biosystems engineering 101 (2008) 425435

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/issn/15375110

Research Paper: PHdPostharvest Technology

Flowability and handling characteristics of bulk solids and powders a review with implications for DDGS5
V. Ganesana, K.A. Rosentraterb,*, K. Muthukumarappana
a b

Department of Agricultural and Biosystems Engineering, South Dakota State University, Brookings, SD 57007, USA USDA, ARS, North Central Agricultural Research Laboratory, Brookings, SD 57006, USA

article info
Much research regarding handling and storage characteristics of bulk solids has been Article history: Received 6 June 2007 Received in revised form 16 July 2008 Accepted 16 September 2008 Available online 7 November 2008 conducted over the years. Physical properties of granular solids play a signicant role in their resulting storage and ow behaviour, and are therefore essential to design appropriate, efcient, and economic bulk solids handling and storage equipment and structures. Distillers Dried Grains with Solubles (DDGS) is a bulk material that has been widely used as a protein source for ruminants and non-ruminants for more than two decades. Distillers grains are energy and nutrient dense, and are often used as a replacement for corn in animal diets. With the exponential growth of the fuel ethanol industry in the last few years, large quantities of distillers grains are now being produced. To effectively utilize these feeds in domestic and international markets, however, these co-product streams are increasingly being transported greater distances, and must be stored until nal use. DDGS ow is problematic as it often becomes restricted by caking and bridging which occurs during transportation and storage. This issue probably results from a number of factors, including storage moisture, temperature, relative humidity, particle size, time, or temperature variations, to name a few. The objective of this study was to review the primary factors affecting owability, handling, and storage of granular solids and powders, as well as appropriate testing methodologies for these materials. Considering these will be helpful when examining granular owability and storage challenges for byproduct feeds, including those surrounding the use of DDGS. Published by Elsevier Ltd on behalf of IAgrE.

1.

Introduction

Maize is the U.S.A.s number one eld crop; maize production measures more than two times that of any other crop. In 2006, U.S. maize production was 10.5 billion bushels (NCGA, 2007). The two most important manufactured maize products are corn sweeteners (i.e., high fructose corn syrup) and fuel

ethanol. As fossil fuel reserves diminish and increase in price, there is growing interest in utilizing biomass and other renewable energy sources, including fuel ethanol. Ethanol from maize is a major fuel additive, and many ethanol plants have been constructed in the last decade in the Midwestern parts of United States. The production of ethanol from maize results in three main co-product streams: Distillers Wet

Mention of trade name, propriety product or specic equipment does not constitute a guarantee or warranty by the United States Department of Agriculture and does not imply approval of a product to the exclusion of others that may be suitable. * Corresponding author. E-mail address: [email protected] (K.A. Rosentrater). 1537-5110/$ see front matter Published by Elsevier Ltd on behalf of IAgrE. doi:10.1016/j.biosystemseng.2008.09.008

426

biosystems engineering 101 (2008) 425435

Grains (DWG), Condensed Distillers Solubles (CDS) which is also known as syrup, and the combined product Distillers Dried Grains with Solubles (DDGS) (which is the most common). These are produced from the non-fermentable maize residues which remain after the starch has been converted to ethanol. DDGS has been widely used as a protein source for ruminants and non-ruminants for several years. DDGS usually contains about 8693% (db) dry matter, 2634% (db) crude protein, and 313% (db) fat (Tjardes and Wright, 2002; Rosentrater and Muthukumarappan, 2006). As the U.S. ethanol industry has grown exponentially over the past several years, the supply of DDGS has also signicantly increased. As a consequence, it is important to continue to increase the use of distillers grains in the marketplace. To achieve this, however, they will have to be hauled greater distances, and will have to be stored in bins and silos until use; sometimes this may be a relatively long time, depending on the animal production setting. DDGS storage and transportation can often be problematic, as it can harden inside storage structures, trucks and railcars. This can lead to damage when the cars are unloaded. Furthermore, marketing of distillers grain products can be hampered by inconsistencies in physical and nutritional properties, both within a single plant over time, as well as between plants. Physical and chemical properties are important, because DDGS storage and ow behaviour will depend in large part on these parameters, as well as environmental variables. At times, small differences in variables such as moisture content, particle size, storage time, temperature, and relative humidity can cause caking/bridging of granular materials. This results in a great disparity in the owability of DDGS. In light of the growth of the industry, the owability of DDGS and their ow behaviour under various conditions are ` -vis handling operations (storage in silos or important vis-a bins) and transportation. As there is currently little information available regarding the handling and ow behaviour of DDGS, there is a critical need for this data. This review discusses various factors affecting the owability and handling of granular solids and powders, as well as appropriate test methods. Hence, it will be helpful for examining DDGS, as well as other co-product feed materials, and should be a useful resource base from which to address and remediate DDGS owability problems.

2.1.

Moisture content

Amid all factors affecting a products storage, moisture is a key factor, as it strongly inuences microbial growth. Moreover, most organic granular materials are hygroscopic in nature, and they gain or lose moisture when they are exposed to various humidity conditions. Moisture sorption is often coupled with increased cohesiveness, chiey because of interparticle liquid bridge formation. Moisture content thus affects the cohesive strength and arching ability of bulk materials (Johanson, 1978). As the moisture content of a powder increases, the adhesion (Craik and Miller, 1958) and cohesion (Moreyra and Peleg, 1981) tend to increase. Even a small change in moisture content can substantially affect the frictional properties (e.g., wall friction angle, internal angle of friction) of material (Marinelli and Carson, 1992). Physical properties of a material are also highly dependent on the materials moisture content, and each material will behave differently. For example, the angle of internal friction of zinc ore was found to be 32 and 56 at 18 and 23% moisture contents, respectively. This material adhered strongly to the steel bin during experimentation, and would have caused a severe owability problem (Johanson, 1978). Duffy and Puri (1994) studied the owability of confectionary sugar and detergent at two moisture contents. As the moisture content increased, the internal angle of friction of both the sugar and detergent decreased. Also the unconned yield strength of sugar increased sevenfold as the moisture was increased by only 3%. It has also been found that with an increase in moisture, bulk density of granular solids generally decreases and the compressibility increases (Moreyra and Peleg, 1981; Yan and Barbosa-Canovas, 1997). Additionally, moisture at high levels can possibly alter the surface properties of particles to such an extent that the adherence pattern may be modied (Hollenbach et al., 1983). Ganesan et al. (2008c) conducted one of the rst studies to characterize the ow properties of DDGS, and found that DDGS owability decreased with increase in moisture content. For example, at 10% soluble solids, angle of repose increased from 43.47 (at 10% moisture content) to 44.13 (at 30% moisture content); at 20% soluble solids, however, angle of repose increased from 42.57 (at 10% moisture content) to 44.77 (at 30% moisture content). At 10% soluble solids, the compressibility of the DDGS increased from 2.30 to 5.77%, while at 20% soluble solids it increased from 1.90 to 6.63%, at moisture contents of 10 and 30%, respectively.

