1 Attachment(s)
Building Steel Frame Gantry Router for wood, MDF, and maybe more
Howdy all. I’ve been lurking for awhile and am finally putting a thread together to get some feedback and hopefully document my build if I can get this project off the ground.
I’m planning to build a welded steel tube frame gantry router with epoxy leveled linear rail beds. I would like to have a working area of around 56”x48”x12”. Those may seem like odd dimensions, especially since the gantry axis (x) is longer than the fixed (base?) axis (y). However, I’d like to be able to cut larger sheets/blocks, without having a full 8-10 foot table, by repositioning my workpiece. I don’t have the space or money currently for a larger table.
My usage of the machine would be for cutting plywood, MDF sheets and laminated MDF blocks for molds, and possibly larger foam blocks for larger molds. Ideally the machine would also be able to do light aluminum work, but I’m not sure that’s a realistic goal. Not sure what I can expect my accuracy to be with Chinese rolled ball screws, but I guess I’ll get what I get and (hopefully) won’t get too upset.
I’ve attached a screenshot of my extremely rough frame design (still have no idea what I’m doing in Fusion360 so sorry for the absolutely garbage model), to give you a vague idea of what I’m envisioning. I’m still determining what steel tube is available locally, so exact structure is subject to change. Broadly, though, I’m envisioning a low gantry and a table with fairly high walls as shown above. I’m really not sure whether having a “flat” 2Dish frame is okay, or if I need to add some sort of triangulated structure below the cutting deck to keep the frame from “racking” if I move it after epoxy leveling the Y axis (and I will 100% have to move it at some point). Advice is appreciated.
Linear Motion: I’m leaning towards rails from BST motion on AliExpress. Any feedback on this option would be appreciated. I’d like to get rails that match the capability of my frame (and vice versa), so that I’m not wasting money upgrading one component when the other is already a limiting factor.
https://www.aliexpress.com/item/3301...24ce67b7Tp5PzQ
Actuation: leaning towards rolled C7 ball screws if I can find ones long enough, possibly also from BST automation? Advice on good vendors would be greatly appreciated. I’ve also seen a recommendation for linearmotionbearings2008 on eBay? I assume I would want something in the 20-25mm diameter range, and a larger pitch? Maybe 10-20mm pitch? I don’t think my cutting forces will be that high on most of the material I plan to cut, so seems to make sense to bias towards faster travel (and larger pitch will also hopefully help keep my screw RPM down to limit any wobble over long distance). One question I have is whether I should be looking at double ballnuts to reduce backlash, or whether that’s going to be overkill for my design. Once I have a better idea of approximate gantry mass, I will do the math to figure out if there is a combination of gantry mass/ball screw pitch/stepper motor torque that seems like it might work, but I’m still crystalizing my frame concept.
Spindle: I haven’t done all that much research yet, since it doesn’t really seem to affect the rest of the design at this stage, but I imagine I will end up going with something along these lines. I’ve had a small amount of experience with the capability of these spindles on a friend’s machine, and it seems to be a reasonable option for the price. I will have to investigate further as to whether a better option exists.
https://www.amazon.com/MYSWEETY-Spin...1282578&sr=8-9
Control: really haven’t gotten here yet, but Mach4 + ethernet smoothstepper + g540 + appropriately sized stepper motors is kind of my fall back. All I know for sure is I really don’t want to deal with a parallel port of any type at this point.
Well thanks for hanging in there for anyone who read all that. At this point I’d greatly appreciate any feedback on whether or not my overall approach is realistic/reasonable, especially with regards to how I plan to do the frame. Thanks again – David.
<a href="https://imgur.com/rH8qF4l"><img src="https://i.imgur.com/rH8qF4l.png" title="source: imgur.com" /></a>
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi David - In principle your going down the right path. I'm not sure about the epoxy levelling because I have not done it. Some threads say it's great some say it was a disaster. I feel its tricky as some epoxies are formulated for a thin film cure and some are not. If you use the thick film type and its only 1mm thick then it may not have enough internal energy to cure correctly, plus when it does it may wrinkle due to the nature of it curing in "cells" and then changing volume due to exotherm etc etc. Sp speak very carefully to your epoxy supplier if this is what you intend to do. I would braze the frame not weld it to side step the distortion issue then you can level the frame and parts with a good level and braze it. It will not change enough to affect your structure. I'd be much more comfortable with that approach vs epoxy levelling. Running your drive through the hollow may not be a good idea as if you want to remove the drive or maintain it you may have to upset your gantry. Try to design things so they can be maintained without disassembly of other systems. Keep modelling until you a very happy with the design. Much easier to play in CAD then the real world. Peter
1 Attachment(s)
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Made a little more progress. Starting to get the hang of Fusion 360 a little. Screenshot below. Integrating the ball screws is going to be interesting. I had thought out most of the frame and rails in my head, but the ball screws will really have to be squeezed in there, especially on the X/Z carriage assembly. I may have to find a way to have those carriages machined so they are not simply flat plates.
