DESIGN AND ANALYSIS OF BRUSHLESS D.C MOTOR FOR …

35

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

DESIGN AND ANALYSIS OF BRUSHLESS D.CMOTOR FOR REDUCE COGGING TORQUE USING

FINITE ELEMENT ANALYSIS

Ananthan N1*, Vimalraj N1 and Ganesh Kumar A1

*Corresponding Author: Ananthan N, [email protected]

The aim of this paper is to evaluate the magnetic field and motor performance of an exterior-rotor brushless DC (BLDC) motor based on two approaches, i.e., the magnetic circuit methodand the Finite-Element Analysis (FEA). An equivalent magnetic circuit model is applied toanalytically estimate the magnetic field of a BLDC motor, while the validity is verified by the two-dimensional FEA. Due to the restriction of the simplified mathematical model, the FEA is furtheremployed to be an assistant tool for the detailed design of the pole shoe of this BLDC motor.Four design cases with different pole shoe dimensions are proposed, and the one that possessesthe largest electromagnetic torque as well as the smallest cogging torque and torque ripple isfurther prototyped for electric bicycle applications.

Keywords: Finite element analysis, BLDC-Brushless DC Motor

INTRODUCTIONToday’s motors for traction in electric vehiclesare most often induction motors. In recentyears, PM-motors have become interesting,as the efficiency can be increased. This isvery important in battery applications. Basedon the computer program, some designs andthe influence of certain parameters as thenumber of poles or the airgap length arediscussed. The design that uses the stator ofthe induction motor that shall be replaced isof special interest. In addition, a compact

ISSN 2319 – 2518 www.ijeetc.comSpecial Issue, Vol. 1, No. 1, March 2015

National Level Technical Conference P&E- BiDD-2015© 2015 IJEETC. All Rights Reserved

Int. J. Elec&Electr.Eng&Telecoms. 2015

1 EEE, Arunai College of Engineering, Tiruvannamalai, India.

design is presented Now-a-days twoapproaches are widely used for the ComputerAided Design of electrical machines,identified as analysis and synthesis method.In the analysis method, the dimensions, typeof construction and details of the materialsused are provided by the designer as the inputdata. The designer calculates theperformance characteristics, and uses hisexperience to alter the design towardsmeeting the specifications by the CADpackage software.

Research Paper

36

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

This process is repeated until the designspecifications are met. In synthesis method,the logical decisions required to modify theinitialdesign are incorporated in the CADpackage. The main advantage of thesynthesismethod is that it allows non-residentexpertise to be utilized. In the thesis,theanalysis method has been used for themotor design.

PERMANENT MAGNETBRUSHLESS DC MOTORThe Permanent Magnet Brushless DC motorhave high power output to size ratio, betterefficiency, high torque per ampere, effectivepower factor and does not use brushes forcommutation. Moreover the power winding ison the stator, where its heat can be dissipatedmore easily, and the rotor loss is extremelysmall. These factors combine to keep thetorque/inertia ratio high in small motors. ThePMBLDC motors are mostly used in fractionalkW applications.The stator of a BLDC motorconsists of stacked steel lamination withwindings placed in the slots that are axially cutalong the inner periphery, and it resembles thatof an induction motor. Rotor is made of

permanent magnet and can vary from two toeight pole pairs with alternate north and southpoles.

Rotor is chosen depending upon theapplication requirements. Based on therequired magnetic field density in the rotor, theproper magnetic material is chosen to makethe rotor. Neodymium Iron boron is used as(permanent) magnet.

CAD PACKAGE OVERVIEWIn computer aided approach the design goeshand in hand with the motor design-engineering environment. In our CAD packageFinite-Element Analysis (FEA) is the basicnumeric-analysing tool. In electromagnetic fieldanalysis FEA is particularly valuable inoptimizing the design of electromagneticdevices such as motors, generators, solenoidsand so on. It is used to study the fieldconfigurations in integrated circuits andelectronic beam devices.

