Study to relate mini-spun yarn tenacity with cotton fiber strength

Abstract

There has been great interest in assessing yarn tenacity directly from available cotton fiber property data acquired by various means, including high-volume instrumentation (HVI). The HVI test is a primary and routine measurement providing fiber properties to cotton researchers. Knowledge about yarn tenacity within a cotton cultivar or between cultivars could be useful with regard to understanding the selection of cotton cultivars. This study examined the effect of cotton growth location, crop year, and cultivar on three relationships (fiber strength versus fiber micronaire, yarn tenacity versus fiber micronaire, and fiber strength versus yarn tenacity), and found great variations in the Pearson correlation and the gradients of respective regression lines. Instead of developing linear regression models from HVI fiber properties to predict yarn tenacity, this study applied a simple ratio method (i.e. normalized fiber strength or yarn tenacity against five HVI fiber properties) to relate fiber strength with yarn tenacity. The short fiber index was found to have a greater effect on the correlation between modified yarn tenacity and modified fiber strength than micronaire, yellowness, upper-half mean length, or uniformity index. This result implied the feasibility of utilizing HVI fiber short fiber index and strength data, as a semiquantitative and fast approach, to compare yarn tenacity performance within a cotton cultivar or between cultivars.

Keywords

cotton fiber fiber property fiber strength yarn tenacity HVI regression

As a raw material, cotton fiber is primarily processed into yarns. Since yarn manufacturing requires specific equipment and uses a large amount of cotton fibers, substantial research has been undertaken to optimize the methods of predicting yarn properties, such as tenacity, evenness, hairiness, and elongation, from available properties of raw cotton fibers.^1–14 In general, two strategies (theoretical versus statistical) have been considered in these studies. The theoretical approach is based on certain assumptions and the generated models provide good information about interactions among different fiber properties and yarn characteristics.^1,2 The statistical approach correlates yarn quality characteristics with fiber properties through simple algorithms and a number of linear regression methods.^3–14 This statistical approach utilizes voluminous fiber property attributes available from routine, standard, and improved fiber property measurements, including those obtained using high-volume instrumentation (HVI); the advanced fiber information system, which measures fiber length, fineness, and maturity parameters; Favimat, which produces single-fiber tensile data; the fineness maturity tester, which generates fiber fineness and maturity data; and fiber cross-sectional image analysis.

Estimation of yarn tenacity directly from raw cotton fiber properties is beneficial to cotton researchers in developing next-generation genotypes, managing cotton crop growth, and spinning fibers into yarns. In the 1960s, Lord³ conceived the use of simple ratios that were derived from three parameters (fiber bundle strength, length, and fineness), and observed the increase in these simple ratios with yarn count-strength product values. Üreyen and Kadoglu⁴ observed a strong positive correlation between fiber and ring spun yarn strength, and found that fiber strength was the most important parameter for yarn tenacity. Also, they noted that fiber elongation, fiber length, length uniformity index (UI, a ratio of mean length to upper-half mean length (UHML, a measure of the upper-half mean length in a fiber sample)), fiber fineness, yarn count, yarn twist, roving count, and unevenness of roving are other parameters that can have significant impacts on yarn tenacity. Thibodeaux et al.⁶ examined the effect of short fiber content or short fiber index (SFI, a representative value for the amount of cotton fibers shorter than 12.7 mm) in raw cotton on yarn quality and found that a yarn strength model developed using the four basic HVI properties (strength, micronaire (MIC, a measure of the air permeability of compressed cotton fibers, often used as an indication of fiber fineness and maturity), short fiber content, and UI) alone was nearly as good as models developed using all 23 fiber properties from the advanced fiber information system, HVI, and Suter–Webb array methods. Liu et al.⁷ proposed seven simple algorithms that utilized different combinations of UHML, UI, short fiber content, strength, MIC, and maturity to model yarn tenacity and evenness. Cai et al.⁹ reported the effectiveness of fiber length parameters and their combinations in predicting yarn properties, such as strength and irregularity. Long et al.¹² assessed alternative cotton fiber property attributes and their relationship with yarn strength, and revealed that the substitution of either fiber fineness for MIC or single-fiber strength for bundle strength improved the prediction of yarn strength. Hequet and collaborators^8,10,11 tried to predict yarn properties from combined HVI and advanced fiber information system data, along with consideration of harvest method and cultivar information. In their continuous efforts, Yang and Gordon^13,14 reported the development and cross-validation of yarn quality prediction models in a program called Cottonspec, which includes a large database of fiber and yarn data, a number of statistical models, and a user interface. Their results demonstrated the predictive ability of the models after applying the criteria of selecting independent and statistically significant variables, correcting yarn data, and introducing a mill correction factor.

In an effort to predict (or correlate) yarn tenacity from (or with) fiber strength and other fiber properties,^1,2,4–14 the multivariable based regressions for yarn skein tenacity (Skein_ten) can be described by the general equation

\begin{matrix} {Skein}_{ten, i} = a_{str, i} \times {HVI}_{str, i} + a_{i 1} \times P_{1} + a_{i 2} \\ \times P_{2} + \dots + a_{in} \times P_{n} \end{matrix}

(1)

where i is the sample number, a is the regression coefficient, n is the number of fiber properties used, and P is a specific fiber property other than fiber HVI strength (HVI_str).

