CNC Production

[Alastair]

Cutting, materials, efficiency, disruption, costs, tricks and design development.

Off the back of this important challenge from Adam in the intro thread, I thought I'd start a thread on on CNC production and its application to housing for all the experts, testers, makers, and put in some thoughts on those issues / questions under discussion.

"Got very excited about this tech and ran it past my best mate who owns a pretty sophisticated CNC. He raised several points:

Plywood is really hard to get a cnc to 'grab', meaning the suction required to hold down the board, making the job difficult.

To cut out one of the example rooms on here would take I think he said, 19 sheets of material, almost 1000kgs, much of which would be waste.

The COST of this compared to a simple-as-pie stick frame is about $1700 compared to about $400. Plus the extra cost of engineering and getting something unusual through council. So of course the question is, how does this tech democratise design in a way that literally anyone can take up?"

1. Efficiency

Based on what we've been doing he's about right, it takes roughly about 5-6 sheets per square metre of internal floor area which if plywood is $30 a sheet (?) works out at least $180 per square metre material plus cost of CNC machine / cutting. Designing down waste, making it easier to cut more, making the structure more material efficient is a big challenge, has a lot to do with nesting of parts on sheets etc.  It might also be we could combine generic lengths of materials with CNC joints to get the best of both worlds. 

2. Materials

It also has a lot to do with what materials. We started with plywood because..it's standardised and widely available at various different prices, but it still has a cost associated with it (£ and Carbon) and isnt that recyclable. So very interesting to look into material alternatives (eg Recycled plastics in sheet form such as ecosheet) which allow a closed-loop material ecology to become a practical possibility, and potentially to radically lower material costs. 

3. Disruption

First, important to point out that the aim of WikiHouse is to democratise design, but it hasn't done it yet! WikiHouse in the end might have nothign to do with CNC as process (mud, straw are still the most open materials around!), but we (London crew!) started working on it because we saw the seed of a disruption we could cause right now. We do have to be completely un-rose tinted about what the disruptions are. You're right, compared with simple stick and frame construction WikiHouse is still more expensive in materials, but what is does do is dramatically lower the thresholds of construction time and skill (which is worth a lot of money if it enables you to invite all your friends over and build your house for you instead of having to do it yourself or pay someone to do it) and also because it can carry embedded complexity in the system, makes it much easier to raise the bar on quality, energy performance, safety etc compared to traditional construction, and to some extent make that quality reliable and replicable. Even if (and ultimately it shouldnt) the material cost is higher, the overall project cost (compared with buying a property) is much lower. 

4.Forms 

Quite a few people have mentioned geodesic domes, round houses etc. We havent done one yet... but would be a fascinating project! Some of the projects in the Construction systems thread are sort of working around it. There are mainly three reasons we havent yet: 1. Obviously CNC cuts at 90 degrees! So if slotting together houses in a simple, modular way, it seemed like a reasonable start point! 2. One of the aims for us is not developing a single open source design as such, but an open source system that can deal with multiple sites, slopes, climates, profiles, tiny little snatched sites etc. Circular houses tend to be quite hard to adapt to difficult / oddly shaped sites and contexts. The linear frames system can change over its length to more or less any unique site profile. 3. One of the problems modernist architecture, and housing prototypes by architects have struggled with in the past is that the thing at the end doesnt look like what people conventionally recognise as a "House" where they live, it looks alien. So there's actually a lot to be said (both in terms of planners and users) in a certain degree of conservatism in form / being stylistically agnostic! 4. A bit like the above, modularity is also about making something that can be used with existing systems, products etc, is very hackable - so seems valuable to not reinvent the wheel form wise. Although geodesic domes are structurally optimised in a 3D sense, they are not necessarily optimised in a 4D sense (integrating time, people, messy reality!). That might all be complete rubbish... but.. those are our thoughts so far...

5.Cutting 

Can anyone comment further on key issues / design challenges to be tackled around CNC cutting? Suction, tolerances, offsets. One thing we know we need is a better calibration test piece (possibly something thats useful afterwards) that helps people work out what offsets to use with the material and machine theyre using. Theres a cool quote from Elvis costello about how he wrote all his songs to be played on the most lo-fi transistor radio. If they worked on that, they'd work on anything...

[Adam Baudelaire]

Hi guys, Adam here. Thanks SO much for the positive and practical response. the Elvis Costello quote is actually a methodology across the recording industry ( old sound engineer here ).

