Journal of Food Engineering 116 (2013) 260–266

Contents lists available at SciVerse ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier .com/locate / j foodeng

Compaction and relaxation characteristics of single compacts producedfrom distiller’s spent grain

Praveen Johnson ⇑, Stefan Cenkowski 1, Jitendra Paliwal 2

Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB, Canada

a r t i c l e i n f o

Article history:Received 9 July 2012Received in revised form 28 November 2012Accepted 29 November 2012Available online 10 December 2012

Keywords:Wet distiller’s spent grainSolublesCompactsDensityStress relaxationAsymptotic modulus

0260-8774/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.jfoodeng.2012.11.025

⇑ Corresponding author. Tel.: +1 (204) 899 6701.E-mail addresses: umjoh526@cc.umanitoba.ca (P.

umanitoba.ca (S. Cenkowski), J_Paliwal@umanitoba.ca1 Tel.: +1 (204) 474 6293.2 Tel.: +1 (204) 474 8429.

a b s t r a c t

Compaction and relaxation characteristics of densified distiller’s spent grain compacts produced at differ-ent levels of compressive pressure (60.3–135.7 MPa), initial moisture content (15%, 20% and 25% wb) andsoluble content (15% and 30%) were analyzed during the study. The compaction levels used in this studycaused up to a 4% wb reduction in the moisture of compacts in comparison to their initial moisture. Thedensity of compacts was analyzed to determine the compaction characteristics of distiller’s spent grainusing Jones model. Analysis of the Jones model showed that there was a significant (P = 0.004) decreasein compressibility with an increase in soluble content from 0% to 30%. The distiller’s spent grain compactswere subjected to relaxation tests and the relaxation data obtained were normalized and analyzed todetermine the asymptotic modulus (EA) of the compacts. The asymptotic modulus was used as a measureof rigidity of the compacts. Distiller’s spent grain compact produced with a compressive force of135.7 MPa and initial moisture of 25% wb possessed the highest EA value.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Distiller’s spent grain is the major by-product of distilleries,where starchy materials such as cereal grains are fermented anddistilled to yield ethanol. There are two types of ethanol processingmethods, i.e. dry milling and wet milling (Belyea et al., 2004). Indry milling process, the starchy grain is first ground into flour(meal) and then processed without separating out the variouscomponents of the grain. Whereas in wet milling process, the star-chy grain is first soaked in water and dilute sulfurous acid, andthen separated into constituent fractions. Dry milling process isless capital intensive and is more common among ethanol plants(Belyea et al., 2004). During the dry milling process in ethanol pro-duction, three major co-products are produced: ethanol, carbondioxide and distiller’s spent grain. During the fermentation pro-cess, starch is converted to alcohol and other fermented products,whereas nutrients such as protein, fiber, fat, vitamins and mineralsremain in distiller’s spent grain.

Distiller’s spent grain has high moisture content (60–85% wb)and low bulk density (380–440 kg m�3), and, therefore, it is verydifficult to handle, transport, store, and utilize in its original form(Zielinska et al., 2009; Rosentrater, 2006; Tumuluru et al., 2010;

ll rights reserved.

Johnson), stefan_cenkowski@(J. Paliwal).

Johnson et al., 2011). During shipping and storage, caking andbridging of distiller’s spent grain are the common problems thathinder its flowability. These problems can in turn induce severedamages to shipping and storage containers. Such damages couldresult in unnecessary expenses related to renting extra machinery,labor, unloading charges, and railcar downtime (Rosentrater, 2006;Tumuluru et al., 2010). Drying and densification of biomass, whichis the process of reducing its volume and moisture content, couldhold the key to the utilization of distiller’s spent grain. In the den-sification process biomass is usually shaped into pellets, briquettesor cubes. Densification reduces the cost of transportation, han-dling, and storage. Also, it provides better flow properties thanthe ground feed, reduces ingredient segregation, produces less feedwastage, improves animal performance and improves palatability(Holley, 1983; Mani et al., 2003; Obernberger and Thek, 2004;McMullen et al., 2005; Franke and Rey, 2006).

