Structural, Optical, Fluid and Thermal (SOFT™) workflows provide a framework for analysing optical systems where optical performance is influenced by fluid flow. Similar in concept to Structural, Thermal, Optical Performance (STOP) analysis, SOFT extends the multiphysics workflow by incorporating Computational Fluid Dynamics (CFD) data directly into the optical simulation process. SOFT leverages fluid-dynamics simulation to quantify how flow-induced density variations affect refractive index distributions and ultimately optical performance. By establishing a connection between CFD and optical analysis tools, the workflow enables engineers to evaluate coupled aero-optical phenomena within a unified simulation framework.
To demonstrate the workflow, this article uses a gas-cooled Yb:YAG multi-slab laser amplifier as an illustrative example. Such systems are representative of many high-power laser architectures where cooling flows are essential for thermal management but can simultaneously introduce optical distortions through refractive index inhomogeneities. The presented workflow shows how CFD-generated density fields can be transformed into spatially varying refractive index distributions and incorporated into wavefront analysis, enabling the investigation of beam degradation caused by the cooling flow. Workflow automation and design exploration are facilitated through Ansys optiSLang (optiSLang), allowing efficient evaluation of multiple operating conditions and design variations.
Authored by Flurin Herren
Introduction
Modern optical systems are increasingly influenced by interactions between multiple physical domains. Structural deformation, thermal loading and fluid flow can all impact optical performance, making isolated single-physics analyses insufficient for many engineering applications. As a result, multiphysics workflows have become essential for understanding performance-limiting mechanisms and supporting simulation-driven design.
Within the optics community, Structural, Thermal, Optical Performance (STOP) analysis is commonly used to evaluate the impact of mechanical and thermal effects on optical systems. However, in many optical applications, optical degradation can also be significantly influenced by fluid flow. Structural, Optical, Fluid and Thermal (SOFT) workflows address this challenge by coupling Computational Fluid Dynamics (CFD) results directly to optical analysis, converting flow-induced density variations into spatially varying refractive-index distributions for optical evaluation.
High-power Yb:YAG laser amplifiers provide a representative example of this type of coupled physics problem. Yb:YAG materials are widely used in high-energy laser systems due to their favorable efficiency, scalability and thermo-mechanical properties [1]. Multi-slab amplifier architectures can achieve high power levels but require effective thermal management to remove heat generated within the gain medium. To accomplish this, cooling gas is circulated through channels between neighbouring amplifier slabs.
While essential for thermal management, the cooling flow also creates density variations within the optical path. These density fluctuations produce refractive index inhomogeneities that can introduce optical path differences and distort the propagating wavefront. Previous studies have shown that turbulence and refractive index fluctuations can significantly influence laser beam quality [2].
Example: Gas cooled Yb:YAG laser amplifier
This article demonstrates a SOFT workflow using a gas-cooled Yb:YAG multi-slab laser amplifier as an illustrative example. The workflow combines CFD simulation with Ansys Discovery (Discovery), density to refractive index conversion with Python and optical wavefront analysis with Ansys Zemax OpticStudio (Zemax OpticStudio) to evaluate flow-induced optical distortions, providing a practical framework for investigating SOFT effects in advanced laser systems.
Nominal Optical Design
The first stage of the workflow involves the development of the baseline optical model within Zemax OpticStudio. The amplifier model is designed using the Sequential Mode, which provides the capabilities to analyse the wavefront error of the laser propagation and to import direct refractive index data via the Multiphysics Data Loader. The representative model has four slabs, leaving three cooling channels as the fluid regions between adjacent slabs, these regions subsequently form the computational domain for the CFD analysis. In the 3D Layout below these cooling channels have been highlight as red rectangular volumes for visualisation purposes.
Once the optical architecture has been established, the complete geometric representation of the amplifier is exported from Zemax OpticStudio as a STEP file.
CFD Simulation and Refractive-Index Generation
Discovery is used to model the gas flow through the Yb:YAG laser amplifier model and to extract the density distribution within the cooling channel regions. Discovery uses a GPU-native meshing and solver technology to perform rapid CFD analyses. In addition, it includes a geometry modelling environment that allows users to modify dimensions and parameterise various geometric aspects of the model.
Using the representative model with the fluid regions between adjacent slabs, the CFD analysis looks as follows:
Python-based scripting capabilities are integrated directly into the software console, so users can carry out specific post-processing or export of data. In this example, a point cloud containing the coordinates and corresponding density values in the regions between the slabs is exported in CSV format (Approximately 70,000 data points per region are exported).
The CFD simulation in Discovery provides a spatially varying density field; however, optical simulation requires a corresponding refractive index distribution. Consequently, an intermediate processing step is required to transform the CFD results into a format compatible with optical propagation analysis.
