[peteeng]

Hi Jon - Is there a stonemason near you? They will the gear to lap and polish your cast blocks. Ah the search for the non workload machine grail continues Peter

[ardenum]

Hi Jon - Is there a stonemason near you? They will the gear to lap and polish your cast blocks. Ah the search for the non workload machine grail continues Peter

ooh that's an interesting idea, machining prices are bonkers but stonemasons, yea that could work, good advice, thank you.

I've seen rails mounted directly on ground cast parts, I'll ask for some quotes, fingers crossed....

[peteeng]

Hi Jon -

https://ericweinhoffer.com/blog/2017...-plates-method

you can always lap your flats. Engineers have been doing that for a very long time.... Peter

[ardenum]

Hi Jon -

https://ericweinhoffer.com/blog/2017...-plates-method

you can always lap your flats. Engineers have been doing that for a very long time.... Peter

you mean lap the steel block inserts myself? I mean it's possible, but the question is how it compares time wise to lapping granite?

[peteeng]

Takes the same time. The time is relative to the starting flatness of the object. You use the same grit steel, iron or granite. Peter

[jackjr-123]

ooh that's an interesting idea, machining prices are bonkers but stonemasons, yea that could work, good advice, thank you.

I've seen rails mounted directly on ground cast parts, I'll ask for some quotes, fingers crossed....

How could that work? Stonemasons don't work with tolerances in the 0.01mm range nor do they have the required precision machinery. Sure they can polish a slab of granite but it won't be flat.

[peteeng]

Hi Jack - My local mason at the cemetery has a large wet cnc with lots of diamond tooling. Plus a large polishing machine. Sure they don't work at engineering tolerances but they can rough the parts then they would need to be manually lapped flat. Peter

[awerby]

Granite is very hard and obdurate stuff. Lapping a large granite surface to be flat within .001" or so by hand would be a major chore. It would also require another large perfectly flat reference surface and a lot of diamond abrasive, or a machine that's set up to produce perfectly flat surfaces in hard and abrasive materials. Granite surface plates are produced on machines like that, but they tend to be relatively small. The difficulty of keeping tolerances on a larger surface increases with size.

[peteeng]

Hi - if you look up some videos of the chinese making granite surface plates its done manually with water grit and laps... Peter

[ardenum]

Granite is very hard and obdurate stuff. Lapping a large granite surface to be flat within .001" or so by hand would be a major chore. It would also require another large perfectly flat reference surface and a lot of diamond abrasive, or a machine that's set up to produce perfectly flat surfaces in hard and abrasive materials. Granite surface plates are produced on machines like that, but they tend to be relatively small. The difficulty of keeping tolerances on a larger surface increases with size.

Good thing the surfaces in question are very small and not as hard(UHPC), mating surfaces are the size of 720x300mm. It's a tiny machine.

How could that work? Stonemasons don't work with tolerances in the 0.01mm range nor do they have the required precision machinery. Sure they can polish a slab of granite but it won't be flat.

Like Peter said, I went through some local ones and they all have big grinding cnc's. If they could grind the surfaces to some tolerance I can then lap them to whatever I need. Also we're talking about UHPC cast parts not granite. Initial grinding at the stonemasons would speed up things considerably.

0.01mm tolerance is orders of magnitude more expensive than what I'm comfortable with finantialy. All machining I price using ISO 2768 – Medium tolerance, check this table :

https://xometry.eu/en/subtractive-ma...ards-overview/

Lapping by hand is my only option to bring the tolerance to manufacturer spec for rails and mating surfaces.

[ardenum]

Hi,

Today we're looking at another granite build, this time an HMC config and the structure is a L shape. 1200x1200x165mm 4um granite plate is cut into 2 and from here we form the L shape. Bridge piece can be made from another granite plate sniped on ebay and lapped on the plate itself or from solid aluminium.

The bellows around the spindle are armored on the front. Spindle travel is X360mm Z460mm(up to 600mm if you keep the whole 1200mm length of the granite plate). Red block is the collision volume with the longest(200mm) and thickest(D.40mm)tool.

