Simulation-Based Analysis of the Thermal Regulation Effects of Traditional Skywell Layouts in Modern Relocation Projects
Year: 2025
Type: article
Abstract: Traditional skywells, relying on the coupled thermodynamic mechanisms of natural ventilation, heat storage and release, and radiative heat transfer, exhibit outstanding passive thermal regulation capabilities and offer valuable references for low-energy building design.However, during modern relocation projects, increasing conflicts emerge between traditional skywell morphologies and contemporary thermal performance requirements as well as urban spatial constraints.Conventional forward design approaches based on iterative "parameter adjustment-simulation validation" are often inefficient and fail to ensure precise thermodynamic performance matching, thereby limiting the scientific rigor and feasibility of relocation practices.To address the coordinated challenge of traditional form preservation, thermodynamic performance optimization, and relocation scenario adaptability, this study aims to quantitatively evaluate and optimize the thermal regulation effects of skywells after relocation.A thermodynamically constrained conditional generative adversarial network (cGAN)-based inverse generation framework is proposed, overcoming the limitations of traditional forward workflows and enabling efficient backpropagation from target thermal environment indicators to optimal skywell design parameters.Using representative traditional skywell relocation projects as case studies, a comprehensive analytical framework is established, integrating thermodynamic mechanism analysis, inverse generative optimization, and performance verification.The framework systematically covers multiple relocation scenarios, including single-building relocation, clustered relocation, and cross-climatic relocation.The results demonstrate that the proposed model achieves highly efficient and accurate optimization of thermal environments in relocated traditional skywells.Compared with conventional forward design methods, the computational time per design scheme is reduced by 99.3%, while the prediction errors of temperature and heat flux are limited to 2.1% and 3.2%, respectively.Across the three typical relocation scenarios, the optimized solutions exhibit consistently superior performance: unit-area heat loss is reduced by 22.2% in single-building relocation, thermal storage efficiency increases by 13 percentage points in cross-climatic relocation, and all generated schemes achieve a 100% compliance rate in terms of traditional morphological compatibility.This study provides a thermodynamically guided and datadriven optimization pathway for the thermal environment design of relocated traditional skywells, enriching the methodological framework for adaptive reuse of architectural heritage and offering critical technical support for traditional building relocation projects in the context of low-carbon urban development.
DOI: 10.18280/ijht.430638
Institutions: Huangshan University, Anhui Jianzhu University
Subfield: Building and Construction
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7125522201
FWCI:
Cites:
Cited by:
Related to: