MASSIVE 2.5M x 4M BED - GANTRY DESIGN HELP NEEDED!!

[woodhouse]

Hi All, and thanks for any advice anyone is able to provide, its greatly appreciated.

We've currently got a project where we are building a massive 2.5M x 4M bed machine with a multi function bridge/gantry.
Its primary use is for a nailgun, for nailing in a pattern but while we are building it it might as well be able to machine as well.

Anyone who isnt sure what we mean just google 'MBA Mobi one' or 'weinmann wms 100' but we intend ours to be much more compact as space is at a premium.

We all ready have the massive overkill NSK 55 profile rails and the bed/frame side of things is not a problem. What we cant decide on is the design for the gantry itself...

so far we have come up with the following...

1) use a large hollow section with the rails tapped into it but would a piece of flat bar be needed under to increase the thickness to get a good tap?

2) use a large c section with the rails bolted through?

3) angle welded flush to face of a hollow section to allow bolting through?

4) smaller hollow welded to produce the gantry?

any ideas and experience of any of these??

I'm assuming that one rail on the front face and one on the top face of the gantry would provide a better design or does it not really matter?

We also still haven't decided on the drive side for it either, whether rack is the way to go or rotating ballscrew?

thanks in advance

Tony

[Blackrat]

option 1 for me please :D, or even use a section that has enough wall thickness , remember this surface must be flat to have an accurate machine, if you weld a plate on, how will you flat it ?

and gearbox with rack and pinion

[D.C.]

Thompson have a ballscrew wizard here you might find useful:
http://www.thomsonlinear.com/website/com/eng/design_tools/linear_motioneering_screws.php

If you don't mind a lot of welding, it might be worth considering a truss type design like pg 178 of this machine design document:
http://www.mech.utah.edu/~bamberg/research/PrinciplesOfRapidMachineDesign/Principles%20of%20Rapid%20Machine%20Design.pdf

[woodhouse]

Thanks guys!
The clearance under the router needs to be quite large so there will be quite a bit of leverage on the gantry from the router.

[Boyan Silyavski]

Take a look at my second build at signature for a strong gantry design. May be that will give you some ideas. I go as far to say that its the strongest for the weight i have seen.

Now if that's your first build, i would be very careful, you will either way overbuild it or way under-build it. Both could cost you money and disappointment.


There would be difficulty driving gantry at 4m with ball screw. While the rotating nut is the solution, i dont know from where you will obtain 4m ball screw. I have to speak with Fred from BST but when i was buying from China, the longest 2510 was 3m .

In fact for such a machine i would pay sb to design the gantry for me if i were you. I or some other people at the forum could do that service for you. Just post it at the relevant work quote section. IMO that would be your best bet and save you big money on the build itself and time also, especially if you value it high.


Dont get me wrong, the things you will have to learn to design properly such a big machine are way too much for an inexperienced person. Not to speak of that even welding CNC structures is time costly and needs patience and skill, that the usual welder at a machine shop does not posses.

If you insist going the DIY way, then make a drawing, show it here and we will help you better it with ideas and critics. But speaking without drawing is a waste of time.

[Gary]

Actually anything 25mm diameter or over is available in up to 6M long, the problem is the shipping.


Take a look at my second build at signature for a strong gantry design. May be that will give you some ideas. I go as far to say that its the strongest for the weight i have seen.

Now if that's your first build, i would be very careful, you will either way overbuild it or way under-build it. Both could cost you money and disappointment.


There would be difficulty driving gantry at 4m with ball screw. While the rotating nut is the solution, i dont know from where you will obtain 4m ball screw. I have to speak with Fred from BST but when i was buying from China, the longest 2510 was 3m .

In fact for such a machine i would pay sb to design the gantry for me if i were you. I or some other people at the forum could do that service for you. Just post it at the relevant work quote section. IMO that would be your best bet and save you big money on the build itself and time also, especially if you value it high.


Dont get me wrong, the things you will have to learn to design properly such a big machine are way too much for an inexperienced person. Not to speak of that even welding CNC structures is time costly and needs patience and skill, that the usual welder at a machine shop does not posses.

If you insist going the DIY way, then make a drawing, show it here and we will help you better it with ideas and critics. But speaking without drawing is a waste of time.

[Boyan Silyavski]

Gary,
could you provide such long screws? Its good to know

of course i am not expecting china prices :beer:

[Gary]

yes of course up to R5050.

[woodhouse]

I'd already spotted zapp had long enough stock ;)

Yes this is my first cnc machine build so inexperienced form that point of view, not in engineering, electonics or CAD though.

What we need is similar (although on a smaller scale) to what is pictured bellow.

15687

As you can see there needs to be a larger distance under the gantry to a normal board type router, which of course increases the leverage force from the tool to the gantry.

As quite rightly said, I don't want to over engineer it for no gain.

[Boyan Silyavski]

So you need the working table to be separate from the gantry and sides?

[woodhouse]

No, that one the gantry is separate so the table can change size. We have no need for that so was planning on the rails being higher up but under the table

[Boyan Silyavski]

To say it better then, you will need to design a table with rails and gantry and then add separate bolt on working area which has to be higher and hide the rails. Otherwise you will be facing an impossible problem of aligning the rails.

Its up to you of course, but maybe for such a machinea separate working table is better idea, and all else modular design using RP, like heavy duty plasma tables.
Cause at the end of the day this is a production machine. Tomorrow you could decide you need 5 meters longer or sell to sb who need 1 meter longer, etc...