I was involved with a guy in the Nederlands that produced these:

http://www.spacebox.nl

We discussed at length and over many years the way one might produce very low cost housing. I think it would be interesting to hear his thoughts on this project.

ALso, I think I'd like to see William Macdonough - Cradle to cradle - cast his brilliant eye over it. Like you started to mention cyclic reuse of materials is crucial to efficiency. It needs to be like a cookie cutter where the offcuts can be mushed back into a flat board for cutting again. Materials science is advancing rapidly. Cheap, endlessly recyclable materials are coming.

Basically. Biomimicry and architecture need to be more common bedfellows.

Adam Baudelaire

www.1620program.tumblr.com

1620p...@mail.com

0401 158250

[nick]

nested sheet. This is an example of a composite wood sheet good available to us here in the United States.  We hold the sheet down with a vacuum plenum that is powered by a regular Shopvac, sucking through a sacrificial surface.  We have cut hundreds of sheets this way.  We also have the ability to have a closed loop recycling system for the cutoffs with the product manufacturer.  You should really take the time to go through the other pages and become more familiar with the history of some of these systems.  Although they are not quite ready to pass the ICC yet. they are well on their way. 

a Segway into residential home building- a brief article on the methodologies of our community that are taking the first steps into the mainstream "Stick built" arena.  Stuff like this will help us all down the road.  A 10 minute video of their procedure if you have the time and desire to see.

Nicholas

[nick]

Some additional thoughts:

I realized that I do not have any designs posted currently in the wikihouse library but I do try and focus on segmenting the parts so that there is a higher yield from a typical 4x8 sheet.

I always leave a one inch border around the perimeter and prefer a 3/8" diameter cutter and usually nest parts with a 1/2" between them.

[Gregory Witkamp]

That sounds to me like way too much space between parts. I work for a furniture company that make upholstered seating in a very similar fashion to how Wikihouse goes together. I use a 3/8 2 flute mortise compression cutter that we run at 18000RPM and 800ipm. For a 3/8 cutter, we leave a .4" space between parts.  This allows you to get a tighter nest and extends tool life since the tool isn't cutting as much material. 

[nick]

Gregory,

Your reply is very telling of your expertise in CNC cutting and part production.  You are right in that a higher tolerance can be achieved.  I must add that Wikihouse is still in its early stages. People are actually able to observe the research and development as it occurs and make helpful contributions.  Personally, I do not have a 15 or 20 HP vacuum hold down system and consequently, It limits me and my part nesting.  Knowing that , especially, with small parts, I will have to tab them or screw them down to keep them from moving during the machining process.  God only knows how many times Something has dislodged and gone flying across the shop!  Furthermore, I am not sure what kind of part hold down will be realistic for folks cutting these parts out without a vacuum system.  Tabbing is what I would assume they will use at first. 

So "baby steps" have to be taken but it will get there withthe help of people like yourself.  Thank you  and I look forward to chatting more with you.

Take care,

Nicholas

[Patrick Beseda]

Hey everyone,

I had a few quick questions as we start to seek out CNC service for our build project.

What is a time estimate for a full sheet of cutting? (I know they are all different, does anyone have an average for a pretty full sheet?)

What kind of specifications am I looking for in a machine to get the cuts right? I'm new to CNC cutting so I'm not familiar with feeds and speeds, vacuum hold down, cutter types etc.

I know it can vary greatly but what can I expect to pay in the US to do around 30 sheets?

Any other info/advice for a novice CNC'er?

[nick]

Patrick,

I looked at the sheet layout on your website.  I would expect to pay $35 - $60 +/- per sheet + material/tooling cost.  Different shops carry different overhead and custom work can carry infinite complexity. 

When bringing work to a shop, there will be some setup required to perform the task. Tool-path generating, selecting the bits and material , any required lead times etc. 

I do not know if you have been in contact with a local shop yet but is is my suggestion on where you could start 100k garages  These are fabbers in the Denver area that were listed on the network that I belong to.

I hope this helps and let me know how it goes.

take care,

Nicholas

[Gregory Witkamp]

Patrick,

To give you a feel for run times you should expect I'll give you some examples from my experience.

For a full sheet of fair small parts (like for upholstered furniture) on an older machine with a max feed rate of 400ipm the run cycle would be around 12 minutes. The run cycle includes loading the material running the machine, and unloading the parts.

