157

FEBRUARY 2005

157

Comparison of the Performance Properties of Carpets Containing Nylon 6 and Nylon 66 Face Yarns P. RADHAKRISHNAIAH Georgia Institute of Technology, Atlanta, Georgia, 30032, U.S.A. ABSTRACT A parametric analysis of the performance properties of a group of twenty-four well balanced carpet samples, half of which represent nylon 6 and the other half nylon 66 face yarns, shows that nylon 66 carpets offer superior performance in terms of texture retention and color fastness to ozone. There are no significant differences in measured performance properties between nylon 6 and 66 carpets in terms of soil repellency, oil repellency, water repellency, color fastness to light, color fastness to nitrous oxides, and thickness recovery from prolonged application of a static load. A nonparametric evaluation of an expert grader’s assessment through a paired comparison of nylon 6 and 66 carpets subjected to wear on a Vettermann drum tester shows that the grader overwhelmingly prefers nylon 66 over nylon 6 (fifty times out of fifty-four). The comparison, however, is between identically constructed carpets and not between carpets whose construction has been optimized for the specific fiber type. Since the texture of the finished carpet is believed to be influenced by a host of construction parameters such as fiber denier, fiber modulus, fiber cross-sectional shape, yarn denier, yarn twist, heat setting conditions, pile height, pile weight, etc., it would be interesting to see how a fiber-specific carpet construction and design optimization would influence the texture comparison. Both pile type and pile yarn weight exert an influence on texture retention properties as measured on the Vettermann tester. In addition, pile type influences soil repellency. Fluoropolymer treatment has a significant positive influence on color fastness to ozone, soil repellency, oil repellency, and water repellency properties.

Historically, studies of nylon 66 and nylon 6 polymers have focused on structure development, physical and mechanical properties, structure-property relationships, and morphology. Thus, many fiber and polymer properties such as melt behavior, tensile properties, thermal and oxidative stability, crystallinity, and orientation have been extensively investigated [2, 5–7, 9, 13]. While differences in some of these properties appear to be significant, nylon 66 and nylon 6 continue to be used more or less interchangeably for many of the same end uses, perhaps due to the similarity of many of their properties. According to Prevorsek et al. [12], while it is possible to impart widely differing properties to both nylon 6 and 66 fibers through process changes and small chemical modifications, the performance of nylon 6 has seldom been equivalent to nylon 66 for the same application. This, however, does not mean that nylon 66 is automatically preferable to nylon 6, because certain end uses may not require the special characteristics of nylon 66, or specific circumstances associated with product design and application may mask any polymer differences. Textile Res. J. 75(2), 157–164 (2005)

When it comes to carpet performance, it is well understood by the trade that in most constructions, nylon significantly outperforms the less resilient polypropylene and polyester fibers. However, the differences in performance properties of carpets made from nylon 66 and nylon 6 fibers have been viewed more or less as subtle differences. Interestingly, there are no significant published data available describing performance differences, if any, between nylon 6 and nylon 66 carpets. In 1995, Werny [16] judged nylon 66 to have a slight texture recovery advantage over nylon 6 in carpets, but his claim has not been supported by data collected from a group of identically constructed carpets. Yet another unpublished report by Beyerlein [1] concluded that no discernable differences in performance exist between carpets representing nylon 66 and nylon 6 fibers. Unfortunately, this work also suffered from the same weaknesses as Werny’s, in that it did not fully match constructions between the fibers; instead, it compared the performance of solution dyed carpets with that of regular dyed carpets. In contrast, our study is a statistically designed, full factorial study to facilitate a scientifically valid and com0040-5175/$15.00

158

TEXTILE RESEARCH JOURNAL

mercially relevant comparison of the important performance properties of carpets made from nylon 6 and nylon 66 face yarns. The main goal of the study is to identify statistically significant carpet performance parameters attributable to the inherent properties of the polymer and to quantify the differences where possible. We believe that the work fills an important gap in understanding the performance differences exhibited by nylon 6 and nylon 66 carpets.

