Has anyone come up with a working solution for X-axis flex on a 1000mm X-beam on their OX? The more I use the machine the more I notice. 2.5D carvings and V-carving are showing the weakness in 2X 20x60 slot configuration. The combination of plunging Z while moving though material allows the2 pieces to flex independently. (The Z-rail is actually rotating around the x axis!) I 3d printed spacers to join the 20x60 extrusions together, but I still have noticeable flex. As expected, it is worse in the center of the beam. Everything on the machine is tight and working. It seems fusing the two 20x60 rails would significantly stiffen the whole OX. I was considering TIG welding the two 20x60 extrusions together, but I am worried the heat would distort the slot the X-carriage wheels run in. Maybe epoxy would be a better solution. Another option I'm considering would be a fitch beam, with a .25" aluminum plate epoxied between the 2 rails. .25" spacers would have to be added to the X carriage wheels to compensate and new holes drilled in the gantry plates. Has anyone tried these options or other ideas?
everyone uses 3 rails on their x, and use 2080 for a 1m length. Dont forget you might just have Z backlash as well.
Yeah, it's definitely not Z backlash. It's the independent movement of the X rails. If you center the spindle on the X axis and apply rotational force about X you get the most movement. The X rails twist more than they bend.
Nice idea Dany. I added another wheel to the Z rail in that area for 4 wheels per side instead of 3. I really don't see any flex there. When I wrap up a couple of other projects, I will try fusing the X rails together with either Epoxy or automotive panel adhesive. Panel adhesive like 3M 08115 or SEM 39747 is incredibly strong. Most auto manufacturers are using this stuff to assemble vehicles. The downside it the cost of the fancy caulking gun that squeezes the two parts through the mixing tube. I have a gun from a restoration project a few years back, so it might be worth trying this mod. I'll post pictures and results when I do.
Just looking at upgrading my mdf built cnc 1200mmx600mm and wondering what size extrusion would be best. The mdf is showing signs of flex and movement and I like the idea of the v slot. Thought I'd ask here rather than open up a new thread Thanks on advance Mike
Start by bolting the two sections together rather than gluing. bit more work but far less mess and risk IMHO Fix the two sections together so they are flush at the ends and can't move (I use joining plates and spare screws and tee nuts on the faces I will not be drilling). Next I drill on 150mm centers with a 4.2mm bit, mark the stacks orientation before splitting out. tap holes in one stick to M5 and and counter bore (use a 10mm spade bit) for bolt head clearance and to create a flat pad for the under side of the bolt head before drilling out the point center to 5.5mm for bolt shaft clearance. This is quite a bit of work and requires good alignment of the sections before drilling or it will not pull up well to the gantry plates. ( I actually think I had to spot drill with a 2-3mm bit so I had a small enough center to keep the tip of my spade bit from wandering before drilling with 4.2mm...) Next build I plan to try and just drill one side with a counter bore and clearance holes and then use shorter screws and tee nuts rather than having to align, drill and also tap the other side. It could be left a little loose until the gantry plates are pulled up saving a lot of alignment issues. I am also keen to try a fletch beam some time in the near future but think that getting it laser cut from 3mm steel with clearance holes/space for cross bolting the beams together would be the way to roll. I think a steel beam would give far better overall performance as the two metals would act as a composite structure. I wonder if the web in a fletch beam needs to be solid or if the middle it could be removed with a triangular cutouts while still maintaining the overall stiffness. Adding more wheels or swapping to a Cbeam 250 actuator and the new XL plates would also be a big upgrade. You could also move to 20/80 or Cbeam for the gantry but you would need to remake your side plates....
The "twisting" of the two beams is the very reason I am rebuilding/strengthening my Ox...I am replacing the the two 20x80's with a C beam. I was thinking of bolting the two 20x80's together, but adding some more upgrades as well (linear rails for the X and a new Z with linear rails and a ball screw. Will end up with 5 1/2" of Z travel and much more clearance from the table to the gantry
I am also planning on doing something similar to this with my X, since my 1000x1000 C-Beam Mahcines X axis seems to keep a little flex in it as well. I noticed that the back of the screws (the head end) was also wallowing out the back of the plate. So for now I have added some Number 10 24x1 1/2 5- screws instead with a wave lock washer and loctite until I get the other plate wheels etc in hand.
@anigeek65 If you are using two C-Beams anyhow, why not put them on either side of the spindle instead of back to back? As long as the spindle is only supported on one side, beefing up that side will only get you so far. The only downside (aside from the cost of an extra cross beam and changing plans) is the reduction in available travel. Or am I missing something? -D
Jonathon, Thanks for reviewing. You are correct I could have placed the other C-BEAM on the other side of the spindle (Straddle the spindle). A good thought. What I was going for was 2 things. Increase the span from V rollers on the front C-BEAM to V rollers on the back C-BEAM. I noticed on the C-BEAM machine I built, the z axis can be pulled forward and back at the spindle end mill (not enough support) . Also wanted as much surface area to mount the side plates to C-BEAM side (2X the C-BEAM surface ends) to help control the parallelogram affect on the x axis. Although I like your idea and it intrigues me, my C-BEAM machine I built to cut plates for this project would probably not make plates large enough to support your design approach. The plates I am making now are about the limit. Maybe in a later and bigger build, after this one I can consider this Idea. Thanks for the feed back.
