The Application of the FibreLux Instrument for Measuring the Diameter of Mohair Fibres

ABSTRACT

The current study investigated the possible application of the FibreLux Micron Meter^TM (FibreLux) instrument for the on-farm measurement of the diameter of greasy mohair staples. Initially, the effects of arrangement of the fibers in the specimen holder and that of solvent cleaning were investigated. It was found that a random arrangement of the fibers in the specimen holder produced more accurate and reproducible results, than when the fibers were aligned parallel (to the length of the specimen holder). The solvent cleaning had a statistically significant effect (p < .05) on the results, with the average of the mean fiber diameter (MFD) for the clean staple samples being 1.21 µm lower (finer) than that for the greasy staples. A good correlation (R = 0.95) was found between the FibreLux greasy staple results and those of the OFDA100 clean staples, the latter being used as the reference. The FibreLux values were, on average, some 0.80 µm lower than those of the OFDA100. It was concluded that the FibreLux holds significant promise for determining the fiber diameter of greasy mohair staples, but this should be validated by further trials, preferably “on farm” field trials, to determine the number of staples that need to be measured for each goat.

摘要

目前的研究调查了FibreLux Micron MeterTM（FibreLux）仪器的可能应用，该仪器最初是为在农场测量羊毛纤维直径而开发的，用于在农场测量油腻的马海毛订书钉的直径. 最初，研究了纤维在试样架中的取向或排列以及溶剂清洗的影响. 研究发现，与纤维平行排列（与样品支架的长度平行）相比，样品支架中纤维的随机排列产生了更准确和可重复的结果. 对油腻的缝钉样品进行溶剂清洁对结果有统计学意义（p < .05），清洁缝钉样品的平均纤维直径（MFD）比油腻缝钉样品低1.21 µm（更细）. 发现FibreLux油腻钉的结果与OFDA100清洁钉的结果之间存在良好的相关性（R = 0.95，n = 26），后者用作参考. FibreLux值平均比OFDA100低约0.80 µm. 得出的结论是，FibreLux在确定油腻马海毛订书钉的纤维直径方面有很大的前景，但这应该通过进一步的试验来验证，最好是“农场”实地试验，以确定每只山羊需要测量的订书钉数量，从而获得足够准确的结果. 此外，还需要定义山羊身上（例如，肋骨中部）应牵引缝钉以及应测量缝钉的位置.

KEYWORDS:

• FibreLux Micron Meter^TM
• mohair greasy staples
• on-farm
• OFDA
• fiber diameter

关键词:

• FibreLux Micron Meter^TM
• 马海毛油腻的主食，
• on-farm
• 在农场
• 纤维直径

Introduction

There is a worldwide trend toward natural fibers, owing to their highly desirable renewable, environmentally friendly and comfort-related properties. Nevertheless, global competition is becoming increasingly severe, with great demands being placed on quality, price, consistency of supply and environmental aspects. Globally, demand is increasing from the textile trade, processors and manufacturers for improvements in fiber quality, notably fiber fineness in the case of animal fibers, including its objective characterization, this being due to the fact that consumers worldwide are becoming increasingly discerning and demanding in terms of the quality, performance and price of the product they purchase. There can be little doubt that the fineness, or more correctly mean fiber diameter (MFD), of animal fibers is one of their most important characteristics from the point of view of both price and quality (application and performance, etc.). Even a small change (e.g. 1 µm) in MFD have a significant effect on mohair price (Hunter 1993), particularly at the finer end of the range. Fibre diameter is the most important factor determining the quality and price of animal fibres such as mohair, the finer the fibre, the higher the demand and price. It determines the processing behavior and performance, as well as product type and quality, the finer the fibre, the finer the yarn which can be spun and the lighter the fabric which can be produced, also finer fibres give a softer handle and also better against the skin comfort. It is, therefore, hardly surprising that the mean fiber diameter, e.g. determined by either Airflow, Projection Microscope (PM), Optical Fibre Diameter Analyser (OFDA) or Laser-Scan^TM, as reviewed by Botha and Hunter (Botha and Hunter 2010) and Cottle and Baxter (Cottle and Baxter 2015), is generally the first objectively measured characteristic of animal fibers, including mohair, with the distribution of fiber diameter and level of coarse fibers, generally also of textile significance in certain applications.

