1 March 2022 in Ultrasonics
1 January 2019 in Latin American Journal of Solids and Structures
Deep beams do not behave according to classical beam theory. The nonlinearity of strain distribution within these elements requires application of strut and tie models (STM) or other alternatives to evaluate the complex stress field. Although the design of these elements is a common task for structural engineers, limited research is found on assessing effectiveness of the results. The purpose of this work is to compare, in a systematic approach, different design solutions for a deep beam using selected performance metrics which are strain energy, reinforcement ratio, maximum load, structural efficiency, safety factor and cracking behavior. A deep beam (2.85 m of height, 4.20 m of length and 0.2 m of thickness) with a square opening (0.7 m x 0.7 m) close to one of the supports was subjected to a uniform loading at the top surface while resting on supports at both ends. A simplified finite element model (FEM) of this beam was developed simulating concrete with elastic linear stress-strain behavior and disregarding steel reinforcements. This model allowed determination of elastic stress fields necessary to subsequent analyses. Four STM were then developed, supposing the total load respectively represented by one (STM-1), two (STM-2), four (STM-4) or eight (STM-8) concentrated loads equally spaced along the top of the beam. Additionally, an in-plane stress field method (SFM) was applied to the design of the same beam subjected to uniform loading on the top surface. After design and detailing of the reinforcement for each situation, nonlinear FEMs were used to predict the ultimate conditions. The strain energy reduced significantly comparing results from STM-1 to STM-2 and subsequently to STM-4 and remained at a low level in STM-8 and SFM. The reinforcement ratio reduced systematically from STM-1 to STM-8, was minimum with the SFM and the same behavior was followed by maximum load. The structural efficiency (maximum load/reinforcement ratio) increased from STM-1 to STM-8, with maximum efficiency at STM-8 and was slightly below with SFM. The safety factor reduced systematically from STM-1 to STM-8 and was slightly lower with SFM, but in all cases was above acceptable limits found in design codes.
1 September 2022 in Revista ALCONPAT
This paper analyzes the impact of corrosion on the global stability of reinforced concrete columns, evaluating the effectiveness of the stiffness reduction criteria proposed by ABNT NBR 6118:2014. It was defined through finite element models that corrosion causes the intensification of the second-order effects. However, in critical situations, the failure in the transversal section led to structural collapse even before the stability loss. The study was conducted through nonlinear geometric analyses and applying the γz, coefficient. It was concluded that the subsequent addition of rheological effects could exceed the limits proposed by the ABNT NBR 6118:2014.
1 January 2024 in Revista IBRACON de Estruturas e Materiais
Nondestructive tests that assess the constitution or degradation of structures are of great interest in Civil Engineering. Among the non-destructive testing techniques, the Ultrasonic Pulse Velocity (UPV) test stands out; however, although its use is widespread, there are still no applications that employ this method to determine the constitution of concrete in situ. Therefore, this article addresses the identification of the coarse aggregate content in concrete specimens by an Artificial Neural Network (ANN) trained with a database of numerical tests that simulated UPV. In this paper, the coarse aggregate content will be described as a percentage of the total area of a two-dimensional concrete model. Three artificial neural network architectures were evaluated. The first two, trained with 13 or 22 paths, solved a classification problem for five aggregate contents, and the third, trained with 22 paths, solved a regression problem. Its performance was compared with those of other regression solutions, namely XGB Regressor, Random Forest, and OLS (Ordinary Least Squares), and showed superior, with -2.55% to +2.17% average deviations. Thus, this paper demonstrated that the use of ANN in combination with UPV test has the potential to identify the coarse aggregate content in concretes. The positive results suggest that this approach is promising and highlights the need for further experimental validation in future research.
6 December 2022 in Revista AIDIS de Ingeniería y Ciencias Ambientales. Investigación, desarrollo y práctica
AMAndré Aparecido de Moraes
André Aparecido de Moraes
Com o crescimento do setor da construção civil, foi observado um aumento considerável dos níveis de geração de resíduos da construção e demolição (RCD), que causam grandes impactos ambientais, ficando evidente a necessidade de soluções para modificação desse cenário. Assim, o emprego de agregados reciclados (AR) provenientes dos RCDs que apresentem aspectos técnicos para sua utilização, de forma a garantir resistência ao novo concreto produzido, vem se destacando por ser considerado um meio sustentável para conservação dos recursos naturais. Entretanto, por se tratar da incorporação de agregados com características distintas do natural, e grande variabilidade, faz-se necessário adequar os modelos para previsão de resistência do concreto curado. Neste sentido, a utilização de Redes Neurais Artificiais (RNAs), que são modelos matemáticos capazes de reconhecer padrões através do treinamento numérico, se colocam como valiosa alternativa para um método de previsão da resistência à compressão de espécimes de concretos com agregados reciclados. Entretanto, em função da variabilidade dos materiais o desafio de se obter resultados satisfatórios com um banco de dados reduzido se impõe. Portanto, para alcançar tal objetivo, foi necessário o levantamento de uma base de dados plural com concretos reciclados de diferentes origens de modo a conduzir a rede neural ao grau de generalização desejado. Por fim, foi proposta uma arquitetura de RNA capaz de prever a resistência final do concreto com precisão aceitável. Os resultados obtidos foram promissores, de modo que a arquitetura proposta, treinada com o banco de dados selecionado, apresentou resultados com um desvio médio de aproximadamente 3 MPa quando comparada ao conjunto de amostras de validação.