Simulation of a Coriolis Flowmeter under Two-Phase Flow Conditions with Signal Parameter Estimation using the Matrix Pencil Method

Authors

  • Dmitry Alexandrovich Goncharov South Ural State University, Chelyabinsk
  • Ivan Igorevich Fedosov South Ural State University, Chelyabinsk
  • Alexander Leonidovich Shestakov South Ural State University, Chelyabinsk

DOI:

https://doi.org/10.14529/mmph260407

Keywords:

coriolis flowmeter, simulation model, random walk model, matrix pencil method, validation

Abstract

Coriolis flowmeters provide high measurement accuracy under single-phase flow conditions. However, gas entrainment in a liquid changes the natural frequency and damping of the tube-flow system, thereby increasing measurement errors and potentially disrupting resonant operation. This study proposes a simulation model for reproducing flowmeter signals under single-phase and gas-liquid two-phase flow conditions. The model uses a second-order discrete transfer function whose time-varying coefficients represent the natural frequency and damping ratio of the tube-flow system. A modified random walk model generates variations in these parameters. The classical matrix pencil method estimates the frequency and amplitude of both simulated and experimental signals. The signal processing algorithm operates on blocks of 50 samples at a sampling frequency of 1 kHz. The model calibration procedure uses experimental data from a DN15 Coriolis flowmeter operating at mass flow rates of 369–1741 kg/h and gas volume fractions from 0 to 42 %. Statistical validation focuses on signal frequency estimates and combines the augmented Dickey–Fuller test, descriptive statistics, empirical distribution functions, the first Wasserstein distance, and quantile differences. The Wasserstein distance remains below the specified tolerance of 0,25 Hz at all investigated operating points. The model meets the additional maximum quantile difference criterion in all cases except at 1037 kg/h and a gas volume fraction of 13 %. These results support the ability of the proposed model to reproduce the distributions of experimental signal frequency estimates over the investigated operating range. This simulation model provides a basis for developing and tuning control algorithms that maintain resonant operation of Coriolis flowmeters under two-phase flow conditions.

Author Biographies

Dmitry Alexandrovich Goncharov, South Ural State University, Chelyabinsk

Post-Graduate Student of Informational and Measuring Technology

Ivan Igorevich Fedosov, South Ural State University, Chelyabinsk

Cand. Sc. (Engineering), Associate Professor of Informational and Measuring Technology

Alexander Leonidovich Shestakov, South Ural State University, Chelyabinsk

Dr. Sc. (Engineering), Professor of Informational and Measuring Technology

Published

2026-11-13