Validation of Artificial Lighting Simulation in CODYRUN: Application to CIE Test Cases

Table of Contents

1. Introduction

This paper presents a validation study of artificial lighting simulation capabilities within the CODYRUN software, a computational tool for aeraulic and thermal simulation in buildings developed by the Laboratory of Building Physics and Systems (L.P.B.S). The study was initiated to assess the software's reliability in simulating this specific physical phenomenon, aiming to identify its limits and potentials for improvement. The validation leverages test cases (specifically scenarios 1 and 3) developed by Task-3 TC-33 of the International Commission on Illumination (CIE), which provides standardized procedures for evaluating lighting simulation software.

2. New Simplified Model for Calculating Indoor Lighting

To quantitatively determine indoor lighting, CODYRUN integrates several combined models accounting for both direct and diffuse artificial light components. The newly introduced simplified model is conceptually similar to those used in established lighting design software like DIALux and CALCULUX.

2.1 Hypothesis of the Simulation in CODYRUN

The model operates under several key assumptions: light scattering is considered Lambertian (uniform in all directions); luminaires are characterized by manufacturer-provided photometric data and are reduced to a point source at their center of gravity; and there is no obstruction between the light source and the illuminated point on the working plane.

2.2 Direct Part of Lighting (from Artificial Light Source)

The direct illuminance at a point on the work plane is calculated based on the source's morphology and the solid angle subtended at the illuminated point relative to the source. Figure 1 illustrates this concept, showing light propagation from a ceiling-mounted source to a point on the work plane.

2.3 Diffuse Part of Lighting (from Indoor Inter-reflection)

The diffuse component results from inter-reflections of the direct light off the room's interior surfaces (walls, ceiling, floor). This component depends on the reflectivity (color) of these surfaces. CODYRUN's model calculates this by weighting the direct illuminance by the average reflection coefficient of the internal walls, as illustrated in Figure 2.

3. Core Insight: Analyst's Perspective

Core Insight: This work represents a pragmatic, engineering-focused approach to validation, prioritizing computational efficiency and integration into an existing multi-physics platform (CODYRUN) over the pursuit of the highest possible physical accuracy. The choice of a simplified, semi-detailed model over more rigorous methods like Radiosity or Ray Tracing is a strategic trade-off.

Logical Flow: The paper's logic is straightforward and defensible: 1) Identify a gap (lack of validated lighting in a thermal simulator). 2) Adopt/develop a computationally lightweight model suitable for integration. 3) Validate it against an industry-standard benchmark (CIE test cases). This is a classic software V&V (Verification & Validation) workflow, akin to methodologies discussed in the ASHRAE Standard 140 or BESTEST procedures for building energy simulation.

Strengths & Flaws: The primary strength is the integration itself. Combining lighting with thermal and airflow simulation is crucial for holistic building performance analysis, impacting energy use for lighting and cooling. Using CIE benchmarks adds credibility. The major flaw, which the authors acknowledge by calling the model "simplified," is the significant simplification of physics. Reducing complex luminaires to point sources and using a weighted-average method for inter-reflections (akin to a crude form factor approximation) will inevitably introduce errors in complex, non-diffuse, or obstructed spaces. This contrasts sharply with the high-fidelity, physically-based rendering techniques used in computer graphics research, such as those built upon the seminal Rendering Equation by Kajiya.

Actionable Insights: For practitioners, this tool is valuable for early-stage, comparative design studies where speed is key. However, for final lighting design compliance or detailed visual comfort analysis, dedicated lighting software (e.g., Radiance-based tools) remains essential. The future path is clear: the model serves as a good foundation. The next step should be a tiered approach—using the simple model for quick iterations and triggering more accurate Radiosity or photon mapping calculations (like those in the open-source Radiance suite) for critical views or final validation, creating a hybrid multi-fidelity simulation environment.

4. Technical Details and Mathematical Formulation

The core calculation, as implied by the paper, involves summing direct and diffuse components. The direct illuminance $E_{direct}$ at a point is governed by the inverse square law and the cosine of the incidence angle, derived from the source's luminous intensity $I(\theta)$ given by its photometric file:

$E_{direct} = \frac{I(\theta) \cdot \cos(\alpha)}{d^2}$

where $d$ is the distance from the source point to the illuminated point, and $\alpha$ is the angle between the light direction and the surface normal.

