[peteeng]

Hi Mogi - Pippin has started a gantry thread, I did some calcs grout vs 200x200x10mm SHS and grout at 200x200 same weight is better then the steel. Damper, can mould in features etc etc. Look up

https://www.cnczone.com/forums/mecha...ntry-beam.html

I think the steel era for machines has ended... Peter

[Mogilogi]

Hi Mogi - Pippin has started a gantry thread, I did some calcs grout vs 200x200x10mm SHS and grout at 200x200 same weight is better then the steel. Damper, can mould in features etc etc. Look up

https://www.cnczone.com/forums/mecha...ntry-beam.html

I think the steel era for machines has ended... Peter

Hi Peter, I red throw all of it and it's interesting topic, did you check CFDST before ? Seems like it combine both of the advantages
Can it be simulated ?

[peteeng]

Hi9 Mogi - What is CFDST? Peter

[peteeng]

Hi Mogi - I looked it up. Concrete filled double steel tube. Yes it can be simulated and I did it for Pippin. I think its a poor solution. I think they missed the point big time1) if you need a stiffer structure you just make it in thicker steel 2) In a machine build case you still have the issue of it changing shape when you machine it I suppose you can weld it heat treat it then fill it with concrete. But that seems to be counter productive to me 3) The concrete and steel do not bond properly. The quoted bond strength is 1Mpa and I'm sure this will become zero over time. If you look at old marinas or structures that use steel and concrete you can see that they are not connected anymore. So the concrete is mechanically locked in place vs "bonded". So its structural efficiency is poor. I looked up the cast friction co-efficient of the bond and used that in Pippins thread f=0.6 . It showed the interface was slipping even under small loads. This contributes to damping the structure. If you use an expanding grout the interface will be stronger. So the concrete and inner tube are along for the ride to gain a very small damping improvement ... pointless.

So the big thing is chasing material dampness does not create enough damping to achieve what machine designers are trying to achieve. If we use the damping ratio zeta to determine this, critically damped machines have a zeta=1.0 most good damping say your car suspension has a zeta=0.6. Structures such as steel and concrete have a zeta=0.03! So its impossible to build conventional machine structures with very good damping compared to what a motion control engineer would call damp. So we have to move along to electronic methods to make truly motion damp machines. This is possible. Servo drives have been doing it forever. Now 3D printers are doing similar things with input shaping and they will go full active damping very soon as if input shaping is implemented its a small step to full active damping.

So in summary if your machine has a vibration issue its not stiff enough. If you designed that said machine then you got it wrong or you where ignorant about the issues and now you have to learn more about that. Make the machine stiff enough and vibration is a second order issue. If you are serious about solving machine vibrations use input shaping (or servo motors with motion feedback) This is available on a few 3D printer motherboards. These are easier to set up then CNC motion control systems and are potentially cheaper then CNC motion control systems because they make millions of them, the system is fully integrated vs via disparate suppliers components with no hope of it ever solving the motion issue and apparently no desire to.

So in summary 1) stay in steel world make if UBER stiff. Have the parts stress relieved and finish machined and your all good 2) make a damper machine then steel, cast in engineering grout, take advantage of being able to make whatever shape you want or need, consolidate parts. Do all this in your garage vs taking things all round the countryside for different processes.
3) consider making a lifting gantry machine, its naturally much stiffer then a typical moving gantry machine. KISS Peter

My next machine will be cast grout with composite inserts. I aim at designing a machine with no metal in it except the rails and bearings and things that I buy. The "structure will be totally composite" and it will use input shaping.

By the way its very difficult to simulate damping. You can identify the vibration modes using modal analysis but unless you know by test the various damping characteristic of each material its difficult to do. Then you also have to be very good at transient dynamic analysis. I have done a bit of this in this sort of area to estimate machine dampness. There's a lot of hoops to jump thru with the data, math and the analysis if there's the need and the $$$ it all can be done by various specialists...I image the big tool builders do this regularly.

