betaFeedback

WorksPhysics and Astronomy

Calculation of force and torque between two arbitrarily oriented circular filaments using Kalantarov-Zeitlin's method

Abstract

In this article, formulas for the calculation of force and torque between two circular filaments arbitrarily oriented in space were derived by using Kalantarov-Zeitlin's method. Formulas are presented in an analytical form through integral expressions, whose kernel function is expressed in terms of the elliptic integrals of the first and second kinds, and provide an alternative formulation to Babic's expressions. The derived new formulas were validated via comparison with a series of reference examples. Also, we obtained additional expressions for the calculation of force and torque between two circular filaments by means of differentiation of Grover's formula for the mutual inductance between two circular filaments with respect to appropriate coordinates. These additional expressions allow to verifying comprehensively and independently derived new formulas.

Introduction

Analytical and semi-analytical methods in the calculation of parameters of electrical circuits and force interaction between their elements play an important

Email address: [email protected], [email protected] (Kirill V. Poletkin) tant role in power transfer, wireless communication, and sensing and actuation, and are applied in different fields of science, including electrical and electronic engineering, medicine, physics, nuclear magnetic resonance, mechatronics and robotics, to name the most prominent. A number of efficient numerical methods implemented in the commercially available software currently provides an accurate and fast solution for the calculation of parameters of electrical circuits. However, analytical methods allow to obtain the result in the form of a final formula with a finite number of input parameters, which when applicable may significantly reduce computation effort. It will also facilitate mathematical analysis, for example, when derivatives of the mutual inductance with respect to one or more parameters are required to evaluate electromagnetic forces via the stored magnetic energy, or when optimization is performed.

Analytical methods applied to the calculation of force and torque between two circular filaments, which is closely related to calculation of the mutual inductance in such the system, is a prime example. These methods have been successfully used in an increasing number of applications, including electromagnetic levitation [1], superconducting levitation [2, 3, 4], magnetic force interaction [5], wireless power transfer [6, 7, 8], electromagnetic actuation [9, 10, 11, 12], micromachined contactless inductive suspensions [13, 14, 15, 16, 17] and hybrid suspensions [18, 19, 20, 21], biomedical applications [22, 23], topology optimization [24], nuclear magnetic resonance [25, 26], indoor positioning systems [27], navigation sensors [28], wireless power transfer systems [29] and magneto-inductive wireless communications [30].

The calculation of force and torque between two circular filaments with running electrical current is reduced to finding first derivatives of a function of mutual inductance corresponding to this filament system. In 1954, using the mutual inductance between two coaxial circular filaments derived by Maxwell [31 page 340, Art. 701], C. Snow obtained the formula for calculation attractive force between them in work [32]. Also, in the same work [32] C. Snow presented formulas for calculation of torque between circular filaments covering the following cases, namely, when axes of circular filaments are intersected, and two concentric circles, and for calculation of force between two parallel circles. The formulas were expressed via series over the Legendre polynomials. In 1996, Kim et al. developed the expression for calculation of the restoring force in a system of two non-coaxial coils based on magnetic potential method [33]. Employing Grover’s formula for calculation of mutual inductance between two filament coils [34], Babič et al. developed the formulas for calculation of force and torque in such filament coil systems, in which circles have the lateral misalignment [35] and whose axes are inclined at the same plane [36, 37], respectively. The developed formulas were applied to the calculation of force interaction between superconducting magnets and coils having a rectangular cross-section. In work [38], Babič et al. presented new general formulas for calculating the magnetic force between inclined circular filaments placed in any desired position based on two approaches, namely, Biot-Savart’s law and the formula for calculation of the mutual inductance [39].

In this article, new formulas for calculation of force and torque between two circular filaments arbitrarily oriented in space presented in an integral analytical form, whose kernel function is expressed in terms of the elliptic integrals of the first and second kinds, were derived based on Kalantarov-Zeitlin’s method and provide an alternative formulation to Babič’s expressions. Kalantarov and Zeitlin showed that the calculation of mutual inductance between a circular primary filament and any other secondary filament having an arbitrary shape and any desired position with respect to the primary filament can be reduced to a line integral [40 Sec. 1-12, page 49]. Adapting this method to the case of two circular filaments, the author derived analytical formulas for calculating the mutual inductance between two circular filaments having any desired position with respect to each other in work [41]. Taking first derivatives with respect to the appropriate coordinates of the formulas for calculation of the mutual inductance obtained by means of Kalantarov-Zeitlin’s method, the derivation of these presented formulas for calculation of force and torque was carried out.

Developed new formulas were successfully verified by the examples taken from Babič et al. work [38] and comparison with results of calculation of force and torque performed by expressions derived from Grover’s formula for calculation of mutual inductance [42 page 207, Eq. (179)]. For the convenience of a reader, all derived formulas including expressions derived by Grover’s method were programmed by using the Matlab language. The Matlab files with the implemented formulas are available from the author as supplementary materials to this article.

Preliminary discussion

Similar to our previous work [41], the general scheme of arbitrarily positioning of two circular filaments with respect to each other is considered as shown in Figure 1. The primary circular filament (the primary circle) and the secondary circular filament (the secondary circle) have radii of RpR_p and RsR_s, respectively. A coordinate frame (CF) denoted as XYZXYZ is assigned to the primary circle in such a way that the ZZ axis is coincident with the circle axis and the XOYXOY plane of the CF lies on circle’s plane, where the origin OO corresponds to the centre of primary circle. In turn, the xyzxyz CF is assigned to the secondary circle in a similar way so that its origin BB is coincident with the centre of the secondary circle.

General scheme of arbitrarily positioning two circular filaments with respect to each other

Figure 1. General scheme of arbitrarily positioning two circular filaments with respect to each other.

The linear misalignment of the secondary circle with respect to the primary one is defined by the coordinates of the centre BB (xBx_B, yBy_B, zBz_B). The angular misalignment of the secondary circle can be defined by using Grover’s angles [42 page 207]. Namely, the angle of θ\theta and η\eta corresponds to the angular rotation around an axis passing through the diameter of the secondary circle, and then the rotation of this axis lying on the surface x′By′x'By' around the vertical z′z' axis, respectively, as it is shown in Figure 2(a). As an alternative manner to Grover’s angles, the same angular misalignment can be determined through the α\alpha and β\beta angle, which corresponds to the angular rotation around the x′x' axis and then around the y′′y'' axis, respectively, as it is shown in Figure 2(b). This additional second manner is more convenient in a case of study dynamics and stability issues, for instance, applying to axially symmetric inductive levitation systems [14, 16] in compared with Grover’s manner. These two pairs of angles have the following relationship with respect to each other such as [41]:

Figure 2. Two manners for determining the angular position of the secondary circle with respect to the primary one: x′y′z′x'y'z' is the auxiliary CF the axes of which are parallel to the axes of XYZXYZ, respectively; x′′y′′z′′x''y''z'' is the auxiliary CF defined in such a way that the x′x' and x′′x'' are coincide, but the z′′z'' and y′′y'' axis is rotated by the α\alpha angle with respect to the z′z' and y′y' axis, respectively.

sin⁡β=sin⁡ηsin⁡θ;cos⁡βsin⁡α=cos⁡ηsin⁡θ.(1)\begin{aligned} \sin\beta&= \sin\eta\sin\theta;\\ \cos\beta\sin\alpha&= \cos\eta\sin\theta. \tag*{(1)} \end{aligned}

The mutual inductance between these two filaments can be calculated by the following formulas, which were derived by using Kalantarov-Zeitlin’s approach in work [41] for two cases. Introducing the following dimensionless coordinates:

x=xBRs; y=yBRs; z=zBRs; s=x2+y2,(2)x=\frac{x_B}{R_s};\ y=\frac{y_B}{R_s};\ z=\frac{z_B}{R_s};\ s=\sqrt{x^2+y^2}, \tag*{(2)}

for the first case when the θ\theta angle is lying in interval of 0≤θ<π/20\le\theta<\pi/2, the formula can be written as

M=μ0RpRsπ∫02πr⋅U⋅Φ(k)dφ,(3)M=\frac{\mu_0\sqrt{R_pR_s}}{\pi}\int_0^{2\pi}r\cdot U\cdot\Phi(k)d\varphi, \tag*{(3)}

where

r=r(θ,η)=cos⁡θsin⁡2(φ−η)+cos⁡2θcos⁡2(φ−η),(4)r=r(\theta,\eta)=\frac{\cos\theta}{\sqrt{\sin^2(\varphi-\eta)+\cos^2\theta\cos^2(\varphi-\eta)}}, \tag*{(4)}
U=U(x,y,θ,η)=Rρ1.5=r+t1⋅cos⁡φ+t2⋅sin⁡φρ1.5,(5)U=U(x,y,\theta,\eta)=\frac{R}{\rho^{1.5}}=\frac{r+t_1\cdot\cos\varphi+t_2\cdot\sin\varphi}{\rho^{1.5}}, \tag*{(5)}
t1=t1(x,y,θ,η)=x+0.5rˉ2tan⁡2θsin⁡(2(φ−η))⋅y,t2=t2(x,y,θ,η)=y−0.5rˉ2tan⁡2θsin⁡(2(φ−η))⋅x,ρ=ρ(x,y,θ,η)=r2+2r⋅(xcos⁡(φ)+ysin⁡(φ))+s2,(6)\begin{aligned} t_1&=t_1(x,y,\theta,\eta)=x+0.5\bar{r}^{2}\tan^{2}\theta\sin(2(\varphi-\eta))\cdot y,\\ t_2&=t_2(x,y,\theta,\eta)=y-0.5\bar{r}^{2}\tan^{2}\theta\sin(2(\varphi-\eta))\cdot x,\\ \rho&=\rho(x,y,\theta,\eta)=\sqrt{r^2+2r\cdot(x\cos(\varphi)+y\sin(\varphi))+s^2}, \tag*{(6)} \end{aligned}
Φ(k)=1k[(1−k22)K(k)−E(k)],(7)\Phi(k)=\frac{1}{k}\left[\left(1-\frac{k^2}{2}\right)K(k)-E(k)\right], \tag*{(7)}

and K(k)K(k) and E(k)E(k) are the complete elliptic functions of the first and second kind, respectively, and

k2=k2(x,y,z,θ,η)=4νρ(νρ+1)2+ν2zλ2,ν=Rs/Rp,zλ=z+rtan⁡θsin⁡(φ−η).(8)\begin{aligned} k^2&=k^2(x,y,z,\theta,\eta)=\frac{4\nu\rho}{(\nu\rho+1)^2+\nu^2z_\lambda^2},\\ \nu&=R_s/R_p,\quad z_\lambda=z+r\tan\theta\sin(\varphi-\eta). \tag*{(8)} \end{aligned}

For the second case when the θ\theta angle is equal to π/2\pi/2 and the two filament circles are mutually perpendicular to each other, the formula becomes

M=μ0RpRsπ{∫−11U⋅Φ(k)dℓˉ+∫1−1U⋅Φ(k)dℓˉ},(9)M=\frac{\mu_0\sqrt{R_pR_s}}{\pi}\left\{\int_{-1}^{1}U\cdot\Phi(k)d\bar{\ell}+\int_{1}^{-1}U\cdot\Phi(k)d\bar{\ell}\right\}, \tag*{(9)}

where

U=U(x,y,η)=Rρ1.5=t1−t2ρ1.5,(10)U=U(x,y,\eta)=\frac{R}{\rho^{1.5}}=\frac{t_1-t_2}{\rho^{1.5}}, \tag*{(10)}
t1=t1(x,η)=sin⁡η⋅(x+ℓˉcos⁡η),t2=t2(y,η)=cos⁡η⋅(y+ℓˉsin⁡η),ρ=ρ(x,y,η)=s2+2ℓˉ⋅(xcos⁡(η)+ysin⁡(η))+ℓˉ2(11)\begin{aligned} t_1 &= t_1(x,\eta)=\sin\eta\cdot(x+\bar{\ell}\cos\eta),\\ t_2 &= t_2(y,\eta)=\cos\eta\cdot(y+\bar{\ell}\sin\eta),\\ \rho&= \rho(x,y,\eta)=\sqrt{s^2+2\bar{\ell}\cdot(x\cos(\eta)+y\sin(\eta))+\bar{\ell}^2} \tag*{(11)} \end{aligned}

and ℓˉ=ℓ/Rs\bar{\ell}=\ell/R_s is the dimensionless variable. The functions Φ(k)\Phi(k) and k=k(x,y,z,η)k=k(x,y,z,\eta) in formula (9) have the same structures as defined by Eq. (7) and (8), respectively. Besides that, in the elliptic module k=k(x,y,z,η)k=k(x,y,z,\eta), the zλz_\lambda function is governed as follows

zλ=z±1−ℓˉ2(12)z_\lambda=z\pm\sqrt{1-\bar{\ell}^2} \tag*{(12)}

Note that integrating formula (9) between −1-1 and 11, Eq. (12) is calculated with the positive sign and for the other direction the negative sign is taken.

