The sales brochures are readily available. The technical papers are not. What the computer software does is readily obvious. What can yet be done to benefit the mining industry is not.

Is this a task for scientific research and development? Or is this a task for the mining philosopher? Maybe this is where we need a Value Engineering Workshop, replete with all those sessions on Function Analysis, Brainstorming, Evaluation, and Ranking.

The basic idea is simple: in the real world particles interact with each other and hence the whole mass of particles behaves in a certain way. I watched my five grandsons playing good old-fashion marbles this weekend. Do you recall the thrill as one large marble hit another and yet another and the whole mass of marbles rolled and tumbled around in unexpected and unanticipated ways. Another example of interacting particles: I took the kids to the local eatery in Keystone, and between repeated visits to the help-yourself-to-as-much-as-you-can-eat, they played billiards on a faded cloth. They took the same delight in watching as one ball, inexpertly hit with a long stick, rolled across the table to send yet other careering around and maybe into a pocket.

Soon we will watch as the harvesters invade the corn fields on this Iowa farm and see the harvesters cut a broad swath through the high, yellow plants and the see the corn come streaming out the spout and into yet another container. The individual corn kernels flow and tumble like a golden stream. And hence, towed by a great green tractor to the silo, to a barge, to a ship, and to another country. Some of the corn, I am told, is passed to the long metal buildings that dot the countryside around the farm. There is no indication of what is inside these long, low metal structures; except in the hot days of early summer, you can smell the hogs. Iowa is the nation’s largest pork producer—and the production “factories” are all around us, supplied by copious corn.

Each year, the newspapers carry horror stories of men who fall into the corn atop a silo and drown in the sea of small particles that respond to the human body as water to a stone. Down goes the victim and he drowns—or is suffocate a more accurate description? Regardless, it is all just a matter of scale and the aggregate behavior of individual particles, each bumping into the other, each supporting or redirecting the other. Mathematically it is just a huge number of simple equations based on equilibrium and conservation of momentum. Most scientists and engineers learn the basic equations in the first semester of applied maths. In those early days, we struggled through complex problems set to test our intellect and teach us mechanical thinking and analysis. We could not, however, analyze complex problems with the paper and slide rules we then had.

Time passed and we bought our first hand-held calculator. We laboriously learnt to program the calculator to do repetitive calculations. My first task was to solve the stresses imposed on a caisson by the footings of a large headgear (head frame as they are called in North America). This involved the repetitive summation of individual stresses from each footing divided into discrete point loads to better replicate the spatial distribution of differential loads.

Then the first real computer arrived. Small by today’s standards, but infinitely more powerful than a slide rule. Suddenly a whole new world of problems was tackled by the repetitive power of the computer. Finite difference and finite element methods became the preferred way to calculate the stresses in a structural slab. Suddenly you could get a reasonable estimate of the distribution of load from backfill on a retaining wall; those highly mathematical integrations from Russian papers were rendered obsolete.

Somewhere along the way I read Harr on particulate mechanics in soil mechanics. He set himself the task of reformulating all soil mechanics theories in terms of particulate mechanics. His insight was simple: soils are just aggregates of particles, and each particle acts on adjacent particles; if you could write the equations of equilibrium of each and every one of the interacting collection of soil particles, maybe you could describe the state and predict the behavior of the soil mass. He nearly succeeded. Certainly he impressed me.

I lost sight of developments in the field for many years as my career took me into uranium mill tailings pile closure, design for a 1,000 years of integrity, hazardous waste landfills and their stability when placed at slopes exceeding 1.5 horizontal to 1.0 vertical, and finally the impact of the 1994 Northridge earthquake on domestic structures.

So it was with some delight that I got an e-mail from a fellow in London, England telling me of a computer code that can do all I have just written about and a lot more. He put me in touch with John Favier and we talked courtesy of the magic of cell phones about his computer code EDEM. This code simulates the behavior of a collection of particles of variable size and shape and their interaction with boundaries that inhibit and control their movement. I quote from a brochure:

“EDEM is an advanced Discrete Element Method (DEM) simulation and analysis software tool for particle-scale modeling of particulate solids handling and process operations.” EDEM can be coupled with finite element and other rigid-body dynamics software to simulate particle-structure and particle-fluid interactions.

The issue I seek to explore in this piece is mining-related uses of this code, alone or when coupled with other software to simulate particle to structure or particle and fluid interactions.

