CNC router build to machine wood and aluminium (1mX1mX0.4m)

Mactec, can you elaborate on the "magic number" for a single under-table ballscrew? Seems like it would be more of a "magic ratio" between spacing of the two carriages on each rail and the width between each rail. Obviously as peteeng says 1:1 is ideal, square arrangement. However, it must scale. So is 1:2 acceptable? say, 250mm carriage spacing and 500mm between rails? Assuming medium preload a la hiwin ZA.

I'm guessing that since the spacing of carriages along the rail of typical routers is around 200mm/8", you are saying that given that, 500mm would be the largest acceptable rail spacing. Which is 1:2.5 ratio

Quote:

---

Originally Posted by catahoula

Mactec, can you elaborate on the "magic number" for a single under-table ballscrew? Seems like it would be more of a "magic ratio" between spacing of the two carriages on each rail and the width between each rail. Obviously as peteeng says 1:1 is ideal, square arrangement. However, it must scale. So is 1:2 acceptable? say, 250mm carriage spacing and 500mm between rails? Assuming medium preload a la hiwin ZA.

I'm guessing that since the spacing of carriages along the rail of typical routers is around 200mm/8", you are saying that given that, 500mm would be the largest acceptable rail spacing. Which is 1:2.5 ratio

---

It's not really a magic number, the word magic gets peoples attention, not to build a Gantry machine, too big with a single Ballscrew

It does not have anything to do with Bearing spacing if you can't get the ( 2 ) linear rails parallel to each other, ( .0002" to .0005" Parallel ) which most can't, unless the mounting surfaces are machined and both in the same set up 500mm x 500mm and anything in that range can work without to much racking with a single Ballscrew, larger sizes is not going to happen without racking no matter how you space the bearings, unless you can get precise alignment of the rails

What causes the Racking and binding is bad alignment and rigidity, not the bearing spacing, Bearing spacing only spreads the load of the Gantry which is important as well

On my builds I have used from 150mm / 200mm on the small machines depending machine size with a single Ballscrew, and 260mm on the mid size machines with a Ballscrew on each side, this does not prevent racking, only alignment can do that

Here is one I built some time ago that I did a video of this had 160mm but was not a very big machine and in the 500mm window for travels, The Gantry was also a complete box, this can also helps to prevent racking, but again alignment is key to any machine build

https://www.youtube.com/watch?v=iTBr_rJXZSw

Hi Catahoula - Once you get very high aspect ratios racking is not preventable by the bearings but has to be managed. Racking is either in the bearing clearance (use high preload bearings) or in the stiffness of the structure (make gantry as stiff as possible) and finally alignment is really important.

I was at a clients shop once and they were setting up a 6ft wide by 16ft long gantry router. The long axis was driven by helix racks each side. The gantry was a massive steel box. I pushed one side and it moved slightly so even at this quality of machine they rack. I think the underside single drive limit is more related to the bed stiffness then the drive. As the machine gets wider the bed has to get stiffer. If you have a machine that does plunges for instance you don't want the bed to flex down. Since the single drive underneath inhibits a large bed depth I think this drives the limit for the configuration. Peter

Ok, yeah alignment and execution is going to be important. I guess I was taking that as a given and looking more at the principles behind the geometry. I wouldn't build an under-table ballscrew machine unless both rails and the ballscrew were mounted to the underside and could be referenced to a single datum surface. Like an upside down surface plate, or a steel frame with precision machined slots and mounting surfaces. And clearly an extremely rigid gantry would be necessary.

My point is just that if a moving table router has good stiffness with one ballscrew because of carriage layout (and lack of a vertical offset to cutter/moment above the rails), that could be applied to single ballscrew moving gantry with the correct execution and carriage spacing. The design I'm playing with would be pretty extreme- 300mm carriage spacing along rails and only 420mm spacing between rails. Of course, there's a lot of lost working space (~100mm) that could be recovered with a dual screw drive and closer carriage spacing. But with an extremely rigid gantry (Like no plate at all, fully boxed with steel tubes for gantry, supports, and the crossbar underneath, all brazed/welded) and a relatively short gantry span (~450 mm), the auto-squaring processes could be a problem. Looking at the manual, Centroid Acorn seems to require the gantry to flex somewhere around 0.01 inches during the auto-squaring process. Also, my research indicates that ballscrew mapping isn't available for a paired/slaved axis (i think this was Acorn, could be wrong and may not apply to all controllers), so to get higher accuracy from cheaper ballscrews, having a single screw compensated with mapping is an advantage for any budget-oriented machine.

