[qwertysimo]

Hi,

after finishing my very first router made of MDF I have decided to move to next level - I opted for aluminium extrusions and plates. Although I like projects like these very much, I would like this to be my final build. I want the router to serve me, not vice-versa. I will reuse components from the first build and buy missing stuff. As the price tag is completly different compared to MDF build costs and I want to build high quality machine, I would love to hear any criticism or suggestions you might have before I buy items for final design.

This is what I have already available:

• 4x NEMA23 381 oz/in 3.5A Dual Shaft Stepper Motor (KL23H2100-35-4B)
• 4x Oldham type motor couplers
• 4x Leadshine´s EM705 stepper motor drivers
• 2x 48 VDC 7.3 A power supplies
• Ethernet Smoothstepper with C25 breakout board
• 2.2 kW water cooled Chinese spindle with VFD inverter
• Water cooling set with radiator
• Misc. HW like emergency push buttons, Hall effect limit switches, spidle speed controler etc.

My shopping list (vendors are selected with respect to availability and reviews from other builders):

• Hiwin ball linear guides with blocks
• Hiwin ball screws, nuts, bearing units
• Misumi aluminium extrusions, brackets, joints etc.
• 6061 T6 aluminium plates

I have reviewed many extrusion designs and I like this and this the most. My metal processing options are rather limited so I have to buy e.g. bearing units instead of making them or to use cutting service to precisely cut extrusions to specified lengths. I definitely don´t want to mess this part by being imprecise.

My idea for this build is as follows:

• Footprint about 1200 x 900 mm, work area about 900 x 600. This might slightly change.
• I want the gantry to be the longest axis. I would like to have option to machine larger parts when needed.
• Supported rails, supported work table.
• 4 mm pitch screw is very little. As Gerry wrote, I need higher pitch, 8 - 10 mm. I choose 10 because of availability.
• The machine must be as rigid as possible - I rather overstep some parameters.
• MDF work table surface.
• I plan to weld steel base frame for machine later.

The very first draft:

So far, the upright will not be straight up but will be sloped to move the center of gravity of the completed gantry between blocks on Y rail as much as possible.

Before I move on with design I would like you to advice me on couple of things, one by one. I will be thankful for any feedback!

Thank you.

[qwertysimo]

I was thinking about mounting linear rails to extrusions - whether to mout rail directly to extrusion or use a support plate. I searched forum and found intersting info:

... Still, I'd bolt it to something else before I'd bolt it to any extrusion.
Extrusions aren't as flat as they look. Rails are also a little flexible.
Smaller sized rails...20mm for example....would conform to the irregularities in the extrusion and confound the heck out of someone for days....and days.....when the blocks/trucks bind in places.
John

The problem is accurately attaching a metric rail to an SAE extrusion and having a good base for it to sit upon.

According to Kent at Automation4Less:
"The bottom line is that slotted extrusions (and I'm speaking of any brand here...T-Slots, 80/20, Bosch, etc.) are not "precise." They are extruded, and often that means that there is some irregularity especially along the slot. Many times there is an observable "slant" to the material when you look at it as it goes into the open slot."

"For this reason and because the rail will only be partially supported on its edges (because the center of the rail would be over the slot), it is my opinion that you cannot get the best, solid connection by placing the rail over the slot and then dropping fasteners through the rail mounting holes into t-nuts below."

"You may get along just fine with your rail sitting over the open slot and by using cap screws and t-nuts. But in my opinion, having the rail sitting on top of open space and resting only on its edges isn't the best way to go."

Kent suggested mounting the rail on a piece of 1/8 aluminum or steel and then mounting the 1/8" to the extrusion.

I modified that idea slightly. I sandwiched a piece of 1/8 aluminum between the rail and the extrusion. I drilled over-sized holes (1/4") in the 1/8" aluminum and attached the rail to the extrusion using the m4 socket head capped screws (shcs) with the 1/8" aluminum between.

In the machines I built, I used CRS flat bar to mount onto the extrusions first and then the rails mounted to this bar. It gave more rigidity and mass and also provided a nice place to add shims should they be needed. ... This extra height also provided extra clearance for the ball screws. I did mill down the first machines extrusions to be flat, but this was really not needed.

On the other hand, when you look at many extrusion builds, mounting rails directly to extrusion is quite common...

This is detail of the draft above - 4 mm plate between rail and extrusion:


Does it really make that difference in respect to long term rigidity?

[qwertysimo]

Hi,

please, can you look at this picture and advice, which design should I continue with?



On #1, 2, 3, there is 80160 used - for profile length 1200 mm and 20 kg centered load the vertical deflection should be less than 0,01 mm.
On #1, there is additional 40x120 for attaching the rail.
On #4, there are two 80x80x40 facing each other.

I do not like #1 because it is very wide.
I do not like #2 and #3 because of the nut housing placement. Center of gravity is lower than on #1.
I like #4 the most - compact, nut housing in between rail blocks. But, 80x80x40 has 1/10th of 80x160´s cross sectional moment of inertia in vertical direction. Reinforce it with backplate bonding both extrusions?

