[Mogilogi]

Hi Mogi - I have made MDF & plywood machine bases that will machine plastics, timber and aluminium easily. You describe the machine as a Mill and mention Steel. These two words really impact what you are trying to do. Steel requires extreme stiffness and damping, Plus it also requires a low speed high torque spindle. The spindle is going to be a large rabbit hole for you and its worth researching this aspect before much longer as it can change how the Z axis is going to work. Your going to need space for pulleys and large motors, probably an AC servo. If you use a high speed spindle like on a router for steel you will need one with very good bearings plus you will have to adopt a light cut very high speed machining technique. There are people doing this, but its early days in this area of work. Or you remove "Mill and steel" from the spec and life becomes very much easier. But aiming at a Mill is not a bad thing, just its a very tough call for your first machine... Look up my Milli thread, years of research and still no result. Peter

Hi Peter, I'm aiming at a mill, so if I could't get what I want I will end up with something strong to be happy with, at the end it's mostly hobby stuff.

[Mogilogi]

What's your budget and what kind of accuracy do you want? Do you want to produce parts for fun? or occasionally as a jobshop? There's plenty of options. We can guide you but we need more data..

Hi Ardenum, About budget, mostly I'm thinking about the steel price and how much it will take to prepare it, hardware I will use what I bought already, at the same time I'm trying to build a machine that is copy to langmuris, to cut parts for fun mostly will be granite engraving, aluminum, it would be good if I can cut steel like langmuris, if I can get it to get better than them would be even better, also occasionally as a jobshop I would like to cut parts for university students.

Stiffness is key to any performance. If you want to cut steel with a langmuirs copy, I have very bad feelings about that

is that because langmuris machine is not stiff enough ? or because I can't copy it good

If you want to cut steel faster you need a different frame for that. If you want absolute best performance at a budget, you need a double column raising gantry.

You mean the Y rails should be supported and connected to the bed ? like this

[Mogilogi]

Hi,
I second that, as soon as you start to cut steel then that is a totally different machine to anything you have discussed thus far.

Take a look at any of the many drill-mills out there, like this one as an example. Its good enough....just....to do steel. If you want to cut steel
you'll need to do at least as well if not better...and this thing is made of cast iron and has ground dovetails....not an easy task to match let
alone exceed.

https://rongfu.com/mill-drill-machin...-machine-rf45/

Craig

Hi Joe, I'm not planing to cut steel all day long, I need the machine to do it with even lower feed, when I need it, to get me started so I maybe after this machine I build a better.

[ardenum2]

.
no I meant when the gantry moves up and down instead of the spindle like in the usual Z axis assembly. Like this:

https://www.youtube.com/watch?v=0yFKaCn9soo

If you want to built from steel, can you get structural square pipes? Like these?(https://stahlshop.de/stahl-produkte/...profile?page=8) It would be straight forward to build such a frame from square pipes and there would be minimal welding(or stich welding) involved or you could braze them to avoid deformation. Unless you can stress relief at a company if thats cheap.

"is that because langmuris machine is not stiff enough?" yes it's barely stiff enough to cut anything

[peteeng]

Hi Mogi - Looked at the link. Yes topology optimisation is the "old" word for it. It's now called generative design or adaptive design. Seems SW does it via FE and Ansys. This is the slow way to do it as FE meshes take a long time to solve and decimate. F360 and others use meshless algorithms that solve very fast. The generative solutions often are unbuildable but give insight into how it could look unless you 3D print. I don't pay for coupons so don't get the full info out of F360 but enough to get clues to how to go fwd with a new design. Peter

Re steel - There seems to be an idea that you can do something like cut steel "occassionally" with an under-par machine. Sorry this does not work. The machine either will cut or won't cut, if under done it won't cut, it will precess, wobble, stall, vibrate etc etc. I've been thru that with aluminium. If you want to cut steel you really need to look at machines that do what you want and gain an understanding of why they work and how they work. Even the spindle is a different rabbit hole. The spindle suppliers will not say that they can machine steel. They say they can engrave steel. So the lifetime of a "router" spindle will be short. So keep researching, long way to go. The year will pass quickly... Peter

https://youtu.be/6dI8_-fguk0 heres a small router cutting steel and it is like dentistry, so its possible if time is not important...

The other aspect is that Mogi has picked a gantry machine. This is usually because of the size of the parts required to be made. Mogi if you don't need such a big platform make a Bridgeport style machine. This will be easier to make as a mill. A gantry machine has an extra axis, gantry alignment and stiffness issues that need to be overcome to cut steel... I'm not trying to talk you out of it just trying to lay out the hurdles..

