-
7 Attachment(s)
150x100cm Steel frame router build
After 4 months since I first thought about building my CNC router, here we go. I'm still at brainstorming/CAD phase so if you have any thoughts and ideas I would love to hear them.
Working surface will be 150x100cm and 18cm Z height (60"x40"x7"). I'm aiming at aroung 10m/min (400IPM) cutting speed (or more if I manage to solve the ballscrew whip problem).
I intend to cut hardwoods and some aluminium. Budget is around 3000€.
Since I have a welding workshop and have access to a milling machine I'll be making the router from steel. I'll mating surfaces and surfaces where I'll bolt linear rails will be machined.
I'm making the Y axis beams from 120x40mmx3mm tubing and I'll weld a strip of 30x15mm steel to have more meat for threading into. My idea is to make the router easily removable from the frame. I'll be bolting the Y beams to the frame with 4 bolts on each corner. Hopefully that will also give me enough room to shim the rails to be in the same plane and in a way disconnect the router from the frame so I'll have to worry less about the welding distortion in the frame. After the router is done I'll resurface the table and any discrepancy in the frame will be solved by that (I hope that is).
I'll probably use 25mm rails since I've got 25mm carriages for cheap.
I thought to use 1610mm ballscrews but since the price isn't that much higher I'll use 2010mm screws. That gives me around 1000rpm before I hit the critical speed in a fixed-supported system. And this is where I have a couple of questions for you guys.
Since the ballscrew is so efficient and almost "friction-less", does it really heat up that much during heavy use? I have no access to a router in a real world so I can't check for myself. The reason I'm asking is if the screw doesn't heat up significantly than it won't expand much and fixed-fixed mounting is possible and that give me 1500rpm before critical speed. Can you give me any feedback on this please.
Another idea I've found here on the forum is to make a spring actuated plastic support for the screw in the middle to prevent it sagging (I attached photos bellow).
I don't intend to hit critical speed but it would be good if I could make even faster machine although it isn't crucial for me.
For X axis beam I'll use 200x80x5mm tubing. In another thread someone mentioned I should use square tubing instead because the cutting force is pushing against the narrower part of the beam. That makes sense but using 200x200 tubing will eat up router space pretty fast. That made me experiment with Fusion360 simulation workspace.
I don't have any idea how high the cutting forces are but I've found on the forum someone mentioned 300N so I used that. In my static simulation where the Z axis is in the middle of X axis and it is at it'slowest point I've got around 20 microns of deflection. I was getting around 30 microns before I've added some diagonal bracing inside the 200x80 beam and side support for spindle mounting plate. (30 micron is 1 thou)
Is this "acceptable" deflecton or should I consider going with 200x200 tubing? At this point the gantry will be over 50kg (110lbs) so I'm afraid to add even more weight.
The Z axis is 25mm aluminum plate and for the spindle mounting plate I'll bend sheet metal around the spindle and bolt it to the plate to make it more rigid.
I found some hybrid NEMA34 8Nm motors (~1000oz/in) and I'll use 2 on the Y axis, one on X and maybe one NEMA 23 4Nm for Z axis since I'll use 1605 screw for it. On MYCNCUK forum I've found spreadsheet for calculating motor size and for 60kg/120lbs gantry I need ~4Nm or torque and since I'm using double 8Nm motors I think I'll have some headroom. (the calculation DID include acceleration load but it is hidden inside the spreadsheet formulas and I can't find what value for acceleration they've used)
I thought about connecting motor with a belt. That would allow me to make a 2:1 reduction since these motors only have 3Nm of torque at 1000rpm and almost 5Nm at 500rpm.
This is it for now. If you had enough patience to read all through this than thank you. If you have any questions/ideas feel free to share. Thanks again.
This is the stepper I intend to use: https://www.vallder.com/en/c/292/38/...hnical-details
-
Re: 150x100cm Steel frame router build
Another question I have in mind, since I'll weld 15mm thick flat bar to the beams for attaching linear rails (stitch weld, not full length) do I need to normalize steel before machining? I'll just skim the surface so I won't remove a lot of material.
