Steel CNC Router Design

Hello to all the cnczone people,

I am keen to make a cnc router, partially to make things and partially to make a router! One of my requirements is to be able to work with Aluminium so I figure I am going to need a strong/stiff machine. As a result I am leaning towards making a welded steel frame and welded steel gantry to get the stiffness. Having never made a cnc router before I am not sure it will be stiff enough for Aluminium work and also if it will be possible to get the gantry straight and square enough after the welding?

Here are some screen dumps of the design....well there will be once I figure out how to do that.....seems images need to be hosted elsewhere...

Any comments and thoughts welcome.

Cheers
Brett

PS I hope this is the right spot for this type of post.

figured out how to upload pics....read the FAQ.

Attachment 171999

Attachment 172000

Attachment 172001

Brett

Great design Brett, the gantry especially should be very rigid while being low weight!

One thing that comes to my mind is perhaps turn the steel channel under the spoilboard, and also the two lengths supporting them, 90 degrees. The reason is that doing this will greatly increase the rigidity when machining. With the existing design you'd find that even if welded, the two steel channel supports would flex under machining load, albeit small, as the two supports have less vertical support than horizontally. Turning them side on vastly increases the vertical support, which means the steel channel under the spoilboard don't need to be vertical anymore.

For practicality it might also be handy to design end boards in, maybe 75 x 25mm steel channel which would cover the tube ends, and make fitting the drivescrew and stepper easier?

cheers,
Ian

Looks good so far... As Ian mentioned, I agree that adding endplates would make for a cleaner appearance as well as add some strength. As to the gantry I'm not sure how well twisting forces are addressed, and the problem may be magnified when cutting at the extents if the table. I'd be inclined to maybe even use profile rail and block on the long axis so the gantry can withstand higher moments when machining at the extremes. Speaking of which, I like to get as much a work envelop as possible within the footprint of my machine, and it looks you can definitely get more travel in the gantry by using longer linear rails; this would put the center of your spindle closer to the edge of the table.

Looks Good,

Make sure you suss out your motor mounting options early on.

If you can move your x-axis ballnut bracket to the rear of the bridge you may be able to hide the x axis motor further under the table out of harms way. (could possibly mount ballscrew bearing (use FK instead of BK)to a plate on one side of the RHS, and the motor to a plate on the other side of the RHS. note that accessing the coupling would be difficult after the spoil board is fitted).

On the same note, if you extend the y axis rails right to the edge of gantry and place the ballnut bracket towards the non-drive end of the plate, then you should be able to y axis travel the full width, and get the whole table as suggested by louieatienza. The plate should be able to travel over the top of the ballscrew bearings without fouling, and then there should be enough room inside your gantry vertical on the drive end to fit the coupling (assuming the motor is mounted to the gantry vertical

Cheers
Paul

Hi Brett;
anotehr way to stiffen the bed is to keep the stringer orientations as they are, but add another cross-ways support in the middle unless it interferes with the drive.

You will probably find it easier to use more steel tube to reinforce the gantry back instead of building up with plate pieces. Less welding involved, and I think it will be stiffer.

What are the dimensions?

If you really want to cut aluminum, going with a fixed bridge/moving table design would be better. That way you can make the bridge as heavy as you'd like, and brace it much better than if it had to move. Moving the table instead of the gantry allows you to spread the sliding load farther, so the whole thing's more rigid. Machines like this take up a larger footprint, but that's the only real drawback.

www.computersculpture.com

Nice modeling! If you're not tight for space, as awerby said it's easier to get high stiffness using a fixed gantry. Some good examples:
Super Duty Series CNC Machining Centers by CR Onsrud, Wood CNC Mill, Metal CNC Machine, Metal CNC Cutting, CNC Machine Vibration Isolation
First 3: Official Home Page of the Roman Black

If you need a moving gantry, here are some well-designed commercial versions:
Multicam: MultiCam - Products - CNC Router - 7000 Series
Vision: http://www.visionengravers.com/image...-machine-1.gif
TechnoCNC: http://cfnewsads.thomasnet.com/image...605/605670.jpg

Some common themes for either approach:
1. Large gantry tubes. For X (long axis) forces on the cutter, the gantry cross-member is put into torsion. Liek Paul said, tube sections (square or rectangular) are much stiffer to resist torsion. This link has a great xls:
cnc machine stiffness calculator - MYCNCUK
In the xls, go to the "Gantry Top Twist" tab and be amazed! :)
2. Lots of separation between the Y-axis rails, and they're mounted to the top and bottom edges of the gantry tube. This minimizes the rail forces and applies those forces into the stiff edges of the tube (not mid-panel).
3. Large tubular cross-section for the gantry uprights/legs. This stiffens the parallelogram mode for forces and acceleration in the Y direction.
4. The gantry tube is as close to the table as possible. Stiffness at the tool is proportional to the inverse square of the z clearance. The clearance affects both torque and deflection, so less clearance helps twice!
5. For moving gantry: long “feet” for the main rail bearings – this minimizes bearing forces and tipping for acceleration in the X direction.

