Illuminating Car Head Light Defrost Simulations – A Digitized Approach 

Introduction

The use of the term digital twin is becoming more and more widespread. It represents a concept in which a digital replica of an object, or even a living being, is created to better and more quickly understand, engineer, and optimize the real-life twin. Common digital twin workflows at Dassault Systèmes for the automotive industry include understanding the vehicle’s aerodynamics, soiling behavior, fan and wind-noise aeroacoustics, as well as engine-bay thermal performance. A, thus far, unexplored application area is the headlight’s thermal defrost characteristic, where an ice layer on the headlight is melted away to a liquid phase. To illustrate that a phase change scenario too can be modeled using one of Dassault Systèmes’ proprietary computational fluid dynamics (CFD) solvers, an exchange between Weldex, a global Tier 1 automotive industry supplier of cameras, LCD monitors and LED lights, was initiated. The headlight CAD geometry, set-up / test conditions, and experimental results are provided by Weldex, while the software used to replicate the defrost scenario in the digital domain is provided by Dassault Systèmes.

Geometry and Software

First, the CAD geometry of the headlight and the CFD software used are described. The headlight is depicted in Figure 1, where Figure 1a depicts the entire headlight housing and Figure 1b illustrates the headlight without the front cover to illustrate the heating elements / wires. Meanwhile, the CFD solver used for the defrost scenario is FMK. It is the finite-volume-based solver that is directly integrated into Dassault Systèmes’ 3DEXPERIENCE Platform. Specifically, made available through the Fluid Dynamics Engineer role, FMK is linked to the CATIA CAD applications on the 3DEXPERIENCE Platform, enabling connected modeling and simulation (MODSIM). This is one of Dassault Systèmes’ strategic pillars: MODSIM left-shifts the product development cycle by shortening design times and allowing CAD geometry changes to be immediately updated in the associated simulation set-up.

Simulation Set-Up

Having introduced the headlight geometry from Weldex as well as the CFD solver that will be used to simulate the headlight’s defrost behavior in the digital domain, the simulation set-up is now described. This starts with the discretization of the fluid domain, which is air, and the discretization of the solid components, which includes the ice layer. As illustrated in figure 2, the fluid domain is meshed using hexahedral elements.

There is refinement on approach to the fluid domain’s stagnation inlet and pressure outlet (Figure 2a and 2b), as well as near the headlight’s boundaries (Figure 2b). Adjacent to solid boundaries, prism layer cells (Figure 2c) are applied in order to more accurately capture the developing boundary layer. With regard to the solid components, they are discretized using a tetrahedral mesh (Figure 3). Based on experimental liquid- / ice-spray data provided by Weldex, the tetrahedral discretized ice layer is initialized to a thickness of about 1mm, as is done in the experimental set-up.

Moving on, the materials chosen for the various components are as follows:

1. The wire element is specified to be copper.
2. The headlight housing is uniformly set to plastic properties in line with the plastic material used in the headlight manufacturing process. It has good vibration and fatigue properties.
3. The ice layer is set to temperature dependent properties of H₂O. Crucially, to model a phase change in FMK, a spike in the specific heat capacity of H₂O is specified such that the integral of specific heat capacity and temperature (i.e. “the area under the curve”) equals the heat of fusion of water (Figure 4).

Additionally, to account for turbulence, the Reynolds-Averaged Navier-Stokes (RANS) realizable k – E model is selected. It is chosen, as the accurate capture of flow separation points is not the target of this defrost simulation, but the more general accurate resolution of thermal behavior in the entire flow domain is. To complete the physics specification, thermal effects, which includes the surface-to-surface ray-tracing radiation model, is enabled and gravity is switched on to account for natural convection in the form of buoyancy. As initial conditions, the air’s general properties are set to -20°C for temperature, 0Pa for gauge pressure, and 0m/s for flow velocity. To initialize the realizable k- E model’s parameters, a turbulence intensity of 0.1 is set, a turbulence viscosity ratio of 100 is specified, and the velocity scale is set at 1m/s. Meanwhile, the headlight’s surface temperature is also initialized to -20°C. For boundary conditions, the following are applied:

1. A heat load of 27.4W is applied to the copper wire elements.
2. The fluid domain’s stagnation inlet is set to a gauge pressure of 0Pa.
3. The fluid domain’s pressure outlet is specified to have a gauge pressure of 0Pa as well.
4. The remaining “walls” of the fluid domain are set as slip and adiabatic.
5. Finally, the headlight’s walls have an emissivity of 1.0 and are set to be no-slip.

