Welcome to chapter five, tissues, skin, and contact with an object. This chapter can be summed up in one sentence: a good fit in CAD does not establish a suitable interface with a person. Everything else is the mechanism behind that sentence: what skin actually is, how it fails, why "it feels comfortable to me" is close to worthless as evidence, and what to ask before you let an object rest against somebody.
The failure this chapter prevents is the most common serious harm an amateur-designed contact object can cause, and it is almost always caused by someone competent. A well-machined splint with a crisp edge. A moulded handle with its parting line running across the point of highest load. A neatly sewn strap whose seam allowance sits under a bony prominence. A strap sized to a measurement taken in the morning on a limb that swells by afternoon. None of those is bad workmanship. Each is good workmanship in the wrong frame.
Every tissue belongs to one of four families. Epithelium covers and lines: the epidermis, the gut and airway linings, the inner lining of blood vessels. Connective tissue supports and binds, and it is defined by the abundant extracellular matrix its cells secrete: bone, cartilage, tendon, ligament, fat, the dermis, even blood. Muscle generates force, in three types: skeletal, cardiac and smooth. And nervous tissue communicates. Learn the families first; the subtypes follow.
Epithelia are named by cell shape and number of layers, and the combination tells you the job: a single flat layer in the lungs keeps diffusion distances short, many layers in skin resist abrasion. Two facts have direct design consequences. First, epithelium has no blood vessels of its own; it is fed by diffusion from the connective tissue beneath, so damage to that layer is what causes real trouble. Second, a continuous epithelial surface is the body's primary barrier. Breach it anywhere, by any means, including a three millimetre pressure-induced break you never intended, and the project changes from an object near a person to an object at a wound. That is a categorically different review.
Connective tissue is where the body's structural materials live, and its properties come mostly from the matrix the cells secrete, not the cells. Collagen fibres give high tensile strength with little stretch, like aramid fibre in a composite, except living, continuously turned over, and able to remodel along the direction of load. Elastin gives large recoverable strain, like rubber, but interwoven with collagen so stiffness rises sharply at higher strain. And ground substance, water held by large sugar-protein molecules, resists compression like a hydrogel. The fluid flowing through it is why tissue is time-dependent.
These are the tissues a maker most often confuses. A tendon connects muscle to bone and transmits force; it is not the motor. A ligament connects bone to bone and stabilises a joint. Articular cartilage covers bone ends, and its load-bearing depends on fluid trapped in its matrix, so it is not a rubber pad. Discs and menisci are partly fluid-pressure structures. Fat varies hugely, and bone remodels continuously. A teaching model that renders tendon, ligament and cartilage as the same rubbery connector has quietly taught three errors. Simplify, and label the simplification. Twelve words prevent the most common misconception about joints.
Now skin, the largest organ by surface area, roughly one and a half to two square metres in an adult.
Skin is a layered composite with active biology. The epidermis is a twentieth to a tenth of a millimetre thick, has no blood vessels, and renews itself from the bottom up about once a month. The dermis, one to four millimetres, is the mechanically dominant layer and carries everything functional: blood vessels, lymphatics, nerve endings, hair follicles and glands. Beneath it the hypodermis of fat and loose tissue varies by a factor of ten or more between sites and people, which is why a design validated on one forearm tells you little about another. And a bony prominence underneath concentrates load deep, before anything shows at the surface.
The epidermis is mostly keratinocytes, which build the barrier, with immune Langerhans cells and Merkel cells for light touch. Melanocytes produce melanin, and their number is broadly similar across human populations; what differs is the amount and type of melanin. That has a direct engineering consequence for any optical sensor, as chapter nine shows for the pulse oximeter. The barrier itself is the outer stratum corneum plus its lipids. Strip it, with solvent, adhesive, repeated tape removal, or soaking from trapped moisture, and the barrier is compromised before any wound is visible. Medical adhesive skin injury is a recognised harm caused directly by design decisions about tapes and straps.
Each function of skin has a design consequence. It is a barrier, so any breach changes the project category, and any chemical contact, whether adhesive, residual uncured resin, textile dye or a nickel buckle, is a chemical exposure. It senses, so the person's discomfort report is a safety mechanism, and if sensation is reduced that mechanism is simply absent. It regulates temperature, so covering skin is a thermal and microclimate decision. And it manages water, so occlusion causes maceration: softened, weakened, more easily damaged skin.
Now the mechanics of contact, which is where the engineering intuition helps and misleads in equal measure.
Average pressure is force divided by area. A bench fixture loaded with twenty newtons over ten square centimetres gives twenty kilopascals, about a hundred and fifty millimetres of mercury. Double the area at the same force and it halves. That is true and useful. Now here is everything it does not tell you. Use the calculation to generate a question about how force is distributed. Never use it to invent a clinical threshold. Published interface pressure figures exist; they are contested, context-specific, and not yours to apply.
What the average cannot establish. Distribution: a single ridge, seam, layer line or fastener head can carry a large fraction of the load, so the local peak can be many times the average, and a two millimetre edge does the damage while the average looks reassuring. Depth: damage from loading often begins deep, near a bony prominence, before anything is visible. Time: a pressure tolerable for two minutes may not be tolerable for two hours, and the relationship is not a simple threshold. The person: tissue condition, perfusion, age, moisture, medication, sensation and mobility all change the outcome. And movement, which brings in two further mechanisms.
Four distinct mechanisms, routinely conflated. Pressure is load normal to the surface; it compresses tissue and can shut off blood vessels. Friction is resistance to sliding at the interface; it abrades the outer layer and causes blisters. Shear is load parallel to the surface, carried into the tissue so its layers slide relative to one another, distorting and kinking the vessels that supply it while the surface may look normal. It is the mechanism people most forget, and why "I reduced the pressure" can coexist with "the tissue got worse". And microclimate: trapped moisture macerates skin and raises friction, and heat raises the metabolic demand of tissue whose supply may already be compromised.
This surprises anyone used to metals. Soft tissue's stress-strain curve is J-shaped: crimped collagen straightens easily at first, then stiffness rises steeply, so there is no single modulus of skin. It is viscoelastic, showing creep, stress relaxation and hysteresis, so a strap tightened comfortably at nine in the morning is at a different tension by eleven, in both directions, because tissue creeps and limb volume changes. It is anisotropic, following its collagen fibres, so a symmetric part is not symmetric in its effect. It is pre-stressed in the body, so data from excised tissue does not transfer directly. And it adapts, so the tissue in week six is not the tissue you measured in week one.
These are wide literature ranges, useful only for knowing which decade you are in. Fat is a few kilopascals. Skin spans three orders of magnitude depending on strain and method. Silicone sits between a tenth and ten megapascals. Rigid engineering plastics, whether moulded, machined or printed, are one to four gigapascals, roughly as stiff as tendon and a million times stiffer than fat. Cortical bone is ten to twenty gigapascals, and stainless steel ten times stiffer again. Every rigid object you place against a person is mechanically much harder than what it touches, and the compliance mismatch is where load concentrates.
Repair runs in four overlapping phases: clotting within minutes, inflammation over hours to days, proliferation of new vessels and collagen over days to weeks, and remodelling over weeks to months or longer. The classic signs of inflammation, redness, heat, swelling, pain and loss of function, are a normal response, but they tell you the tissue has changed, so any geometry fitted before may no longer fit. Swelling in particular is a direct dimensional problem; a limb's circumference can change substantially over hours. And repair produces scar, not replacement, with different strength, elasticity and sensation. Which is why preventing a five millimetre skin break matters far more than being good at responding to one.