2. Factors inuencing the owability of granular solids and powders

Flowability is the ability of granular solids and powders to ow. Flow behaviour is multidimensional in nature, and it depends on many physical characteristics. Flowability, in fact, is a consequence of the combination of a materials physical properties that inuence material ow, environmental conditions, and the equipment used for handling, storing, and processing these materials (Prescott and Barnum, 2000). Because of this, no single test can fully quantify a products owability. Some of the factors that affect owability of bulk solids and powders include moisture content, humidity, temperature, pressure, fat, particle size, and ow agents.

2.2.

Humidity

Relative humidity of the air (interstitial as well as head space) in a storage container, such as a bin or silo, also affects bulk materials properties. Many bulk materials are hygroscopic and thus the exposure to humid conditions results in increased moisture content of the bulk. This can lead to an increase in bulk strength (Marinelli and Carson, 1992), and also to an increase in angle of repose. Flowability of any material reduces with an increase in the angle of repose of that material. Many researchers have observed that higher humidities had signicant effects on the owability and cohesiveness of granular powders (Craik and Miller, 1958;

biosystems engineering 101 (2008) 425435

427

Irani et al., 1959; Peleg and Mannheim, 1973; Johanson, 1978; Stanford and Corte, 2002; Fitzpatrick et al., 2004b). Controlled humidity chambers are employed in many laboratories for studying the physical properties and stability of foods and feeds. Sulphuric acid, glycerol, or saturated salt solutions are often used to control the humidity (Rockland, 1960). Saturated salt solutions are most useful, as the three phase systems (vapourliquidsolid) are independent of changes in their total moisture content. A number of salt solutions suitable for sustaining a wide range of specic relative humidity conditions at various temperatures are available (Spencer, 1926; Hodgman, 1954/55). ASTM standard method E-104 also provides information regarding maintaining constant relative humidity by means of aqueous solutions (ASTM, 2002b).

critical arching dimensions. Overpressure effects are not linear and they vary signicantly with the sample. The crushing of particles due to overpressure can be predicted from the shape of the bulk density versus compacting pressure graph (Johanson, 1978).

2.5.

Fat content

2.3.

Temperature

Temperature also has a substantial effect on bulk solid owability. The most drastic temperature effect is the freezing of the moisture contained within the granular materials and on particle surfaces. The resulting ice bonds weaken the ow (Irani et al., 1959; Johanson, 1978; Fitzpatrick et al., 2004b). Varying the storage temperature from above freezing to 30 or 40  C does not usually have a great impact on powder owability, if there is no melting of components or no component exceeds its glass transition temperature (Teunou and Fitzpatrick, 1999). But severe caking can occur whenever a granular material undergoes a change in crystallinity or other properties due to temperature variations (Johanson, 1978). Temperatures of both the wall material and the bulk material may affect the wall friction angle (Marinelli and Carson, 1992). For example, Fitzpatrick et al. (2004b) investigated the effect of temperature on the ow properties of skim milk powder (SMP), whole milk powder (WMP), and high fat milk powder (HFP); the authors observed that HFP showed highest wall friction at lowest temperatures, whereas SMP and WMP had higher wall friction at higher temperatures. Ganesan et al. (2007a, 2008a) studied both the dynamic and equilibrium water adsorption characteristics of DDGS at four temperature (10, 20, 30, and 40  C) and four relative humidity (60, 70, 80, and 90%) levels, and used common mathematical models to t the adsorption and equilibrium data. The authors observed that DDGS showed characteristics similar to a type III isotherm, which often occurs in high sugar foods; this may be an indication of one of the reasons for DDGS ow problems. The authors did not, however, examine the effect of moisture sorption on resulting ow properties of the DDGS.

Free surface fat is expected to play a key role in granular owability as well, but has not been extensively researched to date. For example, high fat content (20%) spray dried soymilk led to worse ow of the resulting soymilk powders (Perez and Flores, 1997). Further, free fat content, varying from 13 to 74%, had no major impact on the cohesion of a 26% fat milk powder at 20  C (Fitzpatrick et al., 2004b). There are many reports available on the chemical composition of DDGS, but there is no study on the surface fat content of DDGS and its effect on DDGS ow properties. Granular material and powder owability depend on the surface composition as surface free fat plays a key role in determining the owability and stickiness of a powder.

2.6.

Particle size

Particle size, and particle size distribution, both play signicant roles in owability and other properties, such as bulk density, angle of repose, and compressibility of bulk solids. Even a small change in particle size can cause signicant alterations in the resulting owability. Reduction in particle size often tends to decrease the owability of a given granular material due to the increased surface area per unit mass (Fitzpatrick et al., 2004a, b). For example, Farely and Valentin (1967/68) investigated the inuence of particle size distribution on bulk powder properties; they found that particle size was the most important factor governing the structure of the powder compact, and at the same time, the interparticulate force governs the strength of the structure. Particle size also plays an important role in the compressibility of powders. An increase in particle size generally leads to an increase in compressibility (and thus volume reduction) (Yan and Barbosa-Canovas, 1997). In a mass ow scenario, if the particles are less than 1/4 inch in size, then cohesive arching will occur during discharge (Marinelli and Carson, 1992). The ner the particle size and greater the range of particle sizes, the greater the cohesive strength, and lower the ow rate (Marinelli and Carson, 1992). Reduction in size increases the contact area between the particles, thereby increasing the cohesive forces.

2.4.

Pressure

2.7.