Some of the studies reported clogging in compaction units and adecrease in durability of the compacts obtained. The clogging phe-nomenon depends on the type of biomass, moisture content, parti-cle size and its chemical composition. Reece (1966) found that itwas difficult to make wafers when the moisture content of alfalfahay was 25% and above. O’Dogherty and Wheeler (1984) concludedthat increasing the initial moisture content from 20% to 35% wb de-creased the wheat-straw wafer (50 mm diameter) durability fromabout 97% to 85%. Al-Widyan et al. (2002) compacted olive cakeusing maximum pressure (stress) of 15, 25, 35, and 45 MPa andmaterial moisture content of 20%, 25%, 30%, and 35% wb. Their re-sults showed that moisture content in the range of 30–35% wb was

Nomenclature

CDS condensed distiller’s solubleEA asymptotic modulus (MPa)k1, k2 empirical constantsM initial moisture content (% wb)m, b model constantsP compressive pressure (MPa)R2 regression coefficientS soluble content (%)t time (s)

wb wet basisWDG wet distiller’s spent grainY(t) decaying parameter

Symbolsq bulk density (kg/m3)r0 initial stress (MPa)rt stress at time t (MPa)

P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266 261

essential for production of olive cake briquettes with high qualitywhereas moisture content beyond 35% to 40% was not recom-mended for producing stable briquettes.

To reduce the moisture of biomass, rotary or drum driers arecommonly used with hot air at 240 to 550 �C as the drying medium(Stroem et al., 2009). Because of the pasty nature of the spent grain(distiller’s and brewer’s spent grain) exiting the centrifugal separa-tion in the industrial set-up, it is poured onto dried granules(‘seeds’) of the previously dried spent grain and introduced into arotary drier. Therefore, some of the dried spent grain continuouslygoes back to the drier serving as the ‘seed’. Such an operation isneeded to develop a large contact surface area between the pastymaterial and the drying medium. The downside of this process isthat spent grain may be over-dried or even burnt and this mayoccasionally lead to driers catching fire. To avoid the problem offire when hot air is used as the drying medium, Tang et al.(2005) recommended drying the spent grain in thin layer withsuperheated steam (SS) as the drying medium. Zielinska et al.(2009) continued this study by introducing drying of distiller’sspent grain in SS with Teflon spheres as the inert material.

Knowledge of compaction mechanisms is important at the stageof designing energy-efficient compaction equipment and quantify-ing the effects of various process variables on compact quality.Wikstrom and Eliasson (1998) reported that fundamental rheolog-ical measurements can provide detailed information related to thequality of the final product. Among rheological tests, relaxationtesting of viscoelastic material is the most common test used(Bhattacharya and Narasimha, 1997).To produce biomass com-pacts of high rigidity, it is important to understand the fundamen-tal compaction behavior of the material of different conditions(Thomas and van der Poel, 1996; Tabil and Sokhansanj, 1995; Maniet al., 2006). The general concept is to eliminate the common prac-tice of using dried spent grain granules as seed material in rotarydriers and replace them with high moisture compact product.The question remains whether such compacted product can with-stand vigorous movement inside the drier without disintegrating.Therefore, the analysis of factors such as compaction pressure,moisture content, and soluble contents are important for produc-ing good quality compacts of high rigidity. The current study uti-lizes asymptotic modulus as a means for measuring the rigidityof the distiller’s spent grain compacts (Moreyra and Peleg, 1980;Scoville and Peleg, 1981). The objective of the present study is toanalyze the effects of compressive pressure, moisture contentand soluble content on the physical properties of distiller’s spentgrain compacts.

2. Material and methods

2.1. Raw materials

Stillage used for this study was obtained from a local distillery(Mohawk Canada Ltd., a division of Husky Oil Ltd.) in Minnedosa,

MB. The raw material, a non-fermentable residue from starch-to-ethanol fermentation process, was a mixture of corn and wheatin the ratio 9:1.

2.2. Initial sample preparation

The stillage was centrifuged in a Sorvall General Purpose, RC-3centrifuge (Thermo Scientific Co., Asheville, NC) to separate theraw material into different fractions. The centrifuge was operatedat a relative centrifugal force of 790g, with a 1000 mL sample con-tainer, rotating at a speed of 2200 rpm on a radius of 0.146 m for10 min. After centrifugation, the supernatant (thin stillage) wasdiscarded; the remaining part contained the semi-solid fractions(solubles or condensed distiller’s solubles-CDS) and the coarsefractions (wet distiller’s spent grain-WDG). The CDS and WDG frac-tions were bagged in separate airtight containers and stored in afreezer at minus 15 �C. Before each set of experiments, the requiredamounts of samples were thawed at room temperature for 3 h. Theinitial moisture content of the WDG and CDS fractions were 69%and 79.4% wb, respectively. Samples of distiller’s spent grains withsolubles were prepared by blending different proportions of WDGwith CDS.