[peteeng]

Hi Mogi - Here is a large fabricated steel mill by HAAS., They don't fill things with concrete or worry about dampness. They just have lots of big stiff members doing their job. Peter

https://www.haascnc.com/machines/ver...dels/gm-2.html

[Mogilogi]

Hi Mogi - Here is a large fabricated steel mill by HAAS., They don't fill things with concrete or worry about dampness. They just have lots of big stiff members doing their job. Peter

https://www.haascnc.com/machines/ver...dels/gm-2.html

Hi Peter, you are right I just thought it can help with stiffness, you already said if things are stiff enough dampening is not that big of a deal, so I will continue with steel fabricating gantry triangle and I can add rips in the hollow section to make it more stronge
3 parts of the inside the rips will be welded the 4th will be bolted since I can't weld it after it be a closed section, then it can go for heat treatment then machining the rails part and making it parallel and so on.

[peteeng]

Hi Mogi - Going down the conventional path is good for you. I've tried to bring you to the dark side and failed so keep to the high road, it will guarantee you a good result. Does your CAD have FEA? If so do a modal analysis and this will identify any odd vibration modes especially low freq ones. Keep at it Peter

show us your updated design

[Mogilogi]

Hi Mogi - Going down the conventional path is good for you. I've tried to bring you to the dark side and failed so keep to the high road, it will guarantee you a good result. Does your CAD have FEA? If so do a modal analysis and this will identify any odd vibration modes especially low freq ones. Keep at it Peter

show us your updated design

Hi Peter, yes I'm not trying to distract my self with many options, because I will never be satisfied, go with one and try to get the best out of it.
Yes I'm using Solidworks and it can do FEA, also I got simsolid but still trying to learn how to use it.

About new design, I could't have time to do anything yet about the machine design, just searching allot before I play with design again.

Thank you for all this information, it helps allot.

[peteeng]

Regard stress relief. Do not close out any sections of hollows. The section must be able to breath well. If the heat treater feels that the section can't breath he will refuse to treat it, as its a bomb. Lots of damage done when air heats up that much and is trapped. Peter

[Mogilogi]

Regard stress relief. Do not close out any sections of hollows. The section must be able to breath well. If the heat treater feels that the section can't breath he will refuse to treat it, as its a bomb. Lots of damage done when air heats up that much and is trapped. Peter

Hi Peter, would stich welding or drilling holes help with the trapped air ? Or it need to have big area for air to escape?

[ardenum2]

Design the part as a cast iron part and try to replicate it as a fabrication, internal ribs in cast iron machine tools have holes in them for a core, this will work well for you in case of the trapped air. Go over DMG's videos, you can see the frames inside. You can download 3d models from haas too to view them inside, on the up side haas frames aren't as complicated as dmg's inside so could be easier to reproduce as a fabrication.

So much work for a weak design. If I were you I'd just use square pipes, hell even aluminium extrusions. It's almost sad to watch how much quality work will be wasted on a router...

[Mogilogi]

If I were you I'd just use square pipes, hell even aluminium extrusions. It's almost sad to watch how much quality work will be wasted on a router...

Hi arden, I wish I could just pick 1 meter of a square pipe, problem is nobody around me or even far away of me selling it that way, I have to buy the whole 6 meters. that is why I'm trying to fabricate it, it will be cheaper than buying a 6 meter and selling 5 meter as scrap since I have no other use for it.

[peteeng]

Hi Mogi - The bigger the better. Do a sketch and ask your heat treater, he/she's the one that has to be happy. But big holes in the right places like centers of ribs do not affect stiffness.

Mogi - if your having trouble getting big sections then make truss structures from smaller sections. The whole idea of using a single big section is that's its easy to do. But making a bespoke machine is never easy. You have the advantage of local heat treat. This solves so many issues that you can do things that others can't. Have look at trusses... are you doing the welding? Peter

[Mogilogi]

Hi Mogi - The bigger the better. Do a sketch and ask your heat treater, he/she's the one that has to be happy. But big holes in the right places like centers of ribs do not affect stiffness.

