Title: Magnetic Circuits
1Magnetic Circuits
- Machine conversion Elec. lt-gt Mech.
- Mechanical Electrical Terminal Vars.
- Potential difference, current, power
- factor, freq, speed, torque or force
- relationships between these variables is affected
by behavior of magnetic Circuit - A useful concept is the Eq mag. CCT
2Equivalent Mag. Elect. CCTs
- Coil of N turns wound on plastic torus
- ?H.dl?J .dA (1.1)
- H(2?a)Ni A (1.2)
- HNi/2?aNi/l A/m (1.3)
- llength flux path
- H outside torus Zero
- Bavµo Hav T (1.4)
3Eq. Mag. Elec. CCTs ..
- FBav ?d2/4 Wb (1.5)
- ?Ni mmf (1.6)
- ?Ni?J. dA?H.dl A (1.7)
- ?H(2?a)B/µo (2?a)
- F/(?d2/4) .2?a/µo A (1.8)
- R?/F A/Wb (1.9)
- R2?a/µo(?d2/4) A/Wb (1.10)
- Reluctance proportional with l, and inversely
with its S
4Equivalent Magnetic CCT.
- The CCT is analog
- of resistor in an Elec. CCT.
- Acc. To Faraday's law
- eturn dF/dt V (1.12)
- eNdF/dtd?/dt (1.13)
- Where
- ?flux linkage
- ?N2µ?/2?a . ?d2/4 .i wb
5Magnetizing Curve
- B Saturates as H increases in Magnetic Core
- B-H curve named Magnetizing Curve
- Low B represents small R, while High B represents
large R - While in Electric cct. R is Indep. of I
- In Magnetic Circuit R is Dep. to B
- Cast Steel, Cast Iron, Silicon Sheet Steel
different B-H curves
6Magnetic Circuit with Air Gap
- Rotor separated by an air gap
- Cross section of a DC machine shown
- F in air gapf in poles
- Poles made from magnetic core
- To have equal f, air gap associate a higher mmf
- Having high value of B results in core saturation
- While air gap will not saturate as B-H linear
7Magnetic Circuit with Air Gapcontinued
- Core air gap shown by equivalent Reluctances
8Mag CCT with Air Gapcontinued
9Mag CCT with Air Gapcontinued
- lc average length of core
- Lg length of air gap
- Bcfc/Ac and Bgfg/Ag
- In the air gap flux lines slightly fringes
(extended as shown at corners) - Effect simulated by increasing equivalent
cross-sectional - For small air gaps fringing neglected
- AgAc BgBcf/Ac
10Example 1.1 of Mag. CCT with air gap
- Mag cct of primitive relay
- Coil N500 turns, lc360mm, g1.5 mm, B0.8 Tesla
- Core is from cast steel
- The current in coil
- Compute µ and µr of core
- If g0 what is current in coil for B0.8 Tesla in
core
11Solution of Exp 1.1
- Since gap is small fringing can be ignored
- From curve of fig1.7Bc0.8T, Hc510 At/m
- Fc (mmf)Hc lc 510x0.36184 At
- Fg(mmf)Hg 2lgBg/µ?x2xlg0.8/(4?x10-7)
- x2x1.5x10-31910 At
- FFcFg18419102094 At
- gtiF/N2094/500 4.19 amps
- µcBc/Hc0.8/5101.57x10-3
- µrµc/µ?1.57x10-3/4?x10-71250
- FHclc510x0.36184 At
- i184/5000.368 A
12Example 1.2
- In Exp 1.1 if coil current 4 amps and g1 mm,
find Bg? - Since B is not given directly B-H curve can not
be employed, two method can be employed - a-Load line
- NiHg lgHc lcBg/µ? lgHc lc rearranging ?