I know your answer will be yes but I'm interested in how much movement there can be after it cools completely?
Is normalizing really nessecary for a router? I haven't found any even approximate numbers regarding movement after welding and cooling.
What are your thoughts on this? Have you normalized your weldments (if that's the right word)?
Sent from my SM-G970F using Tapatalk
-
Re: 150x100cm Steel frame router build
Is the X axis of your milling machine really long enough to mill those 150cm beams in one go? If so, that's great; I don't think you need to "normalize" afterwards. Nobody here does that when building routers, since nobody (hardly anyone) has heat-treatment kilns big enough. But those beams aren't perfectly straight to start with, and all that welding will warp them further, so you do need to flatten whatever you intend to bolt rails to. If your mill's not big enough, what's Plan B?
-
Re: 150x100cm Steel frame router build
I don't own a big mill myself but I have a friend who does and he'll do it for me.
If I didn't have access to a mill I would use epoxy to level the surface, the process seems easy enough once you find out how to do it properly.
If I were to weld the Y axis beams to the frame I would use epoxy since you can level both beams at the same time and avoid finicking with shims later on.
Thanks for your reply.
Sent from my SM-G970F using Tapatalk
-
Re: 150x100cm Steel frame router build
50kg is not an issue at all for two appropriately sized steppers.
200x200 won't use up any travel, just hang the extra out the back. Travel is determined by your bearing spacing which should be decided on based on a compromise - the wider the spacing the better, but the less travel you get from a set rail length.
I think fixed-fixed is too problematic. Only high end commercial machines would be using this. You need very rigid mounting (not standard BK blocks) and the screw needs to be tensioned.
If you are worried about rpm, get higher pitch screws. I use 15mm on one axis and 16mm on the other two.
I use 1616 screws from BST Automation on AliExpress. The quality seemed higher than my previous (multiple) 1605 and 2005 screws.
-
Re: 150x100cm Steel frame router build
Normalising / stress relieving - I haven't found a good answer on whether something will keep moving after machining.
I believe that most of the movement will happen when you machine the welded part.
One solution to avoid stress relieving is to:
Rough machine, leaving lots of allowance.
Unclamp and reclamp.
Remachine flat.
The more cycles the better I think. It should move less each time.
Important to not clamp / hold down in a way that pulls the welded part down. E.g clamp from the sides. And need to release the clamp to allow the movement then reclamp and remachine.
-
4 Attachment(s)
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
pippin88
50kg is not an issue at all for two appropriately sized steppers.
200x200 won't use up any travel, just hang the extra out the back. Travel is determined by your bearing spacing which should be decided on based on a compromise - the wider the spacing the better, but the less travel you get from a set rail length.
I think fixed-fixed is too problematic. Only high end commercial machines would be using this. You need very rigid mounting (not standard BK blocks) and the screw needs to be tensioned.
If you are worried about rpm, get higher pitch screws. I use 15mm on one axis and 16mm on the other two.
I use 1616 screws from BST Automation on AliExpress. The quality seemed higher than my previous (multiple) 1605 and 2005 screws.
You mean to let the extra hang like in the picture I've attached? Do I gain anything by doing that since I won't be moving the carriages to be below the beam? I intend to use 40mm (1,1/2") thick aluminum riser blocks to connect the beam with carriages and the spacing between the carriages is 10cm (25cm is total length of the riser).
I wanted to make my own mounting blocks since those Chinese blocks use deep-groove ball bearings and I wanted to change them to angular contact anyway. Since I'm doing that I can make the whole block anyway.
Yes, I wanted to tension the screw to remove the sag in the middle. I don't think it takes a lot of force to tension the screw to make a difference, what do you think? A hundred, two hundred pounds of force maybe?
Or is making the plastic support the easiest option? I haven't found any evidence so far that it wouldn't work...