Assuming that the primary and secondary circular filaments carry the currents of IpI_p and IsI_s, respectively, hence, the magnetic force and torque between these two circular filaments can be calculated by taking the first derivatives of the function of the magnetic energy stored in such the system with respect to the appropriate coordinates. Hence, force can be calculated by

Fg=IpIs∂M∂g(13)F_g=I_pI_s\frac{\partial M}{\partial g} \tag*{(13)}

where g=xBg=x_B, yBy_B, or zBz_B. For torque, we can write:

Tg=IpIs∂M∂q(14)T_g=I_pI_s\frac{\partial M}{\partial q} \tag*{(14)}

where g=θg=\theta or η\eta. Thus, to derive formulas for calculation of force and torque between two arbitrarily oriented circular filaments, the first derivatives of formulas of mutual inductance, namely, represented by Eq. (3) and (9) must be taken.

Derivation of Formulas

In this section the first derivatives of mutual inductance with respect to appropriate coordinates are taken for the two considered above cases separately.

The first case: 0≤θ<π/20 \le\theta< \pi/2

For this case, formula (3) for calculation of mutual inductance is considered.

Its kernel is defined as

Kr=r⋅U⋅Φ(k).(15)\mathrm{Kr}=r\cdot U\cdot\Phi(k). \tag*{(15)}

Finding the first derivatives of mutual inductance (3) is reduced to taking the derivatives of kernel function. Following this, firstly five derivatives of the kernel with respect to the defined five coordinates are obtained as follows.

According to the definitions of functions rr, UU, Φ(k)\Phi(k) and kk given in (4), (5), (7) and (8), respectively, the xBx_B-derivative of kernel Kr\mathrm{Kr} can be written as

∂Kr∂xB=∂Kr∂x1Rs=rRs⋅[∂U∂x⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂x],(16)\frac{\partial\mathrm{Kr}}{\partial x_B}=\frac{\partial\mathrm{Kr}}{\partial x}\frac{1}{R_s}=\frac{r}{R_s}\cdot\left[\frac{\partial U}{\partial x}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial x}\right], \tag*{(16)}

where

∂U∂x=(∂R∂x⋅ρ−1.5⋅R⋅∂ρ∂x)/ρ2.5,∂R∂x=∂t1∂x⋅cos⁡φ+∂t2∂x⋅sin⁡φ,∂t1∂x=1,∂t2∂x=−0.5rˉ2tan⁡2θsin⁡(2(φ−η)),∂ρ∂x=(r⋅cos⁡φ+x)/ρ.(17)\begin{aligned} \frac{\partial U}{\partial x}&=\left(\frac{\partial R}{\partial x}\cdot\rho-1.5\cdot R\cdot\frac{\partial\rho}{\partial x}\right)\big/\rho^{2.5},\\ \frac{\partial R}{\partial x}&=\frac{\partial t_1}{\partial x}\cdot\cos\varphi+\frac{\partial t_2}{\partial x}\cdot\sin\varphi,\\ \frac{\partial t_1}{\partial x}&=1,\qquad\frac{\partial t_2}{\partial x}=-0.5\bar{r}^{2}\tan^{2}\theta\sin(2(\varphi-\eta)),\\ \frac{\partial\rho}{\partial x}&=(r\cdot\cos\varphi+x)/\rho. \tag*{(17)} \end{aligned}
dΦ(k)dk=1k2[2−k22(1−k2)E(k)−K(k)],(18)\frac{d\Phi(k)}{dk}=\frac{1}{k^{2}}\left[\frac{2-k^{2}}{2(1-k^{2})}E(k)-K(k)\right], \tag*{(18)}
∂k∂x=2/k−k(νρ+1)(νρ+1)2+ν2λ2⋅ν∂ρ∂x.(19)\frac{\partial k}{\partial x}=\frac{2/k-k(\nu\rho+1)}{(\nu\rho+1)^{2}+\nu^{2}\lambda^{2}}\cdot\nu\frac{\partial\rho}{\partial x}. \tag*{(19)}

For the yBy_B-derivative of kernel Kr\mathrm{Kr}, we have

∂Kr∂yB=∂Kr∂y1Rs=rRs⋅[∂U∂y⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂y],(20)\frac{\partial\mathrm{Kr}}{\partial y_B}=\frac{\partial\mathrm{Kr}}{\partial y}\frac{1}{R_s}=\frac{r}{R_s}\cdot\left[\frac{\partial U}{\partial y}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial y}\right], \tag*{(20)}

where

∂U∂y=(∂R∂y⋅ρ−1.5⋅R⋅∂ρ∂y)/ρ2.5,∂R∂y=∂t1∂y⋅cos⁡φ+∂t2∂y⋅sin⁡φ,∂t1∂y=0.5rˉ2tan⁡2θsin⁡(2(φ−η)),∂t2∂y=1,∂ρ∂y=(r⋅sin⁡φ+y)/ρ.(21)\begin{aligned} \frac{\partial U}{\partial y}&=\left(\frac{\partial R}{\partial y}\cdot\rho-1.5\cdot R\cdot\frac{\partial\rho}{\partial y}\right)\big/\rho^{2.5},\\ \frac{\partial R}{\partial y}&=\frac{\partial t_1}{\partial y}\cdot\cos\varphi+\frac{\partial t_2}{\partial y}\cdot\sin\varphi,\\ \frac{\partial t_1}{\partial y}&=0.5\bar{r}^{2}\tan^{2}\theta\sin(2(\varphi-\eta)),\qquad\frac{\partial t_2}{\partial y}=1,\\ \frac{\partial\rho}{\partial y}&=(r\cdot\sin\varphi+y)/\rho. \tag*{(21)} \end{aligned}

The derivative of elliptic module kk with respect to yy has a similar form to Eq. (19), but in the later one the partial derivative, ∂ρ∂x\frac{\partial\rho}{\partial x}, must be replaced by ∂ρ∂y\frac{\partial\rho}{\partial y} defined in (21). The zBz_{\mathrm{B}}-derivative of kernel Kr\mathrm{K}_r is

∂Kr∂zB=∂Kr∂z1Rs=rRs⋅U⋅dΦ(k)dk⋅∂k∂z,(22)\frac{\partial\mathrm{K}_r}{\partial z_{\mathrm{B}}}=\frac{\partial\mathrm{K}_r}{\partial z}\frac{1}{R_s}=\frac{r}{R_s}\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial z}, \tag*{(22)}

where

∂k∂z=−4νρ⋅ν2zλ((νρ+1)2+ν2zλ2)3/2.(23)\frac{\partial k}{\partial z}=-\sqrt{4\nu\rho}\cdot\frac{\nu^2z_\lambda}{\left((\nu\rho+1)^2+\nu^2z_\lambda^2\right)^{3/2}}. \tag*{(23)}

The θ\theta-derivative of kernel Kr\mathrm{K}_r becomes as follows

∂Kr∂θ=[∂r∂θ⋅U+r⋅∂U∂θ]⋅Φ(k)+r⋅U⋅dΦ(k)dk⋅∂k∂θ,(24)\frac{\partial\mathrm{K}_r}{\partial\theta}=\left[\frac{\partial r}{\partial\theta}\cdot U+r\cdot\frac{\partial U}{\partial\theta}\right]\cdot\Phi(k)+r\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial\theta}, \tag*{(24)}

where

∂r∂θ=−sin⁡(η−φ)2sin⁡(θ)(sin⁡(η−φ)2+cos⁡(θ)2cos⁡(η−φ)2)3/2,(25)\frac{\partial r}{\partial\theta}=-\frac{\sin(\eta-\varphi)^2\sin(\theta)}{\left(\sin(\eta-\varphi)^2+\cos(\theta)^2\cos(\eta-\varphi)^2\right)^{3/2}}, \tag*{(25)}
∂U∂θ=(∂R∂θ⋅ρ−1.5⋅R⋅∂ρ∂θ)/ρ2.5,∂R∂θ=∂r∂θ+∂t1∂θ⋅cos⁡φ+∂t2∂θ⋅sin⁡φ,∂t1∂θ=r⋅y⋅sin⁡(2φ−2η)tan⁡(θ)[∂r∂θ⋅tan⁡(θ)+rcos⁡(θ)2],∂t2∂θ=−r⋅x⋅sin⁡(2φ−2η)tan⁡(θ)[∂r∂θ⋅tan⁡(θ)+rcos⁡(θ)2],∂ρ∂θ=r+y⋅sin⁡φ+x⋅cos⁡φρ⋅∂r∂θ,(26)\begin{aligned} \frac{\partial U}{\partial\theta}&=\left(\frac{\partial R}{\partial\theta}\cdot\rho-1.5\cdot R\cdot\frac{\partial\rho}{\partial\theta}\right)/\rho^{2.5},\\ \frac{\partial R}{\partial\theta}&=\frac{\partial r}{\partial\theta}+\frac{\partial t_1}{\partial\theta}\cdot\cos\varphi+\frac{\partial t_2}{\partial\theta}\cdot\sin\varphi,\\ \frac{\partial t_1}{\partial\theta}&=r\cdot y\cdot\sin(2\varphi-2\eta)\tan(\theta)\left[\frac{\partial r}{\partial\theta}\cdot\tan(\theta)+\frac{r}{\cos(\theta)^2}\right],\\ \frac{\partial t_2}{\partial\theta}&=-r\cdot x\cdot\sin(2\varphi-2\eta)\tan(\theta)\left[\frac{\partial r}{\partial\theta}\cdot\tan(\theta)+\frac{r}{\cos(\theta)^2}\right],\\ \frac{\partial\rho}{\partial\theta}&=\frac{r+y\cdot\sin\varphi+x\cdot\cos\varphi}{\rho}\cdot\frac{\partial r}{\partial\theta}, \tag*{(26)} \end{aligned}
∂k∂θ=[2/k−k(νρ+1)]⋅ν∂ρ∂θ−k⋅ν2zλ∂zλ∂θ(νρ+1)2+ν2zλ2,∂zλ∂θ=sin⁡(φ−η)[∂r∂θ⋅tan⁡(θ)+rcos⁡(θ)2].(27)\begin{aligned} \frac{\partial k}{\partial\theta}&=\frac{\left[2/k-k(\nu\rho+1)\right]\cdot\nu\frac{\partial\rho}{\partial\theta}-k\cdot\nu^2z_\lambda\frac{\partial z_\lambda}{\partial\theta}}{(\nu\rho+1)^2+\nu^2z_\lambda^2},\\ \frac{\partial z_\lambda}{\partial\theta}&=\sin(\varphi-\eta)\left[\frac{\partial r}{\partial\theta}\cdot\tan(\theta)+\frac{r}{\cos(\theta)^2}\right]. \tag*{(27)} \end{aligned}

