Mechanical and Durability Properties of Concrete Incorporating LD Slag

Abstract

Concrete usage in recent decades has skyrocketed, which has led to immense pressure on the natural resources such as river sand and coarse aggregates involved in its production. This continuous consumption of natural resources is due to increased industrialization and urbanization. The circular economy concept, i.e., usage and recycling of waste products, can be incorporated into concrete production to alleviate the overutilization of natural resources. Overutilization of natural resources has impeded the movement towards green concrete in construction, i.e., utilizing waste materials in its production, steel slag waste being one of them. Linz-Donawitz slag (LD Slag) is one of the steel slag generated during the Basic Oxygen Furnace (BOF) steel-making process. Accumulation of LD slag has become cumbersome in India, posing a problem of landfills, aggravating limited land availability and environmental degradation, i.e., abating soil and water quality due to the leaching of toxic and heavy metals. Previous studies employed LD Slag as an alternative to natural aggregates in concrete production. Moreover, fewer experimental studies have examined whether weathered fine LD Slag can substitute for fine aggregates in concrete. However, none of the studies has focused on the comparative behavior of weathered coarse LD slag and fine LD slag incorporated in Normal Strength Concrete (NSC) and metakaolin-based High-strength Concrete (HSC). The present experimental study focuses on determining the ideal replacement level of coarse and fine aggregates with LD Slag, one at a time. Natural aggregates were replaced volumetrically, one at a time, with LD Slag at 25, 50, 75 and 100%. Physico-mechanical and durability properties such as hardened concrete density, ultrasonic pulse velocity, compressive strength, split tensile strength, flexural strength, modulus of elasticity, drying shrinkage, water absorption, sorptivity and abrasion resistance were evaluated. Furthermore, different data sets, one for mixes incorporating LD slag coarse aggregate (LDCA) and another for mixes incorporating LD slag fine aggregate (LDFA), were taken to predict the compressive strength of concrete using artificial intelligence (AI) methods such as Artificial Neural Networks (ANNs), Decision Trees (DT) and Random Forests (RF). The coefficient of determination R2, Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) were computed to assess the performance of generated models. Besides, a sensitivity analysis technique was employed to determine the most influential parameter, among cement, Metakaolin, Fine Aggregate, Coarse Aggregate, LD slag, Superplasticizer, water content and testing age, affecting the compressive strength results. Results showed that replacing 75% of coarse aggregates with LDCA displayed the best results for said properties compared to the control mix for NSC and HSC. Compressive strength, split tensile strength, flexural strength and modulus of elasticity were improved in the range of 7-9%, 19-29%, 6-11% and 7-16% for 75% replacements at 90 days of testing, respectively. The sorptivity coefficient was also reduced compared to control mixes at all testing ages, along with lower water absorption, abrasion, and drying shrinkage. While a 75% replacement of natural aggregates with LDCA yielded the most favourable test results for the examined properties, complete replacements could be contemplated, yielding nearly comparable results to control mixes in NSC and HSC. On the contrary, substituting fine aggregates with LDFA up to 50% for NSC and 25% for HSC yielded superior test results for various properties compared to traditional concrete mixes. At 90 days of testing, there was an approximate 6% increase in compressive strength, a 2-12% enhancement in splitting tensile strength, a 6-16% improvement in flexural strength, and a minor increment of 1- 4% in MOE, showcasing the effectiveness of the LDFA replacements. The durability of concrete incorporating LDFA was also enhanced with the mentioned replacements. In addition, replacements of 75% of fine aggregates with LDFA in NSC and 50% in HSC can be deemed suitable, offering comparable results to control mixes. Among the proposed AI techniques, the DT technique emerged as the top-performing model for both datasets analyzed. For the dataset involving mixes with LDCA, DT exhibited the highest R2 value of 0.957, with the lowest MAE recorded at 4.065 MPa and RMSE of 3.220 MPa. Conversely, for the dataset concerning mixes with LDFA, the R2 value was 0.973, with the least MAE (3.534 MPa) and RMSE (4.409 MPa) compared to other models. Sensitivity analysis conducted on the DT model revealed that testing age was the most influential parameter for predicting compressive strength in both scenarios. Increasing the proportion of LDCA in concrete mixes results in the production of heavyweight concrete due to its higher specific gravity compared to natural coarse aggregates. This heavyweight concrete finds potential applications in various fields such as hospital radiation rooms/bunkers, nuclear waste containers, and tunnels. Furthermore, concretes containing LDCA or LDFA can be explored in unreinforced concrete products like paver blocks, building bricks/blocks, gravity retaining walls, concrete kerbs/curbs, and concrete tiles.