Buying on Aliexpress is not an option for me since the shipping cost is usually way more than the price of the item I was buying. I must shop locally for now but once I gather enough money for my next machine I'll buy proper high quality stuff.
Thanks.
-
3 Attachment(s)
Re: 150x100cm Steel frame router build
I did some quick simulation in Fusion360 to see if there is any noticeable advantage in having 200x200x6mm beam vs 200x100x6mm. I used 16cm (6") Z axis and a force of 300N and I've capped the ends to reduce diagonal distortion.
200x100mm beam had 34 microns of deflection.
200x200mm beam had 18 microns of deflection.
200x200mm beam that's cantilevered out 10cm had 21 microns.
30 micron is 1 thou.
Seeing this, do you think it's worth using 200x200 beam instead of 200x100 one? Is it worth loosing 10cm of travel to fully support the beam? Of course the answer depends, but for a wood/aluminium router?
-
Re: 150x100cm Steel frame router build
If you want to cut aluminium I would use the 200x200. The change nearly halves the deflection. Steel is cheap compared to other costs of a build. Do it now when it is easy.
I can't tell for sure, but it looks like you've got your spindle lined up with the middle of the bearings? You don't have to have it like this. Eg, you could have a bit less cantilever. You could do it based on centre of gravity.
To fully understand the effect of bearing placement / cantilever with the 200x200, you should model the ends more like your earlier CAD pictures. Include the full length of the 40mm spacer (I assume you applied a rigid/fixed constraint on the bottom surface in the FEA) and the bracing piece that will connect to the ball nut.
Re tensioning screws: I don't know the force required. Never seen it done on a DIY machine.
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
Nikolaguca
Another question I have in mind, since I'll weld 15mm thick flat bar to the beams for attaching linear rails (stitch weld, not full length) do I need to normalize steel before machining? I'll just skim the surface so I won't remove a lot of material.
I know your answer will be yes but I'm interested in how much movement there can be after it cools completely?
Is normalizing really nessecary for a router? I haven't found any even approximate numbers regarding movement after welding and cooling.
What are your thoughts on this? Have you normalized your weldments (if that's the right word)?
Sent from my SM-G970F using Tapatalk
Normalizing is different from Stress reliving normalizing uses a different temperature than stress relieving, so you won't want your parts normalized
Yes stress reliving is what is needed for the heat treatment, I do it for every part that needs machining even steel bar will move without any welding if you machine it, when you do your welding stitch weld in the area that the fasteners will be used then you won't get any distortion when mounting Rails Etc, that's a weld every 60mm
If it is sitting on something that is flat in the oven when it is heat treated then it will not move very much
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
awerby
Is the X axis of your milling machine really long enough to mill those 150cm beams in one go? If so, that's great; I don't think you need to "normalize" afterwards. Nobody here does that when building routers, since nobody (hardly anyone) has heat-treatment kilns big enough. But those beams aren't perfectly straight to start with, and all that welding will warp them further, so you do need to flatten whatever you intend to bolt rails to. If your mill's not big enough, what's Plan B?
You get it milled by someone that has a machine big enough for the job stress relieving is needed for a build like this
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
pippin88
Re tensioning screws: I don't know the force required. Never seen it done on a DIY machine.
I do it that way for all ballscrews over 12" / 300mm I install on my builds, ( if you use a rotating nut then you don't need any added tension to the screw ), it works and is worth the extra, very little force is needed from 4lbs to 9lbs ( or around 1Kg to 4Kg ) is the max needed I use a bellevue washer to create the load this way the screw can still expand if it has to
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
mactec54
I do it that way for all ballscrews over 12" / 300mm I install on my builds, ( if you use a rotating nut then you don't need any added tension to the screw ), it works and is worth the extra, very little force is needed from 4lbs to 9lbs ( or around 1Kg to 4Kg ) is the max needed I use a bellevue washer to create the load this way the screw can still expand if it has to
Interesting that so little force is required.
Do you have a diagram / picture of how you do it?