South Africa is presently the major and leading producer of mohair worldwide, accounting for some 50% of global mohair production, its mohair being regarded as of the best and most consistent quality available on account of its classing and preparation standards, consistency and uniformity, fineness and lack of kemp and other medullated fibres (Ammayappan, Jose, and Chakraborty 2016; Atav and Hunter 2023). Its mohair is virtually all produced in the Eastern Cape Province of South Africa, within a 300 km radius of Gqeberha (previously known as Port Elizabeth), by some 1200 commercial and small (communal) farmers (Louw 2018). In light of this and the above, a very valuable tool and aid for mohair growers and breeders, in terms of classing, breeding, trading, etc., would be the ability to cost-effectively and accurately measure fiber diameter (fineness) and its variability “on farm.” Some of the challenges faced by mohair growers and breeders include demand and price being greatly affected by fashion and vulnerability of mohair demand in terms of ethical production, sustainability and environmental friendliness.

Within this context, a great deal of effort has been devoted, over the past 100 years or so, toward developing better technologies, instrumentation, and methods, preferably portable and suitable for “on-farm” use, for the rapid, objective, and cost-effective measurement of the diameter (or fineness) of natural fibers, notably wool. These include direct measurements, such as microscopic, gravimetric, optical diffraction, airflow, photometric, and image analysis. Recent developments in this respect include the Portable Fibre Tester (PFT) (E. C. Quispe et al. 2020), which uses digital image analysis, the FIBRE-EC (M. D. Quispe et al. 2017), also based on digital image analysis, and the FibreLux (FL) (FibreLux, Inc., Johannesburg, South Africa) (see Figure 1), which is based upon light diffraction (Walker et al. 2018; Walker, Pope, and Pfeiffer 2021). More than 50 years ago an instrument, the Mikronmeter, based on the principle of light diffraction, was developed by the National Physical Research Laboratory of South Africa for the “on farm” measurement of wool mean fibre diameter (Boshoff and Kruger 1971). This device was easy to handle, robust, battery operated, and results could be obtained within a few seconds, with more accurate results being obtained with randomized clean fiber specimens (tufts), taken from the mid-rid of the sheep. A significant amount of training was required to obtain accurate results (Hunter 2014), and good precision would be achieved by measuring at least five sub-samples (fiber specimens) (David and Connell 1972). Both the production and marketing of the Mikronmeter were later discontinued due to certain operators, both in South Africa and Australia, finding it difficult to obtain accurate results. Some years later (in the early 1990s) the South African Wool and Textile Research Institute (SAWTRI), commenced with research aimed at the development of a robust, accurate, and reliable electronic version of the Mikrometer, with a first prototype being produced in 1998, a Provisional Patent being lodged in September 2000 in South Africa. This device made use of a laser light source and a linear two diode array, rather than a circular multiple diode array, but was found to produce inaccurate results. This approach was therefore discontinued (Hunter 2014). Further research revealed that a light emitting diode (LED), as a light source, was a better option, than the laser light source. Based on this, a battery operated, opto-electronic instrument (the FibreLux Micron Meter^TM) was developed and commercialized for the on-farm measurement of wool (see Figure 1). It was based upon the diffraction of LED generated light, by a fiber bundle mounted in a specimen holder; the diffraction pattern being detected and measured by an array of sensors and then analyzed using sophisticated algorithms and artificial intelligence. The sensory arrangement was configured to sense the diffraction pattern in at least two different wavelength bands. The processor was configured to evaluate the interference pattern at the respective wavelength bands and therefrom to calculate the mean fiber diameter of the fiber specimen, which consisted of an array of parallel or randomly arranged fibers mounted in a specimen holder. This specimen holder is inserted in a slot in the instrument (as demonstrated in Figure 1), with the mean fiber diameter being displayed digitally within 20 s (Hunter 2014).

Figure 1. Image of the FibreLux Micron Meter^TM with the specimen holder ready for measurement.

