Physical Properties of Filbert Nut and Kernel

ARTICLE IN PRESS

doi:10.1016/j.biosystemseng.2005.11.008PH—Postharvest Technology

Biosystems Engineering (2006) 93 (2), 173–178

Physical Properties of Filbert Nut and Kernel

S. Pliestic; N. Dobricevic; D. Filipovic; Z. Gospodaric

Faculty of Agriculture, University of Zagreb, Svetosimunska 25, Zagreb, Croatia; e-mail of corresponding author: [email protected]

(Received 1 February 2005; accepted in revised form 11 November 2005; published online 23 January 2006)

Various physical properties of filbert (Corylus maxima cv. Istrian long) nuts and kernels were determined as afunction of moisture content. The average length, width, thickness, equivalent diameter, unit mass, volumeand sphericity of nuts were 25�32, 20�54, 17�93, 20�96mm, 3�88 g, 4�88 cm3 and 82�86%, while correspondingvalues for kernels were 20�20, 14�52, 12�64, 15�41mm, 1�70 g, 1�94 cm3 and 77�02%, respectively, at a moisturecontent of 6�19% wet basis (w.b.). In the moisture range from 6�19 to 28�71% w.b., studies on the rewetted nutshowed that the bulk density of nut and kernel decreased from 530 to 454 kgm�3 and 649 to 569 kgm�3,respectively. The true density of nut decreased from 907 to 829 kgm�3, while for kernel, it decreased from 1016to 937 kgm�3. The porosity of nut increased from 41�53% to 45�24%, while porosity of kernel increased from36�18% to 39�44%. The projected area for nut and kernel increased from 423 to 497mm2 and from 246 to283mm2, respectively. In the same moisture range, the static coefficient of friction for nuts on three differentmaterial surfaces varied from 0�233 to 0�450 on aluminium, from 0�319 to 0�531 on plywood and from 0�406 to0�623 on rubber; the corresponding values for kernels were 0�317–0�484, 0�401–0�581 and 0�561–0�731. Themaximum force for nut cracking occurred in the longitudinal direction and the minimum force in thetransverse direction. For all compression directions, the force required to crack the nut decreased withincreasing moisture content.r 2005 Silsoe Research Institute. All rights reserved

Published by Elsevier Ltd

1. Introduction

Filbert (Corylus maxima) is small nut-bearing tree ofthe genus Corylus, much grown in Europe and closelyrelated to hazelnut. Production of filbert in Croatia hasa long tradition, but in the last years of the 20th century,many filbert plantations were abandoned or replantedwith other fruit-trees. However today, filbert has againconsiderable agronomic and economic potential becauseof good agroclimatic conditions for its production, anda highly developed food industry in Croatia. As kernelsof the filbert have a high oil content, they are a highlyprized food and energy source, and also an importantsource of proteins and vitamins (Miljkovic, 1991).In Croatia, there are only a few large producers who

have machinery for filbert harvesting and handling.Many small producers carried out harvesting andhandling of the nuts manually. The physical propertiesof filbert nuts and kernels, like those of other grains andseeds, are essential for the designing of equipment for

1537-5110/$32.00 173

handling, harvesting, processing and storing the grain,or determining the behaviour of the grain for itshandling (Baryeh, 2001). Physical properties of numer-ous grains and seeds have been determined by otherresearchers. Some of them determined physical proper-ties of different nuts such as gorgon nut (Jha & Prasad,1993), neem nut (Visvanathan et al., 1996), cashew nut(Balasubramanian, 2001), bambara groundnut (Baryeh,2001) and arecanut (Kaleemullah & Gunasekar, 2002).Several researchers determined mechanical properties ofnuts such as macadamia nut (Braga et al., 1999), sheanut (Olaniyan & Oje, 2002) and walnut (Koyuncu et al.,2004). Aydin (2002) has studied physical properties ofhazelnut cultivar Tombul produced in Turkey, the mainhazelnut-producing country in the world. He has alsomeasured forces for cracking the nuts. In Croatia, filbertnut cracking is usually carried out manually or withhomemade cracking machines. Damage occurring dur-ing cracking is among the major causes of qualityreduction of filbert kernels. The percentage of the

r 2005 Silsoe Research Institute. All rights reserved

Published by Elsevier Ltd

ARTICLE IN PRESS

Notation

De equivalent diameter, mmFL force in longitudinal direction, NFT force in thickness direction, NF W force in width direction, NL length, mmM mass, gMc moisture content, % w.b.Pa projected area, mm2

