Humidity-Sensor Sorption Hysteresis in Student-Built Engineering Education Drying-Oven Rigs: A Saturated-Salt Reference Evaluation Linking Membrane Conditioning to Food-Safety Margin Reliability

Student-built drying-oven rigs used in engineering education typically rely on low-cost capacitive humidity sensors whose polymer membranes exhibit sorption hysteresis, yet the consequences of that hysteresis for the food-safety margins computed in laboratory exercises have not been quantified. The model treats the sensor response as a rate-independent Preisach operator conditioned on prior sorption history, calibrates its bounds against saturated-salt reference standards, and propagates the resulting interval through the drying-oven mass-balance calculation that students perform. Analytical results establish a monotone relationship between conditioning-path length and worst-case bias, a threshold conditioning protocol below which the safety margin is preserved, and a reliability bound expressed as a function of the number of reference salts used for calibration. The framework is methodological: it specifies the calibration and conditioning protocol that student rigs should adopt, identifies the saturated-salt set that minimises residual bias, and states the boundary conditions under which the bound holds. The contribution is a design rule for humidity-sensor conditioning in educational drying rigs that converts an unexamined sensor artefact into a quantified and manageable reliability constraint.