Peter,
I'm not sure why brazing would be better than welding? The steel tubing will most likely not be sufficiently flat for mounting linear rails from the factory, so it seems that even if I were to braze the frame I would still need to do a machining or epoxy leveling operation at some point? I would like to simply have the frame stress relieved and machined, but I don't think that makes financial sense when you consider the quality of the other components I'm planning on using. As long as I don't do any more welding or removing of material after I complete the epoxy leveling process, I don't see a reason why it would matter whether the frame was internally stressed or not?
I agree that the epoxy leveling seems finicky - I have seen a number of threads where it did not go well. I think the key will be to do some testing prior to the actual leveling pour(s) on the frame.
You are right about getting everything nailed down in CAD. I'm just impatient and want to start buying steel, haha.
Thanks for the comments; they are much appreciated.
David
<a href="https://imgur.com/LNg1n8e"><img src="https://i.imgur.com/LNg1n8e.png" title="source: imgur.com" /></a>
Attachment 429744
1 Attachment(s)
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi David - Personally I would not epoxy level. The epoxy is not stiff enough and the rails will be wavy when you pull them down. There is a lot of debate about this point in the forum. But epoxy is 3.5GPa stiffness and steel is 200GPa so you are creating a rubber gasket under the rails. Which IMHO is not good.
When you weld you melt the parent metal. Steel shrinks 2% by volume when it solidifies. This creates huge tension within the structure. This is why things warp when you weld them. If you braze you do not melt the parent therefore you do not generate much internal tension when the braze solidifies. The braze is a bit stretchier then the parent and if you have snug fits then the fit resists the bead shrinkage. Plus as you braze you heat quite a bit of the parent to red heat which is the stress relief temp. So the parent is soft enough to stretch if it has to. ie you stress relieve as you go. In mig welding you do not heat enough parent for this effect to happen.
Stress relief is a process where the metal is heated to cherry red (650-700C) at this temp steel is like butter and the internal stresses relieve to a very low level ie if the metal has a stress in it higher then the hot strength it stretches and then sets a bit longer reducing the stress. See image attached.
So why is this important? If you weld up your frame as best as possible then machine it, it will warp when you release it from the machine bed so you will defeat the reason for machining it! This occurs due to the change in geometry resulting in a change of stiffness which the internal stresses then balance by warping. So you need to decide on what balancing act you want to follow. One way with welding is to weld it as subframes that bolt together so you can adjust things at the bolted connections. I have suggested in other threads that for machines like these good strength soft solders would be good enough. The metal would not even change colour but the surface area involved would be huge so strength would be adequate for a machine like this. I do a lot of silver soldering of parts and tobin bronzing of parts. As far as I can tell with gauges they do not change shape in the process.
I used to make steel and aluminium bicycle frames. The steel frames where TIGed or brazed. The brazed frames when cool would come out of a jig and you could put them back in the jig no problem., The TIG frames however would spring and you would have to cold set them to return them to the jig even though the fits where machined fits. So think it through there are lots of options available to you at this point. If you where to braze I suggest you tack it in the position you want so it is in a "free" no constrained condition then braze or solder it. ....Peter
Edit You have drawn the small uprights as the same size tube as the horizontals. This means the edges will have large gaps due to the section corner radius. This gap is easy to fill when welding but provides a great place for weld shrinkage to warp the structure. So either a) do not weld bits together at rads or b) use the next size tube down so it touches all surfaces for the weld. From a strength point of view 100% length welds are unnecessary in machines like this. So stitching it together will reduce your shrinkage considerably if this is what you want to do. Only stitch bits that touch...
https://en.wikipedia.org/wiki/Solder_alloys
speak to Kappa Solders I have talked to their people before and they have given excellent advice. https://www.kappalloy.com/ many of their solders are as strong as mild steel...
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi Pete, thanks again for the comments. I will look into brazing a little more. I do like the idea of not having to deal with all the warpage from welding. I only have TIG anyway, so TIG brazing wouldn't really be any more work than I was already thinking. I will have to do some more research on how that compares to other brazing methods.
Epoxy leveling does seem quite bad from a stiffness perspective. Unfortunately, I think machining the rail beds would be prohibitively expensive at this time. I have no idea how much it would cost, but I can't imagine it would be cheap. I guess one option would be to epoxy level now, and perhaps have it machined down the line. I don't see why that wouldn't work.
Quote:
Originally Posted by
peteeng
Edit You have drawn the small uprights as the same size tube as the horizontals. This means the edges will have large gaps due to the section corner radius. This gap is easy to fill when welding but provides a great place for weld shrinkage to warp the structure. So either a) do not weld bits together at rads or b) use the next size tube down so it touches all surfaces for the weld. From a strength point of view 100% length welds are unnecessary in machines like this. So stitching it together will reduce your shrinkage considerably if this is what you want to do. Only stitch bits that touch...
I understand what you are saying here. I had that thought when initially modeling it as well. How do you think it would work if I fabricated it as I have it modeled now, but brazed those large gaps formed by the fillet instead of welding? In my mind, I ideally would replace both of the side "girders" I've modeled with a single larger tube, and then the cross pieces could butt up against flat metal on all 4 sides. I don't know how feasible those large tubes would be for me to obtain though, so this might not be possible.
Do you think this frame design needs more torsional stiffness? One of my fears is that I assemble the frame in one place, move it somewhere else on the floor, and it "racks" out of being a flat plane due to the floor not being flat.
David
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Quote:
Originally Posted by
thegadgetguy
I understand what you are saying here. I had that thought when initially modeling it as well. How do you think it would work if I fabricated it as I have it modeled now, but brazed those large gaps formed by the fillet instead of welding? In my mind, I ideally would replace both of the side "girders" I've modeled with a single larger tube, and then the cross pieces could butt up against flat metal on all 4 sides. I don't know how feasible those large tubes would be for me to obtain though, so this might not be possible.
Do you think this frame design needs more torsional stiffness? One of my fears is that I assemble the frame in one place, move it somewhere else on the floor, and it "racks" out of being a flat plane due to the floor not being flat.
David
My build may be of some interest as my frame design is somewhat similar.
https://www.cnczone.com/forums/diy-c...ded-steel.html
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi Gadget Guy - I would simply not weld or braze anything that does not touch or adjust tube sizes to correct. 100% welding is not necessary....Gaps in welding or brazing welding are your enemy. Peter
for the rails you could get some bright bar (50x12?) and thread and tap to suit the rail. Fit the master to its correct position., Using a machinist level get it level and correct as best as possible on the std section using shims. Then tack it to the section with TIG carefully, then check check check. Once your happy its good, then use epoxy to fill under the bright to the section. Then remove rail and tap right through and complete. Then do same for slave side using master side as reference ? Peter
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi @pippin88
thank you very much for the link! It was very helpful - your machine is quite similar to what I've been envisioning, and it was quite helpful to see something similar that has actually been built. If I may ask, how has your machine performed? Has it lived up to your expectations?
Peter,
Thanks for your input on welding/brazing. I will do my best to ensure good fitup between welded components.
I have considered using precision milled/ground bar in some capacity to establish the flat plane for the rail mounting. However, I'm not confident (just going off intuition here) that I could keep the bar flat enough while attaching it to the frame if I do any sort of welding to it. That being said, I'm not sure it would need to be welded - you could put some metal impregnated epoxy (which I haven't done much research on but I've heard is quite stiff) on the top of the frame, and place the precision bar on top of that, pressing it down by hand into the epoxy. This seems like it could work pretty well for establishing a flat, rigid surface that the rail could be mounted on. This is assuming the precision bar is rigid in it's own right enough to stay flat through the process of pushing it down firmly into the epoxy to ensure a good bond. However, I still don't know how I could ensure that the two rails (which are almost 6' apart from one another) could be kept in the same plane.
With all that said, I'm pretty set on the epoxy leveling at this point. There may be a better way, and I may explore other options in the future, but I think epoxy leveling is the best path forward for me at this time.
Design Update:
I switched out the two frame rails on either side of the frame (the ones made out of sections of 4" square tubing connected by short vertical pieces of tubing) for a single 6"x10"x0.375" rectangular steel tube on either side. A steelyard near me has a piece that's 120" long, so it should be perfect cut in half. This should bring the weight of the frame to around 750 lbs without legs, and gantry around 300, so all in the frame should be somewhere around 1k lbs. I think the single large tube on either side will make fabrication a lot simpler, and easier to fill with epoxy granite or concrete or something down the line if that seems like a good idea.
I've also added some bar stock welded to the frame directly below all the linear rails to help locally stiffen the rail mounts. I've been fleshing out my ball screw mounting, as well as doing some more reading on how other people are doing the Z axis. I think I can probably use a smaller ball screw (like a 1610 or something), which should help with the packaging a lot, as I had been trying to stuff a 25mm ball screw in there up until now.
I'll post some more screenshots once I get the CAD a little more refined.
Thanks all,
David
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
Hi David - I think epoxy levelling is fraught with hazards. I think using epoxy to set a rail or a support plate however is fine. Then drill/tap right through to create a stiff connection.
https://www.diamant-polymer.de/en/
https://www.diamant-polymer.de/en/products/dwh/ info on epoxy fitting of machine parts
On another question from the past. On moving a machine. Many high precision machines have only 3 feet. In this way once setup if you move the machine gravity and the machine balance are the same. They do have outrigger feet but these are just for support not the main load bearers and not involved in set-up.
