Modelling pressure distribution in a rectangular gas bearing using neural networks

Mansour Karkoub, Ali Elkamel

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Gas lubricated bearings are of tremendous use especially in the biomedical and aerospace industries. For that reason, gas bearings have been the subject of much research for the past decade or so. Experimental as well as theoretical work has been done to calculate the pressure distribution inside the bearing. The models available to predict the pressure are primitive and need to be improved. This paper discusses a new modelling scheme known as artificial neural networks. The pressure distribution and the load-carrying capacity are predicted using feedforward architecture of neurons. The inputs to the networks are a collection of experimental data. This data is used to train the network using the Levenberg-Marquardt optimization technique. The results of the neural network model are compared to a theoretical model and the results are promising. The neural network model outperforms the available theoretical model in predicting the pressure as well as the load-carrying capacity.

Original languageEnglish
Pages (from-to)139-150
Number of pages12
JournalTribology International
Volume30
Issue number2
Publication statusPublished - Feb 1997
Externally publishedYes

Fingerprint

Gas bearings
gas bearings
pressure distribution
Pressure distribution
Neural networks
load carrying capacity
Load limits
Gas lubricated bearings
Bearings (structural)
aerospace industry
Aerospace industry
neurons
Neurons
industries
optimization

Keywords

  • Backpropagation
  • Gas bearings
  • Levenberg-Marquardt technique
  • Neural networks

ASJC Scopus subject areas

  • Mechanics of Materials
  • Mechanical Engineering
  • Surfaces and Interfaces
  • Surfaces, Coatings and Films

Cite this

Modelling pressure distribution in a rectangular gas bearing using neural networks. / Karkoub, Mansour; Elkamel, Ali.

In: Tribology International, Vol. 30, No. 2, 02.1997, p. 139-150.

Research output: Contribution to journalArticle

@article{05da80f03ad44b9a98136418243a373d,
title = "Modelling pressure distribution in a rectangular gas bearing using neural networks",
abstract = "Gas lubricated bearings are of tremendous use especially in the biomedical and aerospace industries. For that reason, gas bearings have been the subject of much research for the past decade or so. Experimental as well as theoretical work has been done to calculate the pressure distribution inside the bearing. The models available to predict the pressure are primitive and need to be improved. This paper discusses a new modelling scheme known as artificial neural networks. The pressure distribution and the load-carrying capacity are predicted using feedforward architecture of neurons. The inputs to the networks are a collection of experimental data. This data is used to train the network using the Levenberg-Marquardt optimization technique. The results of the neural network model are compared to a theoretical model and the results are promising. The neural network model outperforms the available theoretical model in predicting the pressure as well as the load-carrying capacity.",
keywords = "Backpropagation, Gas bearings, Levenberg-Marquardt technique, Neural networks",
author = "Mansour Karkoub and Ali Elkamel",
year = "1997",
month = "2",
language = "English",
volume = "30",
pages = "139--150",
journal = "Tribology International",
issn = "0301-679X",
publisher = "Elsevier Inc.",
number = "2",

}

TY - JOUR

T1 - Modelling pressure distribution in a rectangular gas bearing using neural networks

AU - Karkoub, Mansour

AU - Elkamel, Ali

PY - 1997/2

Y1 - 1997/2

N2 - Gas lubricated bearings are of tremendous use especially in the biomedical and aerospace industries. For that reason, gas bearings have been the subject of much research for the past decade or so. Experimental as well as theoretical work has been done to calculate the pressure distribution inside the bearing. The models available to predict the pressure are primitive and need to be improved. This paper discusses a new modelling scheme known as artificial neural networks. The pressure distribution and the load-carrying capacity are predicted using feedforward architecture of neurons. The inputs to the networks are a collection of experimental data. This data is used to train the network using the Levenberg-Marquardt optimization technique. The results of the neural network model are compared to a theoretical model and the results are promising. The neural network model outperforms the available theoretical model in predicting the pressure as well as the load-carrying capacity.

AB - Gas lubricated bearings are of tremendous use especially in the biomedical and aerospace industries. For that reason, gas bearings have been the subject of much research for the past decade or so. Experimental as well as theoretical work has been done to calculate the pressure distribution inside the bearing. The models available to predict the pressure are primitive and need to be improved. This paper discusses a new modelling scheme known as artificial neural networks. The pressure distribution and the load-carrying capacity are predicted using feedforward architecture of neurons. The inputs to the networks are a collection of experimental data. This data is used to train the network using the Levenberg-Marquardt optimization technique. The results of the neural network model are compared to a theoretical model and the results are promising. The neural network model outperforms the available theoretical model in predicting the pressure as well as the load-carrying capacity.

KW - Backpropagation

KW - Gas bearings

KW - Levenberg-Marquardt technique

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=0031077447&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031077447&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0031077447

VL - 30

SP - 139

EP - 150

JO - Tribology International

JF - Tribology International

SN - 0301-679X

IS - 2

ER -