The η\eta-derivative of kernel Kr\mathrm{K}_r can be written as

∂Kr∂η=[∂r∂η⋅U+r⋅∂U∂η]⋅Φ(k)+r⋅U⋅dΦ(k)dk⋅∂k∂η,(28)\frac{\partial\mathrm{K}_r}{\partial\eta}=\left[\frac{\partial r}{\partial\eta}\cdot U+r\cdot\frac{\partial U}{\partial\eta}\right]\cdot\Phi(k)+r\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial\eta}, \tag*{(28)}

where

∂r∂η=sin⁡(φ−η)cos⁡(φ−η)cos⁡(θ)(1−cos⁡(θ)2)(sin⁡(η−φ)2+cos⁡(θ)2cos⁡(η−φ)2)3/2,(29)\frac{\partial r}{\partial\eta}=\frac{\sin(\varphi-\eta)\cos(\varphi-\eta)\cos(\theta)(1-\cos(\theta)^2)}{\left(\sin(\eta-\varphi)^2+\cos(\theta)^2\cos(\eta-\varphi)^2\right)^{3/2}}, \tag*{(29)}
∂U∂η=(∂R∂η⋅ρ−1.5⋅R⋅∂ρ∂η)/ρ2.5,∂R∂η=∂r∂η+∂t1∂η⋅cos⁡φ+∂t2∂η⋅sin⁡φ,∂t1∂η=tan⁡(θ)2⋅r⋅y⋅[∂r∂η⋅sin⁡(2φ−2η)−r⋅cos⁡(2φ−2η)],∂t2∂η=−tan⁡(θ)2⋅r⋅x⋅[∂r∂η⋅sin⁡(2φ−2η)−r⋅cos⁡(2φ−2η)],∂ρ∂η=r+y⋅sin⁡φ+x⋅cos⁡φρ⋅∂r∂η,(30)\begin{aligned} \frac{\partial U}{\partial\eta} &= \left(\frac{\partial R}{\partial\eta}\cdot\rho- 1.5\cdot R\cdot\frac{\partial\rho}{\partial\eta}\right)/\rho^{2.5},\\ \frac{\partial R}{\partial\eta} &= \frac{\partial r}{\partial\eta}+\frac{\partial t_1}{\partial\eta}\cdot\cos\varphi+\frac{\partial t_2}{\partial\eta}\cdot\sin\varphi,\\ \frac{\partial t_1}{\partial\eta} &= \tan(\theta)^2\cdot r\cdot y\cdot\left[\frac{\partial r}{\partial\eta}\cdot\sin(2\varphi-2\eta)-r\cdot\cos(2\varphi-2\eta)\right],\\ \frac{\partial t_2}{\partial\eta} &= -\tan(\theta)^2\cdot r\cdot x\cdot\left[\frac{\partial r}{\partial\eta}\cdot\sin(2\varphi-2\eta)-r\cdot\cos(2\varphi-2\eta)\right],\\ \frac{\partial\rho}{\partial\eta} &= \frac{r+y\cdot\sin\varphi+x\cdot\cos\varphi}{\rho}\cdot\frac{\partial r}{\partial\eta}, \tag*{(30)} \end{aligned}
∂k∂η=[2/k−k(νρ+1)]⋅ν∂ρ∂η−k⋅ν2zλ2∂zλ∂η(νρ+1)2+ν2zλ2,∂zλ∂η=tan⁡(θ)⋅[∂r∂η⋅sin⁡(φ−η)−r⋅cos⁡(φ−η)].(31)\begin{aligned} \frac{\partial k}{\partial\eta} &= \frac{[2/k-k(\nu\rho+1)]\cdot\nu\frac{\partial\rho}{\partial\eta}-k\cdot\nu^2 z_\lambda^2\frac{\partial z_\lambda}{\partial\eta}}{(\nu\rho+1)^2+\nu^2z_\lambda^2},\\ \frac{\partial z_\lambda}{\partial\eta} &= \tan(\theta)\cdot\left[\frac{\partial r}{\partial\eta}\cdot\sin(\varphi-\eta)-r\cdot\cos(\varphi-\eta)\right]. \tag*{(31)} \end{aligned}

Accounting for (16) and (20), the formulas for calculation the first derivatives of mutual inductance between two circular filaments arbitrarily oriented in space relative to the xx and yy axis can be written as

∂M∂g=μ0πRpRs∫02πr⋅[∂U∂g⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂g]dφ,(32)\frac{\partial M}{\partial g}=\frac{\mu_0}{\pi}\sqrt{\frac{R_p}{R_s}}\int_0^{2\pi}r\cdot\left[\frac{\partial U}{\partial g}\cdot\Phi(k)+U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial g}\right]\mathrm{d}\varphi, \tag*{(32)}

where g=xBg=x_B, and yBy_B, the derivatives of functions UU and kk with respect to appropriate coordinates are defined in (16), and (20) for xBx_B and yBy_B coordinate, respectively. The formula for calculation of the first derivative of the mutual inductance (3) with respect to zBz_B by taking into account (22) becomes

∂M∂zB=μ0πRpRs∫02πr⋅U⋅dΦ(k)dk⋅∂k∂zdφ.(33)\frac{\partial M}{\partial z_B}=\frac{\mu_0}{\pi}\sqrt{\frac{R_p}{R_s}}\int_0^{2\pi}r\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial z}\mathrm{d}\varphi. \tag*{(33)}

The first derivatives of the mutual inductance (3) with respect to angular coordinates by accounting for (24) and (28) are

∂M∂g=μ0RpRsπ∫02π[∂r∂g⋅U+r⋅∂U∂g]⋅Φ(k)+r⋅U⋅dΦ(k)dk⋅∂k∂gdφ,(34)\frac{\partial M}{\partial g}=\frac{\mu_0\sqrt{R_pR_s}}{\pi}\int_0^{2\pi}\left[\frac{\partial r}{\partial g}\cdot U+r\cdot\frac{\partial U}{\partial g}\right]\cdot\Phi(k)+r\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial g}\mathrm{d}\varphi, \tag*{(34)}

where g=θg=\theta, and η\eta. Substituting (32) and (33) into (13), the force between two filaments arbitrarily oriented in space relative to the XX, YY and ZZ axis and carrying electrical current IpI_p and IsI_s can be calculated. Substituting (34) into (14), the torque acting on these two circular filaments can be estimated.

The second case: θ=π/2\theta= \pi/2

For the second case, formula (9) for calculation of mutual inductance is used.

Its kernel is defined as

Kr=U⋅Φ(k).(35)\mathrm{Kr} = U \cdot\Phi(k). \tag*{(35)}

Then, the xBx_B and yBy_B-derivative of the kernel (35) are

∂Kr∂g=∂Kr∂g1Rs=1Rs⋅[∂U∂g⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂g],(36)\frac{\partial\mathrm{Kr}}{\partial g} = \frac{\partial\mathrm{Kr}}{\partial g}\frac{1}{R_s} = \frac{1}{R_s}\cdot\left[\frac{\partial U}{\partial g}\cdot\Phi(k)+U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial g}\right], \tag*{(36)}

where g=xBg=x_B, and yBy_B,

∂U∂g=(∂R∂g⋅ρ−1.5⋅R⋅∂ρ∂g)1ρ2.5,\frac{\partial U}{\partial g}=\left(\frac{\partial R}{\partial g}\cdot\rho-1.5\cdot R\cdot\frac{\partial\rho}{\partial g}\right)\frac{1}{\rho^{2.5}},
∂k∂g=2/k−k(νρ+1)(νρ+1)2+ν2zλ2⋅ν∂ρ∂g,(37)\frac{\partial k}{\partial g}=\frac{2/k-k(\nu\rho+1)}{(\nu\rho+1)^2+\nu^2z_\lambda^2}\cdot\nu\frac{\partial\rho}{\partial g}, \tag*{(37)}

for the xx-derivatives of RR and ρ\rho:

∂R∂x=∂t1∂x=sin⁡η,\frac{\partial R}{\partial x}=\frac{\partial t_1}{\partial x}=\sin\eta,
∂ρ∂x=x+ℓˉcos⁡ηρ,(38)\frac{\partial\rho}{\partial x}=\frac{x+\bar{\ell}\cos\eta}{\rho}, \tag*{(38)}

for the yy-derivatives of RR and ρ\rho:

∂R∂y=∂t2∂y=cos⁡η,\frac{\partial R}{\partial y}=\frac{\partial t_2}{\partial y}=\cos\eta,
∂ρ∂y=y+ℓˉsin⁡ηρ.(39)\frac{\partial\rho}{\partial y}=\frac{y+\bar{\ell}\sin\eta}{\rho}. \tag*{(39)}

The zBz_B-derivative is

∂Kr∂zB=∂Kr∂z1Rs=1Rs⋅U⋅dΦ(k)dk⋅∂k∂z,(40)\frac{\partial\mathrm{Kr}}{\partial z_B}=\frac{\partial\mathrm{Kr}}{\partial z}\frac{1}{R_s}=\frac{1}{R_s}\cdot U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial z}, \tag*{(40)}

where the partial derivative of elliptic module kk with respect to zz has the same structure as in (23). We can write the following equation for the η\eta-derivative of the kernel:

∂Kr∂η=∂U∂η⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂η,(41)\frac{\partial\mathrm{Kr}}{\partial\eta}=\frac{\partial U}{\partial\eta}\cdot\Phi(k)+U\cdot\frac{\mathrm{d}\Phi(k)}{\mathrm{d}k}\cdot\frac{\partial k}{\partial\eta}, \tag*{(41)}

where

∂U∂η=(∂R∂η⋅ρ−1.5⋅R⋅∂ρ∂η)1ρ2.5,\frac{\partial U}{\partial\eta}=\left(\frac{\partial R}{\partial\eta}\cdot\rho-1.5\cdot R\cdot\frac{\partial\rho}{\partial\eta}\right)\frac{1}{\rho^{2.5}},
∂R∂η=∂t1∂η−∂t2∂η=x⋅cos⁡η+y⋅sin⁡η,\frac{\partial R}{\partial\eta}=\frac{\partial t_1}{\partial\eta}-\frac{\partial t_2}{\partial\eta}=x\cdot\cos\eta+y\cdot\sin\eta,
∂ρ∂η=y⋅cos⁡η−x⋅sin⁡ηρ⋅ℓˉ,\frac{\partial\rho}{\partial\eta}=\frac{y\cdot\cos\eta-x\cdot\sin\eta}{\rho}\cdot\bar{\ell},
∂k∂η=2/k−k(νρ+1)(νρ+1)2+ν2zλ2⋅ν∂ρ∂η.(42)\frac{\partial k}{\partial\eta}=\frac{2/k-k(\nu\rho+1)}{(\nu\rho+1)^2+\nu^2z_\lambda^2}\cdot\nu\frac{\partial\rho}{\partial\eta}. \tag*{(42)}

Hence, replacing the kernel in formula (9) by Eq. (36), the first derivatives of the formula for calculation of the mutual inductance in the case, when the circular filaments are mutually perpendicular to each other with respect to variables of xBx_B and yBy_B can be written as

