4 February 1997 All that's woolly is the fibre – everything else is solid science

Carpet buyers would be surprised to know the extent to which, until recent years, the basis for estimating the potential performance of their purchase was largely subjective.

Look, feel and educated guessing were essentially the way carpets were assessed for their likely performance in use.

Times have, however, changed and now it can be said with accuracy – "tread confidently, for you tread on a whole lot of science".

What brought about the progress? Basically consumer demand and the challenge from synthetic fibres, says the Professor of Wool Science and Technology at New Zealand's Lincoln University, Errol Wood.

"The marketplace wanted woollen carpets that retained their original appearance throughout their lifetime of use ... it wanted all the special characteristics that wool offers, such as springiness, durability, and a superior ability to retain appearance, but it wanted these on a consistent basis. It was growing intolerant of problems such as excessive matting or compaction of the pile through wear, a problem which also led to alterations in the texture and colour appearance of the carpet because when a carpet's pile collapses, light is reflected differently.

"All of us working in the wool industry knew that in the basic wool fibre we had a marvellous product – absolutely unbeatable in many respects – but the challenge was to ensure consistency and predictability in its end use.

As a senior scientist working in the industry Wood had long been critical of the subjective way carpets were assessed for subsequent performance in use.

In a 1989 paper he lamented that "changes in carpet appearance cannot at present be measured objectively, and so it is necessary to rely on subjective techniques such as a panel of trained observers using a standardised grading system ... (an approach which is)... not entirely satisfactory".

He went on to say that despite the obvious need for instrumental methods of quantifying carpet texture and the changes it undergoes during wear, there had been relatively little progress in developing practicable systems.

Some of the more notable attempts to come up with something were the glass bead displacement test, contour film photography, opto-electronic scanning, densitometry and image analysis.

Professor Wood picks up the story: "Image analysis was starting to become widely used in medical diagnostics, the remote sensing of the earth from satellites and in automated inspection in industry. It attracted my interest and I applied the methods developed in these other fields to make progress with the problem of describing and quantifying carpet pile texture.

"Collaboration with overseas textile scientists tackling similar problems was useful, but as far as concepts from which to start were concerned we were virtually operating in a vacuum. Indeed, it was only through trial and error that we were able to come up with definitions for the key parameters for describing carpets, an essential first step before we could evaluate anything.

"Nevertheless, we were convinced that digital image analysis offered the most promise for developing a rapid and reliable instrumental method to accurately determine the presence of characteristics needed for consistent quality in carpets."

And so it has turned out. Take carpet pile texture, for example. There are at least eight aspects to its objective description – tuft definition, lightness, periodicity, directionality, coarseness, pile coherence, and the presence of localised and extended features. Image analysis methods have allowed the quantification of all of these and the potential application of the technology to quality control in carpet manufacture has been explored with local carpet mills.

So promising and diverse are the potential applications of image analysis to fibre and textile science that within Lincoln University's Department of Wool Science and Technology, it is now a ‘unifying theme’ touching on the bulk of the current research.

"Objectivity is the keynote," says Professor Wood, "and the department is now using the most modern technology in the data collection step of its research activities."

For example, in the measurement of wool fibre diameter, antiquated subjective assessment and error-prone first generation measurement technology are being replaced by the Optical Fibre Diameter Analyser (OFDA). The Lincoln department was one of the early groups to acquire this Australian-developed instrument and it has assisted in the development and ongoing refinement of the instrument by liaison with the inventors and participation in inter-laboratory trials.

OFDA is essentially an optical microscope with an attached video camera which scans fibre snippets sprinkled on a slide. Dedicated software takes the images from the camera and analyses them to extract the fibre diameter distribution statistics, especially the mean fibre diameter.

OFDA is proving versatile for other fibre measurement tasks as well, for instance Masters student Justine Lee is working with Lecturer Dr Peter Maher on its capabilities for assessing medullation levels. A technical word, yes, but if you've ever noticed hairy, undyed fibres in your woollen jacket or some finely woven garment, you'll know its effect.

Medullation is the presence of a hollow core within some wool fibres, especially the coarsest ones. Fibres with medullated cells tend to be more brittle than unmedullated ones and they reflect light differently when dyed, sometimes creating problems for the production of cloth with consistent appearance and colouration. On the other hand, the inclusion of a small proportion of medullated fibres in a carpet can contribute a natural character to the product.

An extension of the OFDA software combined with an illumination technique allows medullated fibres to be isolated and counted and Lincoln's Department of Wool Science and Technology has cooperated with OFDA's inventors, BSC Electronics, Perth, in this development in addition to taking part in international trials.

"Given that there is currently no rapid, commercially viable method for objectively measuring medullation, it's an exciting development," says Justine.

"Such a method will allow the industry to investigate whether current beliefs about the effects of medullation are correct."

Other Lincoln research using OFDA includes profiling variations in the cross-section and diameter of fibres along their length, and investigating the curvature of fibres.