Inverse freeform design of a parallel-to-two-target reflector system Journal: opticajournal Publication type: Research Article

P. A. Braam, 1,* J. H. M. ten Thije Boonkkamp, 1 M. J. H. Anthonissen, 1 K. Mitra, 1 R. Beltman, 2 and W. L. IJzerman 1,2 · 2025

Read Inverse freeform design of a parallel-to-two-target reflector system Journal: opticajournal Publication type: Research Article by P. A. Braam, 1,* J. H. M. ten Thije Boonkkamp, 1 M. J. H. Anthonissen, 1 K. Mitra, 1 R. Beltman, 2 and W. L. IJzerman 1,2 (2025) 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-2503-17199.txt.

Introduction

Pieter Braam and five co-authors present a design method for a two reflector lighting system in a 2025 paper. The goal is to turn a parallel beam of light into two different light patterns at two target planes, using two freeform reflectors. The reflectors must control both where each ray lands and the direction it takes.

The authors derive generating functions and equations that link reflector shape to the desired light maps. A three stage least squares algorithm solves the system numerically. They also derive a condition that guarantees the reflectors do not intersect themselves. Worked examples confirm the condition and show the method can realize complex target patterns.

Catalog record

CREATORP. A. Braam, 1,* J. H. M. ten Thije Boonkkamp, 1 M. J. H. Anthonissen, 1 K. Mitra, 1 R. Beltman, 2 and W. L. IJzerman 1,2
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-17199

