2
3
4
1.1INTRODUCTION ... 9
1.2PRESENT AND FUTURE OF EV’S ... 10
1.3ENGINEER PHILOSOPHY ... 11
1.4FINAL CONSIDERATION ... 13
2.1INTRODUCTION ... 15
2.2ALGEBRAIC EQUATIONS,ORDINARY DIFFERENTIAL EQUATIONS,PARTIAL DIFFERENTIAL EQUATIONS, AND THE LAWS OF PHYSICS ... 18
2.3THE FINITE ELEMENT METHOD FROM THE WEAK FORMULATION:BASIS FUNCTIONS AND TEST FUNCTIONS ... 21
2.4TIME-DEPENDENT PROBLEMS ... 28
2.5DIFFERENT ELEMENTS ... 30
2.6ERROR ESTIMATION ... 31
2.7MESH CONVERGENCE ... 32
3.1INTRODUCTION ... 35
3.2MAGNETIC MATERIALS ... 35
3.3SOFT MAGNETIC MATERIALS... 36
3.4HARD MAGNETIC MATERIALS ... 44
3.5CHARACTERISTICS COMPARISON OF SOFT AND HARD MAGNETIC MATERIALS ... 48
3.6JILES-ATHERTON MODEL ... 49
3.7CONDUCTORS ... 51
3.8MODEL MATERIALS ... 54
4.1INTRODUCTION ... 57
4.2KINETIC ENERGY HARVESTING ... 58
4.3TRANSDUCTOR GENERATOR ... 66
4.4ENERGY HARVESTING BRAKES ... 72
4.5ENERGY HARVESTING DAMPERS ... 75
4.6LINEAR MACHINE DESIGN ... 78
4.7ECDOVERVIEW ... 83
4.8FINAL CONSIDERATION ... 83
5.1INTRODUCTION ... 85
5.2EDDY CURRENT SPRING-DAMPER DESIGN... 85
5.3EDDY CURRENT DAMPER DESIGN ... 92
5.4EDDY CURRENT REGENERATIVE DAMPER ... 97
5.5FINAL CONSIDERATIONS ... 100
6.1INTRODUCTION ... 102
6.2PROJECT ... 102
6.7RESULTS ... 108
6.8OPTIMIZED MODEL SIMULATION ... 116
5
7.1FUTURE WORKS ... 122
A.1AIR POLLUTION ... 126
A.2GLOBAL WARMING ... 126
B.1MAXWELL’S EQUATIONS ... 128
B.2POTENTIALS ... 129
B.3MAGNETIC FIELD ANALYSIS ... 131
B.4THE CURRENT MODEL (MAGNETIC FLUX CALCULATION) ... 132
C.1INTRODUCTION ... 134
C.2THE B-HCURVE,PERMEABILITY, AND DIFFERENTIAL PERMEABILITY ... 134
C.3HOW THE SMOOTHNESS OF THE B-HCURVE AFFECTS THE SIMULATION ... 137
6 𝐻
𝐵⃗
𝜇
𝜇
𝜇
𝑎𝑀
𝐵⃗
𝑡𝑜𝑡𝐵⃗
𝜒
𝐵⃗
𝑠𝐵⃗
𝑟𝐻
𝑐7 𝐵⃗𝐻
𝑚𝑎𝑥𝐵⃗
𝑥𝐵⃗
𝑦𝐵⃗
𝑧𝜀
𝜙
𝑏𝐸
𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑟𝐸
𝑚𝑜𝑡𝑖𝑜nal𝐽
𝜎
𝑣
𝐹
𝑧𝜏
𝜏
𝑚8
𝑜
𝐸 𝑘
𝐾 𝑘
𝛿
𝜔
𝜌
𝜔
9
Figure 1: Growth of population and vehicles 1
10
11
Figure 2: Development trends of EV's and HEV's (courtesy of EVAA)