I would find an exemplar machine and study it... Peter

[Mogilogi]

Hi Mogi - Looked at the link. Yes topology optimisation is the "old" word for it. It's now called generative design or adaptive design. Seems SW does it via FE and Ansys. This is the slow way to do it as FE meshes take a long time to solve and decimate. F360 and others use meshless algorithms that solve very fast. The generative solutions often are unbuildable but give insight into how it could look. I don't pay for coupons so don't get the full info out of F360 but enough to get clues to how to go fwd with a new design. Peter

Hi peter, Yea i got the point, it helps you identify the weak spots in your design and where can you put support to make it better, can i export from solidworks to F360 to use their Generative design ? or my design should be built in F360?

[Mogilogi]

.
no I meant when the gantry moves up and down instead of the spindle like in the usual Z axis assembly. Like this:

https://www.youtube.com/watch?v=0yFKaCn9soo

Hi arden, wow that is impressive machine, i wouldn't even dream to have one like that or even close, this design is good and more stable i'm sure, maybe my next machine would be from this type

If you want to built from steel, can you get structural square pipes? Like these?(https://stahlshop.de/stahl-produkte/...profile?page=8) It would be straight forward to build such a frame from square pipes and there would be minimal welding(or stich welding) involved or you could braze them to avoid deformation. Unless you can stress relief at a company if thats cheap.

There is some companies that are selling that kind of structural square pipes in my country, but they only sell it as 6 meter long and i only need 1 meter, so going with combining plates together is the easiest way, about stress relief i'm searching for a company that can do that, if i could't fine, i'm planning to build my own stress relief furnace maybe that can be an option, similar to this one

yes it's barely stiff enough to cut anything

Is this not good, i don't have good experience with milling steel, that is why what i saw looks good for my use case and what i'm thinking i can be doing with the machine.


by the way my main job is a Mobile App Developer, this CNC stuff is my hobby, i like to build things + i have young boy children i want them to grow seeing that kind of stuff and be interested in it to grow their imagination.

[peteeng]

Seems my response is lost - Damn
Yes Mogi - topological optimisation is called generative design by many now. SW uses FE via ansys. This means its slow as the FE uses a mesh and then after solution it decimates the mesh and rebuilds etc etc. Newer systems use meshless algorithms which are really fast. Peter

[peteeng]

Hi - I have entered here twice and they are lost?? Peter

[peteeng]

hi seems this is not working - peter

- - - Updated - - -

seems this is not working - peter

[peteeng]

seems to be back on line!! Peter

to build an oven the size you need is ambitious within itself. You need to SR the entire machine in one go, then you need someone with a big mill to finish machine it. We have a forge at out mens shed and its hard enough to get small things to red... Getting a machine up to red, do you have a small power station at home? Lots of research to do yet. Peter

[peteeng]

Hi Mogi - Here's a fellow with family built a plywood router to build an aluminium plane. Aluminium you see is no issue, but steel another beast. Also there's no reason a CNC can't be rivetted/bolted together, mine are... most planes are are, even biggest ones.. Peter

Thatcher CX4

[ardenum2]

to build an oven the size you need is ambitious within itself. You need to SR the entire machine in one go, then you need someone with a big mill to finish machine it. We have a forge at out mens shed and its hard enough to get small things to red... Getting a machine up to red, do you have a small power station at home? Lots of research to do yet. Peter

double column gantry would be easier to stress relief since its made from 4 long and relatively thin columns, the bed could be cast from non shrink grout, in the same way you'd cast a machine foundation.

Is this not good, i don't have good experience with milling steel, that is why what i saw looks good for my use case and what i'm thinking i can be doing with the machine.

That actually looks quite good for how the machine is build, you can hear the vibrations from the sides though.

[joeavaerage]

Hi,

Hi Joe, I'm not planing to cut steel all day long, I need the machine to do it with even lower feed, when I need it, to get me started so I maybe after this machine I build a better.

Been there, tried that....and failed. Either a machine is rigid enough to cut steel, or its not. My experience is that no matter how slow and gentle you cut as soon as you start making
steel chips the cutting forces will deflect your machine, a little and acceptably if the machine is stiff, or unacceptably if not. If it flexes it will vibrate and the cuts will be inaccurate.
It just what happens.

If you want to cut steel then make that choice early and then work towards it, its quite a mission and will result in a different machine altogether than what you have proposed.
Otherwise ditch the idea of cutting steel and proceed accordingly. You ideas, if well executed would be fine for aluminum,brass, wood and plastics,.....but always be poor for steel.

