[jono5axe]

Build thread for a large format CNC 5-axis gantry router (composite gantry), similar to the attached photos of commercial models.

General Specs:
- 4000mm (X)
- 4000mm (Y)
- 2000mm (Z)
- 360 degree (A)
- 270 degree (C)
- ATC Spindle 10kW 9Nm 24000rpm
- Linux CNC / Mesa 6i25, 7i77 & 7i76
- Composite Gantry

We are building this machine for use in our business, Solpont Composite Engineering, for the purpose of:
- Machining large composite patterns/molds.
- Machining large aluminium patterns/molds.
- Post processing of composites parts (trimming & drilling etc).
- Scanning/probing of existing parts for importing to CAD for product modifications.

Thank you for your attention, I look forward to your comments and assistance as we progress through this project, hopefully we will make it out the other side.....

Regards Jono

[awerby]

Do you have a plan of your own, or just copying that DMS? How much of it will you be building out of composite materials? Are you using carbon fiber/epoxy or something else? Are you building your own 4th/5th axis head, or sourcing a commercial one? Same question about the ATC - build from scratch or buy?

[jono5axe]

The design started at the tools, >> then toolholder system, >> then spindle.

Decided on HSK63F toolholder system, Hiteco 24000rpm ATC spindle.

[jono5axe]

Making the composite gantry beam first, before steel main frames, to allow the gantry beam extra time to cure / stabilise.

Construction method:

- Composite gantry beam 4500mm long x 600mm high x 400mm deep [RHS - rectangular hollow section].
- Construction: Sandwich construction - marine ply core, internal laminate = 2mm biaxle / double bias glass fibre. External laminate = 4mm carbon/glass fibre. Transverse internal gussets/bulkheads for torsion. Closed ends.
- Resin system = epoxy.
- Method: vacuum infusion. [make up core material as C-Section, lay-up inside C-Section, bond in transverse gussets, close C-Section by bonding on last side panel, vacuum infuse external laminate in one hit].
- Metal threads - 50x20mm m/steel flatbar laminated into the core.


Pic#1 - vacuum test of the core / internal laminate.






Pic#2 - finished composite RHS gantry beam (still wearing peelply).

[mactec54]

Making the composite gantry beam first, before steel main frames, to allow the gantry beam extra time to cure / stabilise.

Construction method:

- Composite gantry beam 4500mm long x 600mm high x 400mm deep

Great choice of spindle, looks like a fun build

[jono5axe]

[jono5axe]

A couple of progress pics.

X-Axis main members finished and ready for filling with epoxy/aggregate. The beams are upside-down in the pics. The bolts/plates are for bolting on to the main legs/frames, leveling adjusters, etc.

(That is the composite gantry beam in the background).

[jono5axe]

Composite Sandwiches.

The below picture shows a sample of a 'composite sandwich structure'. It shows a laminate on each side of a lower density core material. In this case it is a 12mm rigid foam core (approx 80kg/m3) with 5mm structural glass fibre laminate on each side (this sample also has some kevlar in it, which you can see on the face closest to us). The laminate carries the load (stress), and the thicker the core the stiffer the structure.

[peteeng]

Hello Jono - sandwich structures are only stiffer in the thru thickness direction. Their global stiffness is actually less then if you only had the fibres on the outside. Sandwiches are great for boats that have to keep water (thru thickness pressure) out but not good for things like beams such as you are making. The only reason to make a sandwich in that case is to control local buckling. But your structure will not buckle in this application. There is also a comment to add acetone to make the epoxy thinner do not do this! The acetone is non functional and takes up volume within the epoxy. It then desorbs out of the epoxy over time and the epoxy shrinks. So your surface will change!! cheers Peter s

[jono5axe]

Hello Jono - sandwich structures are only stiffer in the thru thickness direction. Their global stiffness is actually less then if you only had the fibres on the outside. Sandwiches are great for boats that have to keep water (thru thickness pressure) out but not good for things like beams such as you are making. The only reason to make a sandwich in that case is to control local buckling. But your structure will not buckle in this application. There is also a comment to add acetone to make the epoxy thinner do not do this! The acetone is non functional and takes up volume within the epoxy. It then desorbs out of the epoxy over time and the epoxy shrinks. So your surface will change!! cheers Peter s

Thanks for the comments Peter. I understand what you are saying, but i can say that:

1) Most fibre is in fact on the outside.

2) the method of constructing a large beam without a mold utilises the core to support for the fibre/laminate during construction.

3) The skins need stiffness themselves, in order to mount precision equipment (linear bearings, gear racks).