As Louie mentioned, consider using profile rails. SBR is 3x more flexible than profile in compression, and much less stiff at resisting "pull-off" (tension) forces as the bearing "opens up". Asian blocks only have 1 ball track to resist tension. http://i01.i.aliimg.com/img/pb/439/4...493439_025.jpg
I think most Thompsons use 2 ball tracks near the opening, but the design just doesn't allow a very strategic orientation of the tracks, so it's inevitably weak in tension / pull. Probably the best way to orient SBR on the gantry is to point them up and down (away from each other); then they don't see tension, and the span between them is maximized, which minimizes forces. Still, this puts the trapezoid base in bending -- supported round is really only good in compression.

IF you can afford better rails Hiwin is a good alternative. SBR is ok but this was the weakest part of my build. I will be upgrading to HIwin.

Make the gantry stiffer. Use channel or RHS. Bolt and epoxy as much as you can to eliminate welding stresses.

Thanks for the feedback guys, some real good points there I hadn't considered.

I am short on space so was hoping to get away with a moving gantry design rather than a moving table. But as I said I want to do aluminium work so I will reconsider that decision... A moving table requires longer rails for the same work area but I could get better quality shorter rails for a moving gantry....I need to do more research here on how to spend my $ and space :-)

I can see now that I need to make the gantry stronger in twisting, the bed stronger horizontally and look at the linear bearing mounting in more detail. Before I address those issues I better resolve the moving gantry vs table question.

The overall size is 1100mm x 800mm and I should get a work area of 700mm x 560mm with the current design - although that will change. I have modelled it in Solidworks and am happy to send out the files once it has matured a little more if anyone is interested.

Cheers
Brett

A moving table can still fit a small area, Just make it wide to get the most of the area. DaTRON make gantry machines for milling aluminium so it is possible.

Making it a moving table vs moving gantry is all about balancing budget, materials, and expectations. I'm building a moving gantry for mostly aluminium work, as the working and machine size is important to me. Especially as I'll be plumbing coolant and I'd like build a water/sound proof cover around it. So I'm putting heaps of mass into the base, and making the moving parts as low weight as possible for the mass and rigidity required to machine aluminium, without heaps of reasonance noise, but still get highly repeatable accuracy. Given in either model the X axis is sliding perpendicular to the Y axis, and introduces it's own accuracy issues, I don't beleive for the home hobbiest that the gantry vs table argument is black and white at all. It all comes down to what works for you, and personal requirements which drive choice of design and most importantly, materials used.

For my machines, I've been more than happy with moving gantries for machining aluminium, so for me that's what works. Everybodies experience and desires will be different, but don't let it bog you down as a showstopper.

cheers,
Ian

Yes, it's certainly possible to have a stiff moving gantry -- mostly more engineering to make it both stiff and reasonably light.

If SBR is used anywhere to save $, I think the best place is the X rails: the cutting force is in between them and so it is shared by both rails, and with long enough feet the bearings would not see tension. The cutting force is cantilevered off the Y and Z rails, so their lower bearing loads are ~2x the cutting force, which means they flex more and let the cutter deflect more. So profile rail is most important for those.

Welding is easy and strong, but it does distort both right away and over decades (stress relief). The commercial machines put the whole frame in an oven to stress relieve, then they machine the rail surfaces flat. The immediate distortion is mostly a function of good design (e.g., if a tube needs to be straight, only weld it near its ends) and the welder's skill (good fit-ups, a zillion small tacks, minimal heat, TIG if possible, avoiding inside corner welds, checking distortion along the way and figuring out a weld order to ~cancel it). Rosette/plug welds are best for distortion, as they tend to self-equalize, but they're more work. Since CNC machines are governed by stiffness (not strength), there's little reason for full-length welds. Many tacks, rosettes, and some stitches will distort a lot less than blasting away. All that said, epoxy will be much straighter.

Quote:

---

Originally Posted by bmanners

Hello to all the cnczone people,

---

HI

Quote:

---

I am keen to make a cnc router, partially to make things and partially to make a router! One of my requirements is to be able to work with Aluminium so I figure I am going to need a strong/stiff machine.

---

You will probably be better off with a fixed gantry, moving table system

Quote:

---

As a result I am leaning towards making a welded steel frame and welded steel gantry to get the stiffness. Having never made a cnc router before I am not sure it will be stiff enough for Aluminium work and also if it will be possible to get the gantry straight and square enough after the welding?

---

Stiffness is an engineering problem.

Also welded steel frames in the automation industry, such frames are machined after welding and preferably after stress relief/ normalization. Usually the frames will incorporate pads or flanges for assembly that are machined square or parallel to bearing surfaces.

