[Vilyoo]

Im designing my router for cutting steel. this is going to be my 4th machine, before that i made 2x routers 2500x1500 working area and 2 small machines. The idea of this machine is to fit a normal wood cnc router inside and be able to flatten it. Working area around 1700x3600x500mm. 4000mm length is the maximum i can machine flat in my area, so the longest steel tube is 4m. The machine must be capable of cutting with facemill diameter 80-100mm. Speed is not important, depth of cut maximum 1mm. Overall accuracy i think 0,1mm per meter is okay.

1. All steel tubes are 300x200x10mm. The bottom 5 tubes are connected with M20 bolts to the long. Everything will be filled with sand.
2. Ballscrew for X 2200mm and Z 600mm 2505.
3. Helical rack 4000mm for Y 25x25mm Module 2.5 ......already bought
4. all motors nema 34 12nm closed loop ..... Have some in stock, servos are too expensive for me but if they are must i will consider it
5. all cars 35mm 3 on each side of Y.
6. all steel plates 15-20mm
7. Weight around 2500kg without sand, with sand 4400kg
8. Spindle.. i have 90mm bt30.. thinking for the motor 3.8kw servo or low price asynchronous/DC motor..?? since it will work only with facemill cutters 50-100mm fixed RPM will work too.
9. Controller DDCSv3.1 + Fusion 360

(total tubes length is around 23 metters. Tube inside is 0.28m x 0.18m this makes total volume 1.15 cubic metters, sand is 1.65tones per cubic meter, or total sand is around 1.9 tones)

all steel tubes will be flat machined after making the welding and holes. Im very impatient and started the first part.. the plate for XZ carriages. Other parts are not bought yet.

My question for now is do i need gear reduction for the rack, for example planetary gearbox 3:1?
What do you think about this design, and what can be upgraded without increasing price too much?

Still working on the CAD..

[peteeng]

Hi Vily - No need to fill with sand. Ballscrew for X at 5mm pitch is too small. Using stepper say 800rpm max gives you 4m/min rapid. Not fast enough, use 10mm or more. 3 cars on column not needed. Peter

what pinion size do you intend to use? I do think you will need gearbox for rack. Once pinion size is known can calculate forces. Take 12Nm motor on X. I usually work at 500rpm max cutting (for steppers) so it will have around 6Nm at half speed. 6Nm with a 10mm screw can push 365kgf (which is huge) so the rack system should be designed to match... at 500rpm this is 5m/min still a bit slow for some tools. What's the planned max spindle speed? Peter

[Vilyoo]

thanks for joining Peter. I saw Haas routers have 4 cars on each side, the machine will cut only steel at 100-200mm/min, also gantry+Z axis weights around 400kg

planned spindle speed max is around 1000-1500, depends on the motor i find.

pineon size pitch diamter 58,53, overall diameter 63.5mm (22 teeth)

[peteeng]

Hi Viny - What is the weight of your gantry? I see 400kg The Co for a 35mm car is around 70kN ie 7000kg so 3 cars can support 21 tonne? They must use 4 for some other reason... They may use 4 for their design life. So the load is 1/4 the static load so the life is 4x Life depends on distance travelled and load. An extra car is cheap so up to you... Peter

so a 58.5mm pitch pinion at 6Nm torque is T=Fxr so F=T/r 6Nm/0.0585 = 102N or 10kgf so at 3:1 you have 30kgf which is very small. At 500rpm the travel speed is 1 perimeter = PIx58.5=3.141x58.5=183.8mm so at 500rpm its moving at 500x183=91.5m/min !! so with 3:1 its 30m/min. Go with a bigger gearbox... 10:1 gives you 9m/min at 500rpm and gives you force 100kgf... plus you have one on each side. 200kgf Check my numbers did this a bit quick.. Peter

[Vilyoo]

calculations are right. so 10:1 is what i need. How about the holding torque for X when we increase pitch size does it become half?

It looks that I dont have enough torque to move the gantry?

[peteeng]

Hi Vily - No the holding torque for the motor stays the same so at zero speed you have the full 12Nm. So double the forces just calculated. So at 12Nm (stall or very slow) and 10mm pitch you have F=Tx2pi x n/(9.81x pitch) so 12x2xPIx0.9/(9.81x0.01) =691kgf at stall... For the X dirn (on gantry) Peter

So the force available is 691kgf at zero and zero kgf at ~1000rpm... With a servo its flat from go to WOW 3000rpm or more.

The torque curve for a stepper motor is about linear and drooping. Max torque at zero speed to zero torque around 1000rpm or if you have the published torque curve then you can use that for this sort of calc. N34 can be slow due to their inertia (bigger dia) and their inductance. Use a high voltage system to overcome these issues.

Consider small servos. Since you have to gear them high for the R&P a small servo will give you large forces, a better speed range, the ability to be overdriven so you can take better advantage of its torque without fear of missing steps and won't be much of a cost up, but a clear, big performance up. On big jobs time saved is worthwhile... Peter

[Vilyoo]

are 750w fine for the job? 2.4nm at 3k rpm or 24nm at 300 rpm with 10:1 gear box. How about X. Can i connect the servo direct to the screw without gearbox.

