애스크로AIPublic Preview
← 학술논문 검색
학술논문한국유체기계학회 논문집2025.02 발행

평균유선 해석법의 손실 모델 계수 보정을 통한 원심펌프의 성능 예측 정확도 개선

Improvement in Accuracy of Performance Prediction for Centrifugal Pumps by Loss Model Calibration of Meanline Analysis

심현석(국립창원대학교)

28권 1호, 16~22쪽

초록

This study presents an improved one-dimensional meanline analysis for predicting the hydraulic performance of a single-stage centrifugal pump. The basic model for the meanline analysis is derived from the Euler equation, defined by the velocity triangle and the Wiesner slip factor. Empirical coefficients from the loss models proposed in previous studies are listed and selected as variables for calibration. Latin hypercube sampling is employed to generate multiple sampling points to calibrate the coefficients in the loss models. Three centrifugal pumps with different specific speeds are selected, and the relative errors of the head and efficiency values obtained from experiments and the meanline analysis are calculated to quantify the prediction model's accuracy. The optimal combination of loss model coefficients that minimizes the mean square error between the experimental values and the meanline analysis is determined from the sampling points. The reliability of the methodology is demonstrated by showing that the predicted values for head and efficiency are within 5% and 10% error ranges, respectively.

Abstract

This study presents an improved one-dimensional meanline analysis for predicting the hydraulic performance of a single-stage centrifugal pump. The basic model for the meanline analysis is derived from the Euler equation, defined by the velocity triangle and the Wiesner slip factor. Empirical coefficients from the loss models proposed in previous studies are listed and selected as variables for calibration. Latin hypercube sampling is employed to generate multiple sampling points to calibrate the coefficients in the loss models. Three centrifugal pumps with different specific speeds are selected, and the relative errors of the head and efficiency values obtained from experiments and the meanline analysis are calculated to quantify the prediction model's accuracy. The optimal combination of loss model coefficients that minimizes the mean square error between the experimental values and the meanline analysis is determined from the sampling points. The reliability of the methodology is demonstrated by showing that the predicted values for head and efficiency are within 5% and 10% error ranges, respectively.

발행기관:
한국유체기계학회
분류:
기계공학

AI 법률 상담

이 논문의 주제에 대해 더 알고 싶으신가요?

460만+ 법률 자료에서 관련 판례·법령·해석례를 찾아 답변합니다

AI 상담 시작
평균유선 해석법의 손실 모델 계수 보정을 통한 원심펌프의 성능 예측 정확도 개선 | 한국유체기계학회 논문집 2025 | AskLaw | 애스크로 AI