From capillary to elastic timescales: Temporal memory in Newtonian and viscoelastic microfluidic jet breakup

Liquid jets formed during microfluidic displacement exhibit markedly different breakup dynamics depending on dispersed phase rheology. Newtonian liquids frequently fragment into droplets, whereas polymer solutions form elongated filaments persisting over several milliseconds. We study this contrast using high-speed imaging of 35 experiments (Newtonian and viscoelastic fluids, 49 000 frames at 10 000 fps). Jet geometries are extracted automatically using a convolutional neural network, enabling systematic analysis. Autocorrelation analysis reveals that Newtonian jets decorrelate rapidly (characteristic timescale 0.54 ms), whereas viscoelastic jets retain temporal correlations for 4.40 ms. The viscoelastic memory timescale (τV=4.40±0.03 ms) falls quantitatively within the range of polymer relaxation times determined independently by shear rheology (λ=1.0−13.7 ms), directly attributing the observed temporal memory to elastic stress storage in stretched polymer chains. This agreement establishes autocorrelation analysis of high-speed imaging data as a noninvasive tool for extracting rheological relaxation timescales from interfacial dynamics in microfluidic complex-fluid flows.