∂M∂g=μ0πRpRs{∫−11[∂U∂g⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂g]dℓˉ+∫1−1[∂U∂g⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂g]dℓˉ}(43)\frac{\partial M}{\partial g}=\frac{\mu_0}{\pi}\sqrt{\frac{R_p}{R_s}}\left\{\int_{-1}^{1}\left[\frac{\partial U}{\partial g}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial g}\right]d\bar{\ell}+\int_{1}^{-1}\left[\frac{\partial U}{\partial g}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial g}\right]d\bar{\ell}\right\} \tag*{(43)}

where g=xBg=x_B, and yBy_B. Accounting for (40), the first derivative of formula (9) with respect to zBz_B becomes

∂M∂zB=μ0πRpRs{∫−11U⋅dΦ(k)dk⋅∂k∂zdℓˉ+∫1−1U⋅dΦ(k)dk⋅∂k∂zdℓˉ}(44)\frac{\partial M}{\partial z_B}=\frac{\mu_0}{\pi}\sqrt{\frac{R_p}{R_s}}\left\{\int_{-1}^{1}U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial z}d\bar{\ell}+\int_{1}^{-1}U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial z}d\bar{\ell}\right\} \tag*{(44)}

The η\eta-derivative of formula (9) by taking into account (41) can be written as follows

∂M∂η=μ0RpRsπ{∫−11[∂U∂η⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂η]dℓˉ+∫1−1[∂U∂η⋅Φ(k)+U⋅dΦ(k)dk⋅∂k∂η]dℓˉ}(45)\frac{\partial M}{\partial\eta}=\frac{\mu_0\sqrt{R_pR_s}}{\pi}\left\{\int_{-1}^{1}\left[\frac{\partial U}{\partial\eta}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial\eta}\right]d\bar{\ell}+\int_{1}^{-1}\left[\frac{\partial U}{\partial\eta}\cdot\Phi(k)+U\cdot\frac{d\Phi(k)}{dk}\cdot\frac{\partial k}{\partial\eta}\right]d\bar{\ell}\right\} \tag*{(45)}

Now the force and torque for this particular case of the configuration of the filament system can be calculated by substituting (43), (44) and (45) into (13) and (14), respectively.

Since the obtained formulas are intuitively understandable for application, they can be easily programmed. For this purpose, the Matlab language was used. The Matlab files with the implemented formulas (32), (33), (34), (43), (44) and (45) are available from the author as supplementary materials to this article. Also, the developed formulas can be rewritten through the pair of the angle α\alpha and β\beta.

Examples of Calculation. Numerical Verification

In this section, developed new formulas (32), (33), (34), (43), (44) and (45) are verified by the examples taken from Babič et al. work [38] and comparison with results of calculation of force and torque performed by expressions derived from Grover’s formula for calculation of mutual inductance [42 page 207, Eq. (179)]. The derivations of these expressions are shown in Appendix A. In all examples below, it is assumed that the carrying currents in both coils are equal to one ampere (Ip=Is=1I_p = I_s = 1 A). All calculations for considered cases proved the robustness and efficiency of developed formulas.

Force and torque between circular filaments with parallel axes

The scheme for calculation of force and torque between circular filaments with parallel axes is shown in Figure 3. The linear misalignment in the Grover notation can be defined by the geometrical parameter, dd, which is the distance between the planes of circles and the parameter, ρ\rho, is the distance between their axes. These parameters have the following relationship to the notation defined in this article, namely, zB=dz_B = d and ρ=xB2+yB2\rho= \sqrt{x_B^2 + y_B^2}. Figure 3 shows the particular case, when ρ=yB\rho= y_B, which is convenient for calculation of the restoring magnetic force Fy=FρF_y = F_\rho and the propulsive magnetic force Fz=FdF_z = F_d corresponding to Example 6 in Babič et al. work [38]. Also, the torque TθT_\theta and TηT_\eta, which are directed along the xx- and zz-axis, respectively, is calculated.

Figure 3. Geometrical scheme of circular filaments with parallel axes denoted via Grover’s notation: ρ\rho is the distance between axes; dd is the distance between the coils’ planes d=zBd = z_B.

zB=dz_B=d mmBabič’s formula, [38 Eq. (10)], Fr=FρF_r=F_\rho, μNGrover’s method, Eq. (A.16), Fρ=FyF_\rho=F_y, μNThis work, (32), Fy=FρF_y=F_\rho, μN
00.07547749710028990.07547749710028980.0754774971002900
1.00.07488583327209790.07488583327209790.0748858332720979
2.00.07313671349198670.07313671349198680.0731367134919868
3.00.07030546181947180.07030546181947190.0703054618194720
4.00.06651032498899320.06651032498899370.0665103249889934
5.00.06190265669551000.06190265669550990.0619026566955097
6.00.05665516867960670.05665516867960670.0566551686796069
7.00.05094972551589430.05094972551589450.0509497255158945
8.00.04496601777908210.04496601777908230.0449660177790822
9.00.03887211239892300.03887211239892330.0388721123989234
10.00.03281745285065930.03281745285065960.0328174528506595
11.00.02692846490788990.02692846490789020.0269284649078900

Table 1. Calculation of the restoring force for Example 1

Example 1: Force (Example 6, page 74 in Babič et al. work [38])

In this example, the restoring force (in Babič’s notation, it is denoted as FrF_r) between the primary filament having a radius of Rp=42.5R_p=42.5 mm and the secondary one having a radius of Rs=20.0R_s=20.0 mm was calculated. The distance between the coils’ centres is ρ=yB=3.0\rho=y_B=3.0 mm, while the distance, d=zBd=z_B, between the coils’ planes is in a range between 0 and 11 mm. For the calculation both the developed new formula (32) and the one derived by Grover’s method (A.16) were used. The results of calculation are summed up in Table 1. As it can be seen from the analysis of Table 1 that all results of calculation of restoring force obtained by both formulas (32) and (A.16) are in an excellent agreement with results of calculation performed by Babič’s formula.

zB=dz_B=dBabič’s formula, [38, (10)], Fz=FdF_z=F_d, μN\mu\mathrm{N}Grover’s method, Eq. (A.17), Fd=FzF_d=F_z, μN\mu\mathrm{N}This work, (33), Fz=FdF_z=F_d, μN\mu\mathrm{N}
mm
00.00.00.0
1.0-0.0510570118824195-0.0510570118824194-0.0510570118824195
2.0-0.101293579295215-0.101293579295215-0.101293579295215
3.0-0.149926462414138-0.149926462414137-0.149926462414138
4.0-0.196243385024538-0.196243385024538-0.196243385024538
5.0-0.239630444882395-0.239630444882394-0.239630444882394
6.0-0.279591228665736-0.279591228665736-0.279591228665736
7.0-0.315756838476827-0.315756838476827-0.315756838476827
8.0-0.347887153545896-0.347887153545896-0.347887153545896
9.0-0.375864540746323-0.375864540746323-0.375864540746323
10.0-0.399681785157597-0.399681785157597-0.399681785157597
11.0-0.419426221842137-0.419426221842137-0.419426221842137

Table 2. Calculation of the propulsive force for Example 2

Example 2: Force (Example 6, page 74 in Babič et al. work [38])

The propulsive force (in Babič’s notation, it is denoted as FzF_z) between the primary and second filament under the same configuration given in Example 1 was calculated. The results of calculation are shown in Table 2. Analysis of Table 2 depicts that all results of calculation obtained by new formula (33) and formula (A.17) derived by Grover’s method have very good agreement with results of calculation performed by Babič’s formula.

Example 3: Torque

Considering the scheme shown in Figure 3, the torques corresponding the generalized coordinates θ\theta and η\eta, namely, TθT_\theta and TηT_\eta, respectively, are calculated by using formula (34). The calculation was performed for the same coil arrangement as in Example 1. Changing the distance, d=zBd=z_B, between the coils’ planes in the same range between 0 and 11 mm, the obtained results of calculation of the torque TθT_\theta are depicted in Table 3. While the results of calculation of the torque TηT_\eta are shown in Table 4. The analysis of the tables show that the results of calculation are in excellent agreement to each other. Also, worth noting that calculation of the torque TηT_\eta by means of Kalantarov-Zeitlin’s method provides the zero result for all considered cases, while the Grover method shows small errors.

zB=dz_B=dGrover’s method,This work,
mmEq. (A.14), TθT_\theta, nN mEq. (34), TθT_\theta, nN m
000
1.0-0.1578740617610107-0.1578740617610108
2.0-0.3116544572018358-0.3116544572018354
3.0-0.4575094468248243-0.4575094468248247
4.0-0.5921003456361885-0.5921003456361873
5.0-0.7127567873444469-0.7127567873444469
6.0-0.8175809895015301-0.8175809895015308
7.0-0.9054774849564481-0.905477484956448
8.0-0.9761154129579873-0.9761154129579885
9.0-1.029838133074231-1.029838133074231
10.0-1.067538737088322-1.067538737088322
11.0-1.090520164249831-1.090520164249831

Table 3. Calculation of the torque TθT_\theta for Example 3

Example 4: Force (Example 6, page 74 in Babič et al. work [38])

Now, considering the coils having the same radii as in Example 1 and 2, but let us assume that the centre of secondary coil has the following coordinates: xB=yB=2.1213x_B = y_B = 2.1213 mm and zB=8.0z_B = 8.0 mm. It means that neither the YY-axis nor XX-axis coincides with the ρ\rho-axis. In this particular case, the values of coordinates xBx_B and yBy_B are corresponding to the Grover parameter

zB=dz_B=dGrover’s method,This work,
mmEq. (A.14), TψT_\psi, nN m(34), TηT_\eta, nN m
0−1.36317281085437×10−16-1.36317281085437 \times10^{-16}0
1.0−2.000664966411278×10−16-2.000664966411278 \times10^{-16}0
2.0−1.781112775806449×10−16-1.781112775806449 \times10^{-16}0
3.0−1.996595700073574×10−16-1.996595700073574 \times10^{-16}0
4.0−1.194158312920886×10−16-1.194158312920886 \times10^{-16}0
5.0−6.21450714145062×10−17-6.21450714145062 \times10^{-17}0
6.0−9.46986487725448×10−17-9.46986487725448 \times10^{-17}0
7.01.10795621370042×10−161.10795621370042 \times10^{-16}0
8.01.548091424456673×10−161.548091424456673 \times10^{-16}0
9.0−2.570095735195105×10−16-2.570095735195105 \times10^{-16}0
10.08.214737035699823×10−178.214737035699823 \times10^{-17}0
11.02.123122223811704×10−162.123122223811704 \times10^{-16}0

Table 4. Calculation of the torque TηT_\eta for Example 3

ρ=xB2+yB2=3.0\rho= \sqrt{x_B^2+y_B^2}=3.0 mm, which has the same value as in the previous examples 1 and 2 (please see Figure 4). The restoring force with its xx- and yy-components and propulsive force between the primary and second filament were calculated. The results are shown in Table 5 and 6, respectively.

Figure 4. Geometrical scheme of circular filaments with parallel axes corresponding to the general case, when neither the YY-axis nor XX-axis coincides with the ρ\rho-axis.