Hi Peter, sure I should do that, I will contact with them and show the design and ask what is required to make it safe for them, then I can confirm it's not effecting the stiffness by running simulation on it

Mogi - if your having trouble getting big sections then make truss structures from smaller sections. The whole idea of using a single big section is that's its easy to do. But making a bespoke machine is never easy. You have the advantage of local heat treat. This solves so many issues that you can do things that others can't. Have look at trusses... are you doing the welding? Peter

Because I have the advantage of local heat treat, it's making more sense to me to design the gantry as big section, I'm not experience designer just trying to learn, so to build a trusses will make it more complicated for me.
About welding, I won't be doing the welding, my Father in law is a welder, he will help me find a good shop that is doing it with a good price, he won't be doing the welds because for now he don't have a bench just moved to a new house.

[peteeng]

Hi Mogi - Here's a good steel gantry design. Is stiff, gives good access to the bearings. Also note that a gantry does not have to be deep at its ends. The ends are under shear but not under bending. The middle has to be stiff in bending the ends stiff in shear. Your FE can prove that out. But this allows say a two piece gantry to be used. Similarly my latest gantry is not deep at the ends which helped with the bearing connection. Plus with the double tube design you automatically get an internal web that helps with the lozenging. Peter

[joeavaerage]

Hi,
yep, stiff alright. Stiff as a schoolboy at a nudist colony!

Craig

[peteeng]

Hi Mogi - Since you have considered using a top and bottom rail design consider a round gantry with rail top and bottom. Under torsion a round shape does not change sectional shape. The model is a 150x150x10 SHS vs a 150dia x10mm thick circular section. The circular does better in the push pull dirn. To make it a bit better in the trans and plunge dirn a bigger tube cam be used. If the comparison was of equal weight tubes that would be the case anyway. Load applied 300mm below section 1000N applied. The z part is rigid.

push pull circ 66um shs 70um
trans circ 8um shs 8um
plunge circ 21um shs 17um

Peter

[Mogilogi]

Hi Mogi - Here's a good steel gantry design. Is stiff, gives good access to the bearings. Also note that a gantry does not have to be deep at its ends. The ends are under shear but not under bending. The middle has to be stiff in bending the ends stiff in shear. Your FE can prove that out. But this allows say a two piece gantry to be used. Similarly my latest gantry is not deep at the ends which helped with the bearing connection. Plus with the double tube design you automatically get an internal web that helps with the lozenging. Peter

Hi Peter, I have this in mind to use double tube design, if the cost is high and gain is low for a fabricated gantry, since anyhow I will have to stress relive them, and weld them, machine them.

- - - Updated - - -

Hi Mogi - Since you have considered using a top and bottom rail design consider a round gantry with rail top and bottom. Under torsion a round shape does not change sectional shape. The model is a 150x150x10 SHS vs a 150dia x10mm thick circular section. The circular does better in the push pull dirn. To make it a bit better in the trans and plunge dirn a bigger tube cam be used. If the comparison was of equal weight tubes that would be the case anyway. Load applied 300mm below section 1000N applied. The z part is rigid.

push pull circ 66um shs 70um
trans circ 8um shs 8um
plunge circ 21um shs 17um

Peter

Hi Peter, as I can see from the result, it's not huge different, but to use a round gantry is harder to install, machine, and fit in the machine, maybe that is why it's not a famous design amount CNC gantries.

[peteeng]

Hi Mogi - I disagree that its hard and there are semi circular gantries out there. And there are very sophisticated machines out there that are "hard " to do.

But that's it for now. You have to wade through many decisions. Every machine part contributes to its total compliance and when you get to modelling your entire machine and aim at 20N/um you will see that every little bit helps. Peter

[Mr-MaW]

I have made a lot of products, if you are interested, you can take a look at my products.

https://grabcad.com/library/drone-cnc-diy-1