This paper presents a physics-based approach aimed to predict the performance curves of PATs (pumps as turbines). The methodology includes a tuning procedure, which allows the calibration of the physics-based model on a given dataset of pumps/PATs, and a prediction procedure, which allows the estimation of PAT performance curves for previously unknown PATs. The methodology is applied to six pumps/PATs, of which the performance curves were experimentally determined at different rotational speeds. A cross-validation process is applied in such manner that one of the six pumps/PATs is not employed for the tuning procedure and is used for validating the reliability of the prediction procedure. The results show that, in the range ±20% with respect to the experimental BEP, the deviation of the predicted PAT performance curves from experimental curves can be acceptable from an engineering point of view, with an average value of 13.4%, 14.0% and 12.4%, for head, power and efficiency, respectively.
Development and validation of a comprehensive methodology for predicting PAT performance curves
Castorino G. A. M.
Primo
;Manservigi L.Secondo
;Losi E.Penultimo
;Venturini M.Ultimo
2023
Abstract
This paper presents a physics-based approach aimed to predict the performance curves of PATs (pumps as turbines). The methodology includes a tuning procedure, which allows the calibration of the physics-based model on a given dataset of pumps/PATs, and a prediction procedure, which allows the estimation of PAT performance curves for previously unknown PATs. The methodology is applied to six pumps/PATs, of which the performance curves were experimentally determined at different rotational speeds. A cross-validation process is applied in such manner that one of the six pumps/PATs is not employed for the tuning procedure and is used for validating the reliability of the prediction procedure. The results show that, in the range ±20% with respect to the experimental BEP, the deviation of the predicted PAT performance curves from experimental curves can be acceptable from an engineering point of view, with an average value of 13.4%, 14.0% and 12.4%, for head, power and efficiency, respectively.File | Dimensione | Formato | |
---|---|---|---|
176.pdf
solo gestori archivio
Descrizione: Full text editoriale
Tipologia:
Full text (versione editoriale)
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
4.03 MB
Formato
Adobe PDF
|
4.03 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.