This work, (32)FxF_x, μNFyF_y, μNFx2+Fy2\sqrt{F_x^2+F_y^2}, μN
This work, Eq. (32)0.0317957760945440.0317957760945440.0449660177790823
Babič’s formula [38 Eq. (10)], FrF_r, μN0.0449660177790821
Grover’s method, Eq. (A.16), FρF_\rho, μN0.0449660177790823

Table 5. Calculation of the restoring force for Example 4

Babič’s formula, [38 Eq. (10)], FzF_z, μNGrover’s method, Eq. (A.17), FdF_d, μNThis work, Eq. (33), FzF_z, μN
−0.347887153545896−0.347887153545896−0.347887153545896

Table 6. Calculation of the propulsive force for Example 4

Example 5: Force (Example 8, page 75 in Babič et al. work [38])

The primary coil has a radius of 1 m, while the secondary of 0.5 m. The centre of the secondary coil with respect to the primary one is located at point BB having the following coordinate xB=yB=zB=2.0x_B=y_B=z_B=2.0 m. The results of calculation are as follows

This workFxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
−2.745371984357345−2.7453719843573493.509473102444028
Babič’s formulaFxF_x, nNFyF_y, nNFzF_z, nN
−2.745371984357346−2.7453719843573463.509473102444032
Grover’s method–FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–−3.8825422940372913.50947310244403
This work,Fx2+Fy2\sqrt{F_x^2+F_y^2}, nN:3.882542294037292–

Table 7.

Example 6. Force, the special case of θ=π/2\theta= \pi/2 (Example 9, page 76 in Babič et al. work [38])

The primary coil has a radius of 1 m, while the secondary of 0.5 m. The centre of the secondary coil with respect to the primary one is located at point BB having the following coordinate xB=1.0x_B = 1.0 m, yB=2.0y_B = 2.0 m, zB=3.0z_B = 3.0 m. The secondary coil is located on the plane x=1.0x = 1.0 m (θ=η=π/2\theta= \eta= \pi/2 rad). For the calculation, Eq. (43) and (44) are used. Results are

This workFxF_x, Eq. (43), nNFyF_y, Eq. (43), nNFzF_z, Eq. (44), nN
1.939241379554505−1.861181718234279−2.202382194552672
Babič’s formulaFxF_x, nNFyF_y, nNFzF_z, nN
1.939241379554508−1.861181718234281−2.202382194552672

Example 7. Torque (the singular case of Kalantarov-Zeitlin’s method)

For the same arrangement of coils as in Example 6 above, the torque TηT_\eta is calculated by formula (45). This case corresponds to the singularity of Kalantarov-Zeitlin’s method for the calculation of the torque TθT_\theta. Due to this fact the value is not available in the table below. The results are

This work, Eq. (45):TθT_\theta, nN mTηT_\eta, nN m
−−6.860953527497655
Grover’s method, Eq. (A.18):TθT_\theta, nN mTψT_\psi, nN m
−6.0364717317884686.860953527497664

However, to avoid this difficulty the angle θ\theta can be chosen enough close to the value π/2\pi/2, but not equal to it. Hence, formula (34) can be applied. For instance, if the angle θ\theta is 1.570 62 rad, the result of calculation of the torque TθT_\theta becomes −6.036 471 614 863 915 nN m with a relative error of −1.93696844616845 ×10−8\times10^{-8} in comparing with the exact result of calculation obtained by means of Grover’s method.

Example 8: Force, the special case of θ=π/2\theta= \pi/2 (Example 10, page 76 in Babič et al. work [38])

The primary and secondary coils have the same radii as in example 5. While, the centre of the secondary coil with respect to the primary one is located at point BB having the coordinate: xB=yB=zB=2.0 mx_B = y_B = z_B = 2.0\,\mathrm{m}. The secondary coil is located on the plane y=2.0 my = 2.0\,\mathrm{m} (θ=π/2 rad\theta= \pi/2\,\mathrm{rad} and η=0 rad\eta= 0\,\mathrm{rad}). For the calculation, as in the previous example Eq. (43) and (44) are used. The results are

This workFxF_x, (43), nNFyF_y, (43), nNFzF_z, (44), nN
4.9013981770523381.9848723132001362.582265710169335
The Babič formulaFxF_x, nNFyF_y, nNFzF_z, nN
4.9013981770523451.9848723132001372.582265710169336

Table 8.

Example 9: Force

The primary and secondary coils have the same radii as in example 5 (the primary coil has a radius of 1 m and the secondary one of 0.5 m). The centre of the secondary coil with respect to the primary one is located at point BB having the coordinates: xB=0x_B = 0 and yB=zB=2.0 my_B = z_B = 2.0\,\mathrm{m} (θ=0\theta= 0 and η=0 rad\eta= 0\,\mathrm{rad}). The results of calculation are as follows

This work:FxF_x, (32), nNFyF_y, (32), nNFzF_z, (33), nN
−2.943923360032078×10−15-2.943923360032078 \times10^{-15}−13.05164071847218-13.051640718472180.58360681028384260.5836068102838426
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–−13.05164071847219-13.051640718472190.5836068102838380.583606810283838

Table 9.

Example 10: Force, the special case of θ=π/2\theta= \pi/2

The primary and secondary coils have the same radii as in example 5 (the primary coil has a radius of 1 m and the secondary one of 0.5 m). The centre of the secondary coil with respect to the primary one is located at point BB having the coordinates: xB=0x_B = 0 and yB=zB=2.0 my_B = z_B = 2.0\,\mathrm{m}. While, the angular orientation of the secondary coil is defined as follows θ=π/2\theta= \pi/2 and η=0 rad\eta= 0\,\mathrm{rad}. The results of calculation are

FxF_x, (43), nNFyF_y, (43), nNFzF_z, (44), nN
This work:
−3.986562883376773×10−16-3.986562883376773 \times10^{-16}9.6248050106798212.81718822413886
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
−-9.62480501067982412.81718822413886

Table 1 (PDF p. 21).

Force and torque between inclined circular filaments with intersect axes

The general scheme of the arrangement of two inclined circular filaments whose axes intersect for calculation of the force and torque is shown in Figure 5. The centre BB of the secondary circle is located on the ZYZY- plane and the xyxy plane of the secondary circle is inclined by the θ\theta angle (η=0\eta= 0).

Figure 5. Geometrical scheme of circular filaments with intersect axes.

Example 11: Force

The primary and secondary coils have the same radii as in examples 7 and 8 (the primary coil has a radius of 1 m and the secondary one of 0.5 m). Also, similar to Example 7 and 8, the centre of the secondary coil with respect to the primary one is located at point BB having the coordinates: xB=0x_B = 0 and yB=zB=2.0y_B = z_B = 2.0 m, while the angel of θ\theta is changed in a range from π/12\pi/12 to 5π/125\pi/12 rad. Hence, the results of calculation of force are as follows.

For the angle of θ=π/12\theta= \pi/12 rad, results are

This work:
FxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
−3.925231146709438×10−15-3.925231146709438 \times10^{-15}−10.83934689818066-10.839346898180664.0853351958496234.085335195849623
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–−10.83934689818066-10.839346898180664.085335195849634.08533519584963

Table 10.

For the angle of θ=π/6\theta= \pi/6 rad, results are

This work:
FxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
−2.943923360032078×10−15-2.943923360032078 \times10^{-15}−7.552692503639927-7.5526925036399277.6331017476933137.633101747693313
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–−7.552692503639934-7.5526925036399347.6331017476933037.633101747693303

Table 11.

For the angle of θ=π/4\theta= \pi/4 rad, results are

This work:
FxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
−2.943923360032078×10−15-2.943923360032078 \times10^{-15}−3.307376210012321-3.30737621001232110.6882270992755210.68822709927552
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–−3.307376210012323-3.30737621001232310.688227099275510.6882270992755

Table 12.

For the angle of θ=π/3\theta= \pi/3 rad, results are

This work:
FxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
−9.813077866773594×10−16-9.813077866773594 \times10^{-16}1.4233673904911711.42336739049117112.6870457957939812.68704579579398
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–1.423367390491171.4233673904911712.6870457957939812.68704579579398

Table 13.

For the angle of θ=5π/12\theta= 5\pi/12 rad, results are

This work:
FxF_x, Eq. (32), nNFyF_y, Eq. (32), nNFzF_z, Eq. (33), nN
4.906538933386797×10−164.906538933386797 \times10^{-16}5.9345426931425685.93454269314256813.3495876036238213.34958760362382
The Grover method:FρF_\rho, Eq. (A.16), nNFdF_d, Eq. (A.17), nN
–5.9345426931425665.93454269314256613.349587603623813.3495876036238

Table 14.

Example 12: Torque

Now, for the same coil arrangement as in Example 10 above, the components of the torque were calculated. The results are as follows.

For the angle of θ=π/12\theta= \pi/12 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−16.91560720972092-16.91560720972092−1.103971260012029×10−15-1.103971260012029 \times10^{-15}
The Grover method,
Eq. (A.14):−16.91560720972094-16.915607209720943.496810701779198×10−153.496810701779198 \times10^{-15}

Table 15.

For the angle of θ=π/6\theta= \pi/6 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−18.05278139644676-18.05278139644676−1.717288626685379×10−15-1.717288626685379 \times10^{-15}
The Grover method,
Eq. (A.14):−18.05278139644677-18.052781396446772.99873918328272×10−162.99873918328272 \times10^{-16}

Table 16.

For the angle of θ=π/4\theta= \pi/4 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−17.23198131459169-17.231981314591694.599880250050123×10−154.599880250050123 \times10^{-15}
The Grover method,
Eq. (A.14):−17.23198131459168-17.23198131459168−2.386186316760677×10−16-2.386186316760677 \times10^{-16}

Table 17.

For the angle of θ=π/3\theta= \pi/3 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−14.34174801596063-14.341748015960633.925231146709438×10−153.925231146709438 \times10^{-15}
The Grover method,
Eq. (A.14):−14.34174801596063-14.34174801596063−1.034553191651082×10−15-1.034553191651082 \times10^{-15}

Table 18.

For the angle of θ=5π/12\theta= 5\pi/12 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−10.05276095457351-10.052760954573511.177569344012831×10−141.177569344012831 \times10^{-14}
The Grover method,
Eq. (A.14):−10.05276095457353-10.052760954573532.72021517637498×10−162.72021517637498 \times10^{-16}

Table 19.

Force and torque between circular filaments arbitrarily positioned in the space

The validation of the developed formulas, namely, for force calculation (32), (33) and for torque calculation (34) between circular filaments arbitrarily positioned in the space as shown in Figure 6 is performed by comparison with the results of calculation obtained by utilizing Grover’s method. The position of the secondary coil with respect to the primary one is determined by the linear and angular misalignment, in particular, the angular one is defined by the η\eta- and θ\theta-angle.

Figure 6. Geometrical scheme for calculation of force and torque between circular filaments arbitrarily positioned in the space: the angular misalignment is determined by the η\eta- and θ\theta-angle.

Example 13: Force. The primary and secondary circles have radii Rp=16.0R_p=16.0 cm and Rs=10.0R_s=10.0 cm, respectively. The centre of the secondary circle is located at xB=0x_B=0, yB=4.3301y_B=4.3301 cm, zB=17.5z_B=17.5 cm and the angle, θ\theta of 60.0°, but the angle η\eta is varied in a range from 0 to 360°. The arrangement corresponds to one considered in Babič et al. work [39] for calculation of mutual inductance. The results of calculation are summed

up below.

For the angle of η=0\eta= 0 rad, results are

This work:
FxF_x, Eq. (32), μ\muNFyF_y, Eq. (32), μ\muNFzF_z, Eq. (33), nN
−9.12816701829942×10−17-9.12816701829942 \times10^{-17}−0.178854052497651-0.178854052497651−0.120726650359416-0.120726650359416
The Grover method:FρF_\rho, Eq. (A.16), μ\muNFdF_d, Eq. (A.17), μ\muN
–−0.178854052497651-0.178854052497651−0.120726650359416-0.120726650359416

Table 20.

