FRICTION IN AND ENERGY REQUIRED FOR EXTRUDING ALFALFA

S.A. Hann and H.P. Harrison

Department ofAgricultural Engineering, University ofAlberta, Edmonton, Alta.

Received 15 July 1975

Hann, S.A. and H.P. Harrison. 1976. Friction in and energy required for extruding alfalfa. Can. Agric. Eng. 18: 21-25.

High levels of friction are usually encountered in the extrusion process. Using a hydraulic wafering press, a factorialexperiment was conducted to ascertain the effects of pressure, moisture content and binders on the friction to extrudealfalfa, the wafering energy and wafer durability of alfalfa. A split-plot experimental design was used with the bindingagents in the main plots and the other variables in the subplots. The addition of binders at high pressures increased thefriction, except at low pressures. At lower pressuresa durable wafer could be extruded when binders were used. The energyrequired by the extrusion process at the lower pressure is considerably less than is normally required by the presentcommercial wafering presses.

INTRODUCTION

With wafers and pellets, the harvesting,storing and feeding of alfalfa can becompletely mechanized, an obviousadvantage over most other forage harvesting systems (Buckingham 1961).Gustafson and DeBuhr (1965) allege thatwafering and pelleting improves animalutilization of low quality forages. Perhapsequally important is the price stability forforages that wafers and pellets canprovide by expanding the market area.The acceptance of wafers (Curley et al.1973), as distinct from pellets, has notbeen as widespread as expected. Theadvantage of the larger particle sizeassociated with wafers (Hironaka andCheng 1974), has been more than offsetby drawbacks associated with the processitself (Dobie 1973; Dobie and Carnegie1973).

Most wafering machines employ theextruding process, which has been usedfor many years in pelleting. Inherent inthis process are high levels of frictionencountered in extruding the foragethrough the die which is additional to theresistance encountered to compress theforage. In order to compress the forageand overcome the friction, extremelyhigh pressures (Reece 1966) are requiredand therefore considerable energy.Equally significant is the requirement forequipment which can withstand suchpressures. It is expensive to manufactureand maintain.

OBJECTIVES AND EXPERIMENTALDESIGN

The causes of friction in extruding hayhave received little attention, and thefeasibility of extruding a satisfactoryproduct with less energy than is presentlyused has not been established. The objectof the investigation, therefore, was toobtain information on how various

factors affect the friction associated with

the extrusion of alfalfa, and to determinethe operating conditions for the waferingprocess, such that durable wafers can be

21

produced with the least amount ofenergy. The variables included in thestudy were extruding pressure (threelevels), back pressure in the die (representing die characteristics, two levels),moisture content (three levels), and theuse of binding agents (two types).

A split-plot experimental design withtwo replications was selected. Alfalfaharvested in the current year was used forone replication or block with alfalfaharvested in the prior year for the otherblock. Binders were randomized in the

main plots in order that other binders orother binder concentrations could betried in the future without the necessityof repeating this experiment. The 18treatment combinations of moisture con

tent, extruding pressure and backpressure were then randomized withineach subplot.

FACILITIES

A single-shot hydraulic wafering press(Figure 1) was used to simulate theextruding-type wafering process (Hann1974). The press consisted of an upperand lower hydraulically driven ram whichmoved inside an electrically heated die1-1/4 inches (32 mm) in diameter. Thedie temperature used throughout theexperiment was 175°F (79°C), which iscomparable to the usual wafering temperatures. The upper ram exerted theextruding pressure. The lower ram exerted a back pressure in the die equivalent tothe resistance normally caused by the diebeing full of compacted forage.

The wafering energy was determinedby measuring the pressure in the uppercylinder with a pressure transducer, andthe displacement of the upper cylinderram with a rotary potentiometer. Thepressure and displacement were recordedon an x-y plotter; therefore, the areaunder the curve was a measure of the

work done or the energy required. Theextruding time was obtained by measuring the time the wafer required to passthrough the die, as indicated by the x-yplotter. Wafer durability was measured as

Figure 1.

