Eng

Accurate early-stage assessment of building energy and carbon performance is essential for informed sustainable design yet remains challenging due to limited design detail and simulation effort. This study presents a Building Information Modeling–Machine Learning (BIM-ML) framework for predicting office building energy and carbon performance at early design stages using simulation-based datasets. A reduced-factorial Design of Experiments (DOE) generated 210 parametric office building models for Orlando, Florida (ASHRAE Climate Zone 2A), complemented by additional climate scenarios. Systematic variations in geometry, envelope, building systems, and operational schedules produced a dataset with 14 independent variables and five performance indicators: Energy Use Intensity, Operational Energy, Operational Carbon, Embodied Carbon, and Total Carbon. Four regression methods—Linear Regression, Model Tree (M5P), Sequential Minimal Optimization Regression, and Random Forest—were trained and evaluated using 10-fold cross-validation. Random Forest showed the strongest overall predictive performance. Feature-importance analysis identified HVAC system type, Window-to-Wall Ratio, and operational schedule as the most influential parameters, while geometric factors had lower impact. Cross-climate analysis and validation with measured data from two university office buildings indicate that the framework is adaptable and generalizable, supporting reliable early-stage evaluation of energy and carbon performance.

To comprehensively evaluate the thermal risk parameters of the HMX synthesis process via the acetic anhydride method, we systematically investigated the safety of raw materials, ingredient mixing, nitration, and crystal transformation processes using DSC, ARC, and reaction calorimetry, which enabled the optimization of feeding strategies based on the exothermic characteristics observed during both ingredient mixing and nitration. Results indicate that the decomposition temperatures of raw materials and products are all above 200 °C, showing excellent thermal stability. Thus, multi-batch feeding is preferred for reaction material preparation. For the nitration process, continuous and stable feeding must be guaranteed during the feeding stage. During nitration, the temperature relationship satisfies T_p < MTSR < MTT < T_D24, wherein the risk of secondary decomposition and overflow is low. Additionally, both the nitration filtrate and crystal transformation filtrate exhibit low thermal hazards. These collective findings indicate that the acetic anhydride-based HMX synthesis process maintains relatively safe operational characteristics under standard processing conditions.

To reduce mold costs in composite forming, multi-point tooling technology has been integrated into the hot diaphragm forming process. However, this approach still faces several challenges, including time-consuming prepreg layup, high energy consumption, and poor surface quality. This study proposes a heating pad-assisted multi-point thermoforming process: the prepreg is embedded in the thermal functional layers, placed on the lower mold, and formed via the downward movement of the upper mold to accomplish mold closure. Instead of the conventional rectangular array, this study adopted multi-point tooling with a hexagonal pin arrangement. Compared to traditional configurations, this hexagonal layout increases the punch support area by 9.8%, while its dense punch arrangement improves the accuracy of the molded curved surface. Taking a saddle-shaped surface as the target, a prototype part was fabricated. Subsequent analysis of the part’s surface quality identified three defects: dimples, fiber distortion, and ridge protrusions. The surface dimples were eliminated by adjusting the distance between the upper and lower molds. Notably, ridge protrusion is a defect unique to the hexagonal pin arrangement. We conducted a detailed analysis of its causes and solutions, finding that this defect arises from the combined effect of the pin arrangement and the saddle-shaped surface. Through a series of height compensation experiments, the maximum deviation at the ridges was reduced from 0.46 mm to approximately 0.35 mm, which is consistent with the deviation of defect-free areas. This work demonstrates that the multi-point hot-pressing process provides a potential, efficient, and low-cost method for manufacturing double-curvature composite components, whose effectiveness has been verified through the saddle-shaped case study.