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Inverse freeform design of a parallel-to-two-target reflector system Journal: opticajournal Publication type: Research Article P. A. Braam, 1,* J. H. M. ten Thije Boonkkamp, 1 M. J. H. Anthonissen, 1 K. Mitra, 1 R. Beltman, 2 and W. L. IJzerman 1,2 Abstract We present an inverse method for transforming a given parallel light emittance to two light distributions at different parallel target planes using two freeform reflectors. The reflectors control both the spatial and directional target coordinates of light rays. To determine the shape and position of the reflectors, we derive generating functions and use Jacobian equations to find the optical mappings to the two targets. The model is solved numerically by a three-stage least-squares algorithm. A feasibility condition is derived, which ensures that the reflectors are not self-intersecting. Several examples validate this condition and demonstrate the algorithm’s capability of realizing complex distributions. 1 Introduction The goal in non-imaging optics is to design optical surfaces that convert a source light distribution to a target distribution. The two most commonly used methods are ray-trace methods and inverse methods. Ray-trace methods entail an iterative procedure where light rays propagate through the optical system and based on the deviation from the desired target distribution, the shape(s) of the optical surface(s) are altered [ 1 , 2 , 3 ] . Although these methods can be computationally expensive and often involve adjusting system parameters to gradually improve performance, recently developed methods have shown ways to improve the accuracy and reduction of computation time [ 4 , 5 ] . On the other hand, inverse methods provide a promising alternative approach by using the principles of geometrical optics and conservation of energy to compute the shapes of optical surfaces directly [ 6 ] . Inverse methods can be classified into two categories: methods that use optimization strategies and methods that directly solve a (generalized) Monge-Ampère equation or generated Jacobian equation; for a comprehensive overview, see [ 7 , p. 133] . Doskolovich et al. used methods from the first category and solved an optimal transport problem by reducing it to a linear assignment problem [ 8 ] . Methods from the second category were used by Brix et al. who solved the Monge-Ampère equation using a collocation method [ 9 , 10 ] . Feng et al. computed solutions to this equation using the theory of viscosity solutions [ 11 , 12 , 13 ] and Kawecki et al. used a finite element method [ 14 ] . Finally, Caboussat et al. developed a least-squares method to compute solutions to the Monge-Ampère equation [ 15 ] , which Prins et al. adapted for transport boundary conditions [ 16 ] . This method was extended by Yadav et al. taking multiple freeform surfaces into account [ 17 , 18 ] , and subsequently by Romijn et al. taking generating functions as input [ 7 , 19 ] and it is therefore also known as the generating least-squares method . The parallel-to-parallel reflector system converts a parallel emittance from a source \mathcal{S} to a parallel irradiance at a target \mathcal{T} using two reflectors. This system was studied by Yadav et al. [ 20 ] and Van Roosmalen et al. [ 21 , 22 ] . To find the desired reflectors, they introduced source coordinates \bm{x} and target coordinates \bm{y} , from which they derived the distance u_{1}(\bm{x}) from the source to the first reflector and the distance u_{2}(\bm{y}) from the target to the second reflector. The generating least-squares method is based on this formulation and uses a generating function of the form \displaystyle u_{1}(\bm{x})=G(\bm{x},\bm{y},u_{2}(\bm{y});V), (1) where V is the optical path length between the source and the target. In the parallel-to-parallel reflector system, the optical path length V is constant. In the parallel-to-two-target reflector system, the directions of rays vary at the first target and hence, V is a function of the target coordinates \bm{y} . The generating function can then be generalized to \displaystyle u_{1}(\bm{x})=G(\bm{x},\bm{y},u_{2}(\bm{y});V(\bm{y})). (2) The reflectors can be used to control the spreading of light rays at the targets and create two different light distributions for two separate targets. f g_{1} g_{2} \mathcal{R}_{1} \mathcal{R}_{2} \mathcal{S} \mathcal{T}_{1} \mathcal{T}_{2} Figure 1 : Parallel-to-two-target reflector system. In this paper we extend the generating least-squares method by modeling the two-target two-reflector system, illustrated in Fig. 1 , where light rays originate from a parallel light source \mathcal{S} with a specified emittance distribution f indicated by the blue dashed line, then reflect via two reflectors \mathcal{R}_{1} and \mathcal{R}_{2} , and finally reach the two targets \mathcal{T}_{1} and \mathcal{T}_{2} . The goal is to find the shapes of the reflectors such that the light rays originating from \mathcal{S} with emittance f reach \mathcal{T}_{1} and \mathcal{T}_{2} with light distributions g_{1} and g_{2} , respectively, indicated by the green and red dashed lines. As not all choices of g_{1} and g_{2} result in feasible reflectors, we will derive a feasibility condition to determine when they are not self-intersecting. Notably, Feng et al. also designed an optical system to control the spatial and directional target coordinates of light rays, but used a point source instead of a parallel source and lens surfaces instead of reflectors [ 23 ] . To find the shapes of the reflectors in the parallel-to-two-target reflector system, we apply a three-stage algorithm. In the first stage of the algorithm, we use energy conservation to determine the optical mapping from \mathcal{T}_{1} to \mathcal{T}_{2} . In the second stage, we use this mapping to compute the optical path length V from \mathcal{S} to \mathcal{T}_{1} . In the final stage, we use the optical path length in a generated Jacobian equation and apply energy conservation to determine both the mapping from \mathcal{S} to \mathcal{T}_{1} and the shapes of the reflectors. This paper is structured as follows. In Sec. 2 we model the parallel-to-two-target reflector system using generated Jacobian equations and the optical path length. In Sec. 3 we derive the feasibility condition and in Sec. 4 we provide an overview of the generating least-squares algorithm. We present results in Sec. 5 and formulate our conclusions in Sec. 6 . 2 Formulation of the parallel-to-two-target reflector system In this section, we formulate a mathematical model for the parallel-to-two-target reflector system. We derive a generating function of the form ( 2 ), find an equation for the optical path length and use energy balances to derive two generated Jacobian equations for the optical mappings. We use these results to compute the shapes of the reflectors. We denote two-dimensional and three-dimensional vectors by a bold symbol, underlining them when they are in \mathbb{R}^{3} , and adding a hat to indicate unit length, e.g., \underline{\hat{\bm{s}}} . In addition, |\cdot| denotes the 2-norm. 2.1 The generating function formulation We consider the parallel-to-two-target reflector system illustrated in Fig. 2 , where a light ray originates from (\bm{x},0) on the source plane \mathcal{S} with direction \underline{\hat{\bm{s}}} and hits the first reflector \mathcal{R}_{1} at point P_{1} , with reflected direction \underline{\hat{\bm{i}}} . Thereafter, it hits the second reflector \mathcal{R}_{2} at point P_{2} and is reflected again with direction \underline{\hat{\bm{t}}} . The ray then reaches (\bm{y},L_{1}) on the first target plane \mathcal{T}_{1} , maintains its direction \underline{\hat{\bm{t}}} and finally reaches (\bm{z},L_{2})

Read the complete text