Phase Change Materials (PCMs) are increasingly incorporated into building envelopes to improve thermal inertia and reduce short-term temperature fluctuations. However, accurately predicting heat transfer in PCM-enhanced walls remains challenging due to the nonlinear behavior associated with latent heat effects. Classical numerical solvers tend to smooth thermal gradients, while traditional data-driven models struggle to reconstruct full spatio-temporal temperature fields.This study proposes a hybrid modelling framework based on a Deep Residual Graph Convolutional Neural Network (DRGCN) that combines physics-based simulation with experimental fine-tuning. The wall is discretised into spatial nodes, allowing heat transfer to be represented as a graph structure. The model is first pre-trained on a large synthetic dataset generated using a one-dimensional parabolic partial differential equation solver and then refined using experimental measurements collected at multiple depths within a PCM-enhanced mortar wall. To the best of our knowledge, this is the first application of a graph neural network to heat transfer in mPCM-enhanced walls, combining PDE-based pre-training with experimental fine-tuning.The results show that the hybrid DRGCN significantly outperforms the physics-based baseline. Across all interior sensor positions, the model achieves a root-mean-square error of 0.4–0.5 °C, compared with approximately 1.1 °C for the numerical solver. In addition, the model also reproduces the curvature of spatial temperature profiles and the thickness of phase change transition zones more accurately than the reference solution.

This study offers a comprehensive evaluation of the energy and hygrothermal performance of compressed earth block (CEB) building envelopes through a combination of experimental characterization and multiscale numerical modeling. At the material level, the thermal and hygric properties of CEB were experimentally measured and used as input parameters in a coupled heat and moisture transfer model based on Künzel's formulation. Validation was performed at the laboratory wall scale, showing close agreement between simulated and measured temperature and moisture profiles across wall depths. At the building level, thermal performance was assessed under representative Oceanic and Mediterranean climates in France. Simulations were conducted for a 98 m3 office space occupied by two adults, comparing CEB wall assemblies to a traditional concrete envelope. Results show that CEB walls reduce heating demand by 15% under oceanic conditions and by 13% under Mediterranean conditions. Additionally, cooling demand is reduced by up to 75%, especially in Mediterranean climates. These improvements are mainly due to the higher thermal inertia and lower thermal conductivity of CEB materials. Dynamic analysis indicates increased time lag and decreased decrement factors, reflecting better attenuation of external thermal fluctuations. Hygrothermal results also demonstrate improved moisture-buffering capacity, less moisture buildup, and a lower risk of mold growth compared to concrete. This leads to more stable indoor conditions and enhanced thermal comfort. Overall, the findings emphasize the potential of CEB as a low-energy, high-performance building envelope suitable for various climatic conditions.

This study investigates the thermal performance and heat transfer mechanisms of a wall constructed with Hollow Date Palm fiber Concrete Blocks (HDPCB) through combined experimental and numerical analyses. Steady state and dynamic laboratory measurements were conducted to determine the wall's thermal transmittance, thermal resistance, time lag, decrement factor, and periodic thermal transmittance. A 3D model was developed to simulate heat transfer through the solid components of the wall and natural convection within the cavities, while the contribution of radiative heat transfer was evaluated separately. Sensitivity analysis was performed to identify the factor with the greatest influence on predicted heat transfer. Experimental results indicate that HDPCB walls provide effective thermal insulation, with a U value of 1.23 W m-2 K-1 and an R value of 0.81 m2 K W-1. Numerical predictions agreed with measurements within 3%, confirming that natural convection is the dominant heat transfer mode within the cavities of the blocks under the tested conditions. These results demonstrate the potential of HDPCB as a sustainable and energy-efficient building envelope solution.

This study offers a detailed analysis of water sorption isotherms in raw earth materials, which are increasingly recognized as sustainable alternatives to traditional building materials. The water retention behavior of Compressed Earth Blocks (CEBs) was examined under steady-state conditions using sorption isotherms, which do not fully capture transient moisture buffering performance at the building scale. In this context, we focus on their ability to adsorb and desorb moisture as a measure of their potential to help control indoor humidity. The Dynamic Vapor Sorption (DVS) device was employed to compare the adsorption and desorption isotherms of 1 cm3 cubic samples and powder fragments (2650 mg) of CEBs. Several factors were also investigated, including temperature (15, 23, and 35 °C), repeated wet/dry cycles, and the effects of time on sorption behavior. Using appropriate isotherm modeling, the isosteric heat of water sorption was calculated for the CEB. Results indicated that cubic samples adsorbed up to 23