Flow conditioners and anticaking agents

Compacting pressure is also an important factor that affects the ow properties of bulk solids (like DDGS). The bulk may be subjected to compaction due to vibration (e.g., during transportation), impact from a falling stream of solids (e.g., during silo lling), or external loading. The effect of increased pressure on owability of powders is twofold: (1) it leads to a larger number of contact points between particles, thus causing more inter-particle adhesion (Irani et al., 1959); and (2) the increased compaction produces a signicant increase in

Caking and stickiness are common problems that almost every industry dealing with granular solids and powders encounters. Caking is dened as when two or more macroparticles, each capable of independent translational movement, contact and interact to form a congregate in which the particles are incapable of independent translations (BarbosaCanovas and Yan, 2003). Flow conditioners and anticaking agents are commonly used as additives that can assist a powder in maintaining a steady ow and/or increase its ow

428

biosystems engineering 101 (2008) 425435

rate. Flow conditioners are usually made from chemically inert substances and are often effective at concentrations up to 2%. Most are insoluble in water, but many of them can adsorb signicant quantities of moisture as a result of their very large surface areas. Many research studies have examined the effect of ow conditioners on ow properties of granular solids and powders in the past few decades (Craik, 1958; Craik and Miller, 1958; Irani and Callis, 1960; Sjollema, 1963; Peleg and Mannheim, 1973; Johanson, 1978; Chen and Chou, 1993; Onwulata et al., 1996; Perez and Flores, 1997; Jaya and Das, 2004). To be effective, a ow agent must be ner than the material to be conditioned. The ner the particle size of the conditioner, the less severe the caking of the original material should be (Irani et al., 1959). Additionally, the conditioners particles must stick to the host powder particles, thereby producing a smoother and less frictional surface by lling the void spaces (Peleg and Hollenbach, 1984). Flow properties of conditioned powders are strongly inuenced by the kind of surface interaction between the host powder and the conditioner particles (Hollenbach et al., 1983). If there is no attraction between the particles, then the conditioner particles themselves could segregate and ll the inter-particle voids instead of reducing the cohesiveness (Peleg, 1983). Above a certain inclusion rate, however, the conditioner can actually retard the overall ow of the material (Irani et al., 1959; Nash et al., 1965; Danish and Parrott, 1971; Hollenbach et al., 1982). If the surface afnity between the host powder and conditioner is strong, then large effects will be present in the bulk density and compressibility at inclusion concentrations between 0.1 and 0.5%. When there is little afnity, the effects become noticeable only at higher concentrations, such as 12% (Hollenbach et al., 1983). There is limited information available on using different ow agents with DDGS. Ganesan et al., 2008b examined the use of calcium carbonate at inclusion levels of 0, 1, and 2%, and determined the effects on Carr properties. They found that without any ow agent, at 10% soluble solids, angle of repose increased from 44.77 (at 10% moisture content) to 45.17 (at 30% moisture content); at 20% soluble solids, however, angle of repose increased from 43.13 (at 10% moisture content) to 44.87 (at 30% moisture content). At 10% soluble solids, the compressibility of the DDGS increased from 1.70 to 4.27%, while at 20% soluble solids it increased from 1.33 to 4.50% at moisture contents of 10 and 30%, respectively. Thus, these results were similar to those of Ganesan et al. (2008c). They also found that the effect of adding calcium carbonate, even at 2%, did not signicantly improve the DDGS ow properties. Furthermore, there is currently no literature available on improving the ow of DDGS at commercial settings (i.e., ethanol plants or livestock feeding operations). Thus further research is warranted.

particles, or along the wall surface of a container (Peleg, 1977). Flow characteristics are of vital signicance in bulk material handling and processing, since they impact conveying, blending, and storage options. DDGS often clumps together and forms cakes when it is stored for a long time. These clumps are sometimes very hard to break and often lead to damage to storage structures and economic losses (due to the use of sledgehammers and labour time and salaries). To ensure steady and reliable ow, it is crucial to accurately characterize the ow behaviour of these granular materials (Kamath et al., 1994). Barbosa-Canovas and Yan (2003) allowed a granular material to ow through a laboratory bin/conical funnel and evaluated its owability based on the mass ow rate. The mass ow rate Vm is denoted as Vm arb D 4d (1)

where rb is bulk density (g cm3); D is orice diameter (cm); d is mean diameter of the constituent particles (cm); a and 4 are empirical coefcients (). This approach should work for a variety of materials. Material owability can also be quantied via several additional parameters, including angle of repose, bulk density, angle of internal friction, cohesion, adhesion, and compressibility.

3.1.

Angle of repose

Angle of repose is dened as the angle between the horizontal and the slope of a heap of granular material dropped from some designated elevation. Angle of repose corresponds qualitatively to the ow properties of that material, and is a direct indication of potential owability. Angle of repose of a bulk solid can be described using the following equation (Fowler and Wyatt, 1960; Mohsenin, 1986; Rao, 1992): tanFr an2 b M csg d Dav (2)

3.

Flowability-related properties

where Fr is angle of repose (degrees); n is shape factor based on specic surface (); M is moisture content (db %); Dav is average particle diameter (cm); sg is specic gravity (); a, b, c, and d are empirical constants. Most often, however, angle of repose is determined experimentally by allowing a sample to ow onto a at surface, and then measuring the angle with respect to horizontal. There is much literature available on the angle of repose of granular materials (like food, grains, industrial powders, pharmaceutical powders, etc.) but not on DDGS. Typically, the lower the angle of repose of a dry material, the more owable the material is, and vice versa (Carr, 1965a). Higher angles (i.e., 5060 ) indicate material with difcult ow, while a lower angle, such as 3040 , represents a material with relatively easy ow. Angle of repose gives a reproducible numerical value, so it has been adopted as a common method to assess ow properties (Craik and Miller, 1958). Generally the magnitude of angle of repose increases with increase in moisture content.

The handling, storage, and ow of particulate materials are important in industries associated with agricultural, food, chemical, ceramic, pharmaceutical, metallurgical, and other bulk solids and powder processing. Flow is dened as the relative movement of a bulk of particles among neighbouring

3.2.