12
13
Figure 3: Interactions among EV subsystems 1
14
15 𝑢 ≈ 𝑢
h𝑢
h= ∑ 𝑢
𝑖𝜑
𝑖𝑖
16
Figure 4: The function u (solid blue line) is approximated with uh (dashed red line), which is a linear combination of linear basis functions (ψi is represented by the solid black lines). The coefficients are denoted by u0 through u7 4
17
Figure 5: The function u (solid blue line) is approximated with uh (dashed red line), which is a linear combination of linear basis functions (ψi is represented by the solid black lines). The coefficients are denoted by u0 through u7.4
18
Figure 6: Finite element discretization, stresses, and deformations of a wheel rim in a structural analysis.4
partial differential equations
19 ρ𝐶
𝑝𝑑𝑇
𝑑𝑡 = 𝑔(𝑇, 𝑡)
𝑇 = 𝑓(𝑡)
4
ρ𝐶
𝑝𝛿𝑇
𝛿𝑡 + ∇ ∙ 𝒒 = 𝑔(𝑇, 𝑡, 𝑥)
20
∇ ∙ 𝒒 = 𝜕𝑞
𝑥𝜕𝑥 + 𝜕𝑞
𝑦𝜕𝑦 + 𝜕𝑞
𝑧𝜕𝑧
𝑞 = −𝑘∇𝑇 → 𝑞 = (−𝑘 𝜕𝑇
𝜕𝑥 , −𝑘 𝜕𝑇
𝜕𝑦 − 𝑘 𝜕𝑇
𝜕𝑧 )
𝜌𝐶
𝑝𝛿𝑇
𝛿𝑡 + ∇ ∙ (−𝑘∇𝑇) = 𝑔(𝑇, 𝑡, 𝑥)
21
𝑇 = 𝑓(𝑡, 𝑥)
22
∇ ∙ (−𝑘∇𝑇) = 𝑔(𝑇, 𝑥) 𝑖𝑛 Ω
𝑇 = 𝑇
0𝑜𝑛 δΩ
1(−𝑘∇𝑇) ∙ 𝒏 = ℎ(𝑇 − 𝑇
𝑎𝑚𝑏) 𝑜𝑛 δΩ
2(−𝑘∇𝑇) ∙ 𝒏 = 0 𝑜𝑛 δΩ
3Figure 7: The domain equation and boundary conditions for a mathematical model of a heat sink4
23
∫ ∇ ∙ (−𝑘∇𝑇)𝜑𝑑𝑉
Ω
= ∫ 𝑔𝜑𝑑𝑉
Ω
ϵ ϵ
24
∫ 𝑘∇𝑇 ∙ ∇𝜑𝑑𝑉 +
Ω
∫ (−𝑘∇𝑇) ∙ 𝒏𝜑𝑑𝑆 = ∫ g𝜑𝑑𝑉
Ω δΩ
25
𝑇
ℎ(𝑥) = ∑ 𝑇
𝑖𝜓
𝑖(𝑥)
𝑖
∑ 𝑇
𝑖∫ 𝑘∇𝜓
𝑖𝑖 Ω
∙ ∇𝜓
𝑗𝑑𝑉 + ∑ ∫ (−𝑘𝑇
𝑖∇𝜓
𝑖) ∙ 𝑛𝜓
𝑗𝑑𝑆 = ∫ 𝑔 (∑ 𝑇
𝑖𝜓
𝑖𝑖
)
Ω 𝑖 𝛿Ω
𝜓
𝑖Figure 8: Finite element discretization of the heat sink model from the earlier figure 4
26
𝐴𝑇
ℎ= 𝑏
27
Figure 9: Tent-shaped linear basis functions that have a value of 1 at the corresponding node and zero on all other nodes. Two base functions that share an element have a basis function overlap 4
28
Figure 10: Two basis functions that share one element vertex but do not overlap in a 2D domain 4
ψ
𝑗29 𝑇
𝑖δ𝑇
𝑖𝛿𝑡 ≈ 𝑇
𝑖,𝑡+∆𝑡− 𝑇
𝑖,𝑡∆𝑡
𝑇
𝑖,𝑡𝑡 + ∆𝑡
𝑇
𝑖,𝑡𝑡 + ∆𝑡(𝑇
𝑖,𝑡+∆𝑡)
30
Figure 11: Node placement and geometry for 2D and 3D linear elements 4
31
Figure 12: Second-order elements. If the gray nodes are removed, we get the corresponding serendipity elements.