For the angle of η=π/6\eta= \pi/6 rad, results are

This work:
FxF_x, Eq. (32), μ\muNFyF_y, Eq. (32), μ\muNFzF_z, Eq. (33), nN
0.07396147213165610.0739614721316561−0.159981513393707-0.159981513393707−0.128840188997851-0.128840188997851
The Grover method:FρF_\rho, Eq. (A.16), μ\muNFdF_d, Eq. (A.17), μ\muN
–−0.159981513393707-0.159981513393707−0.128840188997851-0.128840188997851

Table 21.

For the angle of η=π/3\eta= \pi/3 rad, results are

This work:
FxF_x, Eq. (32), μ\muNFyF_y, Eq. (32), μ\muNFzF_z, Eq. (33), nN
0.1287683544877270.128768354487727−0.108439678353152-0.108439678353152−0.151814417341889-0.151814417341889
The Grover method:FρF_\rho, Eq. (A.16), μ\muNFdF_d, Eq. (A.17), μ\muN
–−0.108439678353152-0.108439678353152−0.151814417341889-0.151814417341889

Table 22.

For the angle of η=π/2\eta= \pi/2 rad, results are

This work:
FxF_x, Eq. (32), μ\muNFyF_y, Eq. (32), μ\muNFzF_z, Eq. (33), nN
0.1501801187212590.150180118721259−0.0376756210066879-0.0376756210066879−0.18544258034698-0.18544258034698
The Grover method:FρF_\rho, Eq. (A.16), μ\muNFdF_d, Eq. (A.17), μ\muN
–−0.037675621006688-0.037675621006688−0.18544258034698-0.18544258034698

Table 23.

For the angle of η=2π/3\eta= 2\pi/3 rad, results are For the angle of η=5π/6\eta= 5\pi/6 rad, results are

This work:
FxF_x, Eq. (32), µNFyF_y, Eq. (32), µNFzF_z, Eq. (33), nN
0.131748018061080.034039013278426−0.222277730943632
The Grover method:FρF_\rho, Eq. (A.16), µNFdF_d, Eq. (A.17), µN
−0.034039013278426−0.222277730943632

Table 24.

This work:
FxF_x, Eq. (32), µNFyF_y, Eq. (32), µNFzF_z, Eq. (33), nN
0.07689300044310870.0873707070288501−0.251739105122414
The Grover method:FρF_\rho, Eq. (A.16), µNFdF_d, Eq. (A.17), µN
−0.0873707070288498−0.251739105122414

Table 25.

For the angle of η=π\eta= \pi rad, results are

This work:
FxF_x, Eq. (32), µNFyF_y, Eq. (32), µNFzF_z, Eq. (33), nN
−2.38736675863216×10−16-2.38736675863216 \times10^{-16}0.107078490953807−0.263132411165356
The Grover method:FρF_\rho, Eq. (A.16), µNFdF_d, Eq. (A.17), µN
−0.107078490953807−0.263132411165356

Table 26.

For the angle of η=7π/6\eta= 7\pi/6 rad, results are

This work:
FxF_x, Eq. (32), µNFyF_y, Eq. (32), µNFzF_z, Eq. (33), nN
−0.07689300044310880.08737070702885−0.251739105122414
The Grover method:FρF_\rho, Eq. (A.16), µNFdF_d, Eq. (A.17), µN
−0.0873707070288499−0.251739105122414

Table 27.

For the angle of η=4π/3\eta= 4\pi/3 rad, results are

This work:
FxF_x, Eq. (32), µNFyF_y, Eq. (32), µNFzF_z, Eq. (33), nN
−0.131748018061080.0340390132784261−0.222277730943632
The Grover method:FρF_\rho, Eq. (A.16), µNFdF_d, Eq. (A.17), µN
−0.0340390132784261−0.222277730943632

Table 28.

For the angle of η=3π/2\eta= 3\pi/2 rad, results are For the angle of η=5π/3\eta=5\pi/3 rad, results are

This work:
FxF_x, Eq. (32), μNFyF_y, Eq. (32), μNFzF_z, Eq. (33), nN
−0.150180118721259−0.0376756210066879−0.18544258034698
The Grover method:FρF_\rho, Eq. (A.16), μNFdF_d, Eq. (A.17), μN
−−0.0376756210066879−0.18544258034698

Table 1 (PDF p. 27).

This work:
FxF_x, Eq. (32), μNFyF_y, Eq. (32), μNFzF_z, Eq. (33), nN
−0.128768354487727−0.108439678353152−0.151814417341889
The Grover method:FρF_\rho, Eq. (A.16), μNFdF_d, Eq. (A.17), μN
−−0.108439678353152−0.151814417341889

Table 2 (PDF p. 27).

For the angle of η=11π/6\eta=11\pi/6 rad, results are

This work:
FxF_x, Eq. (32), μNFyF_y, Eq. (32), μNFzF_z, Eq. (33), nN
−0.0739614721316564−0.159981513393707−0.128840188997851
The Grover method:FρF_\rho, Eq. (A.16), μNFdF_d, Eq. (A.17), μN
−−0.159981513393707−0.128840188997851

Table 3 (PDF p. 27).

For the angle of η=2π\eta=2\pi rad, results are

This work:
FxF_x, Eq. (32), μNFyF_y, Eq. (32), μNFzF_z, Eq. (33), nN
−2.24693341988909×10−16-2.24693341988909 \times10^{-16}−0.178854052497651−0.120726650359416
The Grover method:FρF_\rho, Eq. (A.16), μNFdF_d, Eq. (A.17), μN
−−0.178854052497651−0.120726650359416

Table 4 (PDF p. 27).

Example 14: Torque

For the same arrangement of coils as in Example 13 above, the torque is calculated. Results are shown below.

For the angle of θ=0\theta=0 rad, results are For the angle of η=π/6\eta=\pi/6 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−35.67279850151469-35.67279850151469−2.10650008114602×10−15-2.10650008114602 \times10^{-15}
The Grover method,
Eq. (A.14):−35.67279850151468-35.672798501514687.820828205289753×10−167.820828205289753 \times10^{-16}

Table 29.

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−34.95015525890449-34.950155258904493.2026057047728453.202605704772845
The Grover method,
Eq. (A.14):−34.95015525890452-34.950155258904523.2026057047728393.202605704772839

Table 30.

For the angle of η=π/3\eta=\pi/3 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−32.95959317516494-32.959593175164945.5757985176730815.575798517673081
The Grover method,
Eq. (A.14):−32.95959317516493-32.959593175164935.5757985176730755.575798517673075

Table 31.

For the angle of η=π/2\eta=\pi/2 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−30.16106072467219-30.161060724672196.5029493207492496.502949320749249
The Grover method,
Eq. (A.14):−30.16106072467218-30.161060724672186.5029493207492476.502949320749247

Table 32.

For the angle of η=2π/3\eta=2\pi/3 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−27.19599671483476-27.195996714834765.7048209300628335.704820930062833
The Grover method,
Eq. (A.14):−27.19599671483475-27.195996714834755.7048209300628285.704820930062828

Table 33.

For the angle of η=5π/6\eta=5\pi/6 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−24.8616946084914-24.86169460849143.3295438121870643.329543812187064
The Grover method,
Eq. (A.14):−24.86169460849138-24.861694608491383.3295438121870523.329543812187052

Table 34.

For the angle of η=π\eta=\pi rad, results are For the angle of η=7π/6\eta= 7\pi/6 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−23.96174184275239-23.961741842752395.617333549722721×10−155.617333549722721 \times10^{-15}
The Grover method,
Eq. (A.14):−23.96174184275239-23.961741842752394.791446029057778×10−154.791446029057778 \times10^{-15}

Table 35.

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−24.8616946084914-24.8616946084914−3.329543812187048-3.329543812187048
The Grover method,
Eq. (A.14):−24.86169460849138-24.86169460849138−3.329543812187046-3.329543812187046

Table 36.

For the angle of η=4π/3\eta= 4\pi/3 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−27.19599671483477-27.19599671483477−5.704820930062826-5.704820930062826
The Grover method,
Eq. (A.14):−27.19599671483473-27.19599671483473−5.704820930062823-5.704820930062823

Table 37.

For the angle of η=3π/2\eta= 3\pi/2 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−30.16106072467218-30.16106072467218−6.502949320749249-6.502949320749249
The Grover method,
Eq. (A.14):−30.16106072467218-30.16106072467218−6.502949320749249-6.502949320749249

Table 38.

For the angle of η=5π/3\eta= 5\pi/3 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−32.95959317516493-32.95959317516493−5.575798517673075-5.575798517673075
The Grover method,
Eq. (A.14):−32.95959317516497-32.95959317516497−5.575798517673072-5.575798517673072

Table 39.

For the angle of η=11π/6\eta= 11\pi/6 rad, results are

TθT_{\theta}, nN mTηT_{\eta}, nN m
This work,
Eq. (34):−34.95015525890449-34.95015525890449−3.202605704772844-3.202605704772844
The Grover method,
Eq. (A.14):−34.95015525890452-34.95015525890452−3.202605704772843-3.202605704772843

Table 40.

For the angle of η=2π\eta= 2\pi rad, results are

This work,TθT_{\theta}, nN mTηT_{\eta}, nN m
Eq. (34):−35.6727985015147-35.6727985015147−1.40433338743068×10−15-1.40433338743068 \times10^{-15}
The Grover method,TθT_{\theta}, nN mTψT_{\psi}, nN m
Eq. (A.14):−35.6727985015147-35.67279850151473.112983652841607×10−163.112983652841607 \times10^{-16}

Table 41.

Conclusion

We derived new formulas, namely, (32), (33), (34), (43), (44) and (45) for calculation of force and torque between two circular filaments arbitrarily oriented in space presented in the integral analytical form, whose kernel function is expressed in terms of the elliptic integrals of the first and second kinds. In particular, formulas (43), (44) and (45) are applied for the special case (θ=π/2\theta= \pi/2 rad), when the circular filaments are mutually perpendicular to each other. For calculation of the torque TθT_{\theta}, this special case corresponds to the singularity one. To avoid this difficulty, it is suggested that the angle θ\theta can be chosen enough close to the value π/2\pi/2, but not equal to it and calculation is performed by means of formula (34). Thus, the developed formulas are applicable for all possible arrangements between two circular filaments.

New developed formulas have been successfully validated through a number of examples available in the literature and direct comparison with results of calculation of force and torque performed by expressions derived by Grover’s method. Besides, the obtained formulas can be easily programmed, they are intuitively understandable for application.

Acknowledgment

Kirill Poletkin acknowledges with thanks the support from German Research Foundation (Grant KO 1883/37-1) under the priority programme SPP 2206.

Geometrical scheme of arbitrarily oriented circular filaments: the Grover notations

Figure A.7. Geometrical scheme of arbitrarily oriented circular filaments: the Grover notations.

