Building Steel Frame Gantry Router for wood, MDF, and maybe more

[teamloks1]

thanks for your insight peter i might need to rethink my whole idea again

[wizard]

yeah thanks well need to stick with extrusion then since there no place here to do that

Extrusions are not anymore flat than steel tubing. This is pretty much a function of the supplier. Now that being said extrusions often surpass the manufacture straightness specs but so does steel tubing. Some suppliers like Misumi will fatten certain extrusions for you which can be worth the money. In any event this comes back to what are your expectations, for many people extrusions can be good enough as can wood or steel components.

In any case a surface that hasn’t been machined flat can be a real hassle when it comes to mounting linear rails. You can easily find yourself in a situation where every time you torque down the mounting screws a saddle will bind up. This will be due to the rails mating up with the imperfections in the mounting surfaces. The stiffer you make things the more of a problem these fit up issues become.

[wizard]

thanks for your insight peter i might need to rethink my whole idea again

There are so many options when it comes to building a router that design paralysis can set in. Your best bet at building something that meets your needs is to detail what those needs are and what you need to get there. Once you know what you will need you can decide if a given build material will get you there.

Plywood or other plant based materials are perfectly viable for smaller wood processing machines. The first negative here is exposure to moisture if you should happen to need a coolant. As such I can’t suggest a wood framed machine if you want to process aluminum or even some plastics where a coolant / lube would commonly be used. That doesn’t even get into stiffness required for machining harder materials well.

One of the bigger issues with a router build is the gantry beam which often does not get enough thought. If you have limited resources (to fabricate a beam) this is where extrusions can actually be of great value if you can accept the limitation of what are rather light weight extrusions. An alternative is to find and buy used industrial equipment that gets you 95% of the way there. There is a class of machines call mold machine extractors (or robots) that have beams that would make an ideal gantry beam. If you find one of these used they often have many usable parts that can be repurposed. In any even the gantry on these are often cantilevered a considerable distance so the beams are stiffly made. There are other possibilities when it comes to industrial equipment but the gantry beams on mold machine extractors are almost a perfect fit for our needs.

The idea here is out of the box thinking! If you don’t have a local machine shop able to handle machining a steel beam then you need to look at the alternatives. Refactoring used machinery is always worth considering if you find something that fits the bill. Frankly a clapped out CNC router is another way to a low cost no machining solution.

I might comment on Pete’s suggestion to build a series of machines to get you from Zero to where you want to be. Unless you like building machines as a hobby I wouldn’t do more than two. Build one to learn on is a good idea. This especially if you have both machine building skills and machining skills to develop. Learn enough so that you can spec out a machine that will actually do what you need to do. Even if you need to contract some of the build out it will likely be cheaper than building A series of progressively better machines and a lot quicker. Do however make that first machine the best you can. For one the machine may be around longer than you expect Second as you learn you want a machine that can keep up with your progress.

There are arguments on both sides, if you have the skills and resources doing exactly what you need on the first build can make sense. The effort required to do a machine build is often underestimated so consider that.

[teamloks1]

There are so many options when it comes to building a router that design paralysis can set in. Your best bet at building something that meets your needs is to detail what those needs are and what you need to get there. Once you know what you will need you can decide if a given build material will get you there.

Plywood or other plant based materials are perfectly viable for smaller wood processing machines. The first negative here is exposure to moisture if you should happen to need a coolant. As such I can’t suggest a wood framed machine if you want to process aluminum or even some plastics where a coolant / lube would commonly be used. That doesn’t even get into stiffness required for machining harder materials well.

One of the bigger issues with a router build is the gantry beam which often does not get enough thought. If you have limited resources (to fabricate a beam) this is where extrusions can actually be of great value if you can accept the limitation of what are rather light weight extrusions. An alternative is to find and buy used industrial equipment that gets you 95% of the way there. There is a class of machines call mold machine extractors (or robots) that have beams that would make an ideal gantry beam. If you find one of these used they often have many usable parts that can be repurposed. In any even the gantry on these are often cantilevered a considerable distance so the beams are stiffly made. There are other possibilities when it comes to industrial equipment but the gantry beams on mold machine extractors are almost a perfect fit for our needs.

