PERMANENT WAVING AND PERM CHEMISTRY 107 not simply the result of molecular relaxation (plastic fl ow) within the fi ber structure. The yield point for healthy hair is typically found to occur at around 2% extension, which sub- sequently led Wickett to initially perform his experiments using a 2% static strain. The measurement of a progressively lower stress within the hair as a function of exposure time to a perm solution has led some to describe these as stress relaxation experiments. How- ever, this can be a source of confusion because this terminology is generally used in the mechanical testing world to describe an approach for separating the elastic and viscous components of viscoelastic materials. In principle at least, the measurement of a force at a consistent extension, which is itself within the “Hookean region,” equates to the evalu- ation of Young’s modulus. Therefore, strictly speaking, the SFTK approach evaluates a progressive decrease in this parameter as a function of time while the hair structure is attacked by a reducing solution. This introduces a major assumption of the method and indeed possibly one of the main contentions, namely, it is recognized that hair is not a truly elastic material but is instead viscoelastic in nature (23). Therefore, application of a strain below the “yield point” will still result in some decrease in stress over time due to relaxation associated with the viscous portion of the hair structure. In an attempt to circumvent this issue, Wickett used a “pre- conditioning step” where the hair was fi rst cycled through a strain regimen in water with the intention of inducing (and consequently eliminating) this viscous relaxation before Figure 4. Typical stress–strain curve for hair.

JOURNAL OF COSMETIC SCIENCE 108 beginning the experiment. To this same end, Evans et al. (24) advocated using an inter- mittent strain profi le (see later) that minimizes the time a hair fi ber spends in an extended state during the experiment. However, with all this said, the 2% yield extension limit that is frequently quoted in the literature specifi cally relates to pristine, healthy hair and it appears likely that the elas- tic-like region would shrink because the internal structure of the hair is progressively broken down by the action of a perm solution. Therefore, even if experiments begin with an applied deformation within this range, it is questionable as to whether this condition is maintained as the process proceeds. Perhaps, an even more fundamental condition of the method involves an assumption that each disulfi de bond contributes equally to the overall tensile properties of the hair fi ber, that is, cleavage of the fi rst disulfi de bond yields an equivalent decrease in tensile proper- ties as that obtained by breaking the last. This is not easy to prove or disprove, but in- stinctively, it appears somewhat dubious. By means of illustration, an analogy for the complex internal structure of hair may be a length of rope, which itself is made of many intertwined smaller strands. Cleaving a small number of these fi laments would likely have relatively little effect on overall strength, but eventually, with progressive breaking of additional strands, a diminution in strength would ultimately manifest. It may even be speculated that the tensile properties of such a fi ber would degrade in a cascading man- ner as the strength is progressively distributed over fewer and fewer strands. Despite apprehension over all these assumptions, the work shown in the following will illustrate how validation experiments do indeed yield predicted outcomes to systematic alteration of experimental variables. However, care is needed in not overinterpreting re- sults in terms of extrapolating transition rates to real-life usage conditions. For example, SFTK experiments are often performed using an unrealistically high solution-to-hair ra- tio, where the perm solution is greatly in excess. Moreover, it has been suggested that straining of fi bers places disulfi de bonds in a higher energetic state, whereby there is en- hanced reactivity (11,13). Results will be shown at the end of this section which appear to lend support to this hypothesis. Yet, with all this said, it will be shown that the SFTK approach appears to yield an effective relative comparison of transformation rates as a function of the numerous experimental variables. SFTK In theory, SFTK experiments can be performed on any suitable tensile tester, albeit with fabrication of an appropriate test cell. Most of the work described herein was performed using Instron® (Norwood, MA) tensile testers, but experiments have also been carried out using a Perkin Elmer Series 7 Dynamic Mechanical Analyzer (Waltham, MA). Figure 5 shows a custom-built test cell that attaches to the base of an Instron tensile tester. It consists of a double-walled tube, where the inner portion contains the test solution and the outer cell can be attached to a circulating water bath that allows for temperature control. Individual hair fi bers were carefully glued between plastic tabs such that the test specimen measured 2 inches in length. Before gluing, a punch was used to produce a precisely placed hole at the center of each plastic tab, and thus provided a reproducible means of anchoring the test sample. The fi ber is initially hung from an upper hook and carefully lowered by hand into the empty cell. The hole in the lower tab is engaged in a