Thermal Manikin Digital Twin for Comfort Assessment

Keywords: thermal comfort, indoor environmental quality, digital twin, thermal manikin, CFD, personalized ventilation, energy efficiency, HVAC design

Thermal comfort is often addressed only superficially during building design, despite its strong influence on occupant satisfaction and energy consumption. This article presents an accessible digital twin that couples a heated thermal manikin with a CFD model through a real-time interface, enabling dynamic comfort assessment rather than a one-time evaluation.

Why comfort needs better tools

Buildings account for approximately one third of global energy demand and CO₂ emissions, with HVAC systems responsible for roughly one third of building energy use. Thermal comfort therefore plays a central role in both occupant well-being and energy efficiency. In non-automated buildings, discomfort frequently leads occupants to override controls by opening windows or adjusting thermostats, negating intended efficiency gains.

Despite its importance, comfort is typically evaluated using a single PMV/PPD calculation that assumes a uniform indoor environment. In practice, airflow jets, thermal plumes, and personalized ventilation create strong local variations in temperature and air velocity that whole-body indices cannot capture. More accurate methods, such as experiments with instrumented manikins or human subjects, are reliable but time-consuming and costly, making them impractical for routine design use. A digital twin offers a practical alternative, combining much of the realism of physical testing with the flexibility of simulation.

From thermal manikin to digital twin

A thermal manikin is a heated, human-shaped device that reproduces human heat exchange with the environment. It provides repeatable results without the variability or ethical constraints associated with human testing and is widely used in comfort, clothing, and ventilation studies. The UTCB prototype represents an average adult and is divided into six independently heated segments, each equipped with temperature sensors and Arduino-based PID control.

A digital twin is defined by a continuous, bidirectional link between a physical system and its virtual counterpart. In this case, the physical object is the thermal manikin, while the virtual counterpart is an ANSYS Fluent CFD model of the manikin placed in a room. Real-time sensor data connects the two, allowing each to update the other continuously (Figure 1). This two-way exchange distinguishes a true digital twin from conventional digital models or static “computer-simulated persons” commonly described in the literature.

Figure 1. Two-way data flow between the user, the digital twin, and the physical thermal manikin.

A single interface for the entire system

To ensure practical usability, the system is controlled through a MATLAB App Designer interface that integrates all functions into a single environment (Figure 2). Users input environmental parameters such as air and radiant temperature, air velocity, humidity, activity level, and clothing insulation. The interface displays the six manikin segments in real time, colour-coded by temperature.

Segment temperatures can be set either as fixed values or driven automatically by Gagge’s two-node thermophysiological model. The interface also launches CFD simulations, visualizes velocity and temperature fields around the body, and reports comfort indices (PMV/PPD), equivalent temperature, and energy consumption. All results can be exported for further analysis. This integration allows comfort, physiology, airflow, and energy performance to be evaluated simultaneously and in real time.

Figure 2. MATLAB App Designer interface with five functional blocks.

Performance verification

The digital twin generates airflow and temperature fields using the manikin’s measured surface temperatures as real-time boundary conditions. Figure 3 illustrates a seated occupant, where the buoyant thermal plume rising from the body governs local airflow and comfort—an effect not captured by whole-body indices.

Validation was performed in two stages. First, the thermophysiological model was compared with Gagge’s original results over a temperature range of 10–40 °C and multiple humidity levels, yielding a mean deviation of 8.6%, largely attributable to uncertainties in the original graphical data. Second, the CFD model was validated against experimental measurements of thermal plumes reported in the literature, obtained using particle image velocimetry and infrared imaging. The model accurately reproduced both plume structure and peak velocities of approximately 0.25 m/s.

To ensure responsiveness, the CFD model was intentionally simplified. Mesh reduction and limited iterations significantly reduced computation time while maintaining velocity and temperature errors within approximately 0.2 m/s and 0.2 °C, an acceptable compromise for real-time comfort assessment.

Figure 3. Velocity and temperature fields around the manikin.

Practical implications

Dynamic, local comfort assessment enables several practical applications. The digital twin can support personalized ventilation and heating strategies, inform adaptive control algorithms, and lower the entry barrier for designers without CFD expertise. Because comfort feedback is provided in real time, designers can immediately assess the impact of design decisions, rather than relying on a single PMV value calculated late in the process.

Figure 4. Thermal manikin digital twin in operation.

Outlook

The system presented here provides a foundation for further development. Future enhancements include automated sensor integration, more advanced physiological models, and AI-based control strategies. The approach is readily transferable to vehicles, aircraft cabins, and clinical environments. Even in its current form, the digital twin demonstrates how realistic, real-time thermal comfort assessment can be integrated into everyday engineering practice.

Acknowledgement

This work was supported by the grant of CCCDI – UEFISCDI, project number 60PHE / 01.04.2024 (2024–2026).

References

[1]     UN Environment Programme, “Global Status Report for Buildings and Construction 2024/2025,” 2025.

[2]     L. Yang, H. Yan and J. C. Lam, “Thermal comfort and building energy consumption implications — a review,” Applied Energy, vol. 115, pp. 164–173, 2014.

[3]     I. Holmér, “Thermal manikin history and applications,” European Journal of Applied Physiology, vol. 92, no. 6, pp. 614–618, 2004.

[4]     C. Croitoru, I. Nastase, F. Bode, A. Meslem and A. Dogeanu, “Thermal comfort models for indoor spaces and vehicles — current capabilities and future perspectives,” Renewable and Sustainable Energy Reviews, vol. 44, pp. 304–318, 2015.

[5]     K. Katić, R. Li and W. Zeiler, “Thermophysiological models and their applications: a review,” Building and Environment, vol. 106, pp. 286–300, 2016.

[6]     D. Fiala, G. Havenith, P. Bröde, B. Kampmann and G. Jendritzky, “UTCI-Fiala multi-node model of human heat transfer and temperature regulation,” International Journal of Biometeorology, vol. 56, no. 3, pp. 429–441, 2012.

[7]     Q. He et al., “From digital human modeling to human digital twin: framework and perspectives in human factors,” Chinese Journal of Mechanical Engineering, vol. 37, no. 1, art. 9, 2024.

[8]     V. A. Arowoiya, R. C. Moehler and Y. Fang, “Digital twin technology for thermal comfort and energy efficiency in buildings: a state-of-the-art and future directions,” Energy and Built Environment, vol. 5, no. 5, pp. 641–656, 2024.

[9]     P. Danca, C. I. Coşoiu, I. Nastase, F. Bode and M. R. Georgescu, “Personalized ventilation as a possible strategy for reducing airborne infectious disease transmission on commercial aircraft,” Applied Sciences, vol. 12, no. 4, art. 2088, 2022.

[10]   D. Lemian and F. Bode, “Digital twins in the building sector: implementation and key features,” E3S Web of Conferences, vol. 608, art. 05004, 2025.