Mechanical parts in a deforming interface wear and drift more than a purely electronic display

What it is

Pixels mostly age in brightness uniformity; pins, membranes, pumps, and valves also walk height zero, travel, and seals away. Deforming mechanisms wear and drift more than a purely electronic display means calibration and life have to be done as mechanical parts, not copied from a panel’s MTBF. Whether shape survives power-off is another layer. This one is, while powered, how height consistency among cells scatters across cycles.

Why it happens

Every pin guide, membrane crease, and valve seal is a wear face. Cycles open clearance, zeros crawl, and the same commanded height becomes a tilted field. Grit in the gaps sticks individual cells; the array grows “dead pixels” that will not come back from a software calibration tap the way a panel might—stuck is a mechanical event. Pneumatic leak drops held height and needs continuous makeup air; power and noise both rise. People read geometry from relative height; one pin off, and a ridge bends, a plane bulges. A dead pixel on an electronic display is usually visual noise; a dead cell in deformation is a pit the hand will trip on.

Studying it

Cycle raise/lower to the target life, sampling flatness and single-pin travel along the way, with dust and side loads injected.

Independent variables: cycle count, load (display only / also pressed by a hand), contamination, mid-life calibration or not. Dependent measures: height standard deviation, stuck-cell count, leak of held height, whether people can still read the target shape.

Demo flatness on a new array is not a spec. At mid-life and end-of-life, measure whether “follow a ridge, read stiffness” still completes. If calibration needs a teardown, the field cannot calibrate.

Where it stops holding

An exhibition that rises a few times a day and must not be touched has a wear curve closer to electronics. Trainers pressed thousands of times a day have wear as the main cost. Liquid or powder environments pull gap problems into early life. Some applications can accept “swap the whole board when it fails”; a consumer desk cannot. A single bladder membrane fails as a whole face going soft or bulging on one side, unlike pins’ scattered dead cells; the maintenance policy does not copy.

Applying it

• Write flatness, stuck rate, and leak into a spec bound to cycle count; do not only quote new-part numbers.
• Leave calibration when it can be done in the field; if it needs a teardown, design margin and replacement as if uncalibratable.
• Arrays a hand will press should quote life on loaded cycles.
• Verify: at mid-life, run find-ridge / read-stiffness against a new part. When height SD kills the shape task, that is the replacement point. Stick several cells on purpose and see whether content can degrade. Retest after a dust chamber.

Related

• Same group: C10.15.1 Physical pixels (pin arrays) present shape and texture by locally raising a solid surface · C10.15.2 Deformation output can carry 3D geometry or stiffness that digital content cannot present · C10.15.3 Deformation mechanisms are far coarser and slower than a visual display of the same size
• Adjacent: C10.16 Mechanical Reliability and Lifetime · C10.06 Shape-Changing Interfaces
• Search: actuator drift · pin wear · shape calibration