Source: https://.legendaryalpacas.com/products/1645/fibrelux-micron-meter.

Due to the fact that no similar instrument was available for the on-farm measurement of the fiber diameter (fineness) of mohair, the present study was undertaken to determine the suitability of the FibreLux Micron Meter™ instrument for measuring the mean diameter (fineness) of mohair fibers, preferably when in a staple and greasy form.

Experimental

Materials

In all, 30 samples of greasy staples, covering a wide range $(\approx$ 23.5–35.0 µm) of mean fiber diameter (MFD), considered representative of the MFD of the South African mohair clip, were selected for the study.

Methods

OFDA tests

The Optical Fibre Diameter Analyser (OFDA-100), an automatic image analysis-based fiber diameter test instrument (Qi et al. 1994), was used as the reference method in this study, with the tests being conducted according to the IWTO-0 (IWTO 2002) Appendix B for comparing different methods. Each of the 30 staples was manually (by hand) divided into two halves, each half being identified with the same unique sample number, one half for measurement on the OFDA and the other half for measurement on the FibreLux. Tests were carried out on both greasy and clean fibers. In the latter case, the staples were rinsed in dichloromethane (DCM) to remove dirt and grease, followed by oven drying at 105°C for 5 min and conditioned. During some initial tests on the OFDA, it was observed that the diameter of the fibers along the staple could vary considerably, mostly increasing from the staple tip to the root sometimes by as much as 4 µm, possibly due to the increased age and body weight/size of the goat during the approximately 6-month growth period (McGregor 1987; McGregor, Butler, and Ferguson 2012), such an effect being greatest for the kid and young goats. In light of the above, it was decided to always do the OFDA fiber diameter test on snippets cut (guillotined) some 2 cm from the root (base) end of the cleaned staples, the results being captured in Table 1.

Table 1. OFDA measured MFD and CV results*.

Sample ID	MFD (µm)	CV (%)
1	34.7	22.3
2	30.1	15.1
3	32.7	25.9
4	35.0	19.8
5	24.2	24.9
6	24.1	16.3
7	28.2	19.0
8	26.3	24.3
9	29.4	16.9
10	33.6	21.8
11	27.4	16.7
12	28.2	19.5
13	29.2	24.4
14	31.1	18.8
15	31.2	18.4
16	33.8	22.1
17	29.1	24.3
18	29.3	16.2
19	28.3	21.3
20	26.9	17.3
21	23.5	23.3
22	31.8	25.2
23	24.9	18.2
24	26.4	25.0
25	31.3	26.9
26	24.6	20.6
27	29.4	17.3
28	29.0	29.6
29	29.1	23.1
30	27.9	15.2

*Clean fibers.

FibreLux tests

Sample Preparation

Two different test specimens were prepared for measuring on the FibreLux, the one in which the fibers were approximately parallel to each other and the other where the fibers were more or less randomly arranged. The preparation of the first mentioned test specimen for the FibreLux involved manually teasing the fibers apart and carefully aligning them by hand, such that the large majority (portion) of fibers lie approximately parallel to each other. Thereafter, this was followed by combing them to complete the process of producing an array of individualized and reasonably parallel fibers. The preparation of the second specimen, namely randomized fibers, was also done by hand prior to insertion in the specimen holder (see Figure 2), care being taken to form as randomized as possible fiber arrangement in the specimen holder.

Figure 2. Fibre orientation in FibreLux specimen holder, (a) aligned and parallel with respect to the length of the specimen holder and (b) randomized.

Effect of fibre orientation

The effect of fiber orientation and alignment (see Figure 2), within the specimen holder, was investigated using 15 of the 30 greasy staples, the measurement being done both with the fibers arranged parallel in the specimen holder, as recommended for wool (Figure 2(a)), and also alternatively with the fibers arranged in a randomized fashion (Figure 2(b)) in the specimen holder.

Prior to testing, each staple was cut in the middle whereafter the base end half of the staple was placed in the specimen holder in such a way that only the middle portion of the base part was measured by the FibreLux. Similarly, the tip end measurements were made on the middle portion of the tip end half of the staple (see Figure 3).