This study investigated the hygrothermal behavior in bio-based mortar incorporating Micronized Miscanthus Fiber (MMF) through a combination experimental analysis, numerical modelling, sensitivity analysis and hysteresis effect. In a preliminary step at material scale, sorption and desorption isotherms were measured at 23 degrees C and exhibited pronounced hysteresis, which was fitted using the Guggenheim-Anderson-de Boer (GAB) model. Cyclic wetting-drying tests further confirmed this hysteresis effect, mainly attributed to capillary condensation and the ink-bottle phenomenon in the porous microstructure. At the wall-scale, a rectangular specimen representing a section of wall material was tested under controlled bi-climatic conditions, simulating outdoor moisture cycling and indoor passive environments. The MMF specimen exhibited strong hygric buffering capacity. Under extreme outdoor relative humidity (RH) cycles, the inner side exhibited strongly dampened RH variations compared with outdoor fluctuations. Coupled heat and moisture transfer simulations were subsequently performed using the K & uuml;nzel model in COMSOL Multiphysics. The model showed good agreement with experimental results. A sensitivity analysis was then performed to assess the influence of sensor positioning, sorption isotherm, and water vapor permeability. Results highlighted that both material property uncertainties and small sensor positioning errors can significantly influence simulation accuracy. To account for hysteresis effects, the empirical Pedersen model and the physically based Carmeliet model were tested. The Carmeliet model provided superior performance, particularly at the middle and inner sides of the wall specimen. Overall, this work provides a comprehensive understanding of hygrothermal behavior and hysteresis effects in bio-based mortars. It emphasizes the importance of accurately characterizing material properties and properly hysteresis modelling to ensure reliable predictions for sustainable bio-based wall systems.

This study investigates the prediction of thermal behavior in cement mortar walls incorporating bio-based phase change materials (PCM), using two advanced deep learning frameworks: Universal Differential Equations (UDE) and Fourier Neural Operators (FNO). A physical experiment was conducted on a cement mortar wall containing 11 wt% of microencapsulated PCM. This wall was instrumented with thermal sensors at three different depths and subjected to controlled boundary conditions. The UDE approach integrates an enthalpy-based physical model with a neural correction term to capture latent heat dynamics, while also utilizing learnable thermophysical parameters and an adaptive multiple shooting strategy to prevent numerical drift. Conversely, FNO learns to predict entire temperature fields directly from experimental data by operating in the Fourier domain. Results show that both models achieve high predictive accuracy. The optimized UDE model demonstrates superior performance at the sensor locations with a global RMSE of 0.0399 degrees C and an R2 of 0.9999, compared to an RMSE of 0.1680 degrees C and an R2 of 0.998 for the FNO. These findings highlight the comparable accuracy of both frameworks, while the UDE offers the added benefit of a physically interpretable model with embedded and calibrated thermophysical parameters.

Building envelope walls play an important role in controlling heat transfer and reducing energy demand by moderating outdoor thermal loads. In hot semi-arid climates, enhancing wall thermal inertia is an effective passive approach to limit indoor temperature fluctuations and improve summer comfort. In this study, date palm concrete (DPC) is investigated as a bio-based wall material due to its favorable thermal inertia compared to conventional brick construction. To further enhance the wall's heat-storage capacity and delay heat transfer, phase change materials (PCMs) are integrated into the wall assembly. The thermal performance of brick and DPC walls, with and without PCM layers of different melting temperatures, is evaluated under HVAC ON conditions to assess heating and cooling energy consumption and identify the optimal PCM configuration. Thereafter, the selected PCM is applied to all wall configurations and the analysis is extended to free-floating conditions (HVAC OFF) to examine indoor air temperature regulation during summer. The study further investigates wall surface temperatures and heat-flux behavior for different wall orientations to evaluate thermal damping and heat-transfer attenuation. The results confirm that DPC significantly enhances wall thermal inertia compared to conventional brick construction. Under HVAC-ON conditions, annual heating demand is reduced by 57%, while cooling demand reduced by 35%. The integration of PCM further improves performance, with the interior RT25 configuration providing the highest energy savings. Under HVAC-OFF summer conditions, DPC shows a longer time lag (7 h vs. 4 h) and a lower decrement factor (0.09 vs. 0.38), and the addition of RT25 reduces peak indoor air temperature by up to 3.5 °C. In all cases, PCM outperforms PSM, confirming the dominant contribution of latent heat storage activated near the comfort temperature range.