Calculation of force and torque between two arbitrarily oriented circular filaments using Grover’s formula of mutual inductance

According to Grover’s notations, the linear misalignment of the centre of the secondary circle is characterised by two parameters, namely, d=zBd = z_B and ρ=xB2+yB2\rho= \sqrt{x_B^2 + y_B^2} as shown in Figure A.7. Besides that the angular misalignment is defined in accordance with the first manner as shown in Figure 2(a), but keeping the original Grover’s notation the angle, η\eta, is replaced by ψ\psi. In absence of the angular misalignment, the CF xyzxyz assigned to the secondary circle is oriented in the following way. The zz-axis is directed upward along the dd-line, while the yy-axis is parallel to the ρ\rho-line and directed in continuation of the ρ\rho-line. Then adopting the above considered notations, Grover’s formula for calculation of mutual inductance between two circular filaments can be written as

M=μ0RpRs2π∫02πU⋅Ψ(k) dφ,(46)M = \frac{\mu_0\sqrt{R_p R_s}}{2\pi}\int_{0}^{2\pi} U\cdot\Psi(k)\,\mathrm{d}\varphi, \tag*{(46)}

where

U=U(γ,θ,ψ)=R(γ,θ,ψ)V1.5=cos⁡θ−γ(cos⁡ψcos⁡φ−sin⁡ψcos⁡θsin⁡φ)V1.5,(47)U = U(\gamma,\theta,\psi) = \frac{R(\gamma,\theta,\psi)}{V^{1.5}} = \frac{\cos\theta-\gamma(\cos\psi\cos\varphi-\sin\psi\cos\theta\sin\varphi)}{V^{1.5}}, \tag*{(47)}
V=V(γ,θ,ψ)=1−cos⁡(φ)2sin⁡(θ)2+2γ(sin⁡ψsin⁡φ−cos⁡φcos⁡ψcos⁡θ)+γ2,(48)V = V(\gamma,\theta,\psi) = \sqrt{1-\cos(\varphi)^2\sin(\theta)^2+2\gamma(\sin\psi\sin\varphi-\cos\varphi\cos\psi\cos\theta)+\gamma^2}, \tag*{(48)}
Ψ(k)=2k[(1−k22)K(k)−E(k)],(49)\Psi(k)=\frac{2}{k}\left[\left(1-\frac{k^2}{2}\right)K(k)-E(k)\right], \tag*{(49)}
k2=k2(γ,Δ,θ,ψ)=4αV(αV+1)2+z2,(50)k^2=k^2(\gamma,\Delta,\theta,\psi)=\frac{4\alpha V}{(\alpha V+1)^2+z^2}, \tag*{(50)}

α=Rs/Rp\alpha=R_s/R_p, Δ=d/Rp\Delta=d/R_p, γ=ρ/Rs\gamma=\rho/R_s, z=Δ−αsin⁡θcos⁡φz=\Delta-\alpha\sin\theta\cos\varphi.

The kernel of formula (A.1) is

Kr=U⋅Ψ(k).(51)\mathrm{Kr}=U\cdot\Psi(k). \tag*{(51)}

Accounting for (A.2), (A.3), (A.4) and (A.5), the ρ\rho-derivative of the kernel becomes

∂Kr∂ρ=∂Kr∂γ1Rs=1Rs⋅[∂U∂γ⋅Ψ(k)+U⋅dΨ(k)dk⋅∂k∂γ],(52)\frac{\partial\mathrm{Kr}}{\partial\rho}=\frac{\partial\mathrm{Kr}}{\partial\gamma}\frac{1}{R_s}=\frac{1}{R_s}\cdot\left[\frac{\partial U}{\partial\gamma}\cdot\Psi(k)+U\cdot\frac{d\Psi(k)}{dk}\cdot\frac{\partial k}{\partial\gamma}\right], \tag*{(52)}

where

∂U∂γ=(∂R∂γ⋅V−1.5⋅R⋅∂V∂γ)V2.5,∂R∂γ=−(cos⁡ψcos⁡φ−sin⁡ψcos⁡θsin⁡φ),∂V∂γ=sin⁡ψsin⁡φ−cos⁡φcos⁡ψcos⁡θ+γV,∂k∂γ=2/k−k(αV+1)(αV+1)2+z2⋅α∂V∂γ,(53)\begin{aligned} \frac{\partial U}{\partial\gamma}&=\left(\frac{\partial R}{\partial\gamma}\cdot V-1.5\cdot R\cdot\frac{\partial V}{\partial\gamma}\right)V^{2.5},\\ \frac{\partial R}{\partial\gamma}&=-(\cos\psi\cos\varphi-\sin\psi\cos\theta\sin\varphi),\\ \frac{\partial V}{\partial\gamma}&=\frac{\sin\psi\sin\varphi-\cos\varphi\cos\psi\cos\theta+\gamma}{V},\\ \frac{\partial k}{\partial\gamma}&=\frac{2/k-k(\alpha V+1)}{(\alpha V+1)^2+z^2}\cdot\alpha\frac{\partial V}{\partial\gamma}, \tag*{(53)} \end{aligned}
dΨ(k)dk=2k2[2−k22(1−k2)E(k)−K(k)].(54)\frac{d\Psi(k)}{dk}=\frac{2}{k^2}\left[\frac{2-k^2}{2(1-k^2)}E(k)-K(k)\right]. \tag*{(54)}

The dd-derivative of the kernel is

∂Kr∂d=∂Kr∂Δ1Rp=1Rp⋅U⋅dΨ(k)dk⋅∂k∂Δ,(55)\frac{\partial\mathrm{Kr}}{\partial d}=\frac{\partial\mathrm{Kr}}{\partial\Delta}\frac{1}{R_p}=\frac{1}{R_p}\cdot U\cdot\frac{d\Psi(k)}{dk}\cdot\frac{\partial k}{\partial\Delta}, \tag*{(55)}

where

∂k∂Δ=−k⋅z(αV+1)2+z2∂z∂Δ,(56)\frac{\partial k}{\partial\Delta}=-\frac{k\cdot z}{(\alpha V+1)^2+z^2}\frac{\partial z}{\partial\Delta}, \tag*{(56)}
∂z∂Δ=1.\frac{\partial z}{\partial\Delta}=1.

Note that in Eq. (A.10) the kk-derivative of Ψ(k)\Psi(k) is defined similarly as in Eq. (A.9). The derivatives of the kernel with respect to the angular coordinates are as follows. The θ\theta-derivative is

∂Kr∂θ=∂U∂θ⋅Ψ(k)+U⋅dΨ(k)dk⋅∂k∂θ,(57)\frac{\partial\mathrm{Kr}}{\partial\theta}=\frac{\partial U}{\partial\theta}\cdot\Psi(k)+U\cdot\frac{d\Psi(k)}{dk}\cdot\frac{\partial k}{\partial\theta}, \tag*{(57)}

where

∂U∂θ=(∂R∂θ⋅V−1.5⋅R⋅∂V∂θ)/V2.5,∂R∂θ=−sin⁡θ⋅(1+γ⋅sin⁡φsin⁡ψ),∂V∂θ=−sin⁡θ⋅(cos⁡2(φ)cos⁡θ−γ⋅cos⁡φcos⁡ψ)V,∂k∂θ=[2/k−k(αV+1)]⋅α∂V∂θ−k⋅z⋅∂z∂θ(αV+1)2+z2,∂z∂θ=−α⋅cos⁡θcos⁡φ.(58)\begin{aligned} \frac{\partial U}{\partial\theta} &= \left(\frac{\partial R}{\partial\theta}\cdot V-1.5\cdot R\cdot\frac{\partial V}{\partial\theta}\right)/V^{2.5},\\ \frac{\partial R}{\partial\theta} &= -\sin\theta\cdot(1+\gamma\cdot\sin\varphi\sin\psi),\\ \frac{\partial V}{\partial\theta} &= \frac{-\sin\theta\cdot\left(\cos^{2}(\varphi)\cos\theta-\gamma\cdot\cos\varphi\cos\psi\right)}{V},\\ \frac{\partial k}{\partial\theta} &= \frac{\left[2/k-k(\alpha V+1)\right]\cdot\alpha\frac{\partial V}{\partial\theta}-k\cdot z\cdot\frac{\partial z}{\partial\theta}}{(\alpha V+1)^{2}+z^{2}},\\ \frac{\partial z}{\partial\theta} &= -\alpha\cdot\cos\theta\cos\varphi. \tag*{(58)} \end{aligned}

The ψ\psi-derivative is

∂Kr∂ψ=∂U∂ψ⋅Ψ(k)+U⋅dΨ(k)dk⋅∂k∂ψ,(59)\frac{\partial K_{\mathrm r}}{\partial\psi} = \frac{\partial U}{\partial\psi}\cdot\Psi(k) + U\cdot\frac{\mathrm d\Psi(k)}{\mathrm dk}\cdot\frac{\partial k}{\partial\psi}, \tag*{(59)}

where

∂U∂ψ=(∂R∂ψ⋅V−1.5⋅R⋅∂V∂ψ)/V2.5,∂R∂ψ=γ⋅(cos⁡φsin⁡ψ+sin⁡φcos⁡ψcos⁡θ),∂V∂ψ=γ⋅(sin⁡φcos⁡ψ+cos⁡φsin⁡ψcos⁡θ)V,∂k∂ψ=2/k−k(αV+1)(αV+1)2+z2⋅α∂V∂ψ,∂z∂ψ=0.(60)\begin{aligned} \frac{\partial U}{\partial\psi} &= \left(\frac{\partial R}{\partial\psi}\cdot V-1.5\cdot R\cdot\frac{\partial V}{\partial\psi}\right)/V^{2.5},\\ \frac{\partial R}{\partial\psi} &= \gamma\cdot(\cos\varphi\sin\psi+\sin\varphi\cos\psi\cos\theta),\\ \frac{\partial V}{\partial\psi} &= \frac{\gamma\cdot(\sin\varphi\cos\psi+\cos\varphi\sin\psi\cos\theta)}{V},\\ \frac{\partial k}{\partial\psi} &= \frac{2/k-k(\alpha V+1)}{(\alpha V+1)^{2}+z^{2}}\cdot\alpha\frac{\partial V}{\partial\psi},\\ \frac{\partial z}{\partial\psi} &= 0. \tag*{(60)} \end{aligned}

Using the derivatives of the kernel obtained above, we can write the first derivatives of Grover’s formula of mutual inductance with respect to the appropriate coordinates. Taking into account Eq. (A.7), the ρ\rho-derivative of Grover’s formula (A.1) is

∂M∂ρ=μ02πRpRs∫02π(∂U∂γ⋅Ψ(k)+U⋅dΨ(k)dk⋅∂k∂γ)dφ.(61)\frac{\partial M}{\partial\rho} = \frac{\mu_{0}}{2\pi}\sqrt{\frac{R_{p}}{R_{s}}} \int_{0}^{2\pi} \left( \frac{\partial U}{\partial\gamma}\cdot\Psi(k) + U\cdot\frac{\mathrm d\Psi(k)}{\mathrm dk}\cdot\frac{\partial k}{\partial\gamma} \right)\mathrm d\varphi. \tag*{(61)}

Accounting for (A.10), the dd-derivative of Grover’s formula (A.1) can be written as

∂M∂d=μ02πRsRp∫02πU⋅dΨ(k)dk⋅∂k∂Δ dφ.(62)\frac{\partial M}{\partial d} = \frac{\mu_{0}}{2\pi}\sqrt{\frac{R_{s}}{R_{p}}} \int_{0}^{2\pi} U\cdot\frac{\mathrm d\Psi(k)}{\mathrm dk}\cdot\frac{\partial k}{\partial\Delta}\,\mathrm d\varphi. \tag*{(62)}

Considering Eq. (A.12) and Eq. (A.14), first derivatives of Grover’s formula (A.1) of with respect to angular coordinates can be written

∂M∂g=μ0RpRs2π∫02π(∂U∂g⋅Ψ(k)+U⋅dΨ(k)dk⋅∂k∂g)dφ.(63)\frac{\partial M}{\partial g} = \frac{\mu_{0}\sqrt{R_{p}R_{s}}}{2\pi} \int_{0}^{2\pi} \left( \frac{\partial U}{\partial g}\cdot\Psi(k) + U\cdot\frac{\mathrm d\Psi(k)}{\mathrm dk}\cdot\frac{\partial k}{\partial g} \right)\mathrm d\varphi. \tag*{(63)}

where g=θg = \theta and ψ\psi, respectively.

The derived formulas (A.16), (A.17) and (A.18) can be easily programmed by using, for instance, Matlab language. The file with Matlab code with implemented formulas is available for a reader in supplementary materials to this article.

