Proposal for calibrating laboratory preparation systems and instruments to measure curvature on raw wool

Abstract

The measurement of fibre curvature has attracted increasing interest since it first became widely available using the OFDA (Brims, 1993) and subsequently the Laserscan (Dabbs et al, 1994). Use of the measurements was the subject of the Commercial Technology Forum held at the IWTO conference in Barcelona in 2002.

There have been many investigations into the use of curvature for measuring crimp and related characteristics of wool. Lobb et al (1997) reported on comparisons between OFDA and Laserscan measurements of curvature and highlighted some of the preparation factors that influence the measurement. Fish et al (1999, 2000) reported extensively on various aspects of preparation effects, and highlighted the fact that virtually all aspects of sample preparation may have an influence on mean curvature measurements. These papers also mentioned one problem that has delayed development of a draft test method for curvature - the lack of calibration standards or procedures.

Baxter (2002a) showed that the OFDA 100 and 2000 models are inherently relatively accurate when tested using graticules, and concluded that "Variability between instrument calibrations, together with measurement variability, indicates that the overall accuracy and precision of a single measurement on a graticule is of the order of ± 1 to 7 º/mm, depending on the level and degree of variability of curvature."

In examining the performance of the OFDA2000 instrument operating in 100 mode, Baxter (2002b) noted that the precision of the two models of the OFDA instrument used in a controlled trial in two laboratories was ± 4º/mm on a range of tops, but ± 14 º/mm on scoured wool samples. The large discrepancy in the 95% CL estimates was considered to be due to minicoring differences between the two laboratories. This may have been exacerbated by the fact that because of the trial design, only two laboratories were involved, (which was nevertheless adequate for the purpose of the project). The concerns raised by Raw Wool Group Report: 04 pevious authors about sample preparation effects are immediately apparent in this comparison of precision estimates for the two sample types.

In another paper to be presented to this conference, Baxter and Johnston (2002) indicate that the precision of measurements of curvature using a large number of OFDA2000 instruments is in the order of ± 7 º/mm on full-length top sliver and ± 9 º/mm on greasy wool staples.

It is apparent from this abbreviated history that measurements of mean fibre curvature on tops and greasy staples appear to be a good deal more stable than on minicored snippets prepared from scoured raw wool, and that the reproducibility of measurements on tops using the OFDA instruments is almost as good as may be obtained on graticules of known curvature values.

This report examines the possibility of using measurements on 2mm guillotined IH top snippets to calibrate the preparation and measurement system used for raw wool fibre curvature measurement.

It is concluded that on the basis of this preliminary work, it should be feasible to overcome one of the major constraints in progress on a test method for fibre curvature.

It is recommended that the fibre curvature working group be re-convened to re-examine the working group draft test method and to organise a round trial to test the proposed calibration concept.

Citation

"Proposal for calibrating laboratory preparation systems and instruments to measure curvature on raw wool", Peter Baxter, Raw Wool Group, Report 04, IWTO Nice Meeting, November 2002.