Geldart Group Research Articles

Cold modelling studies of fluidised bed reactors

Experiments performed on pairs of dynamically similar gas-fluidised beds confirm the validity of scaling relationships, based on purely fluid-dynamic interactions, to bubbling regime behaviour over a wider range of conditions than hitherto reported: ranging from fine powder (Geldart group A), to large particle (Geldart group D) systems. The scaling relationships, however, fail to explain observed slugging regime characteristics, thereby supporting the view that slugging involves a degree of defluidisation, and hence direct frictional interaction between particles.

Fluidization behavior of polymeric particles in gas—solid fluidized beds

A systematic study is conducted which illustrates the effects of the average size and size distribution of polymeric particles on the fluidization quality of gas—solid fluidized beds. The fluidization quality is characterized by the bubble flow rate, and evaluated based on a bed collapse technique. Polymeric particles used include polyethylene (PE) and polypropylene (PP) resins with a density of 0.66 and 0.57 g/cm 3, respectively, and sizes ranging from 116–775 μm. The results indicate that there is virtually no effect of the particle size and its distribution on the fluidization quality for PE and PP particles of Geldart group B, provided that the fluidization quality is characterized by the bubble flow rate. These results are consistent with Geldart's finding (Geldart, 1972). For PE particles belonging to Geldart group A, it is found that the particle size has a slight effect on the fluidization quality. However, the size of group A PP particles and the size distribution of group A PE particles have essentially no effect on the fluidization quality. Experiments involving mixing fine polyethylene powders (microthene), having an average size of 32 μm (Geldart group C), with PE particles are also conducted to investigate the fine particle—coarse particle interactions, such as the van der Waals and electrostatic forces, in the fluidized bed. The improvement of the fluidization quality due to the addition of fine powders in a bed of polymeric particles is discussed in the light of the observation of surface structures from an SEM study of the mixture sampled from the fluidized bed.

FBSim: A model for fluidized bed simulation—I. Dynamic modeling of an adiabatic reacting system of small gas fluidized particles

In the present study a dynamic model for fluidized bed simulation (FBSim) is developed. The FBSim model can be applied to adiabatic gas fluidized bed reactors containing Geldart's group A or B particles. The classical two-phase theory is slightly modified for the evaluation of the fluidization parameters of group A particles. Reacting systems with gases and solids undergoing multiple homogeneous and heterogeneous reactions can be accommodated. The FBSim model accounts for the superficial velocity and density changes of gases due to reaction stoichiometry and adiabatic operation of the reactor. Freeboard calculations are also included. The FBSim model equations have been incorporated in a user-friendly computer program for dynamic simulation of gas fluidized bed reactors. A technique is developed to obtain consistent initial values for the iteration variables of the model equations. The application to an industrial FCC regenerator is presented in Part II [Faltsi-Saravelou et al., Computers chem. Engng 15, 647, (1991)].

Influence of pressure on the minimum fluidization velocity

The influence of pressure on minimum fluidization velocity has been measured in a pressurized fluidized bed. Three different bed materials, corresponding to Geldart's group B or group D particles, were investigated at pressures between 0.1 and 1.6 MPa. The onset of fluidization was determined both visually and by pressure drop measurements. The results show a clear decrease in the minimum fluidization velocity with increasing pressure. The experimental results are in good accord with results obtained using the Ergun equation. A number of simplified correlations derived from the Ergun equation were also investigated. While many of these give a good description of the pressure effect, their accuracy varies significantly for different bed materials.

The bubble frequency in a fluidized bed at elevated pressure

A piezoelectric probe has been used to determine bubble frequencies on the axis of a bed of 135 mm diameter, in which particles of sand and fire-clay covering Geldart's Groups B and D were fluidized at pressures up to 1. 3 MPa. Bubble frequency is proportional to the gas flow in excess of that at minimum fluidization under ambient as well as elevated pressure. The ratio of bubble frequency to excess gas velocity above minimum fluidization is 1.2 ± 0.13 m −1 at atmospheric pressure, but increases to 1.9 ± 0.17 m −1 at elevated pressures ranging from 0.4 to 1.3 MPa.

Hydrodynamic behaviour of aerogel powders in high-velocity fluidized beds

The behaviour of silica aeroget powder and iron oxide—silica aerogel powder in fluidized beds has been observed at gas velocities between 0.1 and 1.6 m/s to determine whether aerogels can be treated in turbulent and circulating fluidized bed (CFB) contactors. Pure silica aerogel particles agglomerate at these velocities to create larger particles whose equivalent diameter is a strong function of the gas velocity at which they are formed. Although they retain the high specific surface area of the original powder, the agglomerates have a low permeability. It was never possible to smoothly circulate pure silica aerogel in a CFB unit with an L-valve, because the fine material could not be returned to the base of the riser, and large agglomerates remained suspended in the riser. However, by adding a granular diluent, it was possible both to suppress much of the agglomeration and to form a mixture ideal for CFB contacting. Iron oxide—silica agglomerates were homogeneous in size and insensitive to the gas velocity. Tests in a small CFB unit with a V-valve showed that iron oxide—silica aerogel particles could be circulated smoothly and steadily, because fluidized aerogel particles in the V-valve leg reorganized themselves into agglomerates which behaved like Geldart's Group A or B powders.

On the minimum bubbling voidage and the Geldart classification for gas-fluidised beds

Reported experimental findings for the effect of systematic variation of all the relevant variables on the bubbling point of gas-fluidised beds are shown to be in agreement with the predictions of a general fluid dynamic stability criterion for fluidisation. The criterion is also applied to the evaluation of the stabilising effect of fines addition and, somewhat speculatively, to the prediction of the boundary between the Geldart Group A and Group C powders.