Depending upon what you are trying to achieve using welded steel framing can be difficult if you don't have access to a machine shop to square up and machine the bearing surfaces. Structural steel isn't all that straight to begin with especially if you are use to extruded aluminum sections.

Quote:

---

Here are some screen dumps of the design....well there will be once I figure out how to do that.....seems images need to be hosted elsewhere...

Any comments and thoughts welcome.

Cheers
Brett

PS I hope this is the right spot for this type of post.

---

Probably not a bad start but since this is aluminum related you may want to try one of the metal working threads for home built machines. Aluminum requires a bit stiffer machine than wood but not excessively so.

Finally it is mistake to believe that just using steel makes for a stiffer machine. If you want a stiff machine made out of steel it needs to be designed that way. Further expect that machine to be pretty heavy compared to many wood working routers.

I've done some more reading and there are several examples of moving gantry machines working with Aluminium so it is possible but as pointed out will require some attention to detail to make sure the design is good. For me space is a premium so I am going to stick with the moving gantry design.

I have access to a milling machine...and a machinist :-) so I will be able to machine the gantry mounting points after welding.

The HIWIN linear rails look the business and automation4less have some old models going for about the same $ as the SRB rails you can get on ebay.

Cheers
Brett

Out of curiosity Brett, what sort of prices inc shipping are you getting from automation4less? And for what size rails? Getting the good ones always seems prohibitively expensive in OZ, as you no doubt are very well aware mate! ;) We certainly don't fall in the category of Americans with their multitude of suppliers at economical prices!

So long as you make the gantry rigid enough, and put a decent amount of mass into the table and frame, I don't see you'll have any problems at all doing aluminium with a moving gantry. What are you looking at for the table material itself? I will say with doing aluminium this very important, if you go for a thinner top, with supports spaced more than a few inches apart, you'll get tremendous reasonance noise. For the big one I'm building, i'm going with a 20mm thick aluminium plate, as the T-Slot on the 6090 I bought creates a huge amount of noise because it lacks mass, and is composed of multiple pieces that have minute air gaps that cause sections to vibrate against each other.

Even if you are going with a thick plate, a couple more supports will go a long way to cutting noise and vibration down.

Edit: Actually if you are in Vic, Robot Trading, (and other stores) have 50mm x 50M rolls of double sided sisalation tape for around $25 last time I got some. I'm planning on using this between the contact points on the underside of my table, as it compresses down to nothing, but helps amazingly well in dampening vibration and noise. I'd highly recommend it if your affixing the table to the frame.

cheers,
Ian

Ops, I seemed to miss that Automation4less sell the rails as single items and the SBR guides on ebay are sold as pairs. So the HIWIN are about twice the cost. Not sure about shipping costs as I was just browsing the online store.

Brett

I am looking at 20mm rails/guides and 1000mm long.

Sent from my Galaxy Nexus using Tapatalk 2

FWIW, your design bears significant similarity to one I built about a year ago of wood, and am still using. I used bearing guides and rails very similar to the ones in your drawings as well. Early on, you mentioned twisting forces on the gantry. After experiencing some racking issues, the first mod I made to my machine after I got it running, was to add a second stepper/leadscrew to the Y axis, moving the existing center mounted one to one side, and balancing it with the new one on the other. Racking issues went away completely, and with Mach3, the extra motor was a breeze to setup and I've had no syncing issues. Although I work primarily with wood, I've also used it to machine 1/8" alu, worked great, haven't had a need or request to go thicker, yet. My next upgrade will be to change the y axis over to rack and pinion, as screw whip is now a limiting factor, my y axis uses a 48 inch (1219 mm) screw.

Note, as I've noticed that terminology differs, to me, "Y axis" is the one that moves the gantry assembly fore and aft as you stand facing the gantry.

HTH

Quote:

---

Originally Posted by BuzzCut

FWIW, your design bears significant similarity to one I built about a year ago of wood, and am still using. I used bearing guides and rails very similar to the ones in your drawings as well. Early on, you mentioned twisting forces on the gantry. After experiencing some racking issues, the first mod I made to my machine after I got it running, was to add a second stepper/leadscrew to the Y axis, moving the existing center mounted one to one side, and balancing it with the new one on the other. Racking issues went away completely, and with Mach3, the extra motor was a breeze to setup and I've had no syncing issues. Although I work primarily with wood, I've also used it to machine 1/8" alu, worked great, haven't had a need or request to go thicker, yet. My next upgrade will be to change the y axis over to rack and pinion, as screw whip is now a limiting factor, my y axis uses a 48 inch (1219 mm) screw.

Note, as I've noticed that terminology differs, to me, "Y axis" is the one that moves the gantry assembly fore and aft as you stand facing the gantry.

HTH

---

For me X and Y are relative to how I approach and load the machine. My router was as you said, but I turned it sideways to fit a table/cabinet, and I kept getting confused so I swapped X & Y. No more confusion.