Multi-objective optimization of a centrifugal compressor for ......2. Objective: Increase...

Multi-objective optimization of a

centrifugal compressor for turbocharging

Johannes RatzGas Turbines and Aerospace Propulsion

TU Darmstadt

Source:

MTU Friedrichshafen GmbH

Contents

2Gas Turbines and Aerospace Propulsion | Johannes Ratz 26.09.2019

Introduction

Optimization-Workflow

Results

Introduction


Introduction

Source:


Introduction


Introduction

Source:


Turbocharger

Introduction


Source:

MTU Friedrichshafen

GmbH

Introduction Centrifugal

CompressorTurbocharger

Introduction


Source:


Performance Map

Introduction

Centrifugal

Compressor

isolines

efficiency

Turbocharger

Introduction


Source:


Performance Map

Introduction

Centrifugal

Compressor

isolines

efficiency

Turbocharger

1. Objective: Increase Efficiency

Introduction


full-load curve

Source:


Performance Map

Introduction

Centrifugal

Compressor

isolines

efficiency

Turbocharger


Introduction


full-load curve

Source:


Performance Map

Introduction

Centrifugal

Compressor

isolines

efficiency

Turbocharger


2. Objective: Increase Performance Map Width

Introduction


Introduction


2. Objective: Increase Performance Map Width

Problem: A pure aerodynamic optimization often leads

to a poor structural integrity.

1. Constraint: Von Mises Stresses

2. Constraint: Eigenfrequency

CFD

CSM

Contents


Introduction


Results

Workflow


Workflow

Workflow - Parametrization


Workflow

Impeller

Diffuser

Volute



Workflow

3D Camber Surfacewith variable

blade ruling

Meridional Betaangle - Distribution

Source:

Friendship Systems



Workflow

Thicknessdistribution

3D Blade

Source:

Friendship Systems



Workflow

Geometry Export:

CFD (Meridional Contour, Blades) → geomTurbo (Numeca)

CSM (Impeller Solid) → Parasolid

Solid Final 3D Impeller and Diffuser

Source:

Friendship Systems



Workflow

A

AA-A

Volute

Geometry Export:

CFD → Parasolid

Source:

Friendship Systems



Workflow

Parametrisierung

• Hub / Shroud

• Blades

• Main- and splitterblade

With variable blade ruling and non-linear

Leading Edge

• Diffuser

With variable blade ruling and non-linear

Leading/Trailing Edge

• Impeller back

• Volute

Workflow


Workflow

Workflow - CFD


Workflow

Solver Numeca FINE/Turbo

Objectives:

• Isentropic Efficiency

𝜂𝐷𝑃

• Performance Map Width𝜋ℎ𝑖𝑔ℎ𝑙𝑦 𝑡ℎ𝑟𝑜𝑡𝑡𝑙𝑒𝑑

𝜋𝐷𝑃

Constraints:

• Total Pressure Ratio

𝜋𝐷𝑃

• Choke Mass Flow

Workflow - CSM


Workflow

Von Mises Stress

Main blade

Fillet

Constraints:

• 1. blade eigenfrequency

• Max. von Mises stress:

• Main blade

• Splitter blade

• Fillet

• Bottom

• Drill

low

high Splitter

BottomCalculix

Drill

Workflow


Workflow

Contents


Introduction


Results

Results


Results

• Evolutionary Algorithm + RSM

• 59 Parameters

• 1313 Designs

Results


Results

Choke-Mass flow > Choke_constraint

PI > PI_constraint

baseline

selected DesignsD477

D1808

Results


Results

𝜂𝐷𝑃 → + 2.2 percentage points

𝜋𝐷𝑃 → - 0.1 (3.25%)

Map width (ሶ𝑚𝐶ℎ𝑜𝑘𝑒− ሶ𝑚𝑆𝑡𝑎𝑙𝑙

ሶ𝑚𝐶ℎ𝑜𝑘𝑒)

→ + 5.3 percentage points

D477

𝜂𝐷𝑃 → + 1.4 percentage points

𝜋𝐷𝑃 → - 0.17 (5.5%)

Map width (ሶ𝑚𝐶ℎ𝑜𝑘𝑒− ሶ𝑚𝑆𝑡𝑎𝑙𝑙

ሶ𝑚𝐶ℎ𝑜𝑘𝑒)

→ + 5 percentage points

D1808

Compliance of

structural constraints

The authors acknowledge the financial support

by the Federal Ministry for Economic Affairs and

Energy of Germany (BMWi) in the project

GAMMA (project number 03ET1469).

Thank You for Your Attention!

Multi-objective optimization of a centrifugal compressor for ......2. Objective: Increase...

Documents

Transcript of Multi-objective optimization of a centrifugal compressor for ......2. Objective: Increase...