[peteeng]

Hi Vily - The gantry+Z axis is on rails so your only moving the inertia and the friction. Since the friction coeff of the rails is very low say 0.004 you only need 0.004x400=1.6kgf to move it slowly. This is like you being able to push your 1000kg car in a car park... Lifting the Z axis is a different matter as you have to lift the weight. But a 5mm screw can lift a lot! Once you get your feed speeds and static forces sorted (balanced) you can move onto the dynamic issues eg how much accel do I need or have. In commercial machines minimising air time is important so high accels are needed. Once you get your head around the statics come back to the dynamics. Servos are well worth it here, at least consider them. They will be smaller and lighter, there are gearboxes available etc etc.. Peter

[peteeng]

I would say 750W is overkill with the big gearbox on it for Y's Yes you can attach the servo direct to the X ballscrew same as the stepper Oh use a coupling Once you figure your needed feeds and speeds then you can balance out the components. Peter

[Vilyoo]

im worried about Z with servo because it weights around 150kg. How about 2x servos on Y and steppers on Z and X, or it is not good idea.. I already have 10x nema 34s

[peteeng]

Hi Vily - up to you. Since its a paper exercise at the moment develop both ideas until you are clear for the direction. Mixing servos and steppers is OK. A 5mm pitch screw only needs 150x9.81x0.005/2/pi/0.9= 1.3Nm torque so x2 for service factor 2.6Nm is heaps so your N34 are overkill but you have them in hand.... many decisions to make... Peter For the Z do you have a braked motor? or a double shaft so you can use a brake? or are you going to counterweight it? The 5mm and 10mm pitch screws will allow the Z axis to fall under its own weight if power is off... Peter

seems you have a purpose for the machine which is good. Talk to the tool supplier about your most used tool(s) and get the power and max feeds required so you make sure you optimise around those tools.

Consider dynomotion controllers they have advanced motion control, very good support and good price. Plus more axes then the DDCS.

Dynomotion Motion Control Boards for CNC Manufacturing and Robotics Applications

[Vilyoo]

i dont have motor with brake nor with double shaft. How about a relay that disconnects the stepper motor from the driver when power off and short circuit the motor. Maybe will add counterweight if needed

i like that ddcs is very compact and doesnt require computer

[Vilyoo]

this is for 63mm facemill cutter with 4 cutting edges.. recommended chip load for the cutter is 0,15mm per tooth. depth of cut is 1mm.. with 100mm facemill will be 2 to 3 times faster since it will do only 1 pass but it requres double torque

[joeavaerage]

Hi,
to be honest the recommended chip load and so on for the face mill is just not useful at the moment.

Your machine is going to be rigidity limited, and the cutting forces and thereby the cutting parameters will be limited to those that do not excite undue flexure in your machine.
The gantry particularly is going to move around heaps. Concern yourself with making the gantry as rigid as possible and that will determine cutting parameters not the other way around.

Craig

[peteeng]

Hi Vily - so on the point of rigidity add an extra section to the back of the gantry. Most of your loads will be in the XY plane so this needs to be stiff. It takes up no extra space in the envelope yet will make your machine much much stiffer. Every part must be as big as possible... You have described this as a router but its a MIll - needs to be Uber rigid... Peter

[peteeng]

Hi Vily - The HAAS GR-712 gantry machine with X=2.1m and Y=3.6m (similiar to yours) weighs 10 tonne for a reason.

GR-712 | 40-Taper Mill | Large-Platform CNC Routers |Vertical Mills – Haas CNC Machines Peter the gr-712 has 3 sections stacked up. They have gone for height to lessen the load on the gantry bearings..

[Vilyoo]

added more steel to the gantry. Internal bracing will be added. Do i need to upgrade the gantry tube, it's now 300x200x10 for example to 2x 150x200x10 welded together

[peteeng]

Hi Vily - So its 300 high and 200wide? Internal bracing not needed. External bracing has to be on edges not in the middle of a face. It can deflect / vibrate if loaded in the middle of a face. The webs you have added at the ends are not needed. The middle of the gantry is the bit that deflects not the ends. Peter or can you get 150x200x12?

does your CAD have FE? or simulation? a machine especially a Mill cannot be too stiff.

[Vilyoo]

yes 300 high and 200 wide, i have simulation, those are the steel available at my local store

tell me where to add load how many kN and i will simulate the deflection.

I have 2x 150x250x12.5mm at 2200mm long each. This is 160mm shorter than cad tube, maybe i can try to fit it, if it's needed.. can also extend it by 160mm because on the photo are leftovers 600 and 900mm and can cut from them



like this on the last photo, should i be worried about the weight it will reach 700kg with spindle and motor

[peteeng]

Hi Vily - That small free corner will vibrate so cap it. OK FE is a big subject. I apply 1000N at the spindle nose. Use bonded connections everywhere to begin with. Place the gantry at max depth and in the middle of the table. Make a dummy vice and place under the nose of the spindle. Then restrain the vice and place the load on the spindle nose in the X Y and Z direction. 3 loadcases. What you are looking for is the nose deflection in microns. Then normalise the deflection to N/um. A modern VMC is around 150N/um stiffness. See how you go. You are trying to mimic ASME B5 attached. Peter