betaFeedback

WorksPhysics and Astronomy

Full analytical solution for the magnetic field of uniformly magnetized cylinder tiles

Abstract

We present an analytical solution for the magnetic field of a homogeneously magnetized cylinder tile and by extension solutions for full cylinders, rings, cylinder sectors and ring segments. The derivation is done by direct integration in the magnetic surface charge picture. Results are closed-form expressions and elliptic integrals. All special cases are treated individually, which enables the field computation for all possible position arguments. An implementation is provided in Python together with a performance analysis. The implementation is tested against numerical solutions and applied to compute the magnetic field in a discrete Halbach cylinder.

Introduction

Permanent magnets are widely used in numerous technical applications [1, 2, 3, 4, 5, 6, 7, 8]. Among all conceivable geometries, simple magnet shapes such as cuboids or cylinders are especially popular and widespread as they are mass-produced and available at short notice in various dimensions, magnetization and materials [9, 10]. When the magnet geometry is simple and the magnetization is homogeneous, it is often possible to derive analytical expressions for the field generated by such magnets [11, 12, 13, 14, 15, 16, 17, 18, 19, 20]. These expressions provide excellent approximations to the real fields when modern, high-grade magnetic materials, like SmCo, NdFeB or ferrites with susceptibilities χ<0.1\chi< 0.1 are involved [8, 21].

The main advantage over common numerical methods such as finite element (FE) approaches or direct numerical integration is the fast computation times of the order of microseconds [21, 22], which enables highly efficient multivariate parameter space analysis and solving global optimization problems for permanent magnet arrangements [23, 24, 25].

The analytical formulas for cuboid magnets [8, 11, 26, 27] are particularly useful since cuboids can be trivially subdivided into smaller cuboid cells to which the formulas are again applicable. This allows modeling of the material response with the magnetostatic method of moments [28], a common approach of computing dipole interactions in micromagnetic software [29, 30, 31, 32, 33, 34, 35]. In contrast, the analytical expression of a cylinder [15, 16, 17, 18, 19, 20] is not sufficient for such calculations, as the cylindrical volume element is a cylinder tile, i.e., a cylindrical ring segment. In addition, many geometric shapes used in magnetic position systems, such as the cylindrical quadrupole [36, 37] or rotary encoder wheels [38, 39], can be constructed with such tiles, motivating us to find analytical expressions for this geometry.

Finding an analytical expression for the homogeneously magnetized cylinder tile was attempted in many previous works [40, 41, 42, 43, 44, 45], but to date all solutions contain integral expressions which can only be solved numerically. In this work we present for the first time a complete analytical solution for the magnetic field of a homogeneously magnetized cylinder tile. All expressions contain only basic arithmetic operations, trigonometric functions and special functions like elliptic integrals, which can be computed very efficiently [46, 47, 48, 49, 50, 51, 52, 53, 54, 55]. In previous works, only the most general case is treated, while indeterminate forms, that typically appear at surface extensions, are neglected. We provide expressions for all such cases, which allows us to compute the field at any point in space. In addition, we include the limiting cases with inner radius r1=0r_1 = 0 (cylinder sector), constricting sector angles φ1=0\varphi_1 = 0, φ2=2π\varphi_2 = 2\pi (cylinder ring) and the full cylinder as the combination of the two cases.

The structure of the paper is as follows: In Sec. 2 we establish the notation and describe the surface charge method, which enables us to express the total magnetic field as the sum of various surface integrals. The first and second integration steps are shown in Sec. 3 and Sec. 4, respectively. In Sec. 5 we validate our results by comparison with

[email protected] (F. Slanovc)

ORCID(s): 0000-0002-2335-0076 (F. Slanovc) solutions from numerical integration and finite element method. A discrete Halbach cylinder application example is demonstrated in § 6. We close with a brief discussion of the results and outlook in § 7.