R2 coefficient of determinationT thickness, mmW width, mma angle of tilt, 1

e porosity, %m static coefficient of frictionrb bulk density, kgm�3

rt true density, kgm�3

f sphericity, %Subscripts

a aluminiumk kerneln nutp plywoodr rubber

S. PLIESTIC ET AL.174

damage depends on mechanical force applied to the nut,rotational speed of the cracker, thickness of shell andshape of the nuts, number of sizing grades and efficiencyof sizing (Ozdemir & Ozilgen, 1997).The objective of this study was to investigate some

physical properties of Croatian domestic and the mostproduced filbert cultivar ‘Istrian long’, named accordingto its origin region, Istria (peninsula in Croatia), andshape. These properties are, namely, dimensions, mass,volume, equivalent diameter, sphericity, bulk density,true density, porosity, projected area, static coefficientof friction on three different surfaces and crackingforces.

Lk

Wk

Tk

Tn

Wn

Ln

FL

FW

FT

Fig. 1. Characteristic dimensions of filbert nut and kernel andcompression directions; Ln, Wn, Tn are the length, width andthickness of the nut; Lk, Wk, Tk are the equivalent dimensions ofthe kernel; FL, FW, FT are the forces applied in three directions

2. Material and methods

The filbert (cv. Istrian long) nuts used in this studywere collected in the 2003 season from an agriculturalcompany ‘Orahovica’, located 200 km east from Zagreb,capital of Croatia (451 320 N, 171 540 E). The sampleswere manually cleaned to remove foreign matter, andbroken and immature nuts. The dimensions, size andsphericity of nuts and kernels were evaluated at initialmoisture content found to be 6�19% wet basis (w.b.).The moisture content of the samples was determined byusing the AOAC official method 925�40 (AOAC, 2002).The gravimetric properties of the nut and kernel wereevaluated as functions of moisture content. To obtainhigher moisture contents than initial, the samples wereprepared by adding a pre-determined quantity ofdistilled water and sealing in separate polyethylenebags. The samples were kept at 5 1C in refrigerator for 1week to enable the moisture to distribute uniformly.Before starting the experiment, the samples were takenout of the refrigerator and allowed to warm up to theroom temperature for 2 hours.

2.1. Dimensions, size and sphericity

To determine the average size of the nut and kernel, asample of 100 nuts was randomly picked and their threemajor dimensions namely length, width and thickness(as shown in Fig. 1) were measured using a micrometerwith accuracy of 0�01mm.

To evaluate the mass, each nut and kernel wereseparately weighed on a precision electronic balancereading to 0�001 g. The volume of each nut and kernelwas calculated using the following equation (Mohsenin,1970):

V ¼ pLWT=6 (1)

ARTICLE IN PRESS

PHYSICAL PROPERTIES 175

where: V is the volume in mm3; L is the length in mm; W

is the width in mm; T is the thickness in mm; andsubscripts n and k distinguish between nuts and kernels,respectively.The equivalent diameter and sphericity were calcu-

lated using the equations (Mohsenin, 1970)

De ¼ ðLWTÞ1=3 (2)

f ¼ 100 De=L (3)

where: De is the equivalent diameter in mm; and f is thesphericity in %.

2.2. Gravimetric properties

The bulk density of nuts and kernels (the ratio ofweight and volume) was determined with a weight-per-hectolitre tester, which was calibrated in kg-per-hecto-litre (Deshpande et al., 1993). The nuts were poured intothe calibrated bucket, up to the top, from a height ofabout 15 cm, and excess nuts and kernels were removedby a strike-off stick (Dutta et al., 1988). No separatemanual compaction of nuts was done. The true densitydefined as the ratio between the mass of nuts and kernelsand its true volume was determined using the toluenedisplacement method. The porosity of nut and kernelwas calculated from the bulk, and true density, using theequation (Mohsenin, 1970)

e ¼ ð1� rb=rtÞ100 (4)

where: e is the porosity in %; rb is the bulk density inkgm�3; and rt is the true density in kgm�3.The projected area of each nut and kernel was

measured by placing it under a thin transparent paperand using a planimeter equipped with a magnifying glass(Makanjuola, 1972). The gravimetric properties definedabove were evaluated at four different moisture contentsthrough 15 replications.

