Strain-Gauge Bridge Nonlinearity in Student-Built Engineering Education Load Frames: A Dead-Weight Calibration Evaluation Linking Excitation Voltage to Structural Safety Margin Estimates

Student-built load frames used in engineering education often rely on strain-gauge bridges whose output is assumed linear with applied load, yet the excitation voltage chosen for the bridge can shift the operating point and introduce nonlinearity that propagates into safety-margin estimates. The protocol specifies a four-level excitation sweep and a five-level dead-weight ladder, with each combination replicated to separate gauge hysteresis from excitation-dependent nonlinearity. A variance-component analysis decomposes the observed calibration residuals into gauge, excitation, and interaction terms, and a first-order propagation theorem bounds the safety-margin error as a function of the bridge nonlinearity coefficient and the excitation deviation. The analysis indicates that excitation-induced nonlinearity contributes a systematic bias to safety-margin estimates that is not removed by routine zero-and-span recalibration, and that the bias grows with the square of the excitation deviation from the nominal bridge voltage. The paper concludes that calibration protocols for student-built frames should report excitation voltage alongside span and zero adjustments, and that safety-margin estimates derived from such frames should carry an explicit excitation-dependent uncertainty band.