Method

It is convenient to describe the problem in cylindrical coordinates (r,φ,z)∈[0,∞)×[0,2π)×(−∞,∞)(r,\varphi,z) \in[0,\infty)\times[0,2\pi)\times(-\infty,\infty). The cylinder tile is defined through six boundaries at r1<r2r_1 < r_2, φ1<φ2\varphi_1 < \varphi_2 and z1<z2z_1 < z_2, see Figure 1a. We define the homogeneous magnetization vector M\mathbf{M}, through its magnitude MM, azimuth angle φM\varphi_M and polar angle θM\theta_M, see Figure 1b.

Dimensions of the cylinder tile in cylinder coordinates and magnetization vector

Figure 1. (a) Dimensions of the cylinder tile in cylinder coordinates and the resulting magnetic surface charges σ\sigma. We assume 0≤r1<r20 \le r_1 < r_2, φ1<φ2\varphi_1 < \varphi_2 and z1<z2z_1 < z_2. (b) Magnetization vector M\mathbf{M} described by the three coordinates MM, φM\varphi_M and θM\theta_M.

In Cartesian coordinates the magnetization is then given by

M=M(cos⁡φMsin⁡θMex+sin⁡φMsin⁡θMey+cos⁡θMez).(1)\mathbf{M}=M\left(\cos\varphi_M\sin\theta_M\mathbf{e}_x+\sin\varphi_M\sin\theta_M\mathbf{e}_y+\cos\theta_M\mathbf{e}_z\right). \tag*{(1)}

Field calculation with magnetostatic potential

In [56 Sec. 5.9 C], the scalar magnetic potential ΦM\Phi_M is derived for volumes VV with homogeneous magnetization M\mathbf{M} in the absence of electric currents,

ΦM(r)=14π∮∂Vσ(r′)∣r−r′∣ dA′.(2)\Phi_M(\mathbf{r})=\frac{1}{4\pi}\oint_{\partial V}\frac{\sigma(\mathbf{r}')}{\lvert\mathbf{r}-\mathbf{r}'\rvert}\,\mathrm{d}A'. \tag*{(2)}

with the magnetic surface charge σ\sigma given by the inner product of the magnetization vector with the surface normal vector n\mathbf{n},

σ(r′)=M(r′)⋅n(r′).(3)\sigma(\mathbf{r}')=\mathbf{M}(\mathbf{r}')\cdot\mathbf{n}(\mathbf{r}'). \tag*{(3)}

The magnetic field H\mathbf{H} is then given by

H(r)=−∇ΦM(r)=14π∮∂Vσ(r′)(r−r′)∣r−r′∣3 dA′.(4)\mathbf{H}(\mathbf{r})=-\nabla\Phi_M(\mathbf{r})=\frac{1}{4\pi}\oint_{\partial V}\frac{\sigma(\mathbf{r}')(\mathbf{r}-\mathbf{r}')}{\lvert\mathbf{r}-\mathbf{r}'\rvert^3}\,\mathrm{d}A'. \tag*{(4)}

Magnetic surface-charge density

The tile boundary ∂V\partial V includes three types of surfaces. Each type is denoted by its normal vector, see Figure 1a. We calculate the magnetic surface charge density σ\sigma according to Eq. (3) in cylindrical coordinates for each surface type.

σri(φ′)=(−1)iM⋅(cos⁡φ′ex+sin⁡φ′ey)=(−1)iMsin⁡θM(cos⁡φMcos⁡φ′+sin⁡φMsin⁡φ′)=(−1)iMsin⁡θMcos⁡(φM−φ′)(5)\begin{aligned} \sigma_{r_i}(\varphi') &= (-1)^i \mathbf{M} \cdot\left(\cos\varphi' \mathbf{e}_x + \sin\varphi' \mathbf{e}_y\right) \\ &= (-1)^i M \sin\theta_M \left(\cos\varphi_M \cos\varphi' + \sin\varphi_M \sin\varphi'\right) \\ &= (-1)^i M \sin\theta_M \cos(\varphi_M-\varphi') \tag*{(5)} \end{aligned}
σφj=(−1)jM⋅(−sin⁡φjex+cos⁡φjey)=(−1)jMsin⁡θM(−cos⁡φMsin⁡φj+sin⁡φMcos⁡φj)=(−1)jMsin⁡θMsin⁡(φM−φj)(6)\begin{aligned} \sigma_{\varphi_j} &= (-1)^j \mathbf{M} \cdot\left(-\sin\varphi_j \mathbf{e}_x + \cos\varphi_j \mathbf{e}_y\right) \\ &= (-1)^j M \sin\theta_M \left(-\cos\varphi_M \sin\varphi_j + \sin\varphi_M \cos\varphi_j\right) \\ &= (-1)^j M \sin\theta_M \sin(\varphi_M-\varphi_j) \tag*{(6)} \end{aligned}
σzk=(−1)kM⋅ez=(−1)kMcos⁡θM(7)\begin{aligned} \sigma_{z_k} &= (-1)^k \mathbf{M} \cdot\mathbf{e}_z \\ &= (-1)^k M \cos\theta_M \tag*{(7)} \end{aligned}

