L3viathan's Steel Router

[l3viathan]

I've been reading this site and a few others for almost five years now, and after all that research, I've finally settled on a design that I think will do what I need. I arrived at this design based on the machines and capabilities I have, as well as the parts I already have and what will fit in my budget. I plan to pour leveling epoxy the way fluxion did for the rails and support mounts. I'd like to pour the two gantry support pads first, then weld the Y tubes on, although i'm not sure if the epoxy will make it through the de-stressing temperatures. For an idea of scale, I have a 30" x 30" aluminum table, with full coverage (so I can surface it in place).



Similar Threads:

• New Machine Build- 4' x 6' Steel CNC Router
• Steel Router Mount for Milwaukee 5625 Router - No Longer Available
• diy cnc router in steel, any one try it yet??

[wizard]

Wow it is like you are reading my mind. This is similar in concept to ideas that I trash around in my mind, thinking one day all the bills will go away. I wasn't actually thinking of a square machine and had a slightly different arraignment of the tubes. I hope to see you realize this machine soon. The only concern I have is with the x axis tubes which look a bit thin in the render.

[l3viathan]

All of the tube is .25" wall. I have had concerns about thread engagement, since i'm using 8mm bolts on the columns, although, that is the same thickness as a standard 8mm nut...or is stiffness the concern you had?

The longer tubes that carry the table are 4" x 4" x 54" long. The gantry is 8" x 8" x 48". The columns, and front support are 6" x 6" and the large tube under the columns is 10" x 6" x 54". I will be filling all of the cavities with oiled sand to reduce resonance. The best price I have so far on the steel is $500, although i'm still waiting on another quote before I buy.

[pippin88]

With epoxy make sure you pour a good thick layer, about 1/4" to ensure self leveling.

You will need to pour your epoxy pads much bigger than the rails due to meniscus.

Consider bringing the table rails in a little, so that the table is supported not just on the edge.

By having your table support beams on top of the other beams you have to have a taller gantry, which means more flex. I would suggest a single plane bottom structure.

If you want to cut 4 foot, you'll need rails longer by how wide the Z axis is. E.g 54" rails to cut 48" (8" wide z).

[ger21]

Most epoxies fail at 200-250 degrees F. You don't want to weld after pouring your epoxy.

[l3viathan]

With epoxy make sure you pour a good thick layer, about 1/4" to ensure self leveling.

You will need to pour your epoxy pads much bigger than the rails due to meniscus.

Consider bringing the table rails in a little, so that the table is supported not just on the edge.

By having your table support beams on top of the other beams you have to have a taller gantry, which means more flex. I would suggest a single plane bottom structure.

If you want to cut 4 foot, you'll need rails longer by how wide the Z axis is. E.g 54" rails to cut 48" (8" wide z).

I was planning to trim the edges of the epoxy, although I could probably pour larger pads without interfering with anything if it would be simpler to do so.

I wasn't sure about how much overhang I could get away with on the table, so the rails are 24" center to center. On a 30" table, I could probably pull them in a bit. The columns are 14.5" (including epoxy), although I may lower that a bit depending on how thick the table ends up being. I'm shooting for 7" of clearance, so I can mount my low profile vice and still have room.

As for cutting 4', I only planned to be able to cut the size of the table, so 30" x 30" is the limit. Going any bigger than that increases cost and difficulty exponentially for me.

[l3viathan]

Most epoxies fail at 200-250 degrees F. You don't want to weld after pouring your epoxy.

As I think about it, even if there was an epoxy would take the high temps, it's probably counter-productive to pour leveling epoxy before stress relieving the assembly anyway.

[pippin88]

If you make the base all one level you can do a single epoxy pour for the rails, ballscrew mounting points and gantry upright pads.

Don't try to trim the meniscus off and then mount to that same area, it won't be accurate. You want to be mounting to untouched epoxy. I would leave a minimum 1/2" either side of the rail. Epoxy ain't cheap but it ain't that expensive.

[l3viathan]

If you make the base all one level you can do a single epoxy pour for the rails, ballscrew mounting points and gantry upright pads.