On a newer machine that can run at 1200 ipm cutting larger parts such as those for Wikihouse, I would expect it to not take more than 5 minutes per sheet.

As for CNC services, I have always heard $100 per hour. I would check with some custom millwork/cabinets shops to see if they have a CNC and what they would charge.

Hope this helps,

Greg

--
Greg Witkamp
Phone: 765-491-8539

[l...@autobusesaga.com]

Yo uso el método de Patrick Hood Daniel para el corte CNC 

El sistema de fijación es simple, pues solo requiere tornillos que fijan la lámina. 

Aqui incluyo links para dos videos de trabajo con estas máquinas CNC, en estos videos Patrick no usa el plugin de wikihouse, yo si lo uso habitualmente para mis trabajos. 

Espero que sea de utilidad y os de una idea de como se puede trabajar con plywood. Algo que me sorprendió es que en ellos se hacen dos pasadas, una de desbaste y otra de acabado, esto hace tardar mas el corte pero las piezas quedan muy precisas y prolijas. 

http://www.instructables.com/id/From-SketchUp-to-CNC-Fabrication/

I use the method of Patrick Hood Daniel for cutting CNC

The locking system is simple, it only requires screws securing the blade.

Here two videos include links to work with these CNC machines in these videos Patrick does not use WikiHouse plugin, I use it regularly for my work.

I hope you find it useful and an idea of how you can work with plywood. Something that surprised me is that they make two passes, one for roughing and one more for finishing, this make bigger the dealy, but the parts are very precise and neat.

http://www.instructables.com/id/From-SketchUp-to-CNC-Fabrication/

[Neil Berrow]

A great Kickstarter project for a portable CNC miller, check it out:

http://www.kickstarter.com/projects/1320575205/handibottm-a-smart-digital-power-tool

[Brad Jensen]

Just a few thoughts - 

Suction - You don't have to use suction to hold down the sheet.  You can also use brass screws, double sided tape, or even bondo.  I often use a ShopBot buddy that doesn't have a suction system at all.

Nesting - There are commercial programs out there that do this quite well.  I use VCarve.   I have attached an example for 3.1 that is not perfect but it gives you the idea.  

Environment - I think the most environmentally conscious material for now is used or re- manufactured plywood.  I am just starting to look into this though.  I will have to invest in a metal detector to make sure I have gotten all the nails out if I try to go with used plywood.  

[nick]

Or aluminum screws like brass. They will save your bit in case of an accidental swipe. They do tend to strip out quicker or even break off sometimes but that has happened with hardened steel screws as well.

[Eric Hunting]

On the subject of materials, some I've been wanting to experiment with for sake of their reduced latent toxicity and sustainability.

http://www.kireiusa.com/wheatboard_viewer/wheatboard.html

http://en.wikipedia.org/wiki/Pressboard

http://www.corkstore.com/Products/Wall-and-Ceiling-Coverings

http://www.centriforce.com/en/products/sheet-and-board/

http://www.bondedlogic.com/

http://www.ecobuiltsystems.com/eco-materials/mineral-foam/

http://www.airkrete.com/

http://www.buyaerogel.com/thermalwrap/

http://www.jetsongreen.com/2010/11/enguard-recycled-pet-batt-insulation.html

[Vincent Muller]

"ECOBoard Bio based Panels are made from agricultural residue"

Anyone familiar with eco-boards? [More less the same as kirei board mentioned by Eric in previous message?]

http://www.eco-boards.eu/sample-page/ecoboard-standard/

Is it possible to use for a WikiHouse?

[Eric Hunting]

This looks like a great material. Similar to weatboard in origin, but with an apparently higher density. If the descriptions are accurate, this could be a good 'green' plywood alternative with about the same performance. I guess the key question is how the usual CNC router head deals with the material. Thanks for noting this. A good addition to my materials reference catalog. 

[nick]

Eco-board reminded me of another sheet from Hemp technologies called Hempboardhttp://hemp-technologies.com/page15/page19/page19.html

[Vincent Muller]

Indeed promising.
The CNC router head needs less replacement than the usual plywood I heard so far, but have to verify this... 