Bulk density

Bulk density of granular solids and powders is important when determining the volume of transport vehicles and

biosystems engineering 101 (2008) 425435

429

storage vessels. It mostly depends on particle size, moisture, and chemical composition, but also on handling and processing operations. It dictates the stability of ow and loads on bin walls (Johanson, 1971/72). Bulk density is dened as the mass of particles that occupies a unit volume of a container. Increases in bulk density have been observed when conditioners are added (Peleg and Mannheim, 1973; Hollenbach et al., 1983), which results in modication of density via lowering the inter-particle interactions. Bulk density of food powders has also been observed to decrease with an increase in the particle size as well as with an increase in equilibrium relative humidity (Yan and Barbosa-Canovas, 1997). The bulk density values of DDGS have been shown to range from 389 to 501 kg m3 (Rosentrater, 2006). As of now, however, there is very little known about the inuence of DDGS bulk density on its ow, or the effect of storage conditions on the resultant bulk density. The ratio between tapped (a dened number of taps) and loose bulk density is known as the Hausner ratio, and it is often used as an internal friction index in cohesive powders (Guo et al., 1985; Malave et al., 1985). Porosity, which is related to bulk density, can be expressed as the percentage of voids in a bulk solid:  (3) P V Vp V 100 % where P is porosity (%); V is bulk volume of the bulk (cm3); and Vp is particle volume of the bulk (cm3). P is affected by the ow of the granular material. As porosity decreases, bulk density increases (Sjollema, 1963).

forces, and magnetic forces. The most prevalent inter-particle forces in systems composed of particles are the liquid bridging force and the Van der Waals force (Weber et al., 2004). Cohesion generally increases with a decrease in particle size and with an increase in moisture (Fitzpatrick et al., 2004a), depending on the material. A yield locus plot of failure shear stress versus normal stress for a given consolidating stress is often tted using the Warren Spring equation (Farely and Valentin, 1967/68; Harwood, 1971; Peleg and Mannheim, 1973; Rooda, 1975; Peleg, 1977; Moreyra and Peleg, 1981; BarbosaCanovas and Yan, 2003; Fitzpatrick et al., 2004a):
n s=C s T=T

(4)

3.3.

Frictional forces

The angle of internal friction is a measure of the force required to cause particles to move or slide on each other. Stable slopes and hang-ups in bins are highly dependent on angle of internal friction (Johanson, 1971/72). Internal friction is inuenced by many parameters, including particle surface friction, shape, hardness, size, and size distribution. Angle of internal friction data is needed for calculating the lateral pressure on walls of storage bins and for the design of gravity ow bins and hoppers (Mohsenin, 1986; Rao, 1992). Shear testing is used to measure the angle of internal friction (Peleg and Mannheim, 1973; Teunou et al., 1999; Fitzpatrick et al., 2004b). Wall friction is a key parameter in the design and operation of hoppers, silos, and storage and discharge chutes. It is dened as the frictional resistance to bulk ow that exists between particles and wall material (Iqbal and Fitzpatrick, 2006). Wall friction is a complex phenomenon inuenced by many factors, such as wall surface characteristics, bulk solids properties, and handling conditions (Prescott et al., 1999). Moreover, the wall surface is affected by the surface material, surface roughness, surface wear (Bradley et al., 2000), and surface corrosion. Savage (1967) carried out an analysis of gravity ow of cohesionless bulk solids in a vertical converging channel and observed that the wall friction was more inuential than the angle of internal friction in reducing the ow rate for small values of cone wall half angle. Cohesion is the property of particles adhering together in a bulk. Cohesive forces can occur from a variety of sources, namely liquid bridges, Van der Waals forces, electrostatic

where s is failure shear stress (kPa); C is cohesion (kPa); n is shear index (dimensionless); T is tensile stress (kPa); and s is normal stress (kPa). An increase in cohesion can often be caused by the formation of liquid bridges between the particles due to the melting of lipids (Fitzpatrick et al., 2004b). DDGS contains relatively high fat content (Tjardes and Wright, 2002; Rosentrater and Muthukumarappan, 2006) and at higher temperatures, there is a possibility of fat melting to form liquid bridges between particles. During handling and storage of granular solids, all particles of a given size range can stick together or to container walls. Particle adhesion can be caused by a number of factors, including surface and eld forces (e.g., van der Waals, electrostatic, and magnetic forces), material bridges between particle surfaces (e.g., liquid and solid bridges, occulants), and physical interlocking of the particles themselves (Tomas, 2007). To measure particle adhesion tendency, Ta, Pietsch (1969) developed an empirical relationship: P Bi x i >1 (5) Ta P Fjy x
j

where Bi(x) are the binding forces (N), and Fjy(x) are the components of the environmental forces (N) which are acting upon the particles. The relative effect of adhesion depends on the particle size; to cause adhesion, Ta must be bigger than 1.0. When adhesion is signicant, the angle of repose is increased, and ow is reduced. Adhesion can be suppressed by the addition of ow conditioners (Craik and Miller, 1958). Tomas (2007) modelled single particle adhesion using stiff particles and developed equations for elastic-plastic particle contact behaviour with adhesion, loadunload hysteresis, and adhesion force functions.

3.4.

Compressibility

Much attention has been given to the behaviour of bulk solids under compressive stress. Typically, a set of compression cells is used to obtain the pressuredensity relationship for a given material; the test material is compressed using a compression testing machine, and the force-deformation data must then be converted to a pressuredensity relation (Hollenbach et al., 1983; Barbosa-Canovas et al., 1987; Barbosa-Canovas and Yan, 2003). A number of authors have suggested empirical equations to describe these relationships. Three commonly used models are: Heckel, Kawakita and Ludde, and Sone (Malave et al., 1985; Barbosa-Canovas and Yan, 2003).

430

biosystems engineering 101 (2008) 425435

There are other ways to examine compressibility, however. Hausner ratio, the ratio between tapped and loose bulk density, is also used to quantify the compressibility of bulk solids. The compressibility of a material can also be computed by the following equation: C 100P A=P (6)
3

where C is compressibility (%), P is packed bulk density (kg cm ); A is aerated bulk density (kg cm3). The greater the compressibility of a bulk solid, the less owable it is. In general, the borderline between free owing and non free owing is approximately 2021% compressibility (Carr, 1965b). Compressibility and bulk density are related, and have been correlated using the following empirical equation (Peleg and Mannheim, 1973; Peleg, 1977; Hollenbach et al., 1983; Malave et al., 1985): BD a b log P (7)

Fig. 1 Schematic diagram of Jenike shear cell components.

where BD is bulk density (kg cm3); a is extrapolated BD at 1 kg cm2 pressure (kg cm3); b is the slope of the straight line (compressibility) (kg cm3); P is applied pressure (kg cm2). A decrease in compressibility has been observed with an increase in the particle size (Yan and Barbosa-Canovas, 1997). A decrease in compressibility of various powders and granular solids was observed when conditioners were added (Peleg and Mannheim, 1973; Hollenbach et al., 1983). DDGS is expected to gain moisture when exposed to higher humid conditions. This might increase the compressibility of DDGS and lead to ow problems.