Any thoughts welcome. Thank you.

[RicknBeachcrest]

Hi,

please, can you look at this picture and advice, which design should I continue with?



On #1, 2, 3, there is 80160 used - for profile length 1200 mm and 20 kg centered load the vertical deflection should be less than 0,01 mm.
On #1, there is additional 40x120 for attaching the rail.
On #4, there are two 80x80x40 facing each other.

I do not like #1 because it is very wide.
I do not like #2 and #3 because of the nut housing placement. Center of gravity is lower than on #1.
I like #4 the most - compact, nut housing in between rail blocks. But, 80x80x40 has 1/10th of 80x160´s cross sectional moment of inertia in vertical direction. Reinforce it with backplate bonding both extrusions?

Any thoughts welcome. Thank you.

Linear rails like to be parallel to each other. Like a couple of thousands. I got mine to within .001" Which option will allow you do fine tune the placement of the rails?

[RicknBeachcrest]

[QUOTE=qwertysimo;1780878]I was thinking about mounting linear rails to extrusions - whether to mout rail directly to extrusion or use a support plate. I searched forum and found intersting info:



It depends on the size of your rails. I used 15mm so there wasn't a lot of seat on each side of the slot. Also keep in mind that extrusions are canted inward on each side of the slot. You do not have a flat surface to mount the rail.

[RicknBeachcrest]

Hi,

please, can you look at this picture and advice, which design should I continue with?



On #1, 2, 3, there is 80160 used - for profile length 1200 mm and 20 kg centered load the vertical deflection should be less than 0,01 mm.
On #1, there is additional 40x120 for attaching the rail.
On #4, there are two 80x80x40 facing each other.

I do not like #1 because it is very wide.
I do not like #2 and #3 because of the nut housing placement. Center of gravity is lower than on #1.
I like #4 the most - compact, nut housing in between rail blocks. But, 80x80x40 has 1/10th of 80x160´s cross sectional moment of inertia in vertical direction. Reinforce it with backplate bonding both extrusions?

Any thoughts welcome. Thank you.

Well since you asked.

I would modify option 1. Take two 80X80's and attach a piece of cast aluminum plate to the backs of the 80X80's. Your nut will be in between the two extrusions, your rails will be on the same plane.

[qwertysimo]

Is this what you mean?



I am starting to concern about gantry weight with respect to it´s movement (Y-axis). Gantry #5 above has about 50 kg (110 lb). Without Z-axis motor, motor attachment, X-axis ball screw or gantry uprights...
Y-axis will be dual motor configuration with 2x 381 oz.in steppers coupled to 2010 ball screws. Will this work?

Woud it be reasonable to use thiner alu plates (green arrows; e.g. 10 mm instead of 15 mm) and standard 80x80 extrusions instead of "heavy" version (red arrows). But not at the expense of ridigity.

[qwertysimo]

It depends on the size of your rails. I used 15mm so there wasn't a lot of seat on each side of the slot. Also keep in mind that extrusions are canted inward on each side of the slot. You do not have a flat surface to mount the rail.

20 mm for X and Y. 15 mm for Z. I will use supporting plates - slots in extrusions are half the rail width. I will sandwich them like you did in your build.

[RicknBeachcrest]

Is this what you mean?



I am starting to concern about gantry weight with respect to it´s movement (Y-axis). Gantry #5 above has about 50 kg (110 lb). Without Z-axis motor, motor attachment, X-axis ball screw or gantry uprights...
Y-axis will be dual motor configuration with 2x 381 oz.in steppers coupled to 2010 ball screws. Will this work?

Woud it be reasonable to use thiner alu plates (green arrows; e.g. 10 mm instead of 15 mm) and standard 80x80 extrusions instead of "heavy" version (red arrows). But not at the expense of ridigity.

Yes, that is what I mean.

I think 10 mm for the back plate would work just fine. I sure would not skimp on the bolts that attach it to the extrusion though. The other green arrows; I think I would use the 15mm, since it will be sandwiched in between other components and you may need to counter bore some holes.


As far as standard vs heavy versions of extrusions. I guess that would depend on what the difference were on the Moment of Inertia. As a comparison: This is the thickness of some 75mmX75mm extrusions Moment of Inertia . . . Ix=3.379 in 4 (140.645 cm 4 ) Weight. . . . . . . . . 4.027 lbs/ft (5.993 kg/m) And I'm really happy with the thickness of this extrusion.


A certain amount of gantry weight is good. There always are trade-offs in the cnc world. Weight helps to dampen the vibrations, but then you have to deal with the momentum with too much.

[qwertysimo]

Thanks for your comments. Standard vs. heavy 80x80 mm Misumi extrusion - FYI, the heavier is about 60 % more expensive, 55 % heavier (7.13 kg/m) and provides 40 % higher moment of inertia (183 cm⁴). 80x80 seems to be reasonable sized for my gantry so I probably stick to the heavy version for it´s increased stiffnes.

I will keep the vertical gap between these extrusions exactly 80 mm. If needed, extra 40x80 can be inserted and bolted to plate:


I searched forum and found several posts on motor torque and gantry weight. It should not be any problem here. Acceleration might be lower.

