CubeFrame Costs

[James]

Just for reference, a "625" (600mm cube out of 25mm extrusion) frame costs about $70 for the extrusion and the hardware - the hangup with this size is:

 the corner connectors don't exist - you can use 1" corners, but they run $17.85 apiece! http://www.ebay.com/itm/8020-T-Slot-Aluminum-Tri-Corner-Connector-10-S-4042-N-/220470841032?pt=LH_DefaultDomain_0&hash=item3355160ec8

-holy shit!! there are 8 required!! ($136 + freight!!! ridiculous!!!#$!)

my friendly machinist quoted me $25 each in small quantities - and this is totally reasonable ($200 per cube - ouuuchie)

my answer to this is to make them from bar-stock, they ARE labor intensive, since every side of the stock must be milled, but they are small and only take about $4 worth of aluminum (counting freight, cutting, milling, drilling and material) 

so my offer is this - send me (or dropship me) a piece of aluminum bar-stock 1.25x1.25x1.35 x #number of cube corners you want made) x 2 

the stock is 1 1/4 x 1 1/4, I cut it into 1 1/4 pieces +.100" kerf so it takes 1.135 inches of length per block I figure 10 cuts for 8 blocks since the barstock sometimes needs a cleanup cut on the end so 10 cuts is 1 (10x.100") inch of the bar becomes chips

plus I need something left over to hold on to so: if you send me double the required stock - I can make more blocks - to get more machines built I think you'll find that a piece of stock is pretty cheap compared to any other solution - and I have a $38 blade in the saw and 3 $46 carbide endmills to make these parts with...

IF we can get a couple dozen CubeMills scattered around - then producing these should speed-up and regionalize

your input please ;-)

James

[Bruce Wattendorf]

I will take you up on that offer it is a great way to start having me send you 2 times the material.. 

how about trying to print the connectors??

I could print you a set and send them down?  

Bruce

[Joshua D. Johnson]

I'd like to get a set of connectors in plastic and cast them in aluminum. Alternatively I'd be willing to send metal as well. would you be machining them with CNC equipment? If not I'd want to compensate a little as well.

One of the great aspects of open source is the sharing of ideas. Maybe if this project became a little more open we could all pitch in a little and help you through. I am unemployed as well so I coould put a little time to a good cause.

JOSH

[John Griessen]

On 11/09/2012 05:21 PM, Joshua D. Johnson wrote:
> I'd like to get a set of connectors in plastic and cast them in aluminum.

I have some casting experience for art purposes and this is a good tack to take.

[John Griessen]

On 11/09/2012 03:32 PM, Bruce Wattendorf wrote:
> how about trying to print the connectors??
>

Out of ABS? It would be better to print guide enclosures to strengthen
and allow mini or micro USB connectors to be surface mount assembled onto
small controller boards. or print castable wax positives out of some good hard wax
and investment cast them in aluminum, then send that step to China if popular.

[Joshua D. Johnson]

We're talking about the corner connectors right? I work with Bosch extrusions a lot, the connectors wouldn't have to be super precise, just adjustable. I would think direct burn out of PLA?? I've seen direct pour into foam, it vaporizes instantly but I'd have to play with ABS and PLA.. You're right if it ever became mass produced other methods would be better.

[John Griessen]

On 11/09/2012 07:34 PM, Joshua D. Johnson wrote:
> I would think direct burn out of PLA??

Not sure. Iwas suggesting the machinable wax as direct burnout for investment casting with
USG no1 or 20 minute plaster and various aggregates possible. You could get good
castings with 1/2 plaster and half Texas dirt.

[Joshua D. Johnson]

I've got other parts prepped for the caster. We usually do investment and reclaim wax here then ship. I'd like to try direct casting into plastic. Our foundry uses fine resin bond but I'd compromise at fine green sand. It depends where holes are..

I guess I don't really understand what the corners are for., usually you can just bolt the pieces after drilling two holes.

Josh

[John Griessen]

On 11/09/2012 08:06 PM, Joshua D. Johnson wrote:
> I guess I don't really understand what the corners are for., usually you can just bolt the pieces after drilling two holes.

I think the corners are for strong holding of the extrusions the define the edges of cubes.
a 3D print of a large thread like 3/8" could work with crummy printing resolution, but some now
are very fine and could print a 1/4-20 thread quite well, and then you get 1/4-20 self tapping screws
and presto! Assembly from castings and extrusions with low hassles and high strength.

[Joshua D. Johnson]

These are designed already!? Maybe I'll take a look and see what I can make. Usually you just drill and tap the end of one extrusion, drill and connect the other two and you have a strong corner., no need for castings or gussets/brackets.

