It .3

:5 1%.}:
vs. 2. l2, ‘0.
f I 61...... :2

3. "39%.

a...

 

 

. twin

.-
Law."

 

2
am»?

I mi. 1!

ﬂaw":

.1
.
.u. . .10 .
L23 .nmela. 33.9.,“ n
rvw

. '2. i:

2.: “mos...

t!
a

n. 49......“ x35. .
{31.45: “am: 2
4.32:7}?! (I.

3.7“"! 9: vs
gun“. .67..

 

t "WWI.

....._N.....r%.5.z..§.x...!
at... 2.5
at; ‘h-

a .w.

9%.va a. n

‘1 I .-

o-
.!

 

.Eﬂkgég , , egiﬁcé . ﬁsmfwmimw
.. . . 9.... ...r..,.?...r,..

 

Was

 

2 0 <5 3 ____
LIBRARY .

Etchigaét $‘i‘i2‘:

University

w-v

 

 

This is to certify that the

thesis entitled

Human Sensory Firmness Scale Based on Gelatin Gels

presented by

Robyn Lynn Reynolds

has been accepted towards fulﬁllment
of the requirements for

 

 

M. S . degree in Food Science

JZmJ

Major professor ﬂ

Date 2/74/0’

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDuo.p65—p. 1 5

 

 

HUMAN SENSORY FIRMNESS SCALE BASED ON GELATIN GELS
By

Robyn Lynn Reynolds

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Food Science and Human Nutrition

2001

ABSTRACT
HUMAN SENSORY FIRMNESS SCALE BASED ON GELATIN GELS
By

Robyn Lynn Reynolds

A series of gelatin gels were produced as ﬁrmness analogues of typical
meat samples for use in texture evaluation. Gels of varying concentrations of
gelatin and sorbitol were made to resemble core samples of meat. The maximum
force required to cut through the gel sample was determined with a Wamer-
Bratzler Shear, and a TA.TX2 Model Texture Analyzer equipped with a shear
blade attachment. Samples were scaled with shear force values. An untrained
sensory panel of 30 people and a trained panel of 12 conﬁrmed the sensory scale
using a triangle test and unstructured scales. Untrained panelists were able to
distinguish between three concentration levels of gels representing three
tenderness categories (p S 0.001). Signiﬁcant differences were detected by trained
panelists between six scaled standards. 'Changes in temperature Signiﬁcantly
affected ﬁrmness; therefore, the sample temperature must be set and maintained
during tests. Results suggest that scaled ﬁrmness standards based on gelatin
provide a good model of meat tenderness. Gelatin gels may also be used to

develop ﬁrmness scales for testing many additional food products.

DEDICATION

To my loving ﬁancé, Douglas Jones,

who supported me and encouraged me across the miles.

iii

ACKNOWLEDGEMENTS

I would like to thank Dr. James Steffe for his endless encouragement and
relentless support throughout my graduate career. Dr. Steffe has proven to be a
role model not only through research, but in life’s lessons as well.

I would also like to thank Dr. Kirk Dolan, Dr. Janice Harte and Dr.
ChristOpher Daubert for serving on my committee. Many thanks to the Michigan
State University Meat Laboratories, MS. Carrie Wing and Mr. Richard Wolthuis
for their technical support throughout the course of this study. Thanks also to all
of the trained panelists who sacriﬁced their time for this study.

Most importantly, I would like to thank my parents, Mr. and Mrs. Keith
Reynolds, who have always believed that I could accomplish anything I set out to

do.

Table of Contents

List of Tables
List of Figures
Introduction

1. Literature Review
1.1. Food Gels and Texture
1.2. Sensory Scale Development
1.3. Textural Parameters
1.3.1 . Characterizing Gel Firmness
1.3.2. Evaluating Meat Texture
1.4. Mechanical and Sensory Correlations

2. Materials and Methods

2.1. Procedure for Making Gelatin Gels
2.1.1. Formulations
2.1.2. Sample Preparation

2.2. Characterization of F irmness
2.2.1. Compression Testing and Modulus Calculation
2.2.2. Shear Force Determination
2.2.3. Temperature Effects
2.2.4. Gelatin Brand Comparison

2.3. Sensory Panel Evaluation
2.3.1. Difference Testing
2.3.2. Trained Panelists

3. Results and Discussion
3.1. Compression Testing
3.2. Evaluation of Shear Forces
3.3. Temperature Effects on Gelatin Gels
3.4. Comparison of Brands
3.5 Difference Testing
3.6 Trained Sensory Panel
3 .7 Practical Recommendations

4. Summary

References

Page
vi

vii

35
38
44
49
52
55
61

63

65

Figure
1.1.
2.1.

2.2.

2.3.

2.4.

2.5.

2.6.

3.1.

3.2.

3.3

3.4

3.5

3.6

List of Figures

(Images in this thesis are presented in color.)

Triangular blade of Warner-Bratzler Shear.

Coring gelatin samples using a mounted 8-inch drill press.

Warner-Bratzler Shear instrument with partially sheared
sample.

TA.XT2 Warner-Bratzler blade attachment with gelatin
sample loaded for testing.

Picture and dimensions of TA.XT2 and Warner-Bratzler
blades.

Sensory evaluation booth with red lighting to mask color
differences.

Example of tray served to panel participants.

Average shear force values vs. gelatin concentration for
the Wamer—Bratzler shear instrument and the TA.XT2.

Warner-Bratzler and TA.XT2 correlation.

Effects of temperature on the shear force measurement of
22.5% and 35.0% gelatin gels.

Average Shear force values vs. test temperature for 22.5%
and 35.0% gelatin concentrations.

Average shear force values for the three brands of gelatin
at Six different concentrations.

Trained panel results for scaling of gelatin gels.

vi

Page
16

24

26

27

28

31
32

40

41

45

46

51

6O

Table

1.1.

1.2.

2.1.

2.2.

3.1.

3.2.

3.3.

3.4.

3.5.

3.6.

3.7.

3.8.

3.9.

3.10.

List of Tables

Szczesniak’s original standard hardness scale.

Modiﬁed hardness reference scale.

Gelatin speciﬁcations of the three brands evaluated.
Formulations for gelatin concentrations.

Determination of secant modulus values at 15% strain.
Average shear force values for gels of different gelatin
concentrations evaluated on the Warner-Bratzler and the

TA.XT2.

Comparison of slope values to demonstrate effects of
temperature.

Comparison of average Warner-Bratzler shear force values
of three brands of gelatin.

Determining sensory parameters for difference testing
from shear force results.

Correct responses from difference tests.

Results of the ﬁrst unstructured scale test.
ANOVA results for ﬁrst unstructured scale test.
Results of the second unstructured scale test.

ANOVA results for second unstructured scale test.

vii

Page

11
21
22
37

39

50

53

54
56
57
58

59

INTRODUCTION

Perception of food texture plays a key role in human evaluation of food
products. Although there are several deﬁnitions of the term “texture,” the
International Organization for Standardization deﬁnes texture as, “all the
mechanical, geometrical and surface attributes of a product perceptible by means
of mechanical, tactile and, where appropriate, visual and auditory receptors”
(1995). In the early years of food evaluation, rheologists, who studied the ﬂow
and deformation of food products, measured texture properties. However, texture
properties can be so complex, human sensory evaluation is necessary to explain
the overall aspects of texture only experienced by humans.

Food texture is important in product development, quality control, sensory
testing, and process engineering. Fortunately, there have been efforts to fully
characterize the texture of many food products with methods such as Texture
Proﬁle Analysis (TPA). Knowing how to effectively characterize the texture of a
food product allows one to constantly improve formulations, processing methods,
product stability, and shelf life. Sensory panels along with mechanical evaluation
of food products have led to an advanced understanding of textural properties.
Panels are able to develop terminology necessary to describe certain textural
attributes unique to a Speciﬁc product. Instruments are then used to generate data

that reﬂect levels of intensity for each textural attribute.

Sensory scales, such as those for hardness and brittleness, have been used
to train panelists on intensities of textural attributes and for comparison of food
products. All of these scales are composed of a variety of food products and
designate the sample Size, brand, and serving temperature for each scale item. In
some studies, these Speciﬁc food items are either difﬁcult to obtain, due to local
availability, or time-consuming to prepare and characterize due to Size differences.
Modiﬁcations and improvements to these standard scales can further advance the
knowledge of texture evaluation and improve the accuracy of scaled food items.

Model materials, such as gelling agents, can be used to represent the texture
of food products when evaluating changes in texture due to differences in
formulation, processing, or temperature. Scaled standards covering a desired
range of intensities and comprised of one constituent, or model material, can help
decrease the variation among scaled food items. Such standards could be applied
to the training of sensory panelists, meat tenderness evaluation, instrument

calibration, and dental research.

Given the above considerations, the objectives of this research are:

> to develop a procedure for producing a series of gelatin gels varying
in ﬁrmness;

> to characterize ﬁrmness by measuring rheological/textural behavior;

> to scale ﬁrmness standards for application to meat tenderness, and

verify the scale using sensory evaluation;

‘x
\\

to develop practical recommendations for using gelatin gels as

ﬁrmness standards.

