Skip to main content

Fracture Science on the Micro Part 2: Cement plant application

World Cement,

Read part one of Fracture Science on the Micro here.

Researching on an atomic scale

MIT researchers revisited Griffith’s work at the atomic scale and chose energy density instead of surface energy in Griffith’s equation. They realised that atoms on a material’s surface are less tightly bound than the atoms inside the material. Earlier research using the concept of surface energy completely disregarded what happened deep inside the material as it was fractured.

The choice of energy density as opposed to surface energy has significant consequences. We know that it takes relatively little energy to crack pavement or steel or concrete, but to fracture the material takes a great deal more energy. Consider a sheet of glass: you can take a nail punch to the driver’s side of your car window, gently push, and create dozens of cracks yet the glass hasn’t fractured. A block of ice is another good example: you can take an ice pick and push it against the surface until you generate all sorts of cracks but unless you start digging with the ice pick, you’ll never get the individual pieces to separate. Those examples provide a rough analogy between energy density and surface energy. Provided that the centre of the crack is isolated from the sides, scratching or scoring a material will create a plastic deformation along the sides of the scratch and a brittle fracture in the centre.

MIT researchers then used atomistic simulations to determine the cohesive energy density for each of the four major clinker phases and microscratched both alite and belite phase crystals. What they discovered was that the belite phase crystals had a fracture resistance 2.5 times that of alite phase crystals. Additionally, their experimental results validated a predictive model that can now be scaled. Most importantly, they developed a quantitative model that is repeatable and reproducible – important attributes when it comes to practical use.

Cement plant application

At first glance it is easy to miss how this research fits into the larger picture of crushing and grinding at the plant level. Crushing takes advantage of internal pores and cracks to make them much larger. Depending on the material characteristics, those cracks can become self-propagating: we do not need to input any more energy. From that point, grinding is about finding the weaknesses around the different crystals and within the crystal structure itself. MIT’s research has focused on the weakness of the crystals. That is the new endpoint. But grinding is much more than the grindability of individual phases; grinding is also about separating those phases from one another. Using our existing knowledge of crushing and separation, we can finally start to understand how this new research can be used.

A key result of the MIT research is that the industry now has the ability to evaluate characteristics of individual clinker phases and multi-crystal grains (both of which vary from plant to plant based on raw materials and processing differences) to estimate the fracture toughness and determine the theoretical grinding energy required.

Those determinations could be used to evaluate grinding operations plant-by-plant or even across the industry. MIT’s model is a revolutionary tool that allows plants to compare their grinding energy usage against theoretical requirements, as well as in comparison to others. Plants might use this new research to test the grindabilities of different clinker formulations.

Plants implementing this test method could evaluate their clinkers on an ongoing basis and use the test results to develop a robust database. That database would also include fuel usage, clinkering energy, emissions data, electrical usage and a variety of other metrics. All of this data could then be used to find the product formulations that best optimise quality requirements, energy costs and market demands. In effect, cement production now becomes a problem in mathematical optimisation.

Equipment suppliers tend to provide production guarantees based on tonnage and product fineness. MIT’s research might now shift us away from that model and move us towards production guarantees based upon a plant’s clinker formulation for each one of its products. We may ultimately see energy usage guides similar to those available for electrical appliances based upon some measure that normalises tonnage, energy usage and fracture toughness.

In the meantime, the research has more immediate impacts. The MIT discoveries in grinding point back to pyroprocessing as a key factor in addressing grindability. For example, since grinding is inherently inefficient, it is clear that we’ll need to refocus on clinker phase crystal size and morphology. We will also need to keep the hardness of belite crystals in mind as we develop new cements. This may lead to doping clinkers in an effort to increase their grindability.

Written by Rick Bohan, Portland Cement Association (PCA). This is an abridged version of the full article, which appeared in the December 2013 issue of World Cement. Subscribers can view the full article by logging in.

Read the article online at:


Embed article link: (copy the HTML code below):