Prior to discussing the results, the simulation set-up’s transient solver settings need to be explained. The physical simulation time is set at 1000 seconds to cover the available experimental data. Meanwhile, although FMK is an implicit solver, a Courant–Friedrichs–Lewy (CFL) condition is still specified to dictate the simulation time-step size. Here, the CFL number is set to 5000. As the simulation progresses, the time-step is ramped-up (i.e. it gets larger) until the CFL condition of 5000 is met. The saturation limit occurs at a simulation time-step size of 5 seconds.

Results Correlation

 With the simulation set-up thoroughly described, the headlight defrost results from the digital climate chamber are now compared to the real world defrost progression data provided by Weldex. The data comes in two forms: (1) a 1-dimensional temperature evolution plot, and (2) 3-dimensional temperature contour plots. The 1-dimensional temperature development plot is derived from measurements taken at the probe location depicted in figure 1a. Looking at figure 5, one can discern that the phase change starts marginally earlier (i.e. by about 0.5 minutes) in the experiment than it does in the simulation. However, at the specified probe location, the phase change from ice to liquid water is completed at the identical time of about 9.8 minutes both in the experiment and the simulation. The linear temperature evolution thereafter is a bit steeper in the experiment than in the simulation. Specifically, by 15 minutes since the heating wires are switched on, the probe location in the experiment is about 4°C warmer than in the simulation. These slightly higher temperatures in the experiment are confirmed when looking at the 3-dimensional temperature contour plots of figure 6: at all times, except for the time of 0 minutes, the surface temperatures of the headlight are higher in the experimental images than in the images from the simulation. This discrepancy between experiment and simulation might be attributed to the fact that the ice layer thickness could not be measured to 100% accuracy: the ice layer might have been marginally thicker in the simulation than in the experiment. Nonetheless, the correlation is still good and, more importantly, deemed acceptable by the customer.

Conclusion

In conclusion, a defrost scenario for a headlight geometry from Weldex has been successfully simulated using the 3DEXPERIENCE Platform’s Fluid Dynamics Engineer role. The simulation brings with it two major advantages over its experimental test-bench counterpart: (a) one has access to temperature and flow velocity values at every point in the computational domain’s mesh (i.e. one gathers far more information on the physics of the problem), and (b) the design process is left-shifted due to the 3DEXPERIENCE Platform’s MODSIM capabilities where CAD and simulation set-up are inherently linked. Overall, this customer reference story is a further example of how Dassault Systèmes uses software “to provide business and people with virtual universes to imagine sustainable innovations capable of harmonizing product, nature and life.

Marek Roh holds a Master’s degree in Engineering Mechanics and Biomechanics from Brno University of Technology. To address Weldex’s diverse simulation needs, Roh and his team implemented Dassault Systèmes’ 3DExperience platform, a comprehensive solution covering structural, thermal, durability, and moldflow analysis—all under one roof. Roh has played a critical role in the technical evaluation of headlamp systems from the RFQ stage through to design. His ability to deliver precise insights and optimize processes has made him a key contributor to global automotive lighting projects, where his work continues to push the boundaries of simulation technology. Roh remains dedicated to advancing simulation engineering, using his expertise to drive innovation in every project he tackles. 

---

Interested in the latest in simulation? Looking for advice and best practices? Want to discuss simulation with fellow users and Dassault Systèmes experts? The SIMULIA Community is the place to find the latest resources for SIMULIA software and to collaborate with other users. The key that unlocks the door of innovative thinking and knowledge building, the SIMULIA Community provides you with the tools you need to expand your knowledge, whenever and wherever.