Biscuit dough is a complex material resulting from the mixing of various ingredi- ents, such as wheat flour, water, sugar, and fat. During the mixing process, spe- cific interactions between these constituents (and especially gluten proteins) de- velop, giving their structure to the dough. In the following, the viscoelastic behavior of a biscuit dough is characterized in small-amplitude oscillatory shear experiments. These measurements permit one to follow the development of the structure during the kneading process and to check the influence of the process parameters on dough quality.
2. Materials and Methods
The dough formula we studied corresponds to a semisweet biscuit. On a flour weight basis, it contains 18% fat, 35% sugar, and 2.4% leavening agents. Water content is 23%. All the doughs were made using an experimental horizontal-type mixer with a mixing capacity of 6 kg of dough. The various ingredients of the dough formula are preliminarily homogenized at low speed for 2 minutes, then mixed at high speed for a fixed time. Process parameters are the rotation speed (50–120 rpm), the mixing time (200–700 s), and the regulation temperature of the double jacket. After kneading, the dough is left at rest for 30 minutes before rheological measurement.
The rheological measurements were carried out at 20°C using a parallel- plate rheometer in oscillatory mode. To avoid slip at the wall, which may arise from the presence of fat in the formulation, grooved plates were used. The follow- ing tests were made:
Deformation sweeps, at controlled frequency, to quantify the linear domain Frequency sweeps, in the linear domain, between 0.1 and 100 rad/s Deformation loops, to characterize the dough structure (89)
At a fixed frequency, oscillatory measurements are made for increasing deforma- tions, until a maximum value, then for decreasing deformations. The nonrevers- ibility of the trajectories is characteristic of the structural changes experienced by the dough.
3. Results and Discussion
a. Linear Domain. As for bread doughs (7, 72, 91, 92), the linear visco- elasticity domain is extremely reduced. It can be seen inFigure 7that this domain is below 0.02% deformation. Above this value,G′andG″decrease rapidly, indi- cating a high sensitivity of the structure to the deformation. However, the torque
Figure 7 Linear viscoelasticity domain of a biscuit dough (ω⫽1 rad/s).
signal remains sinusoidal up to 0.2% and thus allows a correct quantification in this range of deformation, as explained by Berland and Launay (72).
b. Frequency Sweeps. These are performed at 0.2% deformation, start- ing from the highest frequencies. An example of a result is presented inFigure 8.Over the whole range of frequencies, the storage modulusG′is higher than the loss modulusG″. Moreover, at low frequencies, both moduli tend toward an equilibrium value, independent of frequency. This behavior is characteristic of a weak physical gel (93) and is evidence of the structured nature of the dough.
Figure 8 Mechanical spectrum of a biscuit dough (γ⫽0.2%).
The complex viscosityη* obeys a power law without any Newtonian pla- teau at low frequency, which is characteristic of a yield stress fluid (94). From the data presented inFigure 8,we can deduce the following values: consistency K⬇15,000 Pa-s, power law indexn ⬇0.29.
c. Deformation Loops. In a previous study (95), Contamine et al. pro- posed considering the crossover ofG′andG″in a deformation sweep as an index of the structural properties of the dough. Because this crossover occurs in the nonlinear domain, it seems more accurate to perform deformation loops (which means an increase in deformation, followed by a decrease and a return to initial conditions) and to characterize, close to the linear domain, the drop of modulus resulting from the imposed deformations.
OnFigure 9,trajectories during increasing and decreasing deformation are different, and the moduli have undergone a drop of approximately 60% (for a maximum deformation of 80%). This means that the interactions between the dough constituents are weak (as van der Waals or hydrogen bond interactions) and can easily be destroyed by small deformations. In fact, the drop in moduli directly depends on the maximum deformation experienced by the dough and increases regularly with the deformation.
Figure 10shows more explicitly the change in the drop of the moduli (that is,∆G′/G′0and∆G″/G″0, whereG′0andG″0are the initial values at 0.2% deforma- tion) with the maximum deformation. Even for limited deformations, the drop is important (for example, 8% onG′for 2.5% deformation), which is evidence of the ‘‘weakness’’ of the structure. The drop continues to increase progressively and seems to reach a stabilized value around 60%, for deformations higher than 60–70%. We may observe that these changes are similar forG′andG″. If the
Figure 9 Deformation loop (up to 80%) of a biscuit dough (ω⫽1 rad/s).
Figure 10 Drop ofG′(䊊) andG″(䊉) moduli as a function of the maximum applied deformation.
dough is allowed to rest for a certain time after the return trajectory, we observe a partial recovery of the viscoelastic moduli, which proves that the structural modifications are not totally irreversible (89). Similar changes have also been reported by Berland and Launay (72) on wheat flour doughs.
d. Influence of Processing Conditions. Such amplitude oscillatory shear measurements allow one to characterize the viscoelastic behavior of dough and its structural properties. We can apply these techniques for understanding the mixing process and the influence of the main parameters on the dough behavior.
For example, we present the influence of kneading time after a preliminary mix- ing of the different ingredients during 3 minutes at low speed.Figure 11shows the evolution of the loss tangent (tan ϕ ⫽ G″/G′) with frequency. It may be observed that the principal modifications concern the behavior at low frequency.
Just after premixing (mixing time⫽ 0), we note an increase of tanϕbelow 1 rad/s. During the mixing (after, respectively, 1.5, 6, and 12 minutes), a plateau appears, which indicates the development of the viscoelastic structure of the dough. Similar results were reported by Amemiya and Menjivar (8) on a wheat flour/water dough. It has been shown that an increase in rotation speed or in mixer regulation temperature has the same qualitative effect as an increase in mixing time (96).