[pminmo]

I just really wonder if there is not a lot of over focus on high cost linear rails and bearings for home machines. By that I mean something that isn't cutting parts constantly on a daily basis. For example, wouldn't a 1" x 3" 6063 aluminum as rails with skate bearings running on it last for most home machines? The aluminum is machineable with carbide tipped wood working machines, you can lap it with fine wet/dry sandpaper for more precision if you wanted. It is dimensionally stronger and more stable than MDF. Of course its more expensive, but at the price people are paying for Thompson et all. it's cheap. Granted if you want precision below .006, but then the MDF doesn't make sense either.

Phil

[balsaman]

IMHO you are correct, unless you can get a hold of some real cheap, in which case why not?

Eric

[ballendo]

Hello, right idea wrong material choice, IMO.

Yes, there is an over-focus on high grade rails. But...

No, aluminum is not a great 2nd choice , when used as you suggest. Why not?

Because steel rod is cheap,widely available, much stiffer, and most importantly, has a higher surface hardness. Which matters because the skatewheel bearings are hard. And they're exerting a lot more force than you think. Remember that the .866 diameter means that the contact point is pretty small. And you don't want to increase it by using a flat surface because that invites "bumps" from rolling over dust and swarf...

Hope this helps,

Ballendo

P.S. As fro cutting, you're only cutting a few pieces to length. hacksaw-able, even

[pminmo]

I would tend to agree but how much force are we talking here, IMHO not a lot on the size of a machine that is just cutting wood. Tempered aluminum alloy such as 6061-T6 while not as hard as steel, will take some abuse. I would concur on the flat surface as a collection point on top, but not so much on the sides and bottom. Just thinking out loud some. Glad you replied.

Originally posted by ballendo
Hello, right idea wrong material choice, IMO.

Yes, there is an over-focus on high grade rails. But...

No, aluminum is not a great 2nd choice , when used as you suggest. Why not?

Because steel rod is cheap,widely available, much stiffer, and most importantly, has a higher surface hardness. Which matters because the skatewheel bearings are hard. And they're exerting a lot more force than you think. Remember that the .866 diameter means that the contact point is pretty small. And you don't want to increase it by using a flat surface because that invites "bumps" from rolling over dust and swarf...

Hope this helps,

Ballendo

P.S. As fro cutting, you're only cutting a few pieces to length. hacksaw-able, even

[ballendo]

Hello,

More force than you think! Been there, done that, have the t-shirt...

Think about not only the weight, but the vibration, and the dynamic loads as the axis moves.

If you want to use aluminum rails, switch your bearing material to UHMW, or one of the exotic plastics designed for use with aluminum rails. But now we're back as expensive as what we were trying to avoid...

Of course, if you're talking 6IPM, and a 12x12 machine with a dremel, then okay. But you're not really saving anything, and you ARE giving some things "away",

Ballendo

Steel rod is cheap. pipe may be even less. Even angle iron's a better choice than 6061 or 6063 aluminum (when used with skate bearings); you'll just need to sand it first

[balsaman]

ballendo,

I have seen you on CAD_CAM... Good to have you here. Welcome.

Eric

[Konrad]

I just have a thought on this railing business, I'm not a router guy jet, still learning from you and have come to conclusion that you need the router table as light as possible for faster acceleration and less wear an tear on your threaded rods and nuts[I only guess], if that's the case, using lighter materials is the way to go.

So using aluminum extrusions, [not solid] is the way to go, and for wear, where your rollers run [scate rollers], you can simply tie a wearstrip, lets say about 1/16 x 3/4 or 1/2" wide springsteel.
I have seen it on precision machines.
All you do is, drill a hole at the end and screw it to the rail.
If you need high accuracy, then you could scrap it first.
Does that make sense!
Konrad

[jphagen]

I agree with Konrad, that it is better to use steel components for the wear surfaces and bearings.
Use aluminum for the structure where moving weight is a factor.

[steveald]

Hi,

I have to agree that some linear bearings are overkill on most homebuilt hobby machines. Accuracy and repeatability are a function of the entire system of components in your machine. Expensive linear bearings won't compensate for deficiencies elsewhere, so IMO money, design consideration and effort should be distributed evenly between all the components of a machine.
As far as weight goes, IMO the gantry should be as light as possible without sacrificing stiffness, but the table itself should be heavy to dampen vibration (assuming moving gantry style).
Back to the linear bearings, the Thomson bearings themselves aren't that expensive, it's the pillow blocks, case hardened shafts and shaft supports that are expensive, and with a little extra work you could use regular steel shafts and build the other parts. CAD files are available from most of the manufacturers, and you could end up with close to Thomson performance relatively cheap.