Uses a fancy toolchanger arm, although haven't quite figured out the mechanism for that(no cam gears and rollers), maybe a belt driven servo and a pneumatic cylinder or a ballscrew spline from THK with two belt driven servos.

On the Y axis travels 4th axis rotary table. Utilizing an YRT three row roller slewing ring. Powered by a homegrown PMSM torque motor(sneak preview inside).

.step(https://www.mediafire.com/file/6144y...mn2_x.stp/file)

[ardenum]

At this point only one final concept remains to be unveiled, the one that that fits all my requirements and also the one that has most of compromises. QoL is a huge goal with this project. It's starts of as a 3 axis and I can use the machine itself to upgrade it up to 7 axis, of course who needs that many?

To go beyond 3th axis there is a bunch of technologies that I need to develop. One of those, that I hinted already is a PMSM motor. I have the basic geometry/structure figured out. Slots are fully separated(to maximize slot fill) and they are directly glued into a water cooling sleeve with steel filled epoxy. Steel filler to maximize the heat transfer. Plastic parts that separate the coils from laminations are printed on a resin printer. The rotor magnets have a square cross section and are arranged in a hallbach array. In practice it will have the the same effect on the flux as having custom and expensive curved magnets.

Ansys has a wizard that generates various types of motors, PMSM's included, I will use that. I'll also try to verify the generated stator with a bunch of maths I've gathered from research papers. Nothing is optimized yet.

Laminations will be either 2mm or 1mm, laser cut, isolated with a 100um kapton tape film. Unless I find a cheap chinese vendor to make them.

Drives will be running either FOC or DTC, running PMSM's with DTC involves a higher torque ripple and back emf so it's better to use FOC, however with DTC you basically don't need an encoder, although I'm not sure if it's precise enough for this type of use scenario.

What you'll also be able to see in the cad file is a carbon sleeve that goes over the rotor for speeds >5 000RPM. Depending on the sleeve thickness you can go up to 60 000 RPM. Wondering why so fast? Because I'm also designing a motorized spindle. Which will be the topic of a future post...

.step(https://www.mediafire.com/file/k9uig...e_mot.stp/file)

[ardenum]

I like this factory tour...wink wink

[ardenum]

Another tech for today: Vacuum table.

2 piece, bottom piece has a vacuum chamber, top piece has the pattern, in my case this is datrons pattern because I'll be strictly using their vacucard method for sheet metal and in case it won't work I can just use the regular grid+seal method. That's mainly why it's 2 piece. A little insurance so to speak and in case of a crash. The top and bottom plate have 4 locating pins.

The pattern on the granite plate(inserts) is machined by the machine itself, so it will act like a zero point system. The vacuum subplate or any plate that goes on the granite will get bolted in with shoulder bolts, that will automatically align anything within the coordinate system. Just requires initial configuration and then you don't have to use the touch probe for any 'attachment', only for the workpieces.

This is the first in one of the many workholding methods I'll need to develop, next time, a custom adjustable vise.

.step(https://www.mediafire.com/file/n2ny5...akuum.stp/file)

[peteeng]

Hi Ard - so for sheet metal work you are not cutting all the way thru the thickness? otherwise you are cutting into the vac pad? Peter

[ardenum]

Hi Ard - so for sheet metal work you are not cutting all the way thru the thickness? otherwise you are cutting into the vac pad? Peter

Hey Peter, I'm cutting all the way through the material and into the vacucard. It's a small surface but I don't know if I can get enough vacuum from the pump. I'd like to try and use a regular pump+cylinder and a venturi to make vacuum from pressure, datron states in one of their video that's not be powerful enough for the vacucard method and that's why they use a dedicated industrial pump. So the top plate is a bit of a gamble.