Materials and Methods TEST MATERIALS This study covered twenty-four different carpet samples, half of which represented nylon 66 yarns and the other half nylon 6 yarns. The yarns were perfectly matched for cross section, filament denier, and yarn denier. Also, to ensure that the results of this study could be applied to a wide range of residential and industrial carpets, we included pile type, pile weight, and pile height as the construction variables and fluoropolymer treatment as a finishing variable. We also kept the levels of these variables similar for the two fiber types, thus providing for a matched pair comparison of the performance properties of the two fiber types across a range of construction variables. Table I, shows that the test materials represent a full factorial design [11] with four different factors, each factor being represented by at least two different levels.

TABLE I. Description of variable factors. Factor

Levels

Polymer Pile type Pile weight Fluorochemical treatment

nylon 6, nylon 66 cut, loop low, medium, high treated, not treated

The complete set of test materials representing these variable parameters is shown in Table II. All necessary precautions were taken during the fabrication of the test materials to keep the physical parameters of the two sets of carpets as close to each other as possible. Pile weights chosen for cut and loop pile carpets were different (30, 36, and 42 ounces for cut pile and 24, 28, and 32 ounces for loop pile). These differences reflect industry practice. There was no statistically significant difference between the average pile weights of the nylon 6 and nylon 66 carpets, and this was true for the average pile heights as well, thus providing for a fully balanced design of nylon 6 and nylon 66 carpets.

Following are some additional construction particulars of the carpet samples that are not listed in Table II: Sample production: the carpets were specially made to the specifications of the study, and the two fibers used the same production equipment. Yarn type: eighteen of the twenty-four carpets used continuous filament yarn and the remaining six carpets used staple fiber yarn; 50% of the filament yarn carpets represented nylon 6 and the other 50% nylon 66; similarly, three of the six staple yarn carpets represented nylon 6 and the other three represented nylon 66. Particulars of filament yarn: the filament yarn was made of 20 dpf trilobal filaments, yarn denier was 1300 with 4.75 turns of twist. Particulars of staple fiber yarn: the staple yarn was made of 18 dpf trilobal fibers, yarn count was 3.4 Ne, singles twist was 5.3 tpi, and ply twist was 5.1 tpi, needle gauge: 1/10⬙, amount of stain blocker on nylon 6 and 66 carpets: zero (no stain blocker), heat setting equipment: the filament yarns used the Superba heat setting equipment and the staple yarns used the Sussen heat setting equipment; yarns were heat set following the commercial practice for nylon 6 and 66 yarns. Coloring method: wet dyeing.

TEST METHODS The following performance properties were evaluated for all test materials: Light fastness: Color fastness to light was measured according to the standard ATTCC test method (method 16), which uses a water-cooled xenon arc lamp that emits continuous light. Higher measured values of fastness indicate better performance. Static loading test: We measured 1 hour and 24 hour thickness recovery values by placing on the surface of the carpet weights close to the average body weight and then measuring the percent thickness recovered exactly 1 hour and 24 hours after the weight was removed. There is no standard ASTM test procedure available for this test, so we used a procedure similar to that of the carpet manufacturers of Dalton, GA. The results of this test are useful in predicting and understanding the indentations made by furniture and help to estimate the recovery time span required for the indented spot to return to the original state. The results of this test bear no relationship to surface texture measurements. Vettermann drum test: We used the standard ASTM test (D 5417-99) for characterizing the texture retention property of carpets, with three replicates for each carpet sample, and ran the test for 5000 cycles. Higher measured values suggest better texture retention. Ozone fastness: We measured colorfastness to atmospheric ozone at high humidities as per method 129 of the AATCC. Higher measured values indicate better fastness properties. NOx fastness: We

