Interpreting the Calibration Report for my Dial Indicator

[s4000853]

hi there, i know that this post is a bit outdated. but i would like to make use of this post to ask some questions.

i received a calibration report for my dial indicator.

MANUFACTURER STARRETT (709A)
RANGE 0 - 0.003 INCH

FEATURE INSPECTED ALLOWABLE ERROR (INCH) MEASURE VALUE (INCH)
REPEATABILITY 0.00005 0.00003
DISCRIMINATION 0.00015 0.00012
LIJMITS OF ERROR FOR
AN INTERVAL OF
ANY ONE DIVISION 0.0001 0.00009
ANY FIVE DIVISION 0.00025 0.00014
ANY HALF REVOLUTION 0.0005 0.00051

in any case of a vernier caliper report, it is quite straight forward to get my resultant error and decide whether tool is good to use compare with my tolerance. but in this case, i am not sure how to read the report and whether it is good to use for my application.

 

[Pezikon]

I see your conundrum. I've never seen that before. Never the less, I will present what it looks like to me.

Assuming; range = 0.03, resolution = 0.0005.
Test indicators (such as your Starrett 709) are usually pre-loaded and zero'd right before use. A user might pre-load the test indicator anywhere throughout the range. A user might pre-load the test indicator half way though the range, leaving only 0.015 of travel remaining. One might say, to give you a test point of "0.010 inch" would be meaningless because it's 0.010 inches relative to some arbitrary zero. Obviously, this is not a problem for a vernier caliper because it's zero is not relative; it is absolute. You cannot reset the zero on a vernier caliper.

So instead of giving you data for 0.010", 0.020", and 0.030", you might get data that tells you how much error is present when the needle moves only one division. How much error is present when the needle moves five divisions. How much error is present when the needle moves a half revolution.

Judging from the excessive precision of the data, I would bet that the standard used to calibrate your test indicator was something like a Mahr 400B-50. The technician zero'd everything and then oh so carefully turned the calibrator up until the the needle on the test indicator read one division. He then looked at the digital display which read 0.00059. That's +0.00009 of error. However, I would say the uncertainty attributed to UUT resolution is at best 0.0001". If the technician had repeated the measurement several times, I'm sure he would have seen numbers like, 0.00056, 0.00061, etc. Or maybe they're good and they reported the average.

I digress. If I were to convert the data into something you are more familiar with, it would look like so:
0.0005 = 0.00059
0.0025 = 0.00264
0.0150 = 0.01551

The last data point is at the edge of its tolerance. If I were to extrapolate the data for 0.02 or 0.03, they would be out of tolerance. How convenient that such data was not given. On the other hand, extrapolated values are merely conjecture.

 

[SATHYABALARAMAN]

hi,

Does your calibration report has a standard reference?

If so, permissible error will be mentioned in the standard, the calibrated readings can be verified against the permissible error & judgement can be made accordingly.

For Dial indicators refer IS standard IS 2092-1983.

Hope this helps..

 

[Hershal]

What is the reported uncertainty by the cal lab?

Given the numbers involved, that may be more important.