[peteeng]

Hi Ard - Use a proper vacuum pump that is used for evacuating air conditioners. Full vac at sea level is 10 tonne per sqm. These are cheap enough these days. So you use a piece of paper card that's good. 200gsm is 0.25mm thick Peter

https://www.vevor.com.au/vacuum-pump...20group%20%231

[Buoyen]

Hi,

Today we're looking at another granite build, this time an HMC config and the structure is a L shape. 1200x1200x165mm 4um granite plate is cut into 2 and from here we form the L shape. Bridge piece can be made from another granite plate sniped on ebay and lapped on the plate itself or from solid aluminium.

The bellows around the spindle are armored on the front. Spindle travel is X360mm Z460mm(up to 600mm if you keep the whole 1200mm length of the granite plate). Red block is the collision volume with the longest(200mm) and thickest(D.40mm)tool.

Uses a fancy toolchanger arm, although haven't quite figured out the mechanism for that(no cam gears and rollers), maybe a belt driven servo and a pneumatic cylinder or a ballscrew spline from THK with two belt driven servos.

On the Y axis travels 4th axis rotary table. Utilizing an YRT three row roller slewing ring. Powered by a homegrown PMSM torque motor(sneak preview inside).

.step(https://www.mediafire.com/file/6144y...mn2_x.stp/file)

For a machine like this. What is your feeling about spindle taper and power.
I asked Olispeed about their SAB150 with HSK63, but at 8500 euros it's a bit expensive.
And HSK63 is expensive tooling.



Some rough exaples (don't really know if they are 100% realistic).
For high speed milling, example in aluminium:
10mm 3 FL carbide end mill.
20mm DOC
2,5mm AOC
20 000rpm (628m/min)
0,1mm/tooth
6000mm/min feed
->
cutting force approx 400N
Spindle power 6,75kW
Torque 2Nm


For high speed milling, example in steel :
10mm 4 FL carbide end mill.
20mm DOC
1mm AOC
6700rpm (210m/min)
0,075mm/tooth
2000mm/min feed
->
cutting force approx 650N
Spindle power 3,64kW
Torque 3,3Nm



It looks like the more popular sizes is ISO30 or HSK40E for high speed spindles with HSK beeing more expensive.
Do you think HSK will have any practical impact vs ISO30?

[ardenum]

For a machine like this. What is your feeling about spindle taper and power.

I wouldn't worry about power in this type of a machine(high speed machining), power becomes needed when you're doing heavy and deep cuts with low <500RPM , then you want power on top of torque. For this design, which is strictly aiming for high speed machining, I won't be using the drive dongs in the BT/SK form, only driving them by the taper and if it were hsk then skipping the drive dongs as well and running HSK E toolholders. The tapers provide enough for the toolholder not to slip when high speed machining. Aiming for about 5000N clamping force in the drawbar.

I asked Olispeed about their SAB150 with HSK63, but at 8500 euros it's a bit expensive.
And HSK63 is expensive tooling.

I agree, I asked them for SAB150 sample price range on July, 2021 and they said:
"Price for SAB150 can between 2900-4000 euro depending accessory and performance."

It looks like the more popular sizes is ISO30 or HSK40E for high speed spindles with HSK being more expensive.
Do you think HSK will have any practical impact vs ISO30?

ISO/SK/BT 30 vs HSK 40 E

with ISO/SK/BT 30 you can get around 25,000 RPM with HSK40E you could theoretically go up to 40000 RPM. As for any practical impact I don't know, I haven't actually machined anything in my life yet but it probably depends on what material you're cutting and how you want to cut it.

For HS machining you have 2 ways, lower RPM (12-15,000) and shallow cuts but engage all flutes, or higher RPM like dental mills where you machine very tiny bits with the tip. Those dental mills run high RPM's over 30,000 so if that's what you want then HSK would be the better choice between the two and probably size 32 or 25 would be better. Most common in those micro centers are actually spindles that don't use toolholders but grab the tool directly with high speed spindle collets like this one:

https://www.acrow-tools.com.tw/EN/pr...20Collets/289/

Datron machines .eg are basically upscaled dental mills.

[Buoyen]

I'm thinking regular HS machining and enough low end torque to do face milling.

Is it worth paying more for HSK40 vs ISO30?