QUALITY CONTROL CHARTS FOR HAZELNUT UNSHELLING OPERATION

Abstract

Özet Kalite güvence çizgelerinden olan p-çizgeleri, fındık kırma işleminde, 3 ayrı boy Akçakoca fındığında, vurgunlu ve geri-dönen özürleri için çizildi. p-Çizgeleri fındık kırma işleminin izlenmesinde, iyileştirme olasılıklarının belirlenmesinde ve gerekli tedbirlerin alınmasında kullanılabilir. Her boy, randıman ve üretim bölgesi için ayrı bir p-çizgesi çizmek gereklidir. Abstract The p-charts were constructed for 3 different diameter grades of Akçakoca hazelnuts for damaged and recyclable hazelnut units for hazelnut unshelling operation. The p-charts satisfactorily monitored the hazelnut unshelling operation, pointed out possible improvements and necessary precaution to be taken both operation and as maintenance. The results indicate that it may be necessary to construct p-charts of damaged and recyclable hazelnut units for each diameter grades and for different percent kernel and production region.

Full text

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 1
QUALITY CONTROL CHARTS FOR HAZELNUT UNSHELLING
OPERATION
KALİTE GÜVENCE ÇİZGELERİNİN FINDIK KIRMA İŞLEMİNDE
KULLANILMASI
Murat ÖZDEMIR
Özet
Kalite güvence çizgelerinden olan p-çizgeleri, fındık kırma işleminde, 3 ayrı boy
Akçakoca fındığında, vurgunlu ve geri-dönen özürleri için çizildi. p-Çizgeleri fındık kırma
işleminin izlenmesinde, iyileştirme olasılıklarının belirlenmesinde ve gerekli tedbirlerin
alınmasında kullanılabilir. Her boy, randıman ve üretim bölgesi için ayrı bir p-çizgesi çizmek
gereklidir.
Abstract
The p-charts were constructed for 3 different diameter grades of Akçakoca hazelnuts
for damaged and recyclable hazelnut units for hazelnut unshelling operation. The p-charts
satisfactorily monitored the hazelnut unshelling operation, pointed out possible improvements
and necessary precaution to be taken both operation and as maintenance. The results indicate
that it may be necessary to construct p-charts of damaged and recyclable hazelnut units for
each diameter grades and for different percent kernel and production region.

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 2
INTRODUCTION
Wild species of hazelnut, genus Corylus, are distributed in nearly all parts of
temperate zones of the northern hemisphere while the major producers are Turkiye, Italy,
Spain, and USA (Ayfer et al., 1986). Economically important varieties in Turkiye are selected
from C. Avellana var. pontica (common hazelnut), C. maxima mill. (the giant hazelnut), and
C. Colurna var. glandulifera. (Turkish hazelnut) (Ayfer et al., 1986; Kasaplıgil, 1972).
Turkiye produces 400,000-450,000 tons of hazelnut annually. It is 73% of the total hazelnut
production of the world. Turkiye exports 83% of its production, and gets about 750 million
US$ annually (Anonymous, 1995). Almost eight million people are involved into production
and processing of Turkish hazelnut (Akdağ and Öztürk, 1993). Hazelnuts provide a definite
flavour in food products and plays a major role in human nutrition and health (Woodroof,
1967; Woodroof, 1973; Woodroof, 1975; Labell, 1983; Villaroel et al., 1987; Mattson, 1989;
Elvevol et al., 1990; Nicolosi et al, 1990; Mehlenbacher, 1989; Labell, 1992; Sabate et al.,
1993; Kinderlerer and Johnson, 1992; Garcia et al., 1994; Alphan et al., 1996; Pala et al.,
1996).
Hazelnut Unshelling
Hazelnuts are unshelled by applying compression force on while passing through a
clearance that is sufficient to avoid injuring the hazelnut kernels (stone, conical, and roller
unshellers). Traditionally, hazelnuts are unshelled with stone unsheller in Turkiye (Figure 1).
Damage occurring during unshelling is among the major causes of quality reduction in
Turkish hazelnuts (Ayfer, 1984; Sabır, 1988). In Spain, hazelnuts are remoistured from 5-6%
to about 10% prior to unshelling hazelnuts with conical unsheller to minimise damaged and
broken. After unshelling, hazelnut shells are then siphoned by air suction during which
undersize and broken kernels are also separated. The remaining hazelnut units are then passed
over oscillating shaker screens and separators which separates recyclable hazelnuts from the
kernels (Fig. 2). (Özdemir, 1996).
Quality Control Charts in Food Processes
Quality control charts e.i., Shewart and attribute charts, are used for statistically representing
the picture of a process, raw material, product quality in manufacturing industries. They are
real-time feedback to manufacturing operators to detect and give immediate response to any
adverse signal in timely and to maintain the process under statistical control for a given lot
(Hubbard, 1990; Levinson, 1992). They are simple in concept, easy to construct and

