Experimental and Machine Learning-Based Performance Evaluation of Ambient-Cured Geopolymer Mortar Using Microstructural Characterisation and ABAQUS Numerical Analysis

The Green Concrete Revolution: A Story of Ash, Alchemy, and AI

Imagine the construction industry as a giant, hungry beast that eats up the planet's resources. For over a century, this beast has feasted on Portland cement, a material that is incredibly strong but also a massive polluter, churning out huge clouds of carbon dioxide every time it's made. Scientists have been searching for a "green" alternative for years, and they've found a fascinating candidate: geopolymer mortar. Think of it not as a new invention, but as a magical transformation. Instead of baking clay to make cement, this new material takes industrial waste—like the ash left over from burning coal or the slag from making steel—and "wakes it up" using a special chemical soup called an alkaline activator. It's like taking the leftover crumbs of a factory's day and turning them into a super-strong, durable brick without the heavy carbon footprint.

But here's the tricky part: making this "magic dust" work perfectly is like trying to bake the perfect cake without a recipe. You have to mix different types of waste (the "flour"), choose the right chemical soup (the "yeast"), and decide how long to let it sit (the "baking time"). If you get the mix wrong, the result is weak and crumbly. If you get it right, you get a material that is stronger than traditional concrete and kinder to the planet. This is the puzzle that researchers are trying to solve: How do we mix these industrial leftovers to create the strongest, most reliable building material possible, all while curing it at normal room temperature instead of in a super-hot oven?

The Experiment: Mixing the Perfect "Green" Recipe

In this study, researchers Rushikesh Khartode and Subhash Patankar decided to play the role of master alchemists. Their goal was to find the ultimate recipe for geopolymer mortar that could be made and hardened right in a normal workshop, without needing expensive heat. They gathered four different types of "flour" (precursor materials): Fly Ash (from coal plants), Ground Granulated Blast Furnace Slag (GGBFS, from steel mills), Metakaolin (a clay derivative), and Steel Slag. They also tested different "yeasts" (alkaline activators), mostly focusing on sodium-based and potassium-based chemical solutions.

They didn't just guess; they ran a massive series of experiments. They created dozens of different mixtures, changing the ratios of the ingredients. Some mixes were 100% Fly Ash, others were 100% GGBFS, and many were hybrids, like a 50-50 split. They poured these mixtures into molds and let them sit at room temperature, checking their strength after 7, 14, and 28 days. They also measured how "flowy" the wet mortar was, because if it's too thick, it's hard to pour into cracks for repairs.

The Big Discovery:
The results were clear. The mix made entirely of 100% GGBFS was the undisputed champion. After 28 days, it reached a crushing strength of 68.51 MPa, which is incredibly high for a material cured at room temperature. The runner-up was a hybrid mix of 50% Fly Ash and 50% GGBFS, which hit 61.44 MPa. This hybrid was a sweet spot, offering a great balance of strength and workability.

On the other hand, the mixes relying heavily on Metakaolin or Steel Slag didn't perform as well under these specific room-temperature conditions. They were weaker, suggesting that without the right amount of calcium (which GGBFS has in abundance) and the right chemical "wake-up call," these materials just didn't harden as effectively. The researchers also found that a sodium-based chemical soup (NaOH + Na2SiO3) worked much better than potassium or calcium-based ones for this specific job.

The Crystal Ball: Using AI to Predict the Future

Making concrete is expensive and time-consuming. You have to mix, pour, wait weeks, and then crush it to see if it worked. To speed this up, the researchers turned to a digital crystal ball: Machine Learning. They fed their experimental data into four different computer brain models: Artificial Neural Networks (ANN), Gaussian Process Regression (GPR), Random Forest (RF), and Support Vector Machine (SVM).

Think of these models as four different students trying to guess the final exam score based on how much homework they did. The Artificial Neural Network (ANN) was the star student. It learned the complex, non-linear rules of the chemistry better than the others. It predicted the strength of the mortar with amazing accuracy, matching the real-world results almost perfectly. The other models were good, but the ANN was the most reliable.

To make sure the computer wasn't just guessing, the researchers used a tool called SHAP analysis. This is like asking the computer, "Why did you think this mix would be strong?" The computer pointed its finger at the most important ingredients: the amount of GGBFS, the concentration of the sodium hydroxide, and the dose of sodium silicate. It confirmed what the experiments showed: these are the keys to the kingdom.

The Microscope and the Heat Test: Looking Inside the Brick

To understand why the GGBFS mix was so strong, the researchers looked at it under a microscope and subjected it to heat.

• The Microscope (SEM & XRD): When they looked at the 100% GGBFS mix, they saw a dense, tightly packed city of gel. It was full of a special glue called C-(A)-S-H gel, which forms quickly because of the high calcium in the slag. There were very few holes (pores) and the particles were glued together tightly. In contrast, the weaker mixes had more gaps and less of this strong glue.
• The Heat Test (TGA-DTG): They heated the samples up to 1000°C to see how they held up. The GGBFS mix was a tough cookie. It lost very little weight as it got hotter, proving that its internal structure was stable and didn't fall apart easily. This suggests that these bricks wouldn't just be strong; they would be fire-resistant and durable over time.

The Structural Test: Saving a Crumbling Beam

Finally, the researchers asked: "Can this stuff actually save a building?" They used a powerful computer simulation called ABAQUS to model a concrete beam that was already cracked and weak. They then wrapped this beam in a "jacket" made of their new geopolymer mortar and reinforced it with layers of hexagonal wire mesh (like chicken wire).

They tested jackets with 1, 2, and 3 layers of mesh.

• The original, weak beam could only hold 48 kN of force before failing.
• The beam with the 3-layer geopolymer jacket could hold 74 kN.
• That's a 54.2% increase in strength!

The simulation showed that the geopolymer jacket acted like a superhero's shield. It stopped cracks from spreading, held the beam together, and made it much stiffer. The more layers of mesh they added, the better the beam performed, with the 3-layer version showing the least amount of damage and the least amount of bending.

The Verdict

This research paints a very promising picture for the future of construction. It proves that we can take industrial waste and turn it into a high-performance building material that is stronger than many traditional cements, all without needing a furnace. The 100% GGBFS mix is the strongest, but the 50-50 Fly Ash/GGBFS mix is a fantastic, balanced alternative.

The study also showed that we don't have to guess the recipe anymore; smart computer models (especially the ANN) can predict the results with high precision, saving time and money. And perhaps most importantly, this new material isn't just a lab curiosity; it works as a powerful tool to repair and strengthen old, damaged buildings, making our infrastructure safer and more sustainable. While the researchers note that more testing is needed to see how these materials handle extreme weather over many years, the evidence so far suggests that geopolymer mortar is a giant leap forward for green construction.