The idea here is out of the box thinking! If you don’t have a local machine shop able to handle machining a steel beam then you need to look at the alternatives. Refactoring used machinery is always worth considering if you find something that fits the bill. Frankly a clapped out CNC router is another way to a low cost no machining solution.

I might comment on Pete’s suggestion to build a series of machines to get you from Zero to where you want to be. Unless you like building machines as a hobby I wouldn’t do more than two. Build one to learn on is a good idea. This especially if you have both machine building skills and machining skills to develop. Learn enough so that you can spec out a machine that will actually do what you need to do. Even if you need to contract some of the build out it will likely be cheaper than building A series of progressively better machines and a lot quicker. Do however make that first machine the best you can. For one the machine may be around longer than you expect Second as you learn you want a machine that can keep up with your progress.

There are arguments on both sides, if you have the skills and resources doing exactly what you need on the first build can make sense. The effort required to do a machine build is often underestimated so consider that.

since im stay near china i can get full steel frame with gantry everything for around 1000-1500usd but the problem is my location i cant bring those big machine to my place because of limited space for transport so that why im going to assemble everything .and for my extrusion supplier they say can get around 0.05mm tolerance for 5 meter extrusion

[peteeng]

Hi Loks - That sounds like an amazing tolerance for a small extrusion. Gravity would bend it more over that distance. But if they can do it go for it. I came across this photo of a riveted structure, should be more of them. I think I'll do this for my next bench in aluminium It also gives me an idea on how to remove a weld in my new column design, excellent... Peter

[machinedude]

I came across this photo of a riveted structure, should be more of them. I think I'll do this for my next bench in aluminium It also gives me an idea on how to remove a weld in my new column design, excellent... Peter

i found it interesting about the welds and internal stresses vs bolted designs. the build i have going at the moment for myself is completely bolted together. as a matter of fact i have a small fortune tied up in SHC screws as a result

i think it might have been you talking about the vibration dampening of bolted vs welded too?

[peteeng]

Hi Machine dud - any photos, we like to watch. Yes I have discussed lots about bolts, rivets, brazing and other ways to build machine frames. The friction in bolted joints make it damper then a welded joint...Peter

[machinedude]

Hi Machine dud - any photos, we like to watch. Yes I have discussed lots about bolts, rivets, brazing and other ways to build machine frames. The friction in bolted joints make it damper then a welded joint...Peter

One of the things i wanted to avoid was welding to avoid having things twist into a pretzel as a result. I even went as far to machine the materials straight on the flat stock i used. Doweled the important stuff in place and made sure things were squared.

I have pictures just not on here unfortunately.

[thegadgetguy]

Looks like there has been some interesting discussion since I last visited. Anyway, not a whole lot of progress, but construction has begun. I changed the design to make the side rails one large piece, bought steel, and have started assembling the frame.

[thegadgetguy]

Can anyone comment on the necessity of double ball nut ball screws? BST automation only has the double ones in 5mm pitch, which is a lot smaller than I wanted to go pitch wise. Should I try and find another supplier that can do double ball screws in 10mm or larger pitch, or should I just get single ballnuts for now and worry about upgrading later if it seems necessary? If double ballnuts are deemed necessary, does anyone have supplier recommendations that can do 10mm pitch or greater?

At this point I'm leaning towards just getting the single ballnut screws, as once I get the rail beds welded on my frame I'm pretty much stuck until I have the screws in hand

Any insight is appreciated.

[peteeng]

Hi Gadget - A lot of your questions would be self answering if you define what you need to do and hence the accuracy /tolerance needed. So think about and understand what you want to do. Accuracy costs $$$, double nuts are more expensive and more accurate then single nuts. But also the screw itself is made to a tolerance so a good single is as good as a poor double as its running on the same tolerance screw. Upgrading machines is easier said then done. Double nuts take up room and so you lose envelope. They may need some space where you don't have space allowed for. Everything's connected, now how can I get the Corona virus into this discussion? Peter

[mactec54]