2.3. Frictional properties

The static coefficients of friction of filbert nuts andkernels on three different structural surfaces, namely,aluminium, plywood and rubber, were determined atdifferent moisture contents. The nuts and kernels wereplaced on an adjustable tilting plate with a screw device.The plate was raised slightly until the nuts and kernelsstarted to slide down and the angle of tilt was read froma graduated scale. This was replicated by taking 15samples of nuts and kernels at four different moisturecontents, and the averages were calculated. The coeffi-cient of friction was calculated from the equation (Singh

& Goswami, 1996)

m ¼ tan a (5)

where: m is the coefficient of friction in decimal; and a isthe angle of tilt in degree.

2.4. Cracking forces

To measure the forces required to crack filbert nuts, auniversal testing machine was used to compress the nut.The nut was placed on the fixed plate with dynamometerand pressed with compression plate. The forces weremeasured by the data acquisition system, which includeddynamometer HBM (Hottinger Baldwin Messtechnik,Darmstadt, Germany) with a capacity of 1000N,amplifier HBM DMC 9012A and personal computer.The nuts at four different moisture contents werecompressed in three directions, presented in Fig. 1; FL

is the force applied in the longitudinal direction (X-axis),FW is the force applied transversely in width direction(Y-axis) and FT is the force applied transversely inthickness direction (Z-axis). For each combination ofmoisture content and compression direction, a sample of15 nuts was tested.

3. Results and discussion

3.1. Dimensions, size and sphericity

The average dimensions, size and sphericity of filbertnuts and kernels are given in Table 1. The frequencydistribution curves of filbert nuts are shown in Fig. 2. Inthe sample, about 65% of the nuts had a length in therange of 25–27mm, about 76% had a width in the rangeof 19–21mm and about 80% had a thickness in therange of 17–19mm. The frequency distribution curves offilbert kernels are also shown in Fig. 2. About 66% ofthe kernels had a length in the range of 19–21mm, about66% had a width in the range of 14–16mm and about63% had a thickness in the range of 12–14mm. Thevalues of dimensions and mass of filbert nuts are foundto be higher than that of hazelnut (Aydin, 2002), butlower than that of cashew nut (Balasubramanian, 2001).The filbert kernels are longer and wider than hazelnutkernels (Aydin, 2002); but hazelnut kernels are thickerand have higher sphericity.

3.2. Gravimetric properties

The bulk density of filbert nut and kernel decreasedfrom 530 to 454 kgm�3 and from 649 to 569 kgm�3,respectively, as moisture content increased from 6�19 to

ARTICLE IN PRESS

Table 1

Dimensions and some properties of filbert nut and kernel, at a moisture content of 6.19% w.b. (standard deviation in parentheses)

Length, mm Width, mm Thickness, mm Mass, g Volume, cm3 Equivalentdiameter, mm

Sphericity, %

Nut 25�32 20�54 17�93 3�88 4�88 20�96 82�86(0�88) (0�62) (0�53) (0�37) (0�37) (0�63) (1�75)

Kernel 20�20 14�52 12�64 1�70 1�94 15�41 77�02(0�92) (1�10) (1�03) (0�19) (0�32) (0�70) (3�24)

b,n = 548.039 − 3.355 Mc

R2 = 0.990

�t,n = 926.803 − 3.433 Mc

�b,k = 668.155 − 3.562 Mc

R 2 = 0.991

R 2 = 0.997

�t,k = 1034.792 − 3.444 Mc

R 2 = 0.988

3000 5 10 15 20 25 30 35

400

500

600

700

800

900

1000

1100

Moisture content Mc , % w.b.

Den

sity

�, k

g m

−3�

Fig. 3. Effect of moisture content on filbert nut (subscript n)and kernel (subscript k) bulk density (&) and true density(J); ———, nut; – – –, kernel; subscripts b and t, bulk and

true, respectively; R2, coefficient of determination

340 5 10 15 20 25 30 35

36

38

40

42

44

46

48

50


Poro

sity

�, %

�k = 35.326 + 0.145 Mc

R 2 = 0.997

�n = 40.739 + 0.163 Mc

R 2 = 0.979

Fig. 4. Effect of moisture content on filbert nut (subscript n)and kernel (subscript k) porosity; ———, nut; – – –, kernel; R2,

coefficient of determination

010 12 14 16 18 20 22 24 26 28

10

20

30

40

50

60

Nut and kernel dimension, mm

Num

ber

of n

uts

and

kern

els

Fig. 2. Frequency distribution curves of filbert nut and kerneldimensions: J, length; &, width; W, thickness; ———, nut;

