Textural and Rheological Properties of Processed Cheese

Abstract

Cheddar cheeses made from ultra filtered (UF) as well as vacuum condensed milks (CM) containing two protein levels (4.5 and 6.0%) were used to manufacture processed cheeses. These processed cheeses were evaluated for instrumental textural profile analysis (TPA), stress relaxation characteristics using Sintech universal testing machine, and visco elastic characteristics (Elastic modulus-G′ and viscous modulus-G′′) using a Haake Viscometer. A small amplitude oscillatory shear test was employed to assess the visco elastic characteristics. Peleg model and six elements Maxwell model were used to assess stress relaxation characteristics of the cheese. Results indicated that the instrumental TPA hardness of UF2 cheeses made from high protein UF milk was highest (149 N) as compared to cheeses made from low protein UF milk (125 N), high and low protein vacuum condensed milks (103 and 62 N, respectively), and control (53 N). The values of elastic and viscous moduli of cheeses made utilizing high protein UF milk were 3.53 and 2.32 MPa respectively, which indicated its higher viscoelastic nature. The UF2 cheese also showed higher values of modulus of elasticity during stress relaxation (21.8 and 16.2 kPa respectively for Peleg and Maxwell models) and required longer time to relax the given amount of stress as compared to rest of the cheeses. The higher values of hardness and visco-elasticity were attributed to higher protein and lower fat contents in the cheeses. The methods of instrumental texture and rheology evaluation had common outcome for hardness and visco-elasticity of cheeses. Results of stress relaxation tests were more useful to differentiate the cheese characteristics as compared to the results of instrumental TPA and dynamic rheology tests.

Keywords:

• TPA
• Dynamic rheology
• Stress relaxation
• Peleg model
• Maxwell model
• Process cheese
• UF
• Vacuum condensed milk

Introduction

Cheese rheology is an important tool to study and identify the textural and structural properties. It deals with deformation of the sample by employing different kinds of instruments. Results of the small and large deformation tests are interpreted to understand the effect of composition, process modification, and storage etc., variables. With introduction of advanced instrumentation texture profile analysis, small amplitude oscillatory shear test—dynamic stress rheology and stress relaxation tests have been employed routinely in cheese research. Park et al.1 correlated effect of emulsifying salts on various textural parameters like hardness, gumminess, cohesiveness, brittleness, and adhesiveness of processed cheese made with ultra filtered (UF) cheese base. Jack et al.2 differentiated various Cheddar cheeses with wide range of textural characteristics using electronmyography, quantitative descriptive profiling and Instron deformation measurements. They obtained good prediction of sensory score from the textural variables of Cheddar cheese. Mistry and Kasperson3 observed increase in hardness and fracturability upon increasing the salt content of low fat Cheddar cheese. However, recently Finney et al.4 showed need for more appropriate practice to obtain high correlation between instrument and sensory data for predicting cheese texture. Ak and Gunasekaran5 determined elastic (G′) and viscous (G′′) moduli of low-moisture part-skim Mozzarella cheese at 10 and 20°C during one month of refrigerated storage. At both temperatures, G′ was always greater than G′′. During storage at 10°C, G′ increased from 90 to 630 kPa and G′′ increased from 44 to 52 kPa, whereas increase in G′ and G′′ at 20°C was from 28 to 190 kPa and from 14 to 53 kPa respectively. Sutheerawattananonda and Bastian6 used dynamic stress rheology parameters to monitor process cheese meltability. Chung and Meullenet7 used fundamental rheological analysis using a stress controlled dynamic rheometer to predict cheese texture attributes. However, they could not establish a good correlation between the large deformation tests and texture attributes. Hassan and Lucey8 characterized melt properties of Cheddar cheese using dynamic low amplitude oscillatory rheology and melt profile analysis. Yun et al.,9 while studying the effect of draw pH and storage on stress relaxation behavior of Mozzarella cheese found that hypothetical equilibrium stress was higher at higher draw pH indicating stronger network structure in presence of higher calcium. Wium and Qvist10 suggested that either stress at fracture alone or together with other parameters from uniaxial compression should be used to describe texture properties of Feta cheese made by UF milk. Achilleos et al.11 used stress relaxation test to study openness of cheese. Venugopal and Muthukumarappan12 studied stress relaxation characteristics during heating and cooling of Cheddar cheese containing various fat and moisture contents. They observed that Peleg model and Maxwell model containing eight elements well described the stress relaxation behavior of melted cheeses. The authors recommended the stress relaxation test as a tool to study cheese functionality. These rheological tests have good correlation with cheese properties, when performed individually. However, study of texture profile, dynamic rheology and stress relaxation behavior of processed cheese together is not available in the literature.

