Grid Beam Fabrication

[Eric Hunting]

Looking for advice for Grid Beam tools, specifically tools for easy centering of drilling on 2x2s and fool-proof precision series drilling. 

I've long been wanting to work with Box Beam/Grid Beam and have been an advocate of it for many years. But I've also always been short on workspace and tools while factory production of the holed beams has long been sporadic. While easy to make with simple woodworking tools, Grid Beam pieces are hard to make well. Simple home-made hole spacing jigs made of wood do not last long and are cumbersome and slow to use when drilling so many holes. And small spacing errors rapidly accumulate along the length of beams, making longer structural elements very difficult to produce. This has long been one of the chief challenges for people who want to experiment with this. So I've long been searching for a means to make this more fool-proof and quick without needing lots of workspace. 

When mass produced, Grid Beam has been made with very large expensive multi-spindle drill presses. Drilling 8 holes at a time makes for a much smaller error potential than drilling one hole at a time. One thing that seems to point in the right direction is the traditional peg-based hole drilling guide or jig clamped to the back fence of a conventional drill press. But these seem to usually be home made, positioned by sight, and attached by crude clamps that will slip out of place after a few holes drilled or if accidentally bumped. One interesting--albeit a bit expensive--tool I recently learned of is an auto-centering jig called the DJ-1 from Bridge City Tools. ( http://www.bridgecitytools.com/blog/category/dj-1-drilling-jig/ ) This is a remarkable device that allows for high precision drilling in odd shapes even with a simple hand drill. But it doesn't really help in precise-spaced series drilling. The company offers add-on flip stops that clamp onto the DJ-1, but these only help for making a few precision holes at the ends of a stock piece. 

A more radical thought has been to make a wooden T-slot based on the observation that the routing of T-slots all along the faces of a square stock is generally a much easier process than drilling many uniformly spaced holes. Grid Beam is, generally, akin to a 'poor man's' T-slot building system, given that the industry of aluminum T-slot extrusions still tends toward high prices based on the 'executive premium'. (the mark-up applied to non-commodity goods sold B2B based on the assumption that one is selling to corporate executives and not thoughtful patient consumers who will actually bother to shop and compare) A wooden T-slot profile would seem to offer many of the advantages of aluminum T-slot but would also have some critical drawbacks. The concept has been tried before, but only with large lumber dimensions because, unlike aluminum, wood does not have uniform tensile strength. It's strength in parallel to the wood grain is high, but weak perpendicular to it. T-slot attachment works by clamping along the relatively thin strips of material on either side of the slot and relying on the strength of the likewise thin connection to the body of the profile--all in tension perpendicular to the length of the profile. That works fine for aluminum with its uniform tensile strength but it's along what is usually the weak axis of milled lumber. So a wooden T-slot profile of the same size as a typical Grid Beam piece would be comparatively very weak and easily torn apart by hand. To approach the same connection strength, you would need a much larger piece of lumber. That's the basic limitation of the material and the use of a clamped instead of through-hole-bolted connection. It's always something... So Grid Beam as designed would still be a more practical option, if the ease of the DIY hole drilling could be improved. Grid Beam and the kinds of structures typical for it have a lot of potential relief applications, but if you can't easily make the beams everywhere it hampers that potential. 

I'd like to hear what suggestions people might have. 

[John Griessen]

On 12/24/2012 10:21 AM, Eric Hunting wrote:
> Grid Beam has been made with very large expensive multi-spindle drill presses.

.
.
.

> auto-centering jig called the DJ-1"

Since you are wanting to make many holes in standard sizes of wood,
(I think, not knowing grid beam details myself), the universal adjusting
features of the DJ tools are not needed, but close accuracy on long spaced
holes is.

