[moseus]

Goals, need to accurately cut 0.16-0.05" or thicker 6061 and 2024 aluminum with accuracy to cut hole matched parts to mate around .001" accuracy with in tolerance range of rivets. Aluminum sheet size needed is 4'x12' with some parts nearly that size. This will include thousands of rivet holes to be drilled. I would like also to be able to wire cut foam (and router) which comes in billets 4'x12' 3' thick to build molds. Anyone seen an add on to a gantry like this? Location, building a basement foundation this summer and will build foundation per spec. I'm considering hiring an engineer to design a 2' or so thick slab to build flush with the floor and mount with recessed rails. Thoughts? Style idealized, moving gantry, but possibly open. I'm new to CNC, read a bit, understand some basics, have an engineer friend, which said my price range was far to low to buy one already built that wasn't worn out, let alone ship to Alaska. Expected problem, ball screw long and accurate enough seems out of the price range vs speed question. Would a linear index for the X axis combined with a 'blind' drive of some sort handle the distance? (gear driven both sides, or maybe a dual cable/belt driven transport ? ideas?) Looking for ideas for a lightweight gantry which could handle a 3' z axis. Is there a design where the y axis cross member changes height? Would it be better to build a moving table super solid gantry? Is there a better approach, Thanks in advance. Raymond.

[ger21]

I don't think a 3' Z axis will give you the accuracy you need on your aluminum sheet. Look at rack and pinion for driving it. Sorry, but don't really have any other answers for you.

[harryn]

That is a big machine for that accuracy goal in Al. Your needs for power, accuracy, and cutter speed are quite a wide range. It might be more practical to consider the needs in terms of two machines. - one for Al, one for softer thick materials.

Holding 0.001 inches over 12 ft length in Al is not trivial, even holding 0.001 inch per foot of length requires thought. Without running the calculations, I am pretty sure this will require you to temperature control the machine and work to within a few degrees. It might makes sense to consider a laser based X /Y positioning system with active feedback so that all holes are drilled to a reference point with less concern for machine related imperfections. I am assuming you are planning to use some kind of machining fluid / coolant for that much drilling and cutting.

Obtaining a flat large area like that might be a good application for either a pure epoxy floor ( www.precisionepoxy.com ) or see the epoxy granite thread in the engineering section of cnczone. For examples, look for member walter.

Hopefully, you have a really large door to get those precision cut / drilled parts out of the basement.

It might not seem like it at first, but if you are able to stay with a rack and pinion design, and you are sufficiently handy to weld up your own frame, the electronics will dominate the build cost. If you are not going to use BOTH a router and aluminum cutter at the same time, MAYBE it makes sense to consider moving the motors and electronics from one machine to the other depending on the work needs. If it is profitable and you need more capacity, then you can just buy more electronics.

[moseus]

Thanks Harryn and Ger21,

My target accuracy was .001 per foot. I also miswrote, .016-.050 in thickness. It'll be a daylight foundation/basement 16' tall with a door nearly that high. The aluminum is for an airplane design, so it's flexible there. The routing would be irregular, primarily for developing molds for new shapes. Floor space is limited, so, I'm keen to make one machine do it all. Funny how writing things down clears up some ideas, Ger21 is right, light weight gantry and 3' z axis isn't realistic. I think I'll have to move the table. Harryn, do you have any suggestions on where to look up the Laser X/Y position system, this sounds like a winner. Shop will have in floor heat, fairly stable. Thanks for the help. Raymond.

[ahren]

Aerotech sells laser-based feedback systems, but you better believe they aren't cheap. You don't necessarily need this for 0.001/foot -- it sounds like you are worried about error build-up vs overall accuracy. To account for this, you could also go with linear encoders (renishaw and heidenhein both make nice "stick on" versions) on each axis to make up for lead error on your screws or motion error on your rack and pinion.

Good luck!

Ahren
www.cncrouterparts.com

[wcarrothers1]

Don't know what kinda of budget you are looking at but I'm a big fan of this design
http://www.inshorepowerboats.com/cnc.../CNCROUTER.htm

where the router gantry sits on side tracks.. Looking at how my home built came together you can see some things similar.

http://www.cnczone.com/forums/showthread.php?t=26768

Thing I like the most of it is how I can clamp things around the edges of the table and not have to worry about moving parts (other then the router and Z) hitting clamps. A good way to go with things if you don't have vacume holding.

Also gives you the ability to lower or raise the table as you can see the grid for the table support bolts to the sides and could be dropped with spacers/longer bolts or what ever if needed.

b.

[harryn]

wcarrothers1 - that is an excellent build and design for this application.

I looked at the plans that you linked to - Pretty reasonable price for what you get.