2.3. Moisture content of distiller’s spent grain samples

Moisture content of WDG and CDS samples were determinedbased on the air-oven drying method (AACC, 2000) using a labora-tory hot air oven (Thermo Electron Corporation, Waltham, MA). Inthis method 2 g of samples were placed in the oven at 135 �C for2 h. Also, oven drying at 135 �C was used to reduce the initial mois-ture content of samples to 15%, 20% and 25 ± 1%.

2.4. Densification using single compaction unit

A single compaction unit was used for producing compacts forthe study. The compaction unit was attached to an Instron univer-sal testing machine (UTM) (Model 3366 Universal Testing Systems,Instron Corp., Norwood, MA). Loading for the sample was madepossible by a 6.4 mm plunger attached to the Instron UTMequipped with a 10 kN load cell. The preset load for the test wasset at 2000, 3000, 4000 and 4500 N, corresponding to pressuresof 60.3, 90.5, 120.6 and 135.7 MPa, respectively, at a crossheadspeed of 50 mm/min. After reducing the moisture content of thesample to required levels, a 0.5 g sample was fed into the die, com-pressed to a preset load, and held at constant deformation for aspecified time of 180 s to arrest the spring back effect. The compactformed was about 13 mm in height and was removed by freelymoving the plunger through the die followed by gentle tappingof the compacted biomass (Tumuluru et al., 2010; Mani et al.,2003).

262 P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266

2.5. Moisture content, single compact density and temperaturemeasurement

The moisture content of the compacts was determined usingASAE S358.2 method (ASAE, 2003), in which single compacts weredried in the convection oven at 103 �C for 24 h. The mass measure-ments of each compact were done using an electronic balance withan accuracy of 0.001 g (Model Adventurer Pro AV313, Ohaus Cor-poration, Pine Brook, NJ). For compact density determination, massand volume of the compact was measured immediately after eject-ing the compact from the die after densification. The volume of thecompact was calculated by measuring its length and diameterusing a Vernier caliper (Tumuluru et al., 2010; Shankar et al.,2008). The temperature of the compact was measured immediatelyafter densification (no compression holding time) using a T-typeneedle thermocouple (Omega Technologies Company, Stamford,CT) to determine the extent of temperature rise of the compactdue to frictional heating.

2.6. Jones model

Jones (1960) expressed the density-pressure data of compactedpowder in the form of Eq. (1).

lnðqÞ ¼ m lnðPÞ þ b ð1Þ

where, q is the bulk density of the compact powder mixture, kg/m3,P is the applied compressive pressure, MPa; m and b are model con-stants that are determined from the slope and intercept, respec-tively, of the extrapolated linear region of the plot of (q) vs. ln(P)(Adapa, 2011; Adapa et al., 2009). The constant m can provide valu-able information about the onset of plastic deformation of theground biomass samples. York and Pilpel (1973) have shown theconstant m to be equal to the reciprocal of the mean yield pressurerequired to induce plastic deformation.

2.7. Stress relaxation analysis

The stress relaxation data obtained from the compaction exper-iment were analyzed by the method given by Peleg (1980) forpowdered materials (Peleg, 1979; Peleg and Moreyra, 1979). Theexperimental stress relaxation curve, as recorded by the Instrontesting machine was normalized using Eq. (2).

YðtÞ ¼ r0 � rt

r0ð2Þ

where Y(t) represents the decay of the stress and is called the decay-ing parameter. rt is the stress at any time t, r0 is the initial stress, tis compression holding time. The characteristic shape of the func-tion Y(t) vs. t suggests the simplified mathematical form given byMickley et al. (1957).

