JCMsuite is based on advanced mathematical methods and technologies from computer science. It leverages the power and flexibility of the Finite Element Method (FEM) to achieve fast and accurate results and uses latest machine-learning technologies to optimize complex optical devices.

CAD and meshing tools

  • The JCMsuite geometry and meshing tools are specially desinged for photonic applications.
    • Shapes and geometries: Various CAD geometries such as 2D and 3D primitives, extrusions, corner-rounded shapes, and free-form shapes can be created using linear or curved elements.
    • Symmetries: By defining periodic or mirror-symmetric meshes or by working in cylindrical and twisted coordinate systems the computation times can be drastically reduced.
    • Infinite structures: Multi-layers, layered exterior domains and waveguide structures are supported.
    • Adaptive meshes: automatic mesh refinements, corner and normal refinements allow for highly accurate computations.
  • Learn more in our tutorials:  Geometry Setup and Mesh Generation

Hp-FEM solver

  • The Finite Element Method (FEM) provides a general, rigorous, versatile, and very fast method to the solution of scientific and technological challenges.
    • Problem classes: JCMsuite solves the time-harmonic Maxwell's equations for optical scattering problems, waveguide design problems, and optical resonance problems as well as linear elasticity problems, heat conduction problems, and any coupled problem classes of these types.
    • Automatic numerical settings: Various numerical settings such as finite-element degrees and PML settings (perfectly matching layer) are chosen automatically relying on residual-based error estimation.
    • Materials and sources: Various material properties, such as complex and anisotropic material permittivity and permeability tensors, dispersion properties, thermal conductivity, and stiffness can be defined. The structures can be excited, for example, by plane waves, periodic or isolated dipoles, beams, and waveguide modes.
    • Post processes: A special focus lies on the support and efficient computation of all necessary post processes in optics such as Fourier transformation, far fields, energy fluxes, overlap integrals, optical imaging, resonance expansions, and Purcell-factors.
  • Learn more in our tutorials:  Electromagnetic Field Computation
  • Related publications:  Light Scattering Computation  Propagation Mode Computation  Resonance Mode Computation  Advanced Finite Element Methods
  • Slideshow:   Advantages of Finite-Element Method for Nano-Optics Simulations


Analysis and optimization toolkit

  • Machine learning technologies enable the efficient analysis and optimization of the properties of optical devices.
    • Optimization: Bayesian optimization is a highly efficient optimization method that enables to develop high-performance devices in shorter computation times. Other supported optimization methods include downhill simplex optimization, particle swarm optimization, differential evolution, and the L-BFGS-B method.
    • Uncertainty quantification: Often, paramerters of optical systems are subject to uncertainties and fluctuations. The toolkit includes several efficient methods to determine parameter sensitivities (Sobol coefficients) and the average performance under fluctuations and its variance.
    • Parameter reconstruction: Reconstructing system parameters like material properties and shape parameters from measured data is a complex numerical task. JCMsuit includes dedicated tools for the time efficient and precise reconstruction of parameter values and their measurement uncertainties.
    • Prediction: After a learning phase the performance of optical devices can be predicted for unknown parameters.
  • Learn more in our tutorials:  Optimization using MatlabĀ®,  Optimization using Python
  • Related publications:  Optimization and Parameter Retrieval Methods  Uncertainty Quantification Methods
  • Slideshow:   A Machine Learning Method for Efficient Design Optimization in Nano-optics
logo