4) The critical consideration for the gantry beam is deflection in torsion (rather than bending) and I consider that my nice stiff sandwich rectangular-hollow-section is going to have less deflection than a floppy thin skin rectangular-hollow-section in torsion (bigger moment of gyration?). [circular section would be more pure solution, but not ideal for mounting rails on.]

5) I wanted thickness/stiffness/substance to counteract resonance. Thin skin would be like a drum.

6) I like my sandwich.


Ps Acetone added to epoxy >> degraded cure and degraded mechanical properties of cured resin, and porosity.

[peteeng]

Hi Jono - I suggest you do a torsion proof test and measure its torsional deflection. I doubt if the bending stiffness will be the weak link as you say. Then back calculate the bulk shear modulus for the beam and see how you do compared to other structures eg riveted aluminium or steel (like aircraft construction) You'll only have one good opportunity to do this before you build it into the machine then you will know if you have made a good structure. The structure you describe has a small shear modulus and being square is not geometrical ideal. Then you'll have good info for when you build your next machine. I've been involved with composite structures for about 30 years and understand why you would build this way. Composite are great but they need to be executed correctly to get high mechanical properties. These days it's quite easy to get aerospace grade laminates in your garage using infusion methods.

The discussion about mass as a damper is an interesting one. Adding mass is the easiest method to damp a structure, its harder to excite a heavy thing then a light thing. But if you want it to move fast this is a poor method as then you need more power to accelerate it. You won;t get a job in an F1 team or a Boeing design team if you say "let's add some mass here" So using the add mass approach we would make suspension elements on cars heavier or at least make the car heavier which is what the old american cars did. But if you want a performance structure then you have to lightweight the hell out of it and solve the resonance problem (if there is one) another way. There are many ways to dampen a thin structure, from adding creases, to adding soft stuff to it, to adding tuned dampers, to active dampers etc etc. If you want to up the game you have to up the technology. Cheers Peter S

[jono5axe]

Hi Jono - I suggest you do a torsion proof test and measure its torsional deflection. I doubt if the bending stiffness will be the weak link as you say. Then back calculate the bulk shear modulus for the beam and see how you do compared to other structures eg riveted aluminium or steel (like aircraft construction) You'll only have one good opportunity to do this before you build it into the machine then you will know if you have made a good structure. The structure you describe has a small shear modulus and being square is not geometrical ideal. Then you'll have good info for when you build your next machine. I've been involved with composite structures for about 30 years and understand why you would build this way. Composite are great but they need to be executed correctly to get high mechanical properties. These days it's quite easy to get aerospace grade laminates in your garage using infusion methods.

The discussion about mass as a damper is an interesting one. Adding mass is the easiest method to damp a structure, its harder to excite a heavy thing then a light thing. But if you want it to move fast this is a poor method as then you need more power to accelerate it. You won;t get a job in an F1 team or a Boeing design team if you say "let's add some mass here" So using the add mass approach we would make suspension elements on cars heavier or at least make the car heavier which is what the old american cars did. But if you want a performance structure then you have to lightweight the hell out of it and solve the resonance problem (if there is one) another way. There are many ways to dampen a thin structure, from adding creases, to adding soft stuff to it, to adding tuned dampers, to active dampers etc etc. If you want to up the game you have to up the technology. Cheers Peter S

Hi Peter. The beam is built and the proof will be in the pudding. Regarding mass and moving fast (accelerating), this is one of the primary design considerations of course. I fully understand the relationship between mass and acceleration. That is why I built a composite gantry beam. Conversely, a component that does not move (the main frames) are made of steel filled with concrete and weigh a great deal. And the floor slab is 200mm of concrete. In any case all the design values are calculated long ago for the purpose of specifying equipment. Aluminium was not considered for the gantry beyond initial stages due to it's thermal expansion coef. The presence of mass in the form of a small quantity of core material is hardly having a major affect on overall mass, especially considering it relative to a steel design with similar stiffness (and probable residual stresses, have to heat relieve a 5.0m component somewhere). Steel and concrete have similar thermal expansion coefs. The steel laminated into the structure to receive the fasteners adds far more mass than the core material. And in any case, as I have said earlier in the thread, this beam is some what experimental. Hopefully I will be installing the gantry in the next short period. Your comments are all very good and I do hope that you won't take this the wrong way, but I do know all that stuff already, and with composites there is always many many ways to skin the cat. At the moment I am fighting with the mundane and time consuming issue of preparing flat and straight machine beds, so the higher level stuff is not at the front of my mind. J

[jono5axe]

Hi I see you are doing what kiwis do best, build things better than others can

I would rethink your Z axes, if you place the linear bearing at the bottom of the Y axes carriage, and the rails mounted on the Z axes moving column you will see that your support of the Z axes will be a lot better than what you have, the rails should move not the linear bearings, the rest looks fine, I build machine like this also, not the carbon fiber, but I am interested in this construction of the beam Mac

Hi there, thank you for your message. I do understand your point, and I have seen and read that a lot on this forum, but there are reasons for doing it the way it is shown on the picture of the DMS machine.