Use of concentrated milk for cheese making has become a regular practice and Cheddar is the most common variety of cheese manufactured from concentrated milk. Various methods, such as vacuum concentration, membrane processing including ultrafiltration, reverse osmosis, micro filtration etc., are employed for concentrating milk, however, vacuum concentration and lately ultrafiltration are more common methods of milk concentration. Use of concentrated milk in cheese making by both the methods has pros and cons; however, comparison of vacuum concentration and ultrafiltration methods for cheese making is not reported so far. Further, utilization of Cheddar cheese made from concentrated milk in processed cheese making is also not reported. Recently, Acharya and Mistry (personal communication) studied effect of concentration method and level of concentration on Cheddar cheese manufacturing parameters. The present experiment was aimed to study effect of milk concentration method and level of concentration on textural and rheological characteristics of processed cheese using instrumental texture profile analysis, small amplitude oscillatory shear test (dynamic rheology), and stress relaxation test.

Materials and Methods

Processed Cheese Manufacture

Processed cheeses were manufactured using different Cheddar cheeses. The Cheddar cheeses were made using ultra filtered (UF) and condensed (CM) milks with two protein levels, viz. 4.5% (UF1 and CM1) and 6.0% (UF2 and CM2). Control Cheddar cheese (C) was made from normal milk.13 Respective blocks of the Cheddar cheeses of 12 and 30 weeks of age were blended in 1:1 proportion for process cheese manufacture. Anhydrous milk fat and di-sodium phosphate, each @ 3% were added to the shredded cheese before processing the cheese at the temperature of 80–85°C.

Chemical Analysis

Each sample of the Process cheese was analyzed for moisture using a moisture balance (model MB200; Ohaus Corp., Florham Park, NJ) by a method of Crosser and Mistry,14 fat by the Mojonnier method15 and ash (method number 33.7.07, 935.42) and total protein contents (method number 33.7.02.991.20) by methods outlined in AOAC.16

Samples Preparation

Samples for instrumental TPA were prepared by cutting the cheese block into cylindrical size of 20 mm diameter using a cork-borer and 20 mm height using a knife.3 Samples for Stress relaxation test were also prepared by the same method as above. For dynamic rheology study a slice of 3.6 mm thickness was cut using a food slicer (Model MS1043-W, The Rival Co., Kansas, MO) considering perpendicular direction to the cheese surface. From this slice a cylindrical specimen of 20 mm diameter was cut with a cork borer. The cheese samples were covered in a petri dish and transferred in cold atmosphere (20°C) to avoid moisture loss and to maintain uniform temperature throughout the test.

Instrumental Texture Profile Analysis

In a two-compression test, the cheese samples were compressed using a universal testing machine (2/D model, MTS Sintech Inc., Research Triangle Park, NC) at 75% compression, 100 N load cell and 50 mm/min cross head speed.3 A personal computer and software (Test work Version 2.10, MTS System Corp., Eden Prairie, MN) were used to obtain a force-distance curve and to determine the texture attributes including hardness, cohesiveness, springiness, gumminess, adhesiveness, and chewiness.3 The maximum force during the first compression was hardness. The extent to which the sample returned to its original height between the first and second compression was springiness. Energy (work done) of cheese during the second compression in relation to the first compression was termed as cohesiveness. Gumminess was calculated as a product of hardness and cohesiveness, whereas, chewiness was termed as a product of gumminess and springiness. Adhesiveness corresponded to the energy (work done) of the first inverse peak.

Dynamic Rheology

A Haake® viscometer (HB Instruments Inc., Paramus, NJ) was used to characterize the rheological properties of cheese samples. Control system RV20 was used to control shear rate, and control system RC20 was used to control amplitude and frequency. CV20 sensor system with a Q20 sensor was used to measure the dynamic rheological properties of the cheese. A personal computer with software (Haake® Software, Roto visco RV20, Oscillation, Version 2.23) was used for data collection and analysis.

The test was performed by placing the specimen between the parallel plate fixtures of CV20 sensor system and oscillating at a constant frequency of 1 Hz. The peripheral surface of the cheese was oiled to avoid moisture loss during the testing. From earlier reports,12 17 it was found that time doesn’t affect the elastic and viscous modulus of cheese, hence, time sweep mode was used to calculate rheological properties. A constant strain of 0.5% was used for testing so that viscoelastic properties of the cheese are obtained in a linear region.12 The data collection included elastic modulus G′ and tangent (tan δ). Viscous modulus was computed using relation G′′ = G′ × tan δ. Variation of G′ and G′′ was recorded as a function of time. Values of G′ and G′′ were used to analyze the visco-elastic properties of the cheese samples.

Stress Relaxation

A plunger of the universal testing machine was allowed to touch the upper surface of the cheese sample so that the stress applied was zero. Subsequently the plunger was set into motion and the sample was compressed instantaneously to 10% deformation whereupon the crosshead was stopped and the sample was held in this position for 3 min. The 10% deformation was decided keeping in view to have minimum damage to the sample and to obtain the data in a linear region. The force was measured as a function of time and the data were interpreted as stress and strain.

The stress relaxation was recorded as a function of time. To analyze the data obtained by the stress relaxation experiment, two rheological models, namely Maxwell model containing 6 elements and Peleg model were considered for representation. These models were based on the time effect on the mechanical properties of the cheese. The characteristic stress and strain curves for the relaxation behavior were used to determine the values of the modulus of elasticity as a function of time. The stress was converted into Modulus of Elasticity by dividing it with constant strain. To observe the time effect on the mechanical properties of a material, a Maxwell body, connected in parallel with a spring, was used in this study. The generalized Maxwell model18 is represented by the following equations:

where, E(t) is the modulus of elasticity for the entire body at any time t, E ₁ is the modulus of elasticity of the first element of the Maxwell body, E ₂ is the modulus of elasticity of second element, in this case a spring, and τ is the relaxation time of the Maxwell element of the body. The relaxation time is the time required for the stress on the Maxwell body to decay to a value that is 37% of the total stress of that body. The coefficient of viscosity for the Maxwell body can be determined using η = τ × E ₁. The Maxwell models were analyzed using successive residuals method suggested by Mohsenin.19 The following equation represented the six-element Maxwell model:

where E(t) is the modulus of elasticity for the entire three-element body at any time t, E ₁, E ₂, and E ₃ are the elastic modulus of springs and τ ₁, τ ₂, and τ ₃ are the relaxation times. The time dependent elastic modulus of the Peleg model was represented as:

where, S(t) is the modulus of elasticity for the entire Peleg model at any time t, S _i is the initial modulus of elasticity, S _e is the equilibrium modulus of elasticity at infinite time, and B is the time necessary for (S ₁ − S _e )/(S _i  − S _e ) to become equal to 0.5.20 21

Statistical Analysis

Each of the five processed cheeses, viz. control, CM1, CM2, UF1, and UF2 were analyzed in triplicate for various attributes. The whole experiment was replicated four times resulting in 4 × 5 × 3 = 60 analyses for each parameter. Mean square values of all the parameters were determined using the GLM procedure on SAS® Statistical software package (Version 6.12, SAS® Institute, Inc., Cary, NC). An interaction was described to be significant only when p < 0.05.