FEBRUARY 2005

159 TABLE II. Construction particulars of the experimental carpet samples.

S. no.

Sample ID

Fiber type

Pile type

Fluoropolymer

Pile weight, oz/sq.yd

Weight rank

Pile height, inches

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

15 16 19 32 34 35 20 21 24 26 30 36 14 17 18 28 29 33 22 23 25 27 31 37

6 6 6 6 6 6 6 6 6 6 6 6 66 66 66 66 66 66 66 66 66 66 66 66

cut cut cut cut cut cut loop loop loop loop loop loop cut cut cut cut cut cut loop loop loop loop loop loop

none none none treated treated treated treated none none none treated treated none none none treated treated treated none treated none none treated treated

30.3 36.1 41.7 30.3 36.1 41.7 28.5 28.5 31.7 23.8 31.7 23.8 29.8 35.7 42.1 36.1 29.8 42.1 32.2 32.2 24.9 27.6 27.6 24.9

low medium high low medium high medium medium high low high low low medium high medium low high high high low medium medium low

0.219 0.25 0.281 0.219 0.25 0.281 0.156 0.156 0.188 0.125 0.188 0.1 0.219 0.25 0.281 0.25 0.219 0.281 0.188 0.188 0.125 0.156 0.156 0.1

measured color fastness to nitrous oxides at high humidities as per method 164 of the standard AATCC. Higher measured values indicate better fastness. Resistance to oil stains: Resistance to oil stains or oil repellency followed AATCC method 118. Higher measured values indicate better repellency. Water repellency: Resistance to wetting or water repellency followed AATCC method 118. Higher measured values imply better repellency or greater resistance to wetting. Soil repellency: We measured soil repellency following a method developed by Solutia Inc., in which a water-based slurry is sprayed on cut samples in a precisely controlled manner using an apparatus designed and built by Solutia. The soil is the standard soil supplied by 3M as per the requirements of the AATCC soiling method 123. After wet soil application, the samples are allowed to dry and then trafficked in the Vettermann drum for 600 cycles. We then measured color on both soiled and unsoiled carpets with a Minolta chroma meter CR-210. The color difference between soiled and unsoiled carpets is reported as delta E, which is calculated as follows: Delta E ⫽ 关共DL*兲2 ⫹ 共Da*兲2 ⫹ 共Db*兲2 兴1/ 2

,

where DL* ⫽ L* soiled ⫺ L* unsoiled, Da* ⫽ a* soiled ⫺ a* unsoiled, and Db* ⫽ b* soiled ⫺ b* unsoiled. Lower measured values of delta E imply better soil resistance.

Results and Discussion PARAMETRIC ANALYSIS

OF

MEASURED PROPERTIES

Table III gives the measured performance properties of the experimental carpets for light fastness, ozone fastness, NOx fastness, oil repellency, water repellency, soil repellency, thickness recovery, and surface texture rating as measured by the Vettermann drum method. Table IV indicates which of the carpet construction variables exert a statistically significant influence (95 % confidence) on the measured performance properties and which variables have no influence on the measured properties. In addition to the statistical significance results given in Table IV, we used a series of box plots to illustrate the influence of different construction variables on performance properties. Box plots display the entire distribution of each measured property parameter side by side for different levels of the variable, as opposed to displaying only the mean or median values. The height of the rectangle of each box plot represents the inter-quartile range (mid 50% values) of the measured property parameter, while the vertical line above and below the rectangle represents the fourth and first quartiles. The horizontal line inside the rectangle represents the median value of the measured parameter. The absence of a horizontal line inside the rectangle suggests that the median coincides with the lines representing either the

160

TEXTILE RESEARCH JOURNAL TABLE III. Measured performance properties of the carpet samples.

