Influence of vibration on structure-rheological properties of highly ...
RheoFuture Young Scientist Award
Paper 1
Influence of vibration on structure-rheological properties of highly concentrated suspension in steady shear flow Dr. B. Ouriev, Prof. N. Ur'ev, Dr. P. Braun This is a part of industrial research project for development of material characterization and implementation of Rheology in material process development.
Problem statement and research goal Characterization of flow effects and their correlation with flow properties of multiphase fluid systems is a key solution for know-how of the flow process development. How to generate and purposeful use flow effects? This is a core question in our research with the goal to accelerate, optimize and control structure formation in flow. Wide range of industrial fluid treatment technologies such as milling, dispersing and disagglomeration, extrusion, casting, calendaring and mixing processes are often have a variety of hidden flow effects. Such flow effects are: slip-stick, memory effect, shear heating, dynamic sedimentation, separation and forced aggregation, fracture and etc. Such uncontrollable flow effects strongly influence energy consumption of the fluid treatment and as most important they disapprovingly contribute to characteristic quality parameters of the fluid. As alternative to actual development procedure of fluid processing we introduce Process - Structure - Rheology - Control concept. As illustrated in FIG. 1 this concept is based on application of the controllable flow effects in combination with structure characterization technique. Actual S ituation
Process
Hidden Effects + Controllable Effects
Product Quality Technological Problem
S olution
Product structural- properties Flow
Hidden Effects
+
Controllable Effects
On-line monitoring Flow control Flow effect control Flow behaviour control
Novel Technology
In-Line Rheometry
UVP-PD and Reference
characterization
FIG. 1: Schematics of conventional development procedure for fluid processing and proposed alternative solution which is based on combination of rheology, rheometry, controllable flow effects and on-line characterization.
Paper 1: 1 / 10
We focus our aim on selection and characterization of the way to generate flow effects. Subsequently, control of desired flow properties narrow selection of flow effects. Finally we introduce selected flow effects as controllable fundamental bases for development of fluid treatment technology. Shown above conceptual idea consider precise characterization of flow properties priory to layout of process flow parameters. Most common way of fluid structure characterization under defined boundary conditions is application of Rotational Rheometry. In fact, while measuring integral value of torque or speed in rotational experiment we impose shear forces on fluid structure and often lead to uncontrollable flow effects whether on micro or macro scale. Thus we influence and observe structure formation simultaneously while varying flow velocity, sample temperature or pressure. Similar situation can be often found in industrial fluid processing. All of the mentioned common sources of energy, eider temperature, pressure or velocity can be hardly considered as individually controllable parameters. Increase of flow velocity is followed with pressure increase and usually followed with shear-heating effect in fluid sample or can also lead to uncontrollable slip-stick effect and unpredictable velocity or stress pulsations and finally to quality problems of treated fluid. There are many examples of such chain reaction between conventional process parameters, flow effects and fluid flow properties can be found in Industry. Our goal is to design Process - Structure - Rheology - Control concept with following content: Rheology: To generate and investigate different flow effects and their influence on time-dependent and time-independent rheological properties. Structure: To study structure formation in combination with particularly selected flow effect and to validate how flow effect influences fluid structure formation. At this step we define fluid structure related control parameters and physical quantities which mostly reflect structural state of processing fluid. Process: Based on rheology-structure background we design flow process geometry and define art of processing with individual process parameters. Control: We select, develop and test on-line characterization technique. Such on-line technique is prepared for process feedback control. In this paper we introduce one of example of Process - Structure - Rheology - Control concept application to highly concentrated multiphase system. To generate desired flow effect we select mechanical vibration and impose it to shear flow. Source of mechanical vibration can be controlled independently on process flow rate, pressure or temperature. Also vibration can be applied at the surface of flow channel or in volume of processed fluid and can be of drag or pressure driven nature [8-14, 18, 19]. Combination of flow type and parameters of oscillations can be selected and adjusted upon requirements of particular application case. We propose to use combination of mechanical vibration, shear and non-invasive on-line characterization technique for design and development of low energy flow processing with controllable fluid structure treatment. Introduction In presented work influence of mechanical vibration on flow properties of highly