During mass flow, material moves faster in the central region of the flow stream, due to friction along the walls and the geometry of convergence that resolves the effect of gravity according to the inclination of the flow path of an element from the vertical. The velocity gradient across the flow stream is greater at high flow rates, with large-sized outlets, and in containers where the wall slope is barely adequate for developing a mass flow form. Nevertheless, mass flow avoids the indefinite period of storage that occurs in the no-flow regions of a bin with a core flow pattern. The re-arrangement of particulate structure that takes place every time the material discharges in mass flow manner also deters the progressive formation of strong particle-to-particle forces, caking mechanisms and other fusion, sintering and chemical bonding processes that progress with time and sustained contact pressure.

Mass flow construction is recommended if there is a possibility that fluidised products in the upper part of the container might flush through to the outlet and cause a sudden high discharge flow (flushing). Freshly-loaded material will have time to settle before reaching the outlet if the discharge is controlled so as to maintain a 'heel' of material which has had time to settle (de-aerate) to provide a stable flow pattern due to the 'first-in, first-out' behavior of mass flow. However, even a container designed for mass flow will not function as such if the bulk material can slip on itself more easily than on the hopper walls. In that case even when a 'heel' of settled bed is retained to prevent the flushing of freshly-loaded material there remains a danger of a preferential flow path forming that allows the loose material to flood out of the discharge port. This is because there is always a higher flow velocity in the central region of a converging flow channel.

Material in a fluid condition exhibits hydrostatic pressure that invariably exceeds the lateral pressure of more settled material. This serves to amplify the radial velocity gradient and reinforce the development of a preferential central flow, as the more mobile product replaces stable flow material replenishing the flow channel. Unless the new material settles to a stable flow condition before reaching the outlet it will form a penetrative path for fluidised material to flush through under the hydrostatic head of the container contents.

Mass flow also tends to redress segregation occurring as the container fills by re-mixing the cross sections of contents during discharge. This again is only partially effective, as the differential flow velocity across the converging section accumulates the outer top surface regions of the contents as the final portion to discharge.

The criteria for adopting or incorporating a mass flow design sometimes tend to be overstated by 'experts' and ignored by industry. The various advantages and drawbacks of both mass flow and core flow are set out in Table 3. Some factors may have an over-riding influence, but the ultimate choice has to take full account of both practical and economic considerations.

In order to avoid confusion with the coherent movement of a body of material in a parallel flow channel, the definition of mass flow is proposed to apply only to flow in a converging channel with slip on the boundary walls. The term mass flow hopper, applies to a complete storage container, in which material slips on all wall contact surfaces during flow, whether some of the internal movement is coherent or has differing velocities in different regions. It should be emphasised that a 'mass flow hopper' is only such with respect to a specific combination of container construction and geometry with a bulk material having flow properties that meet the mass flow design criteria for the storage unit. The important features to be satisfied are that

• the material slips on the wall contact surface
• the outlet allows flow to take place over the total area
• the outlet is sufficiently large to avoid the material forming a stable arch

In general a 'Vee' shape of converging channel is more effective for mass flow than one of conical or pyramid shape. The slope of wall needed to ensure slip is mainly determined by the wall friction (Fig. 10).

The size of opening needed to avoid blockages from lumps coming together is relatively easy to assess and allow for, although this process is stochastic and the critical opening size depends upon many factors of the container geometry and flow circumstances. A commonly cited 'rule of thumb' is that a circular outlet has to be more than eight times the size of the lumps. For a slot shaped outlet the width should be greater than five times the lump size, provided that the slot length is at least three times the slot width and flow takes place over the whole cross-sectional area. Less guidance is available when the lump sizes are variable or indeterminate, but in important cases of uncertainty other steps -- such as grids or flow inserts -- should be incorporated to offer protection against lump blocking.

More difficult to determine is the size of orifice necessary to avoid the formation of cohesive arches. The only internationally recognised methodology is the Jenike shear cell procedure. This technique currently remains within the domain of specialists. The redeeming feature of the situation is that means are available to stimulate flow through a smaller outlet than the 'critical orifice size' dictated by design. It is difficult to correct for inadequate wall inclination, yet wall friction measurements -- the key parameter required for an adequate design -- are relatively easy to secure.

In practice the flow of particulate material is rarely smooth and uniform. Detailed studies have shown that gravity flow takes place in a changing state of dilatation as regions of reduced pressure work through the system. Variable velocity gradients and discontinuities, together with small regions of coherent movement, occur at differing times and with differing boundaries, constantly changing during the flow process. While plasticity theory has been successfully applied to hopper design, planes of weakness, propagation of cracks and detailed forms of unstable complexity abound. Nevertheless, gross flow patterns emerge having relative simplicity, invariably leading to predictable results because of the conservative steps incorporated within the theory.