Don't try to trim the meniscus off and then mount to that same area, it won't be accurate. You want to be mounting to untouched epoxy. I would leave a minimum 1/2" either side of the rail. Epoxy ain't cheap but it ain't that expensive.

I'll do the volume calculations, and see how much extra I'll need for that. It may be easier to do a larger pad, as the mold shell can include the edge radii, rather than having to be formed over them.

I can see the benefit of making it a single level, although I'm not sure that is within my capabilities. It would mean I'd need a larger surface to keep all of that flat for welding, and some interesting cutting to get everything to fit together.

[ger21]

If you're going to actually have the entire thing stress relieved, why not try to find someone that can grind the tw
he rail mounting surfaces of your two assemblies. Then you don't need to mess with the epoxy. If I recall correctly, another member here had his welded assembly ground flat for around $300.

I would weld the Y tubes to the base tubes, and have them ground flat. Have the gantry tube ground flat.

THen shim and use an epoxy grout when bolting the gantry down, rather than mounting your gantry uprights to an epoxy pour.

[l3viathan]

If you're going to actually have the entire thing stress relieved, why not try to find someone that can grind the tw
he rail mounting surfaces of your two assemblies. Then you don't need to mess with the epoxy. If I recall correctly, another member here had his welded assembly ground flat for around $300.

I would weld the Y tubes to the base tubes, and have them ground flat. Have the gantry tube ground flat.

THen shim and use an epoxy grout when bolting the gantry down, rather than mounting your gantry uprights to an epoxy pour.

If that were an option, I would go that route. I live in the very rural south-western New York state. I haven't found a place within a reasonable driving distance that could do it (or wanted to). Most places I've contacted never even responded.
I realize that having the surfaces ground would be the ideal way to go about it, so I'll take another look at a larger distance.

My thought process on this is to spend a large part of my budget on the frame, that way even though I'm starting out with steppers, I can upgrade to servos, higher precision ball screws, and a larger spindle down the line without having to start from scratch. Don't get me wrong though, I am buying components that will allow me to make the parts I want within tolerances I find acceptable. If I never upgrade this machine, it'll do what I want for the foreseeable future.

[wizard]

All of the tube is .25" wall. I have had concerns about thread engagement, since i'm using 8mm bolts on the columns, although, that is the same thickness as a standard 8mm nut...or is stiffness the concern you had?

Actually at the time I was thinking something entirely different. I'm wondering if a 4" square beam is big enough cross sectional area. Especially off the back where you have unsupported beam.

As for wall thickness in an ideal world you would be slightly thicker than the bolt diameter. At the face where you bolt the rails you would normally want to counter bore or counter sink so that the first thread starts below flush. So you would loose half the thickness (pitch) of the thread form. With epoxy leveling you probably don't need to worry.

The longer tubes that carry the table are 4" x 4" x 54" long. The gantry is 8" x 8" x 48". The columns, and front support are 6" x 6" and the large tube under the columns is 10" x 6" x 54". I will be filling all of the cavities with oiled sand to reduce resonance. The best price I have so far on the steel is $500, although i'm still waiting on another quote before I buy.

[l3viathan]

After Ger21's last reply, I called around a bit more, and found a place that has the capability to machine and grind the surfaces that need it (not sure how this will impact budget or schedule yet). Since that option is now available, I'm entertaining the idea of welding plate to the mounting areas, and forgo the epoxy. This will give me double the thread engagement I would have had otherwise.

The table supporting tubes extend about 9" past the base support tubes, and with the linear bearings being 6" on center from the ends of the table, at maximum travel, support will only be 3 inches past the larger supports. In the end, I think it's a trade-off between larger rail beams, and thus taller columns, or accept some amount of deflection, and have shorter columns.

What is the general consensus around mounting the ballscrew on top of the gantry? I think it would put more force on the gantry bearings, however I'm having an issue with pushing the spindle further away from the gantry to keep the screw between the linear rails.

[wizard]

After Ger21's last reply, I called around a bit more, and found a place that has the capability to machine and grind the surfaces that need it (not sure how this will impact budget or schedule yet). Since that option is now available, I'm entertaining the idea of welding plate to the mounting areas, and forgo the epoxy. This will give me double the thread engagement I would have had otherwise.