– – –, kernel


28�71% w.b. (Fig. 3.). This was due to the fact that anincrease in mass owing to the moisture gain in nut andkernel was lower than accompanying volumetricexpansion. The true density was found to decreasefrom 907 to 829 kgm�3 for nut, and from 1016 to937 kgm�3 for kernel. The negative linear relationshipof bulk and true density with moisture content was alsoobserved by Visvanathan et al. (1996) for neem nut,Baryeh (2001) for bambara groundnut and Aydin (2002)for hazelnut.The estimated porosity [Eqn (4)] of both nut and

kernel of filbert was found to slightly increase withincrease in moisture content from 6�19% to 28�71%(Fig. 4). The porosity of nut increased from 41�53% to45�24%, while porosity of kernel increased from 36�18%to 39�44%. The porosity values for filbert nut and kernelare found to be similar to that for bambara groundnut(Baryeh, 2001) and arecanut (Kaleemullah & Gunase-kar, 2002), but lower than that for hazelnut (Aydin,2002), neem nut (Visvanathan et al., 1996) and cashewnut (Balasubramanian, 2001).The projected area of filbert nut and kernel increased

from 423 to 497mm2 and from 246 to 283mm2,respectively, with increase in moisture content from6�19% to 28�71% (Fig. 5). Similar trends have beenreported for bambara groundnut (Baryeh, 2001) andhazelnut (Aydin, 2002).

3.3. Frictional properties

The static coefficient of friction of filbert nuts andkernels was determined on three different surfaces and

ARTICLE IN PRESS

Pa,k = 236.048 + 1.687 Mc

R2 = 0.991

Pa,n = 405.288 + 3.222 Mc

R2 = 0.985

1000 5 10 15 20 25 30 35

200

300

400

500

600


Proj

ecte

d ar

ea P

a , m

m2

Fig. 5. Effect of moisture content on filbert nut (subscript n)and kernel (subscript k) projected area; ———, nut; – – –,

kernel; R2, coefficient of determination

0.20 5 10 15 20 25 30 35

0.3

0.4

0.5

0.6

0.7

0.8


Stat

ic c

oeff

icie

nt o

f fr

ictio

n �

Fig. 6. Effect of moisture content on static coefficient of frictionof filbert nut (———) and kernel (– – – ) on aluminium (J),

plywood (&) and rubber (W)

Table 2

Values for the regression coefficients k1 and k2 for the

relationship between the static coefficient of friction and moi-

sture content, together with the coefficient of determination R2

Surface Regression coefficient R2

k1 k2

NutsAluminium 0�179 0�010 0�993Plywood 0�267 0�009 0�994Rubber 0�350 0�010 0�996

KernelsAluminium 0�277 0�007 0�991Plywood 0�356 0�008 0�996Rubber 0�520 0�008 0�989

PHYSICAL PROPERTIES 177

presented in Fig. 6. It is observed that the staticcoefficient of friction increased with an increase inmoisture content on all surfaces, for both nut andkernel. The reason for the increased friction coefficientat higher moisture content may be due to the waterpresent in the nut and kernel offering a cohesive force onthe surface of contact. At all moisture contents, thestatic coefficient of friction for nut and kernel wasgreatest on rubber, followed by plywood, and the leaston aluminium. This may be due to the smoother andmore polished surface of the aluminium sheet than othermaterials used. The static coefficient of friction has alsobeen found to be higher on rubber than on plywood forhazelnut (Aydin, 2002) and higher on plywood than onaluminium for bambara groundnut (Baryeh, 2001).

The relationship between the static coefficient offriction and moisture contents of the nut and kernel canbe represented by the equation

m ¼ k1 þ k2Mc (6)

where: k1 and k2 are regression coefficients with valuesgiven in Table 2 for three surfaces, namely, aluminium,plywood and rubber.

3.4. Cracking forces

The average forces required for cracking filbert nutsin three directions are presented in Fig. 7. For allcompression directions, force necessary to crack the nutdecreased with increasing moisture content. This de-crease in cracking force may be due to the nut becomingsofter at higher moisture contents. At all moisturecontents, the maximum force occurred in the long-itudinal direction FL and the minimum force transver-sely in width direction FW, contrary to the findings ofAydin (2002) for hazelnut, Braga et al. (1999) formacadamia nut and Koyuncu et al. (2004) for walnut.This contradiction could be due to differences in theshape of filbert nut in comparison to hazelnut and factthat filbert nut has no suture, unlike macadamia nut andwalnut. In conclusion, the width direction would berecommended for the design and development ofcracking machines for filbert nuts.

