Linear rail alignment

[Dom B]

Hi I am new here and am slowly learning how to build a cnc engraving/router machine.

I have come across a couple of questions as I try to figure things out.

All the projects that use linear rails of some knid do not seem to locate them by any accurate means like dowel pins. They all seem to be located by the threaded bolts that hold them down to the table. Knowing that a bolt shank is not accurate at all, how are people managing to stop the tables binding up all over the place by just aligning the rails on bolt shanks. If all the bolts were loose the rail would be able to move quite considerably. Why are people spending big bucks on rails with accuracy of 5micron for them only to bolt them to a board with no accurate location? How is this working?

I tried to build a machine based on bronze bushes running on silver steel. my thoughts were top let everything move and just slide the table to one end to get the spacing correct on that end then slide the table to the other end of the rails to get their spacing also correct whilst I tightened them up.

Total failure, when the bolts were finally tightened down they always jammed no matter what I did sliding the table. But its runs beautifully when the bolts holding the rails and blocks are loose.

So I tried again using pinned blocks and using a milling machine to drill/bore all the mounting holes to ensure accuracy, and this time it was much better.

But some of the guys withe the ball slides put the rail flapping about on bolts, so how do they go about aligning the rails for alignment and spacing without binding.

This is currently something of a mistery.

Many thanks for any replies.

Dom

[dertsap]

sounds as though the blocks on your gantry is toed in , which will cause binding and the rails not be able to align properly

[Dom B]

Originally Posted by dertsap sounds as though the blocks on your gantry is toed in , which will cause binding and the rails not be able to align properly

Yes, I would have thought the rails being out of a perfect parallel line would be the cause, but the question is how do people reliably align the rails, if they are not accurately located.

Imagining replacing the rail on a machine in service and spending most of the day tweaking the rails to run free, in a very hit and miss method. You just want to unbolt the old rail, bolt on the new rail, everything to work and go to the next job.

[mxpro32]

you start with one rail. preferably you have a straight referance edge to butt the rail against to ensure that it is straight. if you don't have a referance edge to butt the rail against you can use an indicator off of a straight edge as a reference. this step ensures that the first rail is parrallel to the intended line of movement and it ensures that any slight warpage in the rail from grinding is straightend out. after you have on rail straight, you use a sensitive dial indicator mounted to a carriage on the master rail to indicate off of the slave rail. slightly tighten the slave rail so that it can be bumped around. now run the indicator along the rail and bump it slightly untill the two are parallel. then carefully tighten the second rail. You can find the hiwin manual online that goes into more detail if you would like.

[BobWarfield]

Originally Posted by Dom B All the projects that use linear rails of some knid do not seem to locate them by any accurate means like dowel pins. They all seem to be located by the threaded bolts that hold them down to the table. Knowing that a bolt shank is not accurate at all, how are people managing to stop the tables binding up all over the place by just aligning the rails on bolt shanks. If all the bolts were loose the rail would be able to move quite considerably. Why are people spending big bucks on rails with accuracy of 5micron for them only to bolt them to a board with no accurate location? How is this working? But some of the guys withe the ball slides put the rail flapping about on bolts, so how do they go about aligning the rails for alignment and spacing without binding. This is currently something of a mistery. Many thanks for any replies. Dom

Dom, good catch!

Just mounting them with bolts is not best practices, though if done carefully, it can be made to work for awhile and in some situations. If you download the manufacturer's literature on these rails, you will learn that they are intended to be butted up against a machined edge on at least one side that locates them correctly.

For example, scroll down this page by Nook and look at the diagrams that show how the rails are mounted:

http://www.nookindustries.com/profil...ilGlossary.cfm

One side of the rail is butted against a precision milled/ground edge for accuracy and straightness.

I have not seen too many projects with them here that took account of that. Some folks don't even work very hard to make sure the surface they're mounted to is flat. JerryFlyGuy has a lengthy thread that talks about his trials in trying to line up the rails exactly and "pot" them into place that's pretty interesting.

BTW, I think they'll take a lot of abuse before the out and out bind, but they sure won't be accurate.

Best,

BW

[JerryFlyGuy]

Originally Posted by BobWarfield [teeny snip] JerryFlyGuy has a lengthy thread ... [teeny snip].. BW

What.. Celebrity status already!? naaaa.. I linked onto this thread yesterday, just to follow along and see where it went. Low and behold.. this AM I find my name in it.. ok... well my nickname.. anyway..

