Machine learning for structural engineering: A state-of-the-art review

Predicting the compressive strength of concrete containing metakaolin with different properties using ANN

This review article proposes the identification and basic concepts of materials that might be used for the production of high-performance concrete (HPC) and ultra-high-performance concrete (UHPC). Although other reviews have addressed this topic, the present work differs by presenting relevant aspects on possible materials applied in the production of HPC and UHPC. The main innovation of this review article is to identify the perspectives for new materials that can be considered in the production of novel special concretes. After consulting different bibliographic databases, some information related to ordinary Portland cement (OPC), mineral additions, aggregates, and chemical additives used for the production of HPC and UHPC were highlighted. Relevant information on the application of synthetic and natural fibers is also highlighted in association with a cement matrix of HPC and UHPC, forming composites with properties superior to conventional concrete used in civil construction. The article also presents some relevant characteristics for the application of HPC and UHPC produced with alkali-activated cement, an alternative binder to OPC produced through the reaction between two essential components: precursors and activators. Some information about the main types of precursors, subdivided into materials rich in aluminosilicates and rich in calcium, were also highlighted. Finally, suggestions for future work related to the application of HPC and UHPC are highlighted, guiding future research on this topic.

https://www.mdpi.com/1996-1944/14/15/4304/pdf

Materials for Production of High and Ultra-High Performance Concrete: Review and Perspective of Possible Novel Materials.

Revealing the nature of metakaolin-based concrete materials using artificial intelligence techniques

Predicting Compressive Strength of Manufactured-Sand Concrete Using Conventional and Metaheuristic-Tuned Artificial Neural Network

Producing high performance concrete with a low reactivity pozzolan such as metakaolin is challenging. On the other hand, high performance concretes could be more prone to shrinkage cracking due to their higher cement content . No consensus exists in the literature on how metakaolin affects concrete's shrinkage. In this study, the possibility of making high performance concrete with a low reactivity metakaolin was assessed. Moreover, the effect of using a low reactivity metakaolin on the mechanical properties, durability, and free and restrained shrinkage of high performance concrete was evaluated. The results showed that increase in the percentage of metakaolin replacement led to reduction in the mechanical properties and durability of high performance concrete. The effect of metakaolin on shrinkage of high performance concrete was observed to be dependent on the water to cement ratio . Overall, it is possible to produce high performance concrete with a low reactivity metakaolin (up to 20% replacement level) only at low water to cement ratios.

Studying the effect of low reactivity metakaolin on free and restrained shrinkage of high performance concrete

Perforations adversely affect the structural response of unreinforced masonry walls (UMW) by reducing the wall’s load bearing capacity, which can cause serious structural damage. In the absence of a reliable procedure to accurately predict the load bearing capacity and stiffness of perforated masonry walls subjected to in-plane loadings, this study presents a novel approach to measure these parameters by developing simple but practical equations. In this regard, the Multi-Pier (MP) method as a numerical approach was employed along with the application of an Artificial Neural Network (ANN). The simulated responses of centrally perforated UMW by the MP method were validated utilizing full-scale experimental walls. The validated MP model was used to generate a simulated database. The simulated database includes results of analyses for 49 different configurations of perforated masonry walls and their corresponding solid masonry walls. The effect of the area and shape of the perforations on the UMW’s behavior was evaluated by the MP method. Following the outcomes of the verified MP method, the ANN is trained to develop empirical equations to accurately predict the reduction in the load bearing capacity and initial stiffness due to the perforation of UMW. The results of this study indicate that the perforations have a significant effect on the structural capacity of the UMW subjected to in-plane loadings.

Application of Artificial Neural Network to Predict Load Bearing Capacity and Stiffness of Perforated Masonry Walls

Evaluating the behaviour of centrally perforated unreinforced masonry walls: Applications of numerical analysis, machine learning, and stochastic methods

In this paper, hydroforming process of bi-layered bellows is investigated experimentally and numerically. The process consists of two steps: tube bulging and tube folding. For this purpose, the effects of two main process parameters i.e. internal pressure and die stroke on characterizations of hydroformed bi-layered bellows are examined. The convolution height and thickness of the top point of each convolution are selected as the main features of metallic bellows. The results show that the numerical simulations are in good agreement with experimental measurements. It is demonstrated from the obtained numerical and experimental results that the thickness reduction is noticeably increased from bottom point to top point of each convolution. In addition, it is concluded that the inner layer has the most reduction in wall thickness while the outer layer has the least thickness reduction due to the supportive role of the inner layer for the outer layer. The experimental and numerical results showed that a bi-layered bellows can successfully fabricated with internal pressure of 230 Bar and die stroke of 14 mm. Also, by changing the internal pressure from 70 to 230 Bar in a situation where the die stroke is constant and equal to 14 mm, the convolution height will increase by about 62% while the thickness of top point of outer layer will decrease by about 20%. Changing the die stroke from 6 to 16 mm in a situation where the internal pressure is constant and equal to 230 bar leads to a 60% increase in the convolution height and a 23% decrease in the thickness of top point of outer layer.

Javid Salimi

Papers

Predicting the compressive strength of concrete containing metakaolin with different properties using ANN

Studying the effect of low reactivity metakaolin on free and restrained shrinkage of high performance concrete

Application of Artificial Neural Network to Predict Load Bearing Capacity and Stiffness of Perforated Masonry Walls

Evaluating the behaviour of centrally perforated unreinforced masonry walls: Applications of numerical analysis, machine learning, and stochastic methods

Experimental and numerical investigation of hydroforming process of bi-layered metallic bellows