YðtÞ ¼ tk1 þ k2 t

ð3Þ

where k1 and k2 are the empirical constants (Bhattacharya, 2010).The linear form of Eq. (3) can be represented by Eq. (4). Linear formsimplifies the verification of the appropriateness of the formula andcalculation of its constants. Peleg and Pollak (1982) stated that be-sides the mathematical simplicity of Eq. (4), its applicability can beverified by testing the fit of data into a linear relationship.

tYðtÞ ¼ k1 þ k2 t ð4Þ

The constants k1 and k2 are independent of the test duration andcalculation procedure. The reciprocal of k1 represents the initial de-cay rate; hence small values of the line intercepts indicate fastrelaxation at initial stages. The slope of the straight line, k2, is

called the solidity index of the material and it can be consideredas an index of how ‘‘solid’’ the compacted specimen is on a shorttime scale.

When time t tends to infinity, the reciprocal of k2 reaches theasymptotic level of Y(t) = Y(1). Since the equilibrium conditionsof biological materials are difficult to determine, 1/k2 can be usedas the representative of the equilibrium condition. The value of1/k2 becomes zero for ideal elastic solids and one for ideal liquids.Therefore, 0 < 1/k2 < 1 represents the asymptotic residual values ofY(1). Hence, r0 (1–1/k2) represents the portion of the stress thatwould have remained unrelaxed at equilibrium. Therefore, whenlong term experiments become theoretically difficult, the asymp-totic value can be used to calculate a residual modulus that is rep-resentative of the short term rheological characteristics. Theresidual (unrelaxed) modulus is given by Eq. (5).

EA ¼r0

e1� 1

k2

� �ð5Þ

where EA is the asymptotic modulus (MPa) and e is the strain(dimensionless). Asymptotic modulus is an indication of an empir-ical index of solidity of the compacts; the higher the value of EA, themore solid the compact is (Mani et al., 2006).

2.8. Statistical analysis

A completely randomized design was chosen for the study withthree factors: compressive pressure (60.3, 90.5, 120.6 and135.7 MPa), initial moisture content (15%, 20% and 25% wb), andsoluble content (15% and 30%). All experiments were conductedin triplicate. The WDG sample with 25% moisture content was cho-sen as reference. One way analysis of variance was used to analyzethe significance of data obtained from the study. All statisticalanalyses were performed using Sigma Stat 3.5 (Systat SoftwareInc., Chicago, IL).

3. Results and discussion

3.1. Compression test and compact density

The influence of compressive pressure, moisture content andsoluble content on compact density is shown in Figs. 1 and 2.

In our experiments, the density of compacts increased signifi-cantly (P < 0.001) with the increase of compressive pressure forall treatments. During the compaction process, rearrangement,sliding and stacking of the particles takes place under low pressureto closely pack the raw material. With the increase in compressionpressure, elastic and plastic deformation of particles occurs to formstrong solid bridges between the particles; this results in higherstrength of the produced compacts. Hence, high compactionpressures will result in reduced porosity and increased compactdensity. Sastry and Fuerstenau (1973) reported that higher com-paction pressures result in the development of bonding forces suchas van der Waal forces providing higher strength to the compacts.

The major factors which affect the densification of a biomasssample are: its nutrient composition, feed moisture content, com-paction pressure, materials used as particle-binders, feed particlesize and densification equipment variables (Turner, 1995). Particlesize of the feed material is a key factor in determining compactsdurability. Generally, tiny particles absorb more moisture and heatwhen compared to larger particles and undergo a higher degree ofconditioning (Kaliyan and Morey, 2009). The major issue involvedin using small particle sizes in feed industries is the high cost in-volved in grinding. A mixture of different sized particles will makepellets of optimum quality, as they form the inter-particle bonding

Fig. 1. Compact density of WDG samples as affected by compressive pressure andinitial moisture content (M- initial moisture content, % wb).

Fig. 2. Compact density of WDG samples as affected by compressive pressure forsamples with and without solubles (initial moisture content of samples is 25% wb;S- soluble content).

Fig. 3. Moisture content of WDG compacts with different levels of compaction andinitial moisture content (M- initial moisture content, % wb).

Fig. 4. Moisture content of distiller’s spent grain compacts as affected bycompressive pressure for samples with and without solubles (initial moisturecontent of samples is 25% wb; S- soluble content).

P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266 263

with nearly no inter-particle spaces (Payne, 1978; Grover andMishra, 1996).