Unlike previous investigations, which developed a set of multiple variable regression models to predict yarn tenacity from HVI properties,^1–14 this study took a different strategy of exploring the relationship between fiber properties and yarn tenacity. The aim was to examine the correlations between modified yarn tenacity and fiber strength, using five different HVI fiber properties (MIC, SFI, yellowness (+b), UHML, and UI). One purpose was to seek a relatively simple and semiqualitative screening method for a rapid determination of yarn tenacity among different cotton cultivars, by analyzing the available HVI fiber properties using a simple ratio method approach instead of a multiple variable regression approach.

Materials and methods

Cotton samples

During the 2011, 2012, and 2014 crop years, four commercial cultivars (Deltapine (DP) 393, Fibermax (FM) 958, Phytogen 72, and UA 48), together with 16 additional breeding lines, were grown in four replicated field tests at the Clemson University Pee Dee Research and Education Center near Florence, SC (Florence), the Clemson University Edisto Research and Education Center near Blackville, SC (Blackville), and the North Carolina State University Sandhills Research Station near Jackson Springs, NC (Sandhills). The fields consisted of a Norfolk loamy sand soil in Florence, a Barnwell loamy sand soil in Blackville, and a Candor sand soil in Sandhills. Each trial was arranged in a randomized complete block design with four replications. Each entry was planted in a two-row plot of 10.7 m in length with a 96.5 cm spacing between rows. Plots were managed conventionally and followed established local practices.

Fifty bolls were hand harvested from each plot and ginned on a 10-saw laboratory gin. Collected cotton fibers were conditioned at a constant relative humidity of 65 ± 2% and a temperature of 21 ± 1℃ for at least 24 hours,¹⁵ prior to subsequent fiber and yarn property measurement. Table 1 summarizes the fiber replications of four cotton cultivars at each trial (three growth locations over three crop years), as well as fiber and yarn property collection dates. Some trials were not scheduled at specific locations or crop years. For the respective 2011, 2012, and 2014 crop year cottons, fiber properties were tested in August 2012, January 2013, and February 2015, while yarn properties were measured in February 2013, February 2014, and January 2015.

Table 1.

Fiber replications of four cotton cultivars at each trial (three growth locations × three crop years)

	Florence	Blackville	Sandhills	Date of fiber test	Date of yarn test
DP 393 (total: 32)
2011	4	4	4	August 2012	February 2013
2012		4	4	January 2013	February 2014
2014	4	4	4	February 2015	July 2015
FM 958 (total: 20)
2011	4	4	4	August 2012	February 2013
2012		4	4	January 2013	February 2014
2014
Phytogen 72 (total: 31)
2011	4	4	4	August 2012	February 2013
2012		3	4	January 2013	February 2014
2014	4	4	4	February 2015	July 2015
UA 48 (total: 32)
2011	4	4	4	August 2012	February 2013
2012		4	4	January 2013	February 2014
2014	4	4	4	February 2015	July 2015

Fiber property measurement

Average HVI properties were obtained from five replicates on each sample using an Uster HVI 1000 system (Uster Technologies Inc., Knoxville, TN), following an ASTM D5867 standard.¹⁶ All measurements were performed at the Southern Regional Research Center of USDA’s Agricultural Research Service (USDA-ARS-SRRC, New Orleans, LA). The same HVI instrument was used for all samples during the experiments.

Yarn tenacity measurement

Because only 70–90 g of fibers were available, a mini-spinning protocol was used to generate yarns. Approximately 60 g of each fiber was carded on a modified Saco Lowell Model 100 card.¹⁷ The carded web was drawn into a sliver on a modified Saco Lowell DF 11 draw frame. Two bobbins of ring spun yarn were then spun to a nominal count of Ne 30/1 with a twist multiple of 3.8. A 54.9 m (109.8 m for the 2012 cottons) mini-skein was produced from each bobbin and tested on an Instron tensile tester according to an ASTM D1578 standard.¹⁸ The yarn tenacity in this work was determined by skein testing, not single-strand testing. Owing to the small amount of fibers available in this study, a skein testing method was chosen because it has been the industry’s de facto method for small-scale fiber processing and for its ability to account for greater than normal yarn unevenness.^17,19,20

Mathematical and statistical analysis

Mathematical and statistical analyses were performed using Microsoft Excel 2016. For mathematical analysis, relationships between fiber strength and yarn tenacity were examined by fitting the data into a linear regression option. For statistical analysis, P and Pearson correlation coefficients between pairs of fiber and yarn properties were acquired using the Excel regression function under the data analysis tools.

