Long Linear Motion, Best Choice?

It was brought to my attention that a normal rail, even at 12mm, is the worst for long parts [in my case, 1000mm].
So I thought I'd ask before getting too far financially into my build, what is the best for longer linear travels?

To my knowledge not even any of the professional machines have gone that large would take a month to print 24 hours a day. You would need a very specialized enclosure and an absolute rocksolid reliable machine. Even The largest professional FDM or FFM machine is 24" x 24" x 24" to my knowledge. What you're talking about is more CNC router size the thing with linear supported rails is the member underneath is for support not the actual rail. Here's a picture of my printer I don't know if you seen it before it's nowhere near a thousand millimeters span. Plus reprap electronics and motors will definitely not be enough to drive anything that big
https://www.facebook.com/photo.php?v=108919115976745&l=8836569305644303332

Edited 2 time(s). Last edit at 04/28/2014 10:20PM by cnc dick.

Go with a supported linear rod solution with open bearings.. or industry standard linear guides like what misumi makes if you can afford them.

A super cheap solution is to do what the diycnc folks did using angle iron and bearings.

You can also look into linear systems based on aluminum extrusions, like Makerslide or OpenBuilds V-slot.

I can't imagine what you'd be printing that would need 1m of bed in any direction, but at least with aluminum extrusion you can support long spans without interfering with the bearings.

If you're looking for precision you should define it so we know what to suggest.
That said 1000 mm, (40 inches) is a long span. You will need a supported rail, and a strong frame, start with Dia 1.0 inch.

If you are not looking for uber precision the following might work.

OpenRail looks promising, but unfortunately they don't provide blueprint specifications that a human can read, so consider other options if that is a deal breaker for you. It's any ones guess how straight they are.

OpenRail
[openbuildspartstore.com]

V-Slot look good, you can read the blueprint specification, granted they were drawn by an amateur, missing straightness tolerance, you have to zoom up on it to see that there is not much for an engineer to use, and it hardly make sense, but better than nothing?
I am seeing more printers using these as of late.

V-Slot 20 x 80mm
[openbuildspartstore.com]

I think the extruder knows the score and is saving OpenRail face in hopes of more business.... {Good guy Greg the extruder }
Unless there is a hidden room with openbuild blueprint specification that I have over looked, any one know?

Edited 1 time(s). Last edit at 04/29/2014 02:49AM by A2.

Unless you want to print very slowly, and with very large prints that is going to take a loooong time, I'd suggest building the machine with architecture similar to the Makerbot machines. Moving a large, heavy print bed in the Z-axis is easier than moving it in the Y axis, though the mechanics are more complicated. My machine has a 12"x12.5" printbed and besides being large, the bed is heavy (1/4" aluminum tooling plate with kapton heater attached). When trying to jerk it back and forth during a print, unless I decrease jerk and acceleration settings which slows printing down to a crawl, the belts stretch ever so slightly each time the Y axis starts/stops, leaving artifacts in the surface of the print. If you have to use a RepRap architecture, consider using a lead screw to move the Y axis.

All that said, I recommend you start going to local scarp yards that take in industrial machinery, disassemble it and sell the parts for scrap. You may be able to get linear motion parts very cheaply that way. That's where the stuff for my printer came from and where I'm going to get stuff for the next one. Right now I'm looking at an XY stage with 18"x26" travel on 1" guide rails with lead screws and DC servo motors.

[mark.rehorst.com]

Most scrappers don't bother trying to sell that stuff as-is. They take it apart, sort it by the type of metal, and sell it for scrap by the pound. If you buy an assembly from him, usually by the pound, it saves him the time and effort of taking it apart and sorting it. If it's a mix of aluminum and steel he'll probably charge you the aluminum scarp price for the whole thing.

Edited 1 time(s). Last edit at 04/29/2014 07:50AM by the_digital_dentist.

It's going to be a 3D printer/CNC in one, so by saying it's more CNC size more or less proves my purpose.
I can afford those massively priced rails from Misumi.. but I would really rather not spend that much if I can acheive the same thing for less.
I plan to use half the bed [about 800X350mm full size] for printing and the full bed for faster machinable materials in CNC.
My mechanics are going to be that of Makerbot/Solidoodle, I knew that before I really started designing it. Interchangeable heads are also in the plans, switching between direct and bowden [and CNC] for different material and mechanical requirements based on what the customer wants.
I've also got steel reinforced GT2 belts, so stretch will be less* so a factor.

Supported rails seems to be the most popular for strength at distance, but that's also a lot of added weight on the moving X carriage.