Nutritional components such as proteins will denature underthe presence of heat and act as a binder during the compactionprocess (Thomas et al., 1998). Water-soluble fibers will increasethe viscosity of the feed whereby positively affecting the durabilityof compacts. But water-insoluble fibers may negatively affect thedurability of the densified products. However, due to their resil-ience characteristics, fibers may not provide good bonding be-tween the particles (Thomas et al., 1998).

Fig. 1 shows a negative correlation between moisture contentand compact density. Though the difference obtained was statisti-cally significant (P < 0.001) only for 60.6 MPa, a decreasing trendfor compact density with increasing moisture content was ob-served for other compaction pressures too. Increasing the moisturecontent of samples increases the amount of water molecules occu-pying the void spaces that cannot be expelled during the compac-tion process. This, in turn, increases the volume of the compactedmass and at the same time decreases its density.

Compacts of different soluble contents showed differences incompact density which were statistically significant (P = 0.004)only for the lowest compaction of 60.6 MPa. Compact density tendsto increase with an increase in soluble levels (Fig. 2) at 60.6 MPa.

This may be due to the fact that solubles have smaller particle sizeand they can effectively occupy the inter-particle void spaces. Pre-vious studies reported by Tabil et al. (2011) showed that compactswith smaller particle sizes have higher density than compacts withlarger particle sizes. However, for other compaction pressures thistrend was not observed and the difference in resultant compactdensity was not statistically significant.

The results of the temperature measurements of the compactsshowed that the increase in temperature of the compact was onlyup to 4 �C. Such a small increase was a result of the slow compac-tion rate and the cooling effect that decreased frictional heating.

3.2. Moisture content of distiller’s spent grain compacts

Moisture present in the sample acts as a facilitator of naturalbinding agents as well as a lubricant. Optimum moisture contentis an essential factor for producing high quality compacts. Ourstudy showed that samples with higher moisture contents (>25%wb) plugged the compaction unit due to the adhesive forces devel-oped between the sample and the metallic surfaces. An oven dryingtemperature of 135 �C was chosen to reduce the initial moisture

Fig. 5. Compact density as affected by compressive pressure in natural logcoordinates for distiller’s spent grain samples.

Table 1Compression characteristics of distiller’s spent grain samples using Jones Model.

Moisture content (%, wb) Soluble content (%) m (kg/N-m) b (kg/m3)

15 0 0.135 6.34520 0 0.138 6.32125 0 0.155 6.23125 15 0.108 6.45525 30 0.078 6.602

264 P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266

content of the sample. This was based on the studies done by Tanget al. (2005) who demonstrated that in a superheated steam drierincreasing the drying temperature from 110 to 180 �C had only asmall effect on the change of the nutrients in spent grain samples.

Figs. 3 and 4 show the experimental results of moisture changesin the distiller’s spent grain compacts as a result of the compres-sion level for different initial moisture content of samples andthe amount of solubles.

It is evident that there was a difference in moisture content be-tween the initial sample and the compact obtained. Increasing thecompressive pressure from 60.3 to 135.7 MPa reduced the mois-ture after compaction by 1% to 4% wb. The reduction in moisturecontent was caused by the expulsion of water from inter-particlepore spaces in the process of compaction.

Fig. 6. A typical printout from the Instron machine of a compression relaxation curve forforce and 180 s compression holding time.

3.3. Jones model

Fig. 5 shows the logarithmic plot between compressive pressureand compact density for distiller’s spent grain. In this graph thedensity-pressure data were connected with straight lines insteadof showing a regression line for each treatment. However, the anal-ysis of the relationship between the compact density and compres-sive pressure was based on the regression line as per Eq. (1).

The constants m and b determined from the slope and interceptof the logarithmic plot are shown in Table 1. Large m value (lowyield pressure) indicates that the material is more compressibledue to the onset of plastic deformation at a relatively low pressure.However, the relatively low compact density obtained at the low-est compressive pressure may have influenced the constants ob-tained from this study. The results indicate that an increase ininitial moisture content increases the compressibility (an increasein coefficient m) of samples. But the increase in compressibilitywith the initial moisture content was not statistically significant.Weir (1952) reported that compressibility increases markedly fornatural polymers having high moisture compared to low moisturepolymers; which may be due to the fact that water exists in hydro-gen bonded form at low moisture and as free liquid at high mois-ture contents.