References

  1. [1]E. Okress, D. Wroughton, G. Comenetz, P. Brace, J. Kelly, Electromagnetic levitation of solid and molten metals, Journal of Applied Physics 23 (5) (1952) 545–552.
  2. [2]Y. M. Urman, Theory for the calculation of the force characteristics of an electromagnetic suspension of a superconducting body, Technical Physics 42 (1) (1997) 1–6. doi:10.1134/1.1258645. URL https://doi.org/10.1134/1.1258645
  3. [3]Y. M. Urman, Calculation of the force characteristics of a multi-coil suspension of a superconducting sphere, Technical Physics 42 (1) (1997) 7–13. doi:10.1134/1.1258654. URL https://doi.org/10.1134/1.1258654
  4. [4]M. W. Coffey, Mutual inductance of superconducting thin films, Journal of Applied Physics 89 (10) (2001) 5570–5577.DOI
  5. [5]Y. Urman, S. Kuznetsov, Translational transformations of tensor solutions of the helmholtz equation and their application to describe interactions in force fields of various physical nature, Quarterly of Applied Mathematics 72 (1) (2014) 1–20.DOI
  6. [6]U.-M. Jow, M. Ghovanloo, Design and optimization of printed spiral coils for efficient transcutaneous inductive power transmission, IEEE Transactions on biomedical circuits and systems 1 (3) (2007) 193–202.
  7. [7]Y. P. Su, X. Liu, S. Y. R. Hui, Mutual inductance calculation of movable planar coils on parallel surfaces, IEEE Transactions on Power Electronics 24 (4) (2009) 1115–1123. doi:10.1109/TPEL.2008.2009757.DOI
  8. [8]S. Y. Chu, A. T. Avestruz, Transfer-power measurement: A non-contact method for fair and accurate metering of wireless power transfer in electric vehicles, in: 2017 IEEE 18th Workshop on Control and Modeling for Power Electronics (COMPEL), 2017, pp. 1–8. doi:10.1109/COMPEL.2017.8013344.DOI
  9. [9]A. Shiri, A. Shoulaie, A new methodology for magnetic force calculations between planar spiral coils, Progress In Electromagnetics Research 95 (2009) 39–57.DOI
  10. [10]R. Ravaud, G. Lemarquand, V. Lemarquand, Force and stiffness of passive magnetic bearings using permanent magnets. part 1: Axial magnetization, IEEE transactions on magnetics 45 (7) (2009) 2996.DOI
  11. [11]S. Obata, A muscle motion solenoid actuator, Electrical Engineering in Japan 184 (2) (2013) 10–19.DOI
  12. [12]R. Shalati, K. V. Poletkin, J. G. Korvink, V. Badilita, Novel concept of a series linear electromagnetic array artificial muscle, Journal of Physics: Conference Series 1052 (1) (2018) 012047. URL http://stacks.iop.org/1742-6596/1052/i=1/a=012047DOI
  13. [13]K. Poletkin, A. I. Chernomorsky, C. Shearwood, U. Wallrabe, An analytical model of micromachined electromagnetic inductive contactless suspension, in: the ASME 2013 International Mechanical Engineering Congress & Exposition, ASME, San Diego, California, USA, 2013, pp. V010T11A072–V010T11A072. doi:10.1115/IMECE2013–66010. URL http://dx.doi.org/10.1115/IMECE2013-66010
  14. [14]K. Poletkin, A. I. Chernomorsky, C. Shearwood, U. Wallrabe, A qualitative analysis of designs of micromachined electromagnetic inductive contactless suspension, International Journal of Mechanical Sciences 82 (2014) 110–121. doi:10.1016/j.ijmecsci.2014.03.013. URL http://authors.elsevier.com/sd/article/S0020740314000897
  15. [15]Z. Lu, K. Poletkin, B. den Hartogh, U. Wallrabe, V. Badilita, 3D micro-machined inductive contactless suspension: Testing and modeling, Sensors and Actuators A Physical 220 (2014) 134–143. doi:10.1016/j.sna.2014.09.017. URL http://dx.doi.org/10.1016/j.sna.2014.09.017
  16. [16]K. Poletkin, Z. Lu, U. Wallrabe, J. Korvink, V. Badilita, Stable dynamics of micro-machined inductive contactless suspensions, International Journal of Mechanical Sciences 131-132 (2017) 753 – 766. doi:https://doi.org/10.1016/j.ijmecsci.2017.08.016. URL http://www.sciencedirect.com/science/article/pii/S0020740316306555
  17. [17]V. Vlnieska, A. Voigt, S. Wadhwa, J. Korvink, M. Kohl, K. Poletkin, Development of control circuit for inductive levitation micro-actuators, Proceedings 64 (1). doi:10.3390/1eCAT2020–08479. URL https://www.mdpi.com/2504-3900/64/1/39DOI
  18. [18]K. V. Poletkin, A. I. Chernomorsky, C. Shearwood, A proposal for micro-machined accelerometer, base on a contactless suspension with zero spring constant, IEEE Sensors J. 12 (07) (2012) 2407–2413. doi:10.1109/JSEN.2012.2188831.
  19. [19]K. V. Poletkin, R. Shalati, J. G. Korvink, V. Badilita, Pull-in actuation in hybrid micro-machined contactless suspension, Journal of Physics: Conference Series 1052 (1) (2018) 012035. URL http://stacks.iop.org/1742-6596/1052/i=1/a=012035DOI
  20. [20]K. V. Poletkin, J. G. Korvink, Modeling a pull-in instability in micro-machined hybrid contactless suspension 7 (1) (2018) 11.DOI
  21. [21]K. Poletkin, Static pull-in behavior of hybrid levitation micro-actuators: simulation, modelling and experimental study, IEEE/ASME Transactions on Mechatronicsdoi:10.1109/TMECH.2020.2999516. URL https://doi.org/10.1109/TMECH.2020.2999516
  22. [22]T. Theodoulidis, R. J. Ditchburn, Mutual impedance of cylindrical coils at an arbitrary position and orientation above a planar conductor, IEEE Transactions on Magnetics 43 (8) (2007) 3368–3370. doi:10.1109/TMAG.2007.894559.DOI
  23. [23]M. Sawan, S. Hashemi, M. Sehili, F. Awwad, M. Hajj-Hassan, A. Khouas, Multicoils-based inductive links dedicated to power up implantable medical devices: modeling, design and experimental results, Biomedical Microdevices 11 (5) (2009) 1059. doi:10.1007/s10544-009-9323-7. URL https://doi.org/10.1007/s10544-009-9323-7
  24. [24]S. Kuznetsov, J. K. Guest, Topology optimization of magnetic source distributions for diamagnetic and superconducting levitation, Journal of Magnetism and Magnetic Materials 438 (2017) 60 – 69. doi:https://doi.org/10.1016/j.jmmm.2017.04.052. URL http://www.sciencedirect.com/science/article/pii/S0304885316319515
  25. [25]D. Hoult, B. Tomanek, Use of mutually inductive coupling in probe design, Concepts in Magnetic Resonance 15 (4) (2002) 262–285. arXiv:https://onlinelibrary.wiley.com/doi/pdf/10.1002/cmr.10047, doi:10.1002/cmr.10047. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/cmr.10047DOI
  26. [26]N. Spengler, P. T. While, M. V. Meissner, U. Wallrabe, J. G. Korvink, Magnetic lenz lenses improve the limit-of-detection in nuclear magnetic resonance, PLOS ONE 12 (8) (2017) 1–17. doi:10.1371/journal.pone.0182779. URL https://doi.org/10.1371/journal.pone.0182779
  27. [27]G. D. Angelis, V. Pasku, A. D. Angelis, M. Dionigi, M. Mongiardo, A. Moschitta, P. Carbone, An indoor ac magnetic positioning system, IEEE Transactions on Instrumentation and Measurement 64 (5) (2015) 1267–1275. doi:10.1109/TIM.2014.2381353.DOI
  28. [28]F. Wu, J. Jeon, S. K. Moon, H. J. Choi, H. Son, Voice coil navigation sensor for flexible silicone intubation, IEEE/ASME Transactions on Mechatronics 21 (2) (2016) 851–859. doi:10.1109/TMECH.2015.2476836.DOI
  29. [29]X. Zhang, C. Quan, Z. Li, Mutual inductance calculation of circular coils for an arbitrary position with electromagnetic shielding in wireless power transfer systems, IEEE Transactions on Transportation Electrification (2021) 1–1doi:10.1109/TTE.2021.3054762.DOI
  30. [30]B. Gulbahar, A communication theoretical analysis of multiple-access channel capacity in magneto-inductive wireless networks, IEEE Transactions on Communications 65 (6) (2017) 2594–2607. doi:10.1109/TCOMM.2017.2669995.DOI
  31. [31]J. C. Maxwell, A Treatise on Electricity and Magnetism, 3rd Edition, Vol. 2, Dover Publications Inc., 1954.
  32. [32]C. Snow, Formulas for computing capacitance and inductance, Vol. 544, US Govt. Print. Off., 1954.
  33. [33]K.-B. Kim, E. Levi, Z. Zabar, L. Birenbaum, Restoring force between two noncoaxial circular coils, IEEE Transactions on Magnetics 32 (2) (1996) 478–484.DOI
  34. [34]F. W. Grover, The calculation of the mutual inductance of circular filaments in any desired positions, Proceedings of the IRE 32 (10) (1944) 620–629. doi:10.1109/JRPROC.1944.233364.DOI
  35. [35]S. I. Babic, C. Akyel, Y. Ren, W. Chen, Magnetic force calculation between circular coils of rectangular cross section with parallel axes for superconducting magnet, Progress In Electromagnetics Research 37 (2012) 275–288. doi:10.2528/PIERB11110508. URL http://dx.doi.org/10.2528/PIERB11110508
  36. [36]S. I. Babic, C. Akyel, Torque calculation between circular coils with inclined axes in air, International Journal of Numerical Modelling: Electronic Networks, Devices and Fields 24 (3) (2011) 230–243.DOI
  37. [37]S. I. Babic, C. Akyel, New formulas for calculating torque between filamentary circular coil and thin wall solenoid with inclined axis whose axes are at the same plane, Progress In Electromagnetics Research 73 (2018) 141–151.DOI
  38. [38]S. Babic, C. Akyel, Magnetic force between inclined circular filaments placed in any desired position, IEEE transactions on magnetics 48 (1) (2012) 69–80.DOI
  39. [39]S. Babic, F. Sirois, C. Akyel, C. Girardi, Mutual inductance calculation between circular filaments arbitrarily positioned in space: Alternative to Grover’s formula, IEEE Transactions on Magnetics 46 (9) (2010) 3591–3600. doi:10.1109/TMAG.2010.2047651. URL https://doi.org/10.1109/TMAG.2010.2047651
  40. [40]P. L. Kalantarov, L. A. Zeitlin, Raschet induktivnostey (Calculation of Inductances), 3rd Edition, Energoatomizdat, Leningrad, 1986.
  41. [41]K. V. Poletkin, J. G. Korvink, Efficient calculation of the mutual inductance of arbitrarily oriented circular filaments via a generalisation of the Kalantarov-Zeitlin method, Journal of Magnetism and Magnetic Materials 483 (2019) 10–20. doi:https://doi.org/10.1016/j.jmmm.2019.03.078. URL https://www.sciencedirect.com/science/article/pii/S0304885318337703
  42. [42]F. W. Grover, Inductance calculations : working formulas and tables, special ed. prepared for instrument society of america Edition, Research Triangle Park, N.C. : Instrument Society of America, 1981, reprint. Originally published: New York : Van Nostrand, 1946. With publisher’s comment.

Paper details

vlnc:W_bp6f8u3k
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents