MU (Measurement Uncertainty) for an Optical Microscope

J

jimmy0ad

#1
Hello Guy,

I am looking for different way of expressing the uncertainty of a measuring instrument, an optical microscope used to measure length, I have come across multiple ways but I am not quiet sure which one is right and which one is wrong.

My overall purpose is to develop a class table which we could refer in our laboratary and give a range to our customer saying that this measurement is X micrometer with a Y+- range.

I would really appreciate if someone could also post some reference with their suggestion for me to read some more regarding this topic and to use it as a reference.

Thanks in advance!!
 
Elsmar Forum Sponsor
U

Umang Vidyarthi

#2
Re: MU for an optical Microscope

Hello Guy,

I am looking for different way of expressing the uncertainty of a measuring instrument, an optical microscope used to measure length, I have come across multiple ways but I am not quiet sure which one is right and which one is wrong.
My overall purpose is to develop a class table which we could refer in our laboratary and give a range to our customer saying that this measurement is X micrometer with a Y+- range.

I would really appreciate if someone could also post some reference with their suggestion for me to read some more regarding this topic and to use it as a reference.

Thanks in advance!!
Hello jimmy0ad, :bigwave:Welcome to the cove:bigwave:

I am not an expert on the subject, so wait for others to chime in. In the meantime could you tell us about the 'multiple ways you have come accross', to help you better.

Umang :D
 
J

jimmy0ad

#3
Thanks for the welcome and reply Umang. Here' some more infro no what I am trying to achive and the ways of calculating uncertainty

1. Student T Test Confidence Interval (Sorry I los the article which referred me)

2. spiff. rit. edu / classes / phys273 / uncert / uncert.html

3. GUM (which I found quiet difficult)

More Info on my project:

1. I work in a lab environment where the optical microscope is used for inspection (not for length measurement but general quality issue like cracks and dings and color and stuff)

2. My project is Measurement System Analysis and to help my supervisor better understand my the topic. During my presentation, he raised the issue of uncertainty/ accuracy

3. From this he asked me to divert my attention towards creating a class table which would list the magnification, max and min length that can be measured, and uncertainty associated with each.

4. From what I understood about MSA is that Measurement System is comprised of Part, Operator, Environment, Tolerance, Instrument, and Method. Anytime one of these criteria changes the whole system changes.
This part confuses me because if this is true then you can never find the uncertainty of just the microscope itself (regardless of all other criteria) and even if I do find a methodology the uncertainty will only be associated with that particular measurement system and it cannot be exptended to others.

I feel like I am running in a circle and thats why I woudl like to get an opinion of somebody. Like I said any help will be deeply appreciated and I would really really like a referenc material as I could back my experiment/ methodolgy during my presentation.
 

BradM

Staff member
Admin
#4
Hello, Jimmy0ad!:bigwave:

Welcome to the Cove! We are sure glad you dropped by here. :yes:

Pardon me for starting at the beginning here. So apologies if anything here seems overly basic.

Uncertainties for a CMM system:
http://elsmar.com/Forums/showthread.php?t=1046

This post has a link to a really good introduction to uncertainty:
http://elsmar.com/Forums/showpost.php?p=381531&postcount=4

Also, you can go up to the Green post attachment lists at the very top of the webpage. Click on it, and do a search under uncertainty. There are several attachments in there that should help you out.

So, let's assume after all your searching and reading that you have gained an understanding of what uncertainty is, and how to tackle estimating it. Do keep in mind, uncertainty budgets can range from very simple to quite complex, depending on the process, need, etc.

So, go ahead and start putting some variable together for your process. I would estimate any environmental conditions, vibration, error among operators, equipment error, etc. Put them all together in a spreadsheet or a budget calculator. Here is one for a balance I put together. I'm pretty sure I got the spreadsheet from here at the Cove.

NOTE:If you are unsure what uncertainties to put next to it, just start with identifying all the variables that may affect that equipment. Then you can research getting estimations for the uncertainties.

Not sure if any of that helps. I'm not an expert with uncertainties, but I'll sure offer what help I can.:)
 
J

jimmy0ad

#5
Hey Brad,

Thanks for your reply. I have few questions regarding GUM and the approach you described.

1. So basically my goal should be to take into account all significant contributor into my uncertainty budget (and then add them to find Expanded Uncertainty). If I want to be contain 68% of the data I take standard deviation. If I want to contain 95% of data. I multiply standard deviation by 2 and so on.

2. I could only think of three variations that contribute significantly to my verall uncertainty budget. They are Equipment Error (Repeatabilty), Operator (Reproducibility), and Calibration. We developed Fish Bone Diagram to find the causes that would lead us to inaccurate measurements and all the other causes (30 or so) seems to contribute not as much as these three. Do we have to include them into Uncertainty budget.