[andy_ck87028]

Hi

There are a few design principles that I think should be mandatory with a gantry router.

1. keep the X axis height as close as possible to the Y axis height
2. if you want to be able to cut deep raise the X axis as per the plans in the previous 2 messages within this thread
3. aim to always cut as close to the X axis height as possible in order to minimise Z extension
4. allow for deep Z axis cut by having a variable height table

In addition to that, I have included these other principles

1. have 2 y rails horizontally to minimize flex
2. build the z assembly to a high quality as that it can be re-used on another machine
3. assume large material does not require precision cutting to the same level as small material
4. assume large material is softer than small material (aly is usually small, foam is usually big)
5. small foam is easy on a big machine .... large aly is unlikely on a large machine.
6. Do not try to design high precision into a large machine; instead re-locate the high quality z axis into a high precision small XY assembly
7. Use acme (or better) for the small xy and rack pinion/timing belt for the large xy .... your maximising your dollars that way
8. build a large machine in modules with rack / pinion or timing belts .... just add on another section if you need to ... why clutter the machine shop
9. design with ample adjustment in mind if no access to precision machinery

I hope the above criteria helps you keep the quality high without going over the top with gear that is used for 1% of the time.

Happy new year

Andy

[fireballcnc]

I think those are Excellent building principals Andy...

Even on my small desktop routers, I use those same "rules" for
designing and building.

John

[ger21]

1. keep the X axis height as close as possible to the Y axis height
2. if you want to be able to cut deep raise the X axis as per the plans in the previous 2 messages within this thread

The downside here is it makes for a difficult to load machine, especially when loading large panels. But if you're mostly cutting large panels, you don't need a 12" Z axis, right. The problem comes in when you want one machine to do everything, compromises have to be made.

[andy_ck87028]

Gerry,

Its a design principle. Nothing to do with compromises.

The reason you would want to keep the z minimum and the x as close as possible to the y is to minimise moments.

If you want good access, have an adjustable height table that can be raised after material is inserted.

Surely as a principle, cutting accuracy is more important than feed efficiency.

andy

[ger21]

Gerry,

Its a design principle. Nothing to do with compromises.

I probably didn't word it the way I wanted to.

The reason you would want to keep the z minimum and the x as close as possible to the y is to minimise moments.

Agree with you 100%.

If you want good access, have an adjustable height table that can be raised after material is inserted.

Wouldn't you want to raise it before inserting the material?

Surely as a principle, cutting accuracy is more important than feed efficiency.
andy

Not in a production environment where you don't want to spend as much time loading as cutting. Or on a large machine where loading large sheets becomes very difficult.

I just don't see any high end commercial machines with the rails above the table. Must be a reason for that?

[andy_ck87028]

Wouldn't you want to raise it before inserting the material?

Very true .... my last post was after midnight.... too much alcohol.... feel a bit seedy now

Not in a production environment where you don't want to spend as much time loading as cutting. Or on a large machine where loading large sheets becomes very difficult.

I just don't see any high end commercial machines with the rails above the table. Must be a reason for that?

I think there is a danger that the original context of this thread is being shifted.

I suspect that production machines are all purpose designed and also throughput is not to be sacrificed. This is achieved through componentry that is beyond the hobbiest reach.

Consequently I could see that the optimum production design would have the x rails at table height.

I don't accept that argument for a budget hobbiest design where maximising bang for buck is a driver.

I believe that hobbiest designs should be very adaptable to broaden its application and often one of the first casualties is throughput.


Happy new year for those who are just about to have it

And if you deserve a hangover .... may it be a big one


Andy

[moseus]

Thanks Ahern,

I was thinking linear encoders, additionally mounting them on aluminum itself with one end fixed and the other allowing it to float with temperature, since my medium for the highest accuracy is the same any temperature change, I think should occur in both equally. I'll look up Renishaw and Heidenhein.

Wow WCarrothers1, Budget, well that where this site helps us, I hope to contribute later with experience. Just what everyone is talking about here is quite reasonable. I very much like this design, and allowing for a re positionable table height eliminates the high z axis deflection, in fact this might allow me to go just a little bit higher for clearance. I like the rack and pinion on the Y axis mounted with the axle mounted at height, that removes a bunch of gantry mass and deflection. excellent. I'll think about height change on the table, itself, to accurately keep it flat.

Andy that sums it up in a list.

Loading, shouldn't be a problem, if I need to, this design would lend itself to a simple carriage with some sort of vacuum lift.

I really also like that a variable height table would allow for other types of tables, such as for torching or welding if I get that far.

Anyone seen a foam wire cutting design which would lend itself to this?

Thanks All, Happy New Year, Raymond