Steve

[limbo]

Its funny to see that indeed most people are thinking in the same direction. Its just as if there are a few standard designs you can choose from, and vary a little in implementation. (OK, me too ) Would be great to see some real out of the box designs. So pminmo, the floor is yours

BTW, somewhere I've seen a pic of grooved cams runnning on a steel wire, laid in a small nut or groove in a square tube. Neat, one wire on top, another on the bottom, bit of pre-tension et voila! I bet this could be simply made with hardware store materials?

greetings,
John

[twombo]

Don't forget HIWIN rails! I bought 20 mm lgh HIWINs for the X and Y axis' and 15 mm THKs for the Z my 36 x 24 router for roughly $1100 from bob campbell. The cost seems high but the quality is real good and I believe I have achieved that good balance that steveald speaks of:
1. Good solid extruded aluminum structure,
2.precision slides
3.minimal backlash in the drive screws using 3/4" zero backlash cam action Supernuts on the X and y axis' (z is 5/8").

Admittedly I have spent nearly 3500 dollars on the machine but it should provide steady, reliable, service since each component is quite robust.

Your goals , I guess, dictate your investment tolerance. I want a rugged reliable machine that will hold up in a light production environment with a failrly high degree of accuracy. I hope i madethe right choices.

The great discussions on this site are very valuable in this project, Thanks.

[pminmo]

TWOMBO, You are a fortunate person that can justify/spend $1100 for rails and bearings and $3500 on your machine. I guess I believe that's a little more than most hobbist want to spend. IMHO. But thanks for your input. Opinions of all varieties are welcome.

[Hobbiest]

There are a lot of good linear motion products made by www.igus.com, that are polymer construction. It is inexpensive, slides extremely well, and is extremely resistant to dust infiltration as well. As far as using recirculating ball bearingson less expensive steel components, a problem may arise with wear. CRS rod is softer than the balls in the bearings, and so the rod itself will wear. This is why all of these assemblies use hardened shaft. You want the stationary parts (ways) to be harder than the moving parts (bearings). Simple bearings, bearings that don't have any moving parts, and work off of low co-efficient of friction, will run all right on softer shafting however. The bearings from Igus run on anodized aluminum!
As far as gantry weight, if the bearings are smooth enough, the gantry glides on them, with minimal force output. Sure you can make the gantry too heavy, but it would have to be built like an aircraft carrier before it made a real difference. After your moving parts are assembled, and easy test is to tilt the table, and see at which angle the parts start to slide. Anything over 45 deg is not optimal, but could still be worked with larger motors, as long as everything rides smoothly (ie, no binding, just a little tougher to move).

[Jan]

Something I haven't seen mentioned is moglice
for use as bearing material.
I have just recieved an info pack about this and am amazed at the things it's used for.
At $38 + for 100 grams, doesn't seem cost effective until you realize that, that will make bearings with an id of .5, od of .750, and about twelve inches in length, don't quote those figures, as it's from memory, and I've slept since talking with the sales rep.
http://www.moglice.com/ for more info
jerry

[pack rat]

I constantly overbuild due to my lock of knowledge and proper tools. I dont know what I am going to end up with so I create a compensation in my design to overcome that. Everything that I designed is adjustable, axis are adjustable, leadscrews are adjustable, squareness is adjustable etc. That is a lot of variables and potential for errors I try to minimize. Linear slides eliminates slew of problems and time building it.If I had monies I could buy completed one, if I had more money I could hire someone to run it but then what I would do with myself? I enjoy building things. To do the best I can with what I can afford. Too many details bogs me down, some people cast their own bearing housing and gantry sides, now that is talent.