Cr

1. Literature Review
1.1. Food Gels and Texture

Determining the texture of food is a complicated process because it
involves both sensory and instrumental measurements, and their complex
relationships. Measurement of food texture must take into consideration the
physiology and psychology of human perception, the physical and chemical
components of food structure, and food behavior when subjected to large
deformation. In addition, visual, tactile and auditory stimuli inﬂuence the texture
perception of a particular food before it even enters the mouth.

Sensory scientists, nutritionists, and product developers, alike are
increasingly interested in relating perceived food texture to measurable physical
and mechanical properties. Some foods, such as surimi and meats (Hamann and
Lanier, 1987), have been studied to ﬁnd correlations between subjective and
objective measurements. Food gels have provided much insight into the perceived
texture of fabricated foods. There are three types of gel structures that represent
all levels of molecular complexity existing in food products (Morris, 1986): single
component gels are in the simplest form, mixed gels are models for more complex
structures, and ﬁlled gels may be used to assess the role of particulates in food
systems. Gels, used as model foods, can provide good descriptions of rheological
and mechanicalbehavior of numerous food systems.

There have been many studies where gelling agents were used to determine

certain attributes of food texture. Agar has been widely used to model gelling

Sl

CC

34

systems (Marrs, 1997). Agar has the ability to retain and support its Shape at low
concentrations, and can have a very rigid structure with good elastic properties
(Matsuhashi, 1990). Carrageenan has been found to be very versatile in the area
of stabilization, where texture properties such as thickening and creaminess may
be enhanced. For example, carrageenan can be added to instant beverage powders
so particles are suspended when milk is added (Stanley, 1990). Pectin is another
alternative for modeling the texture of food systems. The number of esters present
in pectin correlates with the ﬁrmness of the manufactured gel. If a soft, spreadable
jam product is desired, low-ester pectin is normally used in the gelation process
(Rolin and De Vries, 1990). Gelling agents have the properties necessary for
providing valuable information about food texture.

Recent studies involving food gels have focused on the physical properties
of gellan gum, gelatin, and mixed polymer gels. Tang et al. (1997) measured the
stress-strain relationships for gellan gum in tension, compression and torsion to
fully understand how gels form three-dimensional networks, and to predict their
functionality within a food product. Food materials are usually subjected to large
deformations during manufacturing and consumption. Lelievre et a1, (1992)
subjected gellan gum samples to large deformations, similar to conditions found in
normal processing, to determine failure characteristics. Furthermore, gelling
agents, such as gelatin and carrageenan, have been used in different concentrations

to study the ﬂavor release of aromatic compounds in model gel systems (Guinard

and Marty, 1995). It was found that the ﬁrmness of gels affects the rate of ﬂavor
release by entrapping or binding the molecules in gelling agents.

Gelatin, which is the gelling agent examined in this research, is the most
commonly used gelling agent in the food industry. This material has good
‘ nutritional value, a wide range of application, and is relatively inexpensive when
purchased in bulk. Gelatin gels are thermo-reversible, elastic in texture, and do
not need the presence of other reagents such as sucrose or salts for gel formation
(Johnston-Banks, 1990). Gelatin is able to form gels due its triple helical structure
and its ability to immobilize water within complex protein chains. Because of the
unique texture and sensory characteristics of gelatin, there have been some studies
correlating instrumental measurements with standardized sensory data (Johnston-
Banks, 1990). These studies indicate that gelatin gels may be superior to other gel
systems in the areas of ﬁrmness, cohesiveness, and elasticity. Dental researchers
have also used gelatin gels as elastic food models to characterize the variability of
the masticatory process of chewing (Lassauzay et al., 2000). Overall, gelatin has
the textural characteristics necessary for modeling sensory properties of food

systems.

1 .2 Sensory Scale Development
Scaling has been used in a variety of industries for many different purposes.
Ratio scales for parameters such as brightness and loudness were used to study the

effects of excitation at the physiological level (Stevens and Galanter, 1957).

Discrimination between scale levels appears to be based on an additive mechanism
of physiological excitation. Physical properties, such as hardness and
extensibility, of materials like metal and rubber may be measured and compared
by using existing scaling methods. In early food texture research, Raffensperger et
al. (1956) worked on developing a scale for grading toughness and tenderness in
beef in relation to consumer preference. Researchers have evaluated and scaled
attributes assessed during the mastication process such as cohesiveness of mass,
and moisture adsorption (Munoz, 1986). However, hardness and fracturability of
food materials are the most studied and applied standards in the food industry.
Texture scales for food were originally developed as a result of the “lack of
an adequate bridge between theoretical rheology and practical applications”
(Szczesniak, 1963). Szczesniak was the ﬁrst to establish a common frame of
reference for textural properties by developing a nomenclature to describe textural
qualities, and deﬁning terms such as texture and consistency. The primary
purpose of developing standard scales was to start on a common ground so that all
sensory panelists would understand the basic terms before applying the scales to
food products. Szczesniak’s Six standard scales were comprised of perceptual
points that were equidistant from each other and increased in the order of intensity
(Szczesniak, 1963). Scales for properties, such as hardness, brittleness, and
cohesiveness, provided a well-deﬁned basis for the development and
implementation of the Texture Proﬁle Method. Table 1.1 displays the standard

materials selected to represent the scale values for hardness.

Table 1.1 Szczesniak’s original standard hardness scale (1963).

 

 

Panel Product Brand or «Manufacturer Sample Temp.
Rating Type Size
1 Cream Cheese Philadelphia Kraft Foods ‘/2" 45-550F
2 Egg White Hard-cooked N/A V2" tip Room
(5 min)
3 Frankfurters Large, Mogen David V2" 50-65°F
uncooked, Kosher Meat
skinless Products Corp.
4 Cheese Yellow, Kraft Foods ‘/2" 50-65°F
American,
pasteurized
process
5 Olives Exquisite, Cresca Co. 1 olive 50-65°F
giant size,
stuffed
6 Peanuts Cocktail type Planters Peanuts l nut Room

in vacuum tin

7 Carrots Uncooked, N/A 1/2" Room
fresh

8 Peanut Brittle Candy part Kraﬁ Foods N/A Room

9 Rock Candy N/A Dryden & N/A Room
Palmer

 

During the development of the standard scales, several terms were
introduced and deﬁned (Munoz, 1986). A rating scale is a series of intervals used
for the perceived intensity of a sensory stimulus, whereas a scale value is the
number that describes the speciﬁc location of a stimulus material over the allotted
range of intensities. The reference material is carefully selected to represent a
speciﬁc sensory attribute and its intensity. In the early stages of scale
development, groups ﬁrst discussed and suggested certain food items that
represented speciﬁc texture characteristics. Products were evaluated and reference
materials were further considered or eliminated. When the panel felt as though a
material represented a speciﬁc intensity, the product sample was accepted as part
of the scale and subsequently used as a common point of reference.

Scales have proven effective for training panelists for the Texture Proﬁle
Method because there is good correlation with the instrumental measurements
using scaled food items (Szczesniak et al., 1963). The standard scales have also
been used for the preliminary screening of panelists, establishing deﬁnitions of
sensory terms, evaluating instrumental methods, and for the selection of
experimental samples. Standard scales have provided a starting point for
researchers to address their speciﬁc needs and interests in sensory research.

There have been a few modiﬁcations of Szczesniak’s original standard
scales despite assumptions that the published examples were to be adhered to
exactly. Cardello et al. (1982) established ratio scale analogues to the standard

scales to show differences in scaling techniques. Cardello et a1. (1982) also

5U

SC

dz"

ch

If

cl.

It)

substituted several references for speciﬁc standards to compensate for current and
local unavailability of recommended food items. In addition, standard texture
scales were modiﬁed to accommodate the conditions in Columbia (Boume et.al.,
1975). Some sensory data collected using the original scales showed a large
degree of variation among trained panelists (Munoz, 1986). Munoz replaced
certain standards from the existing scales with new reference materials and
developed new scales to reduce the amount of variability among panelists. Table
1.2 displays the modiﬁed scale for hardness covering the entire intensity range of
hardness evaluated by molar compression.

The ﬂexibility of scales must be maintained due to the constantly
changing food industry, importance of texture properties of food products and the
increased interest in this area. Certain brands of food products may need to be
replaced occasionally if there are formulation modiﬁcations or processing
changes, both of which affect texture, thus altering the scale values. Researchers
and sensory scientists need to be aware that changes are expected as the ﬁeld of
texture evaluation develops. Improving scales by modifying components allows
for the generation of more reliable sensory data and enhanced understanding of the

texture properties of food.

10

Table 1.2 Modiﬁed Hardness Reference Scale (Munoz, 1986).

 

 

Scale Product Brand or Manufacturer Sample Temp.
Value ‘ Type Size
1.0 Cream Cheese Philadelphia Kraft l/2” 40-45°F
cube
2.5 Egg White Hard-cooked N/A %" Room
(5 min) cube
4.5 American Yellow, Land O Lakes 1/2" 40-45°F
Cheese pasteurized cube
6.0 Olive Stuffed, Goya Foods 1 piece Room
Spanish type,
pimento
removed
7.0 Frankfurter Beef Franks, Hebrew National ‘/2" slice Room
cooked 5 min. Kosher Foods
in boiling
water
9.5 Peanut Cocktail type Nabisco Brands 1 piece Room
in vacuum tin
1 1.0 Carrot* Uncooked, N/ A V2" Room
fresh, slice
unpeeled
11.0 Almond Planter, Nabisco Brands -1 piece Room
Shelled
14.5 Rock Candy Life Savers Nabisco Brands 1 piece Room

 

,*Area compressed with molars is parallel to cut

11

1.3 Textural Parameters

Food technologists involved in product development and quality assurance
are interested in the mechanical properties of gels in relation to shaping, stand-up,
handling, cutting, slicing or eating characteristics (Smewing, 1999). Evaluating
hardness, or ﬁrmness, of model gel systems can aid in the overall characterization
of these properties. According to Szczesniak (1963), hardness is a primary term in
texture classiﬁcation and terms such as “soft,” “ﬁrm,” and “hard” can be used by
consumers to describe degrees of hardness. The terms “hardness” and “ﬁrmness”
are sometimes not well deﬁned in research studies and are most often used
interchangeably. The most commonly used deﬁnition of hardness is the force
required for a speciﬁed degree of deformation.

In a study investigating how consumers evaluate ﬁrmness using non-oral
sensory methods, Szczesniak and Boume (1969) deﬁned ﬁrmness as the force
required to compress a product to a standard distance. Sherman (1969) proposed
the term “ﬁrmness” instead of “hardness” when referring to a deformation even
though it was already incorporated in Szczesniak’s standard scale for hardness.
Munoz et al. (1986) deﬁned ﬁrmness as “the textural property manifested by a
high resistance to deformation by applied force.” She evaluated the ﬁrmness of
gelatin gels varying in concentration using manual shear, oral shear, and
compression tests. Gels made of low concentrations of gelatin (22 to 45 g/L)
along with ﬂavoring agents showed that yield and maximum forces increased with

increasing concentration of gelatin. Henry et a1. (1971) evaluated the ﬁrmness of

12

commercial desserts, whipped toppings, and marshmallow creme as part of the
Texture Proﬁle Analysis (TPA). Overall, it is important to establish a precise
deﬁnition of the parameter being assessed, in addition to the method of

measurement, so that results can be clearly understood and easily compared.

1.3.1 Characterizing Gel F irmness

There are several ways to characterize the deformation of a food product.
Properties should be measured differently depending on where the food product
falls within the range of hardness. For example, when assessing whole meat
texture, it is logical to measure the force to cut through or break the ﬁbers of
muscle (Lyon and Lyon, 1998). This type of large deformation is just one of
many ways to measure the ﬁrmness of a product. In the area of food gels, there is
little information on their fundamental behavior during large deformation or
failure tests. Deformation parameters are usually measured by constant-speed
experiments such as uniaxial compression, uniaxial tension, and three-point
bending. In most food products, the stress-strain relationship is not constant
during large deformation tests (Smewing, 1999). The same holds true in the case
of gels where the stress usually reﬂects the ﬁrmness of the gel. Compression tests
are most commonly used for evaluating gel properties; yet, tensile tests give a
clearer picture of the stresses in a sample since shear stresses in a tensile test are
negligible. Unfortunately, there are many difﬁculties when performing tensile

tests on gels: samples may not be self-supported, samples cannot be mounted

l3

US;

Ih.

properly in the apparatus, and cracks in the sample can lead to failure at small

strains (Lelievre et al, 1992).

1.3.2 Evaluating Meat Texture

The term “tenderness” is most often used when referring to the texture of
meat products. Tenderness of meat has been a major focus of research for many
years, and information surrounding the cause and effects of tenderness continue to
be vital to the meat industry. Consumers evaluate the texture of meat by biting
through pieces with their front teeth and grinding them with their molars (Boyar
and Kilcast, 1986). Boleman et al. (1997) proved that consumers are able to
distinguish between degrees of tenderness and are willing to pay a premium for
more tender products. Although there have been a number of instruments
(Texturometer, Tenderometer, Wamer—Bratzler Shear) developed for direct
measurement of meat tenderness, the lack of uniformity among preparing, cooking
and evaluating samples has made it difﬁcult to compare studies from different
laboratories. Recently, there have been efforts to standardize the procedures for
evaluating meat tenderness (Wheeler et al., 1997). With improved methods of
evaluation, there will most likely be an increase in consistency among
investigators for comparative evaluation.

Many instrumental methods (physical, chemical, and structural) have been
Used over the years to determine the tenderness of meat products (Chrystall, 1994).

The most widely used methods to determine hardness, ﬁrmness, and tenderness

14

have been physical tests involving a Shearing device such as the Wamer-Bratzler
Shear (Chand, 1986). Figure 1.1 illustrates the Shearing principle of the Wamer-
Bratzler blade. The cylindrical meat sample, either V2" or 1" in diameter, is
compressed by the descending anvil. During the test, the sample changes cross-
sectional shape to conform to the shape of the hole in the blade, then ﬁlls in all the
available area, and is sheared across the blade (Boume, 1982). The sample
experiences a complex stress pattern due to the combination of tension,
compression, and shearing forces present during deformation. A force transducer
measures the maximum force required to cut through the sample. Many factors
may lead to variation of recorded shear force values: the width of the blades and
the position of the triangle; the speed of the test; and the Shape, mass, and
orientation of the test samples (Lyon and Lyon, 1998).

The Wamer-Bratzler Shear and subsequent adaptations of this instrument
have been used in numerous studies dealing with meat tenderness. Wheeler et al.
(1996) studied tenderness differences in meat that was thawed at different
temperatures. Results showed that thawing steaks to a consistent temperature (3 to
6°C) yielded the best results because it prevented protein hardening during long
cooking times. In addition, they found that coring the sample parallel to the
longitudinal orientation of the muscle ﬁbers increased repeatability. Lyon and
Lyon (1998) evaluated three shear test devices to see if information obtained in a
Similar way could be interchangeable. They looked at the differences between the

Warner-Bratzler Shear (BT-WB), the Wamer—Bratzler blade attachment (TA-WB)

15

 

Figure 1.1. Triangular Blade of Warner-Bratzler Shear
(Similar to Figure 19 shown in Bourne, 1982)

and a 45° chisel-end blade attachment (TA-WD) to the Model TA.XT2 texture
analyzer. Results Showed that the devices varied in their measured values;
however, no signiﬁcant difference (P > 0.05) was found between the BT-WB and
TA—WB. Sensory panel results corresponding to Wamer-Bratzler Shear readings
have a correlation in the range of 0.6 to 0.85 (Greaser and Pearson, 1999). In
addition, meat tenderness was related to ﬁrmness in sensory terminology in a
study by Rajalakshmi et a1. (1987). A strong relationship (R2 = 0.97) was found
between ﬁrmness and toughness in mutton when evaluated as part of descriptive
analysis. In that study, ﬁrmness was deﬁned as “the textural property manifested
by a high resistance to deformation by applied force” and was recorded after the
ﬁrst few bites. Overall, results from most studies using the Wamer—Bratzler Shear
suggest the need for more standardization of testing procedures, and the

development of calibration materials to reduce variation in shear force

measurements (Wheeler et al., 1996).

1.4 Mechanical and Sensory Correlations

Although no mechanical test can completely replace the complexity of
human sensory evaluations, there have been many useful studies comparing
mechanical and sensory results. Sensory and instrumental tests may be carried out
Simultaneously so that correlations can be made between these different evaluation
methods. Research has shown that a linear relationship often exists between

Sensory and instrumental tests; however, when non-linear relationships are found,

17

._}._1
'.\‘\r

v
.p‘yv

Mn ‘\

“-41.

'13.":
51““
la

.‘g
40‘

M

[L‘r
l.‘

“A“
"51‘

‘5
WT";
‘u

they are not always explained (Szczesniak, 1987). Many imitative tests have been
developed to mimic mastication. Each of these instruments resembled the
chewing motion of food and measured stress and/or strain outputs. For instance,
Proctor et al. (1955) created a strain-gauge tenderometer, which incorporated
human dentures that rocked back and forth to Simulate human jaw movement and
corresponding strains. To obtain sound statistical correlations when performing
these empirical tests, factors such as the geometry and position of the sensors and
the chewing motion must be carefully considered.

The most recognized and widely used imitative test is Texture Proﬁle
Analysis (TPA) developed by General Foods during the mid 1960’s (Szczesniak,
1963). The creation of the General Foods Texturometer allowed for correlations
to be made between instrumental and sensory data. The Texturometer generated
multiple sensory parameters that could be applied to various types of food items
(Friedman et al., 1963). By combining these two methods, standardized texture
parameters were established and assessed so that others could beneﬁt from deﬁned
levels of textural attributes. Later, Boume (1978) used the double compression
form of TPA using an Instron Universal Testing Machine to evaluate the textural
parameters of food. This procedure has now become a standard for texture
evaluation of solid foods.

There have been many advances in food texture evaluation that have led to
new techniques and equipment, and thus a more in-depth understanding of the

Correlation between objective and subjective measurements. When Henry et al.

18

(1971) evaluated the textural characteristics of semi-solid foods such as dessert
puddings, pie ﬁllings, and whipped toppings, many new physical parameters as
well as sensory tests were established. Some new parameters included stringiness,
maximum tensile force, and tension. Results suggested that physical values could
predict essential sensory attributes. Kalviainien et al. (2000) deﬁned and
quantiﬁed the most important texture and ﬂavor variables of high viscosity gels
with different thickeners and aromas. Panelists could detect the inﬂuence of
strong and weak gels on Speciﬁc ﬂavor intensities.

Other sensory studies involving gels have used the Universal Instron
Machine to correlate the physical and sensory properties. “Compressive
resistance” of six different gel systems was tested and correlated with sensory
analysis (Daget and Collyer, 1984). In that study, the hardness of the gels was
assessed by physical handling and its perception in the mouth. Both evaluations of
hardness correlated closely with rupture deformation and rupture force, which
were measured with an Instron. Similarly, Montejano et al. (1985) compared the
sensory results from both a fundamental torsion failure test and an imitative test
(TPA). The shear stress at failure and the TPA hardness were highly correlated

(R2 = 0.94) when evaluated by a trained panel.

19

MATERIALS AND METHODS

2.1. Procedure For Making Gelatin Gels
2.1.1. Formulations

Gelatin gel samples were comprised of gelatin (Great Lakes Unﬂavored
Gelatin, Grayslake, IL) (Table 2.1), sorbitol (Roquette America, Inc., Gumee, IL)
and distilled water. Sorbitol was added to the formulation because of its ability to
retain moisture and increase ﬁrmness of gels. Table 2.2 displays the actual
formulations of each of the concentration levels prepared. The amount of water
naturally occurring in the sorbitol (30% wet basis from manufaCturer speciﬁcation
sheets) was accounted for in the solution formulation. Sorbitol replaced 20% of
the water in each concentration level. Formulations also accounted for the amount
of moisture present in the commercial gelatin. The moisture content of the gelatin
on a wet basis was determined by drying four S-gram samples in aluminum pans

in an oven (Fisher Scientiﬁc, Itasca, IL) at 105°C for 17 hours (GMIA, 1986).

20

Table 2.1. Gelatin speciﬁcations of the three brands evaluated.*

 

 

Great Lakes Gelatin SKW Gelatin Knox Brand
Type A-Porcine B-Bovine A-Porcine
Bloom 225 grams 225 grams 235 grams
Color Light tan Light tan White/ Off-white
pH 4.5-5.8 5.0-6.0 3.8-5.5
Moisture (%,wb) 12 max 8-12 8-13
Ash (%) 1.0 max 2.0 max 0.3-2.0

 

*All values obtained from speciﬁcation Sheets from manufacturer

21

Table 2.2. Formulations for gelatin concentrations.

 

 

Gelatin (bone dry), % 22.5 25.0 27.5 30.0 32.5 35.0
Sorbitol,% 15.5 N 15.0 14.5 14.0 13.5 13.0
Water, % 62.0 60.0 58.8 56.0 54.0 52.0
Sample Mass, g 400 400 400 400 400 400
Actual Formulations

lRawGelatin,g 99.27 110.30 121.33 132.36 143.39 154.42
2Raw Sorbitol, g 88.57 85.71 82.86 80.00 77.14 74.29
Water, g 212.16 203.98 195.81 187.64 179.46 171.29

 

1 Average Moisture Content = 9.34% (wet basis);
2 Moisture Content = 30% (wet basis)

22

2.1.2. Sample Preparation

Gelatin gel samples were prepared according to a modiﬁed version of the
Gelatin Manufacturers Institute of America, Inc. (GMIA) method for cold water
dispersing (GMIA, 1986). This method of preparing gel samples was modiﬁed by
the addition of sorbitol. Concentration levels of dry gelatin ranged from 22.5% to
35.0% and were chosen after preliminary experiments revealed limitations:
gelatin gels below 22.5% were too soft to register a force response during
experimentation, and gels above 35% were difﬁcult to prepare and lacked
uniformity due to the incorporation of air bubbles in the samples. The desired
amount of sorbitol was added and stirred into the measured amount of water prior
to the dispersion of gelatin powder. Solutions of 400 grams were prepared by
pouring gelatin into an 800 ml Pyrex beaker containing the sorbitol and water at a
temperature ranging from 20°C to 30°C. The solution was stirred quickly with a
glass stirring rod and allowed to stand for approximately 40 minutes so that all
particles were fully hydrated. The solution was gently stirred and heated to 60°C
until uniform. Solutions were poured into 24 oz. plastic containers (13 x 8x 6 cm),
covered with a plastic lid, placed in a 5°C refrigerator, and held overnight.

Gel samples, 1.21 cm in diameter, were cored (Figure 2.1) at 620 rpm using
a V2" diameter coring cutter (G-R Manufacturing Co., Manhattan, KS) attached to
'a mounted 8-inch drill press (Sears, Roebuck and Co., Hoffman Estates, IL; Model
No. 137.219080). Kastner and Henrickson (1969) found that a mounted core

borer produced cores that were uniform in diameter. Cored samples were sliced

23

 

Figure 2.1. Coring gelatin samples using a mounted 8-inch drill press
(Sears, Roebuck and Co., Hoffman Estates, IL; Model No. 137.219080)
and a ‘/2" diameter coring cutter (G-R Manufacturing Co., Manhattan, KS).

24

with a ﬂat-edged blade to a length of 2.58 cm and stored at 5°C in airtight

containers until tested.

2.2. Characterization of F irmness
2.2.1. Compression Testing and Modulus Calculation

Gel samples, 1.21 cm in diameter, were cut to 1.21 cm in length so the
aspect ratio was equal to one. Eight samples. of each gelatin concentration (22.5%,
25.0%, 27.5%, 30.0%, 32.5%, and 35.0%) at 5°C were compressed between
lubricated plates attached to a TA.XT2 Texture Analyzer (Texture Technologies,
Corp., Scarsdale, NY/Stable Micro Systems, Godalming, Surrey, UK) at a test
Speed of 0.5 mm/s to a distance of 25% of the original sample height. Plates were
lubricated with vegetable oil to prevent samples from adhering to plates during
deformation (Bagley et al., 1985). Average force values were calculated for
samples at each concentration. Secant modulus values (determined at 15% strain)

were calculated for each concentration.

2.2.2. Shear Force Determination

To characterize the ﬁrmness of each sample, the maximum force required
to cut through the center of each gel sample was determined with a bench top
Wamer-Bratzler Shear instrument (G-R Elec. Mfg. Co., Manhattan, KS) (Figure
2.2), and a TA.XT2 Model Texture Analyzer equipped with a W-B blade

attachment (Figure 2.3). Both blades had a ‘/2" round bevel and were 1.016 mm

25

 

' ' le.
Figure 2 2 Warner-Bratzler Shear instrument wrth partially sheared samp

26

 

Figure 2.3. TA.XT2 Warner-Bratzler blade attachment with gelatin sample
loaded for testing.

27

25.4mm
diameter

liliCkllLNx ‘
l.()|(r mm 1

90mm

1
?

'l‘hickncss
l.()|(5 mm

70mm

 

Figure 2.4. Picture and dimensions of TA.XT2 and Warner-Bratzler blades.

28

thick (Figure 2.4). Eight core samples at six concentrations (22.5%, 25.0%,
27.5%, 30.0%, 32.5%, and 35.0%) were tested on each instrument and the average
Shear values calculated for each concentration. The temperature of the gelatin
samples was 5°C for the duration of the tests. The blades of both instruments

sheared through samples at a speed of 3.3 mm/s.

2.2.3. Temperature Effects

The Wamer—Bratzler Shear and the TA.XT2 were used to determine the
effects of temperature on the shear force values of gelatin gels. Gel samples of the
lowest and highest concentrations (22.5% and 35.0%) were equilibrated at three
different temperatures (5°C, 15°C, and 25°C) in three Haake water baths: F3/CH,
F 6/C25, and D1/W19 (Haake USA, Paramus, NJ). Five gelatin samples, 1.21 cm
in diameter and 2.58 cm in length, of each concentration, and at each temperature,
were sheared (quickly to keep the sample temperature constant) using the Wamer-
Bratzler and the TA.XT2. The average shear force values were calculated and

compared among the three temperatures.

2.2.4. Gelatin Brand Comparison
Two gelatin brands, in addition to the Great Lakes Gelatin used in the Shear
Force Determination, were tested to compare results for different types of gelatin.

The data in Table 2.1 compares the characteristics of Knox Brand Gelatin

(Nabisco, Parsippany, NJ) and Rousselot ® 225 B 40 Edible Gelatin (SK Gelatin,

29

Waukesha, WI). Eight core samples of each concentration (22.5%, 25.0%, 27.5%,
30.0%, 32.5%, and 35.0%) made from both brands of gelatin were sheared by the
Wamer-Bratzler and the TA.XT2. Average shear force values were calculated for

each concentration on each instrument. Shear force results of all three brands

were compared.

2.3. Sensory Panel Evaluation
2.3.1. Difference Testing

An untrained sensory panel of 30 panelists was used to subjectively
evaluate three concentrations (25%, 30%, and 35%) of gel samples. Volunteer
panelists were comprised of graduate students, staff members, and faculty
members of the Department of Food Science and Human Nutrition at Michigan
State Univerisity. Sample codes, generated from a random table of numbers, were
used for each of the concentrations (Meilgaard et al., 1991). Panelists signed
consent forms and were given rewards for their participation.

A triangle test was chosen to determine if panelists could distinguish the
difference between three levels of ﬁrmness. Individual booths with red lighting
(to mask color differences) in a sensory laboratory were used for the evaluation
(Figure 2.5). Panelists were served a tray that contained a ceramic plate (to
maintain cool temperature) with two labeled sets of three gel samples taken

directly from a refrigerator maintained at 5°C (Figure 2.6). The tray also

30

 

Figure 2.5. Sensory evaluation booth with red lighting to mask
color differences.

 

Figure 2.6. Example of tray served to panel participants.

contained a cup for expectoration, a pencil, and a score sheet. Two out of the
three gel samples of each set were the same and one gel sample of each set was an

odd sample. Panelists were asked to handle the gel samples with toothpicks
provided to prevent tactile evaluation of ﬁrmness, and changes in temperature due
to human contact. Panelists were instructed to bite completely through the center
of each sample with their incisors and expectorate any remaining sample in the
mouth. Panelists were then asked to identify the odd sample in each set and circle
the corresponding code number on the score sheet. Correct responses were
computed for each triangle test. Corresponding “P” values for the number of
correct responses accounting for the total number of panelists were found using
Statistical Chart 2 in Laboratory Methods for Sensory Analysis of Food (Poste et
al., 1991). Statistical Chart 2 determines the probability that the different sample
was correctly identiﬁed by chance alone. If the corresponding “P” values were

greater than 0.05, signiﬁcant differences existed between samples.

2.3.2. Trained Panelists

Twelve panelists (6 men, 6 women, age 22 to 38) participated in a trained
sensory panel evaluating the ﬁrmness of the gelatin gels (Meilgaard et al., 1991).
Panelists were selected based on their responses from the pre-screening difference
test described above or due to their strong interest in the study.

Panel training consisted of one main session held in small groups and

involved mostly explanations of the study, the purpose of training, and how

33

ﬁrmness of the gels was to be evaluated. Eight different concentrations of gelatin,
encompassing the concentration range to be assessed, were used for training
purposes. Panelists were asked to evaluate the ﬁrmness of each sample by biting
through the center of the gel with their incisors and to remember where that
sample rated on a ﬁrmness scale from 1 to 15.

Two evaluations were conducted in the Sensory Evaluation Laboratory at
Michigan State University. Partitioned evaluation booths and red lighting were
used to mask sample differences other then ﬁrmness. Every participant evaluated
six samples each time on a 15 cm unstructured line scale (Poste et al., 1991). The
samples were coded with 3-digit random codes and presented in a random order.
Panelists were asked to record each evaluation by marking a vertical line across
the horizontal line at the point that best reﬂected their perception of the magnitude
of ﬁrmness of each sample. Cups for expectorate and toothpicks were also
available on the tray presented.

The differences in ﬁrmness were studied using one-way analysis of
variance, where the one source of variation was the concentration. Tukey’s tests

were used to determine which concentrations signiﬁcantly differed from the others

on the standard scale.

34

RESULTS AND DISCUSSION

3.1. Compression Testing

Gelatin gels do not behave like Hookean solids; therefore, secant modulus
values needed to be determined to characterize nonlinear stress-strain behavior
(Mohsenin, 1970). Secant modulus is the slope of the line connecting the origin
and the point at which the sample was subjected to 15% strain. Equations for
compressive stress (0,) and compressive strain (strueﬁxial) were calculated as

described by Truong and Daubert (2000):

 

0c: F(H02—AH) (1)
7:12 H,
AH
Erma-axial = — 1n[1 _ E] (2)
Where: (IC = compressive stress (Pa)

8,",me = compressive strain (dimensionless)
F = force (N)
AH = deformation (mm)
Ho = initial height of sample, 12.1 mm
R = sample radius (m)
Table 3.1 displays the average values and related information calculated at
15% strain for six gelatin concentrations compressed to 25% of the original height

(1.21cm). The original sample radius of each gelatin sample was 0.6 cm, which

35

increased Slightly during the test. The values determined for stress (N/mz)
accounted for the change in sample radius as the samples were compressed. The
samples had a tendency to bellow out even with the lubricated plates, therefore
only a 25% deformation was used to obtain the stress and strain information. The
lowest concentration (22.5%) had the lowest modulus value (2.05x 105 Pa); .
however, the 27.5% gelatin samples did not f0110w the increasing trend.

Bi-axial compression was chosen as an evaluation method because it
simulates the act of biting between the molar teeth, and results indicate the
ﬁrmness of a sample. Other studies have successfully used this method to
evaluate the ﬁrmness of food products. However, in this study, there are a few
reasons why results did not accurately reﬂect the true mechanical properties of the
samples. Most of error can be attributed to the sample geometry. Core samples
tended to have a tapered shape due to the process of coring, so the diameter was
not always uniform. In addition, when subjecting a gel sample to a low
deformation (25%), few differences can be observed among concentrations. This
allows for too much variation between concentrations and no basis of comparison.
Also, compression with a ﬂat plate gave a lateral expansion with the applied load,
which required a larger force before reaching the yield point. Overall, data from
compression produced less consistent results. Testing in the Shear mode

(discussed in the next section) was superior as compared to compression testing.

36

Table 3.1. Determination of secant modulus values at 15% strain.

 

 

Concentration *Raw Force *Raw AH *Stress *Secant
(%) (N) (mm) (N/mz) Modulus (Pa)
22.5 4.13 1.69 3.08 E4 2.05 E5
25.0 6.11 1.69 4.55 E4 3.03 E5
27.5 7.71 1.69 5.74 E4 3.82 E5
30.0 6.66 1.69 4.96 E4 3.30 E5
32.5 7.25 1.69 5.40 E4 3.59 E5
35.0 7.65 1.69 5.70 E4 3.79 E5

 

*Values represent averages of eight gelatin samples per concentration.

37

3.2. Evaluation of Shear Forces

The average Shear force values for each concentration of gelatin gels were
calculated for both instruments (Wamer-Bratzler and TA.XT2) and displayed in
Table 3.2. The standard deviations for each average shear force value were also
calculated. Figure 3.1 shows how the shear force values increased as the gelatin
concentration increased when shearing samples on both instruments. The average
Shear force values ranged from 1.5 kg to 6.8 kg for the Wamer-Bratzler and from
3.2 to 9.4 kg for the TA.XT2. Overall, the TA.XT2 produced higher standard
deviations most likely due to the high degree of instrument sensitivity as compared
with the standard Wamer—Bratzler. Higher standard deviations may also reﬂect
Slight variations among samples. Correct and consistent positioning of the gel
sample in the center of the shearing device is important for obtaining good results.
The concentration of the gelatin may vary slightly along the length of the sample
due to the entrapment of air in the setting gel and the surface foam that may have
formed during upon heating. Shearing should occur as close to the center of the
sample as possible. Overall, these results Show that {the concentration of the
gelatin inﬂuences the intermolecular linkages forming the gel network, which
affects the ﬁrmness of the gel. As seen in other studies, Shearing was the most
effective method to discriminate across gel concentrations, and compression was
the least effective (Munoz, 1986).

Figure 3.2 Shows the correlation between the Wamer—Bratzler and the

TA.XT2. A strong correlation (R2 = 0.95) exists between these two instruments

38

Table 3.2. Average shear force values for gels with different gelatin
concentrations evaluated on the Warner-Bratzler and the TA.XT2.

 

Concentration Avg. WB Shear Force Avg. TA Shear Force

 

(%) (kg) (kg)

22.5 1.5 (0.12) 3.2 (0.18)
25.0 2.5 (0.09) 4.2 (0.30)
27.5 3.0 (0.05) 5.4 (0.22)
30.0 3.9 (0.09) 5.5 (0.27)
32.5 5.8 (0.26) 7.5 (0.32)
35.5 6.8 (0.28) 9.4 (0.40)

 

Standard deviations are shown in the parentheses next to the average shear force

value.

39

 

 

 

 

 

10.0
I
9.0 -
8.0 .
a I
5 7.0 - A
a)
g 6.0 _ ‘
Ln. I I
,_ 5.0 .
‘3 I
1.: . .
U) 4 0 A
g0 3.0 - I A
< 3 A l ITA.XT2
2.0 - l
10 A i AWarner-Bratzler
0.0

 

20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5 40.0

Gelatin Concentration (%)

Figure 3.1. Average shear force values vs. gelatin concentration for the
Warner-Bratzler shear instrument and the TA.XT2.

40

12

 

 

 

 

 

10-

’53

i

3.

o

LE

6—

.1:

U)

E 4

K -

<

E- 2 - 22.5% 112:0.95
0 I - r

0 2 4 6

Warner-Bratzler Shear Force (kg)

Figure 3.2. Warner-Bratzler and TA.XT2 Correlation.

41

when evaluating the Shear force of gelatin gels at 5°C. However, given the Slope
of the line in Figure 3.2, there is consistently a 10% increase in TA.XT2 shear
force value per kilogram increase in Warner-Bratzler shear force value. These
differences have also been seen in another study where many Shearing instruments
were compared (Lyon and Lyon, 1998). A possible reason for this discrepancy is
the actual Shearing Speed of the instrument. Although the manufacturer speciﬁes
the speed to be 3.3 mm/s throughout the shearing of the sample, the actual speed
of the Wamer-Bratzler was determined (in the MSU Meat Laboratory) to be
approximately 4.3 mm/s. If the sample is very ﬁrm, it has been suggested the
motor cannot maintain a constant speed while shearing through a sample (Johnson,
2001). In addition, the beveled edges of the blades may be slightly different
depending on the way they were crafted, which may also account for the different
average shear force values. Also, during the downward motion of the anvil on the
Wamer—Bratzler system, the blade will also experience some downward motion
while deforming the spring in the force transducer. Overall, the Warner-Bratzler
and TA.XT2 are both suitable for the evaluation of gelatin core samples because
the results of each are consistent and reproducible.

The values found for the shear forces in the current study are similar to the
three categories (Red, White and Blue) for beef tende’mess deﬁned by Boleman et-
al. (1997). The ranges for the three categories of beef tenderness were based on
shear force values: 1) 2.27 kg to 3.58 kg (Red); 2) 4.08 to 5 .40 kg (White); and 3)

5.90 to 7.21 kg (Blue). The three concentrations of gelatin that fell within or

42

around each of three tenderness categories were 25%, 30% and 35%, respectively.
Since consumers are able to distinguish between different degrees of tenderness
and are willing to pay more for more tender meat, these three gelatin
concentrations were chosen for sensory evaluation to see if panelists could detect
differences using gelatin as a model material (Section. 3.5). The gelatin
concentrations chosen for this study (22.5% to 35%) represent a wide range of
ﬁrmness, yet all fall within the range used to evaluate beef. Therefore, ﬁrmness
standards made of gelatin gels at the concentrations used are potentially good
model materials for tenderness evaluation. In addition, although the Wamer-
Bratzler and the TA.XT2 proved to be sufﬁcient instruments for evaluating gels,
the Wamer-Bratzler shear force values were used in subsequent tests to be

consistent with the Boleman et al. (1997) study.

43

3.3. Temperature Effects on Gelatin Gels

Many foods exhibit changes in rheological behavior as temperature
increases or decreases. Gelatin gels also exhibit this behavior when exposed to
different temperatures. Figure 3.3 shows the correlations between the Wamer-
Bratzler and the TA.XT2 for three different test temperatures (5°C, 15°C and
25°C) at two gelatin concentrations (22.5% and 35%). The ﬁgure displays a
higher correlation (R2 = 0.92) for the lower concentration of gelatin, and a lower
correlation (R2 = 0.82) for the higher concentration of gelatin. This variation in
shear force values at the higher concentration may be due to sample uniformity.
At the higher concentrations of gelatin, it was more difﬁcult to obtain completely
uniform samples due to partial air entrapment.

Each concentration Showed an increase in average shear force as the testing
temperature decreased on both the Warner-Bratzler and the TA.XT2 (Figure 3.4).
Table 3.3 Shows that the slopes of the 35% gelatin samples (-0.19 and —0.1) were
steeper than the slopes of the 22.5% gelatin samples (-0.09 and -0.05). These
results show that the gelatin gels are more affected by changes in temperature at
higher concentrations. Overall, temperature has a strong effect on the textural
properties of gelatin gels. Because of this, the temperature of the gelatin gels must
be clearly established, and carefully monitored, if ﬁrmness standards are to be
used for sensory panel training, instrument calibration, or for any other model

material application. Based on this study, a i 1kg variation in shear force value is

44

12

 

_s
O

 

Warner-Bratzler Shear Force (kg)
0)

 

 

 

y = 1.7x - 0.19

R2 = 0.92

4 - o , 50
22.5%
5 C
2 ..
25C
0 . . .
O 2 4 6 8
TA.XT2 Shear Force (kg)

Figure 3.3. Effect of temperature on shear force measurements of 22.5% and
35.0% gelatin gels.

45

—k

 

A

_ iI.22.5%-TA ‘
EA35.0%-WB.
— .\ L-35.O%-TA .

0 10 20 30

 

2
0
8 ‘ ' 322.5%-w3
6
4
2

 

 

Avg. Shear Force Value (kg)

 

Temperature (C)

Figure 3.4. Average shear force values vs. test temperature for 22.5% and
35% gelatin concentrations. (Each point is an average of ﬁve values).

46

Table 3.3. Comparison of slope values to demonstrate effects of temperature.

 

Material Slope Intercept R2
35% - TA -0.19 11.6 0.99
35% - WB -0.10 7.3 i 0.97

22.5% - TA -0.09 4.5 0.99
22.5% - WB -0.05 2.7 0.99

 

47

recommended. The corresponding allowable temperature range (calculated using
the equation of the line with greatest slope in Figure 3.4) is approximately j: 25°C.
Therefore, a testing temperature of 5°C (i 2.5°C) is recommended to minimize

temperature induced sample variation.

48

3.4. Comparison of Brands

Table 3.3 and Figure 3.5 show the comparison of the three gelatin brands
used in this study. The speciﬁc characteristics of each brand can be found in
Table 2.1. Figure 3.5 shows the variation between each brand of gelatin. These
differences can be attributed to the nature of the gelatin source and the natural
heterogeneity of this substance. For example, the SKW gelatin, which is derived
from a bovine source, did not have as high a shear force as the two brands of
gelatin derived from porcine. The chemical composition of the gelatin and the
way in which the proteins were hydrolyzed also contribute to the ﬁrmness of the
gel sample.

These results suggest that careﬁil preliminary testing is needed before
deciding on which concentration ranges to use. Gelatin gels derived from porcine
were proven easier to prepare and analyze than gelatin gels derived from bovine
sources. Bloom values may also be compared carefully so that the desired
firmness is achieved. The Bloom values for the Great Lakes Gelatin, SKW
Gelatin and Knox Brand Gelatine were 225, 225, and 235 grams, respectively.
Bloom values may vary a little due to the natural variability of the gelatin batch

itself; however, different brands of gelatin can be compared easily by this

characteristic.

49

Table 3.4. Comparison of average Warner-Bratzler shear force values of

three brands of gelatin.

 

 

Concentration *Great Lakes *SKW *Knox

(%) (kg) (kg) (kg)

22.5 1.5 (0.12) 1.5 (0.11) 1.8 (0.08)
25.0 2.5 (0.09) 1.6 (0.17) 2.5 (0.10)
27.5 3.0 (0.05) 2.7 (0.10) 3.3 (0.08)
30.0 3.9 (0.09) 3.2 (0.11) 3.7 (0.21)
32.5 5.8 (0.26) 3.2 (0.23) 4.9 (0.14)
35.0 6.8 (0.28) 4.6 (0.13) 5.2 (0.23)

 

*Refer to Table 2.1 for gelatin speciﬁcations of the three brands evaluated.
Standard deviations are shown in the parentheses next to the average shear force

value.

50

 

 

Warner-Bratzler Shear Force (kg)

 

 

 

O T l l l 1
22.5 25 27.5 30 32.5 35

Gelatin Concentration (%)

,___.. __._._..._ ._
1

l

- '0- Great Lakes
- El- -SKW

+Knox

 

Figure 3.5. Average shear force values for three brands of gelatin at six

different concentrations.

51

3.5. Difference Testing

Table 3.4 displays the concentrations that were chosen for sensory
evaluation. The three concentrations (25%, 30%, and 35% of Great Lakes
Gelatin) were selected for the triangle test because the corresponding shear forces
fell within or around the three shear force categories describing beef tenderness.
Table 3.5 shows the number of correct responses for panelists participating in the
triangle tests. At a critical P value of 0.05, detectable differences existed between
all three sample pairs. The probability that the results happened by chance was
less than 0.001. More panelists (18/20) could detect the difference between the
25% and 35% gelatin samples than the other pairs of samples.

Results show there are substantial differences in ﬁrmness among gelatin
samples. Panelists overwhelmingly gave correct responses when the concentration
differed by only 5% gelatin. Just as consumers can distinguish between categories
of beef tenderness, panelists can detect differences in ﬁrmness when three

different concentrations are used to reﬂect speciﬁc degrees of beef tenderness.

52

Table 3.5. Determining sensory parameters for difference testing from shear
force results.

 

 

Meat Shear Force Range Gelatin A vg. Shear Force
Categories (kg, WB) Concentration (kg, WB)

Red 2.27-3.58 22.5% 1.5
25.0% 2.5
27.5% 3.0
White 4.08-5.40 30.0% 3.9
32.5% 5.8
Blue 5.90-7.21 35.0% 6.8

 

53

Table 3.6. Correct responses from triangle tests.

 

 

Gelatin Correct Total Responses % Correct
Concentration Responses
25% and 30% *16 20 80%
30% and 35% *14 20 70%
25% and 35% *18 20 90%

 

*Panelists were able to detect a signiﬁcant difference between all concentration
levels (p<0.001)

54

3.6. Trained Sensory Panel

Tables 3.6 to 3.9 show the scaling results from the trained sensory panel.
Panelists marked off a ﬁrmness response on a 15 cm unstructured line for each of
the six gelatin concentrations. The distance from the origin of the line to each
mark was measured in centimeters and recorded as ratings on Tables 3.6 and 3.8.
Tables 3.6 and 3.7 display the ﬁrmness ratings from the unstructured scales and
the ANOVA values for the ﬁrst sensory evaluation. Tables 3.8 and 3.9 display the
ﬁrmness ratings on the unstructured scales and the ANOVA values for the second
sensory evaluation. The correlation between the average shear force values
determined by the Warner-Bratzler and the average ratings determined by
panelists from both panels was excellent (Fig. 3.6). There was a strong correlation
(R2 = 0.98) between the panelists’ responses and the shear force values.

