View Full Version : question about thk rail mounting


mwp
03-06-2005, 06:47 PM
Hello,

I am in the process of designing a gantry-type wood/aluminum/steel milling machine. I intend to build the major structural components out of welded steel sub-assemblies bolted together. The overall machine footprint is TBD, but I do have 6X THK HSR 25 40" rails + various carriages to use in its construction. At this point, I'm contemplating making the machine able to take a 48"X48" workpiece and have 18" under the gantry (yes, I will need to join the rails but I have done this before).

My question is in regard to how other builders have ensured that the x-axis rails--the ones that the gantry rides on-- are coplanar with each other. I see a couple problems with this. First, the mounting surfaces for the rails need to be made flat (really flat if you want good accuracy since the rails won't be as stiff as the supports and will conform to them-flat or not). I can probably accomplish this with my mill with some difficulty in several steps as it only has ~18" travel on the long axis. Next, after mounting the master rail against a straightedge of some sort and to the first support, the slave rail/support needs to be squared relative to it. To me, this seems like the toughest part. The easy way that I see to do this would be to bolt/weld the whole assembly together roughly square and have the top surface machined flat. This would probably cost quite a bit although I might be able to get it done as a favor...

I have read through everything that I could find here and have not found all that much regarding this topic. One interesting method presented was to shim the rails into flatness/squareness (using a machinist’s level for reference) and bedding them in JB weld. It seems like this would sacrifice some rigidity of the bed due to the flexibility of the JB weld. Might be the way to go though.

Any ideas?

chuckknigh
03-06-2005, 11:58 PM
http://www.5bears.com/cnc.htm

This site is a journal, kept while a man built the best and most highly precise homebuilt mill I've ever heard of. He deals with, among many other things, exactly what you're asking.

Fair warning...his level of perfectionism is addictive. And, scary!

-- Chuck Knight

Hobbiest
03-07-2005, 11:55 AM
It seems like this would sacrifice some rigidity of the bed due to the flexibility of the JB weld.

JB weld is commonly used to repair internal engine parts at the track, on racing motorcycles and cars. I Doubt it is that flexible to make a difference with the THK.

ViperTX
03-07-2005, 12:13 PM
mwp, check out the Thompson site, when using their shafting they recommend laying the shaft on a bed of epoxy that is on the support rail. You can use the same technique to linear rails...unless you have access to a huge mill where you can machine the surface where the rail will be mounted.

The next problem will be in adjusting the linear rails so that they are co-planar...fix one rail, and make the 2nd adjustable....I would level the 2nd rail using a rigid straight edge with a machinist level.....unless you have a perfectly straight - straight edge mark the spot where you place the level and where the straight edge meets the linear rail's trucks.

JavaDog
03-07-2005, 12:57 PM
http://www.5bears.com/cnc.htm

This site is a journal, kept while a man built the best and most highly precise homebuilt mill I've ever heard of. He deals with, among many other things, exactly what you're asking.

Fair warning...his level of perfectionism is addictive. And, scary!

-- Chuck Knight

Holy Crap! You weren't kidding! That guy is Dedicated, what a gorgeous machine.

WoodSnarfer
03-07-2005, 01:15 PM
Yep, "that guy" knows his stuff!

If I'm not mistaken, that's our buddy Swede :)

-Chris

Gordon Kirkwood
03-07-2005, 05:33 PM
Hello,

I am in the process of designing a gantry-type wood/aluminum/steel milling machine. I intend to build the major structural components out of welded steel sub-assemblies bolted together. The overall machine footprint is TBD, but I do have 6X THK HSR 25 40" rails + various carriages to use in its construction. At this point, I'm contemplating making the machine able to take a 48"X48" workpiece and have 18" under the gantry (yes, I will need to join the rails but I have done this before).

My question is in regard to how other builders have ensured that the x-axis rails--the ones that the gantry rides on-- are coplanar with each other. I see a couple problems with this. First, the mounting surfaces for the rails need to be made flat (really flat if you want good accuracy since the rails won't be as stiff as the supports and will conform to them-flat or not). I can probably accomplish this with my mill with some difficulty in several steps as it only has ~18" travel on the long axis. Next, after mounting the master rail against a straightedge of some sort and to the first support, the slave rail/support needs to be squared relative to it. To me, this seems like the toughest part. The easy way that I see to do this would be to bolt/weld the whole assembly together roughly square and have the top surface machined flat. This would probably cost quite a bit although I might be able to get it done as a favor...

I have read through everything that I could find here and have not found all that much regarding this topic. One interesting method presented was to shim the rails into flatness/squareness (using a machinist’s level for reference) and bedding them in JB weld. It seems like this would sacrifice some rigidity of the bed due to the flexibility of the JB weld. Might be the way to go though.

Any ideas?
I have some input for making the rails parallel once you've got them coplanar. Use the slides: Fix one end of the rails with a single bolt so their centers are some distance X apart, and make a crude but solid slide, perhaps out of a slab of aluminum, which you attach the two sides' slide trucks to so their centers are also X apart. (Bolt the trucks to the slide with the slide all the way towards the already fixed end of the rails). then move the slide to the far end. The far end of the rails will now be the same distance apart. the linear rails themselves are your best tool for transferring measurements from one part to the other...

chuckknigh
03-07-2005, 11:28 PM
Wouldn't that just ensure parallelism, but not guarantee that they were coplanar?

-- Chuck Knight

whiteriver
03-08-2005, 07:13 PM
How to make them Co planer without a big ass mill. I have 50' of thk rails to mount on my new machine i'm building and need to do it right the first time. Best method would be to haul the different axis's down to the local big ass mill shop and have it done. The base where the x axis slides will be mounted will be 3000+lbs. Not a easy or cheap task. Gotta be a better way.. How would one do this in a small shop with limited tools. Got a good dial indicator and caliper need a machinist level. What else do I need. Don't have parallels or guage blocks. :drowning: Any ideas. Be specfic please.

Donny

www.whiterivermfg.com

ger21
03-08-2005, 08:25 PM
This might help:
http://www.lmwatts.com/straightedge.html

WoodSnarfer
03-09-2005, 08:14 AM
I think the trick on a large machine is to design it so the rails are on a substrate that is adjustable (using shims, as an example).

How do the 'big' CNC machine builders do this? Do they rely on precision machining to build the frame of the machine, or do they have an adjustment scheme built into their design?

miljnor
03-10-2005, 12:10 AM
every Professional machine I have looked in just had the guide way area machined flat with no "obviouse way to adjust them (planer).