It is beneficial to have the Z-axis linear rails/blocks mounted as far apart as possible for the purposes of rigidity in the Z-axis at full extension, but, if the rails are moving up and down then the Z-axis column must also be wide, and this wide column will interfere with the work piece. By fixing the rails to the carriage and moving the blocks up and down the column can be smaller and not compromise the workpiece access of the spindle. For example, the rails/blocks may be mounted 600mm apart on the Y-axis carriage for good rigidity and the Z-axis column can then be say 250mm diameter, and will not collide with the workpiece.

But, maybe there are other factors that I am not considering?..

Regards Jono

[NIC 77]

Really ambitious project, is this your first build? I assume you have a bunch of machining, and CNC experience already? Is your design finalized?

So for that length of travel and the performance you want cutting aluminum, Your Z axis tube will need to be a monster! Are you going for a composite Z axis tube, I assume so?

For the Z axis, Pneumatic counterbalance or air over oil pneumatics? Rodless Cylinders? Something else?

Epoxy leveling for the rail mounting surfaces?

So many questions really. Can't really comment on the design unless I see some drawings / pics from your CAD files.

Are you modeling your forth and fifth axis based on an already proven solution? And by that I am not talking about the Hiteco design persay, I am talking about the inertias of the harmonic drives, the gear ratio, and the motor torques. The output acceleration of the drive will be 1/gear ratio x the input acceleration from the motor, so really high gear ratios with planetary gears (or harmonic drives) that have large input inertias may hamper your ability to accelerate quickly at the output side.

Just wondering if you've done the math on that or had someone look at it, or copied those aspects from something that is proven?

Good luck to you on your project. I look forward to stopping in from time to time to see your progress.

If I lived in New Zealand I would be banging on your door asking if I could help.

[jono5axe]

Really ambitious project, is this your first build? I assume you have a bunch of machining, and CNC experience already? Is your design finalized?

Chuckle - yes, first build, no experience........ But you can't let these things hold you back.

Are you modeling your forth and fifth axis based on an already proven solution? And by that I am not talking about the Hiteco design persay, I am talking about the inertias of the harmonic drives, the gear ratio, and the motor torques. The output acceleration of the drive will be 1/gear ratio x the input acceleration from the motor, so really high gear ratios with planetary gears (or harmonic drives) that have large input inertias may hamper your ability to accelerate quickly at the output side.

Just wondering if you've done the math on that or had someone look at it, or copied those aspects from something that is proven?

Yes, have done some calculations, researched proven machine specs, etc. Yes, z-axis c/balance presents some challenges on a longer travel. Preliminary design is for a pulley reduction (4:1) built into the z-axis assembly powered by a 1/4 length ram. Or a full length rodless setup. But, this has been put to the side for now until we get further into the project. The z-axis is also braked.

Good luck to you on your project. I look forward to stopping in from time to time to see your progress.

If I lived in New Zealand I would be banging on your door asking if I could help.

[/QUOTE]

Thanks for the offer.

[NIC 77]

Chuckle - yes, first build, no experience........ But you can't let these things hold you back.

It's not funny. You could loose alot of money on this if you're not careful. LOL. I can laugh because it's not my money.

Yes, have done some calculations, researched proven machine specs, etc. Yes, z-axis c/balance presents some challenges on a longer travel. Preliminary design is for a pulley reduction (4:1) built into the z-axis assembly powered by a 1/4 length ram. Or a full length rodless setup. But, this has been put to the side for now until we get further into the project. The z-axis is also braked.

I don't really know what you mean. What's a 1/4 length ram?

What do you mean by a pulley reduction 4:1? From what? Stepper? Servo? what size? What lead of ballscrew? What kind of speeds do you want to achieve? You're not really giving me any info that's useful. If you feel you've got it all covered, and this is more of a show and tell, and I'm asking too many questions, no problems, I'm happy to get out the popcorn and watch. Don't worry, as Captain Jack Sparrow would say, I'm rooting for you, mate!

Your monster Z axis will weigh alot! So you will need a counterbalance. You need to design this before you build or you will find yourself with a bunch of wasted space and less movement than you anticipated. That is an absolute monster amount of Z travel.

