Larger Format Machine

Hi Everyone,
I posted this as part of a reply in the General Discussion section, however, I think this may be a better place for the question:

I'm planning to build a CNC Router that can work full 8x4 sheets of material so I'm going to need rails that are a little over 8 feet (~2500mm) long, the largest V-Slot available is 1500mm long. If I join two smaller pieces of V-Slot using some Double Tee-nuts to create X-axis rails greater than 8 feet in length, will the point where the two pieces join create a bump or disruption when the Y-axis gantry passes over it? Would such a disruption result in significant imperfections in the work piece being cut? If so, is there a way to mitigate or prevent these imperfections from occurring?

One way to deal with the joint is the put 2 separate runs of vslot together side by side and offset the joints so that the wheels do not contact the joints at the same time. It should reduce the effect of the joint from the carriage.

Bob

I'd like to see openbuilds make some 3m v-slot so we can all make these without headaches. Putting a joint in the middle of a beam weakens it significantly and alignment is very difficult. If you could make an insert that goes into the extruded aluminum about 10 inches on both sides it would align and reinforce the beam enough. Maybe you only need 2 inches into each side if it was cut from aluminum.

Yeah that'd be nice! I asked about this, unfortunately it's not possible right now. Here's the response:

Do you think a snug fitting length of metal pipe would work? Something in the area of a 10mm OD?

Well I've heard of people running m8 threaded rod down the center of T-slot extruded aluminum for support... I don't think it works as intended. If you look at the cross-sectional profile of a 20x80 v-slot, as in this diagram:

What you see are not just places to insert rods, but also hexagons with tabs. So with 4 rods and 3 hexagons yeah it would be strong enough. Maybe 4 rods would work as well. Then just drill through the v-slot sideways and bolt the inserts into place. Or don't bold it, maybe glue will hold.

Actually yeah you could use pipes instead of hexagons+tabs. That would be 4 small pipes plus 3 large pipes and it would give you a beam even stronger than one solid piece of v-slot.

Well that should be easy enough to pull off, I should be able to add some extra support along the run of the rails anyway, not just at the ends, so it should be able to support plenty of weight.

My concern now would just be the point where the two extrusions meet. Can two pieces of 20x80 v-slot be mated end to end with an interface so flush it won't affect the work piece when the wheels of the y-axis gantry run over the mating joint? Thoughts? Anybody out there with first hand experience?

I guess it depends on the materials you get. The machines have some variance in the production line but if you got 2 pieces made by the same machine at the same time they should align perfectly. I'm guessing you'll probably have a 0.01mm bump and you could use a clamp to shift them into alignment while you let some epoxy dry. I'd use that resin they use to make fiberglass so you could sand and polish the rail-joint. This might remove the anodizing so be careful.

Hey you could make a phlox(phlatprinter+ox). Just make it 1.5m wide which would give you 4ft plus room for your tool-head... then load up a 4x8 onto it with a roller support, and let it feed through like a phlatprinter does. Make it 1.5m long too so you can do large heavy materials without the rollers as well as 3D printing and plotting etc.

What if he made an extrusion or a milled piece that fit that inside shape perfectly? 2-4 sections of that, along with bolts or screws from the outside, should make that easy. I'm suggesting this so maybe it could be a stock part in the part store.

The phlox could use stock parts and do pieces of unlimited length. It could also act as a tablesaw and table-router and lubmer mill and etc

They butt together rail for big cnc all the time so surely we can do the same with Vslot....

It may result in a small bump in the motion at that spot BUT is that bump actually worth worrying about?

Would most people working in timber/foam notice a small 0.1m bump in those materials?

Here is what I have been thinking....

If your butting vslot together your building a BIG machine so will likely have longer gantry plates and more vwheels. Maybe adding as many as you can will help smooth any bumps as at any time you will have many sets of wheels running on solid vslot and only the single set of wheels running over the joint.

I would also design my joints to all be staggered within the build so that only one car is ever hitting a joint at one time.

An other idea is to build bigger sections than 20x80. I was planning on trying to bolt two sticks of 80x20 together, staggering the joints and using each section to align the next. Finger crossed it will run smooth enough to cut out a surfboard/skateboard blank.

I would think JB weld/ epoxy could be needed to make a super smooth joint but I hope that its not needed.

These are all good suggestions. I think at first I'll try just the double Tee-nuts (maybe with some additional support running through the 80x20) and some JB weld to smooth out the joint. If there's a considerable bump created by that connection I'll start implementing the additional suggestions until I find something that works. One way or another I'll report the results here.