S. No.

Fiber type

Light fastness

Ozone fastness

NOxfastness

Oil repellency

Water repellency

Soil repellency

1-Hour recovery, %

24-Hour recovery, %

V-drum reading

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

6 6 6 6 6 6 6 6 6 6 6 6 66 66 66 66 66 66 66 66 66 66 66 66

4 4 4 4 3.4 3.5 4 4 4 4 3.5 4 4 4 3.5 3 3 4 4 4 3.5 4 3.5 4

2 1.5 1.5 2.5 2.5 2.5 3 2 2.5 2 2.5 2.5 3 2 2 4 3 3 2.5 3 3 2.5 3.5 4

3 3 3 3 3.5 3 4 3.5 3 3.5 3.5 2.5 3.5 2 3 4.5 4 2.5 3 3 3.5 3.5 4 4

1 1 1 3 3 2 5 1 1 1 3 3 1 1 1 3 3 2 5 1 1 1 4 5

1 1 1 3 4 3 4 1 1 1 3 3 1 1 1 4 4 3 4 1 1 1 4 4

14.6 14.8 14.9 9.9 10.4 10.3 8 11.3 11.4 11 9.5 7.5 15.3 15.5 15.8 10.8 10.8 10.9 8.9 11.5 10.7 9.9 8.5 7.3

92.5 91.9 94 88.6 88.8 92.5 98.2 98.5 96.6 96.7 89.8 86.3 93.1 93.7 96 92.7 88.6 88.6 92.3 93.1 92 93 87.4 91.1

97.7 96.6 97.5 95.3 95.1 95.8 98.9 98.9 98.5 98.5 91.6 92.9 96 97.1 96 96.4 94.4 96 98.7 98.1 97.9 98.3 93.1 93.4

3.5 3.5 3.5 3.5 3.5 3.5 4 4 4 4 4 3.5 3.5 3.5 3.5 4 3.5 4 4.5 4.5 4 4.5 4.5 4

third or first quartiles. The printed numerical value inside the rectangular box represents the mean. The asterisks represent the outliers (values falling above and below ⫾ 3 ␴ limits). Influence of fiber type on surface texture properties as measured by Vettermann drum: From Figure 1, we see that nylon 66 has a better texture retention property compared to nylon 6 after 5000 cycles of loading in the Vettermann drum tester. Table IV suggests that the observed difference in the Vettermann drum reading between nylon 6 and nylon 66 fibers is statistically significant. Figure 2 suggests that pile type influences the measured surface texture and that loop pile construction, as expected, retains texture better than cut pile construction. From Figure 3, we see that pile weight also influences carpet texture and that the measured texture values of the low pile weight carpets are significantly lower than those of the medium and heavy pile weight carpets. However, the measured texture values of the medium and heavy pile weight carpets are identical, suggesting that there may be a critical face weight beyond which pile weight may not influence texture as measured by the Vettermann drum. A separate analysis of the texture data of cut and loop pile carpets shows that low pile weight adversely affects the measured texture of both cut and loop pile carpets, and the texture of the loop pile carpet is more sensitive to low pile weights. Influence of fiber type on ozone fastness: From Figure 4 and the corresponding significance value in Table IV,

it is clear that the ozone fastness property of nylon 66 carpets is superior to that of nylon 6 carpets. The behavior is similar to observations made in a previous work [8]. Figure 5 suggests that carpets treated with a fluoropolymer exhibit better ozone fastness compared to untreated carpets. Influence of fiber type on oil repellency: Table IV shows that the observed difference in the measured values of oil repellency for nylon 66 and nylon 6 carpets is not statistically significant. However, Table IV and Figure 6 suggest that oil repellency is positively influenced by fluoropolymer treatment, with the treated carpets giving a mean repellency rating of 3.08 compared to a rating of 1.33 for untreated carpets. Influence of fiber type on water repellency: Table IV suggests that water repellency is not influenced by fiber type. However, the table also shows that water repellency is significantly influenced by fluoropolymer treatment, and this observation agrees with expectations. Influence of fiber type on soil repellency: Soil repellency is not influenced by fiber type. However, the twoway ANOVA tests clearly reveal that soil repellency is influenced by both pile type and fluoropolymer treatment. With lower values for soil repellency, loop pile carpets show better soil repellency compared to cut pile carpets. Also carpets treated with fluorocarbon polymer show better soil repellency compared to untreated carpets.

