Conventional neural networks tend to fall into local extremum on large datasets, while the research on the strength of rubber concrete using intelligent algorithms to optimize artificial neural networks is limited. Therefore, to improve the prediction accuracy of rubber concrete strength, an artificial neural network model with hybrid algorithm optimization was developed in this study. The main strategy is to mix the simulated annealing (SA) algorithm with the particle swarm optimization (PSO) algorithm, using the SA algorithm to compensate for the weak global search capability of the PSO algorithm at a later stage while changing the inertia factor of the PSO algorithm to an adaptive state. For this purpose, data were first collected from the published literature to create a database. Next, ANN and PSO-ANN models are also built for comparison while four evaluation metrics, MSE, RMSE, MAE, and $R^2$, were used to assess the model performance. Finally, compared with empirical formulations and other neural network models, the result shows that the proposed optimized artificial neural network model successfully improves the accuracy of predicting the strength of rubber concrete. This provides a new option for predicting the strength of rubber concrete.

The addition of rubber to concrete improves resistance to chloride ion attacks. Therefore, rapidly determining the chloride permeability coefficient (D_CI) of rubber concrete (RC) can contribute to promotion in coastal areas. Most current methods for determining D_CI of RC are traditional, which cannot account for multi-factorial effects and suffer from low prediction accuracy. Machine learning (ML) techniques have good non-linear learning capabilities and can consider the effects of multiple factors compared with traditional methods. However, ML models easily fall into the local optimum due to their parameters’ influence. Therefore, a mixed whale optimization algorithm (MWOA) was developed in this paper to optimize ML models. The main strategies are to introduce Tent mapping to expand the search range of the algorithm, to use an adaptive t-distribution dimension-by-dimensional variation strategy to perturb the optimal fitness individual to thereby improve the algorithm’s ability to jump out of the local optimum, and to introduce adaptive weights and adaptive probability threshold values to enhance the adaptive capacity of the algorithm. For this purpose, data were collected from the published literature. Three machine learning models, Extreme Learning Machine (ELM), Random Forest (RF), and Elman Neural Network (ELMAN), were built to predict the D_CI of RC, and the three models were optimized using MWOA. The calculations show that the MWOA is effective with the optimized ELM, RF, and ELMAN models improving the prediction accuracy by 54.4%, 62.9%, and 36.4% compared with the initial model. The MWOA-ELM model was found to be the optimal model after a comparative analysis. The accuracy of the multiple linear regression model (MRL) and the traditional mathematical model is calculated to be 87.15% and 85.03%, which is lower than that of the MWOA-ELM model. This indicates that the ML model that is optimized using the improved whale optimization algorithm has better predictive ability than traditional models, providing a new option for predicting the D_CI of RC.

Rubber concrete (RC) exhibits high durability due to the rubber admixture. It is widely used in a large number of fatigue-resistant structures. Mesoscale studies are used to study the composition of polymers, but there is no method for fatigue simulation of RC. Therefore, this paper presents a finite element modeling approach to study the fatigue problem of RC on the mesoscale, which includes the random generation of the main components of the RC mesoscale structure. We also model the interfacial transition zone (ITZ) of aggregate mortar and the ITZ of rubber mortar. This paper combines the theory of concrete damage to plastic with the method of zero-thickness cohesive elements in the ITZ, and it is a new numerical approach. The results show that the model can simulate reasonably well the random damage pattern after RC beam load damage. The damage occurred in the middle of the beam span and tended to follow the ITZ. The model can predict the fatigue life of RC under various loads.

Foam concrete is a typical cement-based porous material; its special microstructure endows it with excellent properties, such as light weight, energy efficiency, thermal insulation, and fire resistance. Therefore, it is widely used as a thermal insulation material for buildings. The heat transfer modes of foam concrete include conduction, convection, and radiation. However, previous studies considered conduction to be the dominant mode, often neglecting the effects of convection and radiation. In this study, a stochastic numerical model of the foam concrete microstructure is established based on the statistical parameters of the pore structure. With this model, the heat transfer mechanism of foam concrete is analyzed at the mesoscopic level, and the equivalent thermal conductivity is calculated. By comparing four different working conditions, the influence of conduction, convection, and radiation on the heat transfer of foam concrete is analyzed, and the specific contribution rates of conduction, convection, and radiation are calculated. The results show that the convection effect is weak due to the pore size being smaller than 1 $\mathrm{mm}$; so, the influence of convection can be neglected in the heat transfer analysis of foam concrete. The contribution of radiation increases with the decrease in foam concrete density and the increase in temperature difference. When the temperature difference is 40 °C and the density is 300 $\mathrm{kg}/{\mathrm{m}}^3$, the contribution of radiation exceeds 20$\%$. Therefore, for low-density and high-temperature difference situations, the influence of radiation cannot be ignored. The heat transfer in foam concrete is mainly through conduction, but with the decrease in density and the increase in temperature difference, the contribution of conduction shows a downward trend. Nevertheless, the contribution of conduction is still much larger than that of radiation and convection.

The incorporation of rubber can enhance concrete’s durability and effectively reduce the damage caused by freeze-thaw cycling (FTC). Still, there has been only limited research on the damage mechanism of RC at the fine view level. To gain insight into the expansion process of uniaxial compression damage cracks in rubber concrete (RC) and summarize the internal temperature field distribution law during FTC, a fine RC thermodynamic model containing mortar, aggregate, rubber, water, and interfacial transition zone (ITZ) is established in this paper, and the cohesive element is selected for the ITZ part. The model can be used to study the mechanical properties of concrete before and after FTC. The validity of the calculation method was verified by comparing the calculated results of the compressive strength of concrete before and after FTC with the experimental results. On this basis, this study analyzed the compressive crack extension and internal temperature distribution of RC at 0, 5, 10, and 15% replacement rates before and after 0, 50, 100, and 150 cycles of FTC. The results showed that the fine-scale numerical simulation method can effectively reflect the mechanical properties of RC before and after FTC, and the computational results verify the applicability of the method to rubber concrete. The model can effectively reflect the uniaxial compression cracking pattern of RC before and after FTC. Incorporating rubber can impede temperature transfer and reduce the compressive strength loss caused by FTC in concrete. The FTC damage to RC can be reduced to a greater extent when the rubber incorporation is 10%.