Quote
MrDoctorDIV
It's going to be a 3D printer/CNC in one, so by saying it's more CNC size more or less proves my purpose.
I can afford those massively priced rails from Misumi.. but I would really rather not spend that much if I can acheive the same thing for less.
I plan to use half the bed [about 800X350mm full size] for printing and the full bed for faster machinable materials in CNC.
My mechanics are going to be that of Makerbot/Solidoodle, I knew that before I really started designing it. Interchangeable heads are also in the plans, switching between direct and bowden [and CNC] for different material and mechanical requirements based on what the customer wants.
I've also got steel reinforced GT2 belts, so stretch will be less* so a factor.

Supported rails seems to be the most popular for strength at distance, but that's also a lot of added weight on the moving X carriage.

that's the main problem when you make a dual purpose machine you can never get optimal out of both. I myself and many people have attached extruders to routers and although they work it's not optimal. Obviously you're going to need some pretty strong motors as things are going to get heavy but the real fly in the ointment is the steps per millimeter needed for high-speed and high acceleration for good prints is not good for a router and vice versa. Have you looked in the CNC zone they have a lot of routers that have been built

Edited 2 time(s). Last edit at 04/29/2014 10:08AM by cnc dick.

While the speed isn't optimal at first. After I get things working on the printing side I will look into making a way to quickly change steps/mm. I originally thought transmission, but that's a fair amount of power loss and more area for inaccuracy. Right now I'm stuck on the idea of making motor mounts easily removable for changing the pulleys. It won't be optimal, but it'll work enough for my purposes. When I want the best of both and have the means and reasons to get them, I'll simply build two different machines [or four or five for other purposes in mind]
Thank you very much for all of the input!

If you go the route of using OpenBuild V-slot, you can use their build example (videos here)

With two Y-carriages carrying separate X-carriages, you can use different motor/pulley combinations to achieve different steps/mm. The carriage not in use can be stowed at the opposite end of the bed area. This eliminates the need to switch motors.

I just had a duh! moment. I'm going to change from being long horizontally to being high. A long Z-Axis on my already planned two threaded rods/four guide rails would be a heck of a lot easier to manage and keep precision and speeds.
I had already kicked that out of my mind on the thinking that prints are weakest in directions against the layers, but everything would be weaker if the carriage doesn't even move right and a million other reasons.
So basically back to the drawing board, making pretty much a repstrap Makerbot Z18, but larger. Unless I'm still missing something and that's even less feasible.
And the V-Slot looks promising, I'll definitely be looking to incorporate that into my design.

Edited 1 time(s). Last edit at 04/29/2014 12:15PM by MrDoctorDIV.

For up to 1000 mm strokes you better use linear bearings/cylindrical smooth rods (like these in 3D Printers). For 1000+ mm - you will need more support on the movement so you better go for ball carriages/guide rails and slotted linear bearings/supported rods.
For faster movement you will need belt driven system on the X and on the Y axis (if you're not designing a CNC router table).

If you're designing a CNC router table for wood/plastic/metal - you will need ball-screws or trapezoidal (ACME) screws. Be sure that the guided rails are with bigger diameter then the screws because of the cutting forces.
For up to 1000 mm strokes you can use screws (ball-screws or ACME) and for the 1000+ strokes - better go for pinions/gear racks system.
And my final advice - Don't put ball-screws on the Z axis (unless you have motors with brakes) because it will be pretty hard for the motor to hold it's position because of the minimum friction.
Best regards,
Bobby

P.S. In CNC machine designing (as in the 3D printers) there is only one rule - you can't make something good with less money. More money you invest - better results you will get.

Quote
BoByS
More money you invest - better results you will get.

Oh trust me, I don't get something simply because of price. Before I started designing I told myself skimping out on anything was not an option. The whole point of this build to make it as solid and reliable and powerful as needed and extra.
I've got ACME leadscrews for Z-axis. M8 size screws. My guide rails are 12mm [X4].

Hello.I read the thread and all the comments.And whille I understand that if I want to make large 3D printer for printing larger objects I need to reinforce it.My plan was to leave all the parts that I already have(stepper motors,extruder etd.) and replace only rods on construction and instead of it just place rails.Now my question is if I do this and I enlarge the construction so that I can print 2m (2000 mm) objects,will this work and will the quality be highly reduced?? After I make it bigger will there be problems with calibration or any other problem??

Edited 1 time(s). Last edit at 04/30/2014 03:12PM by pontifex500.

I haven't ventured too far I can most definitely tell you that you will have problems keeping the same motors and going that massive. That is a LOT of weight to move. What motors? And what size rails? That far I would say at LEAST 16/20mm rails just to get the job done. To get good quality, as these genius gents have stated, you will have to ditch the standard rails and get a better linear motion type.

