cut parts for the finished house.

At Maker Faire a couple months back we were fortunate enough to get to talk to the guys from Shelter2.0, and they gave us the idea of the power of using templates. Templates let you take the CNC machine’s ability to produce accurate parts and multiply it using unskilled labor. It’s easy to follow a template with a router compared to measuring and cutting parts by hand, but it’s also cheaper and faster than cutting parts directly on a CNC machine. In a disaster relief situation where there is an abundance of people who want to help, but limited technology like a CNC router, templates can let one machine produce many parts quickly.

You may notice that our design relies heavily on full sheets of plywood which have tabbed edges. Cutting tabs into the very edge of a sheet of plywood is difficult on any type of cnc machine, and impossible with our design, so we’ve cut those using a template and a palm router with a follower bushing. The template is applied to each corner of the sheet, and routed. The thickness of the follower bushing means that the cut part does not exactly match the template, however this is easily corrected for in the CAD design of the template.

How the router bushing follows the template

How the router bushing follows the template

The template design we’ve been using

The template design we’ve been using

 

This system is working really well for us. The one drawback is that the template can amplify errors in the length of the sheet of plywood because it is applied from each end so any sheets which are cut down in length before the template is applied need to be measured and cut carefully. I would recommend designing using this method with full sheets only. Some very slight changes to our design would let us use only full sheets in the future, which we plan to do on version two.

 

Structural Insulated Panels:

The next idea we have been experimenting with is bonding rigid insulation foam between the sheets of plywood to provide structure. In each of the walls floor and roof of our POD is 3-½ inches of foam insulation which both insulates the building to R28 and greatly increases the strength of the plywood which can be somewhat flimsy on its own. This idea is borrowed from the concept of structural insulated panels (SIPs)  and so far we’ve been very impressed with how strong the structures we’re making are. When we were loading the tall module onto the truck for transport to the museum we had to lay it on it’s side and at one point it was supported only by the tip of the roof and a forklift lifting from one corner of the floor. It was never designed to support its weight that way and I was sure it would break in half, but it didn’t even flex.

 

We will give a full breakdown of what worked and what didn’t once we’ve got the whole house built.

PlyPAD module 1 under construction showing where the rigid insulation goes

PlyPAD module 1 under construction showing where the rigid insulation goes

 

New Contextual Menu:

New since the last update we’ve added a contextual menu that comes up when you click and hold anywhere on the gcode canvas. The menu currently has options to command the machine to move to the clicked location, or place a marker there which can be used to mark out and avoid previously cut parts of the plywood sheet. This feature was an excellent community suggestion. If you have other ideas for location specific commands, let us know!

New Contextual Menu.JPG

Have a great two weeks everyone, and don’t forget to hang out in the forums!

 

-Bar and Hannah