That is good news, having access to machine shop will allow you more design freedom. All you need is to be willing to pay his going rate.

The table supporting tubes extend about 9" past the base support tubes, and with the linear bearings being 6" on center from the ends of the table, at maximum travel, support will only be 3 inches past the larger supports. In the end, I think it's a trade-off between larger rail beams, and thus taller columns, or accept some amount of deflection, and have shorter columns.

With access to a machine shop you can consider a redesign that allows more uniform tube sizing.

What is the general consensus around mounting the ballscrew on top of the gantry? I think it would put more force on the gantry bearings, however I'm having an issue with pushing the spindle further away from the gantry to keep the screw between the linear rails.

Running down the center of the saddle is the normal accepted way to drive an axis. IF keeping the part height low is important to you you can always do a 3 part gantry beam. Another option would be a thick gantry beam saddle and to bore a hole through it to pass the leadscrew. From the standpoint of conserving space though the leadscrew nut should sit underneath the saddle.

There is no perfect answer to a gantry design. People do box beams because they are easy to understand and analyze. However I'm not to concerned about stick out for a reasonable screw size.

[l3viathan]

The design changes I'd like to make will be minimal, but hopefully improve the machine proportional to the extra expense.

Funnily enough, it was the motor causing clearance problems. The distance to center of the motor was larger than the distance to center of the ballscrew end supports. I can of course shim up the screw mounts to make room for the motor, but I hate pushing the spindle even further away.

I suppose that now I have a fairly good idea of how much weight i'll be pushing around, I need to decide on motors and ballscrews. That will determine how much clearance I need for the Z mount. My first though was a nema24 566 oz-in stepper with a 1610 ballscrew. That would give a theoretical ~100IPM at 300oz-in (254rpm). That seems fast enough assuming a 2.2kw spindle can provide a decent removal rate at that speed.

My second thought was a 600 oz-in nema34 stepper with a 2010 ballscrew. This would allow me to move to a 400w servo in the future without having to worry about critical speed too much. The other reason was that a nema24 motor looked ridiculously small compared to the size of the machine (not the best reason I know).

I haven't found any hard and fast rule on how much torque is required from a stepper at any particular cutting speed though. With so many factors influencing cutting performance, I realize there's likely no catch-all formula, but even anecdotal ballpark figures would be a good starting point.

[pippin88]

100ipm is slow once you've used a machine that can go faster.

I had the same problem with my Y axis motor getting in the way. I cut a rebate in the gantry for it. The most sensible solution is in fact to buy a longer ball screw and have the stepper sticking out from the end of the gantry. Increasingly the length of a screw costs very little.

[l3viathan]

100ipm is slow once you've used a machine that can go faster.

I had the same problem with my Y axis motor getting in the way. I cut a rebate in the gantry for it. The most sensible solution is in fact to buy a longer ball screw and have the stepper sticking out from the end of the gantry. Increasingly the length of a screw costs very little.

That is a concern I had. I have no point of reference for how fast a machine can or should/could run. I just watched a Youtube video of a Datron mill machining an aluminum electronics housing at 6m/min (236IPM)...that looked fast to me. From what I've read, I think it will be the spindle that limits speed more than steppers, but I could be way off.

A longer ball screw was something I considered as well. Adding another 100mm brings me to the end of the gantry where I can mount any size motor I like, and the price difference as you noted is negligible.

[wizard]

The design changes I'd like to make will be minimal, but hopefully improve the machine proportional to the extra expense.

Funnily enough, it was the motor causing clearance problems. The distance to center of the motor was larger than the distance to center of the ballscrew end supports. I can of course shim up the screw mounts to make room for the motor, but I hate pushing the spindle even further away.

You could belt drive the leadscrew. I'm not sure why people avoid belt drives but I've seen the successfully drive leadscrews on everything from high precision lathes to injection mold machines. This would mean a slightly longer leadscrew and buying belts and pulleys instead of anti backlash couplings. The motor mount is different but you need one no matter what so that isn't a factor.

Besides moving the motor out of the way you can run the leadscrew / motor at any reasonable ratio you want. This can help with getting a good match between the motor and the leadscrew.