Figure 3. Staple sections from which tip and root (base) measurements were made on the (a) OFDA and (b) the FibreLux.

Due to the small size of the staple samples, only two FibreLux measurements were performed on each of the staples, one on the base end and one on the tip end as just described. The results are presented in Figure 4 and Table 2. In Table 2, “Base” denotes the results obtained on the base section (half) of the staple and “Tip” the results obtained on the tip section (half) of the staple.

Figure 4. Scatter plots of FibreLux greasy MFD vs OFDA MFD (µm) (clean) for different fiber orientations, (regression and 1:1 lines superimposed).

Table 2. FibreLux measured MFD results for the greasy and clean staples and different fiber orientations.

Sample ID	OFDA MFD clean staples (µm)	Greasy staples (µm)				Clean staples (µm)
		Parallel arrangement		Random arrangement		Parallel arrangement		Random arrangement
		Base	Tip	Base	Tip	Base	Tip	Base	Tip
4	35.0	31.4	26.7	33.9	27.2	30.6	29.0	30.7	29.4
5	24.2	25.8	20.8	25.6	21.5	24.8	35.5	23.9	34.5
7	28.2	28.7	24.9	28.3	28.0	27.3	26.3	27.9	26.1
8	26.3	25.2	24.1	25.5	24.2	25.6	24.7	25.4	24.6
9	29.4	26.8	26.5	28.2	28.6	28.5	28.3	27.3	28.0
10	33.6	30.5	29.3	30.2	30.9	27.0	30.6	30.1	29.8
13	29.2	29.6	22.0	25.8	21.9	21.7	19.0	22.7	20.4
14	31.1	30.4	26.7	30.3	29.5	29.5	29.0	29.5	28.8
15	31.2	30.9	29.6	30.6	30.4	32.6	28.1	32.2	27.2
16	33.8	28.2	28.4	32.1	28.4	27.3	27.7	30.1	28.3
20	26.9	27.2	26.1	27.8	25.8	26.6	27.2	27.1	25.3
22	31.8	28.6	26.3	31.6	29.5	27.8	26.8	27.2	26.2
24	26.4	24.3	23.8	24.3	22.5	22.8	20.1	23.3	19.9
27	29.4	28.6	26.8	28.5	28.9	29.3	29.0	28.8	28.3
29	29.1	25.9	25.6	30.6	25.6	27.4	27.3	28.9	26.5
AVERAGE	29.7	28.1	25.8	28.9	26.9	27.3	27.2	27.7	26.9

Effect of solvent cleaning

The same 15 greasy staples, as above, were selected to investigate whether solvent cleaning had any effect on the FibreLux results. The greasy samples were cleaned, using 99% ethanol, and then given several rinses in distilled water, after which they were oven dried for 10 min at 150°C. The results are summarized in Table 2.

Statistical interpretation and analysis

The greasy staple MFD results of the FibreLux and OFDA were compared following the IWTO-00 (2002) procedure for comparison of methods, where the OFDA was considered to be the “reference method.” The observations that showed a significant difference (i.e., ≥5.0 µm) between the two orientations, the conventional parallel and the random arrangement, were judged as “outliers” and not included in the analysis. Sample results outside of the 95% confidence limit (1.96 × average standard deviation (D) of the regression) of the difference versus average regression were also judged as “outliers” and removed before the final analysis, in accordance with (IWTO-00, 2002). Statistical and comparative analyses of the data were done using the Origin Lab 8.0 (Origin Lab® Corporation, 2007) software package. The correlations between the OFDA100 and the FibreLux results were assessed using the Pearson correlation (R) matrix technique (Table 3).

Table 3. MFD correlation matrix for the FibreLux greasy and clean staples and the OFDA clean staples.