It will be useful for our derivation to omit magnetization and powers of (−1)(-1) and define the dashed (sign reduced) surface charge densities by

σri′(φ′)=sin⁡θMcos⁡(φM−φ′)(8)\sigma'_{r_i}(\varphi') = \sin\theta_M \cos(\varphi_M-\varphi') \tag*{(8)}
σφj′=sin⁡θMsin⁡(φM−φj)(9)\sigma'_{\varphi_j} = \sin\theta_M \sin(\varphi_M-\varphi_j) \tag*{(9)}
σzk′=cos⁡θM.(10)\sigma'_{z_k} = \cos\theta_M. \tag*{(10)}

Surface integrals

To express the the surface integrals (4) in cylindrical coordinates, we use the usual transformations,

x=rcos⁡φ,(11)x = r\cos\varphi, \tag*{(11)}
y=rsin⁡φ,(12)y = r\sin\varphi, \tag*{(12)}
er=cos⁡φ ex+sin⁡φ ey,(13)\mathbf{e}_r = \cos\varphi\,\mathbf{e}_x + \sin\varphi\,\mathbf{e}_y, \tag*{(13)}
eφ=−sin⁡φ ex+cos⁡φ ey.(14)\mathbf{e}_\varphi= -\sin\varphi\,\mathbf{e}_x + \cos\varphi\,\mathbf{e}_y. \tag*{(14)}

To simplify the notation, we do not explicitly write the dependency of the coordinates rr, φ\varphi and zz, although almost every term depends on them. With the abbreviation

ξ(r′,φ′,z′):=∣r−r′∣=r2+r′2−2rr′cos⁡(φ−φ′)+(z−z′)2,(15)\xi(r',\varphi',z') := \lvert\mathbf{r}-\mathbf{r}'\rvert= \sqrt{r^2+r'^2-2rr'\cos(\varphi-\varphi')+(z-z')^2}, \tag*{(15)}

every field component can be expressed as a sum of six terms, each originating from one tile surface, as

References

  1. [1]K. J. Srnat, Proceedings of the IEEE 78 (1990) 923-946.
  2. [2]R. H. J. Fastenau, E. J. van Loenen, Journal of Magnetism and Magnetic Materials 157–158 (1996) 1-6.
  3. [3]Z. J. J. Stekly, IEEE Transactions on Applied Superconductivity 10 (2000) 873-878.
  4. [4]J. M. D.Coey, Journal of Magnetism and Magnetic Materials 248 (2002) 441-456.
  5. [5]R. W. McCallum, L.H. Lewis, R. Skomski, M.J. Kramer, I.E. Anderson Annual Review of Materials Research 44 (2014) 451-477.
  6. [6]S. Hirosawa1, M. Nishino1, S. Miyashita, Adv. Nat. Sci: Nanosci. Nanotechnol. 8 (2017) 013002.
  7. [7]J. M. D.Coey, Engineering 6 (2020) 119-131.
  8. [8]P. Malagò, F. Slanovc, S. Herzog, S. Lumetti, T. Schaden, A. Pellegrinetti, M. Moridi, C. Abert, D. Suess, M. Ortner, Sensors 20 (2020) 6873.
  9. [9]C. Treutler, Sensors Actuators A 91 (2001) 2–6.
  10. [10]U. Ausserlechner, A. Satz, F. Gastinger, US Patent (2014) 8,717,010.
  11. [11]J. Camacho, V. Sosa, Rev Mexicana Fís E 59 (2013) 8–17.
  12. [12]R. I. Joseph, E. Schlömann, Journal of Applied Physics 36 (1965) 1579.
  13. [13]A. Smith, K. K. Nielsen, D. V. Christensen, C. R. H. Bahl, R. Bjørk, J. Hattel, Journal of Applied Physics 107 (2010) 103910.
  14. [14]J. Prat-Camps, C. Navau, A. Sanchez, and D.-X. Chen, Journal of Applied Physics 107 (2010) 103910.
  15. [15]A. Caciagli, R. J. Baars, A. P. Philipse, B. W. M. Kuipers, Journal of Magnetism and Magnetic Materials 456 (2018) 423–432.
  16. [16]N. Derby, S. Obert, American Journal of Physics 78 (2010) 229.
  17. [17]E. Furlani, S. Reznik, A. Kroll, IEEE Trans Magn 31 (1995) 844–51.
  18. [18]R. Ravaud, G. Lemarquand, S. Babic, V. Lemarquand, C. Akyel, IEEE Transactions on Magnetics 46 (2010) 3585–3590.
  19. [19]J. T. Conway, IEEE Transactions on Magnetics 46 (2009) 75–81.
  20. [20]J. Selvaggi, S. Salon, M. Chari, Progress in electromagnetics research (2010) 244–51.
  21. [21]M. Ortner, L. G. C. Bandeira, SoftwareX 11 (2020) 100466.
  22. [22]J. W. Jansen, J. P. C. Smeets, T. T. Overboom, J. M. M. Rovers, E. A. Lomonova, IEEE Transactions on Magnetics 50 (2014) 1-7.
  23. [23]M. Ortner, 9th International Conference on Sensing Technology (ICST 2015) 359-364.
  24. [24]M. Ortner, M. Ribeiro , D. Spitzer, IEEE Transactions on Magnetics 55 (2019) 1-4.
  25. [25]M. Galea, L. Papini, H. Zhang, C. Gerada, T. Hamiti, IEEE Transactions on Magnetics 51 (2015) 1-9.
  26. [26]Z. J. Yang, T. H. Johansen, H. Bratsberg, G. Helgesen, A. T. Skjeltorp, Superconductor Science and Technology 3 (1990) 591-597.
  27. [27]R. Engel-Herbert, T. Hesjedal, Journal of Applied Physics 97 (2005) 074504.
  28. [28]O. Chadebec, J.-L. Coulomb, F. Janet, IEEE Transactions on Magnetics 42 (2006) 515-520.
  29. [29]M. J. Donahue, D. G. Porter, OOMMF user’s guide, version 1.0. interagency report, NISTIR 6376, national institute of standards and technology, gaithersburg, MD (1999).DOI
  30. [30]C. Abert. A mumagnum–fd—A finite-difference/fft package for the solution of dynamical micromagnetic problems. https:// github. com/ micro magne tics/ magnum. fd. (2013).
  31. [31]P. Heistracher, F. Bruckner, C. Abert, C. Vogler, D. Suess, Journal of Magnetism and Magnetic Materials 503 (2020) 166592.
  32. [32]C. Abert, F. Bruckner, C. Vogler, R. Windl, R. Thanhoffer, D. Suess, Journal of Magnetism and Magnetic Materials 387 (2015) 13–18.
  33. [33]C. Abert, L. Exl, F. Bruckner, A. Drews, D. Suess, Journal of Magnetism and Magnetic Materials 345 (2013) 29-35.