A generalized bubble diameter correlation for gas‐solid fluidized beds

AbstractA new bubble diameter correlation is derived to predict bubbling characteristics of fluidized beds of varieties of powders. The present model is founded on the postulate that the steady bubble size, which is often called the maximum stable diameter, observed in a bed of Geldart group A powder is formed as a result of an equilibrium of successive coalescence and splitting. For the cases of group B powders the present correlation automatically converges to the conventional correlation of Mori and Wen (1975), whose predictions are close to those of Rowe (1976) and Darton et al. (1977). For group A powders the present correlation is validated by comparison with experimental data in the literature. Based on this correlation a theoretical explanation is presented for the fact that the maximum bubble diameters observed were up to 50 to 100 times as large as those from the stable bubble theory of Harrison et al. (1961).

A new similarity rule for fluidized bed scale‐up

AbstractA rule of hydrodynamic similarity for a scale change of fluidized beds has been developed based on the governing equations of bubble and interstitial gas dynamics. When geometrically similar scale‐up is to be carried out maintaining hydrodynamic similarity, the proposed similarity rule requires that two conditions be satisfied. The first condition assures a similarity in bubble coalescence. The second assures the similarities in bubble splitting and in the interstitial flow pattern. The present work proves theoretically that these two conditions are the necessary, and almost sufficient, conditions for hydrodynamic similarity. They consider not only bubble coalescence but also bubble splitting. The theory proposed was first tested by previous correlations for bubble diameter and grid zone structure, and second by experiment. It has been proved that as long as the present rule is satisfied, the longitudinal distribution of the average bubble diameter, stochastic variation around it, and the radial distribution of superficial bubble velocity can be maintained similar for a scale change.Application of the present rule to predict the bubbling and solid circulation characteristics in a large‐scale unit by a small‐scale experiment has thus proved promising, at least for Geldart group B particles. In order to further study the possibility of a similarity in mass transfer and chemical reactions, computation was carried out using the three‐phase bubble assemblage model. If the effect of molecular diffusion is negligible, as in the case of fluidized bed combustion, even a chemical similarity is found to be possible.

Comparison of choking velocity correlations in vertical pneumatic conveying

A number of empirical equations are available for predicting choking gas velocity in vertical pneumatic conveying. Five such equations are evaluated against published results of different types of gas—solid system using Geldart's classification. It is found that no equation is suitable for all types of system. The equation of Yousfi and Gau is found to correlate all Geldart Group A and Group B choking results to within a root-mean-square relative deviation (RMSRD) of 10%. The equation is recommended for use for Group A and Group B systems. Yang's latest equations correlate Group D results with a RMSRD of 16% and are recommended for use for Group D systems. The equations of Leung et al. give reasonable agreement with Group D results (RMSRD = 20%) and their equations can handle mixed-size particles. Further work is required to develop a more fundamental basis for choking velocity equations and to provide more experimental results for mixed-size particles in tubes of large diameter.

Fluidization of powders showing degrees of cohesiveness—II. Experiments on rates of de-aeration

Experiments have been done on the de-aeration characteristics of 18 powders with mean sizes of 3–125 μm and densities 364–5000 kg m −3. Powders in Geldart's Group A de-aerate at a constant rate analogous to sedimentation in a liquid, and a standardised de-aeration time has been correlated with the powder and gas properties. Cohesive Group C powders de-aerate exponentially with time similar to the final solids consolidation stage observed in solid/liquid systems.

The effect of thin liquid layers on fluidisation characteristics

Experiments are reported which show that addition of small quantities of a non-volatile liquid to a material in Geldart's Group B causes its fluidisation behaviour to move through Group A to Group C. It is shown that this behaviour is qualitatively consistent with Molerus' analysis of the role of interparticle forces in determining fluidisation characteristics. Interparticle forces have been estimated by determining the failure point of a bed of solids suspended by upward gas flow. The results lead to a numerical adjustment of Molerus' criterion for the B/A transition, and give some insight into the physical mechanisms causing the difference in behaviour between these two Groups.

The measurement of emulsion phase voidage in gas fluidized beds of fine powders

The results of two different experimental methods, X-ray absorption and bed collapse, for the determination of the emulsion-phase voidage of gas fluidized beds are compared. It is shown that good agreement exists between them over the range of gas velocities studied, but that possible differences in voidage in an axial direction are only revealed by the X-ray technique. The bed collapse method applied to a Geldart Group A powder confirms that there is little difference between the surface settling rates of a bubbling bed and of a uniformly expanded bed of the same voidage.

The effect of pressure on fluidized bed behaviour

The influence of pressure on the behaviour of different particles fluidized by nitrogen was experimentally investigated. The powders used were mostly in Geldart's group B and the pressure was varied in the range 100 to 3500 kPa. The results for bed voidage and gas velocity at both the minimum fluidization and minimum bubbling conditions are compared with results of other authors and are discussed in the light of available correlations.

The effect of operating temperature on the velocity of minimum fluidization, bed voidage and general behaviour

Tests have been made of fluidized bed behaviour at operating temperatures up to 950 °C. Bed materials used were sand, ash and alumina falling within the categories of Geldart's Groups B and D. The only wide size distribution material used was the ash. The minimum fluidizing velocity was found to decrease with increase in the operating temperature for Group B materials because of the consequent increase in the fluidizing gas viscosity. However, the decrease was not as large as would have been expected because the voidage at minimum fluidization was also found to vary with temperature. No explanation for the cause of this can be offered. With Group D materials, flow is in the turbulent or transitional regime and gas density is the important factor. U mf then increases with increase in operating temperature. Given appropriate values for ϵ mf and φ, the Ergun equation predicts U mf very well. A transition between behaviour characteristic of Groups B and D material types has been observed when the operating Re mf value passes through ∼ 12.5 and Ar ∼ 26000.