Rails....How to Choose?

[Fish4Fun]

OK, I "jumped in" with one of John's (MicroCrave's) "A4" machines after 20 years of "standing on the sidelines." WOW, very cool to FINALLY be in the CNC world. Of course, once you have a machine, you want a "bigger", "faster", "stronger", "more accurate" machine, and so you start looking, thinking, designing. This immediately brings you (me) to "rails". Obviously if you currently have "unsupported rails", you are curious about "supported rails".

I am familiar with typical mill bed & ways ("Square" & "Dovetail"), but have little confidence I could accurately machine them from raw stock (doh!), so this leaves me looking at various production rails & blocks. VXB and E-Bay jump out as "easy access" suppliers. As far as e-bay goes, there are three or four sellers who seem to "be in the business" of supplying rails & blocks (from China). "SBRxx" seems to be a very common format (ie, SBR12, SBR16, SBR20, SBR25 and SBR30). The "xx" appears to be the effective diameter of the "round" portion of the rail in mm. Obviously the larger the diameter and the "thicker" the support, the more "rigid" the rail system will be. Obviously "supported" rails are MUCH more rigid than simple round rails. Of course there are also "v-bearing blocks" that simply use CRS for the rails.

So, here is where I am: CONFUSED! Obviously nothing short of a solid/massive bed/way system is going to be rigid enough for milling steel. If I want to mill steel I need to simply purchase a Knee Mill and convert it; got that. If I want to cut wood and perhaps aluminum, then I need to consider feed rate with respect to (wrt) deflection, and obviously as the size of the machine increases, the rail/block requirements become more demanding. What I am absolutely failing to grasp is how to find a "happy medium" between price, size and performance.

I am not interested in avoiding calculations wrt deflection/vibration etc, but I want some general guidance based on experience for appropriate first level "guestimates" on sizing of rails wrt travel/load. I understand that the entire machine design ultimately dictates the success/failure, not the choice of rail/block size, but, like everything in engineering, there is a min/max; that is, an SBR12 may do 98% of what an SBR30 will do given a particular length/load/superstructure, but in a different span/load/support, the SBR30 may do 85% better than an SBR12. What I would hope to get from this thread is some advice about Span/Load/Support for various travels with various types of supported rails.

For instance, if I want a Y travel of 36 inches and a distance between rail centers of 12 inches in a fixed gantry machine, should I consider SBR12, SBR16, SBR20, SBR25 or SBR30? Should I consider "V-Bearings" and CRS? Obviously a lot depends on feed and feed related forces, but where do I start? If I change from 36 inches of travel to 48 inches of travel, what changes in my consideration of rails and blocks? If I decrease travel to 18 inches, what considerations should I take into account?

I am wanting to design a machine, and I don't want to "simply use bigger" because, "I don't want to use too small". I want to approach a design with some idea of how to select the proper components. If I can narrow the design choices to a "small range general standards" then I can begin to study the "math" w/o trying to crunch a wide range of numbers.

Finally, I would LOVE some comments about various rail/block suppliers wrt quality. If you ordered some rails/blocks from a supplier and you found them to be excellent, PLEASE brag about them; on the flip-side, if you ordered some rails/blocks from a supplier and were disappointed in them PLEASE BASH them!

I expect there to be some controversy on this topic, and hope that everyone who has an opinion about rails and blocks chips in. I am NOT looking for: "If you plan a 23.25 inch travel, then you should use SBR16 rails spaced 8.375 inches apart." What I am hoping for is more like, "If you are planning 18-24 inch travel, generally an SBR16 rail will be sufficient; if you are planning 30-42 inch travel, you should consider SBR20 or SBR25." If the frame is paramount to the rail, then suggestions about appropriate frames for the various size rails should be paramount to rail size, and some mention of frame design should be given.

I hope this isn't too broad a topic, but I think a great deal of effort goes into the selection of rails in "good designs"; generally build threads simply state the "decision" without much discussion of how the choice was made. What I hope for are some general guidelines on how those decisions are made.

Thanks In Advance for any and All thoughts,

Fish

[ger21]

The loads that the rails are carrying is a much bigger factor in the size than the length.
Personally, I'm not a big fan of the round shaft bearings. But if they are supported, than most likely the 20mm will work fine. But if you want to do the calculation, get some info from the manufacturers. You won't be able to find info from Chinese manufacturers, but you can find comparable stuff from NB or Thompson.
http://www.nbcorporation.com/technical/technical.html

[cyclestart]

SBR/TBR open blocks (supported rail) are strongest when the force is applied in a direction that doesn't tend to spread the blocks. A 20mm SBR block handles a static load of 375 lbs according to Glacern but you wouldn't want to vertically suspend a 375 lb load from that block. What are the specs for a side or torsional force ? I don't know The high end suppliers can no doubt provide those specs and they probably apply in a broad sense to the cheaper products. One router design may place the blocks in a different orientation than the next guys design and surely that's part of the equation. When buying China direct the price difference isn't huge for larger diameter so I just went for size large.