concentrated multiphase food system under conditions of drag shear flow is investigated. Experiment was performed on chocolate suspension with the help of modified conventional rotational rheometer. Rheometer was equipped with self-designed high intensity vibration source and vibration isolation system. Time dependent and time independent rheological tests were performed in combination with orthogonal to shear direction vibration. It was found that imposed superposition of mechanical vibration and shear flow dramatically decreases shear viscosity level. Comparison between shear viscosities at specified shear rates with and without vibration shows differences in viscosity level up to 80 times beside conversion of chocolate properties from strongly shear-thinning to Newtonian type of flow behavior. It was also found that slight shear-thickening effect induced at high shear rates under influence of vibration in flow of shear-thinning suspension. Taylor analyses shows that such viscosity increase at high shear rates could not be explained by Taylor effect which is typical for low viscous systems in flow between coaxial cylinders. Also wall slip effect under influence of vibration was investigated. The authors introduce simplified procedure for wall slip detection. This procedure based on analyses of normal force acting at rotating cylinder in Couette geometry. Paper 1: 2 / 10
The analysis of boundary conditions shows that no side effects as wall slippage or Taylor effect were present during shear experiment with vibration. Thus we exclude side effects from discussion of our results and consider strong influence of shear and vibration on structure of chocolate suspension. Introduced results show reach technological potential of non-conventional influence on flow of concentrated suspensions. Results of shear experiment under vibration were analysed using power law model. A shape of velocity profile was estimated from result of the fit into rotational rheometer data. Calculated velocity profile was compared to results of on-line fit into measured stationary shear flow velocity profile [7]. Comparison shows that vibration can considerably change shape of velocity profile without alternating flow rate, temperature or composition of the suspension. Thus forces acting on structure of suspension in flow can be steered independently of typical process parameters as flow velocity, process temperature, suspension composition or texture. Method A vibration system was adapted to commercial rotational rheometer (MCR 500, Physica), see FIG. 2. It provides focused vibration energy transfer into investigated fluid probe. Measuring geometry consist of two coaxial cylinders. As described in work of Uriev [18, 19], the outer cylinder is mounted at the electromagnetic shaker axis [14]. As discussed above, measurements of shear viscosity function is performed under influence of vibration in the range between 1 Hz and 6 kHz frequency and amplitudes in the range between 0.05mm and 2mm peak-to-peak. For Rheometer protection a broad frequency band damping system with incorporated vibration isolation was invented to prevent transfer of 2 kN vibration force onto rheometer housing and electronics. Two different test sequences are introduced. During first test sequence, shear viscosity measured in a wide shear rate range. During the second test sequence suspension was investigated under constant shear rate conditions. In both cases liquid was exposed to orthogonal vibration during the measurement. Rheology
F M Ω
ω/Α
Geometry
Vibration
Source of vibration
FIG. 2: Schematics of rotational cylinder - cylinder geometry with orthogonal vibration source.
Measured quantities are: force and amplitude of oscillation at the cup and bob respectively. All sensors were synchronized and signals were triggered using high sampling rate data acquisition. Simultaneously with shear viscosity measurements, the force transfer from vibrating cup to the rotating bob was determined. Also phase shift and force damping between stationary cup and rotary bob were recorded and analyzed. Material A milk chocolate was used for rheological tests under vibration. Temperature during the test -1 -1 was controlled at the level of 28°C, shear rates were adjusted between 0.01s ÷ 100 s . The sample was loaded in cylinder-cylinder geometry of Ri = 26.6 mm and Ro = 28.92 mm, where Ri and Ro are inner and outer cylinder radii respectively. Analyses of boundary conditions Simplified procedure to detect wall slippage in rotational rheometer Various techniques are available for recognition and characterization of wall slip [1, 2, 4, 5, 15 – 17, 18] especially if measured with Couette geometry. One of examples of such technique is Mooney method that was used by Qiu and Rao [15], also described by Steffe [17], to evaluate slip in apple souse. The investigators found that wall slip correction increases consistency coefficient of the fluid. In earlier work Paper 1: 3 / 10
