I've been looking at options for railing for steel v wheels such as rm2zz or rm2-2rs. I realize aluminum rails are not the way to go, but most case hardened steel rails are super expensive. PBC linear seems to be the cheapest option, but still expensive. They have rail on amazon and zoro. I've gotten a quote from them about their steel rail IVTAAB series, but that is around $3 per inch. Takes out a lot of the hard work of alignment, though. So, has anyone used v-slot with steel wheels, and adding a 0.75 or 0.5o inch steel angle between the wheel and v-slot? Or would it be slightly better to use cold-rolled square mounted at a 45degree angle in the slot to make a rail (closer to the bottom left photo, but with square rather than angle)? I've heard cold rolled would have slightly tighter tolerances (fyi, I'm thinking on order of 0.01mm precision as a goal rather than 0.0001mm, at least on my machine). Could either option work via jbweld? Or perhaps some properly placed screws? I could see gravity holding in the top I don't think I've seen too many options. The angle iron idea would be similar to the photo second from the left on the bottom. The v-slot would then be underneath, and the same thing would repeat on the bottom. The reason I ask, is because two 1000mm and one 1500mm 2060 vslot pieces, along with the six steel angles, would still be around half the cost of several other options I've looked into. In some cases, 10x less... Would something like this "work" but be an order of magnitude away from a proper linear rail and cause lots of problems? This is a hobby, so don't want to go overboard on one part of the system (nema23 motors, thinking 9mm gt2 belts). Just for more information, I already have two 1000mm and a 1500mm 6090 extrusion I picked up a while ago, so I'm looking for something to bolt to those. I'm not excited about drilling directly into the extrusion, so a linear rail like PBC sells would still need another piece to offset the rail (more cost). My thought was butting vslot to the 6090 is easier to do. Thanks for any thoughts.
JBWelding 1/4" square bar on the diagonal would probably be the cheapest. I can't say as anyone has tried it though.
I don't think you're going to see 0.01mm tolerance from the extrusion or over the length of steel. When combined the tolerances may stack in some spots. Ricks idea with JB weld or an epoxy seems like a good one. You can float the steel and make adjustments before it sets. Cncrouterparts has a system called v-con that uses bar stock and specially made clamp brackets. It gets costly though. http://www.cncrouterparts.com/v-con-linear-motion-parts-c-47_50.html You can find hardened v-rail. That is somewhere around $1 an inch.
Thanks for the responses. I should have clarified, the 0.01mm tolerance was a random guess to get a sense of perspective (cnczone people sometimes like to unknowingly compare $1000 machines to $20,000 machines). I'm not even sure of the tolerances of the vslot itself, but I'm guessing it is no where near the tolerances for most of the linear stuff I'm looking at. Precision isn't cheap. The cheapest v-rail I can find is the 0.437in tall stuff from zoro/grainger/amazon from PBC. It is about $22 for a 36in piece, but you'd need two for each carriage. The 48in was $31. But this also requires quite a bit of alignment, drilling, screwing... I've only had regular extrusion at this point, so always eliminated this as an option. The v-slot at least makes it plausible. That being said, I've seen someone do similar to Rick 2.0's idea, but had screws to hold it down. The screw heads just missed the v-wheel. Not sure of the type of head. 20mm extrusion doesn't really provide a lot of room to work with. I've been brainstorming this a while. It is amazing how many flat bar/angle iron parts are almost workable but just don't have two opposing 45 degrees angles like we need. 3in angle iron would work, but I'm guessing it has its own problem. Plus, I already own extrusion... Thanks again.
check this out.... he used round bar http://www.cnczone.com/forums/diy-cnc-router-table-machines/91481-cnc.html
check with cncrouterparts, they have set up using square stock at right angles with clamps to hold the bar in place. Pat
You can also use steel V wheels on round bar if the diameter is small enough. Mark EDIT: whoops already been said
I bought some cold roll 5/16in square solid rod for my x and y axis, laying them at a 45 degree angle in regular old extrusion. It seems solid enough for me to order a few more steel wheels. Ideally, I'd cut a V channel in a piece of aluminum and place the cold roll on that, but resting them on the slot seems to be acceptable to me so far. Or, I'd just buy those parts from cncrouterparts, but I'm cheap.
I'd be curious to know what you ended up doing. I ended up mounting hardened steel v-rail with 3/8" steel v-bearings to c-beam. You can see my process here, of trying to make the c-beam based CNC machine more rigid:
Wiremonkey... you are developing several inches worth of torque around your X-axis with the pressure being applied to your router... It would be more effective to have a dual X-axis with the router between and connected to both so that the forces are directed vertically in a linear fashion rather than rotating (and twisting) your axis... That will solve most of that issue without having to stiffen the heck out of the single axis channel.