F9

Schematic of wafering press.(A) upper cylinder, (B) lower cylinder, (C) loading column, (D) extrusion die, (E) pressure gauges forupper and lower cylinder, (F) pressure control valves, (G) directionalcontrol valve for both rams,(H) microswitches which activatethe directional control valve, (I) displacement transducer for uppercylinder ram.

a rating, which was determined using acage-type tumbler, as specified in theAmerican Society of AgriculturalEngineering Standard (AgriculturalEngineering Yearbook 1974), S269.1.

PROCEDURES

The hay used in the experiment wasalfalfa which had been chopped to anaverage length of 1-1/2 inches (38 mm).It first was dried to approximately 10%moisture content (wet basis), and wasmaintained at this level until ready foruse. The other moisture contents (20 and

CANADIAN AGRICULTURAL ENGINEERING, VOL. 18 NO. 1, JUNE 1976

TABLE I ANALYSIS OF VARIANCE FOR EXTRUDING TIME TABLE II EXTRUDING TIME

Source of variation df Mean squares F ratio Level Time (sec)

Blocks

Binder - B

Error 1

Moisture - M

1

2

2

2

21.60

101.05

20.97

214.19

1.0

4.8

12.7***

Extruding pressure (psi)1,1003,4005,800

6.3

8.2

14.2

Extruding press. - EMB

ME

BE

MEB

Error 2

2

4

4

4

8

306.31

229.68

48.58

111.34

103 61

18.1***

13.6***

2.9*

6.6***

6.2***

Moisture content (%$10

20

30

12.6

10.4

5.8

24 16.84

Total 53

* 5% level of significance.*** 0.5% level of significance.

30%) were obtained by placing a sampleof the alfalfa in a plastic bag and addingenough water to raise the moisture from10% up to the required level. The bags ofalfalfa were stored in a cool room for 48h to allow the moisture to be distributed

uniformly throughout the sample.Dry binders, either Orzan (Crown

Zellerbach Corp. 1970. The Orzan Products.) or Bentonite (Baroid of CanadaLtd. 1973. Albertabond-Canadian

Bentonite.), were sprinkled over thealfalfa and then sprayed with waterbefore mixing. The water was required toactivate the binders and prevent themfrom separating from the alfalfa. Theratio of water to binder was justsufficient to make a solution which was1:1 for Orzan and 2:1 for Bentonite. The

water plus binder was 5% of the waferweight.

The low back pressure was selected sothat the wafer would pass through the diefairly rapidly, between 3 and 4 sec,whereas the high back pressure wasintended to double the extruding time.The extruding time was variable, however, especially with the high backpressure because of variations in thefriction. The extruding time was used,therefore, as a measure of the friction.The extruding time for each treatmentwas based on an average of 10 trials.

The energy requirement for eachtreatment was obtained from the

pressure-displacement curve recorded bythe x-y plotter. The energy required perwafer was divided by the wafer weightwhich was 0.63 oz (18 g) dry basis. Tenwafers from each treatment were placedin an air-tight plastic bag and left forapproximately 24 h to allow them tofully expand before tumbling for thedurability test.

DISCUSSION OF RESULTS

Separate analysis of the extruding timefor the two back pressures was requiredbecause of the heterogeneity of theresidual variance. For the low level back

pressure, the extruding time was aboutthe same, and only one main effect andone interaction tested significant. For thehigh level back pressure, all the maineffects, except for binders, and all theinteractions, tested significant (Table I).Because of the minimal response with thelow back pressure, discussion is limited tothe results obtained with the high backpressure.

In general the extruding time, or thefrictional resistance in the die, increaseswith an increase in the extruding pressureand with a decrease in the moisture

content (John Deere Co. 1972) (TableII). The increase in friction with higherlevels of pressure is due in part to theincrease in the normal force between the

wafer and the die. A large contributor,however, is the tacky or glutinoussubstances *of alfalfa. These substances,which aid in binding the wafer, ooze tothe wafer-die interface in increasingamounts as the pressure increases, hindering the movement of the wafer throughthe die. These substances, however, areonly tacky at the wafer-die interface atthe lower moisture contents. When the

moisture content is 30%, the lubricatingeffect of the water prevails and theextruding time and friction are minimum.