Cheers Peter S
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
It looks like a very large Z axis to my eyes.I read the original post about the type of work you would like to do with the machine and I can see a potential problem.I have machined quite a lot of jobs for various mouldings-never counted them but certainly hundreds and likely approaching couple of thousand and the vast majority were done on a five axis machine as you could tilt the head to get down the side of a large job and into all the corners.
With a three axis machine you have a line from the tip of the tool to the lower corner of the Z axis backplate that limits where you can go and in the most favourable direction the obstacle is the corner of the motor body.In off road vehicle terms these would be referred to as approach angles and could I ask what value you would expect them to be with something like a 55mm tool length?I raise the point because the big Z axis travel will inevitably allow more wobble at the tool tip than a smaller range and you may be making life unnecessarily difficult.There is also a good chance that you will find yourself manually modifying toolpaths so the tool moves around the job on it's way back to the home position after completing a cut as the direct route could be through the job you just finished machining.
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
I have some comments here with respect to welder frames. Welded frames are use a lot in automation and the stresses welding builds into the frame are real. However there are ways to reduce the impact.
Number one is fit up. You really need precise fit up to avoid pulling in stresses even before welding. By this I mean clamping up your weldment and using clamps to close gaps. This induces stress before you even get started. If you go the welding route critical assemblies need precisely machined components for assembly. This so you can fit them together tightly before welding. Most metal cutting saws don’t come close enough to.the required precision, instead we are talking about the precision one gets from a milling machine. If not a milling machine I’ve spent hours with a file and grinder to get precise 90 degree ends on tubing. This will not completely reduce stress so don’t misunderstand what I’m saying, rather I’m saying you can help you’self here.
Welding shop Bruce’s stress but if you combine precise fit up with low penetration welding you will have components that can resist some of the impacts of welding. This is due to a portion of the joint never melting or even getting excessively hot. There is still stress in the joint but it is resisted to some extent by the fact that the whole joint didn’t melt. Unfortunately machine frame members need a bit of thickness, often more than structurally needed, to do this. Also tacking up and weld sequencing is very important here. Balanced stitch welding helps a lot. Still like above you will get distortion or stress you can’t see.
I tend to disagree with Pete on the distortion caused by brazing mild steel. You will get some due to having to get the metal red hot. As Pete noted steel shrinks after heat cycles, it might not be as bad but it still happens. In fact this property of steel is used to straighten shafts or bend tubing to arbitrary shapes. Bicycles are often using lugs when brazed together which helps significantly in maintaining shape. Note we are talking degrees here of distortion that is lower than welding in many cases but how everything is fitted up and actually brazed together is still a big factor. Another example here is my time spent in a zinc die cast foundry; a little molten (About 860°F) Zinc spilt on a floor plate would warp that plate. That isn’t even brazing temperature. It isn’t a question if it will happen but rather to what degree. I’m not sure which would be worse in a structural weld in mild steel, MIG welding or brazing.
The issue with the corner radius and the high fill required to close it up is very real. I’ve seen and worked on many frames where the frame looked tree like as each piece being welded on got smaller in dimension. This so that each new branch could be welded on without the need to address the radius gap. Of course this is where a milling machine might be of value as you can notch the beams to wrap around that radius. Still the common practice is to mill the ends square and use tube sizing to avoid problematic welding.
There is lots of talk about epoxy leveling in these forums and how well it works. It is pretty obvious that it does work for at least some of the expectations people have for their machine. Exactly how it was done on a specific machine is also a big factor. What doesn’t get much commenting here is epoxy grouting. Grouting is a harder resin composite that gets pumped into flanged and the like to finalize a joint that has been lined up. Assuming proper flange sizing I see nothing wrong with such epoxy use. Basically frame members with Jack screws in their mounting flanges are used to bring the commm
Re: Building Steel Frame Gantry Router for wood, MDF, and maybe more
I think epoxy leveling can work well, but it has its challenges.
One is surface tension - this means you need a thick pour. It's just like water on a flat surface, a small amount will form drops and you need enough to cover the whole surface fully. Epoxy is thicker and has a higher surface tension than water.
So you need quite a thick pour, probably 6mm or more.
It's a decent solution for building a good wood router at home, without needing to send out for machining etc.
My machine performs pretty well. Handles a 6mm bit, 6mm deep in MDF at 6m/min no worries.
I'm going to do some modifications in the future.
1. The gantry beam may not be 100% straight. Got a surface plate coming so I'll be able to check.
2. Epoxy granite bulkheads in the gantry. Will help to stiffen and reduce the ringing noise.
I haven't ever got around to finishing a number of features:
4th axis
Clamping for vertical boards on the front to machine board ends.