Beyond the dipole approximation: A compact operator form to describe magnetizable many-body systems Preprint: APS/123-QED
Dirk Romeis · 2026
Read Beyond the dipole approximation: A compact operator form to describe magnetizable many-body systems Preprint: APS/123-QED by Dirk Romeis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-13647.txt.
Introduction
Romeis moves past the dipole approximation for magnetically soft particle systems. Simple dipole models badly understate forces when particles draw close, while full-field methods cost too much compute. Starting from the exact two-body solution, he builds an analytic scheme for many-body interactions in a compact operator form that looks dipole-like. It captures cluster formation and dispersal in applied fields, and it is easy to implement.
Catalog record
| CREATOR | Dirk Romeis |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22062 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.13647
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Beyond the dipole approximation: A compact operator form to describe magnetizable many-body systems Preprint: APS/123-QED Dirk Romeis Abstract To describe the interactions in magnetically soft particle systems either numerical full-field methods or dipole models are used. Whereas the former are computationally challenging, simple dipole interactions are largely underestimating the actual forces when particles get closer. Based on the full 2-body solution, an analytic approximation scheme for many-body full-field interactions is developed. The concept is formulated in terms of an improved operator that is equivalent to the classical dipole form. The full interaction operator allows to describe cluster formation and dispersion among particles in applied magnetic fields very compactly and highly efficient. In view of its simple ’dipole-like’ form, the implementation is straightforward in many areas where magnetically soft objects are used. Introduction – Magnetic multidomain particles often exhibit paramagnetic behavior 23 ; 28 ; 8 ; 3 ; 6 ; 5 ; 24 . Such soft magnetic particles are common in many areas of research and technologies, where they are used as a part of magnetorheological fluids, gels or elastomers 2 ; 12 ; 13 . They are also important in biomedical applications or soft robotics, as well as for manufacturing soft magnetic composite elements used in power electronics 1 ; 20 ; 28 . Iron-based microparticles are typically employed for reasons such as cost-effective manufacturing, availability, excellent magnetic properties and biocompatibility. Furthermore, they are often of spherical shape 23 ; 28 ; 8 and, thus, most models consider magnetizable spheres 18 ; 22 ; 19 ; 17 ; 25 ; 15 ; 27 ; 11 ; 26 ; 10 ; 29 ; 9 . Due to their magnetization behavior, they are very sensitive to the local magnetic field. When two or more such particles approach, they noticeably influence each other. This is because the local magnetic field depends strongly on the exact arrangement of the particles 3 ; 6 ; 5 . In order to describe the interactions among them, a common way is to assign each particle a single point dipole located in the particle center. It is known that when multidomain particles are getting close to each other, the dipole approximation is no longer valid as it strongly underestimates the true interactions 3 ; 6 ; 5 ; 24 . The discrepancy is caused by magnetization inhomogeneities inside each particle. This is schematically shown in Figs. 1 (a) and 3 (a). Beyond dipole approximations, elaborate full-field methods 18 ; 22 ; 19 ; 17 are used to capture these near-field effects. These more precise calculations are computationally very intensive. Already the simplest case of two linearly magnetizable spheres in close contact, requires hundreds of terms to achieve relative accuracy in the subpercentage range 3 ; 6 . This impressively demonstrates the challenges to be faced in such many-body systems. Therefore, dipole approximations remain a common practical choice to model magnetic interactions in many-body systems 25 ; 15 ; 27 ; 11 ; 26 ; 10 ; 29 ; 9 . Considering monodisperse spherical particles in the linear magnetization regime, and inspired by 30 , a highly efficient alternative description is developed below. It is practically equivalent to the dipole formulation, but takes into account full-field interactions. Self-consistent dipole description – By applying a magnetic field \vec{H}_{0} , a single sphere becomes homogeneously magnetized with magnetization \vec{M}\parallel\vec{H}_{0} . This creates an opposite demagnetization field of -\vec{M}/3 inside its volume 7 . Outside, a dipole field is induced, emanating from the center