I suppose that now I have a fairly good idea of how much weight i'll be pushing around, I need to decide on motors and ballscrews. That will determine how much clearance I need for the Z mount. My first though was a nema24 566 oz-in stepper with a 1610 ballscrew. That would give a theoretical ~100IPM at 300oz-in (254rpm). That seems fast enough assuming a 2.2kw spindle can provide a decent removal rate at that speed.

That is rather slow. Don't forget too that Rapids are important in many machining scenarios.

Of course steppers can only take you so far especially if you want precise positioning. Given that I'd still try to configure an option that gets you above 200 ipm. There are actually techniques that use even faster feeds but it probably isn't worth trying to achieve those rates on a first build stepper based machine.

My second thought was a 600 oz-in nema34 stepper with a 2010 ballscrew. This would allow me to move to a 400w servo in the future without having to worry about critical speed too much. The other reason was that a nema24 motor looked ridiculously small compared to the size of the machine (not the best reason I know).

Small is good.

I haven't found any hard and fast rule on how much torque is required from a stepper at any particular cutting speed though. With so many factors influencing cutting performance, I realize there's likely no catch-all formula, but even anecdotal ballpark figures would be a good starting point.

I don't have an answer for that one as you note it is highly variable. You would need to calculate how much torque is required on the leadscrew to reach a certain force on the cutter. You might look into a speed and feeds calculator or manufactures data sheets to see if there is any useful data to be had.

[wizard]

That is a concern I had. I have no point of reference for how fast a machine can or should/could run. I just watched a Youtube video of a Datron mill machining an aluminum electronics housing at 6m/min (236IPM)...that looked fast to me.

It looks fast until you are stating the twiddling your thumbs waiting for s part to come out. The reality is that that may very well be fast for a stepper based machine.

From what I've read, I think it will be the spindle that limits speed more than steppers, but I could be way off.

Steppers are a big factor in the overall performance of a machine that is why so many switch to servos. Steppers are perfectly fine on lower end machines or machines that can get buy with low precision.

The speed at which you machine varies with materials obviously but you generally manage spindle power required by varying the depth of cut. In most cases you want to use the spindle in such a way that tool life and cut quality are managed, which generally means finding the right feed rates and RPMs. At least that is how many approach it, the spindle will limit how much material you remove at a time not so much the ideal feed rate.

A longer ball screw was something I considered as well. Adding another 100mm brings me to the end of the gantry where I can mount any size motor I like, and the price difference as you noted is negligible.

It also gives you far more flexibility in mounting.

[ger21]

I suppose that now I have a fairly good idea of how much weight i'll be pushing around, I need to decide on motors and ballscrews. That will determine how much clearance I need for the Z mount. My first though was a nema24 566 oz-in stepper with a 1610 ballscrew. That would give a theoretical ~100IPM at 300oz-in (254rpm). That seems fast enough assuming a 2.2kw spindle can provide a decent removal rate at that speed.

My second thought was a 600 oz-in nema34 stepper with a 2010 ballscrew. This would allow me to move to a 400w servo in the future without having to worry about critical speed too much. The other reason was that a nema24 motor looked ridiculously small compared to the size of the machine (not the best reason I know).

I haven't found any hard and fast rule on how much torque is required from a stepper at any particular cutting speed though. With so many factors influencing cutting performance, I realize there's likely no catch-all formula, but even anecdotal ballpark figures would be a good starting point.

What's this machine going to be cutting? If you're cutting wood, you really want to be able to reach at least 400ipm, imo. And even if not cutting wood, you'll want rapids to be at least that.
You don't want to design your machine so that your steppers only run at 250rpm.

I would go with the 2010 screws, so if you upgrade to servos later, you won't have as much concern about the screws whipping, since most servos will be capable of 2000-2500rpm. ANd if you do plan on upgrading, you definitely want to design in a belt drive, so you can have some reduction from the servos.

The biggest factor in torque requirements is acceleration. It takes far more torque to accelerate the machine than to do the actual cutting. But since you have a moving table, you won't need as much force as you would if you were moving the gantry.
You can probably safely figure on no more than 50-70lbs of cutting force.