The use of Simulation in Electric Machine Design Stefan Holst, CD-adapco

Transcription

1 The use of Simulation in Electric Machine Design Stefan Holst, CD-adapco

2 Motivation How often can a machine be started within an hour In Hybrids, what effect has the adjacent combustion drive train Space restrictions opposing power requirements in general In batteries, overheating leads to faster aging A joint simulation approach allows to simulate the coupled electric and thermal problem early on in the design process

3 CD-adapco s Purpose CD-adapco s purpose is to inspire innovation while reducing customer costs through the application of engineering simulation 3

4 CD-adapco: Engineering Success Our purpose is to ensure the customer s success through the use of engineering simulation Enable & inspire innovation Reduce engineering time & costs We provide successful engineering simulation solutions Software products like STAR-CCM+ that are accurate, efficient, and easy to use Flow, Thermal and Stress simulation in a single tool. Local dedicated support Engineering services: technology transfer, burst engineering resources, custom software tools We are a growing and successful engineering simulation company 20% growth in FY2011 global software sales $138m End User Spend in FY2011 >700 employees in ~25 offices 40% of employees involved in Research and Development activities >8000 users worldwide Our independence breeds engineering success Largest independent CAE/CFD provider Heavily invest in employees Continuously invest in development of new technology 4

5 Overview SPEED s own thermal capability consists of simplified circuit models Good for a first shot Circuit model allows for quick analysis of duty cycle simulations Efficiency requirements, space and weight restrictions lead to more complex machine and machine enclosure designs Identifying hot spots becomes tricky STAR-CCM+ can show its strength in these models User should benefit as much as possible from the work done in SPEED Import geometry Import distribution of heat loads electromagnetic losses Future: coupled EMAG and thermal analysis in a single tool

6 Early to mid stages of machine design SPEED is the leading design software for electric machines Detailed analytical analysis with finite-element links or finite-embedded solver for motors, generators and alternators including inverters and other electronic controls Specialized tools per motor type improve design efficiency PC-BDC for brushless permanent magnet and wound-field AC synchronous PC-IMD for induction machines PC-SRD for switched reluctance PC-DCM for direct current (PM) PC-WFC for wound field and PM commut.

7 SPEED The design process

8 Equivalent circuit methods Instantaneous results with todays computers Assumptions on flux path and mathematical models give estimates on circuit parameter Inter phase dependence is difficult to incorporate (SRM) Modeling assumptions have certain validity range only How to keep computational speed but extend applicability Use 2D/3D FEA to calibrate, but limit to as to few key simulation conditions

9 SPEED in use: GoFER Go to Finite-Elements and Return Two FE runs only or use the embedded FE-solver directly (PC-BDC)

10 Data transfer to STAR-CCM+ The B-field variation allows iron loss estimation GoFER of 72 rotor positions /elec. revolution Modified Steinmetz method in SPEED applies also to non sinusoidal currents Element Table tool to view and export cell-wise values

11 Thermal modelling in STAR-CCM+ Simulation goal: steady state temperatures A water cooled casing and end winding regions are added to the SPEED model import Coolant is induced at 50 C, at a volume flow rate of 1 l/s Identify interfaces where thermal resistances model neglected fine details slot liner, stator to housing contact imperfections, magnet to rotor contact Rotation is modeled with a moving referernce approach Source to the momentum equation in a domain enclosing the rotor and shaft parts Example taken from the paper STAR-CCM+ and SPEED for electric machine cooling analysis, CWIEME 2012, Berlin

12 Thermal modelling in STAR-CCM+ Resolving each wire is too costly Model winding as bulk region Gaps between wires and wire insulation are accounted for using anisotropic material properties Local wire orientation is interpolated from easily selectable edge orientation Off directional thermal conductivities is tuned to resemble spread between mean and max winding temperatures Air and coolant flow is solved using standard turbulence model settings in STAR-CCM+

13 Results Using the STAR-CCM+ coupled solver in the flow converges within 120 iterations 1000 rpm 6000 rpm The temperature distribution clearly indicates that the windings closer to the coolant outlet are under higher thermal stress

14 SPEED and STAR-CCM+ the combined workflow for Electrical Machines PC-FEA: Loss table calculation Reading the SPEED geometry and the loss distribution Running the final advanced thermal calculation. Initial design with SPEED Temperatures impact life time, reliability, cost & size

15 E-motor simulations Applications Hydraulics IC engines Electro-magnetic systems Battery control STAR-CCM+ v8.02 couples with 1D codes AMESIM Wave OLGA And via scripting also to Matlab/Simulink Applications Full vehicle simulation Drive cycle simulations Control of the battery Prevent excessive loading on high temperature 15

16 Protocol Design Start simulation Open port Specify reports Specify input: field functions time step size Update circuit time step Macro connects to Simulink port Store list of reports Run Step Send reports

17 Matlab/Simulink

18 Thank you