Looks like these guys make ones up to 480" http://www.parkeroriga.com/pdf_catal...002andP120.pdf
Any idea how much your entire Z axis assembly will weigh, everything that moves up and down? P1V1=P2V2 should help you to size the reservoir. And PV is proportional to T (in Kelvin) will help you to figure out what temp changes will do. So what kind of a counterbalance system do they use on the DMS? Looking at the picture you posted I bet the square thing you can see in between the rails above the Z axis column is a rodless cylinder?

My advice if you haven't got it all designed yet is to stop what you're doing and draw it out in CAD before you spend any more money on parts or do any more building.

If you feel like posting pics of your design I will be happy to comment.

[jono5axe]

I don't really know what you mean. What's a 1/4 length ram?

What do you mean by a pulley reduction 4:1? From what? Stepper? Servo? what size? What lead of ballscrew? What kind of speeds do you want to achieve? You're not really giving me any info that's useful. If you feel you've got it all covered, and this is more of a show and tell, and I'm asking too many questions, no problems, I'm happy to get out the popcorn and watch. Don't worry, as Captain Jack Sparrow would say, I'm rooting for you, mate!

For, say, a 1500mm Z travel, a standard ram would need 1500mm extension and therefore an overall extended length of say 3200mm, which is not very suitable when also considering it is mounted onto a z-axis assembly that is already 2.6m off the ground and typical building stud height is 4.5 m. A 'pulley reduction' means a mechanism consisting of a series of pulleys though which runs a cable, chain, or belt which decreases the travel/stroke required for the ram, while increasing the force. So, a shorter more easily accommodated ram may be used for the counterbalance ram. The statement '1/4 length ram' means a ram that has an extension of 1/4 of the length of the z travel, rather than a ram that has to have extension equal to the full travel (i.e. a full length ram).

J.

[awerby]

I like that spindle too; did you get the toolholders from the same source? How much did it cost (if you don't mind my asking)? Are you planning to use the Doughty Drive Doughty Drive – The 4th, 5th, & 6th Axis solution head for your 4th/5th axes, which is made in New Zealand?

[mactec54]

I like that spindle too; did you get the toolholders from the same source? How much did it cost (if you don't mind my asking)? Are you planning to use the Doughty Drive Doughty Drive – The 4th, 5th, & 6th Axis solution head for your 4th/5th axes, which is made in New Zealand?

The Doughty Drive would not come close to what is need for that spindle

[jono5axe]

Nic 77

Thank you for all your calculations/information. I am indeed familiar with rodless cylinders, as I had said in an earlier post. They have some cons, one of which is cost. You should not discount pulley systems so readily as 'DIY', they are a cornerstone of mechanics. This whole build could be regarded as 'DIY' as we are doing it ourselves, but, as we are reasonably competent mechanical engineers building a commercial use machine I would argue that nothing on the build could be considered as 'DIY' in the context that you have used the term. Another poster said "I have had good luck with gas struts..." well, we do not rely on luck, rather good engineering and intelligent thought.

Air over hydraulic - I do not believe that a hydraulic ram is a good solution for a high speed system (1.0 m/s).

As you have gone to the trouble to provide all that information, I propose that we fully indulge your fetish for counterbalances by giving you correct parameters to work with. I suggest that we should consider the following configurations:

1) Single acting pneumatic cylinder 400mm stroke / 4:1 pulley system.
2) Single acting pneumatic cylinder 500mm stroke / 3:1 pulley system.
3) Single acting pneumatic cylinder 750mm stroke / 2:1 pulley system.
4) Rodless pneumatic cylinder 1500mm stroke.


Parameters:
Z-Axis mass = 120kg.
Z-Axis travel, use 1500mm.
Z-Axis travel velocity: 1.0 m/s.
Available factory air: 80cfm / 100 psi.
Factory air receiver capacity: 0.5m3 + 100m of 1.5"NB Schedule pipe reticulation system.
Additional air (spindle tool change, etc): 2nd stage compression - 175 psi / 5 cfm / 50L air receiver.

Note #1: We can fit an appropriate additional air receiver / accumulator dedicated for the counterbalance system, or,
Note #2: We can consider using the factory air system as the accumulator, or,
Note #3: We can consider using the main frames as the dedicated air receiver / accumulator, capacity = 22 meters of 200x200 RHS.
Note #4: I would suggest a safety factor of 1.5 should be applied to the ram size calculations.

Have I forgotten anything?

Regards
Jono

Ps: merry xmas all