Significant interactions indicate thatthe variables are not independent of oneanother; that is, the response to changesin one variable depends on the level ormagnitude of another variable. To beginwith, the moisture content-extrudingpressure interaction (Figure 2) indicatesthat moisture is an important lubricant,but only at the higher extruding pressures. The binder-extruding pressureinteraction (Figure 3), on the otherhand, indicates that the binders, especially Orzan, are viscous, but only at thehigher extruding pressures. Pressures aslow as 1,100 psi (7,600 kPa) areapparently insufficient in forcing theglutinous substances to the outer edgesof the wafer where the frictional

resistance is affected.

CANADIAN AGRICULTURAL ENGINEERING, VOL. 18 NO. 1, JUNE 1976

24- T

20 *10 *20 *30

Ld112

8

0 1 2 3 4- 5 6

EXTRUDING PRESSURE (PSI/1000)

Figure 2. The moisture content-extrudingpressure interaction for extrudingtime.

24- T

20

Si16

£l2I-

*W/0 BINDER

• ORZAN

•BENTONITE

0 12 3 4-56

EXTRUDING PRESSURE (PSI/1000)

Figure 3. The binder-extruding pressure interaction for extruding time.

24-

20

q

LdE12

8

*W/0 BINDER•ORZAN• BENTONITE

10 20 30MOISTURE CONTENT (PERCENT)

Figure 4. The binder-moisture content interaction for extruding time.

22

TABLE ffl ANALYSIS OF VARIANCE FOR WAFERING ENERGY

Source of variation

Blocks

Binder - B

Error 1

Moisture - M

Back press. - LExtruding press. - E

ML

ME

MB

LE

LB

EBMLE

MLB

MEB

LEB

MLEB

Error 2

Total

** 0.5% level of significance.

df

1

2

2

2

1

2

2

4

4

2

2

44

4

8

4

8

51

107

Mean squares

661.0

99.34

1.39

70.39

10.05

1,395.030.66

6.16

1.01

1.80

0.61

4.820.57

1.06

0.87

0.37

0.73

1.08

F ratio

47.4***

71.2***

65.3***

9.3***

1,294.1***<1

5,72***

<11.6

<14.5***<1<1<1<1<1

TABLE IV WAFERING ENERGY

Variables Levels Energy (hp-h/ton)

Blocks

Binder

Extrudingpressure

(psi)

Moisture

content

(%)

Back pressure(psi)

1

2

None

Orzan

Bentonite

1,1003,4005,800

10

20

30

Low

High

14.3

12.8

11.8

13.7

15.1

6.8

14.6

19.1

14.7

13.9

12.0

13.2

13.8

The differential response of bindersfor different levels of moisture content(binder-moisutre content interaction,Figure 4) is attributed to the level oftackiness of the binders. When alfalfa iswafered without a binder, the tackinessof the glutinous substances is maximumat a moisture content of around 20%. Tominimize friction, the moisture contentshould be either low (10%) or high (30%).The addition of a binder alters thisrelationship except at a 30% moisturecontent.

Orzan in the liquid form resemblessubstances like shellac, which get tackieras they dry out. When this binder isadded to alfalfa at a 10% moisturecontent, much of the water in the binderis absorbed by the alfalfa thus leaving anextremely tacky substance on the outsideof the forage and thereby causingconsiderable friction in the die. At a 20%moisture content, Orzan remains in afluid state like fresh varnish, providing

considerable lubrication. Though Bentonite is a cement rather than a tackysubstance, it also adversely affects thefriction at the two lower moisturecontents.

The analysis of variance for the energyrequired to compress and extrude alfalfa(wafering energy) is given in Table III. Allthe main effects, including blocks, testedsignificant. The energy required decreaseswith the moisture content, and is greaterwhen binders are used (Table IV).Apparently, alfalfa compresses morereadily at high moisture content and lessreadily when containing binders. Themain effect for blocks is attributed tosome problems with the hydraulic systemof the press.

