How can fluctuating fields predict forces between oddly shaped objects?

Turning field fluctuations into force

The Casimir interaction is often introduced using idealized flat plates, where established formulas make the geometry manageable. Real objects can curve, have irregular boundaries, or differ in their electromagnetic response. Davidovich's paper asks how to formulate the calculation for two bodies with arbitrary shapes without starting from a plate-specific expression.

The central objects are electromagnetic Green functions: mathematical descriptions of how a field at one location responds to a source elsewhere. The author uses vacuum tensor Green functions and writes integral equations of the Lippmann–Schwinger type for fields inside and around the bodies. In this formulation, each body's fluctuating polarization currents act as sources, and the fields produced by both bodies are related through the Green functions. The correlations of those random fields—not just their average values—carry the information needed for a dispersion force.

The paper then connects the correlations to two familiar force descriptions. One is the Maxwell stress tensor, which packages electromagnetic momentum flow; integrating its stress over a closed surface around a body gives the force on it. The other is the Lorentz force on charges and currents in the material. The author presents the two routes as consistent with momentum conservation, while noting that the Lorentz-force expression can be used in a medium where magnetic induction is defined.

A dipole-based route

To construct the correlations, the paper also considers exciting the bodies with electric and magnetic dipoles. The response fields are obtained by solving integral equations over the volumes of the bodies. The article mentions standard numerical approaches, including Galerkin and Ritz-type methods, for solving these equations. Correlation relations for the excited fields can then be used in the force calculation. In principle, the actual boundary and volume integrals encode the objects' shapes rather than requiring a universal flat-plate formula.

This is a framework for setting up calculations, not a catalogue of solved shapes. The paper does not report a numerical force for a specified irregular geometry or a benchmark showing how the method compares with another calculation. Its conclusion says that explicit expressions require choosing a particular configuration; flat screens or layers are suggested as cases where simpler results could be derived.

What the assumptions leave open

The derivation makes simplifying choices. The initial treatment uses homogeneous, isotropic bodies and assumes constant material properties within each body; it also neglects self-action and initially leaves out bulk charge densities. The paper says the approach can be generalized to inhomogeneous or anisotropic structures, and discusses magnetic response, but those generalizations are not worked through as detailed examples here.

Thermal radiation and zero-point fluctuations both appear in the formulation. However, interactions between bodies at different temperatures are not treated in detail. The author notes that, for highly heated structures far apart, thermal fluctuations can matter more than vacuum fluctuations, citing carbon nanotubes as an example. That observation is not a calculation of a new force for a specific nanotube pair in this paper.

The proposal therefore broadens how one might formulate Casimir-force calculations across geometries, while leaving much of the practical work to future applications: choose real material responses and a shape, solve the integral equations, and evaluate the resulting force. The paper is an arXiv preprint; its presence there does not establish peer review.