How do I know this?...I tried and failed! I learnt my lesson and have had another go, and much, much better, but still a long way from where I really want to be.

Craig

[peteeng]

Hi Mogi - I have looked a few of Langmuirs videos and the MR-1 seems to cut steel fine. I suggest you copy it and you'll be fine. I found the spindle info and they use an AC servo with a 1:2.67 belt (my guess). The servo runs at 3000rpm and they step it up so it can run to 8000rpm. You can get servos that run 0-6000rpm so no belts needed.

https://youtu.be/wUvRFBXM3JM

Now you can fabricate steel parts and finish machine without stress relief but its a gamble. If your good with oxy-acetylene I suggest braze welding works well or hard soldering both are good off the torch no TSR needed. So pick your poison and let the fun begin...

They use a leadscrew for the Z which I find interesting, use a ballscrew. Make it a bit stiffer then the MR-1 and it will work out. And do a better job at cable management...Peter

or save up and buy the LR-1 much simpler. I'm sure your going to pay $4500USD plus by the time you get it all together. If I built that basic machine here in OZ ($4500USD = $6700AUD) I would have no change from $6500AUD although the machine would be a little bigger...

[Titaniumboy]

Pete,

Do you have any idea how Langmuir made the MR-1 stiff enough to mill steel? Just by the looks of it you would not think it would be steel capable, but videos don’t lie. Have you taken a run at doing any modeling and FEA to tease apart how they are able to do steel?

Do you have any pictures of the 1:2.67 belt drive? I have a small mill similar to a Tormach and my spindle motor is a 3000 rpm AC servo. My pulley ratio is 1:1.5 so I can only get 4500 rpm at the spindle. I would love to get that pulley ratio higher to obtain higher spindle rpm’s, but was told that I couldn’t go much smaller on the diameter of the spindle pulley.

[peteeng]

Hi TiBoy - I'm guessing the ratio, as in the video the belt is closed up in a casing. Unless its a gearbox but unlikely to be a gearbox. The quote max speed is 8000rpm so maybe its a 5000rpm servo and smaller ratio? They do steel with light cuts. The vids look good but there are machine aspects that are poor. Like they run the ballscrews directly to the motors with no thrust bearing, so motor life is at risk specially if shock loaded. They use a leadscrew for the Z for some reason as well. The machine assembly videos are quite good so you see how they are doing various bits of the machine. I suppose if its in a commercial environment and gets a few 1000 hrs on it the owners will know. It does have some vibration as in one vid you can clearly see the chatter in the finish then it gets cleared in the spring cut...

In regard to the pulley diameter you pick a small pulley that works then get a big pulley to suit. 3:1 is common on various things. So you must have a space problem?

There are heaps of routers cutting steel videos and small hobby mills that confuse me. But they are cutting light and fast. Then you maybe able to cut a few things but it must take its toll on the machine. There are even ShapeOKO cutting steel! But not like real mills. I'll have to give my router a shot at steel... Peter

Some thoughts:
1) Its a small machine with a heavy damp base. Many small machines like this are quite light so can get excited easily
2) The gantry rails are across the gantry diagonals. This makes the bearings as far apart as possible plus, it puts the moment centre through the shear centre of the gantry which is quite good. ie this geometry has the least twisting tendency
3) the saddle is highly webbed across the cars so is very stiff. That's if the real saddle looks like the CAD saddles on their site. I just found a picture of the saddle and it looks like an aluminium casting. very solid.

https://youtu.be/-g4_5-AmjpM

[peteeng]

Hi Mogi - By the way if you make a mill and its uses coolant you will never do timber or porous stuff again. The lub and swarf goes everywhere and its a pain to try to keep clean,. So from that point on its a mill... If I cut aluminium on the router it takes me a long time to clean up the swarf and I find it 5m away from the machine over tables on the floor etc etc...

https://youtu.be/3DIkpsMBcIc

[peteeng]

Been reading the MR-1 forum and they are moving away from concrete to epoxy granite. They say this is because the EG cures faster so the build is faster. They say the concrete and EG machines behave equally... But I also expect from watching some clients assembly videos & what they say is the concrete was hard to do, need something that flows better around the steel embedment's. Peter

[peteeng]

heres a ShapeOKO cutting stainless steel. So its possible but slow... I'll have to set something up on Scoot to see what it does... Peter



https://youtu.be/QYkDKkfUuFs