Duffy and Puri, 1994, 1999; Schwedes, 1996; and Fitzpatrick et al., 2004b). The accuracy of the results depends upon the material being tested and the technician performing the procedures, and often has reproducibility problems. To date, only one study has pursued Jenike shear testing of DDGS. Ganesan et al. (2007b) studied the ow behaviour of DDGS at four CDS (10, 15, 20, and 25% soluble solids) and ve moisture (10, 15, 20, 25, and 30%) levels using Jenike shear testing. Unconned yield strength varied from 0.05 to 15.96 kPa, angle of internal friction varied from 12 to 54 , and effective angle of internal friction varied from 25 to 58 . The authors found that the cohesiveness and compressibility of DDGS increased with an increase in moisture content and soluble solids levels.

4. Test methods for measuring bulk ow properties

Knowledge of physical and ow properties of bulk solids is essential for the design of reliable storage systems and equipment for handling these materials. As of now, however, there is very little information available on the test methods to measure the ow properties of DDGS. A review of the test methods available for other bulk materials would thus be useful. Generally, shear testers are used to measure the strength and ow properties of bulk solids (Schwedes, 2002). Several commonly used testers and their principles are discussed below.

4.2.

Carr indices

The handling of granular materials is difcult without knowledge of how they ow in bins and hoppers. Carr (1965a) described a number of procedures that permit the evaluation of

4.1.

Jenike shear cell

Jenike (1964) was the rst to establish the fundamental methods for determining the ow characteristics of bulk materials. The procedures delineated by Jenike have become a standard method D6128 (ASTM, 2000). To analyze the ow of solids in bins and silos, and to develop a model of ow and noow, Jenike used the principles of plastic failure (Fig. 3) with the Mohr-Coulomb failure criteria (Thomson, 1997). Ideally, in free owing powders, the resistance to ow is due to the result of friction; but in cohesive powders, the inter-particle forces are enhanced by compaction, which results in mechanical strength in the bulk (Peleg, 1983). Jenikes direct shear cell tester (Figs. 1 and 2), and procedure for the design of bins, have been commonly used in research and in industrial practice for characterizing a variety of granular materials (Ashton et al., 1965; Schra mli, 1967; York, 1975; Kamath et al., 1993, 1994;

Fig. 2 Jenike shear tester.

biosystems engineering 101 (2008) 425435

431

Fig. 3 Typical Mohr circle failure plot used to analyze Jenike shear cell experimental data.

Fig. 4 Semi-automated tester for Carr indices.

ow characteristics of such materials; and these procedures have become ASTM standard method D6393 (ASTM, 1999); it is widely used in powder and pharmaceutical industries (Kuchii and Tomita, 2002; Yang et al., 2005). The procedures described by Carr provide a realistic and straightforward method for assessing different aspects of owability (Table 1) of any granular material. A tester used to evaluate the Carr ow properties is shown in Fig. 4. More detailed explanations of the Carr test procedures and resulting ow properties can be found in Ganesan et al. (2008b, c) and ASTM D6393 (ASTM, 1999).

and Schulze, 1990). It was rst developed by Bishop and Henkel (1962). For applications in powder storage problems, this method was found to be inadequate as it was designed for a relatively high stress range. Over the years, however, triaxial cells have been redesigned and have been successfully used to investigate agricultural grain properties (Bock et al., 1991). They have also been used to successfully measure powders at low stress ranges (Meerman and Knaapen, 1979; Kolymbas and Wu, 1990).

4.5. 4.3. Ring shear tester

Direct shear cell

The Schulze ring shear tester is also a widely used device to measure ow properties, including angle of internal friction, wall friction, and bulk density (Schulze, 1996; Ro ck and Schwedes, 2005). Detailed explanations of the test procedures for this shear tester can be found in Bromhead (1998) and ASTM method D6773 (ASTM, 2002a). Unlike the Jenike shear tester, this tester requires minimal operating skills and time. Moreover, results with low variability can be obtained with this tester (Schwedes and Schulze, 1990; Schulze, 1996; Schwedes, 2000). This tester, however, does not work well for particle sizes larger than powders.

This apparatus has been used to measure the consolidated shear strength of soil in direct shear. Tsunakawa and Aoki (1982) used a direct shear tester with press loading to measure failure properties of granular materials. When compared to the Jenike shear cell test, the direct shear cell has the advantage that the yield locus can be obtained from only one shear test as a continuous line.

5.

Modelling of bulk solids ow

4.4.

Triaxial cell

The most widely used laboratory equipment for investigating the strength and deformation behaviour of soils at high stresses (i.e., >100 kPa) is the triaxial apparatus (Schwedes

Full scale silo experiments to determine material ow, and resulting effects on wall structures during ow are expensive and difcult to accomplish. Initial work in bulk solids ow by Jenike in the early 1960s was accomplished using numerical methods, which are relatively inexpensive and are amenable to extensive parametric studies. Today, numerical solutions of partial differential equations for stresses and ows are often

Table 1 Carr classication of granular material owability by Carr indices (based on Carr, 1965b) Property
Flow index

Index value
90100 019

Degree of owability/oodability
Very good Very bad

Measures
Bridge breaking measures not required Special apparatus and techniques are required to break the bridge/cake/agglomerations Rotary seal must be used to prevent ushing Rotary seal is not required to prevent ushing

Flood index

80100 024

Very high Wont ush

432

biosystems engineering 101 (2008) 425435

done using either the nite element method (FEM) or the discrete element method (DEM). Many studies have modelled the ow of granular solids and powders using FEM and DEM in the last few decades (see, for example, Langston et al., 1996; Lu et al., 1997; Gyenis et al., 1999; Baxter et al., 2000; Prescott and Barnum, 2000; Guines et al., 2001; Vidal et al., 2004; Zhang and Rosato, 2004). Both FEM and DEM are widely used for simulation of material ow in silos and hoppers, but they each have their own limitations. Most of the models have been used to describe the behaviour of cohesionless powders. Studies on particle mechanics, however, give a better physical understanding of the essential mathematical functions within a bulk solids continuum. Tomas (2001a, b, 2004a, b) examined the fundamentals of cohesive powder consolidation and ow properties using both particle and continuum mechanics. Tykhoniuk et al. (2007) used continuum mechanical models along with DEM to relate macroscopic cohesion and friction to microscopic adhesion and contact friction for ultra ne cohesive powders, and observed that macroscopic cohesion was proportional to microscopic adhesion. Because DDGS owability is a multivariate problem, a comprehensive model to predict the owability of DDGS would be very useful, as it could allow the prediction of when ow problems may occur. Ganesan et al. (2007c) developed an empirical model by combining results from Jenike and Carr testing, then using dimensional analysis and response surface modelling to predict the owability of DDGS. This model was a nonlinear polynomial surface, and incorporated angle of internal frication, effective angle of internal friction, compressibility, dispersibility, and Hausner ratio, and t the data well (R2 0.93). This model, however, requires validation, and could be improved by accounting for other parameters associated with storage and owability, including environmental conditions.

protein content and high drying temperatures are typically used when processing DDGS), capillary condensation, and recrystallization (which may cause bonding at inter-particle contacts) over a period of time.

7.

Conclusions

6.

Implications for DDGS

During the last several years, many ethanol manufacturing plants have been established throughout the Midwest states of the United States, as maize is the major cultivated crop in these areas. Distillers dried grains (DDG) and DDGS have been widely used for cattle feed for several years, and are currently expanding into other major livestock segments as well, because they are an economical and effective replacement for maize, soybean meal, and dicalcium phosphate (especially in swine and poultry feeds). To date, most of the research work on ethanol co-products has focused on chemical and nutritional properties of distillers grains but not on the related physical or ow properties. Moreover, there is currently very little information available for storage and handling operations for these products. To improve the use of these co-product feed materials, there is a vital need to understand their ow behaviour as well as storage and handling characteristics. This discussion should be an essential resource for future research aimed at examining and quantifying the ow properties of DDGS, as ow analysis of DDGS will play a signicant part towards addressing ow problems associated with transport and storage. Remedies should be either economical modications in current DDGS processing, handling, or storage, or DDGS formulation modications to ensure reliable and uniform ow. Successful remedies will, in turn, reduce the limitations imposed by the caking/bridging tendency of DDGS. Additionally, this can help optimize utilization of these materials, and thus improve process economics for ethanol producers, livestock feeders, and ultimately the rural economy.

Today the growth of the DDGS market is limited by its caking/ bridging tendency inside railcars and trucks during shipping, and to some extent bins and hoppers during storage. These occurrences make DDGS hard to unload and can lead to severe damage to shipping and storage containers. DDGS owability problems may arise from a number of simultaneously interacting physical, ow, and chemical properties, environmental variables (i.e., temperature, relative humidity), external variables (i.e., vibration, compaction pressure), and storage time, to name a few. To date, only limited information is available for DDGS. The authors have investigated the ow behaviour of DDGS at various moisture contents, CDS levels, and ow agent levels using both Jenike and Carr test methods (Ganesan et al., 2007b, c, 2008b, c), and have observed that increases in moisture content result in increases in cohesiveness and compressibility. Otherwise, there are no other known studies that have examined DDGS owability. There are many things that must be known to thoroughly understand DDGS owability. These include the chemical nature of particle surfaces (e.g., amorphous sugars, fats, and proteins which coat the particles), glass transition temperature of DDGS, denaturing of protein during production processes (because DDGS has high

Acknowledgements
The authors express their appreciation to the Agricultural Experimental Station (AES), SDSU, and the USDA-ARS for supporting this research.

references

ASTM D6128. (2000). Standard test method for shear testing of bulk solids using the Jenike shear cell. In: ASTM Standards. ASTM. W. ASTM International, Conshohocken, PA, USA. ASTM D6393. (1999). Standard test method for bulk solids characterization by Carr indices. In: ASTM Standards. ASTM. W. ASTM International, Conshohocken, PA, USA. ASTM D6773. (2002a). Standard shear test method for bulk solids using the Schulze ring shear tester. In: ASTM Standards. ASTM. W. ASTM International, Conshohocken, PA, USA. ASTM E104. (2002b). Standard practice for maintaining constant relative humidity by means of aqueous solutions. In: ASTM

biosystems engineering 101 (2008) 425435

433

Standards. ASTM. W. ASTM International, Conshohocken, PA, USA. Ashton M D; Farely R; Valentin F H H (1965). Some investigations into the strength and ow properties of powders. Rheologica Acta, 4(3), 206218. Barbosa-Canovas G V; Lopez J M; Peleg M (1987). Density and compressibility of selected food powders mixtures. Journal of Food Process Engineering, 10, 119. Barbosa-Canovas G V; Yan H (2003). Powder characteristics of preprocessed cereal ours. In: Characterization of Cereals and Flours (Kalentunc G; Breslauer K J eds), pp. 173209. Marcel Dekker, New York. Baxter J H; Abou-Chakra; Tu zu n U; Mills Lamptey B (2000). A DEM simulation and experimental strategy for solving ne powder ow problems. Transactions of Industrial Chemical Engineering, 78, 10191025. Bishop A W; Henkel D J (1962). The Measurement of Soil Properties in the Triaxial Test. Edward Arnold, London. Bock R G; Puri V M; Manbeck H B (1991). Triaxial test sample size effect on stress relaxation of wheat en masse. Transactions of ASAE, 34(3), 966971. Bradley M S A; Pittman A N; Bingley M; Farnish R J; Pickering J (2000). Effect of wall material hardness on choice of wall materials for design of hoppers and silos for the discharge of hard bulk solid. Tribology International, 33(12), 845853. Bromhead E N (1998). Measurement of shear strength. In The Stability of Slopes (second ed.)., pp. 99102 Taylor and Francis Group, New York. Carr Jr. R L (1965a). Evaluating ow properties of solids. Chemical Engineering, 72(3), 163168. Carr Jr. R L (1965b). Classifying ow properties of solids. Chemical Engineering, 72(3), 6972. Chen Y L; Chou J Y (1993). Selection of anticaking agents through crystallization. Powder Technology, 77, 16. Craik D J (1958). The ow properties of starch powders and mixtures. The Journal of Pharmacy and Pharmacology, 10(2), 7379. Craik D J; Miller B F (1958). The ow properties of powders under humid conditions. Journal of Pharmacy and Pharmacology, 10, 136144. Danish F Q; Parrott E L (1971). Effect of concentration and size of lubricant on ow rate of granules. Journal of Pharmaceutical Sciences, 60(5), 752754. Duffy S P; Puri V M (1994). Effect of moisture content on ow properties of powders. ASAE Paper No. 944033. Duffy S P; Puri V M (1999). Measurement and comparison of owability parameters of coated cottonseeds, shelled maize and soybeans at three moisture contents. Transactions of the ASAE, 42(5), 14231427. Farely R; Valentin F H H (1967/68). Effect of particle size upon strength of powders. Powder Technology, 1, 344354. Fitzpatrick J J; Barringer S A; Iqbal T (2004a). Flow property measurement of food powders and sensitivity of Jenikes hopper design methodology to the measured values. Journal of Food Engineering, 61(3), 399405. Fitzpatrick J J; Iqbal T; Delaney C; Twomey T; Keogh M K (2004b). Effect of powder properties and storage conditions on the owability of milk powders with different fat contents. Journal of Food Engineering, 64(4), 435444. Fowler R T; Wyatt F A (1960). The effect of moisture content on the angle of repose of granular solids. Australian Journal for Chemical Engineers, 1, 58. Ganesan V; Rosentrater K A; Muthukumarappan K (2007a). Dynamic water adsorption characteristics of distillers dried grains with solubles (DDGS). Cereal Chemistry, 84(6), 548555. Ganesan V; Muthukumarappan K; Rosentrater K A (2007b). Flow Properties of DDGS with Varying Soluble and Moisture