I'm joining a co-op workshop of sorts in Austin TX -- moving in some equipment and
paying rent soon -- where they have a shopbot and a home built CNC router
under construction. The homebuilt router possibilities are getting
lower priced so much that if you don't count time spent on learning weird
software and setting up and testing machine tools for accuracy and squareness,
the price is approaching just 4) DJ2 metric kits, or $2000.
http://www.bridgecitytools.com/default/tools/drilling/drilling-jigs/dj-2-drilling-jig-metric-kit.html

It's easily possible to combine two machine tools running under the same control
software and have a hole drilling robot (looks like a CNC router with big
table for 4 x 8 sheets of plywood) and a chop saw robot attached to one end,
and a semi-auto feed mechanism with foot switch to use while standing at one end
and unstacking/feeding in and out and re-stacking finished work.

A recent open hardware extruded aluminum beam project used an automated chop
saw to get feasibility of delivering kits of precut T-slot beams.

The feeder part could be a conveyor with end stop detectors and robot clamps to hold
blanks in place before they are drilled.

I'm not sure the best way to have a machine know a piece of
wood is all the way against the stop on the router table,
but I'm sure it can be done several ways.

EMC2 is one kind of software based on PCs, and Smoothie is up and coming based on
microcontroller doing the tool-path following and a laptop is usually
used to load data, trigger events and such. EMC2 probably can do what they refer to as
6 axes, but may be able to deal with on/off controls for clamps and such.

Smoothie is for people who want to do a little programming and make something responsive
and full featured, and probably tailored more to the cutters, clamps, sensors, safety
interlocks you have on hand than using "axis control submodules" like is usually done
with EMC2, and even make specific user interfaces.
But it will need python programming.

John

[Eric Hunting]

Thanks for that great reply. Going the rout of a full CNC setup may be too much for the space I have available, but it definitely points toward some of the automation options. 

[John Griessen]

On 12/26/2012 11:14 AM, Eric Hunting wrote:
> Going the rout of a full CNC setup may be too much for the space I have available

Full. What's full? Why not fully reduced to just what you want,
no X-Y-Z motion more than needed.

Say Z is vertical wrt gravity. And X is across the width of the beams, and
Y is along the length. You could make a table that is 8 or 12 or 16 or 20 feet long
with a Y axis motion system for going along 80% or 90% of that length, just 4 inches in X, and
5 inches in Z for drilling/milling holes. And maybe it is not even permanently
aligned/zeroed in X direction and just checks where the fence is each time and
moves wrt that zero datum in X. Then it could be taken apart, moved around the country
and shared. X axis would need a really long cogged belt, any kind of
frame material, maybe wood...why not? Most of the lowest cost and quite adequate
motor control wants fairly low DC volts like 48V, so you could aim this at off grid
power, which might align with who wants it?

[Eric Hunting]

I see your point. This could be much more compact if two of the axis are reduced just to the small cross-dimentions of stock lumber. But this might be a bit beyond my own skills to build. Are there any more plug-and-play motion control platforms that might suit this? 

[John Griessen]

On 12/29/2012 01:48 PM, Eric Hunting wrote:
> plug-and-play motion control platforms that might suit this?

The controls don't care about dimensions.

But the physical being tailored to your app is never going to be available
as plug/play/off-the-shelf since it is a teeny market.

You'll have to create it.

[CubeSpawn]

And, you may wish to consider that in the example above "Y" is a set of synchronized powered rollers and some photo resistors or photo diodes to sense end positions and now your automatic machine is in the $300 dollar range...

a 100 dollar drill press, some carriage bolts, some 2x6's and a couple springs... yer in bidness ;-)

[John Griessen]

On 01/04/2013 11:45 PM, CubeSpawn wrote:
> And, you may wish to consider that in the example above "Y" is a set of synchronized powered rollers and some photo
> resistors or photo diodes to sense end positions and now your automatic machine is in the $300 dollar range...
> a 100 dollar drill press, some carriage bolts, some 2x6's and a couple springs... yer in bidness ;-)

Hmmm.... mini drill presses from Harbor Freight were $49 each once...
You can use four...
The end positioning could be made very accurate with feedback control
and be the only thing automatic and still be a real speed up.
steps would be: load a board in from a stack above the line, press the align
to first stop button, then turn to face each drill press to bore a hole,
then advance to next stop, bore again, repeat, eject, reload.