Results 1 to 16 of 16
05-17-2010, 12:46 PM #1
Precision level required accuracy
How would a 0.05mm/metre level accuracy go for use as a tool to aid in the refurbishment of machine tools...
The usual recommended level, the Starrett #199 is 0.0005"/ft/graduation which is 8.6 seconds of arc..
While the 0.05mm/metre is 10.3 seconds of arc..
We often have the discussion of levels but as most of the members here are from the US the Starrett #199 is often the only level discussed but they are very rare items here on the second hand market and the suitability of levels of other accuracy is never discussed...
Or indeed what accuracy levels are actually needed to level machine tools and to use for refurbishing machine tools..
05-17-2010, 02:53 PM #2
The 0.05mm/meter level will be just fine for the application. One thing to consider is the surface of the level.
The Starrett level has a flat bottom surface, as does my Scherr-Tumice, no grooves. Most of the imported levels I've seen, have grooves, which would cause me to exercise more caution in their use.
05-17-2010, 02:54 PM #3
The exact calibration of the level doesn't matter.
The sensitivity is given in deflection per division, but there's no standard for the distance between the divisions.
Any instrument accurate to about 10 arc seconds would be fine.
I've seen some that are even more sensitive, around .0002"/ft.
05-17-2010, 03:18 PM #4
Thanks for the advice...
I take it the 0.02mm/metre levels are designed for lab work in a temperature controlled room?
05-17-2010, 03:43 PM #5
You ask a question that I can't answer.
What we in the US call a Master Precision Level (Starrett's Model 199Z is, by far, the most common and most widely known member of that genus, but other makers have included Berger Instrument, Lufkin, South Bend Lathe, Scherr-Tumico, Pratt & Whitney, Queen, Schulte, Taft-Peirce, Universal Boring Mill, and probably a couple dozen more) typically contains a vial in which the bubble will move 1/10 inch -- the distance between the vial graduations -- when the vial is tilted somewhere between 8 and 10 arcseconds. This requires that the inside of the vial be ground and polished to a radius somewhere between 215 feet and 172 feet.
As you might expect, British level vials were traditionally graduated in 1/10 inch increments also.
In contrast, modern European and Asian level vials are graduated in 2 millimeter increments. Having the bubble move 2 millimeters when the vial is tilted 0.05 millimeter per meter requires the vial curvature radius be 40 meters, roughly 131 feet.
In my experience, the makers of precision level vials generally consider a vial to be usable if the measured sensitivity is within 10 percent of the nominal sensitivity. If European and Asian level instrument makers go along with the vial-makers +/- 10 percent tolerancing, a level instrument claiming 0.05 millimeter / meter sensitivity should actually range between 0.045 millimeter / meter and 0.055 millimeter / meter. Stated another way, the vial's radius of curvature should range between 36 meters (roughly 118 feet) and 44 meters (roughly 144 feet).
So . . . the bubble in a standard US Master Precision Level will nominally move 1.6 times as far as the bubble in a 0.05 millimeter / meter level in response to identical tilts. At the extremes of vial sensitivities, the bubble in the standard US Master Precision Level could move as much as 1.8 times or as little as 1.2 times the distance the bubble in the 0.05 millimeter / meter level moves in response to identical tilts.
Yes, the standard US Master Precision Level is a fair amount more sensitive than a level with a vial graduated at 0.05 millimeter / meter of tilt.
Do you need the extra sensitivity when setting up or rebuilding machine tools? My best answer sounds awfully trite: You do if you do, and you don't if you don't.
I'll make a suggestion, though. Go ahead and get started with a 0.05 millimeter / meter level, and when you get to the point that you know you need a better level, you can ask yourself if you can make do with a standard US Master Precision Level or if you need a 0.02 millimeter / meter level.
05-17-2010, 03:47 PM #6
05-17-2010, 06:22 PM #7
But you'll seldom have the opportunity to set up a lathe in a temperature-controlled shop.
Levels are fundamentally comparators, and as such they're self-calibrating. You calibrate the zero indication, then you adjust the machine to achieve zero.
The calibration factor is only an indication of how close you need to be to that zero point.
05-17-2010, 06:36 PM #8
05-17-2010, 09:51 PM #9
05-17-2010, 10:07 PM #10
If I use a 0,05 level on a lathe with a 20cm centerheight checking for twist
1 deviation of that level represents a diameterdifference of 0.02mm
So you can figure out of your own if it is accurate enough for you
I liked to start with a 0,05 one and then move on to a 0.02mm/mtr
Peter from Holland
05-17-2010, 11:12 PM #11
Funnily enough, I've just moved all my machines to a new shed and had to re-level them. I used a stabila builders level to 'rough them out' when I put my Starret 199 on to finish them They where all spot on. Probably just dumb luck I suppose
05-17-2010, 11:54 PM #12
How flat is your level?
One other thing to check is how flat is your level, when I blued my Starret and spotted it, the contact points were at either end. Based on this I would recommend spotting and scraping if you want it to have 'truth' in your measurement
05-18-2010, 04:14 AM #13
The "flat" bottom is in fact scraped "hollow", or concave along the length. The exact amount isn't given, but it may amount to a few tenths.
Many other levels, like Lufkin, etc, were made with two pads at the ends and a distinct recessed and unfinished non-contact area in the middle. Not so Starrett.
If you assume the level has a flat bottom, you may find that you get slightly varying results if using it on any surface significantly smaller than the length of the level base.
05-18-2010, 01:33 PM #14
Level too sensitive
a 0.02mm/m level is so sensitive that most of the time it will have bubble off scale or too far off center.
.... most items sag if supported on ends. Even a 0.05mm/m level can pick up an object sagging in the center if supported on ends. put level off center on object and compare with level in center.
..... it is thought that modern level vials are internally ground with diamond plated wheels on cnc lathes. it probably is faster to grind the more sensitive level vial especially if someone is willing to pay more for it.
...... in China it is hard to find a precision level less sensitive than 0.02mm/m. When I asked for a 0.1mm/m or a 0.4mm/m level it was always answered with it can be ordered but is not stocked in the store. (A Starrett 98 level is close to 0.4mm/m)
....... the reason 99% of field machinist work is done with a 0.005" per ft(0.4mm/m) is it is easier to measure how much something is off or needs adjustment. you cannot easily measure if level is off scale. A Starrett 98 level can measure 0.001"/ft if careful and can easily measure 0.002"/ft. If you are setting or adjusting something to less than 0.001"/ft that would be around 4 full lines off on a 0.02mm/m level. (thats about the max that many levels can measure unless it is a big level).
....... most modern precision levels have a cross level vial. the level vial needs to be parallel in elevation (up and down tilt) AND azimuth (side to side adjustment). if main level vial is off sideways then when cross level vial is off a little, main vial can change 1 to 10 level line graduations. not many know that main level calibration requires not only setting so that it repeats when reversed but also that when level base vee is on a cylinder and cross level vial is deliberately put off center to watch if main level vial changes much.
........refer to Wyler web site
BROCHURES + FLYERS / WYLER AG
.......Wyler in my opinion is the leader in level technology. Most companies wish they had even 1/10 of what's in the Wyler catalog
05-18-2010, 02:40 PM #15
My favourite tool for setting-up is the war surplus clinometer shown in the pictures. Maybe only a 30 seconds graduated vial but better than 5 seconds is easy to achieve. A clino is far handier to use than a master precision level, for starters you can keep the bubble in sight and there are a few handy dandy tricks as well. I figured that if one was good enough to help me get optical demonstrations aligned to silly levels of precision it was well up to the job of setting a machine tool.
05-29-2010, 01:57 PM #16