IMHO I would try everything BEFORE glue/epoxy. If you lay the rails on a good surface and clamp them down in two directions while bolting everything up I would think you should get a good connection with almost no noticeable bump BUT this is assuming you get a dead square and true cut to both ends. Rough cutting then milling true is key for butting linear bearing rails together. I think a good cut with Vslot is also going to be the easiest way to get the most gain for butting rail together too.

If the cut is a little off then maybe the glue is needed more as a gap filler than bonding agent? If I was to glue the rail I would start by masking all but the very end of the rail, add glue, then wrap with a plastic (release ) layer then stuff some soft foam insert onto the join area v groves surfaces and apply pressure with a tight wrap of electrical tape. This should push all of the excess glue away from the rolling surface and might save having to sand them after the glue is set. The key will be getting the stiffness of the foam and pressure just right so it forms a nice rail mold rather than a dent....

For the builder who have used Vslot, what is required to span a 1250mm cut area? I was going to go with two 1.5m sticks of 80x20 back to back with around 250mm spread for the widest roller to build a gantry with. Is this a good size or do I need to go bigger? The Ox and frog seem to use this config as well.... (am i thread jacking ... is this different enough it needs to be moved?)

Well, I haven't used vslot yet but I've read a lot of what other builders have done to make similar spans with t-slot. They found that the weight of the tool determines how much the beam flexes AND the force applied to the perpendicular axis(x? y?) causes the middle of the support to flex as well. What they've come up with to span an 8ft section is what looks like an I-Beam. It turns out that you can get a piece of steel U-channel and put it between 2 vslots. You then use tee-nuts to bolt the steel to the aluminum... giving a sideways I-Beam(H-beam?). Obviously this heavy beam is going to put a big load on your x axis...

It may be possible to do an off-the-shelf build with 3 pieces of vslot and some right-angle brackets to make an H.

FYI, this is the wrong way to do 4x8ft. It costs less than having 2 8ft side rails but it makes up the cost by having to have stronger supports, wheels, motors, etc.

Also, just do a 1500mm span and gain the extra build area and avoid the cutting...

interesting idea with the steel c section, I was actually looking at using a flitch beam as a simple way to add a bit of stiffness but the steel C is also a good option.

https://www.google.co.nz/search?q= ...#q=flitch beam&rls=org.mozilla:en-US:official

that extra steel would work well for an X axis and the added mass is not a bad thing at all, maybe two steel c sections to create a true box and run two rows of wheel (one set on each side of the vslot box)? Strong, stiff and heavy but that's good for a base.

I can't see any need for myself to build an 8ft 2.4m Vslot gantry and think it would be a little lover kill for a 4-5ft (1.2-1.3m) gantry.

Are people having success with say two 80x20 stick strapped back to back for a gantry beam? I would like to run a water cooled spindle so need to move around 12-15kg .....

In all honesty, 2 back-to-back beams should be really strong. I'd give it a try and if it doesn't work then consider the H-beam, flitch beam, or even 3d print a spaceframe flitch reinforcement.

I keep seeing the terms "strength" and "strong enough" appearing repeatedly as one of the main focuses but the issue here is not strength but deflection of the rail. When you double the length of a simply supported rail subject to a single centered point load you increase the amount of deflection by a factor of 8. In other words, the .25 mm deflection you don't notice in a 4' rail becomes a 2 mm deflection at the center of an 8' rail, something that will become fairly obvious as your cutter gouges into your table. Solutions to this are fairly simple. Either reinforce the rail with a more substantial piece of framing (a 4x4x0.1875 steel tube should do it) or solve it the more structurally rational way and support the side rail all along its length from the table framing.

Attaching the rails to the table framing also solves two other problems, lateral sway of the rail caused by the sideways kick of the router movement upon the gantry and torsional sway (outward rolling) of the top of the rail. Again, a 4x4x0.1875 steel tube backing up the rail will largely resolve these but why bother when you have a phenomenal amount of lateral strength available from the MDF reinforced table framing.

As for the issue of wheels snagging or jumping while crossing a rail joint, the solution is fairly simple. Wheels... lots and lots of wheels. Envision a 4 wheel gantry plate (2 wheels top/2 wheels bottom) and envision what happens when the first top wheel goes off the free end of a rail. Chaos, right? Now envision a 12 wheel gantry plate. The first three sets of wheels can run off the free end of a rail and the gantry still has no idea anything is happening. By extension then, as these wheels run onto the next section, as long as there is good alignment there will be no bumps to the system. You could easily jump a 2" gap in the rails as long as the majority of the wheels are in contact with rail and there is fairly good alignment of the system.

