Sintering In Fluidized Bed Essay, Research Paper

Processes for cleaner and more efficient energy generation from feed stocks such as coals, lignites, peats, and waste liquors use fluidized beds that are operated at high temperatures and pressures. These processes involve systems that are multi-phase and have complex chemical reactions. Research work has tackled a number of aspects, including mechanical engineering aspects of the reactors, reaction chemistry and products, characterization and physical properties of the ash, fouling by ash deposits and the phenomenon of defluidization by agglomeration or sintering of the ash particles. It is with this latter aspect, the phenomenon of defluidization, that this contribution is concerned. Defluidization is also a problem in a number of other elevated temperature fluidized bed production processes, including size enlargement by agglomeration, fluidized bed processes for poly-olefin production and metallurgical processes.

Defluidization occurs when the particles in the bed adhere. When two particles touch, material at the point of contact migrates forming a neck that is strong enough to withstand the disruptive forces in the fluidized bed. Two categories of adhesion can be discerned. The first type is visco-plastic sintering and it occurs with glassy materials. With these materials, migration is limited by the ability of the material to flow. With increasing temperature the viscosity of the material is reduced and hence the material flows and the size of the neck is increased. At some point, the necks are sufficiently large and strong enough to cause defluidization. The second type occurs when a large quantity of liquid is formed by melting or by chemical reaction. With an increase in temperature, a phase boundary may be crossed bring about rapid defluidization. The liquid formed has a relatively low viscosity and defluidization is caused by the amount of liquid present. This section will primarily focus on visco-plastic sintering.

Model systems, in which chemical reactions do not occur, have been used to investigate the relationship between process conditions and the tendency to defluidize due to visco-plastic sintering. Materials used have included low-density ploy-ethylene and poly-propylene, soda glasses, metals, and inorganic crystalline salts. Under sintering conditions, the fluidizing velocity has to be increased above the minimum fluidizing velocity, Umf, to a higher velocity, U; at which pronounced bubbling occurs. A fluidizing velocity, Umfs, may be defined as a velocity just sufficient to prevent sintering. This characterizes a different fluidization state from Umf, since the in the latter the bed is in an expanded but not bubbling state. Several methods of defluidization behavior have been presented. The simplest is an empirical approach which relates the excess gas fluidizing velocity needed to maintain fluidization to the difference between the bed temperature, T, and the sintering temperature, Ts. The sintering temperature is the observed temperature at which defluidization occurs at the minimum fluidization velocity.

In the case of single-phase materials, sintering can potentially occur by a number of mechanisms: flow of the material, surface diffusion, volume diffusion, and vaporization/condensation. In fluidized beds, the first of these mechanisms is most important because it is the most rapid. Sintering by diffusion (surface and volume) requires the movement of atoms of the material and is relatively slow. Because of the flow of gas, vaporization/condensation of the material is probably the least significant mechanism. Glassy materials flow by viscous of plastic deformation. In the case of crystalline materials, such as metals, visco-plastic flow occurs by slip at dislocations.

The sintering of two particles has been modeled by representing the particles as smooth spheres. For two such particles in contact, the rate of neck growth is represented by an equation of the form:

(x/r)n = kt (1)

(x/r)2 = 3gt/2rm (2)

Rumpf [4] derived an expression which included the sintering effect of an applied load, F, and in which the proportion constant for the first term was found to be 8/5 for hydrodynamic load instead of the 3/2 found by Frenkel:

(x/r)2 = [(8g/5r)+(2F/5pr2)]t/m (3)

Assuming the stress to be evenly distributed over all particle contacts, the above equation may be rewritten using Rumpf’s equation, in terms of stress, s1, applied to an assembly of spherical particles of porosity e:

(x/r)2 = [(8g/5r)+(8es1/5p(1-e))]t/m (4)

Both Equations (3) and (4) can be used in practical situations in which the applied load is significant. They also can be used to analyze data to obtain an apparent value of the viscosity and its variation with temperature. In this case the importance of the first term may be negligible compared to that of the second term.

In real systems, however, the particles are neither perfectly smooth spheres nor are they monodisperse. Taking this into consideration, Equation (4) is probably best written in the form:

(x/r)2 = k1t/m (5)

Where k1 is a factor dependant on both material properties and environmental conditions. Furthermore, in glassy systems the flow is unlikely to be Newtonian in essence. Hence, in practice, the viscosity value must be regarded as a value used to fit the data. An Arrhenius relationship of viscosity and temperature is often taken as:

m = moexp(E/RT) (6)

The magnitude of the activation energy, E, depends on the activation energy for flow and the dependence of the numbers of flowing molecules on temperature, resulting in moving through a material transition temperature. The activation energy may therefore change with temperature and different gradients may be shown in an Arrenius plot.

Gluckman [8] performed deluidization experiments using a plethora of materials, including copper shot, poly-ethylene beads, poly-propylene particles, poly-ethylene terephthalate particles, and glass spheres. It was found that for a given material a certain gas temperature existed which, when underneath, the bed could always be fluidized at the minimum fluidization velocity, Umf,. This temperature was called the minimum sintering temperature, Ts. Above this temperature, a higher fluidizing velocity, Umfs, was needed to fluidize the bed. The data that was collected correlated to a linear function between velocity and temperature:

Umfs = B + AT (7)

where A and B are constants. Also, it was concluded that there is most likely an influence of bed height.

Siegell [2] conducted defluidization tests on several types of particles and based his results in terms of the excess air velocity (Umfs – Umf) needed to keep the bed fluidized and the excess temperature, (T – Ts). It was found that there was a linear relationship between these two parameters and Siegell also reported that increasing the bed height led to the tendency to defluidize.

The conclusion in the study of sintering in fluidized beds is that the effect of temperature is the dominant factor that can result in defluidization of the system. This correlation is quite different in respect to effect of temperature on dynamic liquid bridge forces, where higher temperatures and reduced viscosity lead to weaker forces.

References:

1. M.J. Gluckman, J. Yerushalmi, A.M. Squires, in: D.L. Keairns (Ed.), Fluidization Technology, Hemisphere, Washington, 1976, p. 395

2. J. H. Siegell, Powder Technology 38 (1984) 13

3. W.D. Kingery, H. K. Bowen, D.R. Uhlmann, in: Introduction to Ceramics, second edition, Wiley-Interscience, New York, 1976, p. 492

4. H. Rumpf, in: K.V.S Sastry (Ed.), Agglomeration, American Institute of Mining, Metallurgical, and Petroleum Engineering, 1977, p. 97

Bibliography

References:

1. M.J. Gluckman, J. Yerushalmi, A.M. Squires, in: D.L. Keairns (Ed.), Fluidization Technology, Hemisphere, Washington, 1976, p. 395

2. J. H. Siegell, Powder Technology 38 (1984) 13

3. W.D. Kingery, H. K. Bowen, D.R. Uhlmann, in: Introduction to Ceramics, second edition, Wiley-Interscience, New York, 1976, p. 492

4. H. Rumpf, in: K.V.S Sastry (Ed.), Agglomeration, American Institute of Mining, Metallurgical, and Petroleum Engineering, 1977, p. 97