Creating a measurement system analysis (MSA) procedure for spectroscopy measurements

Without confidence in your measurement system, you simply cannot be sure that your parts meet specification. Whether gauging composition or measuring thickness, out of spec parts result in rework, extra scrap, in-field failure and potential reputational damage.

Internationally recognised quality management systems (QMS) call for measurement system analysis (MSA) procedures. This means you can’t become certified for some international quality standards unless you have a robust MSA in place.

This, among other factors, will include a calculation of the measurement uncertainty for your spectroscopy measurements. For example, the IATF 16949 standard is the automotive industry’s most widespread standard for quality. The standard is designed to be implemented in conjunction with ISO 9001, and a clause in section 7 of the IATF 16949 deals exclusively with measurement system analysis. And for laboratories which are accredited according to DIN EN ISO/IEC 17025, calculating and applying the measurement uncertainty is essential.

MSA in Spectroscopy

In XRF, LIBS and OES analyses, the measurement system includes the following:

•  The operators of the spectroscopy equipment. This is likely to include their training and level of access to the instrument. For example, some analysers have more features available to more experienced personnel.

• Your testing process For example, selecting the right program when using the analyser and including calculations for measurement uncertainty. Ensuring the right procedures are used, such as taking a fixed number of measurements of the sample.
• The analyser (XRF, LIBS or OES) and associated software. Ensuring the analyser is able to meet the detection limits demanded by the final specification and that the instrument is being maintained properly.
• Calibration (reference) samples and calibration procedure. Setting a calibration frequency and ensuring the reference samples are suitable for the application. For example, the composition of the reference sample must be similar to the target composition of the production samples.
• The components to be measured. Ensuring there is consistency in the condition and preparation of the samples.
  
  
• Environmental factors. These include temperature, pressure and humidity, for example.

Implementing a MSA is similar to any quality management system. Firstly you ascertain what the correct procedures are, document them, and then take steps to ensure those procedures are met in production. Regular management review meetings to discuss whether quality standards are being met and where improvements can be made are necessary to maintain certification.

If you are selling into the automotive industry, or one like it that demands high quality standards, you’ll find that a robust measurement system analysis procedure that includes measurement uncertainty is essential. However, once the new MSA procedures are adopted into your operations, they become second nature. For example, if you’re implementing measurement uncertainty procedures, once these are set up correctly, most spectroscopy instruments have the capability to carry out the relevant statistical analysis calculations for you.

Measurement uncertainty in detail

A key part of an MSA procedure in spectroscopy is accurately calculating the margin of error in your measurements. There are a few methods you can choose for this, such as the ISO recognised GUM method, which requires you to have a calibrated reference sample. Or, if you don’t have a reference sample, you can estimate the margin of error using a different method.

Our Search for True Values guide shows you which method is best for your situation and walks through actual calculations using real results. Download of your copy here

Share this blog