ネットゼロ・モジュラー建築に向けた相変化材料内蔵粘土レンガのAI支援統合

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:AIサロゲート×LCA×建材の統合手法と限界（ヒステリシス未考慮）を参照できる。

🏢実務担当者:PCM建材の省エネ・炭素回収年数の目安を建材選定や改修投資判断に活用可能。

🏛政策担当者:建築物の炭素回収期間や省エネ基準設計にPCM・AI最適化を組み込む根拠になり得る。

Purpose This study develops an exploratory artificial intelligence (AI)-assisted framework for integrating phase change material (PCM)-embedded clay bricks into modular wall assemblies to support net-zero building performance. The work addresses the current absence of scalable, climate-adaptive PCM–modular integration strategies and evaluates their potential for energy savings, thermal stability and lifecycle carbon reduction. Design/methodology/approach A multi-scale methodology was adopted, combining EnergyPlus v26.1 building simulations, a representative laboratory thermal-cycling dataset and a hypothetical Toronto case study. PCM-embedded bricks containing 10 wt% microencapsulated paraffin (melting point 24°C) were modeled within modular wall panels. Annual simulations were conducted across diverse Köppen climate zones, including per-zone re-optimization of PCM melting point, loading fraction and wall placement using a Physics-informed neural network (PINN) surrogate model coupled with a genetic algorithm (GA). A streamlined ISO 14040/44 lifecycle assessment (LCA) estimated embodied carbon, operational savings and carbon payback. AI-assisted outputs are treated as qualitative due to limited empirical validation. Findings Simulations indicate cooling energy savings of approximately 25% ± 5% in temperate climates (London) and up to 30% ± 5% in hot-arid climates (Dubai), with peak indoor temperature reductions of 3–4°C. All energy savings are upper-bound estimates, as PCM thermal hysteresis (1–4°C band, reducing effective storage by 15–30%) was not modeled. Laboratory cycling demonstrated a 66% improvement in thermal stability with no leakage across 100 cycles. The modeled LCA projects an additional 8.8 kg CO2eq/m2 of embodied carbon and an estimated carbon payback of approximately 3.4 years under Toronto conditions. Economic payback was estimated at approximately 10.6 years based on current Toronto electricity pricing. AI-assisted optimization reduced design iteration by approximately 40% compared to conventional parametric sweeps. Originality/value This study is the first to link AI-assisted PCM optimization (PINN surrogate + GA), modular precision manufacturing and PCM-embedded clay bricks within a unified framework for net-zero construction. It advances the field by integrating multi-climate simulation with per-zone parameter re-optimization, a small-scale laboratory cycling dataset (100 cycles), lifecycle modeling with inflation and end-of-life scenarios, and a formal sensitivity analysis, establishing a preliminary pathway toward scalable, low-carbon and climate-adaptive modular building envelopes.

• openalex https://doi.org/10.1108/sasbe-05-2026-0401first seen 2026-10-09 04:53:56
• semanticscholar https://doi.org/10.1108/sasbe-05-2026-0401first seen 2026-10-10 05:47:32