Content Using Jenike Shear Testing. ASABE Paper No 076063. ASABE, St. Joseph, Mich. Ganesan V; Rosentrater K A; Muthukumarappan K (2007c). Modeling the ow properties of DDGS. Cereal Chemistry, 84(6), 556562. Ganesan V; Muthukumarappan K; Rosentrater K A (2008a). Sorption isotherm characteristics of distillers dried grains with solubles (DDGS). Transactions of the ASABE, 51(1), 18. Ganesan V; Muthukumarappan K; Rosentrater K A (2008b). Effect of ow agent addition on the physical properties of DDGS with varying moisture content and soluble levels. Transactions of the ASABE, 51(2), 591601. Ganesan V; Rosentrater K A; Muthukumarappan K (2008c). Effect of moisture content and soluble level on the physical, chemical, and ow properties of distillers dried grains with solubles (DDGS). Cereal Chemistry, 85(4), 464470. Guines D; Ragneau E; Kerour B (2001). 3D nite element simulation of a square silo with exible walls. Journal of Engineering Mechanics, 127, 10511057. Guo A; Beddow J K; Vetter A F (1985). A simple relationship between particle shape effects and density, ow rate and Hausner ratio. Powder Technology, 43, 279284. pvo lgyi J; Tsuji Y (1999). Discrete particle Gyenis J; Ulbert Zs; Sze simulation of ow regimes in bulk solids mixing and conveying. Powder Technology, 104, 248257. Harwood C F (1971). Compaction effect on ow property indexes for powders. Journal of Pharmaceutical Sciences, 60(1), 161163. Hodgman C D (1954/55). Handbook of Chemistry and Physics. Chemical Rubber Publications, Cleveland, Ohio. pp 23092310. Hollenbach A M; Peleg M; Rufner R (1982). Effect of four anticaking agents on the bulk characteristics of ground sugar. Journal of Food Science, 47, 538544. Hollenbach A M; Peleg M; Rufner R (1983). Interparticle surface afnity and the bulk properties of conditioned powders. Powder Technology, 35, 5162. Iqbal T; Fitzpatrick J J (2006). Effect of storage conditions on the wall friction characteristics of three food powders. Journal of Food Engineering, 72(3), 273280. Irani R R; Callis C F (1960). The use of conditioning agents to improve the handling properties of cereal products. Cereal Science Today, 5(7), 198201, 214. Irani R R; Callis C F; Liu T (1959). Flow conditioning and anticaking agents. Industrial and Engineering Chemistry, 51(10), 12851288. Jaya S; Das H (2004). Effect of maltodextrin, glycerol monostearate and tricalcium phosphate on vacuum dried mango powder properties. Journal of Food Engineering, 63, 125134. Jenike A W (1964). Storage and ow of solids. Bulletin No. 123, Utah Engineering Station. Bulletin of the University of Utah. Johanson J R (1971/72). Modelling ow of bulk solids. Powder Technology, 5, 9399. Johanson J R (1978). Know your material-how to predict and use the properties of bulk solids. Chemical Engineering917, (Desk book issue). Kamath S; Puri V M; Manbeck H B; Hogg R (1993). Flow properties of powders using four testers measurement, comparison and assessment. Powder Technology, 76, 277289. Kamath S; Puri V M; Manbeck H B (1994). Letter: ow property measurement using the Jenike cell for wheat our at various moisture contents and consolidation times. Powder Technology, 81(3), 293297. Kolymbas D; Wu W (1990). Recent results of triaxial tests with granular materials. Powder Technology, 60(2), 99119. Kuchii S; Tomita Y (2002). Flushing phenomenon of ne powder. Powder Technology, 126(3), 275282.