FEBRUARY 2005

161

TABLE IV. Table of statistical significance for 2-way on measured properties. Description of 2-way ANOVA test V-drum reading versus fiber type and pile weight V-drum reading versus fiber type and fluoropolymer V-drum reading versus fiber type and pile type Ozone fastness versus fiber type and pile type Ozone fastness versus fiber type and fluoropolymer Soil repellency versus fiber type and pile type Soil repellency versus fiber type and fluoropolymer Oil repellency versus fiber type and pile type Oil repellency versus fiber type and fluoropolymer Water repellency versus fiber type and pile type Water repellency versus fiber type and fluoropolymer NOx fastness versus fiber type and pile type NOx fastness versus fiber type and fluoropolymer 1-Hour recovery versus fiber type and pile type 1-Hour recovery versus fiber type and fluoropolymer 24-Hour recovery versus fiber type and pile type 24-Hour recovery versus fiber type and fluoropolymer Light fastness versus fiber type and pile type Light fastness versus fiber type and fluoropolymer

ANOVA

test

P-value of influencing variables

Significance at 95% confidence

fiber type-0.043 pile weight-0.049 interaction-0.908 fiber type-0.063 fluoropolymer-0.781 interaction-0.408 fiber type-0.003 pile type-0.000 interaction-0.159 fiber type-0.006 pile type-0.223 interaction-0.859 fiber type-0.000 fluoropolymer-0.000 interaction-0.448 fiber type-0.828 pile type-0.001 interaction-0.567 fiber type-0.828 fluoropolymer-0.001 interaction-0.565 fiber type-0.684 pile type-0.229 interaction-0.684 fiber type-0.612 fluoropolymer-0.002 interaction-0.400 fiber type-0.679 pile type-0.890 interaction-0.890 fiber type-0.504 fluoropolymer-0.000 interaction-0.504 fiber type-0.496 pile type-0.310 interaction-1.00 fiber type-0.475 fluoropolymer-0.160 interaction-0.280 fiber type-0.438 pile type-0.397 interaction-0.197 fiber type-0.368 fluoropolymer-0.004 interaction-0.600 fiber type-0.864 pile type-0.660 interaction-0.836 fiber type-0.827 fluoropolymer-0.002 interaction-0.528 fiber type-0.252 pile type-0.207 interaction-0.582 fiber type-0.231 fluoropolymer-0.058 interaction-0.949

yes yes no no no no yes yes no yes no no yes yes no no yes no no yes no no no no no yes no no no no no yes no no no no no no no no no no no yes no no no no no yes no no no no no no no

Influence of fiber type on color fastness to nitrous oxides: Table IV suggests that the differences between nylon 6 and 66 carpets are not statistically significant.

FIGURE 1. Influence fiber type on measured texture retention property.

FIGURE 2. Influence of pile type on measured texture retention property.

FIGURE 3. Influence of pile weight on measured texture retention property.

However, in dealing with colorfastness, it is important to remember that what applies to one dye need not be true for other dyes. The results are therefore applicable only for the acid dye system used in this study (Telon Blue BRL 200 and Telon Red 2BN 200).

162

TEXTILE RESEARCH JOURNAL on thickness recovery, but it is significantly lower for fluoropolymer treated carpets, and this is true for both 1-hour and 24-hour thickness recoveries. Therefore, fluoropolymer treatment has a significant negative influence on thickness recovery from an applied static load. As stated earlier, the scope of this study is limited to identifying and quantifying the performance differences, and we did not attempt to explore the underlying reasons for the observed differences. New studies focusing on the factors contributing to the reduced thickness recovery of fluoropolymer treated carpets may be needed. FIGURE 4. Influence of fiber type on ozone fastness.