Besides this and a bit alarming, I heard some rumours that a lot of the plant residue might be gained in China. Not sure about this, so will check first. I did send an email. [phonecall not answered]

From the website:

High Strength, Lighter Weight

The excellent strength to weight ratio of the ECOBoard panels eases handling on the construction site. ECOBoards were benchmarked against wood-based particleboard in terms of strength and specification and can be cross referenced to any international particleboard standard such as BSEN 312 P4 to P1, ANSI A208.1, NEN, etc. ECOBoards can replace plywood, chipboard, OSB, multiplex and conventional fibreboards (including MDF, hard-board and soft-board) for a wide range of applications.

Screw Tightness

ECOBoards have a tenacious screw holding ability that adds to the choice of fitments that can be used.

Machineable

ECOBoards are made in a single layer process giving homogenous consistency throughout its thickness, similar to MDF. Its superior internal bond (cross tensile) allows smooth strong profiling and sharp machined edges



Op woensdag 6 november 2013 17:05:08 UTC+1 schreef Eric Hunting:

[Vincent Muller]

Talked to someone from Eco-boards [Delft, Netherlands].
It's correct, they produce the boards in China at the moment, but they expect to produce it local [in the Netherlands] end of next year.
It has to do with certain certifications...
Still very interesting to use this material for WikiHouse, I think.

Op donderdag 7 november 2013 15:10:29 UTC+1 schreef Vincent Muller:

[PEEL livingprojects]

Hi all!

I just arrived and am very glad to join such na enthusiastic and creative atmosphere.

My name is Paulo Costa, architect, and live in Porto, Portugal.

I'm looking forward to contribute with my view and be able to promote the Wikihouse project around here. Actually I'm very much interested about building a prototype adapting a wiki design, in a short/mid-term plan.

But first things first. So far I didn't manage to make a proper DXF file with cutting sheets..!

I did install the Plugin in Sketchup, Set the Camera to Parallel and did extract 2 files: na empty DXF and a valid SVG, that I can VIEW in a browser or "open" in a program such as Inkscape (but nothing there to see..)

So you see, a rookie beating around the bushes here.

Probably very simple, or so I hope.

Needless to say, HELP needed here, thanx.

All for now, best regards to @ll.

[Eric Dunn]

Hello all,

Just joined, and after reading had this question:

For shipping and storage, wouldn't it be best to leave all the parts in 4x8 form?
All the parts would remain in the full sheet form attached by tabs. To remove the parts, a simple hand saw could be used.

Doesn't that also solve issues with holding the workpiece in place on the CNC table?

[nick]

Eric,

Generally speaking, that is a good setup. Handsaws, chisels, and palm routers are all ideal to remove parts from the skeleton sheets before they are recycled.

There are some instances where removing the parts for transport does make sense. I actually have parts for a single room structure 8x12x10 single pitch roof that totaled about 80 full sheets. It is condensed and stored on a single axle trailor that is 7x12. Minus the extra volume and weight of the unused sheet goods.

[*-*]

Nick,

If the reduction in shipping size is that great, then I really need to look into projected costs to validate the logic and suitability of the wikihouse approach.

Seems like an awful lot of waste material.

[nick]

I agree with your original point of view Eric.  Without having produced Wikihouses on a large scale and shipped them over great distances of course.  I was just offering a single example from my own personal experience in addition.  I should also add that my project has been cut out, assembled, broken down, packed up and moved by myself a few times. In this specific instance it works for a single individual and predetermined size trailer for transport. Waste material was averaging 10%+/- per sheet.  I feel that these are some of the real challenges we face as Wikihouse progresses beyond a concept and becomes a reality.  How does it scale? Is it high production or is it individual construction or both?  Personally, I am interested in the individual empowerment.  As I already have the CNC and tooling I understand that it may be a bit easier for someone like myself.  There is a gap that is closing between design and fabrication here that is almost seamless but not quite yet. 

[Patrick Beseda]

Just throwing in my $0.02 but wouldn't the idea be to transport the material as small of distance as possible (or not at all?). The barrier here, as Nick said, is the availability of of machines and knowledge to run it.

Having moved a small wikihouse ~700 miles, on a 14' trailer, it was not worth the transportation cost. We would have saved time, money, and energy if we had been connected to a shop that could provide the machining more locally. Making the connection between individuals and shops with capability is very lacking (especially outside urban areas).

--

Patrick Beseda

303.518.6920

patrick...@gmail.com

[nicholas godfrey]

That is what the solar powered, cordless, portable CNC router is for! ;-)