Search

The desire of using sustainable materials has reignited the interest in timber-based construction. Researchers and practitioners are developing novel timber-based structural solutions. Cross-laminated timber (CLT)-coupled wall is a recently proposed system for potential use in mid- and high-rise timber construction. The National Building Code of Canada, however, does not include this system and, consequently, the seismic force modification factors are not available. This study evaluated the ductility-related force modification factor ([R.sub.d]) using the FEMA P-695 procedure. Nine archetype buildings were designed considering different design parameters: building storey height, CLT wall configuration, and coupling ratios. Using 30 ground motion records (bi-directional), rigorously selected for seismicity of Vancouver, BC, Canada, incremental dynamic analyses were performed. Collapse margin ratios were calculated to assess the adequacy of the trial Rd factors. Using an over-strength factor of 1.5, [R.sub.d] = 4 is found to be acceptable for this system. Key words: cross-laminated timber (CLT), CLT-coupled walls, seismic performance of timber systems, coupling beam, coupling ratio (CR), over-strength factor ([R.sub.o]), ductility factor ([R.sub.d]), 1. Introduction In the move toward sustainable construction, use of timber and timber-based products is considered as a viable solution (Foster et al. 2016). Due to its structural rigidity in [...]

While urban heat islands (UHIs) have been thoroughly studied in cities worldwide, the specific risks and vulnerabilities related to urban heat in Ethiopia and Africa as a whole has given less attention. Urbanization often replaces green spaces with impervious surfaces, which diminishes natural cooling, precipitation, and water infiltration. This change can significantly affect land surface temperatures (LST) and contribute to UHI formation and its impacts. This study aimed to identify and assess the risk factors linked to UHIs, focusing on pinpointing the most vulnerable areas within cities using principal components explanatory factor analysis (HV-PC-EFA) and the urban heat vulnerability index (UHVI) model. The analysis utilized 19 composite indicators under well-established categories such as exposure, sensitivity, and adaptive capacity to assess potential risk zones. The results from the two models were compared, and their variations were examined. In the HV-PC-EFA model, indicators like urban density and roof type, along with adaptive factors such as vegetation cover, urban thermal field variances, and relative humidity, were not distinctly separated as individual components. This may cause discrepancies in the final outcomes, impacting the spatial distribution and extent of vulnerability. Despite observing some spatial variations in identifying risk areas, the study provides a broad perspective essential for developing evidence-based policies and strategies to enhance cities' resilience to high temperatures and promote sustainable environments. Given the challenges in modifying existing infrastructure, it is practical to regularly implement adaptive measures, such as preserving and restoring urban water bodies, planting trees, creating green public spaces, and raising public awareness about these risks. [ABSTRACT FROM AUTHOR]

Corrosion poses a great risk to the integrity of oil and gas pipelines, leading to substantial investments in corrosion control and management. Several studies have been conducted on accurately estimating the maximum pitting depth in oil and gas pipelines using available field data. Some of the frequently employed machine learning techniques include artificial neural networks, random forests, fuzzy logic, Bayesian belief networks, and support vector machines. Despite the ability of machine learning methods to address a variety of problems, traditional machine learning methods have evident limitations, such as overfitting, which can diminish the model's generalization capabilities. Additionally, traditional machine learning models that provide point estimations are incapable of addressing uncertainties. In the current study, a Bayesian neural network is proposed to include uncertainty in estimating the corrosion defect of a pipeline exposed to external pitting corrosion. The results are then incorporated into a Bayesian belief network for evaluating the probability of failure and its corresponding consequences in terms of social impact, thus forming a comprehensive risk assessment framework. The results of the Bayesian neural network are validated using field data and achieved a testing accuracy of 90%. The framework of the study offers a powerful decision‐making tool for the integrity management of pipelines against external corrosion. [ABSTRACT FROM AUTHOR]