PREDICTIVE MODELS FOR SURFACE TEXTURE PROPERTY AND OZONE FASTNESS PROPERTY Since we noticed major differences in performance properties between nylon 6 and nylon 66 carpets in terms of texture retention and ozone fastness properties, we attempted to develop predictive models for these two properties. The two models and their respective parameters are as follows: V-Drum Reading ⫽ 3.73 ⫺ 0.29 ⫻ Fiber Type ⫹ 0.54 ⫻ Pile Type ⫹ 0.13 ⫻ Pile Weight

FIGURE 5. Influence of fluoropolymer treatment on ozone fastness.

.

(1)

This model produced an R2 value of 0.78, with the variable Fiber Type being significant at a p-value of 0.0012, Pile Type at 0.0001, and Pile Weight at 0.0158. For the purpose of developing the models, we considered all three independent variables as categorical variables and coded them as follows: Fiber Type ⫽ 0 (nylon 66), Fiber Type ⫽ 1 (nylon 6), Pile Type ⫽ 0 (cut pile), Pile Type ⫽ 1 (loop pile), Pile Weight ⫽ ⫺1 (low weight), Pile Weight ⫽ 0 (mid weight), and Pile Weight ⫽ ⫹1 (high weight). Model 1 predicts a better surface texture (higher Vettermann drum reading) for nylon 66 fiber type, for loop pile type, and for higher pile weights. Ozone Fastness ⫽ 2.42 ⫺ 0.71 ⫻ Fiber Type ⫹ 0.79 ⫻ Fluorocarbon ⫹ 0.29 ⫻ Pile Type ⫺ 0.16 ⫻ Pile Weight

FIGURE 6. Influence of fluoropolymer treatment on oil repellency.

.

(2)

2

Influence of fiber type on light fastness: Table IV shows that light fastness is not influenced by fiber type. Table IV also suggests that pile type has no influence on the measured value of light fastness. Here again, it is important to remember that the observed light fastness behavior applies only to the dyes used in this study. Influence of fiber type on thickness recovery from applied static load: Table IV shows that neither 1-hour recovery nor 24-hour recovery is influenced by fiber type. The table also shows that pile type has no influence

This model produced an R value of 0.77, and the variable Fiber Type was significant at a p-value of 0.0001, while Fluorocarbon was significant at 0.0001, Pile Type at 0.0549, and Pile Weight at 0.0895. The variable Fluorocarbon was coded as follows: No Treatment ⫽ 0

,

Treated ⫽ 1

.

This model, similar to the texture property model, accounts for the advantages offered by nylon 66 carpets.

FEBRUARY 2005 NONPARAMETRIC EVALUATION

163 OF

TEXTURE

One of the major limitations to texture evaluation by the CRI texture rating scale is that it represents an ordinal system, where carpet texture changes are compared to a standard “best match” photograph. In addition, since this is an ordinal scale, a rating difference of x at the lower end of the scale cannot be assumed to have the same practical significance as the rating difference of x at the higher end of the scale. Thus, even though the numerical difference between ratings 4 and 5 and ratings 2 and 3 is the same, the practical significance of the two rating differences could be very different. For this reason, quantitative treatment (parametric evaluation) of ordinal data can be somewhat misleading because they do not fully reflect the actual quality differences. In addition, there are situations where nonparametric evaluations may serve the purpose better than parametric evaluations [3– 4, 14 –15]. The superiority of human judges over measuring instruments for picking up fabric streakiness is described by Davis [4], “The most amazing fact about streakiness is the small absolute lightness variation that people perceive as objectionable in fabrics. People object to streaks that are hardly detectable with instruments.” Human judges often manage to pick up even minute differences in texture between samples, thus making the CRI texture rating less effective at resolving differences between samples. Also, in the real world, consumers make purchase decisions based on their own senses, and so capturing what a human perceives is critical to commercial development. Therefore, in this work, we chose to use the more discriminating characterization of paired comparison [3] of each nylon 66 carpet against its nylon 6 counterpart. We used the paired comparison approach to answer the question: Are there significant differences in wear between matched nylon 6 and nylon 66 carpets? We believe the method of paired comparisons is a powerful statistical technique [3] to characterize head-tohead contrasts, and it is superior to the CRI texture rating scale in terms of relevance and usefulness. Three replicate samples of each of the twenty-four carpets that were subjected to 5000 wear cycles on the Vettermann drum tester were used in the paired comparison tests. Samples representing the same construction parameters were chosen as a pair for the paired comparison. Each nylon 6 replicate was labeled 1, 2, or 3, and each nylon 66 replicate A, B, C. All nine possible combinations of the nylon 6 versus 66 contrasts (1 versus A, 2 versus A, . . . 3 versus C, etc.) were presented to the grader (a person skilled in the art of judging wear) to judge the extent of wear (or the texture grade), following a blind procedure. Within a particular pair being tested, the nine comparisons were selected randomly, and after