The black, white, and gray nodes are all present in the Lagrangian elements 4
𝑒 = 𝑢 − 𝑢
ℎ𝑂(ℎ
𝛼) 𝛼
32 𝑒 = 𝑢
ℎ1− 𝑢
ℎ𝑢
ℎ𝑢
ℎ1Figure 13: Benchmark model of an elliptic membrane. The load is applied at the outer edge while symmetry is assumed at the edges positioned along the x- and y-axis (roller support) 4
33
Figure 14: Rectangular elements used for the quadratic base functions 4
34
Figure 15: Relative error in σx at the evaluation point in the previous figure for different elements and element sizes (element size = h). Quad refers to rectangular elements, which can be linear or with quadratic basis functions 4
35
Figure 16: Magnetic domains before and after alignment with the external magnetic field 5
36 𝜇
𝑟= 𝜇
𝜇
0𝜇 𝜇
0= 4𝜋 × 10
−7 𝐻𝑚
𝜇
𝑟𝐵⃗
𝑡𝑜𝑡𝐵⃗
𝑡𝑜𝑡= 𝐵⃗
0+ 𝐵⃗
𝑟= 𝜇
0𝐻 + 𝜇
0𝑀 = 𝜇
0(𝐻 + 𝑀)
𝐵⃗
0𝐵⃗
𝑟37
𝐻
𝑐𝜇
𝑟< 1
Figure 17: Field lines distribution of diamagnetic materials placed in a uniform magnetic field H 6
𝜇
𝑟= 1
38
Figure 18: Field lines distribution of paramagnetic materials placed in a uniform magnetic field H 6
Figure 19: Field lines distribution of ferromagnetic materials placed in a uniform magnetic field H 6
𝜇
𝑟≫ 1
39
Figure 20: Ferromagnetic material in a magnetic field. Flux lines pass through the mate- rial and the material is attracted to the source of the field 5
40
Figure 21: Magnetic circuit used to obtain the magnetic characteristics of a ferromagnetic sample 5
𝐻 = 𝑁𝐼/𝑙 𝐵⃗ = Φ/𝑆 𝐼
Φ
𝐻 𝐵⃗ 𝐼 Φ
Figure 22: A typical B(H) for a ferromagnetic material 5
𝐵⃗(𝐻)
𝛷(𝐼)
41
Figure 23: Permeability curve 8
µ
042
Figure 24: Hysteresis Loop curve 8
43 𝐵⃗ ≠ 0
−𝐵⃗
𝑟Figure 25: Comparison between hysteresis loop of a hard material (a) and of a soft material (b) 7
44
𝐻
𝑐Figure 26: Demagnetization curve, recoil loop, energy of a PM, and recoil magnetic 11
45
46 𝐵⃗
𝑔= 𝜇
0𝐻
ℎ= √ 𝜇
0𝜎
𝑙𝑀𝑘
𝑠𝑎𝑡−1𝑉
𝑀𝑉
𝑔𝐵⃗
𝑀𝐻
𝑀𝐵⃗
𝑟𝐻
𝑐𝐵⃗
𝑟= 𝐵⃗
𝑟20[1 + 𝛼
𝐵100 (𝜗
𝑃𝑀− 20)]
𝐻
𝑐= 𝐻
𝑐20[1 + 𝛼
𝐻100 (𝜗
𝑃𝑀− 20)]
𝜗
𝑃𝑀𝐵⃗
𝑟20𝐻
𝑐 20𝛼
𝐵< 0 𝛼
𝐻< 0
𝐵⃗
𝑟𝐻
𝑐47
Figure 27: Descending branches of hysteresis loops determined at various temperatures 13
48
Figure 28: BH curve comparison on Soft and Hard magnetic materials 8
𝐵⃗
𝑟𝐻
𝑐49
50 𝑀
𝑟𝑒𝑣𝑀
𝑖𝑟𝑟𝑑𝑀
𝑖𝑟𝑟𝑑𝐻 = 𝑀
𝑎𝑛− 𝑀
𝑖𝑟𝑟𝑘𝛿 − 𝛼(𝑀
𝑎𝑛− 𝑀
𝑖𝑟𝑟)
𝑀
𝑟𝑒𝑣= 𝑐(𝑀
𝑎𝑛− 𝑀
𝑖𝑟𝑟)
𝑀
𝑎𝑛(𝐻) = 𝑀
𝑆[coth ( 𝐻 + 𝛼𝑀
𝛼 ) − 𝛼
𝐻 + 𝛼𝑀 ]
𝑎, 𝛼, 𝑐, 𝑘 𝑀
𝑠𝛿
±1 𝑓𝑜𝑟 𝑑𝐻/𝑑𝑡 ≷ 0
51
52
◦
◦
Table 3.1: Electric resistivity, conductivity, and temperature coefficient at 20◦C 12
° °
𝑅(𝜗)𝑅
20[1 + 𝛼
20(𝜃 − 20)]
𝜌(𝜗) = 𝜌
20[1 + 𝛼
20(𝜃 − 20)]
𝜎(𝜗) = 𝜎
201 + 𝛼
20(𝜃 − 20)
𝑅
20, 𝜌
20, 𝜎
20, 𝛼
20◦
53
° β
