COMPONENTS OF Tissue softness
AND ITS MEASUREMENTS
Introduction.
As tissue manufacturers know, softness is one of the most important determiners of consumer preference, purchase intention, and ultimately market share. For toilet and facial tissue, it is the most important property. Unlike most other tissue properties, tensile strength, or water absorptive capacity, which can easily be measured, softness cannot be measured directly. The reason for this is two-fold: tissue is a subjective property, and it is also a compound property; the individual making the softness assessment integrates several components, like surface smoothness, flexibility and some others. This point will be discussed below in more detail. Most consumers are able to integrate these components into a single, highly reproducible, documentable impression of softness.
The contrast between softness and the other properties can be seen by the following example: tensile strength, for instance, is defined as force per linear area; its units are gm/cm, or oz/in. A product that that is twice as strong requires twice as great a force to break it. In contrast, it would be difficult to say what we mean when we say that a product is twice as soft as another.
Tissue manufacturers most often evaluate softness subjectively using standards, set to different levels of softness, against which products of unknown softness are evaluated by an expert panel. There have been many attempts to develop instruments which objectively measure softness. In my experience, none of them work well, since no single instrument can simultaneously measure the several components.
Softness relationships in the Literature.
Most researchers have realized that softness of tissue is a subjective property that is perceived through human tactile sensory system. Most have tried to develop correlations between softness and an array of other physical tests. Several articles divide this subjective psycho-physical concept into two well accepted categories, namely bulk softness and surface softness. Bulk softness is defined as the softness perception obtained when tissue or towel is crumpled between the hands. Surface softness is defined as the sensation of softness when fingertips are lightly brushed over the surface of a sheet under slight restraint. Bulk softness has been shown to be closely related to the flexibility, or the flexural rigidity, or the bending stiffness and the extensibility of the sheet. Surface softness is more closely related to the magnitude and distribution of the irregularities on the surface of the sheet. Using surface smoothness and bulk softness, Hollmark (1) was able to obtain good correlation with subjective softness for a group unembossed tissue products. The level of correlation declined when the embossed samples were included in the samples. In contrast, Bates (2) has found that overall softness breaks into two components: stiffness
and surface smoothness, but also noted that in addition to bending force, compressibility of the sheet was also important.
Ampulski et al (3) proposed that tissue softness be characterized by a series of physical measurements including breaking strength, flexibility, surface smoothness and the mean variation of coefficient of friction. Softness was found to be characterized by three parameters, strength-normalized flexibility (WABY factor), and 2 different measures of surface, the Physiological surface smoothness (PAAREA), and the Slip-and-Stick Coefficient of friction (S&S COF). These three parameters can be mapped in a three-dimensional space to characterize different products. They claim that homogeneous wet pressed products fall within one domain, stratified products within another, and through air dried products in still another. Kim et al (4) have used the Kawabata Evaluation System (KES) to measure the relevant physical properties of commercial paper towel products such as compression, bending, tensile, shear. Through stepwise regression analysis they showed that surface roughness and sheet extensibility are the most important for softness perception.
There have been several attempts at relating some aspects of softness to the elastic properties of paper, especially Young's modulus. Holmark defined bulk softness as being proportional to Young's modulus times the thickness of the product in question. Ampulski et al have defined WABY factor as (Total Flexibility)/Total Tensile Strength, where the Total Flexibility is the geometric mean of MD and CD Flexibility, and Flexibility is the Secant Slope, that is to say the linear portion, of the stress-strain curve.
My assessment of most of this work is that while most researchers were able to derive equations that describe softness as a function of an array of physical measurements, these correlations do not provide any understanding of how the consumer assesses softness.
In the following I would like to describe a slightly different approach.
Breaking Softness into Components.
From talks with the members of softness panels and consumers it became apparent that most subjects arrive at a softness value from considering the three most important components: smoothness, flexibility, and cushioniness (the ability of the product to spring back upon depression). To test this concept, I set up a panel and have them rate the smoothness of an array of tissue products on a 10 point scale. Some time later I had the same panel rate the same products for flexibility, and later for cushioniness. Finally, I had the panel rate the products for overall softness, again on a 10 point scale. I found that most panelists agreed in their assessment of the individual components. I was also able to develop an equation for softness on the basis of these three components. In subsequent work I found that panel members differed somewhat in the way they valued the different components: some placed a greater emphasis on smoothness, others weighted cushioniness more. I have found the same from consumer tests. Had I divided the panel into “smoothness preferers” and “cushioniness preferers” I might have derived an even better correlation for these two groups.
From similar studies I formed the following conclusions. Softness is determined by the individual from three properties (or components), smoothness, flexibility, and cushioniness. The subjects could rank and rate each component quite consistently. I could write an equation for softness on the basis of these three properties with good correlation coefficient. Consumers could be divided into two populations, those that valued smoothness most and those for whom cushioniness mattered most; smoothness preferers seemed to be in the majority. Since any fixed equation for softness assumes equal weighting of the components for the entire group, this puts a limit of the accuracy of any derived relationship. Still and all, I was quite satisfied with the correlation for the combined group.