The diffuse illuminance $E_{diffuse}$ is approximated as a function of the direct component and the room's surface reflectances. A common simplified method (hinted at by "weighting") is using an average reflectance $\rho_{avg}$ and an inter-reflection factor, often derived from the "lumen method" or simple form-factor approximations:

$E_{diffuse} \approx E_{direct} \cdot \frac{\rho_{avg}}{1 - \rho_{avg}}$ (or a similar formulation accounting for room geometry).

The total illuminance $E_{total}$ is then: $E_{total} = E_{direct} + E_{diffuse}$.

5. Experimental Results and Chart Description

The paper applies CIE test cases (Scenarios 1 & 3 from TC-3-33) to CODYRUN. While specific numerical results are not detailed in the provided excerpt, the purpose of such test cases is typically to compare software-calculated illuminance values at specified grid points against reference values or results from other validated software.

Figure 1: Direct Light Source – This schematic depicts a simplified room cross-section. A point light source is shown on the ceiling. A straight line (ray) connects this source to a specific point on the horizontal work plane (e.g., a desk). The angle of incidence is indicated. This figure visually defines the variables (distance, angle) used in the direct illuminance calculation.

Figure 2: Diffuse Light – This diagram illustrates the inter-reflection concept. It likely shows the same room, but now with multiple arrows bouncing between walls, ceiling, and floor before eventually reaching the work plane point. This represents the diffuse component that is not coming directly from the source but from reflections, emphasizing its dependence on surface colors (reflectivity).

6. Analysis Framework: Example Case

Scenario: Evaluating the lighting performance and associated cooling load impact of switching from fluorescent troffers to LED panels in a standard 5m x 5m x 3m office room.

Framework Application using CODYRUN's Model:

1. Input Definition: Create two model variants in CODYRUN. Variant A: Use photometric data (IES/LDT file) for the existing fluorescent luminaire. Variant B: Use photometric data for the proposed LED panel. Define the same work plane height (0.75m) and grid of calculation points.
2. Simulation Execution: Run the lighting simulation for both variants. The simplified model will calculate $E_{total}$ at each grid point. Simultaneously, CODYRUN's thermal engine will calculate the heat gain from the lighting systems (based on their wattage and radiant fraction).
3. Analysis:
   • Lighting Metrics: Compare average illuminance, uniformity ratio (min/avg), and compliance with standards like EN 12464-1.
   • Energy Impact: Compare the lighting power density (LPD).
   • Thermal Impact: Analyze the difference in sensible cooling load due to the change in lighting heat gain.
4. Validation Check: For critical points (e.g., under a window, in a corner), spot-check the illuminance values against a quick calculation using DIALux or a manual formula to gauge the error introduced by the simplification.

This integrated analysis, while approximate, provides a rapid, co-simulated assessment of the design change's multi-domain effects.

7. Application Outlook and Future Directions

The integration of lighting simulation into whole-building performance tools like CODYRUN opens several future avenues:

• Daylighting and Artificial Lighting Hybrid Control: The next logical step is integrating a validated daylighting model (e.g., based on the Perez sky model). This would enable simulation of dynamic control strategies for electric lights based on available daylight, crucial for predicting real-world energy savings.
• Visual Comfort and Non-Visual Effects: Moving beyond simple illuminance to predict metrics like Daylight Glare Probability (DGP), Unified Glare Rating (UGR), and circadian stimulus. This aligns with the growing focus on health and wellbeing in buildings, as seen in WELL Building Standard.
• Model Fidelity Scaling: Developing an adaptive simulation framework where the level of detail in the lighting model adjusts based on the analysis phase—simple models for parametric exploration, and automated triggering of high-fidelity Radiance simulations for final design validation.
• Integration with BIM and Real-Time Control: Using the simulation core to inform real-time building management systems (BMS) or for generating data to train machine learning models for predictive lighting control.

8. References

1. CODYRUN Software. Laboratory of Building Physics and Systems (L.P.B.S).
2. CIE. (Year). Test Cases for the Evaluation of Lighting Software. International Commission on Illumination, Technical Committee TC-3-33.
3. Reinhart, C. F. (2014). Daylighting Handbook I & II. Building Technology Press.
4. Kajiya, J. T. (1986). The Rendering Equation. ACM SIGGRAPH Computer Graphics, 20(4), 143–150.
5. DIALux. DIAL GmbH.
6. CALCULUX. Philips Lighting (Signify).
7. ASHRAE. (2019). Standard 140-2017, Standard Method of Test for the Evaluation of Building Energy Analysis Computer Programs.
8. Ward, G. J. (1994). The RADIANCE lighting simulation and rendering system. Proceedings of the 21st Annual Conference on Computer Graphics and Interactive Techniques (SIGGRAPH '94), 459–472.