All joking aside.. You can find the thread that Bob is talking about here Layout of centerlines

I've finally got my rail all aligned and mostly level using this method. However.. I've come across one little glitch that will mean my moving the jig's and what not to solve. [more below]

One thing to note is simply that if you CAN use or get a machined edge and mounting surface, then do it! The reason I went to all this convulsion to get my rails straight, is because they are 20ft long and composed of two 10ft rail sections. Thats a bit of a stretch to get machined straight and level [ at least in my budget ]. My idea works very well, the only issue I've come across is the joint in the rails. I've not figured out yet how to make that joint true and flat. I'm working on it however.. I've got a couple idea's and we'll see what happens. Right now its close but not as perfected as I'd like.

Again... I'd recommend getting a machined edge.. if you can. You'll be glad you did. If you can get it machined in 1 setup it would be even better, however depending on your machine that might not be possible.

Jerry [ anyone else have a wife w/ one of these signs.. - except.. hers has the arrow pointing sideways... and she only walks on 1 side of me?? hrmm.. ]

[BobWarfield]

Aw Jerry, still no pix of happily potted rail or even the messy goary process in action.

Sniff.

Don't you love us any more on CNCZone?

BW

PS Jerry, I know you don't have it to do over, but I came across a fascinating PhD thesis that has been used to make at least one very successful machine. Lots of ideas for fabricating super high precision machines in novel and inexpensive ways. Google "principles of rapid machine design". It's quite a worthwhile read.

In particular, one of several concept machines they work through is a "tube miller". It's a purpose-built machine that can be used to mill rail locating edges in very long tubes.

There is also quite a lot of thought on 5-axis machine design, which I know is also up your alley.

[JerryFlyGuy]

Heck.. i'd send them... If I had them The other issue I keep getting to is the weather. It's getting cooler up here in the great white north [40miles over the US border into Canada ] and I don't have my garage heated.. [yet but it will be w/in the month] so I have to do the potting on a nice warm day/morning where it will stay warm enought to Exotherm and set.. that basically means on a weekend.. so its rush all week, then find an issue on the weekend.. figure out a fix, rush all week to impliment it.. find another issue.. needless to say.. its getting tiring.. frustrating and boring.. so I need to get this step @#$@^% done.. NOW..

Btw.. I love the Zone.. but you guy's??? um..how to put this "politically correct..".. um.. I'm married.. ??? but you know I think ya'll are great
I'd still be slopin' around w/ a twisted up pc of useless junk if I hadn't come on here and got all-ya'lls help.. ya-heear?

Thanks for pointing out that site.. I'll have to go looking fo' it later tonight..

Later gater..

Jerry [so you sayin' your wife doesn't have one of those shirts/signs? hrmm.. why am I so lucky??]

[BobWarfield]

Originally Posted by JerryFlyGuy Jerry [so you sayin' your wife doesn't have one of those shirts/signs? hrmm.. why am I so lucky??]

My wife has simply determined that this hobby is the cheapest of the many I've gotten into over the years, and therefore she supports it. It is the cheapest because I am so darned slow to actually get something done!

Best,

BW

[JerryFlyGuy]

Well.. maybe thats my problem.. this isn't one of my cheaper hobbies.. anyway... I kidd about my wife, in reality she's more understanding and open about my hair brained scheme's than I am about hers.. [ her wonderful, inteligent, smart and un-canny idea's that is.. ]

Thanks for pointing me out that book, I've printed it out and been reading it through.. a very interesting read for sure. I've got some idea's about my rail issue's but I'll leave that for another post later. Plus.. I don't want to hijack this thread so I'll put it back on the thread I started.. needless-to-say.. I'm going in circles.. [not that that's new or anything..]

Jerry

[Zumba]

Dom, two words for you.

Set screws.

[Walt@SGS.Inc]

I was working at Ingersol Milling (service call) in Roacford, Ill., they were aligning a rail very very long. They did it with a wire pulled very tight with weights attached to the ends and the weights hanging in oil to keep them from swinging. I remember they were looking at the wire with a very strong magnification. The object was to make the wire appear to be lined up in the eye piece. A tight piece of wire is within everyones budget. Now the piece to look at the wire might be another story....
As I remember, their wire was about 60 feet long.
Regards Walt.