Results and Discussion

Cheese Composition

Composition of various process cheeses is shown in Table 1. CM2 cheese contained maximum (p < 0.05) moisture among all the cheese samples, whereas the moisture of rest of the cheeses was at par (p > 0.05). Fat content of control and CM1 cheese was at par and highest among all the cheeses. CM2 and UF2 cheeses had lowest fat content. UF2 cheese contained the highest protein (p < 0.05) followed by CM2 and UF1, CM1 and control. The proteins contents of CM2 and UF1 were statistically similar. The CM2 and UF2 cheeses had highest ash content (p < 0.05) followed by CM1 and UF1 and then control. Control cheese had the least amount of ash among all the cheeses. The ash contents of CM2 and UF2 and CM1 and UF1 cheeses did not differ significantly (p < 0.05). Overall it was observed that the cheeses made from concentrated milks possessed higher level of protein and ash as compared to control. Further, the protein and ash contents were higher and fat was lower in the cheeses made from high protein milks (UF2 and CM2) as compared to those made from low protein milks (UF1 and CM1).

Table 1 Composition of processed cheeses

Parameter (%)	C	CM1	CM2	UF1	UF2
Moisture	39.33b	39.46b	40.00a	39.10b	39.23b
Fat	34.92a	34.72a	31.76c	33.79b	31.47c
Protein	19.49^d	20.05c	21.54b	21.34b	22.64a
Ash	4.67c	5.00b	5.40a	4.94b	5.26a

^a,b,cValues within a row not sharing a common superscript significantly differed (p < 0.05).

Dynamic Rheology

Table 2 shows values of visco-elasticity parameters in terms of elastic and viscous moduli of various cheeses. The values of G′ and G′′ were maximum in CM2 and UF2 cheeses and minimum in control cheese, which indicated that the former cheeses were more visco-elastic compared to the rest of the cheeses, i.e., these cheese had greater values of elastic as well as viscous moduli. The protein network structure is responsible for making cheese viscoelastic.21 When fresh, the proteins of the cheese tend to maintain strong protein-to-protein interactions and have high G′ values even though expressible water and fat are present within the fat channels. High protein and ash (particularly calcium) and low fat contents in these cheeses are attributed for high values of G′ and G′′. The values of elastic modulus (G′) were higher than viscous modulus (G′′) in all the cheeses. The higher values of elastic modulus as compared to their respective viscous modulus indicate that all the cheeses behaved more like an elastic material. The dynamic rheological test differentiated the cheeses made using vacuum concentrated milk (i.e., CM1 and CM2) but not the cheeses made using ultrafiltered milk (i.e., UF1 and UF2).

Table 2 Elastic (G′) and viscous (G′′) moduli of processed cheeses

Parameter	C	CM1	CM2	UF1	UF2
G′ (MPa)	2.40c	3.06b	3.70a	3.38a b	3.53a
G′′ (MPa)	1.51d	1.83c	2.58a	2.12b c	2.32a b

^a,b,cValues within a row not sharing a common superscript significantly differed (p < 0.05).

Instrumental Texture Profile Analysis

Results of instrumental TPA analysis presented in Table 3 indicated that UF2 cheese possessed maximum hardness, gumminess, and chewiness among all the experimental cheeses. Hardness and adhesiveness values of UF2 cheese were statistically similar to those of UF1 cheese. Hardness of UF1 was statistically similar to that of CM2 cheeses (P > 0.05). CM1 and control cheeses had similar hardness (P > 0.05), which was the least among all the cheeses. It was also observed that all the instrumental textural parameters of UF2 and UF1 cheeses were statistically similar but it was always not true in case of CM2 and CM1 cheeses. CM2 cheese had greater hardness and chewiness than CM1 cheese. CM2 cheese had greater concentration of protein and ash and lower fat content as compared to CM1, which could be the reason for difference in their instrumental textural properties. Further, difference in the extent of protein denaturation during vacuum concentration process is also attributed for greater values of hardness and chewiness in CM2 cheese than CM1 cheese. Table 3 also indicated that the UF cheeses were harder than their respective CM cheeses. Denaturation of whey proteins during heat treatment of milk and their subsequent inclusion in the cheese and higher moisture and less proteins as compared to UF cheeses could be the reasons for lower hardness in CM cheeses as compared to UF cheeses. The cohesiveness, adhesiveness, and springiness values of all the cheeses were statistically at par (p < 0.05) indicating no differences in these characteristics of all the cheeses. Overall the instrumental TPA was useful to characterize difference between CM cheeses as compared to UF cheeses.