Results and discussion

Fiber and yarn property characteristics

Table 2 summarizes the range, mean, and standard deviation (SD) of selected fiber and yarn properties for the four cultivars in three growth locations over three crop years. In general, the mean fiber MIC of FM 958 and Phytogen 72 cultivars were almost identical (4.63 versus 4.63), both were lower than those of DP 393 and UA 48 cultivars, which had average MIC values of 4.84 and 4.99. Fiber HVI strength ranged from 30.60, through 30.99 and 32.76, to 34.07 g/tex for DP 393, FM 958, Phytogen 72, and UA 48 cultivars, respectively. Coincident with fiber strength, UHML ranged from 28.7 to 30.7 mm, UI ranged from 83.75 to 84.71%, SFI ranged from 6.70 to 6.00%, +b ranged from 7.60 to 7.02, and yarn tenacity ranged from 44.24 to 48.77 g/tex. The tendency of the strength and tenacity values is not surprising, since cultivars, crop years and growth locations are known to affect values of many traits, including fiber strength and yarn tenacity. The trend in which tenacity values are always greater than strength values for a given sample is due to a clear difference between yarn tenacity determination and fiber strength measurement.

Table 2.

Comparison of range, mean, and standard deviation (SD) of selected high-volume instrumentation (HVI) fiber and skein yarn properties of four cultivars in three growth locations and three crop years with those of other breeding lines. Yarn tenacity is reported in g/tex units, which can be simply converted into typically reported units in either cN/tex (single-end) or mN/tex (skein). HVI typically reports fiber strength as g/tex, which is based on the estimated mass of fibers in the beard that has been broken by the HVI

	DP 393	FM 958	Phytogen 72	UA 48	Other lines
Micronaire (units)
Range	3.73–5.46	3.60–5.68	3.81–5.14	4.07–5.60	3.44–5.62
Mean	4.84	4.63	4.63	4.99	4.73
SD	0.48	0.60	0.35	0.42	0.45
Strength (g/tex)
Range	27.46–32.75	27.19–33.36	29.60–35.91	30.30–37.96	25.18–37.24
Mean	30.60	30.99	32.76	34.07	31.76
SD	1.25	1.60	1.91	1.69	2.05
Upper-half mean length (mm)
Range	26.4–32.2	25.9–32.2	26.7–31.8	27.4–33.8	25.4–33.5
Mean	28.7	29.2	29.7	30.7	29.5
SD	1.52	1.78	1.27	1.78	1.59
Uniformity index (%)
Range	81.44–85.56	81.20–86.52	82.24–85.97	82.48–88.45	80.27–87.64
Mean	83.75	83.86	84.00	84.71	83.88
SD	1.16	1.51	0.97	1.52	1.30
Short fiber index (%)
Range	4.56–8.70	3.91–10.95	4.52–8.92	2.57–8.52	3.06–13.36
Mean	6.70	6.53	6.58	6.00	6.75
SD	1.25	1.92	1.02	1.78	1.62
+b
Range	6.51–8.68	6.11–7.70	6.52–8.81	5.64–8.33	5.36–9.95
Mean	7.60	7.10	7.41	7.02	7.32
SD	0.58	0.36	0.50	0.54	0.74
Yarn tenacity (g/tex)
Range	33.71–54.17	36.27–50.24	42.54–56.94	36.39–58.48	25.87–76.06
Mean	44.24	45.45	49.29	48.77	46.50
SD	5.09	3.64	3.56	5.58	5.29

Fiber and yarn properties for 506 additional breeding lines grown in three locations and three years are included in Table 2 for comparison. The examined properties for the four cultivars are well within those usually obtained for these breeding lines.

Univariate correlation coefficients

Table 3 summarizes the Pearson correlation coefficients (R) among a total of 115 fibers. MIC showed positive and significant correlation with SFI, and negative and moderate correlations with UHML and UI, as well as minimal correlations with fiber strength, +b, and yarn tenacity. This observation echoes a previous report that cottons with identical MIC values can have different yarn strengths.²¹

Table 3.

Univariate correlation coefficients, R, for eight fiber and yarn properties. Absolute R values ≥0.50 were objectively considered to have significant correlation, values between 0.50 and 0.20 to have moderate correlation, and values ≤0.20 were minimally correlated

	Micronaire	Upper-half mean length	Uniformity index	Short fiber index	Strength	+b	Tenacity
Micronaire	1.00
Upper-half mean length	−0.47***	1.00
Uniformity index	−0.48***	0.82***	1.00
Short fiber index	0.54***	−0.74***	−0.85***	1.00
Strength	0.12*	0.57***	0.43***	−0.24***	1.00
+b	−0.08*	−0.04*	0.21**	−0.20**	−0.11*	1.00
Tenacity	−0.20**	0.34***	0.32***	−0.31***	0.37***	0.10*	1.00

***P < 0.001 (significance); **0.001 < P < 0.05; *P > 0.05 (no significance)

UHML showed a strong negative correlation with SFI and a strong positive correlation with UI and strength. Notably, UHML did not correlate strongly with yarn tenacity, nor did fiber strength correlate with yarn tenacity. Both UI and SFI had a strong but negative correlation. Yellowness, indicated by +b, did not exhibit any significant correlations with any of the fiber and yarn properties.

