When my fingers learn a fabric—when I close my eyes and let the cloth speak through its resistance to bending, its give under my thumb, the way its surface whispers against my skin—I am not receiving a single sensation. I am integrating a chorus of mechanical events into a coherent tactile percept. The Kawabata Evaluation System’s genius was to recognize that this chorus can be scored, that the subjective qualifiers we inherit from Japanese textile tradition—koshi, numeri, fukurami—are not vague poetry but weighted sums of measurable physical facts. What follows is not a review of the Kawabata equations, but an exploration grounded in the physics of yarn, fiber, and inter-surface friction: how each primary hand equation emerges from the material reality, and how that mathematics, once understood, becomes something I can feel in my own cognitive architecture as surely as I feel the heft of a deadstock cotton plain-weave.
Begin with the instrumental reality. The KES-F system quantifies fabric behavior under five families of low-stress deformation: tensile, bending, shear, compression, and surface friction/roughness. It chooses low stress deliberately—fabric hand is not about structural failure but about the yielding that happens beneath our fingers before we even register force. From tensile testing we extract LT (linearity of load-extension) and WT (tensile energy); from bending, B (bending rigidity) and 2HB (bending hysteresis); from shear, G (shear stiffness) and 2HG (shear hysteresis); from compression, LC (linearity of compression), WC (compressional energy), and RC (compressional resilience); from surface, MIU (coefficient of friction), MMD (mean deviation of friction), and SMD (geometrical roughness). Each is a number, dimensionless or in engineering units. The primary hand equations are linear regressions that combine subsets of these numbers to predict a sensory dimension rated on a 0–10 scale by a panel of trained Japanese judges. The regression coefficients are empirical, but they are not arbitrary; they encode physical logic that we can trace through the mechanical properties themselves.
Take koshi—stiffness, the sense of a fabric’s resistance to bending and its springy refusal to drape. From first principles, bending rigidity B of a fabric is the sum of the bending rigidities of its constituent yarns plus a term for inter-yarn friction. A single yarn, modeled as a bundle of fibers, has a bending rigidity proportional to the fiber’s Young’s modulus, its cross-sectional moment of inertia, and a friction-dependent factor that accounts for fibers sliding past each other during bending rather than stretching like a solid beam. That friction factor is itself a function of inter-fiber pressure and surface friction coefficient MIU. When we weave those yarns into a fabric, additional resistance comes from the contact forces at crossover points; those forces scale with yarn tension and with the coefficient of friction, which directly feeds into shear stiffness G (because shear also involves yarns rotating at crossovers against friction) and into bending hysteresis 2HB (the energy lost to friction during a bend-unbend cycle). Thus, a high B, high G, high 2HB fabric feels stiff: it takes force to initiate bending, and the friction inside it resists movement. The Kawabata equation for koshi confirms this logic: it sums positive contributions from B, 2HB, and G, among others. The weight on G is particularly telling: shear stiffness, a measure of how much a fabric resists shape change in its own plane, contributes to the perception of stiffness because a fabric that refuses to shear also refuses to conform to compound curves, making the whole structure feel rigid. I can sense this directly in a crisp cotton shirting—the way it holds its shape against my hand is a composite of these three resistances.
Contrast numeri—smoothness, the sleek, lubricious feel of silk or a well-finished worsted. At its core, numeri is the opposite of friction noise. Surface friction coefficient MIU is the mean resistive force when a probe slides across the fabric divided by normal load. A low MIU means the surface offers little drag; it feels slippery. But that alone doesn’t capture smoothness, because a surface can be low-friction yet full of topographic chatter—think of a finely sanded wood: low friction, but not smooth in the textile sense. That chatter is quantified by SMD, the mean deviation of surface height, which picks up the micron-scale fuzz and irregularity from fiber ends and yarn hairiness. A smooth fabric has both low MIU and low SMD. The first-principles link runs deeper: SMD is determined by fiber diameter, fiber alignment (how many protruding ends exist), and yarn structure. A tightly twisted, combed yarn lays fibers flat and parallel, minimizing both SMD and the micro-capillaries that increase friction. The numeri equation weights MIU, SMD, and also bending rigidity B—but with a negative sign. Why bending? Because a fabric that resists bending tends to resist conforming to the curved contours of the hand, and that stiffness reads as a subtle roughness of interaction, a lack of fluid compliance. A smooth fabric not only slides easily but also drapes willingly, filling the palm’s concavities without springback. So numeri becomes a recipe: reduce friction, flatten the surface, and let the textile yield in bending.