Anyways, don't mean to derail this thread, and am not at all arguing against the general wisdom of the dual-ballscrew arrangement. Tony_Stark's machine looks like it's coming along nicely.

Hi catahoula - Really big gantry mills have two rails each side and a square bearing pattern each side ie 4 bearings on each side. I think this is to counter the racking issue. Plus the columns are lifted on using cranes and this way they land on something that balances. But I think the racking issue is the bigger one as they would sort out assembly if it was worth saving two rails and cars etc etc. Peter

Peter, interesting. Your point about table thickness/stiffness is good- adding 100-150 mm to the height of the gantry supports is certainly a concern if not done with great care

Quote:

---

Originally Posted by catahoula

Peter, interesting. Your point about table thickness/stiffness is good- adding 100-150 mm to the height of the gantry supports is certainly a concern if not done with great care

---

Hi Catahoula,

I am a learner. Don't worry about derailing the thread. I am trying to decipher what the conversation is about and what can I learn from it and make a better machine.
I am getting a lot of help here from Mactec and Peter and accordingly my design is improving each day.

Hi Peter,

I didn't get a chance to work on the CAD on the weekends. I will work on it details like how I am going to secure the C section to column. The C section will be 10mm in thickness and I planned to bolt it with M4 nut all along the C section and column meet points.

I will work on the height on gantry to accommodate tool and wasteboard thickness plus my vice. Also, I was thinking to work on the forces and torques/moments that will act on while the machine does various jobs. I will come up with some representation and seek people advice here. The intention is to counter any such movement with a design improvement.

I have a option to replace this C section with a 40X120 Aluminium Profile. That means It would be a box support and should resist any bending/moments over vertical columns. I thought by having C section securely bolted to columns, I will make a box which will resist bending torque but this resistance will be provided by the column itself which betrays the main idea of not transferring such moment to it.

What if a I section is placed behind? The one that we see on rail tracks. That would be the strongest but finding such Al beam is difficult and also i was thinking if I am over designing. Reaching the sweet spot is the target keeping manufacturing, time, availability and cost in mind.

On Limit switch, do I require Normally open or normally closed ones?

Hi Rupesh - M4's are not very stiff even if you use a lot of them and their preload is small. I think you will need to cut your channel to make a bracket and use big screws/bolts. or buy brackets that get used in construction extrusions these are cheap and stiff. Over stiffness is not possible in a cutting machine always choose or make the stiffest solution....Peter

https://aluminiumprofile.com.au/prod...g-angle-59x59/

Got some time today and readjusted the vertical column height. Now with full Z travel there will be 50 mm gap. I am planning to go to the aluminium shop soon just to check out if they have C section. I highly doubt but based on the availability I will have to change the design. The aluminium extrude/profile is hard to find in low quantity. Every one wants to sell me like a ton online. Will search in local market.

As some parts have arrived, I will start machining ball screw ends. The ballscrew I got locally is 2005 with unmachined end.

Good morning Rupesh - 1) You need to move the router down in its clamp. This will allow you to shorten the columns. As drawn the corners of the Z plate will catch on things like clamps, vices and bolts used on the bed. Chamfering the corners is a good idea as well 2) The columns will need webs like the Z plate will. cheers Peter

You should flip around your Z axis so the linear slide carriages are on the X (?) axis plate, and the rails are on the moving part of the Z axis. A small change that will improve rigidity

Hi Rupesh- If you have trouble getting the channel consider making a "box" out of plate. There is an issue with channels and hollows in that when you place the rail on the edge of these extrusions (or steel tubes/angles) the rail its on "air" if you look at commercial mills they never have the rails on air but are on substantial chunks of metal. May not seem much but if you do FEA on these things you will see that locally they do deflect a little in this situation and it's a little springy so can vibrate. If you build the gantry from plate you can mount the rail directly on a thick edge. This is locally much stiffer then when mounted on an extrusion corner. Cheers Peter

This also means you have thick edges to screw to the column... maybe use 15mm rails so they are a little smaller...