Ok,so I understand that I need better rails,that is not the problem.What bothers me is that if I will construct such large printer I need stronger motors as you suggested.But if I put stronger motors I need more power this means that I need to reprogram the whole stuff and or change the entire electronics,if im correct???
Or could I somehow use the same electronics(motherboard) from small 3D printer without reprograming and messing around with it??
If I will work with ABS this means that I should make heated chamber for entire 3D printer is that correct?
I dont own 3D printer jet but I plan to construct one.I want to go with the small one for the time beeing but my goal is after I will know more about 3D printers to upgrade that printer somehow to print 2m large objects-ABS plastic-car bumbers and other plastic car parts.

A car bumper that would take over two weeks 24 hours a day I don't think that's a good idea. With the standard electronics using Nema 23's rated at about an amp and a half should be around 170 ounce inch or so quite a bit more than the nema 17 because of the mechanical advantage of the size of the motor. Still may not be powerful enough depends on how large and how heavy you make it. I would suggest you guys Cut the size down a little bit most professional FDM printers are not even as large as you people are talking about maximum that I might even think about would be 24" x 24" x 24"

Edited 4 time(s). Last edit at 04/30/2014 05:43PM by cnc dick.

I've lowered my X and Y axis down to less than 450mm. I just have a long Z axis, as far as I can tell, that shouldn't be a problem for two lead screws, 12mm from Misumi.

Quote
cnc dick
A car bumper that would take over two weeks 24 hours a day I don't think that's a good idea. With the standard electronics using Nema 23's rated at about an amp and a half should be around 170 ounce inch or so quite a bit more than the nema 17 because of the mechanical advantage of the size of the motor. Still may not be powerful enough depends on how large and how heavy you make it. I would suggest you guys Cut the size down a little bit most professional FDM printers are not even as large as you people are talking about maximum that I might even think about would be 24" x 24" x 24"

You say 2 weeks,what about whit larger extrusion head that extrudes 0,5mm of fillament or if i make 1mm head with Lathe Machine at work(note It is no problem for me to make any part from any metall material on my job)

I said over two weeks of course that's just a estimate but two weeks would be minimum. With Prints that big even with a heated chamber there is a lot of warping you would be better off making them in small sections gluing it together. But it's really not the right thing to do for what you want most masters anything that size are machined. Increasing your nozzle size and layer height is going to give you terrible resolution

And there is no solution for solving the warping problem???Even if I leave smaller nozzle and layer heihgt,make the heated chamber and I let it print more then 2 weeks???
(I cant make it with smaller parts and then gluing pieces together because bumpers would not be hard enough and elastic enough to pass minimum technical needs for car registration)

Quote
pontifex500
And there is no solution for solving the warping problem???Even if I leave smaller nozzle and layer heihgt,make the heated chamber and I let it print more then 2 weeks???
(I cant make it with smaller parts and then gluing pieces together because bumpers would not be hard enough and elastic enough to pass minimum technical needs for car registration)

FFM or FDM just is not the right method to make a car bumper it is just too large. Offhand I don't really know of any even professional machine that can do one. Maybe some of the largest jet type resin printers out there might. What you're talking is in the $400,000 range. Possibly machining a master plug Or buck out of wood or plastic and then using a vacuum forming machine

Edited 1 time(s). Last edit at 05/02/2014 03:17PM by cnc dick.

So...Those are all the problems IF I make the bumpers with ABS plastic(that is standard plastic for bumpers) and then I need heated chamber etc.
But what if I make car bumper with PLA plastic and make it more thicker???(As I understand I don't need heated chaber or heated bed when working with PLA)
What is plastic & elastic caracteristics of PLA plastic? Does it bend more or similar?
Would the printed bumper with PLA be strong and elastic enough to put it on the car that is used daily in trafic and would it be safe enough.(meaning it would not fall apart in the middle of driving or something like that XD )

Quote
cnc dick

Quote
pontifex500
Possibly machining a master plug Or buck out of wood or plastic and then using a vacuum forming machine

I know about that method,my friend works that way on his job everyday,but we want to avoid that method or any other method and use 3D printer for making parts.
As we calculated if we would make 3D printer that would print good parts we would save allot of time,effort and material.(Even if it would took 2 weeks or more to print such large parts)
I will buy smaller printer no matter what and I will try to play with it a little bit so any advice is more than wellcome.
If nothing else I will have a printer that will be able to print small plastic parts for car interior or similar.

Edited 1 time(s). Last edit at 05/02/2014 04:21PM by pontifex500.