Computational predictions for estimating the performance of flexural and compressive strength of epoxy resin-based artificial stones

The main goal of this study is to predict the flexural and compressive strength of artificial stones consisting of epoxy resin and a combination of travertine stone powder, fine sand, and travertine stone sludge. Three groups of stones with different mixtures were fabricated, including a combination of stone sludge and powder, sand and stone powder, and a combination of all of them. Then, we conducted flexural and compressive tests based on ASTM standards. All tests were repeated three times for each mix proportion. Subsequently, we present simple prediction models. The volume fraction of stone powder, sand, epoxy resin, stone sludge, the tensile strength of the epoxy resin, and the curing temperature were selected as their input variables. Based on the number of variables and their levels, the Taguchi approach was employed to design mix proportions in three groups, and the results of these groups were developed for all variables. Five prediction models, including step-by-step regression (SBSR), the combination of stronger variable creator machine (SVCM) and SBSR, gene expression programming (GEP), the combination of SVCM and GEP, and adaptive neuro-fuzzy inference system (ANFIS), were utilized to predict the compressive and flexural strengths of the stones. All models were compared using statistical parameters and error terms. ANFIS performs better than all other models. Moreover, the samples based on epoxy resin 828 are the strongest, while the samples using epoxy resin 557 are more flexible.

Automated summary

The main goal of this study is to predict the flexural and compressive strength of artificial stones made of epoxy resin and various combinations of stone powder, sand, and stone sludge. Through experiments and modeling, it was found that samples using epoxy resin 828 have the highest strength, while samples using epoxy resin 557 are more flexible.