Composition-partitioned volumetric dynamic temperature imaging

A holographic tomography technique is presented for real-time 3-D imaging of dynamic combustion fields. The technique combines a three-angle Mach-Zehnder interferometer with off-axis multiplexed recording, enabling simultaneous multi-path acquisition within a single shot. To mitigate angular sparsity, ART+TV regularization is applied, improving imaging accuracy and structural fidelity. For quantitative temperature reconstruction, a composition-partitioned model is introduced that overcomes the fixed-air assumption of the conventional Gladstone-Dale model, enabling more accurate reconstruction of temperatures in combustion fields and stratified combustion environments. Quantitative validation against thermocouple measurements shows relative deviations within 1.14-1.93%, demonstrating improved reconstruction accuracy. Experiments with propane flames demonstrate successful 3-D temperature imaging at 6 × 10³ fps, with a peak temperature of 1750 K, providing a potential tool for rapid diagnosis of transient combustion or dynamic flow fields.