John sent me technical papers on the use of the code to examine the behavior of rock in mills. Most of the papers are copyrighted and thus I cannot give and e-reference. I am sure there are more uses in mining of the code than are written about in the unreferenced papers, for if you think about it, by one definition, mining is but the preparation and management of a series of collections of individual particles with the object of separating the more valuable from the less valuable. The particulate mining process starts with the blasting and breaking; the intact rock is drilled and blasted to reduce the intact mass to a collection of manageable sized particles. These particles are collected, excavated, moved, and one way or another relocated to the mill. Such relocation may involve placing in trucks, trains, pipes, or on conveyors and moving them from one point to another. Once the individual particles get to the mill, they may be further stored, crushed, ground, stored, placed, and treated chemically and/or physically to separate and/or remove valuable constituents. Hence the resulting non-valuable particles are relocated to the waste disposal facilities, by pipe [pumped] or via trucks to tailings impoundments or rock dumps or similar. And with time these waste disposal facilities may be subject to the forces of water that often result in erosion.

At each of these stages, I submit, DEM may be useful in modeling the process and seeking new and better ways of doing things. Let us explore a few of these. Starting with blasting. There are many codes to assist in blast design, but maybe DEM can be programmed to replicate the blasting process: after all the essence is application of concentrated force at one or more place to induce particle size reduction in what is generally a mass of larger particles. The issue is the optimum layout of the blast holes—the locus of initial force application.

Then we have the excavation process. Force is applied to a pile of individual particles to push them and pick them up. John tells me that DEM has already been tailored to the excavation process. May we use DEMs to replicate specific mine applications associated with specific equipment and hence to improve processes: reduce load time, increase cycles, induce some additional particle size reduction, and place the particles most efficiently in the receiving receptacles.

Could we reduce the size of the mine mass particles to the extent that it is cost-effective to pump via air or even water the resulting mass to the mill. Maybe DEMs can be used to study and evaluate options.

Once the particles reach the mill, there are a plethora of problems that have and yet can be modeled by DEMs. I refer to papers John sent me looking at all the comminution and treatment processes involved in the typical mill. I am sure not all the problems have been solved.

I referred earlier in this piece to Harr’s efforts to model standard geotechnical problems via particulate mechanics processes. Maybe it has been done, but I suspect that the behavior of tailings in deposition and beach formation, waste rock dumps in their falling and development, and covers in their response to the deformation and consolidation of the underlying wastes, could well be analyzed by DEMs.

These are research opportunities to be evaluated by others with more skill and time than I have. I offer them here as ideas. Let me know if you agree, have better ideas, or succeed in solving some of these problems.

To further stimulate your thinking I record that DEMs is being used to study the behavior of particles in the absence of gravity. Apparently flakes of skin, hair, fibers, and other small particles float around at will in the space vehicle and aggregate in ways unknown in the presence of gravity. What better way to model the behavior of these pesky particles in a spaceship than with DEMs. Once you get to the moon, the landing craft has to roll up and down dunes of fine particles that are not subject to the same gravity encountered on earth. The craft’s tires interact in strange and different ways with this “light” soil. What better way to study the process than by modeling with DEMs?

Maybe we can undertake the same process to study light craft on tailings impoundment surfaces? I recall a crude study many years ago where we undertook to study access to the top of the tailings piles that would result from oil shale mining in Colorado. Falling oil prices put and end to the study, but I see the study of oil shale mining is again booming. What of the oil shales in Canada? There are computer codes to model the development of topographic landscapes after reclamation of the vast areas disturbed by oil shale mining. But maybe DEMs can be used to study the interaction of surface flows and the meander of rivers, the need for more gravel at the base of new creeks to control erosion, and the development of differential slopes down the erosion profile. Could DEMs be the basis for setting new standards for mine reclamation that account for natural geomorphic processes? Can we move away from a set of closure guidelines that demand 1,000 geomorphic stability and instead move to a more dynamic equilibrium approach?

But who is to do all this research? The individual mine is unlikely to have the resources to undertake such studies. I am not aware of any consultants who are poised to undertake the work. I see from the papers sent me by John Favier that academics in South Africa at the universities of Cape Town and the Witwatersrand are using DEMs to study process-related problems. Can we look to them to expand their work to include the world-wide mining scene? I doubt it. Maybe this is an opportunity for North American universities seeking new research avenues. But where is the money to come from? I cannot in this piece go into the questions of academic research funding. The University of British Columbia is organizing a conference in 2007 to better explore these issues and we will defer further deliberation on this topic until then. I simply proffer these questions now to get discussion going.

As I said earlier, if you have comment, idea, suggestion, or case history on these questions, please contact me at