𝑅(𝜗) = 𝑅
20[1 + 𝛼
20(𝜗 − 20) + 𝛽
20(𝜗 − 20)
2]
°
Figure 29: Variation of resistivity ρ of metals with temperature: 1,2,3| various mild steels, 4|stainless and acid resistant steels, 5|brass, 6|aluminium, 7|copper. 12
54
Figure 30: HIGH INDUCTION MAGNETIZATION CURVE FOR ARMCO® PURE IRON
55
Figure 31: BH-curve of NdFeB N-50H magnet 16
56
57
Figure 32: Energy consumption distribution of an ordinary automobile in the urban transport
1758
59
Figure 33: Generic model of kinetic energy harvesters
11𝑦(𝑡) =
𝑌 sin 𝜔𝑡,
60 𝑚 ∙ 𝑑
2𝑧(𝑡)
𝑑𝑡
2+ 𝑏 ∙ 𝑑𝑧(𝑡)
𝑑𝑡 + 𝑘 ∙ 𝑧(𝑡) = −𝑚 ∙ 𝑑
2𝑦(𝑡) 𝑑𝑡
2𝑚 ∙ 𝑠
2∙ 𝑧(𝑠) + 𝑏 ∙ 𝑠 ∙ 𝑧(𝑠) + 𝑘 ∙ 𝑠 ∙ 𝑧(𝑠) = −𝑚 ∙ 𝑎(𝑠)
𝑎(𝑡) = 𝑑
2𝑦(𝑡) 𝑑𝑡
2𝑧(𝑠)
𝑎(𝑠) = 1
𝑠
2+ 𝑏 𝑚 𝑠 +
𝑘 𝑚
= 1
𝑠
2+ 𝜔
𝑟𝑄 𝑠 + 𝜔
𝑟2𝑄 =
√𝑘𝑚𝑏
𝜔
𝑟= √𝑘/𝑚
−𝑚 ∙ 𝑎(𝑠) = 𝑠 ∙ 𝑍(𝑠) (𝑚𝑠 + 𝑏 + 𝑘
𝑠 )
61
−𝐼(𝑠) = 𝐸(𝑠) (𝑠𝐶 + 1 𝑅 + 1
𝑠𝐿 )
𝐼(𝑠) = 𝑚 ∙ 𝑎(𝑠), 𝐸(𝑠) = 𝑠 · 𝑍(𝑠), 𝐶 = 𝑚, 𝑅 =
1𝑏
, 𝐿 =
1𝑘
Figure 34: Equivalent circuit of a kinetic energy harvester11
𝜁
𝑇= 𝑏
2𝑚𝜔
𝑟= 𝑏
𝑚+ 𝑏
𝑒2𝑚𝜔
𝑟= 𝜁
𝑚+ 𝜁
𝑒𝜁
𝑚=
𝑏𝑚2𝑚𝜔𝑟
𝜁
𝑒=
𝑏𝑒2𝑚𝜔𝑟
𝑄
𝑇= 1
2𝜁
𝑇62 1
𝑄
𝑇= 1 𝑄
𝑂𝐶+ 1
𝑄
𝑒𝑄
𝑂𝐶=
12𝜁𝑚
𝑄
𝑒=
1 2𝜁𝑒
𝑦(𝑡) = sin 𝜔𝑡
𝑧(𝑡) = 𝑚𝜔
2𝑌
𝑘 − 𝑚𝜔
2+ 𝑗𝜔𝑏 ∙ sin 𝜔𝑡
𝑧(𝑡) = 𝜔
2√(𝜔
𝑟2− 𝜔
2)
2+ ( 𝑏𝜔 𝑚 )
2
∙ 𝑌 sin(𝜔𝑡 + 𝜑)
𝜑 = tan
−1( 𝑏𝜔
𝑘 − 𝜔
2𝑚 )
63 𝑃 = 𝑏 ( 𝑑𝑧(𝑡)
𝑑𝑡 )
2
𝑃(𝜔) =
𝑚𝜁
𝑇𝑌
2( 𝜔 𝜔
𝑟)
3𝜔
3[1 − ( 𝜔
𝜔
𝑟)
2]
2
+ (2𝜁
𝑇𝜔 𝜔
𝑟)
2𝜔 = 𝜔
𝑟𝑃 = 𝑚𝑌
2𝜔
𝑟34𝜁
𝑇𝑃 = 𝑚𝑌
2𝜔
𝑟34(𝜁
𝑚+ 𝜁
𝑒)
𝑃
𝑒𝑃
𝑚𝑃
𝑒𝑃
𝑚𝑃
𝑒= 𝜁
𝑒𝑚𝑌
2𝜔
𝑟34(𝜁
𝑚+ 𝜁
𝑒)
64 𝑃
𝑚= 𝜁
𝑚𝑚𝑌
2𝜔
𝑟34(𝜁
𝑚+ 𝜁
𝑒)
𝜁
𝑒= 𝜁
𝑚𝑃
𝑒= 𝑃
2 = 𝑚𝑌
2𝜔
𝑟316𝜁
𝑚𝑎 = 𝑌𝜔
2𝑄
𝑜𝑐=
12𝜁𝑚
𝑃
𝑒= 𝑚𝑎
28𝜔
𝑟65
Figure 35: Power spectrum of a Kinetic energy harvester with various Q-factor
1166
67
Figure 36: Piezoelectric generators: (a)33mode and(b) 31 mode
1168
Figure 37: Electrostatic generators: (a) in-plane overlap; (b) in-plane gap closing; and (c) out-of-
plane gap closing
1169
Figure 38: Electromagnetic generators
11𝑒. 𝑚. 𝑓. = −𝑁 ∙ 𝑙 ∙ 𝐵⃗ ∙ 𝑑𝑧
𝑑𝑡
70 𝑒. 𝑚. 𝑓. = −𝑁 ∙ 𝑆 ∙ 𝑑𝐵⃗
𝑑𝑧 ∙ 𝑑𝑧 𝑑𝑡
𝑑𝐵⃗/𝑑𝑧
𝑒. 𝑚. 𝑓. = 𝑘 ∙ 𝑑𝑧 𝑑𝑡
−𝑁 · 𝑙 · 𝐵⃗ −𝑁 ∙ 𝑆 ∙
𝑑𝐵𝑑𝑧
71
Figure 39: Comparison of different transduction mechanism of kinetic energy harvesters
1172
73
Figure 40: Total traction energy and energies consumed by drags and braking in an FTP 75 urban driving cycle
23Figure 41: Maximum Speed, Average Speed, Total Traction Energy, and Energies Consumed by
Drags and Braking per 100 km Traveling Distance in Different Drive Cycles
2374
75
Figure 42: A typical road vehicle front suspension system
1876
77
78
79
Figure 43: Basic Geometric Arrangements of Linear Electric Machines. Left: Planar, Right: Tubular
80
Figure 44: Permanent-Magnet Linear Machine Configurations.