Hi Gadget - A lot of your questions would be self answering if you define what you need to do and hence the accuracy /tolerance needed. So think about and understand what you want to do. Accuracy costs $$$, double nuts are more expensive and more accurate then single nuts. But also the screw itself is made to a tolerance so a good single is as good as a poor double as its running on the same tolerance screw. Upgrading machines is easier said then done. Double nuts take up room and so you lose envelope. They may need some space where you don't have space allowed for. Everything's connected, now how can I get the Corona virus into this discussion? Peter

Using a double nut does not make it more accurate the Grade of Ballscrew is how accurate a Ballscrew will be, a double nut is only needed for high loading and the low accuracy ballscrews that have a lot of backlash with a single nut, the C7 ballscrews can be better with a double nut to eliminate backlash the accuracy of the ballscrew does not change it is still a C7

[mactec54]

Can anyone comment on the necessity of double ball nut ball screws? BST automation only has the double ones in 5mm pitch, which is a lot smaller than I wanted to go pitch wise. Should I try and find another supplier that can do double ball screws in 10mm or larger pitch, or should I just get single ballnuts for now and worry about upgrading later if it seems necessary? If double ballnuts are deemed necessary, does anyone have supplier recommendations that can do 10mm pitch or greater?

At this point I'm leaning towards just getting the single ballnut screws, as once I get the rail beds welded on my frame I'm pretty much stuck until I have the screws in hand

Any insight is appreciated.

A double Nut is not needed unless you are using a low accuracy Ballscrew and that will help only with backlash

If you can get C5 Rolled screws they should be a better choice than the C7 that are everywhere 10mm pitch for a build like yours would be a minimum

[thegadgetguy]

Wow, I thought I had updated this more recently, but apparently not. I've made some more progress on the frame. I also ordered linear rails and ball screws from BST Automation. They came recommended by some people on here. Fred was very good to work with, and very patient with me haha. I paid an arm and a leg for shipping due to the Coronavirus, but everything arrived intact.

I bought 20mm linear rails, and 25mm ball screws with 25mm pitch.

I'm currently working on sourcing epoxy, looking at Precision Epoxy's SC-15P epoxy which sounds like it is designed for this application. Also figuring out ball screw mounting, and trying to figure out how to drill all the holes for the linear rails accurately enough that they can be bolted on straight and parallel to one another. I have a few ideas, but will need to do some research.

[thegadgetguy]

Using a double nut does not make it more accurate the Grade of Ballscrew is how accurate a Ballscrew will be, a double nut is only needed for high loading and the low accuracy ballscrews that have a lot of backlash with a single nut, the C7 ballscrews can be better with a double nut to eliminate backlash the accuracy of the ballscrew does not change it is still a C7

A double Nut is not needed unless you are using a low accuracy Ballscrew and that will help only with backlash

If you can get C5 Rolled screws they should be a better choice than the C7 that are everywhere 10mm pitch for a build like yours would be a minimum

Hi Gadget - A lot of your questions would be self answering if you define what you need to do and hence the accuracy /tolerance needed. So think about and understand what you want to do. Accuracy costs $$$, double nuts are more expensive and more accurate then single nuts. But also the screw itself is made to a tolerance so a good single is as good as a poor double as its running on the same tolerance screw. Upgrading machines is easier said then done. Double nuts take up room and so you lose envelope. They may need some space where you don't have space allowed for. Everything's connected, now how can I get the Corona virus into this discussion? Peter

Thanks for the comments guys, I really appreciate it. I ended up going with the single ballnut, 25mm pitch. The C7 grade were not expensive, so ended up going with them. I will be interested to see how they end up working out, or if I'll need to upgrade to something better.

[thegadgetguy]

It's been awhile, but progress is still being made, albeit slowly. I wasn't satisfied with the torsional stiffness of the frame so I added some bracing.

[thegadgetguy]

Okay this thread is super out of date, but here is a quick video of the Y axis moving.

[thegadgetguy]

Thread is still super out of date but I did get this thing working (though not finished). Sadly I moved and it is currently in storage. Will hopefully update this thread eventually.

[harryn]

Looks like success to me.

[routalot]

If it works,you've succeeded and the rest is just tidying up.