Full Text Available

Note: Clicking the button above will open the full text document at the original institutional repository in a new window.

The analysis and optimization of an axial compressor

Thesis (MSc)--Stellenbosch University, 2015.

Saved in:
Bibliographic Details
Main Author: Hamman, Richard Alan
Other Authors: Venter, Gerhard
Format: Thesis
Language:en_ZA
Published: Stellenbosch : Stellenbosch University 2015
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1867614017970765824
access_status_str Open Access
author Hamman, Richard Alan
author2 Venter, Gerhard
author_browse Hamman, Richard Alan
Venter, Gerhard
author_facet Venter, Gerhard
Hamman, Richard Alan
author_sort Hamman, Richard Alan
collection Thesis
dc_rights_str_mv Stellenbosch University
description Thesis (MSc)--Stellenbosch University, 2015.
format Thesis
id oai:scholar.sun.ac.za:10019.1/97936
institution Stellenbosch University (South Africa)
language en_ZA
last_indexed 2026-06-10T12:45:21.489Z
license_str Other — see source repository
provenance_str_mv Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository
publishDate 2015
publishDateRange 2015
publishDateSort 2015
publisher Stellenbosch : Stellenbosch University
publisherStr Stellenbosch : Stellenbosch University
record_format dspace
source_str SUNScholar — Stellenbosch University Repository
spelling oai:scholar.sun.ac.za:10019.1/97936 The analysis and optimization of an axial compressor Hamman, Richard Alan Venter, Gerhard Van der Spuy, Johan Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering. Axial compressor -- Analysis Axial compressor -- Optimization Meta-model based design optimization (MBDO) UCTD Thesis (MSc)--Stellenbosch University, 2015. ENGLISH ABSTRACT: Axial compressors are widely used in a variety of contexts. When compared to centrifugal compressors, axial compressors offer higher pressure ratios and efficiencies. Due to the complex relationship between blade shape and compressor performance, as well as the sensitivity of axial compressors to massflow rates, the design of an axial compressor is a challenging problem. These design difficulties can be circumvented using numerical design optimization. In this work, meta-model based design optimization (MBDO), a variant of numerical design optimization, is used to develop replacement blades for a low speed axial compressor. Two numerical models were developed: a low fidelity, computationally inexpensive single stage model and a high fidelity, computationally expensive three stage model. For reasons of computational cost, the single stage model was used to evaluate the objective function in the optimization process. An optimized blade design was developed, which delivered a 9:83% increase in pressure coefficient over the original design, when evaluated with the single stage model. This increase was not sustained when evaluating the optimized design with the three stage model. After investigation, the cause was found to be a high incidence angle near the hub and shroud, just outside the range in which incidence angle was constrained in the design problem. To compensate for this, the blade angles were manually adjusted, in order to lower the incidence angle. The adjusted blades delivered an increase of 8:54% over the original design, when evaluated with the three stage model. AFRIKAANSE OPSOMMING: Aksiale kompressors word benut in 'n wye verskeidenheid omgewings. In vergelyking met sentrifugale kompressors bied aksiale kompressors hoër werksverrigting en doeltreffendheid. Die komplekse verhouding tussen lemvorm en kompressorwerkverrigting sowel as die sensitiwiteit van aksiale kompressors vir massavloeitempo maak die ontwerp van 'n aksiale kompressor 'n uitdagende taak. Hierdie ontwerpsuitdagings kan omseil word deur van numeriese ontwerpsoptimering gebruik te maak. In hierdie werk word metamodel gebaseerde ontwerpsoptimering (MBDO), 'n variasie van numeriese ontwerpsoptimering, gebruik om nuwe lemme vir 'n lae-spoed aksiale kompressor te ontwerp. Twee numeriese modelle was ontwerp: 'n lae- akkuraatheid, lae-berekeningskoste enkel stadium model, en 'n hoë-akkuraatheid, hoë- berekeningskoste drie-stadium model. As 'n gevolg van berekeningskostes is die enkel stadium model gebruik om die doel funksie in die optimeringsproses te evalueer. 'n Optimale lemontwerp is ontwikkel wat 'n verhoging van 9.83% in die druk-koëffisiënt toon teenoor die oorspronklike ontwerp indien dit met die enkel stadium model ge-evalueer word. Hierdie verbetering is nie sigbaar met die evaluasie van die drie-stadium model nie. Die oorsaak hiervoor was 'n hoë invalshoek naby die middelpunt en deelspanmantel wat net buite die omvang van die invalshoek se beperkinge in die ontwerpsprobleem was. As kompensasie hiervoor, is die lemhoeke aangepas om die invalshoek te verminder. Die aangepaste lemme toon 'n verhoging van 8.54% teenoor die oorspronklike ontwerp as dit met die drie-stadium model ge-evalueer word. 2015-12-14T07:43:18Z 2015-12-14T07:43:18Z 2015-12 Thesis http://hdl.handle.net/10019.1/97936 en_ZA Stellenbosch University 142 pages : illustrations application/pdf Stellenbosch : Stellenbosch University
spellingShingle Axial compressor -- Analysis
Axial compressor -- Optimization
Meta-model based design optimization (MBDO)
UCTD
Hamman, Richard Alan
The analysis and optimization of an axial compressor
title The analysis and optimization of an axial compressor
title_full The analysis and optimization of an axial compressor
title_fullStr The analysis and optimization of an axial compressor
title_full_unstemmed The analysis and optimization of an axial compressor
title_short The analysis and optimization of an axial compressor
title_sort analysis and optimization of an axial compressor
topic Axial compressor -- Analysis
Axial compressor -- Optimization
Meta-model based design optimization (MBDO)
UCTD
url http://hdl.handle.net/10019.1/97936
work_keys_str_mv AT hammanrichardalan theanalysisandoptimizationofanaxialcompressor
AT hammanrichardalan analysisandoptimizationofanaxialcompressor