Our study showed that addition of solubles adversely affects thecompressibility of samples. A significant (P = 0.004) decrease incompressibility was observed when the soluble content increasedfrom 0% to 30%. While reducing the moisture content of the dis-tiller’s spent grain samples by oven drying, it was observed thatthe spent grain particles agglomerated and hardened with theaddition of solubles. This hardening may in turn reduce the com-pressibility of the samples. A significant difference (P = 0.002) be-tween b values was observed between samples with 30% solublesin comparison to 0% solubles. But for different moisture contents,the b values obtained were not significant.

3.4. Stress relaxation analysis and residual modulus

A typical compression relaxation curve for the distiller’s spentgrain samples is shown in Fig. 6. A plot of r0 t/(r0-rt) vs. timecan also be used for testing the fit of a linear relationship (Fig. 7).Higher R2 values (P0.98) were obtained for all treatments whenthe relaxation data was linearized and presented as a linear func-tion of time. The slope, k2, ranged between 1.2 and 3.5 for distiller’sspent grain samples. A value of k2 greater than unity is an indica-tion of the existence of stresses that will eventually remain unre-

a WDG sample at 20% wb initial moisture content, 4000 N (120.6 MPa) compressive

Fig. 7. Linearization curve for WDG sample at 15% wb initial moisture content,4500 N (135.7 MPa) compressive force and 180 s compression holding time.

Fig. 8. The effect of compressive pressure on asymptotic (EA) modulus of WDGsamples for different levels of initial moisture content (M- initial moisture content,% wb).

Fig. 9. The effect of compressive pressure on asymptotic (EA) modulus of distiller’sspent grain with and without solubles (initial moisture content is 25% wb; S-soluble content).

P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266 265

laxed. Asymptotic modulus was calculated using the slope ob-tained from the linearized plots. The EA values indicate the abilityof the compressed powder to sustain unrelaxed stresses; the high-er the value of EA the more solid the compact is.

An increase in compressive pressure from 60.3 to 135.7 MPasignificantly increased EA values for all treatments (P < 0.001). Also,increasing the moisture content of samples significantly (P < 0.002)increased the EA values; this may be due to the fact that moisture isacting as a binding agent providing more strength to the compacts(Fig. 8). Moreyra and Peleg (1980) reported that moisture contentmodified the stress relaxation pattern of the compact product dueto reorientation of liquid bridges and plasticity of the bed solid ma-trix. For compacts having different soluble levels, a significant dif-ference in the EA values were obtained for compacts producedusing 60.3 (P < 0.005), 90.5 (P < 0.001) and 135.7 (P = 0.002) MPa(Fig. 9).

Results showed that compacts produced using a WDG samplewith moisture content of 25% wb, and compressive pressure of135.7 MPa possessed the highest EA value. Since EA values are theindication of rigidity of the compacts, it can be interpreted thatthose compacts with the highest EA value have the highest rigidity.Moreover, results obtained from the study showed that asymptoticmodulus can be used to characterize the distiller’s spent grain com-pacts affected by different compressive pressures and moisture

contents. But EA was unable to be used to characterize distiller’sspent grain compacts based on soluble levels for specific compres-sive pressures. A specific relation was not observed for EA valueswith respect to different soluble contents for specific compressivepressures. It is anticipated that the results obtained from this studywill enable to generate new knowledge/technology for the effectiveutilization of distiller’s spent grain. Though the study was done forsingle compacts, the results from the study can be utilized commer-cially for distiller’s spent grain products such as pellets, briquettesand wafers involving a compaction process.

4. Conclusions

Density of compacts increased significantly with the increase ofcompressive pressure for all treatments. Increasing the compres-sive pressure from 60.3 to 135.7 MPa reduced moisture after com-paction of the compacts by 1% to 4% wb. Analysis of the Jonesmodel showed that, there was a non significant increase in thecompressibility of the distiller’s spent grain samples when itsmoisture content was between 15% and 25% (wb); whereas, a sig-nificant decrease in compressibility was observed when 15% and30% of solubles were added to distiller’s spent grain. Resultsshowed that asymptotic modulus can be used to characterize thedistiller’s spent grain compacts affected by different compressivepressures and moisture contents for all treatments. However, EA

was unable to characterize distiller’s spent grain compacts basedon soluble levels for specific compressive pressures. Among theother treatments, a compressive force of 135.7 MPa and moisturecontent of 25% wb produced compacts of highest EA value.