A small loss of energy occurs when theback pressure is increased. As for theextruding pressure, the energy is afunction of the pressure; therefore, itmust test significant. The relationshipbetween the two variables, however, is

§22

TiePLX

Ld

CD

*W/0 BINDER•ORZAN

• BENTONITE6

0 12 3 4-56EXTRUDING PRESSURE (PSI/1000)

5. The binder-extruding pressure interaction for wafering energy.

Figure

i23T

<3

0 12 3 4-56EXTRUDING PRESSURE (PSI/1000)

The moisture content-extrudingpressure interaction for waferingenergy.

Figure 6.

100 T

P 35Ld

E£ 30LdP-~S5

i-« 80

§75

2 70

♦20 #30

0 12 3 4-56

EXTRUDING PRESSURE (PSI/1000)

Figure 7. The moisture content-extrudingpressure interaction for durability.

non-linear; that is, the energy requiredincreases at a decreasing rate for anincrease in the extruding pressure. Thedifferential response for moisture contentand binders for different levels ofextruding pressure (Figures 5 and 6)indicates that there is little or nodifference between 10% and 20% moisture content or between Orzan andBentonite except at high extrudingpressures.

The analysis of variance for thedurability of the wafers is given in TableV. If a durability rating of 90% or betteris considered adequate, then Table VI

23 CANADIAN AGRICULTURAL ENGINEERING, VOL. 18 NO. 1, JUNE 1976

TABLE V ANALYSIS OF VARIANCE FOR DURABILITY OF WAFERS

Source of variation df Mean square F ratio

Blocks - R 1 255.76 1.14

Binder - B 2 389.60 1.74

Error 1 2 223.96

Moisture - M 2 1,233.53 38.06***

Back press. - L 1 108.40 3.34*

Extruding press. - E 2 1,428.31 44.06***

ML 2 54.26 1.67

ME 4 484.40 14 94***

MB 4 273.73 8.44***

LE 2 87.31 2.69

LB 2 31.23 <1EB 4 260.56 8.04***

MLE 4 29.73 <1MLB 4 8.82 <1MEB 8 164.23 5.06***

LEB 4 20.66 <1MLEB 8 8.17 <1

Error 2 51 32.41

Total 107

* 5% level of significance.*** 0.5% level of significance.

TABLE VI DURABILITY

Variables Levels Durability (%)

Extrudingpressure (psi)

Moisture

content (%)

Back pressure (psi)

1,1003,4005,800

10

20

30

Low

High

84.4

95.1

95.5

90.0

98.2

86.8

90.7

92.7

indicates that durable wafers can be

produced except at high moisture content(30%) and low extruding pressure (1,100psi, 7,600 kPa). With regard to the latter,the moisture content-extruding pressureinteraction (Figure 7) indicates that thelowest level of extruding pressure, orenergy used in the experiment, produceswafers with a durability rating greaterthan 90% if the moisture content is 20%.

Spoilage, however, will be a problem instoring wafers at this moisture content.

The binder-extruding pressure interaction (Figure 8) indicates the feasibilityof producing a durable wafer at a lowextruding pressure if the binder Orzan isadded. The binder-moisture content inter

action (Figure 9) also indicates the samefeasibility at the minimum (and safestorage) moisture content. The second-order interaction of binder-moisture con

tent-extruding pressure reveals littledifference in the durability of wafersextruded at the lowest extruding pressureat 20% moisture content without binder

(Figure 10) and at 10% moisture contentwith the binder Orzan (Figure 11). The

simple effects with the binder Bentoniteare not illustrated.

At present energy costs, the higherextruding pressure (3,400 psi, 23,000kPa) does not incur a large cost penalty,about 10 - 20c per ton of dry alfalfa (6hp-h/ton, 5 kw-h/tonne). On the otherhand, adding Bentonite increases the costby $1.00/ton and by $5.00/ton if Orzanis used. These costs are an oversimplification, however, because they ignoresavings in capital and maintenance costsof less rugged equipment if the extrudingpressures were limited.