3. If we do include into the uncertainty budget. What are the different ways to calculate the uncertaity due to those specific causes that are hard to define and apply any statistical method to.
 

BradM

Staff member
Admin
#6
1. So basically my goal should be to take into account all significant contributor into my uncertainty budget (and then add them to find Expanded Uncertainty). If I want to be contain 68% of the data I take standard deviation. If I want to contain 95% of data. I multiply standard deviation by 2 and so on.
Yes. 95% seems to be the most common expression. However, it is up to you, and what level of confidence you wish to have.

2. I could only think of three variations that contribute significantly to my verall uncertainty budget. They are Equipment Error (Repeatabilty), Operator (Reproducibility), and Calibration. We developed Fish Bone Diagram to find the causes that would lead us to inaccurate measurements and all the other causes (30 or so) seems to contribute not as much as these three. Do we have to include them into Uncertainty budget?
Good question. I would say yes, but there are no values assigned. You would document that you considered a particular variable, but given XXX reason or whatever, it was considered insignificant. That way, if someone ever examined your budget, all the considerations have been accounted for. Too, having that listed may help contribute to a better budget. Someone may look over it and have some better information to contribute. All the information there together helps that process. :)

3. If we do include into the uncertainty budget. What are the different ways to calculate the uncertaity due to those specific causes that are hard to define and apply any statistical method to.
Of the 3 you have listed, they should be pretty easily estimated. However, you have asked the million dollar question.:agree1: One way that will work is to determine the variable and estimated based on the parameters. So say you are talking about vibration. You have measured the vibration, so you contact the mfg. and ask about how much vibration can effect the reading. They may can give you a value within a range or something that it would contribute.

The best way? Control or eliminate those variables where they are not of concern. :D Like temperature and humidity (for some equipment). Find out from the mfg. what range the T/RH can be within without affecting the equipment's performance, and control it where it is not a factor.

As you are building your budget, put everything in the same designation. So have everything expressed in microns, diopters, degrees C, etc. You don't want to mix things in your budget; so convert them as best you can.
*******

There are many here who are way better versed in uncertainty budget than I. What I have found though, is this is a much better way to approach measurement systems. The budget gives you a start, and you begin with your "best assessment" of the measurement system you have. And... it will change. The values will change, and possibly other considerations might need to be considered. Also, you can lower your uncertainty (if needed; desired) by gaining better control over some of the contributors in the budget.:agree1:
 
T

tomvehoski

#7
OK, it has been about a dozen years since I dealt with using microscopes to perform measurements, but here is what I remember.

The application was measuring coating thickness of Titanium Nitride, Titanium Carbide, and similar thin film PVD/CVD coatings. The thickness was in the 1 to 10 micron range. We had two ways to look at the coating. First was cut and polish, where we would mount a part, cut it in half, and then polish the face so we could directly see the coating. Accurate, but very time consuming. Second was a ball cratering method where we rolled a carbide ball bearing with a diamond slurry against the part which created a crater in the coating. When looked at from above, you would see a bowl shape in the coating. By measuring the outer diameter (coating surface), inner diameter (substrate surface) and the diameter of the ball bearing, you can calculate the thickness with trigonometry. This method was faster, but less accurate.

As for the measurement, we had a camera and computer connected to the microscope. We had a certified slide that had a scale on it. You would snap a picture of the slide, and then calibrate the software by teaching it how many pixels on screen equaled the length of the scale. Once calibrated, you could draw circles on a ball crater image and the computer would do the math. Cut and polish you could just do a direct measurement.

We did do a gage R&R - probably a 3 operator, 3 trial, 10 samples for both ball crater and cut/polish. Vibration was definitely an issue for cut/polish because we had to use a much higher magnification than with the ball crater.

I never did a complete uncertainty budget, but some other sources would be:

- accuracy of the calibration standard (slide scale)
- accuracy of the micrometer we used to measure the ball bearing
- probably some type of rounding error on screen pixels vs. actual measurement.
 
