Uncertainty Structure
Systematic documentation and quantification of all measurement error sources establishes the combined uncertainty associated with micro-fabricated thin film transducers. Assembling a thin film sensor uncertainty budget involves identifying individual variance components, establishing mathematical error propagation models and calculating expanded uncertainty. The analysis incorporates uncertainty contributions from calibration references, signal conditioning electronics, sensor non-linearity and environmental sensitivity.
Metrology standards mandate the inclusion of both Type A statistical evaluations and Type B systematic assessments. The resulting expanded uncertainty defines the measurement confidence interval at a specified coverage factor.
Variance Allocation
Individual variance components are categorized according to their origin within the measurement chain. Gauge factor calibration uncertainty and leadwire resistance corrections represent primary contributors to mechanical strain measurement error. Temperature-induced apparent strain, sensor drift and transverse sensitivity cross-talk form significant environmental uncertainty components.
Piezoresistive grid thickness variations and deposition non-uniformities introduce manufacturing variance across sensor batches. Data acquisition analog-to-digital resolution and bridge excitation stability contribute electronic noise components.
Propagation Model
Mathematical measurement models link individual input quantities to the final output measurand through functional equations. Partial derivatives of the model function yield sensitivity coefficients that scale each input variance to its output contribution. Summing individual variance components in quadrature calculates the combined standard uncertainty under the assumption of independent inputs.
Correlation coefficients must be calculated and applied when input parameters, such as temperature and bridge output, exhibit mutual dependence. Multiplying the combined standard uncertainty by a coverage factor of two yields an expanded uncertainty reflecting an approximate ninety-five percent confidence level.
Budget Verification
Validation of the uncertainty budget requires experimental testing using certified reference standards under controlled environmental conditions. Inter-laboratory comparisons and proficiency testing verify that empirical measurement errors fall within the predicted uncertainty envelope. Environmental changes outside evaluated operational boundaries invalidate stated budget figures and require reassessment.
Uncertainty budgets must be revised whenever deposition processes, signal conditioning hardware or calibration algorithms undergo modification. Final measurement reports must declare the expanded uncertainty along with the coverage factor and degrees of freedom.