Flow Rate Computation of Highly Turbulent Pipe Flows using Wide-Band Signals and Matched Filter-Based Approach in Ultrasonic Multi-Element Configuration - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Flow Rate Computation of Highly Turbulent Pipe Flows using Wide-Band Signals and Matched Filter-Based Approach in Ultrasonic Multi-Element Configuration

Cornel Ioana
Angela Digulescu
Ion Candel
  • Fonction : Auteur
  • PersonId : 928771
Gabriel Vasile

Résumé

Flow profile analysis and flow rate estimation with ultrasound Non-Destructive Testing (NDT) tools were performed in a pipe flow varying the Reynolds number from 4.33x10E5 to 1.67x10E6. This study investigated the flow rate computation right downstream a 90 deg curvature pipe with a clamp-on multi-path ultrasonic flow meter (UFM) and a compensation scheme for the time-of-flight (TOF) errors due to the asymmetrical flow profile. The challenge is to model through the CFD analysis the separation region in the intrados of the pipe's elbow and its influence on the flow velocity profile, in order to calibrate the flow rate computations. The use of the matched filtering technique improves the performances in terms of measurement precision, which is validated by the numerical models. The experimental measurements on TOF differences in the pipe's elbow prove the interest of such approach for the improvement of the flow rate computations in the case of relative high turbulent flows.
Fichier non déposé

Dates et versions

hal-01065851 , version 1 (18-09-2014)

Identifiants

  • HAL Id : hal-01065851 , version 1

Citer

Teodor Petrut, Cornel Ioana, Angela Digulescu, Ion Candel, Gabriel Vasile. Flow Rate Computation of Highly Turbulent Pipe Flows using Wide-Band Signals and Matched Filter-Based Approach in Ultrasonic Multi-Element Configuration. IGHEM 2014 - 10th International Conference of Hydraulic Efficiency Measurement, Sep 2014, Itajuba, Brazil. pp.12. ⟨hal-01065851⟩
415 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More