From left to right: Moving Coil, Moving Iron, Moving Magnet
𝐶
𝑋𝑖81
Figure 45: Different PM configurations in the mover. (a) Axially, (b) axially and radially,(c) radially (inward and outward), and (d) radially (outward) magnetized PMs in the mover 45
82
Figure 46: Tubular Machine Slot-less (left) and Slotted (right) Stator Illustrations.
83
84
85
86
Figure 47: Schematic view of the proposed eddy current spring–damper
87
𝐽
𝑚= ∇ × 𝑀 𝑗
𝑚= 𝑀 × 𝑛̂
𝑀 𝑛̂
𝑀 = 𝑀𝑧̂)
𝐽
𝑚= 𝑀𝜑̂
𝐽
𝑚= 0
𝜑
(𝑟, 𝑧)
𝐵⃗
𝑟= 𝜇
0𝐼
2𝜋𝐿 ∫ (𝑧 − 𝑧
′)
𝑟[(𝑅 + 𝑟)
2+ (𝑧 − 𝑧
′)
2]
1/2[−𝐾(𝑘) + 𝑅
2+ 𝑟
2+ (𝑧 − 𝑧
′)
2(𝑅 − 𝑟)
2+ (𝑧 − 𝑧
′)
2] 𝑑𝑧′
𝐿/2
−𝐿/2
𝐵⃗
𝑧= 𝜇
0𝐼
2𝜋𝐿 ∫ 1
𝑟[(𝑅 + 𝑟)
2+ (𝑧 − 𝑧
′)
2]
1/2[−𝐾(𝑘) + 𝑅
2− 𝑟
2− (𝑧 − 𝑧
′)
2(𝑅 − 𝑟)
2+ (𝑧 − 𝑧
′)
2] 𝑑𝑧′
𝐿/2
−𝐿/2
𝐾(𝑘) 𝐸(𝑘)
88
𝐾(𝑘) = ∫ 𝑑𝜃
√1 − 𝑘
2𝑠𝑖𝑛
2𝜃
𝜋/2 0
𝐸(𝑘) = ∫ √1 − 𝑘
2𝑠𝑖𝑛
2𝜃
𝜋/2 0
𝑑𝜃
𝑘
2= 4𝑅𝑟[(𝑅 + 𝑟
2) + (𝑧 − 𝑧
′)
2]
−1𝐼 = 𝑀𝐿
𝐼
1𝑑𝐿
1𝐼
2𝑑𝐿
2𝑑𝐹
21= 𝜇
0𝐼
2𝑑𝐿
24𝜋 × [ 𝐼
1𝑑𝐿
1× 𝑅
21𝑅
213]
89
𝑅
21𝑅
1𝑅
2𝐹
21= 𝜇
04𝜋 ∫ 𝐼
2𝑑𝐿
2× ∫ 𝐼
1𝑑𝐿
1× 𝑅
21𝑅
213𝑐1
𝑐2
(5.6)
Figure 48: Schematic of two current-carrying loops, illustrating the variables used in Eq. 58
𝐹 = ∫ 𝐼
𝑐
𝑑𝐿 × 𝐵⃗
𝑅
1= 𝑅
2= 𝑅 𝐹
𝑧(𝑧
𝑑)
90 𝐹
𝑧(𝑧
𝑑) = 𝜇
0𝐼
1𝐼
2𝑅
24𝜋 ∫ ∫ −𝑧
𝑑cos(𝜃
2− 𝜃
1) 𝑑𝜃
1𝑑𝜃
2(2𝑅
2− 2𝑅
2𝑐𝑜𝑠(𝜃
2− 𝜃
1) + 𝑧
𝑑2)
3/22𝜋 0 2𝜋 0
𝑧
𝑑𝐹
𝑧= 𝜇
0𝐼
1𝐼
2𝑅
24𝜋𝐿
2∫ ∫ ∫ ∫ −[𝑧
𝑑− 𝑧
′− 𝑧′′] 𝑐𝑜𝑠(𝜃
2− 𝜃
1) 𝑑𝜃
1𝑑𝜃
2𝑑𝑧′𝑑𝑑𝑧′′
(2𝑅
2− 2𝑅