-
1 Attachment(s)
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
pippin88
Interesting that so little force is required.
Do you have a diagram / picture of how you do it?
That is all the manufacturers of machining centers use they mostly don't use the bellevue washers though, they do it just like you are torqueing up a bolt / nut, they use a lock nut or double lock nut to keep the ballscrew under tension
Yes the numbers are not very high to tension something like this, it is also to load what would be the free end support bearing which also has to be contained, that bearing also has to be a thrust load carrying bearing
It is very simple, I normally do a internal thread on the end of the ballscrew a ( fine thread ) with either a stud or use a cap screw and custom washers the first washer pushes against the bearing inner race, the bellevue washer is contained in that washer that is against the bearing the second washer is against the bellevue washer and the nut or capscrew loads the bellevue washer which is loaded against the inner bearing face
If you design it correct when the Bellevue Washer is loaded the ( 2 ) Washer Faces are together and locked the Bellevue Washer does all the preload and the lock nut or capscrew are just locked together ( the ( 2 ) washers faces together ) a simple drawing shows how I do it the Bearing housings also have to be suitable to make this work and the frame that the bearing housings are mounted to also have to be able to handle the extra loading
-
3 Attachment(s)
Re: 150x100cm Steel frame router build
I've got a couple of questions for you guys. Hope everyone is doing good in these tough times.
I finally started the build. I bought 200kg/400lbs of steel today.
My Y axis beams are connected on 2 points to the rest of the frame. I'll bolt them down directly to the main legs of the frame (I'll weld the end caps to the top of the legs for threaded holes). Since I can't machine the whole frame once welded, I'll just machine the Y beams (the red part where the linear rail will be attached) before assembly and then I'll try to make them flat and parallel by having some wiggle room to shim this bolted connection.
This evening I was re-reading the famous MadVac CNC build (Machine frame - MadVac CNC) and in the build he used epoxy as a sandwich between the parts he'll bolt together to make that connection flat and to avoid the bolts pulling the connection out of square. He used epoxy putty but also mentioned that for critical joints you could pour the epoxy into the joint after alignment.
That gave me an idea. For each joint I could add 2 more bolts that could act as a screw jack and use them to level the Y beams, after that close the whole joint with some tape and pour epoxy into the cavity and hand tighten the bolts until epoxy cures. If I weld the frame as precisely as I can I don;t think the cavity between the mating parts will be more than 1mm/3/64" maximum.
EDIT: lol, I had a brain fart. It wasn't my idea. I've read it on that MadVac page I've linked above but somehow I forgot about it and rediscovered it later as "my idea".
"On the more critical alignment joints, some sections were tapped with 10-32 threads and the fine adjustment/alignment between sections took place with 10-32 thumb screws to get the <0.0001" movement on the master precision level or dial indicator. Some sections were aligned, dammed with clear packing tape and the epoxy slurry was injected later into the joints with air escaping through pin holes punched in the corners of the "dam""
Or just forget about epoxy, and use shims instead?
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
Nikolaguca
I've got a couple of questions for you guys. Hope everyone is doing good in these tough times.
I finally started the build. I bought 200kg/400lbs of steel today.
My Y axis beams are connected on 2 points to the rest of the frame. I'll bolt them down directly to the main legs of the frame (I'll weld the end caps to the top of the legs for threaded holes). Since I can't machine the whole frame once welded, I'll just machine the Y beams (the red part where the linear rail will be attached) before assembly and then I'll try to make them flat and parallel by having some wiggle room to shim this bolted connection.
This evening I was re-reading the famous MadVac CNC build (
Machine frame - MadVac CNC) and in the build he used epoxy as a sandwich between the parts he'll bolt together to make that connection flat and to avoid the bolts pulling the connection out of square. He used epoxy putty but also mentioned that for critical joints you could pour the epoxy into the joint after alignment.