Figure 1: Structure of BLDC

Figure 2: BLDC Motor

37

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

ANALYTICAL DESIGNHere concentrates on the analytical design ofa SMPM. Specific parameter coherences areshown and general design reflections areoutlined. An adequate analytical loss model isderived and used in the design process.Additional considerations on the armaturereaction and the magnet protection are alsodiscussed. the design of the surface-mountedPM-motor. It is the simplest type of the PMmachine design. There are several aspectsthat must be taken into considerations whendesigning a PM-machine.

Essential criterions such as the choice ofmagnets, their arrangement (salient/non-salient rotor) and the protection againstdemagnetisation (regarding overload andthermal capability) are discussed.The designprocess starts with the definition of theconstraints and the requirements of operation.In this study, the design specifications are: Arated torque of 60 Nm at a rated speed of 1500rpm is required.

The field-weakening range should be up to3, corresponding to a maximum speed of 4500rpm. Some constraints and target values arelisted below: The inverter output line-to-linevoltage is roughly limited to a rms-value of UL-L = 35 V.

That corresponds to a peak value of thephase voltage Û of about 28.6 V (Û U L-L =2/3). The outer dimensions of the SMPMdrive are restricted to the dimension of theinduction motor that shall be replaced. the totallength is restricted to l = 0.34 m the outer statordiameter is restricted to Dy = 0.24 m. Theframe and the bearings are included in theouter dimension and reduce the effectivemotor dimension to l = 0.165 m.

The magnet characteristics are assumedas follows:

remanence flux density Br = 1.1 T

demagnetisation flux density BD = –0.2 T

relative magnet permeability r = 1.05

A Reasonable Design has the FollowingFlux Densities: Fundamental airgap fluxdensity B̂≈ 0.85-0.95 T-maximum flux densityin the rotor yoke Bry ≈ 1.4 T-maximum fluxdensity in the stator yoke Bsy ≈ 1.4 T-maximumflux density in the stator teeth Bst ≈ 1.8 T (nearto saturation).

To prevent high temperatures and insulationproblems, the maximum current density Jshould be lower than 7 A/mm2. This value isrelevant for a motor without forced cooling.Depending on the way the motor is cooled,higher current densities can be possible.

DESIGNING OF SM BLDCThe finite element method is a numericalmethod for solving electromagnetic field

Figure 3: Angle of Rotor and Stator Pole

38

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

problems, which are too complex to be solvedusing analytical techniques, especially thoseinvolving non-linear material characteristics.FEA is a computer based numerical techniquefor calculating the parameters ofelectromagnetic devices. It can be used tocalculate the flux density, flux linkages,inductance, torque emf, etc. In finite elementmethod, the large electromagnetic device isbroken down into many small elements. Thebehaviour of an individual element can bedescribed with a relatively simple set ofequations. Just as the set of elements wouldbe joined together to build a whole device, theequations describing the behaviours of theindividual elements are joined into an extremelylarge set of equations that describe thebehaviour of the whole device.

The computer can solve this large set ofsimultaneous equations. From the solution, thecomputer extracts the behaviour of theindividual elements. The spatial variation of

magnetic potential throughout the motor isdescribed by a nonlinear partial differentialequation derived from Maxwell’s equations.Application of the finite element method tomachine design involves three stages,

It involves in the division of the rotor cross-section into a set of triangular elements (2-Dsolutions) or the division of the motor volumeinto bricks. Modern mesh generation is carriedout using the internal specialist draftingfacilities of the finite element software.Specialist mesh generation softwarecalculates the coordinates required to definethe motor geometry. The cross section isusually split up into regions representingdifferent materials such as current carryingconductors, air, steel, and magnets. Eachregion may define a different component usedin the construction of the motor for example,the shaft, rotor core, magnets, statorlamination, airgap, etc. In most cases it isbeneficial to split the components further intosmaller polygons along the lines of symmetry.