- BgBc-µ? lc/lg HcNiµ? / lg ym x c
- m-µ? lc/lg-4? 10-7x360/2-2.26x10-4, cNi/lg x
µ?1.256T - Load line intersect B-H curve at B1.08 T
- another method of building load line is as
follows - if all mmf acts on core (i.e. Hc0)
- Hc0gtBgNi/lg x µ?500x4x4?x10-7/2x10-3
- 1.256 Tesla
- if all mmf acts on the core (i.e. Bg0),
intersection with H axis is - HcNi/lc500x4/(36x10-2)5556 At/m
13Solution 1.2 continued
- Hc is intersection of load line on H axis
- b-Trial Error
- Assume a flux density, then Hc found from B-H
curve HgBg/µ? - Then FcHc lc , FgHg lg gtFFcFg
- Then iF/N if result in a different current,
another B is chosen continue to reach i4 amps - i.e. assuming B1.1 gt i4.08 amps
- And B1.08 gt i4 amps
14Example 1.3
- In mag cct of Fig E1.3 µr of core is 1200 ,
neglecting mag leakage fringing, dimensions in
cm, core with square cross section - Find air gap flux, air gap flux density, Hg
- F1N1xI1500x105000 At
- F2N2xI2500x105000 At
- µc1200µ?1200x4?x10-7
- Rbafelbafe/(µc Ac)3x52x10-2/(1200x4?x10-7x4x10
-4) - 2.58x106 At/Wb
15Fig of example E 1.3
16Solution Example 3
- Because of symmetry RbcdeRbafe
- Rglg/(µ0 Ag)5x10-3/(4?x2x2x10-4)
- 9.94x106 At/Wb
- Rbe(core)lbe(core)/(µc Ac)
- 51.5x10-2/(1200x4?10-7x4x10- 4)
- 0.82x106 At/Wb
- Loop equations are
- F1(RbafeRbeRg)f2(RbeRg)F1
- F1(RbeRg)f2(RbcdeRbeRg)F2
- Or f1(13.34x106)f2(10.76x106)5000
17Solution Ex.3 continued
- And f1(10.76x106)f2(13.34x106)5000
- gt f1f22.067x10-4 Wb
- air gap flux is fgf1f24.134x10-4 Wb
- Bgfg/Ag4.134x10-4/(4x10-4)1.034T
- HgBg/µ01.034/(4?10-7)0.822x106 At/m
18INDUCTANCE
- Coil wound on a mag. Core
- This coil equivalent ideal model is inductance
- Defined as flux linkage per ampere
- Flux linkage ?Nf Inductance L?/I
- LNf/iNBA/iNµHA/(Hl/N)N2/(l/µA)
- LN2/R
19Example 1-4
- in mag. Cct. Of fig 1.9 N400 turns
- lc50 cm , lg1.0 mm, AcAg15 cm2
- Realitive permeability of core µr3000
- i1.0 A
- Find f and B in air gap, also coil inductance
- Solution
- Rclc/(µr µ0Ac)50x10-2/(3000x4?10-7x15x10-4)8
8.42x103 At/Wb - Rglg/(µ0Ag)1x10-3/(4?10-7x15x10-4)530.515x10
3 At/Wb
20Ex. 1-4, solution continued
- FNi/(RcRg)
- 400x1.0/(88.42530.515)x103
- Bf/Ag
- 0.6463x10-3/(15x10-4)0.4309 T
- LN2/(RcRg) 4002/(88.42530.515)x103
- 258.52x10-3 H
- or
- L?/iNf/i
- 400x0.6463x10-3/1.0258.52x10-3 H
-
21Example 1.5
- Coil of fig 1.4 with N250 turns, wound on a
silicon steel sheet (toroidal) - Inner and outer radii are 20 25 cm
- Toroidal core has a circular cross section
- if coil current were 2.5 A find
- a- B at mean radius of toroid
- b- L of coil, assuming B within core uniform
equal to that of mean radius
22Solution of Ex. 1.5
- Mean radius1/2(2025)22.5 cm
- HNi/l250x2.5/(2?22.5x10-2)442.3 At/m
- From B-H curve of silicon steel sheet (fig1.7)
B1.225 T - Cross-sectional area A?(core radius)2
- ?(25-20)/22x10-4?6.25x10-4 m2
- FBA1.225x?6.25x10-424.04x10-4 Wb
- ?250x24.04x10-40.601 Wb. Turn
- L?/i0.601/2.50.2404 H24.04 mH
- Or can be calculated as follows
23Solution of Ex. 1.5continued
- µ of coreB/H1.225/442.3
- Rcorel/(µA)
- 2?22.5x10-2/(1.225/442.3)x?6.25x10- 4
- 2599.64x102 At/Wb
- LN2/R2502/2599.64x1020.2404 H