Table 3 Instrumental texture profile analysis of processed cheeses

Parameter	C	CM1	CM2	UF1	UF2
Hardness (N)	52.95c	62.32c	103.11b	125.29a b	148.81a
Cohesiveness	0.13a	0.13a	0.12a	0.12a	0.12a
Springiness	0.12a	0.12a	0.12a	0.10a b	0.09b
Gumminess (N)	7.18d	9.44c d	12.47b c	15.19a b	17.66a
Adhesiveness	2.96a	7.55a	7.25a	6.05a	7.00a
Chewiness (N)	0.82b	0.93b	1.61a	1.47a	1.69a

^a,b,c,dValues within a row not sharing a common superscript significantly differed (p < 0.05).

Stress Relaxation

Peleg Model

The stress required to deform a substance per unit strain is known as Modulus of elasticity. Higher value of modulus of elasticity indicates that the cheese possesses more hardness. The Initial modulus of elasticity (S ₁, in Peleg model) was related to the strength of cheese.17 The visco-elastic behavior in terms of Peleg Model of the cheeses as a function of stress relaxation is presented in Table 4. It is noted from the table that UF2 cheese showed significantly higher (p < 0.05) S ₁ value as compared to other cheeses. The cheeses made from UF milk had higher values of S ₁ than their counterparts made from condensed milk. This indicates that the UF cheeses were harder than CM cheeses. CM2 and CM1 cheeses contained higher moisture and lesser protein as compared to their respective counterparts, i.e., UF2 and UF1 cheeses, which might be the reason for reduced hardness in the former cheeses. The outcome of Peleg model confirms the results obtained from instrumental TPA analysis of the cheeses, which indicated higher values of hardness in UF cheeses.

Table 4 Viscoelastic parameters of Peleg model in stress relaxation of processed cheeses

Parameter	C	CM1	CM2	UF1	UF2
Peleg B	0.4689a	0.3838b	0.4286a b	0.4233a b	0.4577a
Correlation (γ)	0.9783b	0.9806b	0.9859a	0.9849a	0.9865a
S1 (Pa)	8,489c	6,671d	12,788b	14,253b	21,816a
S2 (Pa)	2,758c	2,028c	5,696b	6,295b	10,160a

^a,b,cValues within a row not sharing a common superscript significantly differed (p < 0.05).

Value of B indicates the time taken by a cheese to dissipate the stress to a value of 50% of the initial amount of stress. The results of Table 4 showed that the value of B was statistically at par in all the cheese samples, except the CM1 cheese, in which the B was lower. However, it was noted that the values of B were higher (though statistically not significant) in UF cheeses as compared to CM cheeses, indicating that the UF cheeses took longer time to relax the given amount of stress to the value of initial stress as compared to the CM cheeses. Although the springiness values of all the cheese were statistically same, the higher values of hardness and gumminess in UF cheeses as compared to CM cheeses are matched with their respective B values. When the B values were compared for variation within UF as well as CM treatments, it was observed that the protein content of the cheeses was responsible for their B values, i.e., cheeses with higher protein (UF2 vs. UF1 and CM2 vs. CM1) had more hardness. The same behavior is also reflected from the data of instrumental TPA analysis, i.e., UF2 and CM2 cheeses had higher values (though statistically similar in some cases) of hardness as well as gumminess as compared to UF1 and CM1 cheeses respectively.

The residual elastic modulus (S ₂) is considered as a factor that determines the viscoelastic nature of the substance.17 Higher S ₂ values means the material behaves more like a viscoelastic in nature. The S ₂ values of cheeses were different and the UF2 had maximum, followed by UF1/CM2, and C/CM1 cheeses. This behavior indicated that UF2 cheese was more viscoelastic in nature as compared to other cheeses. The higher values of S ₂ in UF2 and CM2 as compared to UF1 and CM1 respectively reflect their viscoelastic nature. This behavior is also confirmed from their respective values of G′ and G′′, which were also higher in UF2 and CM2 cheeses. The mean correlation coefficient γ for the model was more or less the same in all the cheeses, which indicated that the residual factor S ₂ alone determined the viscoelastic nature of the cheese. Values of both B and S ₂ together usually characterized the viscoelastic nature of material. Overall it was observed that Peleg model was useful to characterize stress relaxation behavior of both the types of cheeses and the results differentiated between CM1 and CM2 as well as between UF1 and UF2 cheeses.

