Teodosio, Bertrand ORCID: 0000-0002-3909-4054, Pallewela Liyanage, Piyal wasantha ORCID: 0000-0002-7016-4524, Yaghoubi, Ehsan ORCID: 0000-0003-0639-0225, Guerrieri, Maurice ORCID: 0000-0001-7916-7003, Fragomeni, Salvatore ORCID: 0000-0002-0733-4770 and van Staden, Rudi ORCID: 0000-0002-7339-7702 (2023) Environmental, economic, and serviceability attributes of residential foundation slabs: a comparison between waffle and stiffened rafts using multi-output deep learning. Journal of Building Engineering, 80. ISSN 2352-7102

Abstract

The design of residential foundation slabs is commonly based on standards that emphasise the structural aspects and safety of the structure. Factors related to environmental and economic criteria are seldom given due consideration in the design phase. Considering the growing demand for sustainable approaches driven mainly by climate change concerns, this study developed a smart tool called Multi-OUtput Non-linear Design of Slabs (MOUNDS), which simultaneously predicts embodied energy, carbon emission, life cycle cost and deflection of waffle and stiffened rafts. MOUNDS considers the environmental, economic, and serviceability criteria of waffle and stiffened rafts on soils having varying reactivities. The standard deemed-to-comply design code for residential slabs and footings in Australia was investigated to determine the most advantageous foundation type in terms of both sustainability attributes and serviceability performances. The developed MOUNDS algorithm has shown accurate predictions. The predicted values of the: embodied energy of the residential slabs, greenhouse gas emission of the residential slabs, life cycle cost of the residential slabs, and maximum deflection of the residential slabs of waffle rafts were found more sustainable and serviceable than stiffened rafts in slight to moderate reactive sites. When sites are highly reactive, the difference between the environmental and economic of waffle rafts and stiffened rafts was minimal and did not conform to the serviceability limits of the Australian design code. This novel study linked and predicted the multi-disciplinary relationship between the environmental, economic and structural design aspects of residential slabs through machine learning. This is valuable in decision-making throughout the design phase considering the multi-faceted aspects of residential footing systems on reactive soils.

Search Google Scholar

Repository staff login