The Tukey’s test, for each set of panel results, determined that a signiﬁcant
' difference existed among gelatin concentrations if their average values differed by
a value of 2.24 for the ﬁrst panel and 2.33 for the second panel. Panelists detected
signiﬁcant differences between all concentration levels in the ﬁrst panel except
between 22.5% and 25.0%. In the second panel, panelists detected signiﬁcant
differences between all concentration levels except between 25.0% and 27.5%.
Overall, panelists agreed on where each gel sample fell on the ﬁrmness scale. A11
panelists followed the training instructions and used the entire scale when

evaluating the ﬁrmness of the gel samples with their incisors; therefore, overall

ﬁrmness ratings fell very close together on the scale.

55

Table 3.7. Results of the ﬁrst unstructured scale test.

 

Gelatin Concentration (Code)

 

Panelist 22.50% 25.00% 27.50% 30.00% 32.50% 35.00%
(656) (157) (138) (536) (396) (868)
1 0.4 2.4 4.5 7.5 11 14
2 0.2 5.5 7.2 9.5 12 14.8
3 1.4 3.8 5.5 7.8 11.5 13.9
4 1 3.4 6 5 11 15
5 0.2 3.4 7.5 5.4 9.3 15
6 0.8 2.1 4.1 7.3 14.3 11.7
7 0.6 3 3.8 5.5 10.7 14.2
8 1.3 2.7 3.5 11.9 8.2 10.2
9 1 0.8 1.2 10.5 12.6 14.5
10 0.5 2.7 3.7 9.7 13.2 15
11 0.4 1.4 9.2 11.4 14.5 15
12 0.2 1.8 4.8 3.5 7.6 13.5
AVG 0.7 2.8 5.1 7.9 11.3 13.9
Std Dev. 0.4 1.2 2.1 2.7 2.2 1.5

 

56

Table 3.8. ANOVA results for ﬁrst unstructured scale test.

 

Source of Variation SS df MS F P-value F crit

 

Between Groups 1547.8 5 309.6 89.5 4.89E-28 2.35

Within Groups 228.3 66 3.46

Total 1776.1 71

 

57

Table 3.9. Results of the second unstructured scale test.

 

Gelatin Concentration (Code)

 

Panelist 22.50% 25.00% 27.50% 30.00% 32.50% 35.00%
(288) (996) (972) (738) (112) (611)

1 0.8 6.3 3.2 9.4 12.2 14.5
2 1.5 3.4 6.9 4.8 10 12.3
3 0.2 3.1 4.8 8.1 13 14.6
4 0.1 1.7 3 6.6 10.7 14.9
5 0.1 1.6 6.7 9.5 12 13.4
6 2.2 4.3 5.6 11.5 10.5 12.4
7 0.5 5.3 7.5 8.8 10.5 14.7
8 l 2.7 5.4 6.6 12.1 8.8
9 0.9 0.8 6.7 7.8 12 14.4
10 0.1 8.5 3.6 12 14.9 14.2
11 0.9 8.4 9.2 13.8 12.5 14.8
12 0.2 1.8 4.4 9 13 14.9
AVG 0.7 4.0 5.6 9.0 12.0 13.7

Std Dev. 0.7 2.6 1.9 2.5 1.4 1.8

58

Table 3.10. ANOVA results for second unstructured scale test.

 

Source of Variation SS df MS F P-value F crit

 

BetweenGroups 1464.7 5 292.9 78.7 1.87E-26 2.35

Within Groups 245.7 66 3.72

Total 1710.4 71

 

59

 

6 ‘ 35%

   
 

O
30%

3 _

27.5%
2‘ . y=0.40x+1.0
1 . 22.5% R2=0-98

 

 

0 .

 

Warner-Bratzler Shear Force (kg)
A

5 A 10 15
Panel Firmness Rating (avg.)

Figure 3.6. Trained panel results for sealing of gelatin gels (results include
ﬁrst and second unstructured scale tests).

60

3.7. Practical Recommendations

There are several recommendations that can be made regarding the
preparation and use of ﬁrmness standards made from gelatin gels. First, the
speciﬁc brand of gelatin and its source (either porcine or bovine) may inﬂuence
the ﬁrmness characteristics and/or the variability of the standards. From this
study, gelatin derived from a porcine source is recommended due to ease of
preparation and the accuracy of results. Certain gelatin speciﬁcations, such as
moisture content, are needed so that they can be factored into formulations. The
Bloom value of the gelatin will also affect the ﬁrmness of the standards. A Bloom
value within the range of 225-250 grams is suggested if very ﬁrm gels are desired.
Method of dispersion should be taken into account if preparation time is a crucial
factor. Overall, preliminary experimentation with many concentration levels and
several brands is necessary to achieve the desired ﬁrmness standards.

Firmness standards must maintain a speciﬁc temperature during testing.
Holding gelatin gel standards at cold temperatures show less sample variation and
are easy to achieve with a water bath or a refrigerator. A temperature of 5°C (i
25°C) is recommended to minimize ﬁrmness variation. Samples sheared within
this temperature range produce the most repeatable results. Although aging effects
were not speciﬁcally evaluated in this study, a shelf life of less than four days is
recommended due to potential mold growth and possible changes in texture.
Experience in the current study suggests that solutions be made one day and the

gels used on the following day. In addition, batch-to-batch variations may occur

61

due to ﬂuctuating storage conditions and/or the natural heterogeneity of the
gelatin.

Based on the gelatin gel procedure determined in this research, there are
some speciﬁc practical suggestions that can be made in applications involving
meat tenderness. Porcine gelatin (225 Bloom), with a pre-determined moisture
content, should be used in the sample formulations. Gelatin concentrations
ranging from 22.5% to 35.0% should be prepared by dispersing the gelatin in a
solution of water and sorbitol (20% replacement of water) held at room
temperature. Solutions should be completely melted by heating to 60°C, poured
into a mold, and allowed to gel in a refrigerator. Gel samples should be cored and
tested quickly or stored (in an air-tight container) at 5°C to prevent temperature
ﬂuctuation.

A standard scale of six benchmark levels can be established by preparing
the following gelatin concentrations and the corresponding Warner-Bratzler shear
values (kg): 22.5% (1.4-1.8), 25.0% (2.3-2.6), 27.5% (2.9-3.0), 30.0% (3.1-3.5),
32.5% (3.9-4.5), and 35.0% (5.1-5.6). Once these ﬁrmness standards are
established and characterized by a Wamer—Bratzler or a TA.XT2, sensory panelists
can be trained on each of the gelatin concentration levels. Difference tests can be
used to screen panelists prior to formal training and unstructured scales can be
used to evaluate panelists during the course of training. Panelists can then
evaluate meat samples (or similar materials) with corresponding shear force values

when they have a good feeling for the ﬁrmness scale established with gelatin.

62

SUMMARY

Six scaled standards of ﬁrmness (22.5%, 25.0%, 27.5%, 30.0%, 32.5%,
35.0%) composed of gelatin gels were chosen to represent six distinct degrees of
ﬁrmness and compared to tenderness of meat. The range of ﬁrmness was
restricted to these concentration levels due to the limitations of the testing method
at low concentrations and lack of uniformity at high concentrations. This range
can be expanded if other testing methods are implemented and uniform gels can be
achieved.

Compression testing and the calculation of secant modulus values for
gelatin gels produced inconclusive results. The results produced by shearing
samples on both the Warner-Bratzler and the TA.XT2 proved to be superior in
accuracy, and a better reﬂection of the texture (ﬁrmness) of each standard.
Untrained panelists were able to distinguish among three concentration levels of
gels representing three tenderness categories (p _<_ 0.001) in three consecutive tests.
Signiﬁcant differences were detected by trained panelists between all six scaled
standards. Overall results suggest that signiﬁcant differences exist between all
. six ﬁrmness standards, thus supporting the instrumental results.

It was found that variations in temperature caused changes in the ﬁrmness
of samples. The temperature of the gel samples must be closely monitored during
testing and should not vary any more than i2.5°C to prevent any textural changes.

Different brands of gelatin may also contribute to changes in ﬁrmness among

63

samples of the same concentration level. Hence, preliminary experiments are
needed to achieve desired ﬁrmness levels to ensure prOper testing conditions.

A simple procedure for producing ﬁrmness standards made from gelatin
gels has many food and dental research applications. Results suggest that scaled
ﬁrmness standards made from gelatin and sorbitol are good model materials of
meat tenderness. Using gelatin gels as an indicator of meat tenderness may lead to
a more standardized method of texture characterization, and may reduce the
variability from human or instrumental assessment. In addition, ﬁrmness
standards made from gelatin gels can be used in place of current scaled food items
in the training of sensory panels. Because these standards are made of one
naturally abundant primary constituent, problems with local unavailability of
brands and numerous sample size and temperature speciﬁcations can be
eliminated. Lastly, dental researchers trying to determine the effectiveness of
denture designs, implants, and cutting surfaces can use gelatin gel standards as

food models.