of the particle 7 . Linear magnetization behavior is typically described as \vec{M}=\chi\vec{H} , where \chi is the magnetic susceptibility and \vec{H} the total magnetic field 7 . If N linearly magnetizable spheres are exposed to \vec{H}_{0} , the system of equations in the dipole approximation becomes 26 : \vec{M}_{i}=\chi_{\text{eff}}\bigg(\vec{H}_{0}+\sum\limits_{\underset{j\neq i}{j=1}}^{N}\hat{g}(\vec{r}_{ij})\cdot\vec{M}_{j}\bigg). (1) Here, \chi_{\text{eff}}=\chi/(\chi+1/3) denotes the effective susceptibility 7 ; 26 , \vec{M}_{i} the homogeneous magnetization of particle i , and the spatial vector \vec{r}_{ij} connects the centers of particles i and j . The central quantity is the dipole operator \hat{g} . In dimensionless form it reads: \hat{g}(\vec{r})=\frac{v_{\text{p}}}{4\pi}\left(-\frac{\hat{I}}{r^{3}}+\frac{3\vec{r}\vec{r}}{r^{5}}\right)=a^{3}\left(g_{I}(r)\hat{I}+g_{R}(r)\hat{R}\right). (2) The volume of the sphere with radius a is v_{\text{p}}=4\pi a^{3}/3 . Furthermore, \hat{I} denotes the identity tensor, and \vec{r}\vec{r} the outer product. In the following, the particle radius is set to unity, i.e., a=1 . On the right-hand side in Eq. ( 2 ) we introduce the shorthand notation \hat{R}:=\frac{\vec{r}\vec{r}}{r^{2}} , and the coefficients are g_{I}=-\frac{1}{3r^{3}} and g_{R}=\frac{1}{r^{3}} . The solution of Eq. ( 1 ) yields the self-consistent \vec{M}_{i} for a given particle arrangement. With known \vec{M}_{i} the total magnetic energy and the subsequent magnetic forces acting on each of the particles are determined straightforwardly 19 . In the case of just two interacting spheres, where due to symmetry \vec{M}_{1}=\vec{M}_{2}=\vec{M} , the equations reduce to: \vec{M}=\chi_{\text{eff}}\left(\vec{H}_{0}+\hat{g}(\vec{r})\cdot\vec{M}\right). (3) Note, all fields, like \vec{M} or \vec{H} , are considered homogeneous over the entire volume of each individual particle i . Thus, per definition they equivalently represent the volume average, i.e., \vec{M}_{i}=\langle\vec{M}_{i}\rangle or \vec{H}_{i}=\langle\vec{H}_{i}\rangle . This assumption is valid if the spheres are not too close to each other 3 ; 6 ; 5 . Full-field solution for two interacting spheres – When two spheres in applied magnetic field come close to each other, pronounced inhomogeneities in the magnetization field appear. This is exemplified by the darker shading on the spheres in FIG. 1 (a). The exact solution for a system of two linearly magnetizable spheres Figure 1: (a) Schematic visualization of magnetization inhomogeneities inside the volumes of two closely spaced spheres when an external magnetic field \vec{H}_{0} is applied. The direction of \vec{H}_{0} defines the x -axis. (b) Equivalent illustration in terms of the average formulation. The identical shade of red indicates that both spheres have the same average magnetization \langle\vec{M}\rangle . Vector \vec{r} connects the centers of the spheres. is obtained in terms of a slowly converging series expansion. Biller et al. 3 ; 6 developed a concise function that consists of just a few terms and reproduces the magnetic energy of the two linearly magnetizable spheres in applied field. In general, the magnetic energy of a system of linearly magnetizable particles in external field is given as 5 ; 25 : \displaystyle U_{\text{mag}} \displaystyle=-\frac{\mu_{0}}{2}\int_{V_{\text{S}}}\!\!d^{3}r\sum_{i=1}^{N}\vec{M}_{i}(\vec{r})\cdot\vec{H}_{0} (4) \displaystyle=-\frac{\mu_{0}}{2}H_{0}v_{\text{p}}\sum_{i=1}^{N}\langle M_{i\,x}\rangle \displaystyle\overset{N=2}{=}-\mu_{0}H_{0}v_{\text{p}}\langle M_{x}\rangle. (5) Here, \mu_{0} is the vacuum permeability, and the integral in the first line is taken over the entire system volume V_{\text{S}} where the applied field \vec{H}_{0}=H_{0}\vec{e}_{x} is assumed to be homogeneous. Outside of the particles the material, e.g., polymer or carrier liquid, is non-magnetic. Thus, the integral over V_{\text{S}} turns to N separate integrals over the respective particle volumes. With homogeneous \vec{H}_{0} , each of these integrals yields the average magnetization \langle\vec{M}_{i}\rangle in the respective particle times the volume v_{\text{p}} . The scalar product with \vec{H}_{0} then returns the x -component \langle M_{i\,x}\rangle . Due to symmetry in the case N=2 , see Eq. ( 5 ), the average magnetization in each particle must be the same, i.e., \langle\vec{M}_{1}\rangle=\langle\vec{M}_{2}\rangle=\langle\vec{M}\rangle . Comparing the function to resample the magnetic energy for two particles in 3 ; 6 with the general form in Eq. ( 5 ), we identify: \langle M_{x}\rangle=\chi_{\text{eff}}\,\mathcal{F}_{xx}H_{0}. (6) Here, we introduced the solution factor