Edit/ Also the spacing of the blocks along a rail is a bigger factor than the length of the rail. If the machine's frame has flex the rails won't save you.

[louieatienza]

OK, I "jumped in" with one of John's (MicroCrave's) "A4" machines after 20 years of "standing on the sidelines." WOW, very cool to FINALLY be in the CNC world. Of course, once you have a machine, you want a "bigger", "faster", "stronger", "more accurate" machine, and so you start looking, thinking, designing. This immediately brings you (me) to "rails". Obviously if you currently have "unsupported rails", you are curious about "supported rails".

I am familiar with typical mill bed & ways ("Square" & "Dovetail"), but have little confidence I could accurately machine them from raw stock (doh!), so this leaves me looking at various production rails & blocks. VXB and E-Bay jump out as "easy access" suppliers. As far as e-bay goes, there are three or four sellers who seem to "be in the business" of supplying rails & blocks (from China). "SBRxx" seems to be a very common format (ie, SBR12, SBR16, SBR20, SBR25 and SBR30). The "xx" appears to be the effective diameter of the "round" portion of the rail in mm. Obviously the larger the diameter and the "thicker" the support, the more "rigid" the rail system will be. Obviously "supported" rails are MUCH more rigid than simple round rails. Of course there are also "v-bearing blocks" that simply use CRS for the rails.

So, here is where I am: CONFUSED! Obviously nothing short of a solid/massive bed/way system is going to be rigid enough for milling steel. If I want to mill steel I need to simply purchase a Knee Mill and convert it; got that. If I want to cut wood and perhaps aluminum, then I need to consider feed rate with respect to (wrt) deflection, and obviously as the size of the machine increases, the rail/block requirements become more demanding. What I am absolutely failing to grasp is how to find a "happy medium" between price, size and performance.

I am not interested in avoiding calculations wrt deflection/vibration etc, but I want some general guidance based on experience for appropriate first level "guestimates" on sizing of rails wrt travel/load. I understand that the entire machine design ultimately dictates the success/failure, not the choice of rail/block size, but, like everything in engineering, there is a min/max; that is, an SBR12 may do 98% of what an SBR30 will do given a particular length/load/superstructure, but in a different span/load/support, the SBR30 may do 85% better than an SBR12. What I would hope to get from this thread is some advice about Span/Load/Support for various travels with various types of supported rails.

For instance, if I want a Y travel of 36 inches and a distance between rail centers of 12 inches in a fixed gantry machine, should I consider SBR12, SBR16, SBR20, SBR25 or SBR30? Should I consider "V-Bearings" and CRS? Obviously a lot depends on feed and feed related forces, but where do I start? If I change from 36 inches of travel to 48 inches of travel, what changes in my consideration of rails and blocks? If I decrease travel to 18 inches, what considerations should I take into account?

I am wanting to design a machine, and I don't want to "simply use bigger" because, "I don't want to use too small". I want to approach a design with some idea of how to select the proper components. If I can narrow the design choices to a "small range general standards" then I can begin to study the "math" w/o trying to crunch a wide range of numbers.

Finally, I would LOVE some comments about various rail/block suppliers wrt quality. If you ordered some rails/blocks from a supplier and you found them to be excellent, PLEASE brag about them; on the flip-side, if you ordered some rails/blocks from a supplier and were disappointed in them PLEASE BASH them!

I expect there to be some controversy on this topic, and hope that everyone who has an opinion about rails and blocks chips in. I am NOT looking for: "If you plan a 23.25 inch travel, then you should use SBR16 rails spaced 8.375 inches apart." What I am hoping for is more like, "If you are planning 18-24 inch travel, generally an SBR16 rail will be sufficient; if you are planning 30-42 inch travel, you should consider SBR20 or SBR25." If the frame is paramount to the rail, then suggestions about appropriate frames for the various size rails should be paramount to rail size, and some mention of frame design should be given.

I hope this isn't too broad a topic, but I think a great deal of effort goes into the selection of rails in "good designs"; generally build threads simply state the "decision" without much discussion of how the choice was made. What I hope for are some general guidelines on how those decisions are made.

Thanks In Advance for any and All thoughts,

Fish

You're a little vague as to what you eventually want to mill... What I will say is that if you have the desire to mill some metal in a larger format, then you might want to consider surplus or used profile rail as opposed to round rail; they're far more ridgid at the same size as round rail, though they might take a bit more work to mount. The round rail probably easier to install, and less expensive.

The frame is probably most important, as it is what the rail is attached to. I don't think there's a hard and fast rule as to what rail size to use for axis length, it all depends on your application. You could probably use a 12 or 16 series round rail on a plasma or laser table that's 4 feet long, but maybe that's the small end for a wood router. Profile rail could probably handle 5-10 times the force of round rail at the same size.