A Python-based post-processing script is developed to automate this conversion. The script reads the mesh-based density data exported from Discovery and calculates the local refractive index at each computational node using the wavelength dependent Gladstone–Dale relation. This relationship links gas density to refractive index through the following expression:
(𝜆) = 1+𝐾𝐺𝐷(𝜆)𝜌
- where 𝑛 is the refractive index, 𝐾𝐺𝐷(𝜆) is the wavelength-dependent Gladstone–Dale coefficient, and 𝜌 is the local gas density.
The wavelength dependence of the Gladstone–Dale coefficient is included to ensure consistency with the operating wavelength of the Yb:YAG amplifier. Following the refractive index calculation, the Python script reformats the data into the direct index format required by Zemax OpticStudio.
Import into Zemax OpticStudio and Optical Analysis
The next stage of the workflow evaluates the optical consequences of the CFD-based refractive index variations. The direct index data generated with the conversion script are imported into Zemax OpticStudio via the Multiphysics Data Loader and assigned to the Surfaces which represent the cooling channels within the Sequential Mode. Once imported, the CFD-derived refractive index field acts as a spatially varying optical medium through which the laser beam propagates. Ray-tracing and wavefront analysis tools available within Zemax OpticStudio are then used to assess the impact of the flow-induced refractive index gradients on the beam quality.
Wavefront error is selected as the primary performance metric because it provides a direct measure of the optical distortion introduced by the cooling flow and by comparing results obtained with and without the imported refractive index field, the contribution of the cooling gas can be isolated from the intrinsic optical behaviour of the laser amplifier system.
- The workflow therefore enables direct correlation between CFD-based flow characteristics and optical performance degradation.
Because all stages of the process can be linked through automated data exchange with optiSLang, design modifications can be propagated rapidly throughout the entire simulation chain. This enables efficient exploration of cooling-channel geometries, operating conditions, and system-level design trade-offs, providing a practical framework for simulation-led development of advanced systems.
Automation and Design Space Exploration
To systematically explore the design space, the workflow is automated using optiSLang, integrating Discovery, the Python-based data conversion script, and Zemax OpticStudio within a single parametric workflow.
The following parameters are defined as Input Parameters:
- Volumetric heat source applied to the slabs:
- Internal heat generation rate within the slab material.
- Gas mass flow rate at the inlet:
- Mass of Helium entering the domain per unit time.
- Gas molar mass:
- Gas composition parameter influencing thermophysical properties and flow behaviour
- Operating pressure:
- System pressure affecting gas density and transport characteristics.
These parameters are varied within the defined range upon each run and for each design iteration, the density field exported from Discovery is converted into a refractive-index distribution and reformatted from CSV to the direct index text format required by Zemax OpticStudio before the optical analysis is executed.
The resulting wavefront error is captured as the Response parameter and used within an optiSLang Sensitivity Analysis to quantify the relative influence of each Input Parameter on the performance. The Sensitivity Analysis (Design Space Exploration) within optiSLang can be summarised with the following three steps:
- Design of Experiment (DoE) to explore the design space (DoE techniques include algorithm such as Latin Hypercube Sampling (LHS), Advanced LHS or Sobol Sequences)
- Build an Adaptive Metamodel of Optimal Prognosis (AMOP), which automatically balances model accuracy and sampling effort by iteratively adding new design points where additional information is required.
- Perform a variance-based sensitivity analysis to identify and quantify the most influential input parameters on the wavefront error.
As shown on the figure above, which shows the most important post-processing data plots from the Sensitivity Analysis, the dominante contributors on the performance (metric: wavefront error) are the molar mass of the gas (in kg/mol; 47%), the operating pressure (in Pa; 28.4 %) and the mass flow rate (in kg/s; 24.8 %).
Conclusion
This work has demonstrated an automated multiphysics workflow that couples computational fluid dynamics (CFD) and optical simulation to evaluate aero-optical effects in a gas-cooled Yb:YAG multi-slab laser amplifier. By integrating Ansys Discovery, Python-based data processing, Ansys Zemax OpticStudio, and Ansys optiSLang within a single automated framework, CFD-derived density fields can be transformed into refractive index distributions and directly incorporated into optical propagation analyses with minimal manual intervention. In addition, the automated design space exploration study provided physical insight into the relationship between fluid-dynamic and optical performance metrics and demonstrated the ability of the workflow to efficiently explore the multiphysics design space.
References
- -[1] David, S. P., Jambunathan, V., Lucianetti, A., and Mocek, T., "Overview of ytterbium based transparent ceramics for diode pumped high energy solid-state lasers," High Power Laser Science and Engineering 6, e62 (2018); https://doi.org/10.1017/hpl.2018.58.
- -[2] Bellec, M., Bordenave, C., Chériaux, G., and Lascoux, N., "Effect of turbulence on the wavefront of an ultra-high intensity laser beam," Proc. SPIE 11299, Laser Beam Shaping XX, 112990L (2020).
Acknowledgements
- Supported by Tino Dannenberg (Automation/ Ansys optiSLang), Nathan Abera and Achilleas Krikas (Fluid Dynamics/Ansys Discovery/ Ansys Fluent).
- Presented at SPIE Optics + Photonics 2026 by Sara Karami