the nine comparisons were made, another nine items were judged for the next pair. All pairs of cut pile were judged before proceeding to the loop pile pairs. The comparison schemes for the cut and loop pile carpets and the grader’s preferences for each comparison pair are shown in Tables V and VI:

TABLE V. Paired comparison scheme for cut pile carpets. Cut pile comparisons N6 v N66, item #s 32 v 29 34 v 28 35 v 33 15 v 14 16 v 17 19 v 18

Pair: preference 1vA: 2vA: 3vA: 1vA: 2vA: 3vA: 1vA: 2vA: 3vA: 1vA: 2vA: 3vA: 1vA: 2vA: 3vA: 1vA: 2vA: 3vA:

same same A A A A A A A A A A A A A A A A

Nylon (1–3) & nylon 66 (A–C)

Pair: preference

Pair: preference

1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B

1vC: same 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C

6/9 for N66 9/9 for N66 9/9 for N66 9/9 for N66 9/9 for N66 9/9 for N66

TABLE VI. Paired comparison scheme for the loop pile carpets. Loop pile comparisons

Nylon 6 (1–3) & nylon 66 (A-C)

N6 v N66, item #s

Pair: preference

Pair: preference

Pair: preference

36 v 37

1vA: A 2vA: A 3vA: A 1vA: A 2vA: A 3vA: A 1vA: A 2vA: A 3vA: A 1vA: same 2vA: same 3vA: A 1vA: A 2vA: A 3vA: A 1vA: A 2vA: A 3vA: A

1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B 1vB: B 2vB: same 3vB: same 1vB: B 2vB: B 3vB: B 1vB: B 2vB: B 3vB: B

1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C 1vC: C 2vC: C 3vC: C

20 v 31 30 v 23 26 v 25 24 v 22 21 v 27

9/9 for N66 9/9 for N66 9/9 for N66 5/9 for N66 9/9 for N66 9/9 for N66

Table V shows that in the blind comparison involving matched pairs, the expert grader preferred nylon 66 car-

164

TEXTILE RESEARCH JOURNAL

pets 50 out of 54 times in the case of cut pile carpets and showed equal preference for both nylon 6 and 66 carpets 4 out of 54 times. Table VI shows that the expert grader preferred nylon 66 carpets 50 out of 54 times in the case of loop pile carpets and showed equal preference for both nylon 6 and 66 carpets 4 out of 54 times. Thus, the grader overwhelmingly preferred nylon 66 carpets compared to nylon 6 carpets, and the preferences were identical for the cut pile and loop pile carpets.