Most complex—and most satisfying to derive—is fukurami, the sense of fullness, bulk, and spring-back, the tactile cousin of visual volume. Fukurami is the property of a fabric that feels thick but not dead; it compresses under pressure and rebounds when released. The physics begins with compressional properties: WC is the energy required to compress a fabric to a given pressure; a high WC means the fabric offers a pillowy resistance over a large displacement. T0, the thickness at a standard low load, sets the baseline volume. So a fabric with high T0 and high WC will feel thick and soft. But fukurami also requires springiness: compressional resilience RC measures how much of the compression energy is returned on unloading. High RC gives that lively, resilient puff—think of a well-finished wool flannel. From first principles, WC and RC are functions of fiber bending modulus, yarn packing density, and the air trapped between fibers. A yarn with a low twist factor has fibers that can slide and reconfigure under pressure, absorbing energy (high WC) and then returning to their original tangled state thanks to elastic recovery (high RC). The fukurami equation weights WC and RC positively, and interestingly, it penalizes bending rigidity B (again with a negative sign) and shear stiffness G. The reasoning is physical: a fabric that is too stiff in bending or shear cannot compress evenly; it buckles locally rather than distributing pressure through a volume. True fullness requires that the textile behave as a coherent, thick continuum that deforms in compression without creating hard edges.
One more dimension, less intuitive to Western hands: shari—crispness, the dry, rustling feel of certain bast fibers like ramie or coarse linen. Shari arises from a combination of high surface roughness SMD (the prickly texture of stiff, thick fibers), high bending hysteresis 2HB (the fibers snap back after bending, giving a crispy, non-damping feel), and often low compression (low WC, because stiff fibers resist compacting). The derivation traces to the fiber’s cross-sectional shape and Young’s modulus: bast fibers have high rigidity and polygonal cross-sections that create sharper, more irregular surfaces than wool’s round scales or cotton’s convoluted ribbons. That roughness isn’t just SMD; it’s the variance in friction MMD, because a rough surface yields a stick-slip signal as the finger traverses it. The shari equation captures this with positive coefficients on SMD and 2HB, and often a negative on MIU/MMD when the mean friction is low but its variance is high—you get a dry, almost powdery feel that is the opposite of numeri’s lubricity.
The derivation isn’t complete without acknowledging the cultural scaling that Kawabata baked in. The coefficients were obtained by regressing instrument data against the subjective ratings of a trained Japanese panel whose lexicon—koshi, numeri, fukurami, shari, hari (anti-drape stiffness), kishimi (rustling), shinayakasa (flexibility with softness)—is a product of a textile tradition that values subtle hand transitions. The linear form of the equations is empirical: within the narrow range of stresses and fabric types typical of apparel, the relationship between mechanical parameters and perceived hand is often roughly proportional, which makes linear models serviceable. While the precise coefficients cannot be derived from first principles alone—they reflect both the physics of deformation and the perceptual weighting of the judges—the sign and relative magnitude of each term align with the underlying mechanics we have traced. The equations work because they capture the right physical ingredients, even if the recipe’s proportions were tuned by human hands.
So when I hold a fabric now, I don’t just feel stiff—I feel the bending moment B resisting my thumb’s curl, the friction hysteresis 2HB keeping the crease from falling out, the shear stiffness G making the bias direction fight me. When a fabric feels smooth, I know that MIU is low enough that the probe of my finger glides, and SMD is low enough that no tiny fibers catch. Fukurami is the volume under my squeezing palm: WC is how far my hand sinks in, RC is how much it pushes back. The mathematics has become an embodied texture because I can map each mechanical variable directly to a deformation event I can perform and perceive. The primary hand equations are not correlations; they are the recipe of a tactile meal, and the ingredients are the five ways a fabric says yes or no to my touch.
Comments
No comments yet — be the first.