Quote:

---

Originally Posted by peteeng

Good morning Rupesh - 1) You need to move the router down in its clamp. This will allow you to shorten the columns. As drawn the corners of the Z plate will catch on things like clamps, vices and bolts used on the bed. Chamfering the corners is a good idea as well 2) The columns will need webs like the Z plate will. cheers Peter

---

Peter, moving the router down in clamps will result into even bigger columns? isn't it? Like say in current configuration if I move the router down in clamps, it will touch the bed and now to keep 50mm gap between bed and router collet i will have to increase the column height. M i missing anything.

Your suggestion on router base plate hitting the corner/bolts and stuff is to the point. I have given it a chamfer now.

Quote:

---

Originally Posted by jones

You should flip around your Z axis so the linear slide carriages are on the X (?) axis plate, and the rails are on the moving part of the Z axis. A small change that will improve rigidity

---

Jones, I understand what you said. I agree doind that will make it more rigid but I have a constraint on the length of rails here. In this configuration I can get away with a 360mm long rails but if I do otherwise as you suggest, I will need longer rails.

I have 3meter rails already with me out of which 2 I am using on X axis(for side rails). The remaining one I will split it into 2 parts 640mm for Y axis and 360 for Z axis. budget is the constraint you can say.

Hi Rupesh - Your the conductor of this lot and that's always more then one way to do the job. Keep at it... Its a very big puzzle...Peter

Quote:

---

Originally Posted by peteeng

Hi Rupesh- If you have trouble getting the channel consider making a "box" out of plate. There is an issue with channels and hollows in that when you place the rail on the edge of these extrusions (or steel tubes/angles) the rail its on "air" if you look at commercial mills they never have the rails on air but are on substantial chunks of metal. May not seem much but if you do FEA on these things you will see that locally they do deflect a little in this situation and it's a little springy so can vibrate. If you build the gantry from plate you can mount the rail directly on a thick edge. This is locally much stiffer then when mounted on an extrusion corner. Cheers Peter

This also means you have thick edges to screw to the column... maybe use 15mm rails so they are a little smaller...

---

I seriously miss my engineering days. I use to have Ansys and perform FEA on mechanical buggies/ATVs. Its been more then a decade now but I do understand what you said. I will look for such options, even if I get a C section with wider base and thicker thickness, that will make up isn't it? Like I could mount the rails way inside and will have enough metal to secure it to the columns.But I highly doubt now that i will get any aluminium C section here so may have to look for such alternatives.

Quote:

---

Originally Posted by peteeng

Hi Rupesh - Your the conductor of this lot and that's always more then one way to do the job. Keep at it... Its a very big puzzle...Peter

---

Hahaha, its fun actually. thinking all of these and then making it and see it running will be awesome feel.

Hi Rupesh - Autocad are closing out various functions on Fusion360 so have a look at FreeCAD. It will always be free and has free FEA and CAM. It gets better all the time as far as I can see. I thought there was a bit of an engineer in there from your work. I do FEA nearly every day. Mining trucks, medical equipment, boat structures, routers and mills. So that's my lot...Peter

Quote:

---

Originally Posted by Tony_Stark

I seriously miss my engineering days....

---

Hi.
Yes, you miss. Your gantry is a really high, so you must consider the torque of whole gantry ;-)
The fact - if you want to cut some harder wood (or aluminum!) with any router bit more that 2mm in diameter, you loose all precision - and whole right cutting properties - because your gantry will screw like a paper in the wind. Especialy in Yaxis movement.
Make the X linears a little more distant to each other, there is no reason why not to do that.
If you think that this is not neccesary, try to cut your gantry from a cardboard, just glue a lolipop as a spindle in the middle, like your future router, and try to press at the end of this lolipop. You will see.

The height really gets the attention, isn't it. :P

I think that is why Peter suggested to make web for the columns to prevent that twist action. I think for my application the total Z travel can be cut a bit and the horizontal spread can increased. But I am still figuring it out.

I have no access to my laptop untill next mid week. meanwhile, I practiced threading for ball screw end on a aluminum round. The end nut at the coupling block side is M15 for 2005. The nut is M15X1. So I set my lathe to cut a pitch of 1mm. can any one tell if there is any formula to calculate the shaft diameter for cutting a M15 thread. Does it has to be 15mm exactly or less and then thread is made over it. Cause the nut is fitting over it but its very tight. I have used 60 deg thread cutter tool.