  34. [34]A. Vansteenkiste, B. Van de Wiele, Journal of Magnetism and Magnetic Materials 323 (2011) 2585-2591.
  35. [35]R. Bjørk, E. Poulsen, K. K. Nielsen, A. R. Insinga, Journal of Magnetism and Magnetic Materials 535 (2021) 168057.
  36. [36]C. Huber, C. Abert, F. Bruckner, M. Groenfeld, S. Schuschnigg, I. Teliban, C. Vogler, G. Wautischer, R. Windl, D. Suess, Scientific Reports 7 (2017) 9419.
  37. [37]N. Dupré, O. Dubrulle, S. Huber, J. W. Burssens, C. Schott, G. Close, Proceedings 2 (2018) 763.
  38. [38]U. Ausserlechner, Progress In Electromagnetics Research B 38 (2012) 71-105.
  39. [39]Requirements for the technical representation of magnetic measurement scales in design drawings, E DIN SPEC 91411:2021-03 (2021).
  40. [40]R. Ravaud, G. Lemarquand, V. Lemarquand, C. Depollier, IEEE Transactions on Magnetics 44 (2008) 1982-1989.
  41. [41]R. Ravaud, G. Lemarquand, Progress In Electromagnetics Research B 24 (2010) 17–32.
  42. [42]R. Ravaud, G. Lemarquand, Progress In Electromagnetics Research 94 (2009) 327–341.
  43. [43]R. Ravaud, G. Lemarquand,, V. Lemarquand, IEEE Transactions on Magnetics 45 (2009) 2920–2926.
  44. [44]S. Song, B. Li, W. Qiao, C. Hu, H. Ren, H. Yu, Q. Zhang, M. Q.-H. Meng, G. Xu, IEEE Transactions on Magnetics 50 (2014) 1-11.
  45. [45]K. K. Nielsen, R. Bjørk, Journal of Magnetism and Magnetic Materials 507 (2020) 166799.
  46. [46]T. Fukushima, Journal of Computational and Applied Mathematics 236 (2012) 1961-1975.
  47. [47]T. Fukushima, IEEE 22nd Symposium on Computer Arithmetic (2015).
  48. [48]T. Fukushima, Numerische Mathematik 116 (2010) 687–719.
  49. [49]T. Fukushima, Journal of Computational and Applied Mathematics 235 (2011) 4140–4148.
  50. [50]T. Fukushima, Journal of Computational and Applied Mathematics 236 (2011) 1961–1975.
  51. [51]T. Fukushima, Celestial Mechanics and Dynamical Astronomy 105 (2009) 305.
  52. [52]T. Fukushima, H. Ishizaki, Celestial Mechanics and Dynamical Astronomy 59 (1994) 237–251.
  53. [53]T. Fukushima, Mathematics of computation 80 (2011) 1725-1743.
  54. [54]T. Fukushima, Celestial Mechanics and Dynamical Astronomy 105 (2009) 245.
  55. [55]T. Fukushima, Journal of Computational and Applied Mathematics 282 (2015) 71-76.
  56. [56]J. D. Jackson, Classical Electrodynamics, Third Edition, John Wiley & Sons, Inc. (1999).
  57. [57]E. Jones, T. Oliphant, P. Peterson, SciPy: Open source scientific tools for Python, 2001-, http://www.scipy.org/.
  58. [58]Ansys® Academic Research Mechanical, Release 18.1.
  59. [59]H. Raich, P. Blümler, Concepts in Magnetic Resonance Part B Magnetic Resonance Engineering 23B (2004) 16-25.
  60. [60]P. V. Trevizoli, J. A. Lozano, G. F. Peixer, J. R. Barbosa Jr., Journal of Magnetism and Magnetic Materials, Volume 395 (2015) 109-122.
  61. [61]D. Zhu, S. P. Beeby, J. Tudor, N. R. Harris, Smart Materials and Structures. 21 (2012) 075020.
  62. [62]Wolfram Research, Inc., Mathematica, Version 10.1, Champaign, IL (2015).
  63. [63]A. Rich, P. Scheibe, and N. M. Abbasi, Journal of Open Source Software 3 (2018) 1073.
  64. [64]I. S. Gradshteyn, I. M. Ryzhik, Table of Integrals, Series, and Products, Seventh Edition, Academic Press (2007).
  65. [65]R. Bulirsch, Numerische Mathematik 7 (1965) 78-90.
  66. [66]R. Bulirsch, Numerische Mathematik 13 (1969) 305 – 315.

Paper details

vlnc:W_jnpyaz9f
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