hello all !
Is there a standardized measurement how to measure machine rigidity ? I mean whole setup: frame, spindle, tool holder, workpiece holder.
I was thinking generally about putting a solid bar in tool holder and "pushing" a load cell fixed in workpiece holder and measuring deflections.
Please share your experience !
Not experience but more of a brain fart really to address dynamic vs. static loading.
I like your load cell idea but how about you combine a ball bar tester with a the load cell idea i.e. a ball bar tester you can seriously load and measure stress strain. [I think the ball bar tester may have a micro form of strain gauge in it.]
So lets say your 100 mm ball bar test makes a circle with roundness of 5 micron like +3.5 - 2.0 (kind of thing) and then you keep cycling but you command the machine to make progressively smaller circles stepping in 2 micron at a time. So you would end up with a spiral plot or set of circles (polar plot) that show under what forces the machine maintains the commanded diameter it's making and then when that starts to drift as increasing loads are applied.
OR you could have a meaty ball bar tester when using pizzio (sp)-electric micro actuators to progressively enlarge the circle the machine is making/ trying to maintain and measure forces and deflections accordingly (could be kinda fun/ dangerous). [Pizioelectric crystals + actuators can exert huge forces over a small distances using minimal current].
After a while you start to get a pre-load and lateral force (center seeking) towards the ball/hinge point/ joint of the strain guage + ball bar tester. So when you apply 50 , 100, 200, 400, 800, 1600, 3200 etc Newtons of thrust you can simultaneously measure the degree to which the table VS the spindle moves away / deflects / distorts.
HERE's THE RUB:
A super rigid machine will destroy it's spindle bearings if you keep going 3200, 6400, 12800 Newtons...
Whereas a more "Supple" machine will distort and deflect and in some cases return back to original position unloaded.
Which kinda lends credence to more "Gangly machines" and linkages like a DMG-Mori DMU 65 mono block. That machine makes GOOD parts but it's linkages are very ponderous and half of the machine relies on the mass of floating or dynamic castings. In other words if you give it a good smak around the head it will "give" and rubber bands/ industrial belts to drive the ball screws are used so at least also there is an "elastic" give in the system that protects the other drive systems and yet does not compromise too much in response/accuracy.
Conversely : Go look at the recent seminar/ Webinar on "Tool Bending moment" from Makino.
Machine Tool Webinars and Events | Makino
In that kind of scenario they give you basic "Math" to calculate tool bending moment and present really interesting research about dynamic cutting conditions under pretty serious loads.
They have a special device for that (need to re-watch it).
Basically it's kind of a load cell that you can physically cut !
(far out right ??? !)
Kistler? Kind of a vice that has strain gauges built into it.
Their premise is that a small fraction of their user base is not calculating tool "Bending moment" correctly and hence are TRASHING spindles (like 4th axis horizontals) cutting particular Titanium alloys. So they have some clients they never have to replace a spindle for and some other clients that have four spindles replaced over a relatively short period of time. Citing underestimation of tool bending moment and not being a function of "Speeds and feeds" tool engagement and DOC.
The key point being is that the Machine is EXTREMELY rigid and the weak point is the spindle bearings / wear if you exert excessive lateral forces i.e. too longer tool.
They also present some really almost counter intuitive data about surface
speeds for Titanium alloys [like Ti 6Al-4v ). (worth watching to end).
__________________________________
++ Vibrations during cutting are also monitored during cutting and spindle loads. Peak force spiking not smoothed out data (on X, Y and Z axes separately).
*****---> For some reason can't link to Kistler dynamometers and the like ? It's a "Russian" inter-web thing ? maybe tomorrow.