Acknowledgements

The authors thank the Natural Sciences and Engineering Re-search Council of Canada for their financial support.

References

AACC, 2000. Air-oven method, drying at 135 �C. In: Approved Methods of theAmerican Association of Cereal Chemists: AACC, vol. II, Method 44-19. St. Paul,MN.

Adapa, P.K., 2011. Densification of selected agricultural crop residues as feedstockfor the biofuel industry. Unpublished Ph.D thesis. Saskatchewan, SK,Department of Chemical and Biological Engineering, University ofSaskatchewan.

Adapa, P., Tabil, L.G., Schoenau, G.J., 2009. Compaction characteristics of barley,canola, oat and wheat straw. Biosystems Engineering 104 (3), 1–10.

266 P. Johnson et al. / Journal of Food Engineering 116 (2013) 260–266

Al-Widyan, M.I., Al-Jalil, H.F., Abu-Zreig, M.M., Abu-Hamdeh, N.H., 2002. Physicaldurability and stability of olive cake briquettes. Canadian BiosystemsEngineering 44 (3), 41–45.

ASAE, 2003. Moisture measurement – Forages. In: American Society of Agriculturaland Biological Engineers Standards, 50th ed., Method S-358.2, St. Joseph, M.I.

Belyea, R.L., Rausch, K.D., Tumbleson, M.E., 2004. Composition of corn and distillersdried grains with soluble from dry grind ethanol processing. BioresourceTechnology 94 (3), 293–298.

Bhattacharya, S., 2010. Stress relaxation behaviour of moth bean flour dough:product characteristics and suitability of model. Journal of Food Engineering 97(4), 539–546.

Bhattacharya, S., Narasimha, H.V., 1997. Puncture and stress relaxation behaviour ofblack gram (Phaseolus mungo) flour-based papad dough. Journal of FoodProcess Engineering 20 (4), 301–316.

Franke, M., Rey, A., 2006. Pelleting quality. World Grain, 78–79.Grover, P.D., Mishra, S.K., 1996. Biomass briquetting: technology and practices. Field

Document No. 46. Regional Wood Energy Development Program in Asia, Foodand Agriculture Organization of the United Nations, Bangkok, Thailand.

Holley, C.A., 1983. The densification of biomass by roll briquetting. In: Proceedingsof the Institute for Briquetting and Agglomeration. IBA 18th BiennielConference, Colorado Springs, USA.

Johnson, P., Cenkowski, S., Paliwal, J., 2011. Bulk density and angle of repose ofdistiller’s spent grain under different drying methods and solubleconcentrations. In: Proceedings of the Canadian Society for BioengineeringConference, Winnipeg, Canada.

Jones, W.D., 1960. Fundamental Principles of Powder Metallurgy. Edward ArnoldPublishers Ltd, London, UK, pp. 242–370.

Kaliyan, N., Morey, R.V., 2009. Factors affecting strength and durability of densifiedbiomass products. Biomass and Bioenergy 33 (3), 337–359.

Mani, S., Tabil, L.G., Sokhansanj, S., 2003. An overview of compaction of biomassgrinds. Powder Handling and Processing 15 (2), 160–168.

Mani, S., Tabil, L.G., Sokhansanj, S., 2006. Effects of compressive force, particle sizeand moisture content on mechanical properties of biomass pellets from grasses.Biomass and Bioenergy 30 (7), 648–654.

McMullen, J., Fasina, O.O., Wood, C.W., Feng, Y., 2005. Storage and handlingcharacteristics of pellets from poultry litter. Applied Engineering in Agriculture21 (4), 645–651.

Mickley, H.S., Sherwood, T.K., Reed, C.E., 1957. Applied Mathematics in ChemicalEngineering, second ed. McGraw-Hill Book Co., New York, pp. 8–14.

Moreyra, R., Peleg, M., 1980. Compressive deformation patterns of selected foodpowders. Journal of Food Science 45 (4), 866–868.

Obernberger, I., Thek, G., 2004. Physical characterisation and chemical compositionof densified biomass fuels with regard to their combustion behavior. Biomassand Bioenergy 27 (6), 653–669.