Unlike strong correlations between fiber tenacity (strength) and yarn tenacity (coefficient of determination, R² = 0.73–0.84) in a recent report on mill yarns,¹⁴ a moderate and positive correlation (R = 0.37) was noted between fiber strength and yarn tenacity in this study on mini-spun yarns. This is not unexpected, because cultivar, location, and year all affected the correlation between yarn tenacity and fiber strength in this study. It appears that the degree of the relationship between yarn tenacity and fiber strength differs by cultivar, location, year, and interactions among them. Meanwhile, fiber strength does not equate directly to yarn tenacity. When we measure cotton fiber strength, we break the fibers. When we measure yarn tenacity, the individual fibers slide past each other (function of length, fineness, length distribution, twist, surface morphology) and do not necessarily break.

Besides a strong correlation (R = 0.57) between fiber strength and UHML, neither fiber strength nor yarn tenacity were found to have a significant relationship with fiber MIC, UI, SFI, or +b. This observation echoes well a previous statement that yarn tenacity and fiber strength could be influenced by a number of factors, such as variations in the elastic properties of the fibers, coefficient of friction, mean fiber length and cross-sectional area, yarn evenness, and twist, as well as the poor status of quality control procedures in a mill.¹⁴

Variations in relating fiber HVI strength and yarn tenacity to MIC, also yarn tenacity to fiber HVI strength

To examine the effect of cotton growth location, crop year, and cultivar on the relationships between fiber HVI_str and MIC and between yarn skein tenacity (Skein_ten) and MIC, as well as between yarn Skein_ten and fiber HVI_str, fibers from each cultivar were further partitioned by crop year (2011, 2012, and 2014) or growth location (Blackville, Florence, and Sandhills). Table 4 summarizes the relationships of linear regressions for individual fiber sets. Throughout the manuscript, a negative value of R means that as one variable increases, the other tends to decrease in the data set, and vice versa. With regard to the relationship of HVI_str versus MIC, for the DP 393 cultivar, R ranges from −0.55 to −0.39 among fibers grouped by crop year and from −0.34 to 0.18 within fibers allocated by growth location. For the FM 958 cultivar, R varies from −0.36 to 0.13 among fibers grouped by crop year and from −0.57 to 0.04 within fibers grouped by growth location. For the Phytogen 72 cultivar, R ranges from −0.32 to 0.06 among fibers grouped by crop year and from 0.25 to 0.51 within fibers grouped by growth location. For the UA 48 cultivar, R ranges from 0.15 to 0.51 among fibers grouped by crop year and from 0.04 to 0.51 within fibers grouped by growth location. Meanwhile, the gradients vary from −2.7 to −0.7 among DP 393 crop year fibers and from −3.5 to 0.5 among DP 393 growth location fibers, −1.9 to 0.5 among FM 958 crop year fibers and −5.2 to 0.1 among FM 958 growth location fibers, −1.7 to 0.3 among Phytogen 72 crop year fibers and 1.5 to 3.5 among Phytogen 72 growth location fibers, and also 0.4 to 3.0 among UA 48 crop year fibers and 0.3 to 7.1 among UA 48 growth location fibers.

Table 4.

Comparison of regressions relating either fiber strength or yarn tenacity to micronaire (MIC) and also yarn tenacity to fiber high-volume instrumentation (HVI) strength for four cultivars between fibers, dependent on crop year and growth location