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 3
understand for everyone. Shewart control charts consist of means and range charts, and are
used with measured properties (Grant and Leavenworth, 1980; Miller and Freund, 1985;
Hubbard, 1990; Montgomery, 1991). Attribute charts (p-, np-(m-), c- and u-charts) are
commonly used when each unit in a sample set is assessed on the basis of either conforming
or non-conforming according to the pre-determined quality attributes. The p-charts are based
on the normal curve approximation of the binomial distribution model. This procedure is used
when the defective units are distributed around a mean value following the binomial
distribution model, and the shape of this distribution curve resembles to that of the normal
distribution, therefore the properties of the normal distribution model may be used to
construct the p-chart (Özilgen, 1997). The p-charts for a attribute may be constructed using
the following equations 1, 2, 3 (Hubbard, 1990).
Central line CL PTotal number of defectives in the samples
Total number sample units
( ) = =
(1)
Upper control it UCL p p p
n
lim ( ) ( )
= + −








31
(2)
Lower control it LCL p p p
n
lim ( ) ( )
= − −








31
(3)
Where
n
is the average sample size. Central line is actually an approximation of the
population mean value of p. Equations 2 and 3 describes the upper and the lower control
limits, such that the values of p determined in the individual experiments fall between these
limits with about 99.7 % probability (Özilgen, 1997).
In the present study, we will use p-charts to statistically represent fraction of
recyclable and damaged hazelnut units of 3 diameter grades of traditional Turkish unshelling
operation.
MATERIALS AND METHODS
The hazelnuts
Akçakoca hazelnuts were used in the study. The hazelnuts consisted of a mixture of
various local cultivars (Mincane, Karafindik, Foşa and Çakıldak) (Ayfer et al., 1986) and are
supplied from large district, including Bolu, Adapazarı and Samsun provinces (Turkish

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 4
Standards, 1978a). The hazelnuts contained 5 to 6% m.c. (d.b.). The percent kernel (the ratio
of the kernel weight to the whole hazelnut weight) was 48%. This study was conducted in a
commercial hazelnut processing plant.
Analysis of the unshelled hazelnuts
Quality attributes of the hazelnuts after unshelling are depicted in Table 1. These quality
attributes are defined in the Turkish Standards (1978b) except for recyclable which were
defined to be sum of the cracked, left-in-the-shell, and unbroken hazelnut units as a measure
of capacity usage of the unsheller. Recyclable units were recycled to the unsheller
continuously. A kernel may actually be damaged, but be regarded undamaged in the
commercial sense if the diameter of the damaged area is less than 2 mm and its depth is less
than 1 mm. For the diameter grades of 19-20 mm, 17.5-18 mm, 17-17.5 mm, samples of 1 kg
weight were taken from 10 consecutive unshelling operations of acceptable quality level and
analysed for the visible defects as defined in Table 1. Then, the p-charts of the diameter
grades were constructed using Equations 1, 2, 3 for damaged and recyclable hazelnuts.