4. Conclusion

(1)
The average length, width, thickness and equivalentdiameter of filbert nuts at 6�19% moisture content(w.b.) were 25�32, 20�54, 17�93 and 20�96mm, whilethe corresponding values of kernels were 20�20,14�52, 12�64, 15�41mm, respectively.

ARTICLE IN PRESS

FL = 720.563 − 6.592 Mc

00 5 10 15 20 25 30 35

100

200

300

400

500

600

700

800


Forc

e F

, N

R2 = 0.973

FT = 440.005 − 4.407 Mc

R2 = 0.980

FW = 326.119 − 4.140 Mc

R2 = 0.946

Fig. 7. Effect of moisture content on forces required to crackingfilbert nuts: the longitudinal direction FL (J), the widthdirection FW (&), and the thickness direction FT (W); R2,

coefficient of determination


(2)
The average unit mass, volume and sphericity ofnuts were 3�88 g, 4�88 cm3 and 82�86%, while thecorresponding values for kernels were 1�70 g,1�94 cm3 and 77�02%, respectively, at a moisturecontent of 6�19% w.b.
(3)
The bulk and true densities of filbert nut and kerneldecreased with increase in moisture content, whileporosity and projected area increased.
(4)
The static coefficient of friction of filbert nut andkernel increased with moisture content. This coeffi-cient was highest on rubber, followed by plywoodand aluminium.
(5)
The maximum force for nut cracking occurred inlongitudinal direction and the minimum forcetransversely in width direction. For all compressiondirections, force necessary for cracking the nutdecreased with increasing moisture content.
(6)
The results of this study could be applied to optimisethe threshing performance, pneumatic conveyingand other transportation processes, storage, and todevelop machines for filbert handling and cracking.
References

AOAC (2002). Official Methods of Analysis, 17th Edn.Association of Official Analytical Chemists, Gaithersburg,Maryland, USA

Aydin C (2002). Physical properties of hazel nuts. BiosystemsEngineering, 82, 297–303

Balasubramanian D (2001). Physical properties of raw cashewnut. Journal of Agricultural Engineering Research, 78,291–297

Baryeh E A (2001). Physical properties of bambara ground-nuts. Journal of Food Engineering, 47, 321–326

Braga G C; Couto S M; Hara T; Almeida Neto J T P (1999).Mechanical behaviour of macadamia nut under compres-sion loading. Journal of Agricultural Engineering Research,72, 239–245

Deshpande S D; Bal S; Ojha T P (1993). Physical properties ofsoybean. Journal of Agricultural Engineering Research, 56,89–98

Dutta S K; Nema V K; Bhardwaj R K (1988). Physicalproperties of gram. Journal of Agricultural EngineeringResearch, 39, 259–268

Jha S N; Prasad S (1993). Physical and thermal properties ofgorgon nut. Journal of Food Process Engineering, 16,

237–245Kaleemullah S; Gunasekar J J (2002). Moisture-dependedphysical properties of arecanut kernels. Biosystems Engi-neering, 82, 331–338

Koyuncu M A; Ekinci K; Savran E (2004). Crackingcharacteristic of walnut. Biosystems Engineering, 87,

305–311Makanjuola G A (1972). A study of some of the physicalproperties of melon seeds. Journal of Agricultural Engineer-ing Research, 17, 128–137

Miljkovic I (1991). Suvremeno vocarstvo [Modern OrchardProduction]. Nakladni zavod Znanje, Zagreb, Croatia

Mohsenin N N (1970). Physical Properties of Plant and AnimalMaterials. Gordon and Breach Science Publisher, NewYork, USA

Olaniyan A M; Oje K (2002). Some aspects of the mechanicalproperties of shea nut. Biosystems Engineering, 81, 413–420

Ozdemir M; Ozilgen M (1997). Comparison of the quality ofhazelnuts unshelled with different sizing and crackingsystems. Journal of Agricultural Engineering Research, 67,219–227

Singh K K; Goswami T K (1996). Physical properties of cuminseed. Journal of Agricultural Engineering Research, 64,

93–98Visvanathan R; Palanisamy P T; Gothandapani L; Sreenar-ayanan V V (1996). Physical properties of neem nut. Journalof Agricultural Engineering Research, 63, 19–26

Documents

Physical Properties of Filbert Nut and Kernel