Cultivar	HVI_str vs. MIC	Skein_ten vs. MIC	Skein_ten vs. HVI_str
DP 393
Crop year
2011	HVI_str = −2.7 × MIC + 44.0 R = −0.55	Skein_ten = −10.2 × MIC + 97.0 R = −0.66	Skein_ten = 2.2 × HVI_str − 21.0 R = 0.69
2012	HVI_str = −0.7 × MIC + 33.2 R = −0.39	Skein_ten = −9.4 × MIC + 85.0 R = −0.66	Skein_ten = 2.5 × HVI_str − 28.3 R = 0.33
2014	HVI_str = −1.5 × MIC + 38.7 R = −0.44	Skein_ten = −9.2 × MIC + 88.4 R = −0.49	Skein_ten = 3.8 × HVI_str − 78.3 R = 0.69
Growth location
Blackville	HVI_str = 0.4 × MIC + 29.1 R = 0.18	Skein_ten = −7.4 × MIC + 79.4 R = −0.50	Skein_ten = 3.0 × HVI_str − 46.5 R = 0.46
Florence	HVI_str = −3.5 × MIC + 48.7 R = −0.34	Skein_ten = −13.8 × MIC + 115.0 R = −0.48	Skein_ten = 0.0 × HVI_str + 41.5 R = 0.00
Sandhills	HVI_str = 0.5 × MIC + 28.3 R = 0.13	Skein_ten = 0.5 × MIC + 42.2 R = 0.07	Skein_ten = 1.1 × HVI_str + 11.6 R = 0.56
All	HVI_str = −0.2 × MIC + 31.7 R = −0.08	Skein_ten = −5.2 × MIC + 69.6 R = −0.50	Skein_ten = 1.6 × HVI_str − 5.5 R = 0.40
FM 958
Crop year
2011	HVI_str = 1.9 × MIC + 40.2 R = −0.36	Skein_ten = −3.8 × MIC + 63.8 R = −0.31	Skein_ten = 1.1 × HVI_str + 11.9 R = 0.45
2012	HVI_str = 0.5 × MIC + 29.8 R = 0.13	Skein_ten = 1.6 × MIC + 40.8 R = 0.21	Skein_ten = 0.3 × HVI_str + 37.8 R = 0.14
Growth location
Blackville	HVI_str = 0.1 × MIC + 32.0 R = 0.04	Skein_ten = −2.4 × MIC + 57.6 R = −0.56	Skein_ten = −2.1 × HVI_str + 114.6 R = −0.76
Florence	HVI_str = −5.2 × MIC + 55.0 R = −0.57	Skein_ten = −22.0 × MIC + 152.4 R = −0.65	Skein_ten = 3.6 × HVI_str − 62.2 R = 0.94
Sandhills	HVI_str = −0.4 × MIC + 32.9 R = −0.29	Skein_ten = −0.6 × MIC + 48.6 R = −0.16	Skein_ten = −0.7 × HVI_str + 68.4 R = −0.33
All	HVI_str = −1.2 × MIC + 36.6 R = −0.46	Skein_ten = −2.5 × MIC + 57.2 R = −0.42	Skein_ten = 1.1 × HVI_str + 10.6 R = 0.50
Phytogen 72
Crop year
2011	HVI_str = −1.7 × MIC + 39.9 R = −0.32	Skein_ten = −7.5 × MIC + 86.2 R = −0.47	Skein_ten = 1.7 × HVI_str − 3.4 R = 0.54
2012	HVI_str = −0.3 × MIC + 31.9 R = −0.31	Skein_ten = −12.5 × MIC + 101.0 R = −0.69	Skein_ten = 7.6 × HVI_str − 185.8 R = 0.35
2014	HVI_str = 0.3 × MIC + 33.5 R = 0.06	Skein_ten = 1.6 × MIC + 40.1 R = 0.15	Skein_ten = 1.1 × HVI_str + 88.0 R = 0.41
Growth location
Blackville	HVI_str = 2.6 × MIC + 20.8 R = 0.51	Skein_ten = −5.9 × MIC + 76.0 R = −0.68	Skein_ten = −0.2 × HVI_str + 43.8 R = −0.38
Florence	HVI_str = 3.5 × MIC + 16.6 R = 0.37	Skein_ten = −4.9 × MIC + 74.1 R = −0.15	Skein_ten = −0.3 × HVI_str + 60.0 R = −0.15
Sandhills	HVI_str = 1.5 × MIC + 25.6 R = 0.25	Skein_ten = −1.8 × MIC + 56.8 R = −0.16	Skein_ten = 0.5 × HVI_str + 32.6 R = 0.23
All	HVI_str = 2.4 × MIC + 21.5 R = 0.44	Skein_ten = −2.8 × MIC + 62.4 R = −0.28	Skein_ten = 0.0 × HVI_str + 50.1 R = 0.01
UA 48
Crop year
2011	HVI_str = 1.6 × MIC + 24.9 R = 0.21	Skein_ten = 1.1 × MIC + 46.1 R = 0.07	Skein_ten = 0.7 × HVI_str + 28.9 R = 0.33
2012	HVI_str = 0.4 × MIC + 32.3 R = 0.15	Skein_ten = 15.3 × MIC−18.9 R = 0.71	Skein_ten = 6.3 × HVI_str − 166.0 R = 0.74
2014	HVI_str = 3.0 × MIC + 19.2 R = 0.51	Skein_ten = 10.1 × MIC − 7.2 R = 0.47	Skein_ten = 0.9 × HVI_str + 15.5 R = 0.24
Growth location
Blackville	HVI_str = 1.2 × MIC + 28.4 R = 0.51	Skein_ten = −1.1 × MIC + 51.0 R = −0.07	Skein_ten = 0.7 × HVI_str + 22.5 R = 0.12
Florence	HVI_str = 7.1 × MIC − 4.2 R = 0.47	Skein_ten = −2.3 × MIC + 63.5 R = −0.07	Skein_ten = 0.0 × HVI_str + 51.9 R = 0.01
Sandhills	HVI_str = 0.3 × MIC + 32.4 R = 0.04	Skein_ten = −0.2 × MIC + 51.2 R = −0.01	Skein_ten = 0.1 × HVI_str + 45.2 R = 0.07
All	HVI_str = 0.8 × MIC + 29.8 R = 0.21	Skein_ten = 2.1 × MIC + 38.3 R = 0.16	Skein_ten = 0.1 × HVI_str + 44.9 R = 0.03

On the relationship of Skein_ten versus MIC in Table 4, the tendency of negative or positive values of R from Skein_ten versus MIC is similar to that of HVI_str versus MIC for all fibers for a given crop year, regardless of cotton cultivars, whereas this tendency changes for fibers for a given growth location, from DP 393 and FM 958 cultivars (i.e. Blackville) to Phytogen 72 and UA 48 cultivars (i.e. Blackville, Florence, and Sandhills). Again, the gradients vary from −10.2 to −9.2 among DP 393 fibers grouped by crop year and from −13.8 to 0.5 among DP 393 fibers grouped by growth location, −3.8 to 1.6 among FM 958 fibers grouped by crop year and −22.0 to −0.6 among FM 958 fibers grouped by growth location, −12.5 to 1.6 among Phytogen 72 fibers grouped by crop year and −5.9 to −1.8 among Phytogen 72 fibers grouped by growth location, and 1.1 to 15.3 among UA 48 fibers grouped by crop year and −2.3 to −0.2 among UA 48 fibers grouped by growth location.