Suggestions for making it happen: Put at least 2" spacing between all the rows of wheels. The greater the spread, the better. Set up and carefully tune the wheels on a single piece of rail and then mount that rail. Slide half the wheels off the end and feed in the next length until the ends meet. Clamp the rail off and carefully re-position it until there is no discernible feeling of the joint when sliding the gantry plate back and forth across the joint. Additional prep suggestions include slightly feathering both sides of the V-groove at the mating ends and inserting and JB Welding some 4 mm dowels into the center holes of extrusion. Feathering about 3/4" of the V-groove surface both sides of the joint will allow the wheels to gradually come back into contact with the groove, eliminating any potential snags and JB Welding in the dowels and allowing them to set reduces the potential of knocking the joint out of alignment when installing the final mounting screws.

Sorry to be so long-winded.

Rick.

Thanks Rick! More wheels and feathering down the joint sounds like a winning approach. You weren't long winded, just perspicuous.

I agree with Rick2.0. I posted this up in another thread, but I'll add it here as well since it seems relevant. This method is used in other designs around the web, but I haven't seen it used here yet. Take some 3/4 angle iron the same length as your extrusion. Drill some 5mm holes in along its length, then bolt the whole thing down to a sturdy table or torsion box. This will give you a solid X axis that can be as long as you like without deflection. Bonus points for making a template that spaces the holes correctly. You can also use the same technique for making sturdy carriages from scrap V-Slot. Like this:




and yes, my CNC build is totally overcompensating for its tiny,tiny motors...

Thanks Mark. I'm actually just getting around to posting my progress from last year... I've been distracted with my Delta build. Besides all that steel reinforcing the Y axis was really heavy. Its solid, but really heavy. Your phenolic plates are a better way to go for the Y support. I've got a co-worker that is theoretically cutting me some plates this weekend. That said, the angle iron trick is great for rail support and quick carriages. I've been meaning to suggest that it would make a great addition to the Openbuilds Store. Drilling all those holes is getting old. If we could get it in quantity, I think it would make a great addition. Just cut some to length to add support, create any size angle bracket, or as a quick way to make carriages. I can post more images/uses if your interested.

I think aluminum would have better tolerances than the steel angle that I have available to me. It's not as strong as steel but much lighter. I think the original problem was to prevent the rail from flexing so aluminum would be strong enough. An even better solution I've seen is adjustable legs every foot or so. The kind you see on the bottom of refrigerators that you just turn a screw to adjust the height.

You could make it from aluminum (and obviously that would have better tolerances) but I'm not sure that it would have the necessary strength to support the full weight of the X and Z axis. If you could reinforce the aluminum angle then it would definitely be strong enough. Perhaps an extruded right triangle? I've never seen one, but it would also make it more difficult to attach to the table. Regardless, when using this method to mount the Y axis rails, weight and precision don't matter much so steel angle iron seemed like a reasonable choice.

Dude, why do you think adjustable feet every ft is better than a fully supported rail? Seems like this would be more expensive, and complex to setup, but perhaps I'm missing some major advantage.

On the subject of bridging longer spans, I considered some form of suspension bridge cable contraption. For example running a tensioned cable down the extrusion between the two X supports... However, I decided that it would be simpler to just use additional V-Slot beams to share the load, and then create some form of bonding between the two beams. To that end I made these:
http://www.openbuilds.com/resources/v-slot-connector.40/

Here is the resulting 60x60 beam on my prototype CNC machine (still not cutting...) I think this method could be used to create a fairly rigid 1500mm span.


Let me know what you guys think.

Hey that looks plenty strong. The connectors are a godsend. I don't know whether foot-pegs would be better than steel angle but on larger format machines you may see the metal expand or contract due to temperature. If you connect 2 different metals, they bend instead of expanding. Again, I don't know if it applies to your situation but if you're going for 0.001" in an unheated garage...

Hi, realy like the V-slot connectors, would you consider making them to sell, I would sure be interested in 8 of them.

Hi Velocette,
I don't have any plans to sell the V-slot connectors, I just made a few (12) of them locally on my printer for this project. You should be able to take that stl file to a local maker space and get some printed out for a reasonable price. If not, there are a fair number of online locations that will make you parts (likely more expensive). Its also possible that Openbuils has something in the works to bond sections of V-slot together since it seems like a common problem when trying to make larger format machines. Sorry I can't directly help, I'm getting crushed at work trying to get our game out the door.
-Sage