Test Indicators - resolution vs accuracy

[dbb-the-bruce]

Yes, thanks. This does look like a good place to buy indicators.

Mahr MarTest Dial Test Indicator Model 801SGE MAHR 4308985 Top Mount grads 0.00005" Dial Dia. 1.5"

This is very much in the realm of what I'm looking for. However still having trouble with they way the throw around Accuracy, Resolution and mix other stuff in.

The Mahr site has technical specs for each model.

The catalog helps quite a bit - starting on page 102. On page 110, in the table I can see that the 801 SGE is probably the best choice.

And less than what I was looking at at MSC.
Again thank you.

 

[Jim F]

Are you making parts for NASA ?

 

[mksj]

If you are measuring things like run out, then a 0.0001" indicator is much more informative then say trying to read between the line with a 0.0005" indicator. It has nothing to do with NASA or being OCD, and everything to do with the specifications you are trying to measure. The measurement accuracy of these types of indicates is usually +/- 1 indicator line when using it for measurement purposes, but if you are looking at runout and the pointer stays between two lines that is informative that the runout does not exceed the division between the lines but you cannot infer an amount. As I mentioned, a multi-turn test indicator is more useful in my opinion if you are starting with a wider deviation and dialing it down, a single turn indicator is fine for starting out with very small deviations. The Compac 215GA is one revolution = 0.00004" with a 6 revolution range, so very easy to read at or between the lines as well as a wider range. See discussion below. If you do any work with electronic or mechanical dial edge finders, a 0.0005" test indicator is pretty useless for alignment purposes. YMMV.

Mitutoyo-513-443-10A are also excellent test indicators, but a lot of fakes out there.

Mitutoyo 513-443-10A Dial Test Indicator Mid Set Tilted .016 in

www.msi-viking.com

Compac Test Indicator inch reading

Compac P-Line horizontal and vertical model test indicators with metric reading dials made in Switzerland by TESA are once again availalble.

www.longislandindicator.com

 

[dbb-the-bruce]

Are you making parts for NASA ?

Surprised this type of comment didn't show up earlier in the thread - I hope that you mean it mostly as a good natured ribbing.

When I make stuff, I work to the tolerances required for the part. I make mostly small stuff with features as small as .005

My CNC mill has a spindle that tops out at 24000 RPM and I use end mills as small as .010. I've been breaking end mills more often lately. Spindle runout == death for small end mills at high speeds. A run out of .001 is 10% of the diameter of a .010 end mill.

Turns out that, yes in fact the lower bearing on my spindle is damaged. I have a new spindle on the way. With my new .00005 resolution indicator I now have the ability to "see" small differences when I need to.

I've been fascinated with precision since I was a kid - as a teenager I got into astronomy, ended up more interested in making telescope optics. Making a precise optical surface by hand requires being able to measure it. Making a set of optical flats by hand to 1/4 wave length was a great experience.

 

[Ulma Doctor]

If you are not in a temperature constant environment, not a single measurement finer than a thousandth means anything anyway.
I think people often work outside of reality in their minds, and love to tell fishing tales having no basis in reality

 

[RJSakowski]

If you are not in a temperature constant environment, not a single measurement finer than a thousandth means anything anyway.
I think people often work outside of reality in their minds, and love to tell fishing tales having no basis in reality

Steel has a coefficient of linear thermal expansion of 1.2E-6 per ºC or .66E-6 per ºF. My shop temperature variation is less than 10ºF at the extremes and less than 5ºF for something like 95% of the time. A 10ºF change would result in just over a half a thou in an inch change in length.

Add to that, for steel at least, the dimensional change in a steel work piece will be matched by the change in the measuring instrument.

Finally, a dial test indicator isn't used to make absolute measurements but rather differential measurements as in measuring taper in a two collar test or runout of a cutting tool ore workpiece in a spindle. Each of these reasons can justify the use of a higher resolution measuring instrument.

I am a firm believer at being able to measure to a higher resolution.. If the measuring instrument can only measure to a "just get by" resolution and you are making a string of measurements, measurement errors can stack up to ultimately take you out of tolerance. On the other hand, if you are using a high resolution instrument, you have the option of disregarding or rounding that last digit.

 

[pontiac428]

Those of you with true (pun intended) interest in measurement have lots of reading to do.