- 240.4 mH
24Hysteresis 1-characteristic
- Coil-core assembly of fig 1.12 a
- Assuming core initially is demagnetized
- The resulting B-H characteristic shown
- When H becomes zero still a residual flux density
is left Br - Hysteresis loops are very narrow, for such cores
if hysteresis effect can be ignored, then - B-H characteristic represented by
magnetization or saturation curve - Deltamax,a special ferromagnetic alloys has a
almost square B-H loop - Alloys with 50 iron, 50nickel has B-H loop of
fig 1.13 - Coil wound on Deltamax core used as a switch(very
low current when unsaturated and large current
when saturated
25Hysteresis 2- Losses
- As shown in hysteresis loops of fig 1.12c over a
cycle(where I of coil slowly varying) - energy flows to coil-core from source
- Howeverenergy flowing ingtenergy returns
- The net energy flow from source to coil is the
heat in core - The loss due to hysteresis called Hysteresis loss
- Size of hysteresis loop ? hysteresis loss
- To prove this assume coil has no resistance
26HysteresisLoss , continued
- Voltage e across the coil eN df/dt
- Energy transfer during t1 to t2 is
-
- VcoreA l, volume of core
- Power loss due to hysteresis in core
- PhVcore Wb f
- f freq. of variation of i
27HysteresisLoss , approximate relation
- Steinmetz of G.E. through large no. of experiment
for machine magnetic materials proposed a
relation - Area of B-H loop
- Bmax is the max flux density
- n varies from 1.5 to 2.5, K is a constant
- Therefore the hysteresis power loss
- PhKh (Bmax)n f
- Kh a constant depends on ferromagnetic material
and core volume
28EDDY CURRENT LOSS
- Another power loss of mag. Core is due to rapid
variation of B - In cross section shown voltage induced and ie
passes and due to resistance of core - Peie2 R
- this loss can be reduced
- a- using high resistive core material, few
Si - b- using a laminated core
- Eddy current loss PeKeBmax2 f2
- Ke constant depends on material lamination
thickness which varies from 0.01 to 0.5 mm
29CORE LOSS
- PcPhPe
- If current I varies slowly eddy loss negligible,
separation in fig 1.16 - Total core loss determined from dynamic B-H loop
-
- Using a wattmeter core loss can be measured
- It is not easy to know what portion is eddy
hysteresis
30SINUSOIDAL EXCITATION
- Ac electric machine work with ac supply have
sinusoidal voltage flux fig1.17 - F(t)fmax sin?t
- e(t)N df/dtN fmax ? cos?tEmax cos?t
- ErmsEmax/v2N?fmax /v24.44Nffmax
31Example 1.6
- A 1 phase 120 V, 60 Hz supply and coil with 200
turns on core with - core length100 cm, cross-section area20
cm2, µr2500 - Find a-expression for B in core
- b-expression for I in coil
- Fmax120/4.44x200x600.002253 Wb
- Bmax0.002253/20x10-41.1265 T
- B1.1265 sin2?60t
- Hmax1.1265/2400x4?10-7358.575 At/m
- imax Hl/N358.575x100x10-2/2001.7928 A,
i1.79sin?t
32Example 1.7
- A square wave voltage E100 V f60 Hz applied
coil on a closed iron core, N500 - Cross section area 0.001 mm2, assume coil no
resistance - a- max value of flux sketch V f vs time
- b- max value of E if Blt1.2 Tesla
33Ex 1.7 - Solution
- e N df/dt gt N.?fE.?t