When relating yarn Skein_ten to fiber HVI_str in Table 4, all values of R except four fiber sets dependent on growth location (i.e. Blackville and Sandhills for the FM 958 cultivar and Blackville and Florence for the Phytogen 72 cultivar) were observed to be positive. The gradients range from 2.2 to 3.8 among DP 393 fibers grouped by crop year and from 0.0 to 3.0 among DP 393 fibers grouped by growth location, 0.3 to 1.1 among FM 958 fibers grouped by crop year and −2.1 to 3.6 among FM 958 fibers grouped by growth location, 1.1 to 7.6 among Phytogen 72 fibers grouped by crop year and −0.3 to 0.5 among Phytogen 72 fibers grouped by growth location, and 0.7 to 6.3 among UA 48 fibers grouped by crop year and 0.0 to 0.7 among UA 48 fibers grouped by growth location.

Both the values of R and the gradients from yarn Skein_ten to fiber HVI_str in Table 4 were more enhanced when the four cultivar fibers were reclassified according to crop year, rather than growth location. For example, the overall value of R of 0.40 for all DP 393 fibers represents a range of 0.33 to 0.69 among fibers grouped by crop year or a range of 0.00 to 0.56 within fibers grouped by growth location, while the overall gradient of 1.6 from all DP 393 fibers reflects a range of 2.2 to 3.8 among fibers grouped by crop year or a range of 0.0 to 3.0 within fibers grouped by growth location. For the FM 958 cultivar, the overall value of R of 0.50 reflects a range of 0.14 to 0.45 among fibers grouped by crop year and a range of −0.76 to 0.94 within fibers grouped by growth location, while the overall gradient of 1.1 originates from a range of 0.3 to 1.1 among fibers grouped by crop year or a range of −2.1 to 3.6 within fibers grouped by growth location. For the Phytogen 72 cultivar, the overall value of R of 0.01 originates from a range of 0.35 to 0.54 among fibers grouped by crop year and a range of −0.38 to 0.23 within fibers grouped by growth location, while the overall gradient of 0.0 represents a range of 1.1 to 7.6 among fibers grouped by crop year or a range of −0.3 to 0.5 within fibers grouped by growth location. For the UA 48 cultivar, the overall value of R of 0.03 represents a range of 0.24 to 0.74 among fibers grouped by crop year and a range of 0.01 to 0.12 within fibers grouped by growth location, while the overall gradient of 0.1 reflects a range of 0.7 to 6.3 among fibers grouped by crop year or a range of 0.0 to 0.7 within fibers grouped by growth location. In general, the values of R and the gradients were improved when regrouping the Phytogen 72 or UA 48 fibers by respective crop year fibers, as opposed to the DP 393 and FM 958 cultivars.

Large variations in the values of R and the gradients of fiber strength versus MIC, yarn tenacity versus MIC, and yarn tenacity versus fiber strength underscore the challenge of exploring a unique response between the targeted properties, even from the same cultivar. This demonstrates that genotype and environment, as well as their interactions, greatly influence fiber and yarn property characteristics. Therefore, adding diverse fibers to a database might not provide the desired results.

Predicting yarn tenacity from fiber HVI strength and properties

Both yarn tenacity and fiber strength reflect the external force-induced breaking of cotton yarn and fiber, respectively, but the breaking mechanism between the two differs. In general, yarn breaking is merely forcing fibers to slide past each other, whereas fiber material breaking is breaking individual fibers in a fiber bundle. As discussed, yarn tenacity and fiber strength could be affected by cultivar, location, and crop year. A number of collective factors, including those discussed before,¹⁴ result in a moderate correlation of R = 0.37 (P < 0.001) between yarn Skein_ten and fiber HVI_str, as shown in Figure 1. In other words, only one fiber property (HVI_str) is considered to contribute to Skein_ten in Figure 1. The average value of HVI_str and the SD for all samples are 32.21 g/tex and 0.96 g/tex, respectively, whereas the average value of Skein_ten and the SD were 47.07 g/tex and 3.60 g/tex. Overall, the SD obtained for Skein_ten measurement is greater than that for HVI_str.

Figure 1.

Yarn skein tenacity (Skein_ten) as a function of fiber HVI strength (HVI_str) for four cultivars. Error bars represent the standard deviation (SD) of five HVI_str replicates along the x-axis and the SD of two Skein_ten replicates along the y-axis for each sample.

