Gallery, Projects and General > The Design Shop
Portable milling/grinding machine for machine way reconditioning
PekkaNF:
Hannu? Jorge? :lol:
It's not even nearly DIY...but when I'll take it appart I'll might change my mind. Specially if it turns out that I have whole lot less than 12 surfaces to plane.
Today I checked 20*20 mm and 22*14 mm keystock. Unbelievable straight. All of them had a bow, max. 0.05 mm / m, about the same than linear bearings. That sort of bow straighten easy with finger pressure. Unbelivable luck. Would need superstructure and where it would be easy to bolt down.
Young's modulus seems to be max. 4GPa, or 4 KN/mmē for MDF. P6 MDF (floor grade) 20-25 mm thick is 2,5 KN/mmē
Nylon 2-4
Al 70
Steel 210
Concrete 10-40
Therefore MDF would need to be pretty deep structure to resist deflection.
Epoxy concrete Young Modulus is pretty good in comparison 60 to 80 GPa (KN7/mm2)
Did I got the numbers on ballpark? Looks about there...
Pekka
hanermo:
The most rigid, specialist epoxies with titanium or stainless steel paste in them are something like Hysol (locktite) 3478 et al.
About 1100 kg/sq cm crush resistance.
But strength =/= load =/= crush resistance.
What happens is that the load is spread over a largish area (technically) and the point load or kgf push is low .. approx 3-5 kg, imo.
= 500 N or less.
So, even wood carriages work well (on rail grinder), because the linear guides do geometric averaging on them.
For the same reason - geometric averaging..
I have overconstrained rails on the front, impact side, of the new VMC.
Ie 4 rails, of Hiwin 35 mm profile, 1000 mm tall, on front. 2 at back. 2 cars each rail, 12 total.
A single car has 4000 kgf load capacity, +/-, for 50.000 kg or 500.000 N rated capacity.
I cannot, easily enough, make the assy to much better than 0.01 mm (this level is easy. Assemble so cars donīt bind, by hand. Shim as needed)).
The geometric averaging makes the actual accuracy be about 2-4x better.
I donīt have 50 metric tons push load on my VMC ( I wish).
I hope to do 3 kW spindle power, at most, => 300 kgf push, or 3000 N.
2 kW is fine.
Using 32/4 mm screws, 1200 (1600 mm max) for now.
32 screws are max 1500 kgf max, +/-.
Similar to modern medium VMCs with 1-1.5 m table travels (they do 25 kW).
X axis is the long one, 2200 mm long end-end with 1600 free travel.
Beams are 2200x20x200 mm, with 35 mm rail on top, for 235 mm total height.
Over 1600 mm free span, a 235 mm flat steel beam, 2 of, can probably carry way more than 50 metric tons without failure, in theory.
(Its actually not on top of the beam, but..)
Hannu.
PekkaNF:
Huh! That is a big plan. Very nice to read on how differently people approach different problems. Hope to hear more from that project progress.
I have been doodeling on paper some spindle designs. Goal is to make a simple spindle that will allow a little of axial and radial grinding.
Basically I need a fixed bearing near grinding stone and free bearing near pulley at the other end of the shaft.
I considered first classical pair of angular contact bearings, but very few are offered with any dust shields. I'm setling on dual angular contact bearing with 2RS shields, this I might just get away with the loading and RPM I need here.
I draw two slightly different basic designs, which one is prefered and why?
Is there better, easy to make design?
* Addition: The square on very left is a polyvee belt pulley - free bearing end - ball bearing. Grinding stone will be on right hand side (not drawn) on locating bearing side.....just to clarify my doodle.
Thank you,
Pekka
PekkaNF:
I'm thinking on using fat shaft design....toss of a coin. Grinding stone on right next to double angular contact bearing and poly-vee belt on left to load the free tail bearing.
Pekka
RotarySMP:
Does a surface grinder really need the thrust control of paired AC bearings? My Clarkson T&C grinder only has deep groove ball bearings. Since you are grinding a surface, with no precise shoulders, I'd think you only need radial stiffness, and could make the spindle more simple.
Mark
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version