2𝑐𝑜𝑠(𝜃
2− 𝜃
1) + [𝑧
𝑑− 𝑧
′− 𝑧′′]
2)
3/22𝜋 0 2𝜋 0
−𝑔/2
−𝐿−𝑔/2 𝑔/2+𝐿
𝑔/2
Figure 49: Geometric definition of the eddy current spring–damper
91 𝑉 = 𝑉
𝑡𝑟𝑎𝑛𝑠+ 𝑉
𝑚𝑜𝑡𝑖𝑜𝑛= − ∫ 𝛿𝑩
𝛿𝑡 ∙ 𝑑𝒔 + ∫ (𝜈 × 𝑩) ∙ 𝑑𝒍
𝑠 𝑠
𝜈 𝜈
𝒗 × 𝑩
𝑧= 0
𝑱 = 𝜎(𝑣 × 𝑩)
𝑭 = ∫ 𝑱 ×
𝛤
𝑩𝑑Γ = −𝒌 ̂𝜎𝛿𝑣 ∫ ∫ 𝑟𝐵⃗
𝑟2(𝑟, 𝑧
0)𝑑𝑟𝑑𝜃
𝑟𝑜𝑢𝑡 𝑟𝑖𝑛 2𝜋 0
Γ
𝐵⃗
𝑟,𝑡𝑜𝑡𝑎𝑙= 𝐵⃗
𝑟(𝑧
0) + 𝐵⃗
𝑟(𝑧
1)
92
𝑧
1= | 𝐿 + 𝑔 − 𝑧
0𝑖𝑓 𝑧
𝑜< 𝐿 + 𝑔
−𝐿 − 𝑔 + 𝑧
0𝑖𝑓 𝑧
0> 𝐿 + 𝑔
93
Figure 50: Schematic view of the proposed ECD
94
Figure 51: Configuration of the proposed ECD
𝐸 = 𝐸
𝑡𝑟𝑎𝑛𝑠+ 𝐸
𝑚𝑜𝑡𝑖𝑜𝑛𝑎𝑙= ∫ (𝑣 × 𝑩) ∙ 𝑑𝑙
𝑐
𝐽 = 𝜎(𝑣 × 𝑩)
𝜎
𝐹 = ∫ 𝐽 × 𝐵⃗𝑑Γ = −𝑘̂𝜎(𝜏 − 𝜏
𝑚)𝑣
𝑧× ∫ ∫ 𝑟𝐵⃗
𝑟2(𝑟, 𝑧
0)𝑑𝑟𝑑𝜃
𝑟𝑜𝑢𝑡 𝑟𝑖𝑛 2𝜋 0 Γ
Γ 𝑟
𝑖𝑛𝑟
𝑜𝑢𝑡𝜏 𝜏
𝑚95 𝐶 = 𝜎 ∫ 𝐵⃗
𝑟2𝑑Γ
Γ
Figure 52: Schematic view of two PMs with like poles in close proximity
𝐵⃗
𝑟(𝑟, 𝑧)|
𝑅,𝜏𝑚= 𝜇
0𝐼
2𝜋𝜏
𝑚∫ (𝑧 − 𝑧′)
[(𝑅 + 𝑟)
2+ (𝑧 − 𝑧
′)
2]
1/2𝜏𝑚/2
−𝜏𝑚/2
× [−𝐾(𝑘) + 𝑅
2+ 𝑟
2+ (𝑧 − 𝑧
′)
2(𝑅 − 𝑟)
2+ (𝑧 − 𝑧
′)
2𝐸(𝑘)] 𝑑𝑧′
𝐾(𝑘) 𝐸(𝑘)
𝐾(𝑘) = ∫ 𝑑𝜃
√1 − 𝑘
2sin
2𝜃
2𝜋 0
96
𝐸(𝑘) = ∫ √1 − 𝑘
2𝑠𝑖𝑛
2𝜃
2𝜋 0
𝑑𝜃
𝑘
2= 4𝑅𝑟[(𝑅 + 𝑟)
2+ (𝑧 − 𝑧
′)
2]
−1= 𝑀𝜏
𝑚𝐵⃗
𝑟𝐵⃗
𝑟= 2(𝐵⃗
𝑟(𝑟, 𝑧)|
𝑙𝑚+𝑠,𝜏𝑚− 𝐵⃗
𝑟(𝑟, 𝑧)|
𝑠,𝜏𝑚97
Figure 53: Diagram of the linear regenerative electromagnetic shock absorber.60
Figure 54: Equivalent circuit model of the energy harvester.61
98
𝑅
𝐿2Figure 55: Linear regenerative electromagnetic shock absorber dimensions and parameters60
99 𝑃 = 𝑉𝐼
𝑅
𝑙𝑜𝑎𝑑𝑒. 𝑚. 𝑓.