64

LIST OF REFERENCES

BAGLEY, E.B., WOLF, W.J., and CHRISTIANSON, DD. 1985. Effect of
sample dimensions, lubrication and deformation rate on uniaxial
compression of gelatin gels. Rheo. Acta 24, 265-271.

BOLEMAN, S.J., BOLEMAN, S.L., MILLER, R.K., TAYLOR, J .F ., CROSS,
H.R., WHEELER, T.L., KOOHMARIE, M.,.SHACKELFORD, S.D,
MILLER, M.F., WEST, R.L., JOHNSON, DD, and SAVELL, J.W. 1997.
Consumer evaluation of beef of known categories of tenderness. J. Ani.
Sci. 75, 1521—1524.

BOURNE, MC. 1982. Food Texture and Viscosity, Concept and Measurement,
Academic Press, New York.

BOURNE, MC. 1978. Texture proﬁle analysis. Food Tech. 32, 62-66,
72.

BOURNE, M.C., SANDOVAL, A.M., VILLALOBOS, M.C., and BUCKLE, TS.
1975. Training of a Sensory Texture Proﬁle panel and development of
standard rating scales in Columbia. J. Texture Studies 6, 43-52.

BOYER, M.M and KILCAST, D. 1986. Review: food texture and dental science.
J. Texture Studies 16, 221.

CARDELLO, A.V., MATAS, A., and SWEENEY, J. 1982. The standard scales
of texture: rescaling by magnitude estimation. J. Food Sci. 47, 1738-1742.

CHAND, N. 1986. Textural characteristics of foods based on Warner-Bratzler
' shear. J. Food Sci. and Tech. 23, 49-54.

CHRYSTALL, BB. 1994. Meat texture measurement. Adv. in Meat
Research 9, 316-336.

DAG ET, N. and COLLYER, S. 1984. Comparison between quantitative
descriptive analysis and physical measurements of gel systems and
evaluation of the sensorial method. J. Texture Studies 15, 227-245 .

FRIEDMAN, H., WHITNEY, J ., and SZCZESNIAK, AS. 1963. The

Texturometer: A new instrument for objective texture measurement. J.
Food Sci. 28, 390-396.

65

GMIA. 1986. Standard methods for the sampling and testing of gelatins. Ofﬁcial
procedure of the Gelatin Manufacturers Institute of America, Inc.

GREASER, ML. and PEARSON, A.M. 1999. Flesh Foods and Their Analogues.
In Food Texture: Measurement and Perception, (Rosenthal, A.J., ed.)
Aspen Publishers, Inc, Gaithersburg, MD.

GUINARD, Ji-X. and MARTY, C. 1995. Time-intensity measurement of ﬂavor
release from a model gel system: effect of gelling agent type and
concentration. J. Food Sci. 60, 727-730.

HAMANN, D.D. AND LANIER TC. 1987. Instrumental methods for predicting
Seafood sensory texture quality. In Seafood Quality Determination,
(Kramer, DE. and J. Liston eds). Elsevier Applied Science,

England.

HENRY, W.F., KATZ, M.H., PILGRIM, F.J., and MAY, A.T. 1971. Texture of
semi-solid foods: sensory and physical correlates. J. of Food Sci. 36, 155-
161.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. 1992.
Sensory analysis: Vocabulary (ISI 5492). Available from the ISO Central
Secretariat, Case postale 56, 1211 Geneva, 20, Switzerland.

JOHNSON, M. 2001. Personal communication. Texture Technologies Corp.
Scarsdale, NY.

JOHNSTON-BANKS, FA. 1990. Gelatine. In Food Gels. (P. Harris,
ed.) Elsevier Applied Science, London.

KASTNER, CL. and R.L. HENDRICKSON. 1969. Providing uniform meat
cores for mechanical shear force measurements. J. Food Sci. 34, 303-303.

KALVIAIN EN, N., ROININEN, K., and TUORILA, H. 2000. Sensory
characterization of texture and ﬂavor 0f high viscosity gels made with
different thickeners. J. Texture Studies 31, 407-420.

LASSAUZAY, C., PEYRON, M.-A., ALBUISSON, E., DRANSFIELD, E., and
WODA, A. 2000. Variability of the masticatory process during chewing
of elastic food models. Eur. J. Oral. Sci. 108, 484-492.

LELIEVRE, J ., MIRZA, I.A., and TUNG, MA. 1992. Failure testing of gellan
gels. J. ofFood Eng. 16, 25-37.

66

 

LYON, B.G. and LYON, CE. 1998. Assessment of three devices used in shear
tests of cooked breast meat. Poultry Sci. 77, 1585—1590.

MARRS, M. 1997. Melt-in-mouth gels. World of Food Ingred, June, 39-40.

MATSUHASHI, T. 1990. Agar. In Food Gels. (P. Harris, ed.) Elsevier
Applied Science, London.

MEILGAARD, M., CIVILLE, G., and CARR, T. 1991. Sensory evaluation
techniques. 2nd Ed., CRC Press, Boca Raton, FL.

MOHSENIN, N.N. 1970. Physical Properties of Plant and Animal Materials,
Gordon and Breach Science Publishers, New York.

MONTEJANO, J .G., HAMANN, DD, and LANIER, TC. 1985. Comparison of
Two instrumental methods with sensory texture of protein gels. J. Texture
Studies 16, 403-424.

MORRIS, V.J. 1986. Multicomponent Gels. In Gums and
Stabilisers for the Food Industry 3, 1St Ed. (G.O. Phillips, Wedlock,
DJ, and RA. Williams, eds.) Elsevier Applied Science, London.

MUNOZ, A.M. 1986. Development and application of texture reference scales.
J. Sensory Studies 1, 55-83.

MUNOZ, A.M., PANGBORN, R.M., and NOBLE, AC. 1986. Sensory and
mechanical attributes of gel texture. I. Effect of gelatin concentration. J.
Texture Studies I 7, 1-16.

PROCTOR, B.E., DAVISON, S. and BRODY, AL. 1955. A recording strain-
gauge denture tenderometer for foods. 1. Instrumental evaluation and
initial tests. Food Tech. 9, 471.

POSTE, L.M., MACKIE, D.A., BUTLER, G., and LARMOND, E. 1991.
Laboratory Methods for Sensory Analysis of Food, Research Program
Service, Canada.

RAFFENSPERGER, E.L., PERYAM, DR, and WOOD, KR. 1956.
Development of a scale for grading and toughness-tenderness in beef. Food
Tech. 10, 627-630.

67

RAJALAKSHMI, D., DHANARAJ, S., CHAND, N., and GOVINDARAJAN,
VS. 1987. Descriptive quality analysis of mutton. J. Sensory Studies 2,
93-118.

ROLIN, C. and DE VRIES, J. 1990. Pectin. In Food Gels. (P. Harris,
ed.) Elsevier Applied Science, London.

SHERMAN, P. 1969. A texture proﬁle of foodstuffs based upon well-deﬁned
rheological properties. J. Food Sci. 34, 458-462.

SMEWING, J. 1999. Hydrocolloids. In Food Texture: Measurement
and Perception. (Rosenthal, A.J, ed.) Aspen Publishers, Inc.,
Gaithersburg, MD.

STANLEY, NE 1990. Carrageenans. In Food Gels. (P. Harris, ed.)
Elsevier Applied Science, London.

STEVENS, SS. and GALANTER, EH. 1957. Ratio scales and category scales
for a dozen perceptual continua. J. Exper. Psych. 54, 377.

SZCZESNIAK, AS. 1963. Classiﬁcation of Textural Characteristics. J.
Food Sci. 28, 385-389.

SZCZESNIAK, AS. 1987. Review paper: correlating sensory with instrumental
texture measurements- an overview of recent developments. J. Sensory
Studies 18, 1-15.

SZCZESNIAK, A. S. and BOURNE, M. C. 1969. Sensory evaluation of food
ﬁrmness. J. Texture Studies], 52- 64.

SZCZESNIAK, A.S., BRANDT, M.A., and FRIEDMAN, H.H. 1963.
Development of standard rating scales for mechanical parameters of texture
and correlation between the objective and the sensory methods of texture
evaluation. J. Food Sci. 28, 397-403.

TANG, 1., TUNG, M.A., LELIEVRE, J., and ZENG, Y. 1997. Stress-strain
relationships for gellan gels in torsion, compression, and torsion. J. Food
Eng. 31, 511-529.

TRUONG, V.D. and DAUBERT, CR. 2000. Comparative study of large strain

methods for assessing failure characteristics of selected food gels. J.
Texture Studies 31, 335-353.

68

WHEELER, T.L., S.D. SHACKELFORD, L.P. JOHNSON, M.F. MILLER, R.K.
MILLER and M. KOOHMARAIE. 1997. A comparison of Wamer-Bratzler

shear force assessment within and among institutions. J. Ani. Sci. 75, 2423-
2432.

WHEELER, T.L., SHACKELFORD, SD, and KOOHMARIE, M. 1996.

Sampling, cooking, and coring effects on Warner-Bratzler shear values in
beef. J. Ani. Sci. 74, 1553-1562.

69

IIIIIIIIIIIIIIIIIIIIII

11111111111111.111111111111911