[Bruce Wattendorf]

I am not able to print in ABS because i dont have a heated build platform. I can do PLA

[Joshua D. Johnson]

If you are talking about using it for casting PLA is perfect. Otherwise it is too weak.,

[Data Pathway]

This is the part:

http://cubespawn.com/old-site/600mm.html

only the links for the "625" part are valid

And, yes, I built the first cubes with aluminum angle brackets in the corners - works fine, but there are minor alignment and part rotation issues...

and yes casting would be better, but if everybody built a CubeMill and made 10 sets of these the harder way It'd be a huge geometrically increasing number, there will never be a shortage... ;-)

so there would be distributed mass production, but locally, only a few sets would be needed at any given time...

[James]

'nuther observation/data spew...

The incomplete, but important system elements are many, but the notable ones are:

Pallet Mover - without it how can you get work from one cube to another AUTOMATICALLY? well, you can't

the pallets I'd like to see are 450mmx 20mm thick octogons with a grid of tapped M6x1 holes at 25mm spacing, Pretty fancy, huh? like this tooling plate I had made:

https://plus.google.com/photos/103828779781480193226/albums/5768863819249015441/5768865444639710482?banner=pwa

they also need index holes very precisely placed on thier lower sides for indexing and lockdown 

modular fixturing is needed to mount stops and clamps in a predictable grid to lockdown and edge guide the stock for machining operations similar to this stuff: 

http://www.carrlane.com/catalog/index.cfm/26805071F0B021118070C1C510D020609090C0015480013180B041D1E173C1B08535240

a very limited set of fixturing that can be made reliably should come first

Fun!! no?

[John Griessen]

On 11/09/2012 09:51 PM, Data Pathway wrote:
> This is the part:
> http://cubespawn.com/old-site/600mm.html
> only the links for the "625" part are valid

So, is that hole with dimensions .236 in. and .315 in. a 1/4-20 tapped thread?

I guessed right?

On 11/09/2012 10:44 PM, James wrote:
> a very limited set of fixturing that can be made reliably should come first
>
> Fun!! no?

You bet!

[James]

welll... no.

its actually 6 and 8 mm through holes for the shank and head of the M6x1 socket head cap screw (click on the mm dimension version of the drawing...)

And here is why the drawings need versioning and so on - these were drawn BEFORE I bought the Seig mill and the actual production dimensions are subtly different, and (some of) my tooling is fractional inch, some of it is metric, it doesn't matter on edge milling - since you just change the tool offset, it DOES matter on drilling, since the diameters end up at odd dimensions in mm and with runout are rarely nominal anyhow... ;-)

So laments the poor machinist...

[Data Pathway]

Printed connectors MIGHT work ok, but I'd probably test one (rather savagely - cubes should not recieve rough treatment, but shit happens!!) I know with printed parts that de-laminating along the layers limits thier strength substantially....

I'd be game to try out a set -you should print TWO sets in that case and try out a set yourself too

plus an extra or two to destructivley test (overly tight bolts, strong torsion, strong tension, yadda yadda)

the two-fer offer is an attempt on my part to jump start machines into the hands of others - its the only way to get this started that I know will work... ;-)

James

[Data Pathway]

Yeah,  that'd be good too, I made up corners, because they help pull the cube into alignment (but only if the ends of the 550mm bars are square!) they are simple and uniform - 12 bars, 8 corners, 24 bolts... 3 part numbers and 1 tool - can be assembled in about 10 minutes.

Now decent assembly guides are back on the menu - so you can populate the bars with the correct number of weld nuts during assembly for all the stuff that attaches to the frame AFTER its assembled...

[John Griessen]

On 11/10/2012 10:50 PM, Data Pathway wrote:

[John Griessen]

On 11/10/2012 10:50 PM, Data Pathway wrote:

I made up corners, because they help pull the cube into alignment (but only if the ends of the 550mm
> bars are square!) they are simple and uniform - 12 bars, 8 corners, 24 bolts... 3 part numbers and 1 tool - can be assembled in
> about 10 minutes.

Here is a sketch of the same function, but with extra metal carved away -- aimed at low cost 3DP production
with the least amount of metal needed. Maybe it can be shopped around and quoted from shapeways, or ??
No harm in asking, right?

Shapes machine on every face are time consuming even with CNC...so how to
sidestep those steps is what I'm thinking.

Cube to cube attachment for making a production line is best done with the rails, right?

http://cottagematic.com/images/cube-corner-sintering-minimal-1.png
http://cottagematic.com/images/cube-corner-sintering-minimal-2.png

John

[John Griessen]

On 11/09/2012 08:35 PM, Joshua D. Johnson wrote:
> Usually you just drill and tap the end of one extrusion, drill and connect the other two and you have a strong corner., no need
> for castings or gussets/brackets.
>

> On Nov 9, 2012 9:20 PM, "John Griessen" <jo...@industromatic.com <mailto:jo...@industromatic.com>> wrote:
>
> On 11/09/2012 08:06 PM, Joshua D. Johnson wrote:
>
> I guess I don't really understand what the corners are for., usually you can just bolt the pieces after drilling two holes.

Maybe the extrusions could be made such that a rectangular chunk put between pieces just before assembling as above
would align them? I think they already are... The only thing not as provided by aligner-cube-corners
would be getting the holes drilled accurately and getting the end face flush with eh other two rails's sides.