𝑙 (𝑚) 𝐵⃗
𝑟𝑣
𝑉 = 𝐵⃗
𝑟𝑣 𝑙
𝐼 = 𝑉
𝑅 = 𝜎𝐵⃗
𝑟𝑣
𝑧𝐴
𝑤𝜎 𝐵⃗
𝑟𝑣
𝑧𝐴
𝑤𝑃 = 𝑉𝐼 = 𝐵⃗
𝑟2𝑣
𝑧2𝜎𝑙𝐴
𝑤𝐵⃗
𝑟𝐵⃗
𝑟100 𝑉 = 𝜋𝑁𝐵⃗
𝑟𝑣
𝑧𝐷
𝑐𝑃 = 𝜋𝑁𝜎𝐵⃗
𝑟2𝑣
𝑧2𝐷
𝑐𝐴
𝑤𝐷
𝑐𝐴
𝑊𝑁 = 𝐹 ∙
𝐴𝐶𝐴𝑊
𝐴
𝐶𝐴
𝑊𝐵⃗
𝑟𝐷
𝑐𝐴
𝑐𝐵⃗
𝑟101
102
103
Figure 56: Initial model geometry
𝑅
𝑂𝑅𝑜𝑑
𝑟𝐷
𝜏=
104
𝑅
𝑂𝑅𝑜𝑑
𝑟𝐵⃗𝐼
𝑡)
𝐴
𝑔= 0.5 𝑚𝑚
𝜏
𝐷
𝜏= 2 ∙ 𝜏
𝜏 𝑀
ℎ𝐼𝑃
ℎ𝜏
105
𝜏
𝑚/𝜏 = 0.55
𝐶
ℎ= 𝐷
𝜏/3
𝑅𝑜𝑑
ℎ𝑆𝑡
ℎ𝑅𝑜𝑑
ℎ= 𝑆𝑡
ℎ= 𝐷
𝑃𝑃Figure 57: Optimal damping characteristics of the passive damper
106
Figure 58: Displacements as function of frequency: 1 – relative displacement of shock absorber between the two mounting points, 2 – relative displacement between the piston rod and cylinder tube 62
107
Figure 59: Input function of the system
108
109
𝐶
ℎ= 𝐷
𝜏/3 𝐶
𝑟𝑖𝐶
𝑟𝑖110
Figure 60: How Voltage Induced changes with Coil Internal radius, 0.5mm resolution
0.95 < 𝐶
𝑟𝑖(𝑛𝑜𝑟𝑚) < 1
111
Figure 61: How Voltage Induced changes with Coil Internal Radius, 0.1mm resolution
𝐶
𝑟𝑖(𝑛𝑜𝑟𝑚) ≈ 0.99
𝐷
𝜏= 2 ∙ 𝜏 = 2 ∙ (𝑀
ℎ+ 𝐼𝑃
ℎ)
𝜏
𝑚/𝜏 𝜏
𝜏
𝑚)
112
Figure 62: How Voltage Induced changes with taum/tau, resolution of 0.5 mm
0.6 < 𝜏
𝑚/𝜏 < 0.8
113
Figure 63: How Voltage Induced changes with taum/tau, resolution of 0.1mm
𝜏
𝑚/𝜏 = 0.69
𝑁 = 𝐹 𝐴
𝐶𝐴
𝑊= 𝐹 4𝐴
𝐶𝜋𝜙
2114
𝐴
𝑤𝜙
𝜙
𝑁 = 𝐹 4𝐴
𝐶𝜋𝜙
2≈ 274
𝑅
𝐶= 𝜌 𝐿
𝐴 = 𝜌 𝜋 (𝐶
𝑟𝑖+ 𝐶
𝑟𝑜) 2
𝐴
𝐶= 86.055 ∙ 10
−6Ω
𝐶
𝑟𝑖𝐶
𝑟𝑜𝐴
𝐶𝑅
𝐿𝑅
𝐿𝑜𝑎𝑑115 𝑟 =
𝑅𝐿𝑜𝑎𝑑𝑅𝐶𝑜𝑖𝑙
𝑟 = 0.5 𝑟 = 1 𝑟 = 2
Figure 64: How power out change with load resistance
𝑟 = 1 → 𝑅
𝑙𝑜𝑎𝑑= 𝑅
𝑐𝑜𝑖𝑙116
𝑓 = 1 𝐻𝑧
Figure 65: Three-phase coil voltage at 1Hz
117
Figure 66: Three-phase coil power at 1Hz
20 𝐻𝑧 < 𝑓 < 80 𝐻𝑧
118
Figure 67: Three-phase power output for different frequencies
𝐵⃗
𝑟𝐵⃗
𝑟= 2.15𝑇
119