This study investigates the thermal behavior of two cement mortar walls: one incorporating 15 wt% bio-based Phase Change Material (PCM) and a reference wall without PCM. The thermal performance was first assessed experimentally using a bi-climatic chamber under cyclic temperature variations between 40 degrees C and 15 degrees C, while monitoring temperature profiles at various depths. Two modeling approaches were then employed to describe dynamic thermal behavior: the Piecewise AutoRegressive eXogenous (PWARX) model for identifying discrete thermal states and the Long Short-Term Memory (LSTM) deep learning model for predicting temperature variations. The PWARX model successfully identified distinct phase transition states in the PCM wall: five states at 5 cm and 7.5 cm depth, and four states at 2.5 cm, confirming the presence of latent heat effects. In contrast, the reference wall exhibited only three to four states, highlighting distinct thermal behavior due to the absence of PCM. The LSTM model achieved a high predictive accuracy with an R2 of 0.99, closely matching experimental data. The integration of PCM reduced temperature fluctuations, leading to a 1.2 degrees C lower peak temperature during heating and a 1.0-1.2 degrees C higher temperature during cooling at mid-depths. These findings demonstrate the potential of PCM-enhanced mortar walls to improve building energy efficiency by stabilizing indoor temperatures, reducing thermal fluctuations, and minimizing heating and cooling energy consumption. Furthermore, this study highlights that while the PWARX model provides deeper insights into PCM phase transitions, the LSTM model excels in forecasting thermal behavior, offering complementary tools for optimizing PCM applications in construction.

Raw earth is a material of interest in the building sector. It has the advantages of being sustainable, being readily available and having a low-energy manufacturing process. Unfortunately, only a few case studies of instrumented earthen houses exist in the literature. Thus, a raw compressed earth brick (CEB) test house was specifically built and extensively instrumented for research purposes in Sense-City equipment at Champs-sur-Marne, France. A wide variety of sensors were deployed in the walls and in the indoor and outdoor environments for monitoring energy efficiency, hygrothermal behavior and thermal comfort. The CEB was studied from the material scale with chemical and hygrothermal characterization tests to the building scale during different seasons. After the instrumentation and practical techniques for the placement of temperature and humidity sensors in the walls and of heat flux sensors on wall surfaces for long-term monitoring were presented, an experimental analysis of the sensor outputs was carried out. Although the low wall thermal resistance was estimated at 0.34 m2.K.W-1 via the ISO 9869-1 standard method, the CEB hygroscopic wall had the abilities to regulate the indoor temperature and humidity and dampen and delay hot outdoor temperatures during summer heat waves. Especially in winter periods, concerning the humidity in walls, we observed different humidity levels depending on wall orientation and height-dependent humidity variation in walls exposed to rain. All the measurement data are provided in the associated data article.

In this work, we numerically investigate coupled heat and moisture transfer in a bio-based mortar wall containing micronized miscanthus fibers. The physical property inputs and their dependencies on temperature and moisture content were determined from literature data. A heat and moisture transfer model, based on the Künzel approach, was implemented in COMSOL Multiphysics. Simulations were compared with experimental data, demonstrating accurate estimations of temperature and relative humidity variations at different material depths. A maximum temperature deviation of 0.6 ℃ between experimental and numerical data was observed at a depth of 5 cm, while a maximum relative humidity deviation of 5

Wood-based biowaste represents a plentiful biomaterial on our planet, and harnessing this industrial waste for sustainable construction products is strongly advocated due to its insulation and hygroscopic properties. The primary aim of this work is to implement innovative bio-based gypsum materials by incorporating a mixture of untreated wood and paper wastes. Furthermore, the study aims to comprehensively analyze the hygroscopic properties of these bio-based gypsum composites, comparing them to traditional gypsum boards. To better understand the hydric behavior and particularly the relationship between the porous microstructure and the different properties, several tests including porosity, capillary absorption, sorption isotherm, water vapor permeability, and moisture buffer value (MBV) were performed for various formulations. Moreover, it shows that the biobased composite exhibits a remarkable moisture absorption capacity, nearly 15 times higher than that of the reference sample, and displays a high-water vapor permeability, owing to a significant enlargement in critical pore size. Likewise, all gypsum formulations display MBV values exceeding 2.00 g/(m2.