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 5
Table 1
Quality attributes of the hazelnuts unshelling operation (Turkish Standards, 1978b)
Attributes
Definition
Damaged
kernels with damage greater than 2 mm in diameter, and 1 mm in
depth
Undamaged
kernels with no damage or damage smaller than 2 mm in diameter
Broken
kernels which lost at least one-third of their edible part
Cracked
hazelnuts whose shell was intact with only a minor crack
Left-in-the-shell
kernels, trapped inside a partially broken shell (at least one-sixth of
the shell is broken)
Unbroken
hazelnuts that have a completely unbroken shell
Recyclable1
total percentage that may be recirculated to the unshellers (sum of
the percentages of left-in-the-shell, cracked and unbroken hazelnuts)
1Not defined in Turkish standards (See material and methods section)
RESULTS AND DISCUSSION
The p-charts highlight quality and capacity problems, and standardise the operation.
The operating conditions of the unsheller are determined by the operator of the unshelling
plant. The aim is to have lowest possible damaged hazelnuts with maximum undamaged for
the given diameter grade at a acceptable level recyclable hazelnut units. It depends on
experience of the operator of the hazelnut unshelling plant (Özdemir, 1996; Özdemir and
Özilgen, 1997). So as to monitor the unshelling operation for damaged and recyclable
hazelnuts, fraction of the damaged or the recyclable hazelnuts (values of p) were determined
in each experiment, then the central line, upper and lower control limits of the p charts were
determined by using Equations 1, 2 and 3 respectively (Table 2). LCL of the 19-20 mm and
17.5-18 mm hazelnuts had a negative value, therefore they were set to be zero. Separate p-
charts for the damaged and recyclable hazelnuts were constructed for the 19-20 mm, 17.5-18
mm, and 17-17.5 mm diameter grades as shown in Figs. 3, 4 and 5. The percent variation of
the sample size, calculated by dividing the difference of maximum and minimum sample sizes
to average sample size, was smaller than the recommended value (20%) in all of the trials
(Table 2) (Hubbard, 1990). The unshelling operation is regarded under statistical control to
the given standard, when the experimental data fall between the upper and the lower control
limits of the predetermined quality factors. Figs. 3a, 4a and 5a indicate that the operation was
within statistical control, when the fraction of the damaged hazelnuts are chosen as the quality

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 6
indicator. The average of damaged hazelnuts for 19-20 mm, 18-17.5 mm and 17-17.5 mm
hazelnuts were ranging from about 0 to 1.4%, 2 %, 3 % respectively.
Table 2
CL, UCL and LCL values for the damaged and recyclable units
Diameter (mm)
% Variation
Defects
CL
UCL
LCL
19-20
16.2
Damaged
0.005
0.014
0.0
Recyclable
0.375
0.439
0.312
17.5-18
15.5
Damaged
0.009
0.020
0.0
Recyclable
0.242
0.289
0.195
17-17.5
14.7
Damaged
0.016
0.030
0.003
Recyclable
0.266
0.313
0.220
17-17.51
18.10
Damaged
0.048
0.070
0.026
Recyclable
0.176
0.215
0.137
1 Data gathered from Özdemir and Özilgen (1997)
Moreover, Figs 3b, 4b and 5b shows that the fraction of the recyclable hazelnuts was within
the statistical control when recyclable units was chosen as a quality indicator. But, the
averages of recyclable hazelnuts for 19-20 mm, 18-17.5 mm and 17-17.5 mm hazelnuts were
ranging from 32.2 to 43.3%, 19.5 to 28.9%, and 12% to 31.3% respectively. This means about
1/3 capacity reduction during unshelling operation. Sizing hazelnuts into 20 grades instead of
14 may help to reduce recyclable percentage, and consequently increase the capacity. A
acceptable quality is a compromise between the damaged and recyclable hazelnuts. For
example, increasing recyclable fraction will probably result in lower fraction of damaged at
the expense of reduced capacity, or vice versa. A higher damaged or recyclable hazelnut units
than UCL will require tuning of the unsheller. Fractions of hazelnuts which experience
permanent damaged above UCL may indicate enlarged active unshelling distance in the
stones (Fig. 1), which may occur as a results of stones and pieces of metal found a way into
the feed units (Özdemir, 1996).
Besides unsheller characteristics, amount of damaged hazelnuts depends on the
clearance between the shell and the kernels, thickness of shell and shape of the hazelnuts
(Ayfer, 1973; Nalbant, 1991; Pınar and Beyhan, 1991). The clearance varies from year to
year, and can be related to percent kernel (Özdemir and Özilgen, 1997). Özdemir and
Özilgen (1997) constructed p-charts for Akçakoca hazelnuts with 50-52% kernel for 17-17.5