O’Dogherty, M.J., Wheeler, J.A., 1984. Compression of straw to high densities inclosed cylindrical dies. Journal of Agricultural Engineering Research 29 (1), 61–72.

Payne, J.D., 1978. Improving quality of pellet feeds. Milling Feed and Fertilizer 162(5), 34–41.

Peleg, M., 1979. Characterization of the stress relaxation curves of solid foods.Journal of Food Science 44 (1), 277–281.

Peleg, M., 1980. Linearization of relaxation and creep curves of solid biologicalmaterials. Journal of Rheology 24 (4), 451–463.

Peleg, M., Moreyra, R., 1979. Effect of moisture on the stress relaxation pattern ofcompacted powders. Powder Technology 23 (2), 277–279.

Peleg, M., Pollak, N., 1982. The problem of equilibrium conditions in stressrelaxation analysis of solid foods. Journal of Texture Studies 13 (1), 1–11.

Reece, F.N., 1966. Temperature, pressure, and time relationships in forming densehay wafers. Transactions of the ASABE 9 (6), 749–751.

Rosentrater, K.A., 2006. Some physical properties of distillers dried grains withsoluble (DDGS). Applied Engineering in Agriculture 224 (4), 589–595.

Sastry, K.V.S., Fuerstenau, D.W., 1973. Mechanisms of agglomerate growth in greenpelletization. Powder Technology 7 (2), 97–105.

Scoville, E., Peleg, M., 1981. Evaluation of the effects of liquid bridges on the bulkproperties of model powders. Journal of Food Science 46 (1), 174–177.

Shankar, T.J., Sokhansanj, S., Bandyopadhyay, S., Bawa, A.S., 2008. A case study onoptimization of biomass flow during single-screw extrusion cooking usinggenetic algorithm (GA) and response surface method (RSM). Food andBioprocess Technology 3 (4), 498–510.

Stroem, L.K., Desai, D.K., Hoadley, A.F.A., 2009. Superheated steam drying ofBrewer’s spent grain in a rotary drum. Advanced Powder Technology 20 (3),240–244.

Tabil, L., Adapa, P., Kashaninejad, M., 2011. Biomass feedstock pre-processing – Part2: Densification. In: Bernardes, M.A.S. (Ed.), Biofuel’s Engineering ProcessTechnology. In Tech, Brazil, pp. 439–464.

Tabil, L.G., Sokhansanj, S., 1995. The use of binders in pelleting of alfalfa. In:Proceedings of the Canadian Society for Agricultural Engineering Conference,Mansonville, Canada.

Tang, Z., Cenkowski, S., Izydorczyk, M., 2005. Thin-layer drying of spent grains insuperheated steam. Journal of Food Engineering 67 (4), 457–465.

Thomas, M., van der Poel, A.F.B., 1996. Physical quality of pelleted animal feed 1.Criteria for pellet quality. Animal Feed Science Technology 61 (1-4), 89–112.

Thomas, M., van Vliet, T., van der Poel, A.F.B., 1998. Physical quality of pelletedanimal feed. 3. Contribution of feedstuff components. Animal Feed Science andTechnology 70 (1–2), 59–78.

Tumuluru, J.S., Tabil, L., Opoku, A., Mosqueda, M.R., Fadeyi, O., 2010. Effect ofprocess variables on the quality characteristics of pelleted wheat distiller’sdried grains with solubles. Biosystems Engineering 105 (4), 466–475.

Turner, R., 1995. Bottom line in feed processing: achieving optimum pellet quality.Feed Management 46, 30–33.

Weir, C.E., 1952. Effect of moisture on compressibility of natural high polymers.Journal of Research of the National Bureau of Standards 49 (3), 135–139.

Wikstrom, K., Eliasson, A.C., 1998. Effects of enzymes and oxidizing agents on shearstress relaxation of wheat flour dough: additions of protease, glucose oxidase,ascorbic acid and potassium bromate. Cereal Chemistry 75 (3), 331–337.

York, P., Pilpel, N., 1973. The tensile strength and compression behavior of lactose,four fatty acids and their mixture in relation to tableting. Journal of Pharmacyand Pharmacology 25, 1–11.

Zielinska, M., Cenkowski, S., Markowski, M., 2009. Superheated steam drying ofdistillers’ spent grains on a single inert particle. Drying Technology: AnInternational Journal 27 (12), 1279–1285.