[pminmo]

Limbo,
Here are my thoughts. I've started a small moving gantry router, as I've gone along, I think I had the wrong approach for me. I've wanted to keep the size to a 2' x 4' footprint, minimize cost, but shoot for accuracies/repeatability to .008 or so. I really think it should be a fixed gantry router same foot print, with the long axis getting the accurate linear rail parts. To be able to cut longer narow pieces the ends would have a cutout for pieces to move through. These a rought thoughts: this would be the front looking on. i.e. length 4' http://webpages.charter.net/pminmo/frontview.png
This would be an endview 2' in width:
http://webpages.charter.net/pminmo/fullendview.png
This would be an endview of the assembly on the lond axsis:
http://webpages.charter.net/pminmo/endview.png
This is layed out as 1" x 3" .125 6063 rectangular tubing, with a piece of 1/4" x 1 1/2" x 48" 6061-T6 bar. Both of which machine great with carbide cutters on woodworking equipment. The red is roller bearings, the dark grey would be 5/16" rod which would mount flat to plate for the assembly. Green is acetel (Delrin)Lateral bearing would be acetel, with the adjustment being acetel with threads cut on it. I'm thinking out loud here. By mounting the bearings on rod that would be held flush to the plate of the moving assembly, by buying ground tooling aluminum plate you would have a true reference. For moving dust/cuttings off the bearing rail, you could use a piece of acetel formed in a v similar to a cow catcher on a steam engine. Just thoughts.

Phil

[Hobbiest]

Phil, Your ideas sound great, but I would caution you against using aliminum plate for the way. It simply is not hard enough to take the wear. Bearings ride very nicely on cold rolled steel, and with a little hand scraping, cold rolled steel will be extremely accurate. I think I would also consider using two pieces of delrin to form the adjustable gibs, instead of one, but that is probably just personal paranoia! If you really want to use aluminum for the ways, why not make all of your sliding parts out of delrin and eliminate the bearings? I am building a small mdf machine right now that uses pieces of cutting board riding on aluminum channel for the ways, and even with the increased friction of nylon verses delrin, everything seems to move niceley! If you don't already know what gibs and ways and scraping and all that stuff is, go to the library and get a book on machine tools. Get an old one...'50's or older...they seem to have the best stuff for low tech designs.

[limbo]

pminmo,

Nice setup! Now thats something different ! About the frontview: Is that the leadscrew above the rails? I've found that it is better to have the leadscrews as close as possible to the work-area. Imagine that your block with the bearings and delrin can "rotate" a bit due to flex. If the block is held rigid by the nut above the rails, the tool (being far below the rails) can move sideways quite a lot.

I think Hobbiest is right about the aluminium running in. I havent tried it personally but so many people have warned me beforehand about this, they can't be all wrong. Maybe some clear heatshrink tube around the bearings? Very strong. Don't know if it would distribute evenly though...

greetings,
John

[ballendo]

[QUOTE]Originally posted by steveald
[and with a little extra work you could use regular steel shafts and build the other parts.

Hello,

There's a reason the thomson shafts are hardened...

Using regular steel you will get grooves very quickly, and any preload/precision will be gone. I do have some thomson 5/8 ball bushings which use delrin balls; and these will work with regular steel shafting. But the load rating is pretty low(compression strength of the delrin)

Even shafting that "looks" good will groove. I learned this the hard way about a decade ago. I got a WHOLE BUNCH of printer rods surplus. Chrome shiny, and made by a company called treffers precision grinding. I bought a WHOLE BUNCH of linear bearings to go with them, only to find the results--grooving-- i mentioned above. Tried 303 stainless steel next. No dice.

A thomson style bearing will be really hard on soft shaft materials, as the design directs the force to a very small contact point, the tangency between the balls and the shaft radius. This is why the THK tracks with their greater contact area have higher load ratings.

If you're gonna use cheap shafts, you need to change to a plane type bearing; whether it's bronze, ZA12, or pacific or simplicity, or IGUS, or... (all the other hi-tech plastics)

Hope this helps,

Ballendo

P.S. Thank you for the welcome, balsaman.

[pminmo]

Limbo, yes that was were I originally thought of placement for the leadscrew, but your right it would probably work better below the beam.

Phil

Originally posted by limbo
pminmo,

Nice setup! Now thats something different ! About the frontview: Is that the leadscrew above the rails? I've found that it is better to have the leadscrews as close as possible to the work-area. Imagine that your block with the bearings and delrin can "rotate" a bit due to flex. If the block is held rigid by the nut above the rails, the tool (being far below the rails) can move sideways quite a lot.

I think Hobbiest is right about the aluminium running in. I havent tried it personally but so many people have warned me beforehand about this, they can't be all wrong. Maybe some clear heatshrink tube around the bearings? Very strong. Don't know if it would distribute evenly though...

greetings,
John