IMO there's no such thing as a happy medium, you should pick components that will do the job you need to with a little bit of headroom and not more, because otherwise you're wasting money. If you use a smaller rail, it'll be cheaper, but won't be as ridgid, therefore you're taking shallower and slower cuts. If you use too large a stepper, you've turned your machine into the most expensive electric heater. The larger the rail, generally, the more expensive. Larger rails/motors/screws will weigh more. I think the only exception to this is, for example you require 20 series rail, but can get 25 series rail cheaper. If you can work it into the design that's fine.

I find with profile rail, most people look and bid for 15 and 20 series rail, but not as much with 25 and 30 series. Maybe a bit overkill, but for the right price it might be the ticket.

[LazyMan]

Its important to consider what you are going to be mounting the rails on to and the how rigid your machine frame is. I use Chinese 20mm supported round rails on my extrusion framed machine with very nice results. Profile rails are certainly more capable for the same size but they really need to be mounting to a sturdy and accurate machine frame to take full advantage of them. However, if building a machine with single drive x axis and a long gantry, the rigidity of the profile rails and blocks will definitely shine on the x.

[jsheerin]

There are a few things going on with rails, machines and what you can cut (and how well). You need a machine with a certain level of stiffness between the part (or what the part is mounted to) and the tool. You also need some damping. The stiffness of the machine is made up of all the components - the frame, the bearings, the drive screws (or other linear motion system), the spindle, the vise, etc. Consider all these things like a bunch of springs connected end to end between the part and the tool. If one of the pieces is really flexible, it doesn't matter if the rest are stiffest springs every made. The entire chain will still be flexible. To relate this back to rails, rails and bearings have a certain stiffness. It usually is different in different directions, although that depends on the design. Some companies publish the stiffness values and some don't, but if you want to know, you can typically ask and get the values from the support department. THK HSR values that I got from THK are posted in my mill build thread, for example.

Then you have the load rating of the rail. If you pick a tiny bearing and then use it slide dump trucks (27 tons or so) around on, the bearings will wear out quickly. If you look in the technical catalogs for most bearings, you can find examples of how to do the calculations to figure out how long certain bearings will last under certain loads and with the loads applied to various bearing spacings.

A few other random thoughts - in terms of deflection, your frame is what you should be designing to have minimal deflection. The bearings and rails get mounted to the frame. If your rails are contributing a large amount of the stiffness to the structure, the structure is probably too light.

Larger rails do not necessarily have larger stiffness values. I have some graphs in my mill thread showing stiffness values for several sizes of THK rails that I put together. As they get larger they get a bit stiffer, but not a lot. Mostly the load capacity goes up. Preloading the bearings is what makes the stiffness go up. This is done (as far as I can tell) by putting oversize balls in the bearings, like with ball screws. I asked THK if they could provide any guidance on doing this and they declined to comment. I did some calculations in my mill thread about how much bigger the balls might have to get, and as I remember it was a pretty small amount. In other words, you'd have to be capable of some pretty precise measuring. You can buy preloaded bearings and rails though.

[Fish4Fun]

ger21, cyclestart, louieatienza, lazyman, jsheerin,

Thank you all for your input! I notice no one yet has jumped on the "v-bearings" with CRS rails. From what I can tell this option seems most popular in larger machines where wood products are the primary target stock for a router, and in machines for plasma cutters. While this style rail system intrigues me, I see many potential pitfalls.

You're a little vague as to what you eventually want to mill... What I will say is that if you have the desire to mill some metal in a larger format, then you might want to consider surplus or used profile rail as opposed to round rail; they're far more ridgid at the same size as round rail, though they might take a bit more work to mount. The round rail probably easier to install, and less expensive.

I would like the capability of milling aluminum with shallow cuts, various plastics and wood. I really do not have a definite size machine in mind, nor do I have any specific machine requirements yet. Intuitively I am worried about the supported round rails wrt rigidity, resonance and vibration, but this worry is based 100% on speculation, not experience or engineering. I am somewhat surprised that you all seem to agree that the length of the axis is not an important consideration in the rail sizing, though your collective reasoning makes perfect sense. Obviously in my mind I was thinking/assuming that a larger axis implied larger loads and this is not nessicarily the case.

I have been reading, reading and doing some more reading, LOL! I guess what I really need to do is build a bigger shop so I can build several machines of varying sizes and gain some experience with what works best in different applications, but sadly that is a few years down the road, so, for now, I guess I have to be satisfied about reading, designing and asking the occasional question. Thank you all for helping point me in the right direction!

Fish

[jsheerin]

Just to emphasize why you don't need bigger rails for a longer span, what you're probably intuitively thinking is that a beam needs to be bigger so that it won't sag over a longer span. And you're right - the frame member needs to be bigger when it spans a longer distance in order to not sag more under a given weight or deflect more under a given load. But the frame is what should get bigger - not the rails. If the load you need to work under goes up (like you decide to change from milling foam to milling steel, or you change from the beam carrying a wood router to a 500 pound milling head), then you need to increase both the size of the frame and probably the rails.