				FibreLux Greasy				FibreLux Clean
				Parallel		Random		Parallel		Random
			OFDA	Base	Tip	Base	Tip	Base	Tip	Base	Tip
		OFDA	1
FibreLux Grease	Parallel	Base	0.77	1
		Tip	0.75	0.56	1
	Random	Base	0.88	0.67	0.72	1
		Tip	0.70	0.63	0.90	0.72	1
FibreLux Clean	Parallel	Base	0.55	0.57	0.75	0.74	0.80	1
		Tip	0.11	0.18	0.23	0.37	0.35	0.51	1
	Random	Base	0.73	0.64	0.90	0.85	0.85	0.90	0.45	1
		Tip	0.19	0.26	0.21	0.41	0.34	0.50	0.98	0.45	1

Results and discussion

Effect of fibre orientation

The effect of the fiber arrangement (parallel and random) within the specimen holder was investigated using the OFDA generated MFD results as a basis of reference. As is evident from Table 3 and Figure 4, the random arrangement produced a slightly better correlation with the OFDA results, and it was therefore decided to use this particular fiber arrangement for all further tests.

According to the results given in Table 2, the average MFDs of the clean root (base) and tip section are very similar, while those of the greasy staple are not. Hence, it appears as if the observed differences in MFD between the root and tip sections of the greasy staples may be due to the differences in the grease levels between the base and the tip (due to weathering, etc.) rather than due to the effect of goat age/weight/size. This, however, requires further investigation.

Effect of solvent cleaning of greasy staples

The results of the investigation of the effect of solvent cleaning of the greasy staples on the FibreLux readings are summarized in Table 4.

Table 4. Effect of solvent cleaning of the greasy staples on the FibreLux measured MFD values for the “base” section of the staples (random fiber arrangement).

Statistic	Before cleaning	After cleaning
Number of observations (N)	15	15
Mean (µm)	28.9	27.7
Standard Deviation (µm)	2.77	2.82
SEM (µm)	0.72	0.73
Average Difference (µm)	1.21

As can be seen from Table 4, a statistically significant (p = .008) difference of 1.21 µm was observed between the solvent cleaned and greasy staple base sections, the MFD of the solvent cleaned fibers being on average 1.21 µm lower than that of the greasy fibers, presumably due to the removal of a layer of grease and other contaminants. All further tests were carried out on the greasy staples since this would be preferable for on-farm measurement purposes.

Measurement of greasy mohair staples

The accuracy of the FibreLux measured values for the 30 greasy mohair staples was investigated. First of all, the outlier analysis was conducted according to the IWTO-0 (IWTO 2002) procedure (see Figure 5) which indicated that four data points fell outside the 95% confidence limit (1.96 × standard deviation of the regression) of the difference versus average regression. Nonetheless, the application of the normality test based on the Kolmogorov–Smirnov test proved that the data was drawn from a normally distributed population.

Figure 5. Difference between reference method (OFDA) and alternative method (FibreLux) versus average of both methods.

After the four outliers were removed, the average difference between the MFD values of the FibreLux and OFDA was found to be 0.80 µm (p < .05), which is statistically significant at the 95% confidence limits, with the FibreLux readings, on average, lower than that of the OFDA (see Table 5).

Table 5. Summary statistics for comparing the greasy FibreLux MFD and the clean OFDA100 MFD results.

Statistic	FL	OFDA
Number of observations (N)	26	26
Overall Mean Fibre Diameter (µm)	27.9	28.7
Standard Deviation (µm)	3.00	3.19
SEM (µm)	0.59	0.63
Average Mean Difference (µm)	0.80 (µm)

The slope of the regression of the instrument differences versus the instrument averages was 0.06 ± 0.07, with a p-value of 0.38, indicating that the slope does not differ significantly from zero.

As can be seen from Table 5 and Figure 6, the overall average of the FibreLux and OFDA results are very similar, the average difference being only 0.80 µm.

Figure 6. Scatter plot of greasy staples FibreLux vs OFDA clean MFD, with the corresponding regression and 1:1 lines being superimposed.

As shown in Table 6, the estimated geometric mean slope was 1.06 ± 0.07, the p-value indicating that the slope does not differ statistically (p = .18) from unity.

Table 6. Geometric mean regression using the IWTO-0 appendix B method, after removal of outliers.

Statistic		Outliers removed
N		26
Estimated geometric mean slope		1.06
Standard error of slope:		0.07
Significance of slope	t-value	0.03
	Significance	0.97*
Significance of Correlation	R-Value	0.95
	t-Value	16.6
	Significance	<<< 0.001**

*Not statistically significant, **Statistically significant.