SUMMARY AND CONCLUSIONS

The friction associated with the

extrusion of alfalfa reaches a maximum at

high levels of back pressure in the die andextruding pressure when the alfalfacontains either 20% moisture with and

without the binder Bentonite or 10%

moisture content with the binder Orzan.

The addition of either binder increases

the friction, but it lessens with decreasingextruding pressure. At the lowest extrud-

CANADIAN AGRICULTURAL ENGINEERING, VOL. 18 NO. 1, JUNE 1976

100

£95}Ld

£ 90LdP-

w 85

!j 80

9 75

DO 70

0

100 T

Ldag90 tQ-

85 \

^60

S2 75

70

'500 1 2 3 4- 5 6

EXTRUDING PRESSURE (PSI/1000)

Figure 10. The moisture content-extrudingpressure interaction for durabilitywithout binder.

100

g90&£80 }

£70I—I

-J

560

*W/0 BINDER•ORZAN

•BENTONITE

1 2 3 4- 5 6

EXTRUDING PRESSURE (PSI/1000)

Figure 8. The binder-extruding pressure interaction for durability.

*W/0 BINDER•ORZAN

•BENTONITE

10 20 30

MOISTURE CONTENT (PERCENT)

Figure 9. The binder-moisture content interaction for durability.

•10 #20 *30

500 1 2 3 4- 5 BEXTRUDING PRESSURE (PSI/1000)

Figure 11. The moisture content-extrudingpressure interaction for durabilitywith Orzan.

24

ing pressure, friction is largely independent of the moisture content and the use

of binders.

A durable alfalfa wafer can be

produced at lower extruding pressuresand energy inputs than are commonlyused, if the alfalfa has a 20% moisturecontent or by the addition of the binderOrzan if the moisture content of the

alfalfa is 10%. The required energy forthese conditions is approximately 6hp-h/ton of dry hay (5 kw-h/tonne). Withincreased energy costs, the extrudingprocess will be at a disadvantage to otherprocesses (Molitorisz and McColly 1969)unless lower extruding pressures are used.

ACKNOWLEDGMENT

The authors wish to acknowledge the contribu

25

tion to the study by Alberta Agriculture withthe loan of a hydraulic wafering press.

AGRICULTURAL ENGINEERING YEAR

BOOK. 1974. Amer. Soc. Agric. Eng., St.Joseph, Michigan.BUCKINGHAM, F. 1961. Is wafering theanswer? Implement and Tractor, March 1:24-26, March 15: 28-29, 72, 74.CURLEY, R.G., J.B. DOBIE, and P.S. PARSON. 1973. Comparison of stationery and fieldcubing of forage. Trans. Amer. Soc. Agric. Eng.14(2): 361-364, 366.DOBIE, J.B. 1973. Dry versus liquid binders forcubing straw. Trans. Amer. Soc. Agric. Eng.16(3): 508-509.DOBIE, J.B. and E.J. CARNEGIE. 1973.Cubing and storage of moist alfalfa. Trans.Amer. Soc. Agric. Eng. 16(4): 766-768, 772.GUSTAFSON, B.W. and H.E. DEBUHR. 1965.

John Deere '400' hay cuber. Paper No. 65-639,Amer. Soc. Agric. Eng., St. Joseph, Mich.HANN, S.A. 1974. A study of the friction andenergy requirements in extruding hay. Unpublished M.Sc. Thesis, Department of Agricultural Engineering, University of Alberta,Edmonton, Alta.HIRONAKA, R. and K.J. CHENG. 1974.Influence of feed particle size on feedlot cattle.Feedstuffs 46(30).JOHN DEERE CO. 1972. 390 StationaryCuber, Operator's manual OM-E42627.MOLITORISZ, J. and H.F. MCCOLLY. 1969.Development and analysis of the rolling-compressing wafering process. Trans. Amer.Soc. Agric. Eng., 12(4): 419-422, 425.REECE, F.N. 1966. Temperature, pressure andtime relationships in forming dense hay wafers.Trans. Amer. Soc. Agric. Eng. 9(6): 749-751.

CANADIAN AGRICULTURAL ENGINEERING, VOL. 18 NO. 1, JUNE 1976