Aligner-cube-corners make assembly of any size cube easy if you can chop saw rails to the same length.
Chopping saw rails to work with aligner-cube-corners is easy to do with low skill
and time spent, since a clamped stop on a saw table handles it.

[Data Pathway]

Well I'm of the school of though that "anything that works well enough" is a valid solution

I like the corner cubes with the alignment tabs, AND my utopic thinking about parts is this:

the near term goal is to automate part production, some of the parts I'm currently making are "labor intensive" because I am doing the operations (or the steps between them) manually.

the concepts of "Labor intensive" and "automated" are contradictory, so I see these parts as a good mechanical solution, and once thier production is automated I don't really care how "hard" the machines have to work to make them - as such, I may not have chosen the "easiest" way to do this - as you are all aware: we are making this system up from thin air and Ideas - so whatever works best for you is great, if it fits in with the rest of the system, even better ;-)

James

[James]

In the very near term I plan to joint the cubes side to side with short pieces of 1" x 2" aluminum strap with 2 holes at 50mm spacing

Crude, simple and TEMPORARY

what I'm shooting for in the near-term is either a pin and latch engagement mechanism - or a plate and rotary latch (like a dzus fastner, but bigger)

these would be externally actuated (manually or robotically) the module/pallet carrier should be able to engage a low speed rotary tool to actuate whichever mechanism is chosen to lock a module into its frame.. and  a very similar arraingment will be used to link the cubes together

[Robert Teeter]

Guys - here is an example of what can be done with PLA and Aluminum casting:

http://3dtopo.com/lostPLA/

I hope that this will help.

Bob Teeter

[Robert Teeter]

James can you post where you got the parts for the cube as checking in my Misumi book ((HFS5-2525-xxx page 2241) it looks like the extrusion is about $7.90 per meter.  Just checking up on this.  part HFS5-2525-550 should be about $4.35 if my book is correct.  To build it out of 2040 would be about (HFS5-2040-550 $8.90 per meter) $4.90 per piece.

Bob Teeter

[James]

looks like your getting a better price than I did... if I buy 10 sticks of this

http://www.ebay.com/itm/80-20-T-Slot-Aluminum-Extrusion-25-S-25-2525-x-24-N-/370515348184?pt=LH_DefaultDomain_0&hash=item56446f8ad8

at 22.54 per and add the 4.39 per stick of freight that comes to a raw stock price of 6.73 each at my door 

add tapping, cutting etc I'll guess another 2 bux a stick or about $108 per frame for extrusion only - still  pretty cheap for such a robust frame... ;-)

So its either gone up a bit since I last checked, or I got a break on the price first time around.

[James]

Thats a very nicely documented sequence!!! great stuff!

[Robert Teeter]

Ok here is the cost for a cube made from 25mm square extrusions from Misumi USA, Inc.

12 pieces of 25x25x550     HSF5-2525-550   $4.35ea  total $52.14
8   pieces of 3-way corner bracket  HBLCR5-S   $3.76ea  total $30.08  comes with  3 screws per bracket  It also has a cap to look purdy.....
or
24 pieces of thin brackets HBLSS5  $.68  total $16.32  plus screw CBM5-8 2ea per bracket plus nut HNTT5-5 2ea per bracket
CBM5-8 @ 0.19ea  total $9.12     HNTT5-5 @ $.37ea  total $17.76

Total with corner bracket $82.22 plus shipping
Total with thin brackets  $79.02 plus shipping

This is just the cube with nothing else added.


Bob Teeter

[kc koellein]

James,

Bob's quote here seems to indicate that you may have settled on a half-meter cube size since you and I last talked about CubeSpawn at the 10bitworks hackerspace? Is that right? I remember you we into something like Imperial Metric system, or something I wasn't familiar with... but you sold it confidently, that I remember!

:-)

kc

--
kc
KF5QVL

- When dealing with the dishonest, BE dishonest. Honestly, this is the best policy.

[Data Pathway]

Bob

You Rock!

thanks for the quote and part numbers I think that should go into the wiki in the "cube" pages...

and I'm glad to see the "corners" problem isn't one.. ;-)

kc, nope! not exactly...

I STARTED with 1/2 meter then went to the british metric foot (300mm) due to its general compatability with metric units AND imperial units + plus its adherence to ISO2848

(as shown here - scroll down to "buildings"

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

So the 550mm bars + the 50mm of the two edges gets you to 600 mm overall dimension in Bobs quote.

Since the cube is 600mm and the material is 25mm the "model" number for this implementation is "625"

if you make it using imperial stock (1") then its a "610" model as shown in this breakdown:

Right side 4th paragraph

http://cubespawn.com/old-site/600mm.html

So that people can use whatever they have easiest access to - dilligence is needed to examine generic standards for the cubes to make them keep thier compatability - or migrate toward it when big variances are discovered, (IMO at least)

hope thats a good enough answer!! ;-)

James