Figure 6 shows a scatter plot of the greasy FibreLux MFD results versus the clean OFDA MFD results, with the solid black OFDA 1:1 line and the red FibreLux versus OFDA regression line superimposed. As evident from Figure 6 and Table 6, the FibreLux and OFDA MFD results correlated very well (R = 0.95; p <<< 0.001), which appears to be accurate enough for practical purposes. Nevertheless, this needs to be confirmed on a larger sample population and preferably also by field trials.

According to the results reported above, it appears that the FibreLux Micron Meter^TM can determine the mean fiber diameter of greasy mohair staples with a fair degree of accuracy (±1 µm). The authors recommend, however, that further trials, preferably “on farm” be undertaken to validate these results under on-farm conditions and possibly also involving a larger and more diverse sample population, before any final recommendations can be made. The current study has managed to establish an appropriate and consistent procedure for preparing a properly randomized and representative sample in practice.

Conclusions

The accuracy and precision of the FibreLux Micron Meter^TM in determining the fiber diameter (fineness) of mohair in staple form has been investigated, with the OFDA100 being used as the reference method. Two different fiber specimen preparation and alignment procedures, i.e., parallel and random, in the FibreLux specimen holder, were investigated to determine which was better; it was found that the random fiber arrangement was better. It was also found that the solvent cleaning of the greasy staples had a statistically significant effect on the FibreLux mean fiber diameter (MFD) results, with the solvent cleaned staples reading on average 1.21 µm lower than the greasy ones, ascribed to the removal of the grease and other contaminants covering the greasy fiber surface. The FibreLux MFD results were, on average, 0.80 µm lower than those of the OFDA, with this difference being found to be statistically significant (p < .05). Despite the observed difference, the FibreLux MFD results correlated reasonably well (R = 0.95) with those of the OFDA, which may be good enough for practical applications. This, however, requires validation, preferably by field trials, before a final recommendation for the “on farm” application of the FibreLux for the measurement of mohair can be made.

Highlights

• The FibreLux Micron Meter^TM was originally developed for the measurement of the mean fiber diameter (MFD) of wool fibers. In the measurement of wool MFD, the fibers are prepared in such a way that the majority/large portion of fibers lies approximately parallel to each other prior mounting/inserting in the specimen FibreLux holder. Wool MFD measurements have been done on the parallel arrangement of fibers (i.e., the fibers lying parallel to the length of the specimen holder). This study investigated the best preparation for mohair staples and found that the random arrangement (fibers arranged in a random manner) rather than when the fibers were parallel to the length of the specimen holder.

• The study also investigated the effect of solvent cleaning on the MFD measurements, with the results showing that the solvent cleaning does not have any positive effect on FibreLux MFD results, with the solvent cleaned fibers measuring 1.21 µm finer (lower) than the greasy staples, hence further measurements were carried out on greasy staples. These results are proof that the FibreLux could be a suitable tool for on-farm application.

• In this research study, the authors found that the FibreLux base-end half MFD and tip-end half MFD measurements of the staples statistically significantly differ, with the tip-end half underestimating or measuring significantly lower than the OFDA100, hence further comparison between the FibreLux MFD values with those of the OFDA100 was done using the measurements taken from the base-end half of the staples.

• The current research study confirms the previous findings that the FibreLux measures lower than the OFDA, the FibreLux MFD results being, on average, 0.8 µm lower than the OFDA MFD values. Despite this difference, the FibreLux MFD correlated (R = 0.95) very well with that of the OFDA100, proving that the FibreLux holds a significant potential for “On-Farm” measurement of MFD of greasy mohair fibers. This, however, requires validation by field trials, preferably “On-Farm” using a larger set of greasy staple samples.

Acknowledgments

The authors gratefully acknowledge the financial support of Mohair South Africa (MSA) and the mohair samples supplied and OFDA tests done by Gubb & Inggs as well as the provision of certain research facilities by the Council for Scientific and Industrial Research (CSIR) and Post-doc scholarship by the Nelson Mandela University (NMU).

Disclosure statement

No potential conflict of interest was reported by the author(s).