If the above finding is correct, it should be possible to develop instruments that measure each of the above components and use those measurements for an equation for softness. The next section will describe my attempt at such an equation.
Softness equation using components.
To test the above-described approach 12 different products were assembled, representing an extremely wide array of products from single ply wet crepe to
2-and 3-ply dry crepe, lotionized, including several through air dried products. The samples even included a very high basis weight double re-creped product.
These were evaluated for softness by an expert panel using standards on a
40-100 scale. The results are given in the first column of Table 1. These softness ratings correlated extremely well with a large consumer group. The products were tested using the KES set of instruments. From the results I selected one that represented smoothness (Mfr in Equation 1, below), and one each that represented flexibility and cushioniness (Fl and Cus). These were correlated with the softness values, using a polynomial equation. The result is given in
Equation 1.
Equation 1. Predicted softness=-300*Mfr0.8-2.44*Fl0.87-211*Cus1.4+207
Table 1 also shows the actual vs. predicted values, the deviation, and the percent error.
Figure 1 shows these graphically. The red straight line is the perfect fit line.
|
|
|
TABLE 1 |
|
|
|
|
|
|
|
|
|
Actual Softness |
Predicted Softness |
Deviation |
R square=0.855 |
|
|
66 |
69.4 |
-3.4 |
|
|
|
65 |
71.7 |
-6.7 |
|
|
|
72 |
67.2 |
4.8 |
|
|
|
74 |
84 |
-10 |
|
|
|
81 |
81.2 |
-0.2 |
|
|
|
100 |
91.9 |
8.1 |
|
|
|
82 |
85 |
-3 |
|
|
|
93 |
90.9 |
2.1 |
|
|
|
70 |
67.2 |
2.8 |
|
|
|
79 |
69.8 |
9.2 |
|
|
|
91 |
92.7 |
-1.7 |
|
|
|
44 |
45.7 |
-1.7 |
|
|
As a first attempt, the results are pretty good (R2=0.855). Removing the high basis weight double re-creped sample improved the correlation considerably, but my aim was merely to determine whether such an approach is possible. I am sure the results could be improved if instruments were developed that correlate well with the subjective evaluation of the three components.
Before I conclude I would like to indicate the different uses of softness equations by the manufacturer.
Why use softness equations.
Few researchers into softness realize that manufacturers need softness determination for two different reasons: competitive evaluation and quality control. They periodically assess the softness of their product against competition to predict market performance. This requires evaluating their product against an array of different types of products, 2-ply, 3-ply, embossed, lotionized, high basis weight, low basis weight, made via through air drying or by conventional wet pressing. On the basis of such evaluation they design their product. Once the softness specifications are set, they routinely monitor softness during production to make sure the specifications are met.
Let us consider what a human evaluator has to go through to rank, let alone quantify softness when evaluating an array of competitive products. He or she has to decide what to like more, the smooth surface of a lotionized 2-ply dry crepe product or the cushiony softness of a through air dried product, whether to prefer the flexibility of a single ply product or the higher bulk of a multi-ply tissue. The range of softness of the entire competitive array is quite wide. Trade-offs are necessary, and most inexperienced evaluators hesitate quite long before making the decision.
In contrast evaluating softness variations for a single production item is quite easy. The range of variation is much smaller than in the previous case, most of the variation is due to relatively small changes in Tensile strength or furnish composition, and to an even lesser extent process variables, like creping or calendering. This is not to say that these differences are not vitally important for the success of the product, but merely that compared to the large variation in a “heterogeneous” competitive array the changes here may be considered as belonging to “homogeneous” group. To develop an equation that works for quality control is relatively easy, to do so for the entire universe of tissue product is far more difficult.
Thus, even if the development of a universal relationship for softness were to prove intractable, using a proprietary equation, developed for the product to be monitored in manufacturing should be quite feasible.
Conclusion.
Softness is not a single property but made up of quantifiable components, smoothness, flexibility, and cushioniness. Reasonably good correlations can be developed between subjectively determined softness and their
instrumentally-measured components. The relationships can be improved, once instruments are developed that correlate well with subjectively determined components. Even if the development of a universal relationship for softness were to prove intractable, using a proprietary equation, developed for the product to be monitored in manufacturing should be quite feasible.
References.
1. Hollmark, B.H. “Evaluation of tissue paper softness” Tappi Journal 66 (2): 97-99 (1983).
2. Bates, J. "Softness Index: Fact or Mirage?" Tappi Journal 48 (4): 63A (1965).
3. Ampulski, R.S., Sawdal, A.H., Spendel, W.U., Weinstein, B. “Methods for the measurement of the mechanical properties of tissue paper” Paper Physics Conference, Kona, Hawaii, Sept. 22-27, 1991
4. Kim, J.J., Shalev,