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 7
mm and 16.5-17 mm hazelnuts. They stated that the p-charts should be used with extreme
caution for different region hazelnuts and different percent kernels.
The samples of 17-17.5 mm hazelnuts with 50-52% kernel were not within statistical
control when they were analysed with p-charts of the same grade but 48% kernel (Fig. 6a).
However, the recyclable hazelnut units of the 17-17.5 mm hazelnuts with 50-52% kernel
were within statistical control when they were analysed with p-charts of the same grade but
48% kernel (Fig. 6b). Recyclable hazelnut fraction may be a characteristic of unsheller, and
may not be effected from the percent kernel as much as the damaged hazelnuts. Nevertheless,
it may be necessary to construct different p-charts of damaged and recyclable for each
diameter grades and kernel percent, and for different production region.
The charts can be used by taking 1 kg sample from the shaker/separator inlet (Fig. 2)
after adjusting the unsheller system to the new diameter conditions, and analysing the sample
in terms of damaged and recyclable hazelnuts. If fraction of the attributes of the sample are
within the UCL and LCL of the chart, the operation is within the statistical control for the
given diameter grade.
CONCLUSION
The p-charts satisfactorily describes hazelnut unshelling operation. They enable to
standardise the process and give immediate responses to the changes in product quality.
However, it may be necessary to construct p-charts of damaged and recyclable hazelnut units
for each diameter grades and kernel percent, and for different production region.
REFERENCES
Alphan, E.M. Yılmaz, T.M. Pala, M. and Açkurt, F. 1996. Tekli Doymamış Yağ Asidi İçeriği
Fındıkla Zenginleştirilmiş Diyetin Karbonhidrat ve Lipid Metabolizmasına Etkisi.
Gıda Teknolojisi, 1, 58-63.
Anonymous. 1995. Fındık Ekonomik Raporu. Fiskobirlik, Giresun.
Ayfer, M. 1973. İç Fındıklarda Gizli Vurgun Üzerine Bir Araştırma. Annual of Faculty of
Agriculture of Ankara University, 23(3), pp. 269-284.
Akdağ, Z. and Öztürk, İ. 1993. Meyve-Sebze Işleme Sanayi Özel İhtisas Komisyonu Fındık
İşleme Sanayi Alt Komisyon Raporu. In “Yedinci Kalkınma Planı,” TC DPT,
Ankara.
Ayfer, M. 1984. Dünya’da ve Türkiye’de Fındık. In “Türkiye Ekonomisinde Fındığın Yeri ve
Önemi,” İktisadi Araştırma Vakfı, İstanbul, pp. 33-55.

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 8
Ayfer, M. Uzun, A. Baş, F. 1986. Türk Fındık Çeşitleri. Black Sea Hazelnut Exporters Union,
Ankara.
Elvevol, E.O. Moen, P. Olsen, R.L. and Brox, J. 1990. Some Possible Effects of Dietary
Monounsaturated Fatty Acids in Cardiovascular Disease. Artheriosclerosis, 81, 71-4.
Garcia, J.M. Agar, I.T. and Streif, J. 1994. Lipid Characteristics of Kernels from Different
Hazelnut Varieties, Tr. J. Agric. and Forestry, 18, 199-202.
Grant, E.L. and Leavenworth, R.S. 1980. Statistical Quality Control. 5th ed., Singapore,
Mcgraw-Hill, pp. 1-29.
Hubbard, R.H. 1990. Statistical Quality Control for the Food Industry. Von Norstrand
Reinhold, New York, pp. 29-51.
Kasaplıgil, L. 1972. A Bibliography on Corylus Betulaceae. 63. Ann. Rep. Of the Northern
Nut Growers Associations.
Kinderlerer, J.L. and Johnson, S. 1992. Rancidity in Hazelnuts Due to Volatile Aliphatic
Aldehydes. J. Sci. Food. Agric., 58, 89-93.
Labell, F.M. 1983. Hazelnut Paste Provides Sweet, Delicate Flavour. Food Processing USA,
44, 80.
Labell, F.M. 1992. Hazelnuts Supply Flavour and Crunch, Food Processing USA, 53, 52, 54.
Levinson, W.A. 1992. Make the most of Control Charts, Chemical Engineering Progress, 86-
91.
Mattson, F.H. 1989. A Changing Role for Dietary Monounsaturated Fatty Acids. J. of the
American Dietetic Association, 89, 387-391.
Mehlenbacher, S.A. 1989. Hazelnuts, in Genetic Resources of Temperate Fruit and Nut
Crops. Acta Horticulture 290, International Society for Horticultural Science.
Miller, E. and Freund, J.E: 1985. Probability and Statistics for Engineers. Ed., E. Cliffs, 3rd
Edn., Prentice-Hall Inc., pp.426-431.
Montgomery, C.C. 1991. Introduction to Statistical Quality Control. 2th Ed., John Wiley &
Sons, pp. 101-146.
Nalbant, M. 1991. Fındık Kabuğunun Sıkıştırma ve Çarptırma Yöntemiyle Kırılması. Annual
of Samsun 19th May University, pp. 115-130.
Nicolosi, R.J. Stucchi, A.F. Kewal, M.C. Hensesig, L.K. Hegstein, P.M. and Schefer, E.J.
1990. Effect of Dietary Fat Saturation and Cholesterol on LDL Composition and
Metabolism. Arterisclerosis, 10, 119-128.
Özdemir, M. 1996. Statistical Assessment of Quality Parameters of Turkish Akçakoca
Hazelnuts During Cracking Operation, MS thesis (Unpublished), METU, Ankara.