That gave me an idea. For each joint I could add 2 more bolts that could act as a screw jack and use them to level the Y beams, after that close the whole joint with some tape and pour epoxy into the cavity and hand tighten the bolts until epoxy cures. If I weld the frame as precisely as I can I don;t think the cavity between the mating parts will be more than 1mm/3/64" maximum.
EDIT: lol, I had a brain fart. It wasn't my idea. I've read it on that MadVac page I've linked above but somehow I forgot about it and rediscovered it later as "my idea".
"On the more critical alignment joints, some sections were tapped with 10-32 threads and the fine adjustment/alignment between sections took place with 10-32 thumb screws to get the <0.0001" movement on the master precision level or dial indicator. Some sections were aligned, dammed with clear packing tape and the epoxy slurry was injected later into the joints with air escaping through pin holes punched in the corners of the "dam""
Or just forget about epoxy, and use shims instead?
After you weld the plates on the top of the tube they are easy to machine then fab the rest they may move after with the rest of the fabrication but at least you have a flat surface to work with, injecting epoxy in a joint like this is good and works well, but only for joints like this, machine for the rail mounting there is no substitute for rail mounting
-
3 Attachment(s)
Re: 150x100cm Steel frame router build
Right after I bough almost all the steel I need I decided to completely change the design of the router.
I found out that the factory near me has a gantry router. I could weld and assemble the whole frame and than have all the mating surfaces machined in one setup. The only problem is that their machine has the 1800x1000mm capacity. The router I was building was 1200x1500mm. I wanted to have at least 1000mm width so I had to change the design.
I've decided to lose the raised Y axis beams and instead make the table flat and bolt the rails on the side. This way I could make the X axis beam as long as I need to use the whole table.
One part of the design I'm not 100% confident about, the risers that the X axis stands on (parts in red). Most of the machines I've seen online use aluminium plate (15-20mm) for this. Is that enough? I can't see that setup as rigid as something I've came up with. But maybe this is overkill.
I expect I've lost a lot of rigidity after I removed those raised Y axis beams so I want to compensate by making these risers rigid and strong.
What do you think?
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
Nikolaguca
Right after I bough almost all the steel I need I decided to completely change the design of the router.
I found out that the factory near me has a gantry router. I could weld and assemble the whole frame and than have all the mating surfaces machined in one setup. The only problem is that their machine has the 1800x1000mm capacity. The router I was building was 1200x1500mm. I wanted to have at least 1000mm width so I had to change the design.
I've decided to lose the raised Y axis beams and instead make the table flat and bolt the rails on the side. This way I could make the X axis beam as long as I need to use the whole table.
One part of the design I'm not 100% confident about, the risers that the X axis stands on (parts in red). Most of the machines I've seen online use aluminium plate (15-20mm) for this. Is that enough? I can't see that setup as rigid as something I've came up with. But maybe this is overkill.
I expect I've lost a lot of rigidity after I removed those raised Y axis beams so I want to compensate by making these risers rigid and strong.
What do you think?
It may fit there machine with 2 setups not ideal but most likely can be done see if it can fit your size machine on there table, put pads on the feet which they would machine first then they have a reference plain to mount on there table
-
Re: 150x100cm Steel frame router build
Oh I'm sorry I wasn't clear. By removing the raised Y beams I've made the machine 1600x1000mm which they can machine and I haven't lost any working area...
I think they can machine the table top and the side where I'll mount the rails in one setup, what do you think?
Sent from my SM-G970F using Tapatalk
-
Re: 150x100cm Steel frame router build
Quote:
Originally Posted by
Nikolaguca
Oh I'm sorry I wasn't clear. By removing the raised Y beams I've made the machine 1600x1000mm which they can machine and I haven't lost any working area...
I think they can machine the table top and the side where I'll mount the rails in one setup, what do you think?