Figure 2 suggests a strong linear correlation (R = 0.78, P < 0.001) for the same data set as Figure 1, when both yarn tenacity and fiber strength were normalized by corresponding fiber MIC. Probably, MIC-modified yarn tenacity is strongly related to MIC-modified fiber strength. Averages of HVI_str/HVI_mic and SD for all samples are 6.79 g/(tex) and 0.25 g/(tex), respectively, whereas averages of Skein_ten/HVI_mic and SD are 9.96 g/(tex) and 0.74 g/(tex). In the 1960s, Lord calculated the ratios from the simple formula S × L/H (S, fiber bundle strength; L, effective fiber length; H, standard fiber fineness) and demonstrated that the yarn count-strength product values increased with fiber algorithm ratio.³ During the model development in the Cottonspec program, normalization of yarn tenacity values with fiber tenacity (or strength), yarn evenness, and twist values was considered.¹⁴

Figure 2.

Micronaire-corrected yarn skein tenacity (Skein_ten/HVI_mic) as a function of fiber high-volume instrumentation (HVI) strength (HVI_str/HVI_mic) for four cultivars. Error bars represent the standard deviation (SD) of five HVI_str/HVI_mic replicates along the x-axis and the SD of two Skein_ten/HVI_mic replicates along the y-axis for each sample.

Next, both yarn tenacity and fiber strength were divided by fiber MIC at elevating powers or exponents (i.e. or Skein_ten/HVI_MIC ⁿ , n = 1, 2, 3, 4, 5); a plot of R against the power n is given in Figure 3. Along with n = 0 to 5, R increased, as expected, from 0.37 to 0.98, implying the sensitivity of fiber or yarn mechanical properties to the inherent fiber maturity and fineness attributes that determine fiber HVI MIC.²²

Figure 3.

Pearson correlation R (normalized yarn tenacity against normalized fiber strength) as a function of power of one high-volume instrumentation (HVI) property. HVI properties include short fiber index (SFI), micronaire (MIC), yellowness (+b), upper-half mean length (UHML), and uniformity index (UI).

To verify whether other fiber properties have similar effects on modified fiber strength or yarn tenacity as MIC, a further four HVI properties (UHML, UI, SFI, and +b) were analyzed using the same procedure; the results are also compiled in Figure 3. The trend of R indicates that, in general, SFI has the most impact on the correlation between modified fiber strength and yarn tenacity, followed by MIC, +b, UHML, and UI. This observation is in good agreement with previous results showing that fiber maturity, an attribute that determines fiber MIC, is a “second tier” fiber property that has a small effect on yarn property within a certain range.¹⁴

With an increase of power n from 0 (i.e. no normalization process) to 5 (normalized five times by the same parameter), R increased in nearly identical patterns but differing increments for SFI, MIC, and +b, decreased initially prior to a rapid ascent for UHML and exhibited the least variation for UI. It is likely that this suggests more significant effects of SFI, MIC, +b, and UHML on modified fiber or yarn strength than that of UI.

The value of R easily reached over 0.90 when power n = 1 for SFI (Figure 3), while it needed a power of two for MIC and a power of more than three for +b, UHML, and UI. Therefore, SFI normalized yarn tenacity is the most highly correlated with SFI normalized fiber strength, of the six fiber properties examined. Probably, fibers with a lower SFI will have a higher yarn tenacity and are normalized by a smaller number, and vice versa. Clearly, further study is needed to understand the SFI determination mechanism and more diverse fibers should be studied to confirm this observation. In a previous study,²³ SFI was observed to correlate well with short fiber measurements made using other instruments (advanced fiber information system and Suter–Webb array), and the reciprocal of SFI was reported to have a significant contribution from two HVI variables (UI and UHML). Figure 4 shows a plot of SFI normalized yarn tenacity against normalized fiber strength (R = 0.95, P < 0.001), regardless of fiber cultivar, growth location, and crop year. The relationship between SFI normalized yarn tenacity and normalized fiber strength for all fibers was estimated using a linear regression function in Microsoft Excel 2016; this is inserted in Figure 4.

Figure 4.

Short fiber index (SFI) normalized yarn skein tenacity (Skein_ten/HVI_sfi) as a function of fiber HVI strength (HVI_str/HVI_sfi) for four cultivars.

The equation Skein_ten/HVI_sfi = 1.4 × HVI_str/HVI_sfi + 0.22 in Figure 4 was applied to predict Skein_ten/HVI_sfi, from measured HVI_str and HVI_sfi readings of 63 cotton fibers. These samples represented three breeding lines grown under identical conditions to the four cultivars over three locations and three years, but they were not used to generate the equation. A comparative scatter plot of predicted against measured Skein_ten/HVI_sfi, shown in Figure 5, is very promising, with R = 0.93 and P < 0.001. As a comparison, Figure 6 shows a plot of the measured Skein_ten values against the predicted Skein_ten values from the equation Skein_ten = 0.88 × HVI_str + 18.7 given in Figure 1, resulting in R = 0.26 and P = 0.042. There are 27 and 48 of 63 samples out of the 95% confidence interval in Figures 5 and 6, respectively. This result could imply the feasibility of using fiber SFI and strength to estimate yarn tenacity performance semiquantitatively and rapidly. Clearly, more diversified fibers are necessary to strengthen the robustness and efficiency of this simple approach.

Figure 5.

Correlation between predicted and measured Skein_ten/HVI_sfi values for three breeding lines over three crop years and three growth locations. The equation Skein_ten/HVI_sfi = 1.4 × HVI_str/HVI_sfi + 0.22 shown in Figure 4 was used to predict Skein_ten/HVI_sfi.