This study investigates the impact of incorporating construction and demolition waste (CDW) aggregates and Alfa natural fibers on the performance characteristics of asphalt mixtures, with a focus on mixing temperature. Several formulations were developed and evaluated through multiphysics property measurements, including density, ultrasonic pulse velocity, rutting resistance, thermal conductivity, and spectral reflectance. The results indicate that Alfa fibers enhance thermal resistance and spectral reflectance. Notably, incorporating 1% Alfa fiber and 20% CDW while reducing the mixing temperature to 150 °C significantly improves rutting resistance. These combined effects result in an optimized formulation that is more resistant to thermal stress during service, thereby enhancing its performance at elevated temperatures. These findings highlight the potential of integrating CDW and natural fibers into asphalt mixtures to develop environmentally friendly and thermally resilient materials, particularly for warming climate regions.

A non-insulated raw compressed earth brick test house was built and deeply instrumented in Sense-City equipment at Champs sur Marne, France. A wide variety of sensors were deployed in the room, inside the walls and outside to monitor the hygrothermal behaviour, the thermal comfort and the heating energy consumption. The measurement campaigns were performed over several weeks during winter, spring and summer seasons in 2024. The test house was unoccupied and exposed to natural weather conditions. During winter tests, different heating scenarios were considered. The provided dataset of sensor outputs can be useful for a better understanding of earthen construction and for the experimental validation of building physics models at both the wall and building scales.

This study experimentally and numerically analyzes the hygrothermal performance of a wall constructed from compressed raw earth blocks under different thermal and hygric conditions. The goal is to evaluate the ability of unfired, unstabilized raw earth to regulate indoor temperature and humidity levels. Initial tests revealed strong thermal and moisture-regulating properties at the material level, justifying further investigation at the wall scale. A bi-climatic setup exposed the wall's outer surface to fluctuating temperatures and humidity cycles while maintaining the inner surface at 20 degrees C with varying relative humidity (30-50 % RH). Data on temperature and humidity were collected using sensors at different depths, and a numerical model based on K & uuml;nzel's approach was applied to predict and analyze temperature and humidity variations. The results showed that the raw earth wall effectively dampened temperature fluctuations and regulated humidity, displaying a temperature gradient of 6 degrees C between depths of 2.5 cm and 7.5 cm during temperature cycles. The wall also functioned as a moisture barrier, with internal moisture levels not stabilizing after 20 days at 80 % relative humidity. The numerical model accurately predicted temperature variations. Designed specifically for internal partitions, this type of wall shows great potential to improve indoor environmental quality by passively regulating temperature and humidity.

In the context of sustainable construction, the demand for low-carbon building materials has increased interest in raw earth due to its availability and ecological benefits. Bio-based additives such as cellulose fibers and starch are often introduced to enhance mechanical and durability properties, although their impact on hygrothermal performance requires further investigation. This study examined the hygrothermal and mechanical properties of three raw earth adobe brick formulations: a reference sample, one with cellulose fibers, and one with starch. Key properties, including water vapor permeability, thermal conductivity, volumetric heat capacity, moisture buffer value, and elastic modulus, were analyzed. Notably, moisture regulation and air permeability remained excellent across all formulations, with minimal impact from the bio-based additives. These findings underscore the potential of raw earth adobe bricks, with or without bio-based additives, as a viable low-carbon material for sustainable construction.

Miscanthus fibres are regarded as a sustainable, eco-friendly material with high hygroscopicity and low thermal conductivity. Although promising at the material scale, further wall-scale evaluations are needed. This study investigates the hygrothermal performance of a biobased mortar wall containing micronized miscanthus fibers, compared to a conventional mortar wall. An experimental bi-climatic device was used to simulate distinct outdoor and indoor environments by exposing each side of both walls to different hygrothermal conditions. Temperature and relative humidity (RH) data were collected using a network of sensors, enabling a thorough analysis of the walls' hygrothermal response at various depths under both dynamic and steadystate regimes. Additionally, a numerical model based on the K & uuml;nzel approach was employed to predict and compare temperature and humidity behavior within the biobased wall. Results showed good agreement between experimental data and numerical predictions, with a maximum temperature deviation of 4 degrees C at 5 cm depth. However, RH predictions revealed discrepancies, notably a 5 % deviation at 7.5 cm depth. Statistical validation indicated model accuracy. Furthermore, tests involving cyclic moisture changes under isothermal conditions demonstrated that the miscanthus wall acts as an effective hygric regulator, maintaining stable indoor condition as outdoor relative humidity increases. Higher prediction errors observed at boundary depths (0 and 10 cm) were attributed to external airflow during experiments. Overall, the numerical model reliably predicted temperature variations, and RH estimations were acceptable considering material heterogeneity. These findings confirm the potential of miscanthus fiber-reinforced materials for improving hygrothermal performance in building construction.