434

biosystems engineering 101 (2008) 425435

Langston P A; Tu zu n U; Heyes D M (1996). Distinct element simulation of interstitial air effects in axially symmetric granular ows in hoppers. Chemical Engineering Science, 51, 873891. Lu Z; Negi S C; Jofriet J C (1997). A numerical model for ow of granular materials in silos. Part 1: model development. Journal of Agricultural Engineering Research, 68(3), 223229. doi:10. 1006/jaer.1997.0196. Malave J; Barbosa-Canovas G V; Peleg M (1985). Comparison of the compaction characteristics of selected food powders by vibration, tapping and mechanical compression. Journal of Food Science, 50, 14731476. Marinelli J; Carson J W (1992). Solve solids ow problems in bins, hoppers and feeders. Chemical Engineering Progress, 88(5), 2228. Meerman W C P M; Knaapen A C (1979). A high pressure triaxial testing cell. Powder Technology, 22(2), 271278. Mohsenin N N (1986). Friction. In: Physical Properties of Plant and Animal Materials, Vol. 1. Gordon and Breach Publishers, New York. 556642. Moreyra R; Peleg M (1981). Effect of equilibrium water activity on the bulk properties of selected food powders. Journal of Food Science, 46, 19181922. Nash J H; Leiter G G; Johnson A P (1965). Effect of antiagglomerant agents on physical properties of nely divided solids. Industrial and Engineering Chemistry, 4, 140145. NCGA (2007). World of Corn-production. National Corn Growers Association, Washington D.C. <http://www.ncga.com/ WorldOfCorn/main/production1.asp> (accessed 23.03.07). Onwulata C I; Konstance R P; Holsinger V H (1996). Flow properties of encapsulated milk fat powders as affected by ow agent. Journal of Food Science, 61(6), 12111215. Peleg M (1977). Flowability of food powders and methods for its evaluation a review. Journal of Food Process Engineering, 1, 303328. Peleg M (1983). Physical characteristics of food powders. In: Physical Properties of Foods (Peleg M; Bagley E eds), pp. 293 323. AVI, New York. Peleg M; Hollenbach A M (1984). Flow conditioners and anticaking agents. Food Technology, 38(3), 93102. Peleg M; Mannheim C H (1973). Effect of conditioners on the ow properties of powdered sucrose. Powder Technology, 7, 4550. Perez M F; Flores R A (1997). Particle size of spray dried soymilk. Applied Engineering in Agriculture, 13(5), 647652. Pietsch W B (1969). Adhesion and agglomeration of solids during storage, ow and handling-a survey. Journal of Engineering for Industry, 5, 435449. Prescott J K; Ploof D A; Carson J W (1999). Developing a better understanding of wall friction. Powder Handling and Processing, 11(1), 1926. Prescott J K; Barnum R A (Oct 2000). On powder owability. Pharmaceutical Technology 6084. Rao M A (1992). Transport and storage of food products. In: Handbook of Food Engineering (Heldman D R; Lund D B eds), pp. 199246. Marcel Dekker, New York. ck M; Schwedes J (2005). Investigations on the caking Ro behaviour of bulk solids-macroscale experiments. Powder Technology, 157(13), 121127. Rockland L B (1960). Saturated salt solutions for static control of relative humidity between 5  C and 40  C. Analytical Chemistry, 32(10), 13751376. Rooda E J (1975). A numerical method for the calculation of the powder ow properties obtained with the Jenike ow factor tester. Powder Technology, 12, 97102. Rosentrater K A; Muthukumarappan K (2006). Corn ethanol coproducts: generation, properties, and future prospects. International Sugar Journal, 108(1295), 648657.

Rosentrater K A (2006). Physical Properties of Distillers Grains with Solubles (DDGS). ASABE paper No: 066164. ASABE, St. Joseph, MI. Savage S B (1967). Gravity ow of a cohesionless bulk solid in a converging conical channel. International Journal of Mechanical Science, 9(9), 651659. mli W (1967). On the measurement of the ow properties of Schra cement. Powder Technology, 1, 221227. Schulze D (1996). Flowability and time consolidation measurements using a ring shear tester. Powder Handling and Processing, 8(3), 221226. Schwedes J (1996). Measurement of ow properties of bulk solids. Powder Technology, 88, 285290. Schwedes J (2000). Testers for measuring ow properties of particulate solids. Powder Handling and Processing, 12(4), 337354. Schwedes J (2002). Consolidation and ow of cohesive bulk solids. Chemical Engineering Science, 57, 287294. Schwedes J; Schulze D (1990). Measurement of ow properties of bulk solids. Powder Technology, 61, 5968. Sjollema A (1963). Some investigations on the free owing properties and porosity of milk powders. Netherlands Milk and Dairy Journal, 17, 245259. Spencer H M (1926). Laboratory methods for maintaining constant humidity. In: International Critical Tables of Numerical Data, Physics, Chemistry and Technology, Vol. 1. McGraw-Hill Publications, New York, NY. 6768. Stanford M K; Corte C D (2002). Effects of humidity on the ow characteristics of PS304 plasma spray feedstock powder blend. NASA/TM-2002-211549. Ohio. Teunou E; Fitzpatrick J J (1999). Effect of relative humidity and temperature on food powder owability. Journal of Food Engineering, 42(2), 109116. Teunou E; Fitzpatrick J J; Synnott E C (1999). Characterisation of food powder owability. Journal of Food Engineering, 39(1), 3137. Thomson F M (1997). Storage and ow of particulate solids. In Handbook of Powder Science and Technology, pp. 389486. Chapman and Hall Publications, New York, NY. Tjardes K; Wright C (2002). Feeding corn distillers co-products to beef cattle. Extension Extra SDSU Extension Cooperative Service. Available from: http://agbiopubs.sdstate.edu/articles/ ExEx2036.pdf. Tomas J (2001a). Assessment of mechanical properties of cohesive particulate solids. Part 1: particle contact constitutive model. Particulate Science and Technology, 19, 95110. Tomas J (2001b). Assessment of mechanical properties of cohesive particulate solids. Part 2: powder ow criteria. Particulate Science and Technology, 19, 111129. Tomas J (2004a). Fundamentals of cohesive powder consolidaton and ow. Granular Matter, 6, 7586. Tomas J (2004b). Product design of cohesive powders-mechanical properties, compression, and ow behavior. Chemical Engineering and Technology, 27(6), 605618. Tomas J (2007). Adhesion of ultra ne particles: a micromechanical approach. Chemical Engineering Science, 62, 19972010. Tsunakawa H; Aoki R (1982). Measurements of the failure properties of granular materials and cohesive powders. Powder Technology, 33, 249256. Tykhoniuk R; Tomas J; Luding S; Kappl M; Heim L; Butt H J (2007). Ultrane cohesive powders: from interparticle contacts to continuum behaviour. Chemical Engineering Science, 62, 28432864. Vidal P; Guaita M; Ayuga F (2004). Simulation of discharging processes in metallic silos. ASAE Paper No. 044151. Weber M W; Hoffman D K; Hrenya C M (2004). Discrete particle simulations of cohesive granular ow using a square-well potential. Granular Matter, 6(4), 239254.

biosystems engineering 101 (2008) 425435

435

Yan H; Barbosa-Canovas G V (1997). Compression characteristics of agglomerated food powders: effect of agglomerate size and water activity. Food Science and Technology International, 3(5), 351359. Yang J; Sliva A; Banerjee A; Dave R N; Pfeffer R (2005). Dry particle coating for improving the owability of cohesive powders. Powder Technology, 158(13), 2133.

York P (1975). The use of glidants to improve the owability of ne lactose powder. Powder Technology, 11, 197198. Zhang N; Rosato A D (2004). Analysis of instantaneous dynamic states of vibrated granular materials. Mechanics Research Communications, 31, 525544.