Conclusions A full factorial evaluation of the performance properties of nylon 6 and nylon 66 carpets involving the most common construction variables has revealed that nylon 66 carpets provide better texture retention and better colorfastness to ozone compared to nylon 6 carpets. In a blind comparison test involving matched pairs, an expert grader overwhelmingly prefers nylon 66 carpets (50 times out of 54) over nylon 6 carpets. However, we must remember that the texture comparisons are not between optimally designed nylon 6 and 66 carpets, and preferences can differ substantially when two sets of optimally designed carpets are compared. It is also important to remember that nylon 66 fibers generally cost more than nylon 6 fibers, and the possibility exists for nylon 6 carpets to match the texture preference of nylon 66 carpets on the basis of fiber cost parity. There are no significant differences in performance properties between nylon 6 and nylon 66 carpets in terms of colorfastness to nitrous oxides, thickness recovery from prolonged static load application, oil repellency, water repellency, and soil repellency. As expected, fluoropolymer treatment of nylon significantly improves water repellency, oil repellency, and soil repellency properties. However, fluoropolymer treatment also accounts for reduced thickness recovery from prolonged static load application. This work does not focus on understanding the reasons for the reduced thickness recovery shown by the fluoroploymer treated carpets, and further work may be necessary to understand this behavior. ACKNOWLEDGMENTS I wish to gratefully acknowledge the help of Invista娂 nylon commercial flooring in providing test materials and partial financial support for this work.

Literature Cited 1. Beyerlein, A., Nylon Fiber Facts, Clemson University, Clemson, SC, unpublished report. 2. Danford, M. D., Spruiell, J. E., and White, J. L., Structure Development in the Melt Spinning of Nylon 66 Fibers and Comparison to Nylon 6, J. Appl. Polym. Sci. 22, 3351 (1978). 3. David, H. A., “The Method of Paired Comparisons,” 2nd ed., Oxford University Press, NY, 1988. 4. Davis, H., McGregor, R., Pastore, C., and Timble, N., Human Perception and Fabric Streakiness, Textile Res. J. 66, 533–544 (1996). 5. Dumbleton, J. H., and Buchanan, D. R., A Comment on the Crystal Moduli of Nylon 6 and Nylon 66, Polymer 9, 61 (1968). 6. Han, L., Wakida, T., and Takagishi, T., Changes in Fine Structure and Dyeing Behavior of Nylon 6, Nylon 66, and PET Fibers Treated with Superheated Steam, Textile Res. J. 57, 519 –522 (1987). 7. Horsfall, G. A., Factors Influencing the Daylight Photodegradation of Nylon 66, Nylon 6, and Polyester in Commercial Fabrics, Textile Res. J. 52, 197–205 (1982). 8. Jellinik, H. H. G., and Choudhury, A. K., Inhibited Degradation of Nylon 66 in the Presence of Nitrogen Dioxide, Ozone, Air, and Near Ultraviolet Radiation, J. Polym. Sci. Part A-1 Polym. Chem. 10, 1773–1788 (1972). 9. Leung, W. P., Ho, K. H., and Choy, C. L., Mechanical Relaxations and Moduli of Oriented Nylon 66 and Nylon 6, J. Polym. Sci. 22, 1173 (1984). 10. Makansi, M., Perception and Control of Fabric Streaks, Textile Res. J. 57, 495–502 (1987). 11. Montgomery, D. C., “Design and Analysis of Experiments,” 3rd ed., John Wiley and Sons, NY, 1991. 12. Prevorsek, D. C., and Chin, H. B., Intrinsic Differences Between Nylon 6 and Nylon 66 Industrial Fibers: Micromechanical and Molecular Analysis, Int. J. Polym. Mater. 25, 161 (1994). 13. Schmitz, F. P., Mroszewski, K. D., and Rossbach, V., On The Thermal Mobility of the Amino End Groups of Polyamides 6 and 66, Makromol. Chem. 184, 184 (1983). 14. Slater, K., Subjective Textile Testing, J. Textile Inst. 88, 79 –91 (1997). 15. Wang, J., and Wood, E. J., A New Method for Measuring Carpet Texture Change, Textile Res. J. 64, 215–224 (1994). 16. Werny, F., The Floor Performance of Synthetic Polymers Currently Used in Carpet Fibers, Carpet Rug Ind. 6, 38 (1995). Manuscript received January 16, 2004; accepted March 12, 2004.