Raw earth, considered a low-carbon material, is increasingly utilized in the construction sector as a replacement for conventional construction materials. By reducing the demand for cooling and heating and providing comfortable indoor air quality, raw earth emerges as a material that meets these criteria. This study is part of the DuReTerre project, which aims to address the sustainability of earthen buildings by evaluating earth at the material, wall, and building scale. Specifically, this part evaluates the capacity of compressed earth blocks (CEB) to maintain a comfortable indoor ambiance while managing temperature and humidity levels within a building. To achieve this, thermophysical properties (thermal conductivity during variations in temperature and humidity levels) and hygric properties (sorption/desorption, capillary ab-sorption, Moisture Buffer Value, water vapor permeability and water con-tent) were assessed at the material scale. Results indicate that these blocks have moderate thermal conductivity but are capable of effectively regulating indoor moisture levels due to their remarkable hygric properties.

This study tends to optimize and develop an innovative gypsum material by incorporating hybrid waste compounds. Using a Doehlert design, this work explores the effects of introducing paper (X1), polystyrene (X2), and polyester fibers (X3), as well as their interactions, on gypsum's thermophysical and mechanical behavior. Bulk density, thermal conductivity, flexural strength, compressive strength, and the fire resistance test were performed on the developed formulations. Statistical analysis emphasizes the significant influence of all process variables on gypsum properties. Polyester fibers notably enhance flexural strength when combined with the cooperative effect of paper and polystyrene wastes, which strike a balance between physical and mechanical characteristics. The optimum formulation parameters lead to a significant 33% decrease in bulk density, a 43% reduction in thermal conductivity, a notable 42% increase in flexural strength, and a significant reduction in compressive strength of around 50%, but still meet the 2 MPa standard. The new samples show ductile behavior, highlighting gypsum's potential as a load-bearing material with excellent thermal insulation properties and impressive fire resistance. This research significantly improves our understanding of the effectiveness of a gypsum compound incorporating waste in a hybrid manner, bringing this approach into line with sustainable development objectives.

The lively international debate on the future of the built environment has placed the emphasis on the possibilities offered by bio and geo-based building materials. Among these, raw earth-based materials offer several advantages associated to their reusability and low embodied energy.Nowadays, several companies are basing their production lines on prefabricated raw earth products, as is the case of compressed earth blocks (from now on CEBs). CEBs are commercialized for the construction of massive vertical envelopes, characterized by a high thermal inertia. Nevertheless, in order to compete with conventional building materials, it is also necessary to guarantee a high thermal resistance.In this work, this issue was solved by the design and testing of full-scale uninsulated and bio-based thermal insulated CEB walls. In this way, the thermal performance of CEB walls can be increased to meet the high energy requirements currently adopted in European Countries. Furthermore, the choice of bio-based insulations represents the main novelty of the study, aimed at finding hygrothermal compatible solutions at a low environmental cost.More in detail, this work reports the results of the thermal and physical material characterization of CEBs and of two innovative bio-based thermal insulations (lime hemp and sugarcane bagasse panels), and compare them with measurements made on full-scale uninsulated and insulated CEB walls. For this purpose, the walls are tested inside a double-room climatic chamber where they are subjected to variable temperatures on their two faces, reproducing typical indoor and outdoor conditions during summer and winter conditions in a continental climate.Results show the enhancement of thermal performances of compressed earth blocks walls when thin layers of bio-based thermal insulations are added. The thermal resistance of weakly bio-based insulated CEB walls is found to be nine times (for the sugarcane bagasse insulated CEB wall) and four times (for the lime hemp insulated CEB wall) higher than that of uninsulated CEB walls. Moreover, the addition of the insulation layers enhances the time lag and the decrement factor of compressed earth block walls.