[LeeWay]

The gantry on my home made router weighed about 150 pounds. I drove it with a single centered 5/8" rolled ballscrew and it could rapid 300 IPM. About the same size motor. Just slower acceleration as you mentioned..
You should have zero trouble with two motors.
I would make the gantry as stiff and rigid as possible. I would use 40 by 80 and a thicker plate behind it. I would actually use steel flat bar rather than aluminum. Use the extrusion simply for ease of mounting. This is assuming you can find a nice straight piece of flat bar. You may need to back it up with a peice of angle or channel, but you need someplace for your Echain to ride anyway.

[qwertysimo]

Thanks Lee, I am calmed down concerning the gantry movement.

Is this what you propose for gantry?

Steel is out of question, I have no idea where to source it localy and processing and handling it might be a problem for me.

Actualy, what is the purpose of backplate? Primary to reduce deflection in Z-axis because of longer gantry (X-axis), right? Z-deflection will be reduced by bounding both extrusions so they deflect less.
While rectangular flat bar is great for this purpose, it´s moment of inertia in Y-axis is poor. For example, calculator shows moments of innertia for 240x10x1200 mm solid bar as 1152 and 2 cm⁴.
Doubled thickness of 20 mm gives 2304 and 16 cm⁴. This makes me worry about deflection in Y-axis.

You may need to back it up with a peice of angle or channel

That angle or channel added to backplate, it is mainly to reinforce the Y-axis, right? Like this?

Isn´t this similiar concept to my gantry design #1 above? The bottom extrusion was Y-axis reinforcement...

[LeeWay]

Right about the extra piece being to stiffen up the plate. It would have added use to carry the wiring for you if you turned it 90 degrees left. Could be angle or channel. In your first concept, the two extrusions are acting independently and do not help each other out. When you ties two together using a third member, it becomes much stronger and more rigid as an assembly.

[qwertysimo]

Understood.

OK, I have to include e-chain support to my design, completly forgot about it.

[qwertysimo]

Somehow I am still not quite happy with the gantry design. Adding backplate here, adding C-channel there, attaching extrusion here... It looks already like a patch work at the beginning of project

The best would be to have a beam made of a single piece. If I exclude steel tubes, C-, U- or L-shaped beams, I need massive aluminium extrusion. I did some calculations. Considering static load for gantry beam, cutting force, acceleration force and targeted deflection max 0.01 mm, I got to Misumi 90x180 extrusion, but I like 100x200 even more - it provides spacing 150 mm for X-axis rails. It´s weight is roughly the same compared to any of designs listed above (several extrusions, backplates, channels). I will not need any other enforcening.

What worries me now are the gantry uprights. I planned to use 20 mm aluminium plates without any other enforcening but I feel they can be flimsy compared to gantry beam. I cannot lower the gantry. I consider to replace plain table with U-shaped one and reorient Y-axis rails so I can make uprights shorter. I will post some pics for review later.

[qwertysimo]

I made two versions.



I like the upright on right image - basically the gantry beam 100x200 lies on Y-rail blocks. This setup provides maximum rigidity.

What I do not like are vertical 80x160 extrusions on work table sides. This changes flat work table to U-shaped table, preventing placing larger objects on it (and does not match my taste at all). This setup also decreases X-axis travel.

For these reasons I prefer setup shown on the left image. Gantry uprights are as high as neccessary to provide 130 mm (5") clearance under gantry.

Unsupported part of plate, between the rail block and gantry beam, is about 130 mm. Is this the weakest point?
If I use 20 mm thick aluminium plate, will this be sturdy enough?

[LeeWay]

Just a note or two. You won't be able to load larger parts with either machine. Depending on the end design, you may be able to load longer parts with either.
Also don't forget you will need a table top and a spoil board most likely. They take up space below the gantry beam. The lower gantry beam is more desirable from a rigidity stand point though. Less tendency for the Z axis to act as a lever.
The single extrusion is probably a better choice, but they can deflect some too. It is strengthened some by your rails. The beauty of that design is you can always add a structural member somewhere later on if needed.
My personal choice would be the one on the right.

[qwertysimo]

You are right about large parts - this is design limitation of each finished machine.

Z-axis clearence of 130 mm is from bottom extrusions to gantry extrusion. I will use 18 mm MDF as spoil table top for a start so actualy I am getting 110 mm clearance (sheet stock processing oriented design). I might change this later to T-slot table...
If I will ever need to increase working height, I just replace two small upright plates with higher ones (cut by this machine of course ).

Hm, I like "design on right" more and more...

[qwertysimo]

Hi. Please, do you see any design errors or possible improvements? It looks like a pretty standard design to me... Any comments appreciated. Thank you.

[LeeWay]

That is looking pretty good. The only suggestion I have would be to add another support for the table at your home position under your spindle. Parallel to the gantry.
That will lessen that span and make a more solid support, because it is smaller and under the actual cutting area.

I added a couple braces to the router I just built, but wish I would have added one along the gantry at home position too.