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 9
Özdemir, M and Özilgen, M. 1997. Comparison of the Quality of Hazelnuts Unshelled with
Different Sizing and Cracking Systems. Accepted for publication in J. Agric. Engng.
Res.
Özilgen, M. 1997. Chemical Engineering Applications for Food Process Modelling and
Control, Gordon and Breach Publishing Group UK. (in press).
Pala, M. Açkurt, F. Löker, M. Yıldız, M. and Ömeroğlu, S. 1996. Fındık Çeşitlerinin
Bileşimi ve Beslenme Fizyolojisi Açısından Değerlendirilmesi. Tr. J. of Agriculture
and Forestry, 20, 43-48.
Pınar, Y. and Beyhan, M.A. 1991. Samsun and Ordu Yöresinde Fındık Tarımının
Mekanizasyon Durumu. Annual of University of 19th May, Samsun, pp. 99-114.
Sabate, J. Fraser, G.E., Burke, K., Knutsen, S.F. Bennett, H. and Lindsted, K.D. 1993. Effects
of Walnuts on Serum Lipid Levels and Blood Pressure in Normal Men. The New
England J. Medicine. 328, 603-607.
Sabır, S. 1988. Türkiye’de Fındık Pazarlaması İhracat. In “Türkiye’ de Fındık Politikasının
Esasları,” İktisadi Kalkınma Vakfı, İstanbul, pp. 103-116.
Turkish Standard. 1978a. Unshelled Hazelnut TS 3074. Institute for Turkish Standards,
Ankara.
Turkish Standard. 1978b. Shelled Hazelnut TS 3075. Institute for Turkish Standards, Ankara.
Villarroel, M. Biolley, E. Schneeberger, R. Ballester, D. Ramirez, S. 1987. Amino acid
composition of Chilean hazelnuts. Food Chemistry, 25, 155-158.
Woodroof, J.G. 1967. Tree Nuts, Production, Processing, Products. Avi Publishing Company
Inc., Westport, Connecticut, Vol. 1, pp. 277-312.
Woodroof, J.G. 1973. Peanuts, Production, Processing, Products. 2nd Ed., Avi Publishing
Company Inc., Westport, Connecticut.
Woodroof, J.G. 1975. Tree Nuts, Production, Processing, Products. Vol. 1. 2nd Ed.,. Avi
Publishing Company Inc., Westport, Connecticut.

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 10
Figure 1: Stone unsheller (Adapted from Özdemir and Özilgen, 1997)
Figure 2: Hazelnut unshelling operation
stationary upper cone
hazelnut
output
rotating lower cone
active cracking region
hazelnut input
Kernels
Further Processing
Packaging and Storage
Hand Picking/
Electronic Sorting
Shaker
Recyclable
Broken/
Undersize
Sizing (4 to 20 grades)
Unshelling
Pneaumatic Separation
Shells

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 11
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(a)
0.275
0.325
0.375
0.425
0.475
12345678910
Sample number
p
UCL
CL
LCL
(b)
Figure 3 : P - chart for the (a) damaged (b) recyclable 19-20 mm hazelnuts
0.18
0.20
0.22
0.24
0.26
0.28
0.30
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(b)
0
0.005
0.01
0.015
0.02
0.025
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(a)
Figure 4 : P - chart for the (a) damaged (b) recyclable 17.5-18 mm hazelnuts
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(a)
0.20
0.22
0.24
0.26
0.28
0.30
0.32
0.34
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(b)
Figure 5 : P - chart for the (a) damaged (b) recyclable 17-17.5- mm hazelnuts

Basımı - Published in: Gıda Teknolojisi 2 (11), 56-62
Sayfa: 12
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(a)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
1 2 3 4 5 6 7 8 9 10
Sample number
p
UCL
CL
LCL
(b)
Figure 6 : P - chart for the (a) damaged (b) recyclable 17-17.5- mm hazelnuts (CL, UCL, LCL is
of 48% kernel, samples are from 50-52% kernels (data of 50-52 % k ernels were gathered from
Özdemir and Özilgen (1997))