Figure 68: Magnetic flux density with optimized geometry model
120
Figure 69: Magnetic flux density with optimized geometry model and material feasibility
Figure 70: Magnetic flux density with optimized geometry model and material feasibility
121
Figure 71: Three-phase power output for different frequencies
122
123
Figure 72: How power changes with Coil Phases60
Figure 73: Coil design. (a) Two-phase coil. (b) Three-phase coil. (c) Four-phase coil60
124
125 𝐵⃗
𝑟𝐻
𝑐𝐵⃗
𝑟= 𝐵⃗
𝑟20[1 + 𝛼
𝐵100 (𝜗
𝑃𝑀− 20)]
𝐻
𝑐= 𝐻
𝑐20[1 + 𝛼
𝐻100 (𝜗
𝑃𝑀− 20)]
𝜗
𝑃𝑀𝐵⃗
𝑟20𝐻
𝑐20𝛼
𝐵< 0 𝛼
𝐻< 0
𝐵⃗
𝑟𝐻
𝑐126
127
128
𝜌
∇ × 𝑯 = 𝑱 + 𝜕𝑫
𝜕𝑡
∇ ∙ 𝑩 = 0
∇ × 𝑬 = − 𝜕𝑩
𝜕𝑡
∇ ∙ 𝑫 = 0
129 𝑩 = 𝜇
0(𝑯 + 𝑴)
𝐷 = 𝜀
0𝑬 + 𝑷
𝑱 = 𝜎𝑬
𝜇
0, ε
0𝜎
∇ ∙ 𝑩 = 0 → 𝐵⃗ = ∇ × 𝑨
130
∇ × (𝑬 + 𝜕𝑨
𝜕𝑡 ) = 0 → 𝑬 + 𝜕𝑨
𝜕𝑡 = −∇𝜑
∇
2𝐴 − 𝜇𝜀 𝜕
2𝐴
𝜕𝑡 = −𝜇𝐽
∇
2𝜑 − 𝜇𝜀 𝜕
2𝜑
𝜕𝑡
2= − 𝜌 𝜀
∇ ∙ 𝐴 = −𝜇𝜀
𝛿𝜑𝛿𝑡
∇
2𝐴 − 𝜇𝜀 𝛿
2𝑨
𝛿𝑡
2= −𝜇𝑱
∇
2𝜑 − 𝜇𝜀 𝛿
2𝜑 𝛿𝑡
2= − 𝜌
𝜀
131
∇
2𝑨 = −𝜇𝑱
∇
2𝜑 = − 𝜌 𝜀
∇
2𝐴(𝑥, 𝑡) = 𝜇
04𝜋 ∫ 𝑱(𝑥
′, 𝑡)
|𝒙 − 𝒙
′| 𝑑Γ′
Γ
𝜑(𝑥, 𝑡) = 1
4𝜋𝜀
0∫ 𝜌(𝒙
′, 𝑡)
|𝒙 − 𝒙
′| 𝑑Γ′
Γ
𝛿𝑫 𝛿𝑡
=
𝛿𝑩𝛿𝑡
= 0
𝑩 = ∇ × 𝑨
132 𝐵⃗(𝑥, 𝑡) = 𝜇
4𝜋 ∫ 𝒋(𝒙
′, 𝑡) × (𝒙 − 𝒙
′)
|𝒙 − 𝒙
′|
𝟑𝑑𝑠′
∇𝑨 = 0
∇
2𝑨 = −𝜇
0(𝑱 + ∇ × 𝑴)
𝑱
𝑚= ∇ × 𝑴
𝑩(𝑥) = 𝜇
4𝜋 ∫ 𝑱
𝒎(𝒙
′) × (𝒙 − 𝒙
′)
|𝒙 − 𝒙
′|
𝟑𝑑Γ + 𝜇
4𝜋 ∫ 𝒋
𝒎(𝒙
′) × (𝒙 − 𝒙
′)
|𝒙 − 𝒙
′|
𝟑𝑑𝑠
𝐽
𝑚𝑗
𝑚133 𝑱
𝒎= ∇ × 𝑴 ( 𝐴
𝑚
2) 𝑉𝑜𝑙𝑢𝑚𝑒 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 𝑑𝑒𝑛𝑠𝑖𝑡𝑦
𝒋
𝒎= 𝐌 × 𝒏 ̂ ( 𝐴
𝑚 ) 𝑆𝑢𝑟𝑓𝑎𝑐𝑒 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 𝑑𝑒𝑛𝑠𝑖𝑡𝑦
𝒏
̂
𝑴 = 𝑀
𝑠𝒛̂
𝑱
𝒎= 0
𝒋
𝒎= M𝝋 ̂
Figure B-74: Permanent magnet and its equivalent surface current density40
134
135
Ref. 1
136
Figure 75: How the Extrapolation of the B-H Curve Affects the Simulation