Adding then ATHC and plasma torch is all needed to convert it to plasma , and so on. means easier to sell if needed an upgrade.

Of course your priorities could be different.

Using ball srew and rotating ball nut on such a machine could make it very high value machine that can precisely cope with big stuff. Combine that with strong gantry and you will have a real beast

[woodhouse]

Thanks boyan.

Well that was my thinking too.
The project started out as just a nailgun bridge and was going to size components accordingly as the forces involved from the small nails is minimal and the there's no real deflection forces from the tool other than upwards.
Then I managed to acquire 55mm thk profile rails and carriages off an old machine which got me thinking why not design a more capable machine with a spindle.
Will enable us to do so much more!

[Boyan Silyavski]

ha ha, i see you finishing with 2 spindles and drilling head , so design the z cleverly, better oversize the z plate so you could fit this and that.

[woodhouse]

Haha you know me so well already!!

[woodhouse]

I know this is a very ambiguous question, but does anyone know the maximum horizontal force that can be expected from a spindle?
From this I can obviously work out the resulting force on the gantry rails for different designs and keep the gantries deflection below spec

[Boyan Silyavski]

Depends what spindle you are going to fit. 2.2-3kw water cooled or 10Hp and up using BT holders...

But you have it all wrong there. What would be the biggest longest tool you want to fit and desired deflection. Download HSM adviser it calculates tool deflection nicely.


Cause the tool will be the weak point there.


Also there is very nice stiffness calculator here in the forum. Thats what you need.


Dont worry about the bearings of the spindle, they will take much more than you could push them even routing aluminum.

[woodhouse]

Thanks boyan, you're being really constructive. I appreciate it.
It wasn't the actual tools deflection I was concerned about.
What I am trying to work out is obviously the gantry design.
Now I could go crazy and use a 500x500 square 10mm RHS and would be incredibly heavy but would do the job but waste a fortune trying to accommodate it.
To far the other way and my two 100x50s spaced apart 250mm would be much lighter but probably twist too much to be useable.
Add to the fact that on this build the rails will probably be spaced apart half the distance the spindle is from the bottom rail (eg 250mm apart rails, 500m from bottom rail to spindle)
Then the force on the bottom rail will be 3x that from the tool on the bottom rail and 2x in the opposite direction for the top.
Likewise any deflection in the gantry is magnified at the tool.

If I can work out the force produced by the tool then the gantry can be sized properly for them forces and not twist.
But that's the million dollar question

[Boyan Silyavski]

The key being here is how long the Z axis movement will be. That's how to start , then you will decide on the gantry box section. So how long the z axis movement down will be?

[woodhouse]

We could probably get this down to 500mm

[woodhouse]

But say 600mm to be on the safe side

[Boyan Silyavski]

We could probably get this down to 500mm

Now next is my common sense, later you could prove it in the calculator.

-Then you will need a gantry vertically wide so that 500mm spacing between bearings is achieved.

-2x 150x150x4mm /18kg/m/ separated 200mm and connected every 500mm with braces from 100x100x4mm that all form a leather like structure. Similar to my gantry. Could be reinforced with bolted later steel plate of needed in some direction, though i doubt that you will need for wood.

or simply weld together 2x 200x200x4 or 5mm, though careful engineering here is very important


- You must decide at this point how you will move Z left right, RP or rotating ball nut, how you will move the gantry on long axis, all that before continuing with the gantry design, cause you will need space in between for the rotating nut to pass and so on.

[woodhouse]

Well, I can get my hands on a 50mm ball screw and nut from the same machine the rails came off but the components I will need For it I feel will probably cost more in the long run than RP. So I am tending towards RP at the moment.

[michaelw]

I know this is a very ambiguous question, but does anyone know the maximum horizontal force that can be expected from a spindle?
From this I can obviously work out the resulting force on the gantry rails for different designs and keep the gantries deflection below spec


There are three maxima for the largest forces on tool tip :

(1) The thrust load provided by the ball screw et al . Stall torque of axis motor referred through drive mechanism and ball screw and corrected for any leverage effects .

May have components in more than one direction .

(2) The worst case cutting load . This is usually a max depth cut on one side of cutter only . Stall torque of drive motor referred through drive mechanism and any leverage effects .

May have components in more than one direction .

(3) Crash load . Can be estimated but if you design on (2) that is probably ok .

Simple quasi static loads are not the complete answer . Any cutting is unstable and vibration problems can be more troublesome than the simple loads .

Ned to consider masses , geometry and elastic stiffness of all components to make sense of this .

Generally apart from being just stiff a simple support structure needs to be relatively heavy to bring natural frequency of structure down into a safe zone .

Other methods are available to make lighter structures vibration insensitive .

Beam needs in any case to be stiff in three senses - vertical , horizontal and in twist .

Generally a beam structure which is of large cross section dimensions is desireable . A fabrication of relatively light sections at big spacings is a way to go .

Hollow sections may be filled with structural foam to aid damping and add a little stiffness .

An alternative way to go is to use space frames made of relatively light components arranged very much like a roof truss or lattice bridge .

An unorthodox alternative - which is already being used in some real machines - is to cast a hollow lintel in either concrete or epoxy concrete .

The above all assumes relatively slow movements of tool/table . If rapid moves are likely then there are further design considerations .

michaelw