Stator lamination can be created byreflection followed by multiple rotational copiesof half a slot pitch. This procedure reduces theamount of data needed to 30 specify thegeometry, and reduces the chance of errors.When rotation of the motor rotor is to bemodelled, it is essential to define the airgapusing a sliding surface is to be defined and issplitted at least two layers. One of these layersis fixed to the rotor and one to the stator. Thenode spacing on the central sliding surface isset to a constant such that it is possible torotate the rotor by any multiple of this constant.Figure 4.1 shows a mesh in which the airgapregion is divided into two layers and the slidingsurface is central to the air gap.

Figure 4: Structure of Proposed Module

39

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

RESULTOnce the model has been completed the fieldsolution package is invoked and the programautomatically assembles the stiffness matrix,modifies it to include the boundary conditionand solves the system of N-equation in the N

unknown potential values. The solution of thediscretized partial differential equation uses aspecialized mathematical algorithm. Thealgorithm is often based on the minimizationof energy functional. The discretisationtransforms the partial differential equation into

Figure 5: Flux Density

Figure 6: Efficency Versus Speed

0.00 500.00 1000.00 1500.00 2000.00n (rpm)

0.00

20.00

40.00

60.00

80.00

(%

)

ANSOFT

Curve InfoEfficiency

Figure 7: Torque

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00Time [ms]

-1.25

0.00

1.25

2.50

3.75

5.00

6.25

Mo

vin

g1

.To

rq

ue

[N

ew

ton

Me

ter]

Maxwell2DDesign1Torque ANSOFT

Curve InfoMoving1.Torque

Setup1 : Transient

40

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

a large number of simultaneous non-linearalgebraic equations containing the unknownnode potentials. Iterative methods like Newtonraphson and conjugate-gradient proceduresare widely used. With linear elements thepotential is assumed to vary linearly betweenthe nodes and the flux density is constant withineach element. Current density is also assumedto be constant within each element associatedwith a winding.

The field solution is expressed in terms ofmagnetic vector or scalar potential, but thedesign engineer needs quantities such as fluxdensities, force and torque. The extraction ofthese quantities from the potential solution iscalled post-processing. Ab good interactivegraphics facility is important for that theessential information and b parameters canbeextracted from the large number of nodepotentials effectively and quickly.

CONCLUSIONThe paper identifies the features of the powerfulsimulation FEM software for the analysis of thefield. Analysis of various parameters like co-energy, flux linkages and torque plot revealsthat different types of configurations of FSMcan be modelled, and analysed for improvingthe performance. The simulated solutions haveshown that the model is advantageoushowever in practical situations various otherparameters have to be taken intoconsiderations.

REFERENCES1. IEEE Transaction on “New Class Magnet

and Brushless DC Motors” (2008-10).

2. Krishnan R (2001), “Switched ReluctanceMotor Drives Modelling, Simulation”, CRCPress, New York.

3. Miller T J E (1999), “Brushless PermanentMagnet and Reluctance Motor Drives”,Clarendon Press.

4. Nicola Bianchi (1999), “ElectricalMachine Analysis Using Finite Element”,Clarendon Press.

5. Sawhney A K (1998) “Electrical MachineDesigns”, Naveen Shahdra, Delhi.

Figure 8: Cogging Torque

0.00 125.00 250.00 375.00Electric Degree

-0.25

-0.13

0.00

0.13

0.25

(N

.m)

ANSOFT

Curve InfoCogging Torque

Figure 9: Winding Current

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00Time [ms]

-7.50

-5.00

-2.50

0.00

2.50

5.00

7.50

Y1

[A

]

Maxwell2DDesign1Winding Currents ANSOFT

Curve InfoCurrent(PhaseA)

Setup1 : TransientCurrent(PhaseB)

Setup1 : TransientCurrent(PhaseC)

Setup1 : TransientCurrent(PhaseC)_1

Setup1 : TransientCurrent(PhaseB)_1

Setup1 : Transient

41

Int. J. Elec&Electr.Eng&Telecoms. 2015 Ananthan N et al., 2015

6. Upadhyay K G and Mittle Arvindmittal VN (2009), “Design of Electrical Machines”,Nai Sarak, Delhi.

7. www.uread_books.com/ BLDC/Ansoft’sMaxwell