Figure 6.

Correlation between predicted and measured Skein_ten values for three breeding lines over three crop years and three growth locations. The equation Skein_ten = 0.88 × HVI_str +18.7 shown in Figure 1 was used to predict Skein_ten values.

Conclusions

Fiber property data obtained from routine and standard HVI fiber testing procedures have been used to predict yarn tenacity. One aspect of this study was to examine the effect of cotton growth location, crop year, and cultivar on the relationship between fiber strength and fiber MIC, mini-spun yarn tenacity, and fiber MIC, as well as between fiber strength and yarn tenacity. The observation addressed the complexity of understanding the unique response between yarn tenacity and fiber strength even within one cultivar. Instead of developing linear regression models from available HVI fiber properties to predict yarn tenacity, this study examined the effects of five HVI fiber properties on correlations between modified fiber strength and yarn tenacity. The results indicated that SFI could have a more significant effect on the correlation between normalized yarn tenacity and normalized fiber HVI strength than MIC, +b, UHML, and UI. This finding suggests the feasibility of applying fiber SFI and strength properties, as a semiquantitative and fast tool, to compare yarn tenacity performance in cotton cultivars.

Footnotes

Acknowledgments

The authors would like to acknowledge all technical supporters at both ARS locations (New Orleans, LA and Florence, SC) for their diligent work in field, mill, and laboratory operations.

Mention of a product or specific equipment does not constitute a guarantee or warranty by the US Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Frydrych

. A new approach for predicting strength properties of yarns. Text Res J 1992; 62: 340–348.

Pan

Hua

Qiu

. Relationship between fiber and yarn strength. Text Res J 2001; 71: 960–964.

Lord

The characteristics of raw cotton. In: Coulson

AFW

Tordoff

(eds). Manual of cotton spinning 1961; Vol. 2, part 1Manchester: The Textile Institute, pp. 310–311.

Üreyen

Kadoglu

. Regressional estimation of ring cotton yarn properties from HVI fiber properties. Text Res J 2006; 76: 360–366.

Ramey

Jr Lawson

Worley

Jr . Relationship of cotton fiber properties to yarn tenacity. Text Res J 1977; 47: 685–691.

Thibodeaux

Senter

Knowlton

, et al. The impact of short fiber content on the quality of cotton ring spun yarn. J Cotton Sci 2008; 12: 368–377.

Liu

Yang

Wang

, et al. Comprehensive characterization model of integrated cotton fiber quality index. J Donghua Univ 2011; 28: 379–383.

Faulkner

Hequet

Wanjura

, et al. Relationships of cotton fiber properties to ring-spun yarn quality on selected high plains cottons. Text Res J 2012; 82: 400–414.

Cai

Cui

Rodgers

JE III

, et al. A comparative study on cotton fiber length parameters’ effects on modeling yarn property. Text Res J 2013; 83: 961–970.

10.

Kelly

Hequet

. Breeding for improved yarn quality: Importance of non-HVI fiber properties. ICAC Recorder. 2013; 31: 13–20.

11.

Kelly

Hequet

Dever

. Breeding for improved yarn quality: Modifying fiber length distribution. Ind Crop Prod 2013; 42: 386–396.

12.

Long

Bange

Delhom

, et al. An assessment of alternative cotton fibre quality attributes and their relationship with yarn strength. Crop and Pasture Sci 2013; 64: 750–762.

13.

Yang S and Gordon S. Spinning performance of Australian long staple cotton Sicala 340BRF. In: Beltwide Cotton Conferences, San Antonio, USA, 7–10 January 2013, pp. 1069–1073. Cordova: National Cotton Council.

14.

Yang

Gordon

. Accurate prediction of cotton ring-spun yarn quality from high-volume instrument and mill processing data. Text Res J 2017; 87: 1025–1039.

15.

ASTM D1776/D1776M-12:2016. Standard practice for conditioning and testing textiles.

16.

ASTM D5867-12:2016. Standard test method for measurement of physical properties of cotton fibers by high volume instruments.

17.

Manandhar

Delhom

. Miniature spinning: An improved cotton research tool. J Cotton Sci 2018; 22: 126–135.

18.

ASTM D1578-12:2016. Standard test method for breaking strength of yarn in skein form.

19.

Landstreet

Edwald

Hutchens

. The 50-gram spinning test: Its development and use in cotton-quality evaluation. Text Res J 1962; 32: 665–669.

20.

Landstreet

Edwald

Kerr

. 1959. Miniature spinning test for cotton. Text Res J 1959; 29: 699–706.

21.

Abbot

Higgerson

Long

, et al. An instrument for determining the average fiber linear density (fineness) of cotton lint samples. Text Res J 2009; 80: 822–833.

22.

Lord

. Air flow through plugs of textile fibers: Part II. The micronaire test for cotton. J Text Inst 1956, pp. T16–T47.

23.

Thibodeaux

Senter

Knowlton

, et al. A comparison of methods for measuring the short fiber content of cotton. J Cotton Sci 2008; 12: 298–305.