Maxwell Model

UF2 cheese showed maximum (p < 0.05) value of E ₁ followed by UF1/CM2 and C/CM1 cheeses (Table 5). The value of E ₁ indicates hardness of cheese.17 Accordingly, the UF2 cheese was harder as compared to rest cheeses. The same behavior is also reflected from the results of TPA analysis, which indicated that hardness of the UF2/UF1 cheeses was the maximum as compared to rest of the cheeses. This result is also confirmed from the outcome of Peleg model (S ₁) of UF2 cheeses. The stress relaxation behavior of various cheeses as shown in Fig. 1 also indicates that UF2 cheese possessed maximum value of stress and it required longer time to relax the stress as compared to rest cheeses. Peleg and Normand22 pointed out that the Maxwellian equation containing at least three exponential decay terms has excellent fit of normalized and linearized form. If the number of elements in the Maxwell model is increased, the prediction of data would be more precise but at the same time precise interpretation of such data would be more difficult. The calculated parameters of the six-element Maxwell model represented functional variability of the cheese, since these are functions of time, temperature, and storage period.23 Correlation between the experimental and predicted models was very high in all cheeses (Table 5). The elastic modulus E ₂ and E ₃ constitute residual modulus of elasticity and hence their higher values indicate that they are more viscoelastic solid in nature. The value of E ₂ was highest in UF2 followed by UF1/control and CM1/CM2 cheeses. However, UF2, CM2, and control cheeses possessed statistically similar value of E ₃, which were greater (p < 0.05) than UF1 and CM1 cheeses. This indicated that the former cheeses were more viscoelastic solid in nature as compared to the later cheeses. The Maxwell model very well differentiated the UF and CM cheeses with regard to their modulus of elasticity.

Table 5 Viscoelastic parameters of six elements Maxwell models in stress relaxation of processed cheeses

Parameter	C	CM1	CM2	UF1	UF2	SE
E 1 (Pa)	6,254c	5,596c	11,907b	11,386b	16,277a	1525
E 2 (Pa)	2,872b	432c	731c	2,742b	4,437a	666
E 3 (Pa)	104a	36b	105a	80b	116a	26
t 1 (s)	151b	114c	175b	224a	222a	13.5
t 2 (s)	2.58b	1.42c	0.99c	6.24a	6.50a	0.51
t 3 (s)	0.63a	0.66a	0.41a b	0.43a b	0.14b	0.13
Correlation (γ)	0.985	0.982	0.981	0.990	0.988

SE = Standard error.

^a,b,c,dValues within a row not sharing a common superscript significantly differed (p < 0.05).

Figure 1. Stress relaxation behavior of various processed cheeses. 1 = UF2 cheese, 2 = UF1 cheese, 3 = CM2 cheese, 4 = control, and 5 = CM1 cheese.

Conclusions

High protein processed cheeses (UF2 and CM2) possessed comparatively higher ash and lower fat contents. These cheeses also showed higher values of viscoelastic characteristics in terms of elastic and viscous modulus. However, other textural and rheological characteristics were not consistent in CM2 cheese, these characteristics were repeatedly matched in UF2 cheese. The UF2 cheese showed higher values of instrumental TPA hardness (149 N), and modulus of elasticity S ₁ and E ₁ respectively in Peleg model (21.8 kPa) and Maxwell model (16.2 kPa). It was also revealed that the results of instrumental TPA hardness well matched with the results of elastic and viscous moduli obtained from dynamic rheology test and the values of modulus of elasticity obtained from Peleg and Maxwell models of stress relaxation test. Other TPA attributes like cohesiveness, chewiness, springiness, gumminess etc. were not well matched with the outcome of rheological tests. The results of dynamic rheology test and instrumental TPA were useful to differentiate between the two CM cheeses only, however the results of stress relaxation tests (Peleg and Maxwell models) were more useful because the outcome of the latter tests was also useful to differentiate between UF cheeses.

Abbreviation

TPA	=	Texture profile analysis
UF	=	Ultra filtered
CM	=	Condensed milk