Intact, compromised and breached skin are three different projects. No body contact, such as a tabletop teaching model, needs a workshop review and cleaning for handling. Brief contact with intact skin needs a human factors review of edges, forces and cleaning. Prolonged contact, such as a splint, strap or seat, needs clinical assessment and is not a maker's decision alone. Compromised skin always needs specialist clinical involvement. And breached skin, mucosa or anything internal is a different regulatory and clinical universe. Work out which row your project is in before you build, and write it in the brief. A project that moves up a row has become a different project.
Ask all of these before you commit geometry, and get the answers from the clinical team and the user, not from this book. Where exactly does it contact, over what area, for how long and how often? What forces act? Is the skin intact, and who assessed that? What lies beneath: a bony prominence, a superficial nerve, a vessel? Does the limb change size? What are the edges, seams, parting lines and fasteners doing at the boundary? Can the user feel discomfort and act on it? How is it cleaned, and who re-checks it next week? What happens when it fails, and how is it removed in a hurry? And is there an existing product, and why is it unsuitable?
Padding is not a universal solution, and every modification is a new design. Increasing area lowers average pressure but adds bulk and new edges. Padding distributes load but changes grip diameter, adds seams, absorbs fluid, and can bottom out over a hard edge it now conceals. Rounding edges is almost always good. A softer material reduces mismatch but may lose the function that justified the part. Adjustability accommodates swelling but can be set wrongly. And changing the manufacturing method, from a print to a machined or moulded part, or from a stitched seam to a bonded one, removes a feature the process forced on you. The honest statement is always: this changes X, and I have not established Y.
Skin supports a large device industry. Wound dressings are a moisture-management problem. Negative pressure wound therapy is a genuine engineering device: pump, canister, seal and alarms. Pressure redistribution mattresses and cushions spread load, and pressure-mapping mats show surface pressure only, not the deep stress that fails. Slide sheets attack shear directly, and seating is foam, upholstery and textiles. On the software side, wound imaging apps struggle with lighting and colour, risk scores are prompts rather than predictions, AI lesion triage has a documented weakness on darker skin tones, and orthotic CAD and 3D scanning capture body geometry without radiation.
Training technology is the most accessible part of this area. Suture and skin pads are consumed by the hundred in surgical skills teaching, commercial ones are expensive, and a layered silicone pad with a realistic skin, fat and fascia feel is squarely within a maker's competence. The skill is in mould making: machined or printed moulds, controlled pours, degassing, and fabric layers for tear resistance, and in making the hundredth pad match the first. Moulage, instrumented interface rigs with force sensors under a cast silicone block, and sewn demonstration straps that vary one thing each all teach the chapter's ideas without touching a patient.
An occupational therapist says a user finds a handle uncomfortable, and your hands reach for foam. Stop and ask what uncomfortable means. It might be force concentrated at an edge or seam, too little contact area, an awkward wrist position, too much grip force needed, slipping, or heat and moisture. Each has different evidence and a different fix. For three of them padding makes things worse: it increases the diameter, which raises grip force, it can make slipping worse, and it traps moisture. The design move is to find out which mechanism is operating, which needs the therapist and the user, not to apply the default fix.
Exercises. Build a skin cross-section from card and foam, mark vessels, nerves and glands in the right layers, add a bony prominence, and label it schematic, not to scale. Sketch a narrow and a broad contact carrying the same load, calculate both averages, and mark where an edge or seam would concentrate force. Stack foam sheets and push the top sideways to see shear, then write three ways it is unlike skin. And draw every contact your elbow model makes; you will reuse it later. One warning: do not run pressure or prolonged-contact experiments on yourself or anyone else. Damage develops without warning and cannot be undone by stopping.
The limits. You cannot set a pressure limit; you now understand why the question is hard, and the answer belongs to the clinical team and the specialist literature. You cannot assess anyone's skin. A bench measurement of interface pressure characterises one loading of one person at one moment and is not a safety assessment. "It is skin-safe" is a claim requiring evidence about a finished object in a specific use. And nothing here authorises prolonged-contact or wound-contact work. Next, bones, joints, and movement.