Computer simulations particle packing

The objective of AxelTs [11] experimental study is twofold (1) to develop methods to study the combustion process of a packed-bed of biomass (2) to study the effect of mass flow of air on the combustion process in different conditions with respect to fuel particle size, density, and shape. The results are planned to be applied to computer simulations of packed-bed combustion of wood fuels as well as design data for construction of PBC systems. 

However, by examining the adsorption behavior of polypeptides and proteins with comparable porous and nonporous particles in finite baths, packed columns and expanded or fluidized beds, an iterative simulation approach based on the heuristic principles described earlier and along the lines of the flow diagram shown in Fig. 32 can be developed, leading ultimately to the implementation of useful scale-up criteria. Along the way, computer simulations, generated from the analysis of the concentration-time... 

For reliable application of the free volume concept of disperse systems one must have dependable methods of determination of the maximum packing fraction of the filler tpmax. Unfortunately, the possibility of a reliable theoretical calculation of its value, even for narrow filler fractions, seems to be problematic since there are practically no methods available for calculations for filler particles of arbitrary shape. The most reliable data are those obtained by computer simulation of the maximum packing fraction for spherical particles which give the value associated with possible particle aggregation, so that they are probable for fractions of small particle size. Deviations of particle shape is nearly cubic. At present the most reliable method of determination of [Pg.142]

Recently, Miller and Cacciuto explored the self-assembly of spherical amphiphilic particles using molecular dynamics simulations [46]. They found that, as well as spherical micellar-type structures and wormlike strings, also bilayers and faceted polyhedra were possible as supracolloidal structures. Whitelam and Bon [47] used computer simulations to investigate the self-assembly of Janus-like peanut-shaped nanoparticles and found phases of clusters, bilayers, and non-spherical and spherical micelles, in accordance with a packing parameter that is used conventionally and in analogy to predict the assembled structures for molecular surfactants. They also found faceted polyhedra, a structure not predicted by the packing parameter (see Fig. 8). In both studies, faceted polyhedra and bilayers coexist, a phenomenon that is still unexplained. 

Both laboratory experiments and computer simulations have shown that the random close packing of equal-sized, spherical-shaped particles occupies a volume percentage of 64%. It follows that, if we want to eliminate porosity in such a system, then we must fill the remaining 36% of space (the dead volume ) with an inert binder. By contrast, if we want to retain some porosity, then a certain fraction of the dead volume must be kept free of binder. On the basis of a 50/50 compromise between electrode cohesion and solution access, one may therefore reasonably conjecture that the volume percentage of inert binder should be 18%, and indeed this is a good starting point in the laboratory development of inks from equalsized, spherical-shaped particles. 

The theory summarized in this chapter and elsewhere [l-3a,18] has enabled the development of computer programs that allow so-called computer simulation. In this approach, two to four experimental gradient runs are first carried out for the reversed-phase or ion-exchange separation of a biological macromolecular sample. After the results of these initial experimental runs are entered into the computer, separation times of analytes can be predicted based on initial and final %B, gradient time and shape, column dimensions, flow rate, column-packing particle size, and temperature [42-47]. 

The simulation results are consistent with what mi t reasonably be expected, namely that the less stiff the particles axe, the easier the packed bed is squashed. However, it is noticed that there is 10 fold difference between the values of KN =8E3 and 8E2, while the difference berween 8E2 and 4E2 is only double. There is not much difference between the voidage -bed depth curves of KN =8E3 and KN =8E2, although there is a significant difference between the ones of KN -8E2 and KN =4E2. The sharp change in behaviour with a relatively small variation in KN suggests that it may be convenient to categorise the particles into "sensitive and insensitive groups with respect to their behaviour under compression. The value can be found by computer simulation. 

A collection of hard, identical spheres is the simplest possible model system that undergoes a first order phase transition. For low packing fractions the particles are in a liquid state, but when the packing fractions exceeds a value of 49.4% a ordered solid state becomes more stable. This was first shown in computer simulations by Hoover and Ree [27] in 1968. The experimental realization of a colloidal suspension that closely mimics the phase behavior of hard spheres followed about 20 years later and was a milestone in soft matter physics [28, 29]. More recently the phase transition kinetics of hard sphere colloids has been studied extensively in experiments [5, 30, 31]. However as mentioned in the introduction the interpretation of the data with CNT was rather indirect. 

---