This is an interesting idea. I think I've read this comment before somewhere. Are you indicating two separate X beams, one above the other, each with a rail? I've definitely seen that type of design, but the single beam is not just a twisting, it's actually bending, there is no question about it based on the measurements I took all over the machine. Certainly a second X beam would help. However, I do think I'm headed down the steel closed beam route, it makes the most sense to me after much contemplation. I just ordered a 2" x 6", 3/16" walled piece of mild steel tube. I'll mount linear rails on the face. There will also be twist forces, but the steel tube is vastly more rigid than the aluminum c-beam, heavy too! Presently, that is my destiny. I'm going to have the face of the tube milled flat. Eventually I'll switch to a Y table with a fixed gantry. This just seems like better design. I'll post my current design idea here soon, once I mock it up in fusion 360, my hand drawings are laughable. I'm happy to get feedback. I'm not an engineer, more of a carpenter/mechanic and I'm happy to hear from folks who have genuine engineering experience. I'm itching to figure out how to do stress models in Fusion 360! Thanks for weighing in Ed!
Looks interesting. Are you building it or are you still considering options? Two c-beams are certainly better than one! And I do like the square stock rails. I've seen folks do it with round stock as well. Although steel square stock ain't cheap either. I think round stock is cheaper for hardened polished stuff. What's your timeline? For getting parts, putting it together, etc? I just just got a 2" x 6" 3/16" walled piece of steel that I plan to make my X beam. It's a bit of a monster, but I have no intention of moving fast and already have very beefy steppers. I'll try and mock my idea up in Fusion today and get your input. You used fusion 360 for this?
I did a mock up with wood plates, a piece of Cbeam, Steel wheels, and stainless steel square bar (.25). Feels good. the square bar gets pressed firmly into the Cbeam and doesn't move when the wheels are pre loaded. No need for any bonding agent. I got some ideas from this guy:
Yes, fusion 360. I have too much flex and chatter when milling aluminum. I forces me to take too small of a chip load. If I had unlimited funds I would get these: Linear Guideways- Huntley, Illinois- HIWIN Corporation
I would suggest bonding the two C-beams into a single square tube section. Tubular sections offer more torsional rigidity than the sum of the two C-sections individually. If you need to maintain the width shown, bonding a piece of 20x80 between the two C-beams will also work and make the assembly even more rigid.
I still favour 2 CBeams, one above the other. So, one to the lower part of the Z Axis, and one near the top of the Z axis. The lower beam to take the Vertical weight, and the upper to remove the risk of twist, forward and back.
Mr. Gray UK. I think you're correct, although, closing the structure, attaching the two beams together, perhaps with through bolts, would also go a long way to making things more rigid. Although, if you've seen my video, there seems to be as much general bending upward of the c-beam as there is twist. So, putting one above the other doesn't necessarily address that issue. M90, if you have the parts lying around, why not try it out and report back. Otherwise, I'd start to look at more rigid materials than aluminum for a long span, it's simply not stiff enough unless you start getting into massive thicknesses and then, you may as well go steel.
Yes, linear guides are a great improvement, but rigidity of your structure is just as important. I think I'm going to try these linear rails on the new steel X beam: discounted automation products: BLH Linear Guideway, BLH 20mm Linear Guideways The profile is hiwin compatible and I've heard good things about them. If you want to go hiwin in the future you can keep the same rails. Anyway, I'm going to give them a whack.
What would you recommend as a bonding agent? There are a ton of variations. Maybe plate them together?
Loctite Epoxy Metal/Concrete seems to get the best results when squeezing it down to thin layers between materials. Might be a good idea to rough up the mating surfaces a bit. The plate is also a good idea.
OK gents, I did a quick mock up of the steel 2" x 6" x 3/16" wall X beam that I'm going to construct. The six holes in the end are for mounting to the current gantries. I'll have the bottom surface toward the ends, made flat and perpendicular to the front face so that when I upgrade to a fixed gantry, I can reuse it. Thoughts?
I think I'll try the 2" x 6" like this, it is so sturdy! I think If I did two square tubes, as you've illustrated, I'd do a bit thicker wall, maybe .25". I have the material, so I'm going for it. I'm sure there is a way to test whether or not the 2" x 6" vs two tubes would yield more rigidity. I'm used to all of the upward spindle force being on the bottom of the beam, due to the poly carb v-wheels sandwiching the beam. With the linear rails, both top and bottom rail would be taking the upward forces equally. It's difficult to wrap my brain around that. Certainly it would be cheaper to buy 2" x .25" square tubing! And you could decide how tall you want the beam based on the end plate size. I'm going to do a weight comparison of the two versions and get back to you. Because this 2" x 6" will weigh about 25lbs! I do believe the thicker stock 2" square tube would be close to that. Hmmm. Any engineer types want to weigh in before I pick up my welding helmet?
Mr. Ranger, you mind if I post your image in another forum. I'll send the link. There are some engineers over on MYCNCUK who have been helping me think this through.
Sure, Post it. I think the screw in the middle would drive the rails with more equal horizontal force. You could do the two 2 by's and join them with a plate on the back too. OR You could do the 2x6 with 2 lead screws, one on the top and one on the bottom which would drive them with equal horizontal force too. Endless Possibilities....... I'm no engineer, I'm a Mechanic for over 35 years with some college engineering classes from way back when. I have a good Idea of what works from experience, not calculus
Ha! I was a helicopter mech in the marine corps, but sometimes the slide rule can tell us just a bit more... I like your ideas. I think there will be enough space behind the X axis assembly worth linear real blocks, to get a lead screw in there. Motor on the back with a belt. I'll mock it up...