Just be aware that a large amount of the play in these systems is the fastners bending. Longer fasteners = more bending.
Great food for thought with these ideas. I think I'm going to try bonding/bolting/welding the 20x60 rails together to see if there is improvement. Then I'll cut new plates and move to either a flitch beam or double c-beam. That would probably be a good time to consider a switch to ball screw drive instead of belts...
Image by Evan F posted Jun 23, 2016 at 10:07 AM Finally got some time in the shop time. I bolted the 2 X-rails together with 15 - #8-32 x 1.25" cap screws. I drilled from front to rear and tapped the rear rail. The holes are equally spaced with 8 in the upper groove and 7 in the lower. (My local hardware store only had 15 capscrews). The results are mixed. There is a noticeable change in X rail stiffness. I believe this modification is worthwhile for anyone building a standard OX. Unfortunately, there is still too much flex in the X/Z interface. I believe it is a combination of each of the items pointed out above (fasteners, wheels and their location.)
I am going to bolt the x rail as well as make a new connector for the motor. The flex coupler flexes way too much . I installed a 2.2k spindle and already seeing where I need to strengthen the x & z. I'm going to slap a steel clamp on the connector and see if that takes the sag out. I know I might have overweighted the z axis but I have it hooked up. V rails and v wheels look like a good idea as well. Next design will be Stronger than the Strongest ox. Great for first timers but I need to modify it and with the gained knowledge I will be using a combination of linear steel rails and Aluminum extrusions.
I routinely bolt together the two pieces of 20 x 60 mm in my OX builds. I usually just put in 3 bolts evenly spaced on the center-line of the rails, one in the center and one about a third of the way across on each side. I use standard low profile screws, putting the heads on the back side, with a t-nut on the front side. Forget the length of the bolt but I think its a 30 or 35mm. These do not interfere with the carriage in any way and stiffen it considerably. It's easy to drill the v-slot because of the little groove in the middle. I just clamp the two pieces together and drill it on the drill press. I think your huge number of bolts is a bit overkill, but that's just my opinion. Can't hurt anything, anyway, except for the time involved... I've been waiting for somebody to do OX plates for 20 x 80 mm cross members so I can use either 2 x 20 x 80's or a piece of C-Beam... That, and upgrading the back cross member to a 20x60 or 20x80, or even a 40x40 as well... The ox cross member is just nowhere near stiff enough for a 750 x 1000 build. It's barely adequate for a 500 mm x axis... As soon as I get my new and improved C-Beam machine up and running I'll have to try my hand at making my own... Upgrading the “old” C-Beam machine C-Beam Machine Too
I made plates and I did two 2080s on mine bolted together. Works great. I pretty much have it done and have been drawing circles, squares, and hexagons with it because...well I haven't mastered sketchucam yet and I am scared I'll break stuff. The squares are square and the circles are round and the pen starts and stops at the same spot so I must have done something right!
I designed them myself by modifying the original Ox plates and making them 20 mm taller. I cut them by hand and drilled them on my drill press since I had no access to a CNC, and because it was fun to figure out how to do it as accurately as possible. Lead Screw Driven Ox Derivative (850x1500)
i bolted the 20-60 x axis together in a staggered pattern , ' , ' , ' , and it really improved the rigidity.
Wow, Craig, you have amazing skills! I wouldn't even attempt to hand make all the parts needed for a router build... It must have taken a lot of time. But the good news is that now you have one, the next build will be a piece of cake! And, as Justin says, SketchUCam is really easy to use. Much easier than learning SketchUp itself... And a friend of mine always used to say "If it works, you have to find out what's wrong with it"
Thanks. I am trying to figure out if Sketchucam is 90 or 270 degrees off from what I expected on mine. For all I know at this time, it is possible I have it set up so it is inside out. But back to the question, the double 2080 bolted together at about 840 mm seems very stiff. As I mentioned in my build, a friend printed out the v-slot-connector parts (V-Slot connector | OpenBuilds ) and I used those to line up the two pieces of 20x80 v slot before drilling. I drilled though one, then drilled and tapped the second. I used 12 m5 bolts (actually 11 because the tap broke off inside the twelfth).
No I checked 'overhead'. I am embarrassed to say it was operator error. Everything is set up temporarily with my computer sitting on my belt sander and rotated 180 degrees off from my cnc. So you can probably guess why I thought the machine was off by 180 degrees. the first time i had the computer on the machine and 90 degrees off from the x axis. I realized why I was having these problems last night when I was playing with it again.