Thread starter Similar threads Forum Replies Date
M Measurement Uncertainty in Optical Microscopy Measurement Uncertainty (MU) 1
B IEC 60601-2-10 - Accuracy of Pulse Parameters - Required Measurement Uncertainty IEC 60601 - Medical Electrical Equipment Safety Standards Series 3
G Does pitch/increment/resolution of a ruled scale apply to measurement uncertainty as line item? Measurement Uncertainty (MU) 10
R Uncertainty in measurement larger than tolerance Measurement Uncertainty (MU) 2
K ISO 17025:2017 clause 7.6.2 - Performing calibration of its own equipment shall evaluate the measurement uncertainty ISO 17025 related Discussions 6
G Reporting measurement uncertainty for custom items Measurement Uncertainty (MU) 2
K Measurement Uncertainty Budget (MU) of for Ultrasonic Thickness gauge Measurement Uncertainty (MU) 1
G Can Measurement Machine Bias be part of Uncertainty? General Measurement Device and Calibration Topics 1
D Difference between uncertainty and expanded uncertainty of measurement General Measurement Device and Calibration Topics 1
J Internal Laboratory Scope and Measurement Uncertainty (MU) IATF 16949 - Automotive Quality Systems Standard 1
C MU (Measurement Uncertainty) in an Internal Calibration Laboratory Measurement Uncertainty (MU) 1
W Dimensional Measurement of 10% accuracy and 25% uncertainty standard General Measurement Device and Calibration Topics 6
B Measurement Uncertainty using Minitab Measurement Uncertainty (MU) 2
A Uncertainty Measurement - What this accuracy specification "1% + 10^5": means? General Measurement Device and Calibration Topics 2
Marc Measurement Uncertainty Discussion Sub-Forum General Measurement Device and Calibration Topics 0
C Measurement Uncertainty fluctuates half a tick mark (20 millionths of an inch) Measurement Uncertainty (MU) 1
R MU (Measurement Uncertainty) of a Tensile Tester (e.g. Zwick Roell) Measurement Uncertainty (MU) 4
H Uncertainty Measurement Calculation in Chemical Testing Measurement Uncertainty (MU) 1
S MU (Measurement Uncertainty) for Oscilloscope Calibration Measurement Uncertainty (MU) 2
T Method Repeatability for MU (Measurement Uncertainty) Measurement Uncertainty (MU) 1
S Combined Uncertainty of an Indirect Measurement Measurement Uncertainty (MU) 8
P Measurement Uncertainty and Equipment Accuracy Measurement Uncertainty (MU) 2
N Uncertainty calculation for a voltage measurement Measurement Uncertainty (MU) 3
T Measurement of uncertainty for standard mechanical properties General Measurement Device and Calibration Topics 1
K New Gauges calibration certificate shows deviations from the measurement uncertainty IATF 16949 - Automotive Quality Systems Standard 5
A Calculating Combined Measurement Uncertainty - VDA 5 Measurement Uncertainty (MU) 1
T Uncertainty of Measurement vs. Tolerance Measurement Uncertainty (MU) 5
A Calibration Method and Measurement Uncertainty General Measurement Device and Calibration Topics 5
Geoff Cotton Measurement Uncertainty (MU) of HR-150A Rockwell Hardness Tester Measurement Uncertainty (MU) 1
Jerry Eldred When Does Resolution of Measurement Standard Apply in Uncertainty Calculations General Measurement Device and Calibration Topics 14
E Uncertainty Measurement for Water COD, TDS, TSS & pH Measurement Uncertainty (MU) 1
D General Procedure for Uncertainty Measurement Measurement Uncertainty (MU) 2
S Measurement Uncertainty for Neutral Salt Spray test to BS EN ISO 9227:2006 General Measurement Device and Calibration Topics 4
S Organizing Uncertainty Budgets for different Measurement Equipment Measurement Uncertainty (MU) 4
B Length Measure - Interpreting Calibration Results and Measurement Uncertainty (MU) Measurement Uncertainty (MU) 3
R Procedure for Measurement Uncertainty Estimation for Electrical Measurements Measurement Uncertainty (MU) 2
R Signal Conditioning Card - Calibration Measurement Uncertainty Measurement Uncertainty (MU) 5
K AATCC 81 pH test - How to calculate MU (Measurement Uncertainty) Measurement Uncertainty (MU) 2
optomist1 Definition Measurement Uncertainty - Standard definitions for Definitions, Acronyms, Abbreviations and Interpretations Listed Alphabetically 8
S Uncertainty Estimation in Temperature Measurement by K type Thermocouples Measurement Uncertainty (MU) 7
DietCokeofEvil ILAC-g8:03/2009 - Reporting Results with Measurement Uncertainty General Measurement Device and Calibration Topics 7
N Rounding the Numerical Value of the Measurement Uncertainty Measurement Uncertainty (MU) 2
A Uncertainty of Measurement for Calibration Laboratory? Measurement Uncertainty (MU) 8
C How to interpret Measurement Uncertainty (MU) Measurement Uncertainty (MU) 5
S ISO 19036 - Measurement Uncertainty for Microbiology Measurement Uncertainty (MU) 7
P Measurement Uncertainty Calculations Spreadsheet help General Measurement Device and Calibration Topics 4
L Measurement Uncertainty using Control Chart Measurement Uncertainty (MU) 2
P Uncertainty Measurement for Colour, Odour, Taste, Turbidity, Total Suspended Solids Calibration Frequency (Interval) 2
T Uncertainty Measurement on Kinematic Viscosity of Fuel Oil General Measurement Device and Calibration Topics 1
N Measurement Uncertainty - When is Normal and when is Triangular Distribution used? Measurement Uncertainty (MU) 3

Similar threads

Top Bottom