- E constant gt 500(2fmax)Ex1/120
- Fmax100/(100x120)Wb0.833x10-3 Wb
- From fig. gt Bmax1.2 T
- FmaxBmax x A1.2 x 0.0011.2 x 10-3 Wb
- N(2fmax)E x 1/120
- E120x500x2x1.2x10-3144 V
34Exciting Current
- Current which establish the flux in the core
- The term If if B-H nonlinear, non-sinusoid
- At first consider not to have Hysteresis
- B-H curve ? f-i curve (or the rescaled one)
- Knowing sine shape flux, exciting current
waveform by help of f-i curve obtained - The current non-sinusoidal, if1 lags V 90
- Or no loss (since Hysteresis neglected)
35Considering Hysteresisin Exciting Current
determination
- Now if determined from multi-valued f-i
- Exciting current nonsinusoid nonsymmetric
- It can split 2 component ic in phase with e
rep. loss, im in ph. With f symmetric -
36PERMANENT MAGNET
- Permanent magnet large B-H loop high Br
- Also high Hc , material iron, cobalt, Nickel and
called hard iron vs soft iron - P.M. magnetization
- Assume magnet material
- Initially demag.
- large mmf applied
- removed, remain at Br
- If reverse H1 applied
- To hard iron move to b
- By removing reapply recoil
- Through bc line parallel with xay,
- µrec, for alnico 3-5 µ0
37Approximation Designof Permanent Magnets
- Per. Mag. of last fig. magnetized to Br-point a
- Removing keeper air gap activated
- Assuming no fringing no leakage, also if no mmf
- required for soft iron
- Amperes circuit law
- Hm lm Hg lg0
- Hm-lg/lm Hg
- Continuity of flux requires fBmAmBgAg
- Also Bgµ? Hg gtBmµ? Ag/Am lm/lg Hm
- This equation a straight line, intersection with
demag. Curve, at b, is operating point
38Design of Permanent Magnetscontinued
- Demagnetizing curve shear line
- ? intersection shear line- demagn. Curve result
in operating values of B and H hard iron, keeper
removed - ? if keeper inserted, o.p. moves up recoil line
bc. - ? this analysis gt O.P. of a
- permanent magnet with
- air gap, determined by
- demag. Portion B-H loop
- Dimensions of magnet and air gap
-
39Permanent Magnet DesignContinued
- From last Eqs gt volume of P.M.
- VmAmlmBgAg/Bm x Hglg/Hm
- Bg2
Vg/(µ0BmHm) - Where VgAglg
- To have Bg in gap of volume Vg,
- a min volume of hard iron required if final
o.p. located such that BmHm is a max - BmHm known as energy product of hard iron
40Permanent Magnet Materials
- Alinco (Alloys of Al, Ni, Co)used for for P.M.
- Having a Demagnetization curve fig22
- Br 1.25 Tesla , and Hc -50 kA/mF
- From 1950 Ferrite P.M. also are used, with lower
B (0.35 T) and higher Coercive force(-350 kA/m) - Since 1960 rare-earth P.M. material
developed having high residual B as alinco type,
with a greater coercivity than ferrite - Demagnetization for samaritum-cobalt and
neodymium-boron magnet fig 1.24
41Example 1.8 P.M.
- P.M. in fig 1.21 made of Alinco 5
- B of 0.8 to be stablished in air gap, keeper
removed - Air gap Ag2.5 sq. cm, lg0.4 cm
- O.p. correspond to point with max HmBm
- Or Bm0.95 T , Hm-42 kA/m
- Determine lm, Am of P.M.
42Solution to example 1.8
- Lmlg/Hm Hglg Bg/(Hmµ0) 0.4x0.01x0.8/(42x1000x4?
x0.0000001) - 0.0606 m6.06 cm
- From flux continuity AmBg Ag/Bm
- 0.8x2.5x0.0001/0.952.105 sq. cm
-