Looking for Practical Advice in Managing Measurement Uncertainties

[Rmack]

Hello,

I'm trying to manage the implementation of 17025 in a calibration lab that is part of a testing lab. I'm hoping to find someone up here who has been in the same boat and wouldn't mind offering a little practical advice.

A bit about where we're at. We're confident that we can do the measurement uncertainties to an accrediting body's standards. What I'm not sure about though is HOW to efficiently apply this process to a lab that does > 8k calibrations in a variety of disciplines each year. The number of calculations this represents quite daunting.

Fortunately we're not in a position where we have to drive our uncertainties down into the dirt. I'm simply looking for a healthy idea of where we're at and be able to meet the accrediting body's requirements. Currently we're generating one uncertainty budget for each point in a calibration. Then we take the budgets for that calibration and apply them to all instruments of that model number. Seems like we could be more efficient if we could somehow use one budget to cover a range of points or apply the budgets to more than one specific model of instrument. Does anyone know if this is acceptable?

Another question I have is about when to issue the full blown accredited certificates. All of our customers are part of the same company and most aren't concerned with measurement uncertainty. Is there a strategy to doing accredited vs. unaccredited cals? I ask this thinking that not having to do uncertainties for each calibration would be quicker and easier.

 

[Hershal]

You may consider doing your uncertainty budgets for each TYPE of calibration, rather than each specific calibration. Also, electronic usage may help such as a spreadsheet or another program. As an example, a budget could cover calipers over the range, with some appropriate breakdowns to manage the uncertainties, regardless of model number, instead of each specific point.

Redo the budget when your standards are re-calibrated.

If customers are all inside, then you will have to ask their needs. If they need an accredited certificate then give them one. If they don't, you still maintain the information so you can, but otherwise may not need to give an accredited cert.

Also, if you have not yet selected an accrediting body, then you can look at www.ilac.org and find the accrediting body or bodies for your location. If in the US, there are six that accredit cal labs.

Hope this helps.

 

[Rmack]

Thanks Hershal.

This gives me a whole bunch of new ways to look at this. Seems a lot more manageable looking at it this way.

Thanks again!

 

[jerryb1426]

Understanding and Expressing Measurement Uncertainty associated with Thermodynamic and Low Frequency Metrology ​

​

​

Introduction
Periodic calibration of electronic systems is required due to the large number and wide variety of components in basic measurement and sourcing systems. In metrology today a calibration laboratory needs to understand how to quantify all the influences and components of measurement uncertainty as they pertain to the laboratory’s environment. These uncertainty influences and components are as varied as the cables used in a measurement system to the training level of the technicians tasked with operating this system. Consistent results and confidence in the reported values of a measurement can be achieved with due consideration to all the contributors of uncertainty. This paper is intended to help the laboratory professional begin evaluation of the possible sources of uncertainty and how to formulate a measurement uncertainty budget.
Common terms:
Accuracy is a qualitative expression of the closeness of a measurement’s results to the true value.
Precision is a measure of repeatability. A high precision indicates the ability to repeat measurements within narrow limits.
Resolution is the smallest change that can bedetected. Generally today with modern instruments this is the smallest increment that can be displayed or LSD.
Uncertainty is a quantitative term that represents a range of values wherein the true value may lie. Uncertainty and confidence is determined using statistical techniques.

Traceability is the ability to relate individual measurement results to international SI Units or accepted measurement systems through an unbroken chain of comparisons.
Requirements for sound analysis:
Stable Environment; before a laboratory can begin to evaluate the components of uncertainty in a measurement or calibration system, data must be collected to determine if the system is stable. This data may be as simple as monitoring the ambient environment using a temperature and RH logger or as complex as repetitive measurement schemes ofthe equipment under evaluation.
Proper Training of Personnel; all personnel tasked with performing measurements to assimilation of collected data should beproperly trained and evaluated on their understanding of the tasks assigned to them.
Traceable Standards; all standards used in anuncertainty-testing scheme must be traceable for the results to be meaningful.
Uncertainty considerations; all possible sources of uncertainty should be considered from AC line voltage fluctuations to there solution of the measurement system. A source of uncertainty such as cable EMF may be discarded after determining that the uncertainty is insignificant. It may be appropriate in some test schemes to combine all of the insignificant uncertainties and create a label for this combined uncertainty Instrument specifications are the most common source of uncertainty data, however proper consideration must be given to the manufacturer’s stated confidence level. If the manufacturer did not specify a confidence level, then a rectangular distribution should be assumed, more on distributions later
Sources of Uncertainty
Uncertainty in the results of a measurement can be affected by many factors, some considerations:
Reference standards and measurement equipment: Uncertainty in their calibration; long term drift; resolution; vibration; electromagnetic interference; sensitivity to change during transportation and handling.
Measurement Setup: cables; shielding; warm up time; thermal voltage influences measurement probes.
Measurement Process: duration of the measurement; number of measurements conditioning of standards.
Environmental Conditions: temperature temperature oscillations; humidity electromagnetic influences; transients in power source.
Measurement errors
A measurement is subject to many sources of error, some of which can cause an over or under statement of the measurement quantity. While the goal of any metrology lab is to keep these errors small, they cannot be reduced to zero. The challenge for any metrology lab is to find out the quantity of these errors and how large they may be. Measurements are affected by three types of errors; Random, Systematic, and Gross.
Random errors are due to unknown causes and are only detectable when repeated measurements are made with a stable measurement setup and consistent measurement technique. This type of error will result in readings that, when repeated, are not always the same. If the reason for the variation is not obvious, then it falls into the category ofa random error. Note: Random errors cannot be quantified without a stable environment and consistent measurement technique.
Systematic errors relate to the equipment used in the measurement process or external influences on the equipment. Examples include: loading effects, thermals, drift-rate, leakage currents, and external noise.
Gross errors are caused by the technician and can be strictly controlled with proper training. Examples include: misreading of instrument results, incorrect adjustments, using the wrong instrument, errors in recording calibration data, and computational errors. All of these errors can be avoided with proper training and attention to detail.
Classifications of uncertainty
Type “A” evaluation method; the method of evaluation of uncertainty of measurement by the statistical analysis of a series of measurements. An example would the standard deviation of a series of measurements taken by a laboratory technician.
Type “B” evaluation method; the method of evaluation of uncertainty of measurement by means other than the statistical analysis of a series of observations. An example would be the manufacturer’s published specifications for an instrument.
Methods of determining uncertainty
Published specifications; as mentioned earlier published specifications are the most common source of uncertainty data used by commercial calibration laboratories. This method is the most appropriate for laboratories that take only simple measurements”. Simple measurements can be defined as any measurement that is within the common functional capabilities of an instrument. This method of determination is considered a Type “B” uncertainty.
Statistical methods; this method requires the taking of a series of measurements over a specified length of time. This method is the most robust and is appropriate for any laboratory that requires high confidence in their measurement uncertainty statements.
This is a Type “A” uncertainty.
You also can look at the NIST and NASA web sites
As a Good example Google NISTIR6919 on scale calibration.
ISO Lead Assessor