Sent from my SM-G970F using Tapatalk
That is always the best if they can do it in one setup, make sure where they are going to machine it is thick enough
-
Re: 150x100cm Steel frame router build
Your design is much better than flat plates
-
2 Attachment(s)
Re: 150x100cm Steel frame router build
Hi Nico - Thin square tubes are very stiff globally but not locally. You are better off having a very thick angle in this position then a thin tube. This area is shear dominant and you need thickness to stop this area deflecting. If you do want to use those tubes then they need capping so they do not lozenge... but I'd go a thick angle, thickest I could get.
Similiar theme - the pads that the rails are on need to go out to the sides. The little bit in between the radius and the pad deflects quite a bit relatively when you analyse it. Now is your chance to stiffen it.... Cheers Peter
-
4 Attachment(s)
Re: 150x100cm Steel frame router build
I haven't really thought about local vs global rigidity. Thanks for that idea.
I did some quick simulation in F360 to see what will happen if I used angle iron vs tubing. The results confirm what you said. I didn't cap the ends to exaggerate the deformation. I think I'll go with 10mm thick angle iron with capped ends (10mm is enough for tapping holes in the ends for m5-m6 screws so I don't have to weld the caps). And a big bonus is that angle iron is a lot easier to work with in this situation.
For simulation I used 2000N of side force (along the X axis). 2000N in Y direction didn't show any noticeable deflection in the side risers, it was mostly in the gantry tube.
I'm getting a bit worried about the ever increasing weight. As it stands now, the gantry without the Z axis weighs 70kg/150lbs. I thought about using dual Nema34 9Nm motors in a 2:1 belt reduction setup to have at least 10 meters/m and without going over 500-600rpm. But whatever... I'll design for stiffness and worry about the drive part later.
-
1 Attachment(s)
Re: 150x100cm Steel frame router build
Morning Nico - 2000N is 200kgf. This is an unlikely tool load. I use 1000N as a design load a nice round number. To help optimise your design using F360 you need to pick a target machine static stiffness. VMC's run at 150N/um small routers are <10N/um. Some small hobby mills are ~2N/um. If you aim at 25N/um then you will probably in reality achieve >10N/um which is quite stiff and will cut aluminium. So taking your current numbers 2000N in "push" direction deflects 158um. So the static stiffness is 2000/158= 12.7N/um which is quite stiff for a router but your model does not include bolts, bearings etc which are compliant. So you may get 6N/um in reality. So the point is pick a machine stiffness target and this will allow you to decide where you need to gain stiffness.
The next thing to understand is structural loops. The usual loops on this sort of machine are the Z axis, the saddle, the gantry, the base. Each bit or loop will contribute to the global stiffness. The Z axis usually contributes most as its a cantilever. I hold the Z axis bearings and load that to calculate the loop deflection, then hold the saddle bearings (release the z axis restraint) do same, then the gantry bearings do same. In this way you can figure the proportional stiffness of each loop. Then the loop with the most deflection is the area you need to put work into. For instance my current design the Z axis deflects 30% the gantry is 30% and the base is 30%. Initially the Z was 50% of total so had to work on that. Keep making :) Peter
Re: weight - you do have the tools to reduce weight, define what you want to cut then you will be able to define a design machine stiffness then you have a target to chase to reduce weight yet achieve stiffness. You did pick steel which is the "heaviest" material!!
Its been very difficult getting info on modern machine stiffnesses especially routers. The image attached is from a very old research doc and modern machines are much stiffer then his chart suggests. I have sent inquiries to various toolmakers but they don't answer the question. They specify everything else it seems odd that they would not specify machine stiffness when its such an important thing. One large machine maker actually replied that they do not know!!
Looking back at your #1 post you want to cut hardwood and Al. Al is the killer so you will need a machine of at least 5N/um and I would aim at a machine design stiffness of 15N/um maybe 25N/um up front and see how you go. You will not achieve commercial mill stiffness values.
-
Re: 150x100cm Steel frame router build
Hi Niko - 10m/min is easy to achieve even with small motors. Its the acceleration to get there that requires the torque. Can't do the motor sizing until you know the structure masses and what accel you want/need so keep at it :) Peter