Oldroyd [6] suggested to consider wall slip in terms of general expression for angular velocity Ω that is a function of wall shear stress at the bob and the cup In our work detection of wall slippage was simplified in order to reduce number of tests and to remain experiments with single cylinder geometry [7, 14]. At first wall slippage was experimentally simulated. The walls of cup and bob were lubricated with silicon coating. Chosen lubricant is Silicon spray (GE Bayer Silicones). Shear viscosity of lubricating layer varies slightly around 0.4 ± 0.01 Pas -1 -1 and fluid exhibits Newtonian flow behaviour between 1 s and 1000 s . Lubricant is slightly sensitive to temperature increase in the range between 25 and 40 °C. At 25 °C η = 0.51 Pas and linearly decreases to η = 0.32 Pas at 40 °C. For investigation of the wall slippage we choose strongly shear-thinning fluid as bulk material and low viscous Newtonian fluid as boundary layer. We do not consider subject of particle migration within low viscous boundary layer. The only usage of the layer is to experimentally simulate wall slip effect and it influence on bulk rheological properties of the concentrated suspension. This idea reflects widely used procedure in industrial polymer extrusion applications [1]. For wire coating processes extrusion dies are lubricated with low viscous Newtonian fluid with the purpose of pressure loss reduction. As already mentioned, we generate a wall slippage using modified boundary layer at the bob and cup geometry. For that silicon lubricant was sprayed at the bob and cup surfaces prior measurement of chocolate suspension. Using normal force controlled positioning of measuring geometry the inner rotary bob was introduce into the chocolate sample without damage of the boundary layer. Result of thixotropic loop experiment is introduced in FIG. 3. Shear rate gradually increases -1 -1 -1 from 0.01 s to 1000 s (marked with "Shear flow, up") and decreases again down to 0.01 s (marked with "Shear flow, down"). Shear viscosity is recorded simultaneously with normal force measurement. Pushing normal force F acts axially on rotary shaft and shows positive values and pulling normal force shows negative values, see FIG. 3. Following shown in FIG. 3 data we conclude sudden decrease of normal force which is -1 followed with abrupt shear viscosity decrease at shear rate 500 s . The authors consider this shear rate level as a starting point for wall slippage between suspension sample and boundary layer. From -1 shown data we also conclude that induced wall slippage remains irreversible unless 2 s shear rate is reached. For consideration of wall slippage we calculate ∆F from data shown in FIG. 3 and FIG. 5. For data shown in FIG. 3 ∆Fslip is calculated as ∆Fs = Fup - Fdown. Where Fup and Fdown, slip are normal forces measured during thixotropic loop experiment during increase and decrease of shear rates, respectively. Slip
1000
1 Shear flow, up Shear, wall slip, down Fup, shear flow, up Fdown , wall slip, down
100
0.8 0.6
0.2
F , [N]
η, [Pas]
0.4
0
10
-0.2 -0.4
1 0.01
-0.6
0.1
1
.
γ, [1/s]
10
100
1000
FIG. 3: Shear viscosity as a function of shear rate with and without presence of wall slip effect. Normal force signal indicate appearance of wall slippage and flow fracture at critical sear rates.
Below our conclusions regarding experiment with lubricant layer are summarized: -1 -1 Wall slippage begins at 500 s and remains irreversible unless 2 s limiting shear rate is reached.
Paper 1: 4 / 10
Shear viscosity at downstream shear rates decreases stronger than viscosity values measured at upstream shear rates. Sudden decrease of normal force together with abrupt decrease of shear viscosity reflects appearance of wall slippage.
Whether shear-thickening behaviour under shear and vibration is a result of Taylor effect As we have already discussed, shown in FIG. 6 curve (2), increase of shear viscosity above -1 10 s shear rate can be referred to Taylor instabilities. For low viscous fluids this effect is postulated at Ta numbers Ta > 41. Therefore taking lowest shear viscosity level shown in FIG. 6 curve (2) and using Taylor analyses of flow instabilities in Couette geometry we can calculate Taylor number, Ta, as:
ΩRi 2 ρ (Ro − Ri ) Ta = η∞ 1
3
2
< 41 ,
(1)
where ρ denotes density, Ri and Ro are radii of inner and out cylinders, respectively, and η ∞ is a Newtonian shear viscosity. To calculate rotational velocity for known cylinder geometry and shear rate, the following expression should be used:
γD =
π ⋅ Ω 1+ δ 2 , ⋅ 30 δ 2 − 1
(2)
where, δ = Ro/Ri = 1.0872. Using above equations we conclude that at calculated Ta = 1.77 [.] no flow instabilities can be generated (Ta < 41). We can also conclude with high probability that increase of shear viscosity (See FIG. 6, curve 2) coursed by hydrodynamic forces leading to forced aggregation under influence of shear flow and reduced viscosity. This effect can be presented as vibration induced thickening under shear and vibration. Remaining explanation of slight viscosity increase can be in development of wall slippage. Wall slip effect could be considered as a function of flow velocity and vibration frequency. Therefore an analysis of possible slippage at the wall of vibrating and rotating cylinders was performed. Wall slippage in rotational rheometer under influence of vibration We already considered evidence of wall slip effect and it influence on normal force measurement in Couette geometry. For consideration of wall slippage we calculate ∆F from data shown in FIG. 3 and FIG. 5. For data shown in FIG. 5 ∆F is calculated as ∆F = Fnv - Fv. Where Fv and Fnv are normal forces measured during shear viscosity measurements with and without applied mechanical vibration, respectively. The comparison between calculated ∆F values is introduced in FIG. 5. Vibration start
1'000
η, [Pas]
-2 Shear without vibration Shear with vibration: 60 Hz, 0.5 mm Fnv : no vibration Fv : 60 Hz, 0.5 mm
-2.2 -2.4
100 -2.6
10
1 0.0001
F, [N]
10'000
-2.8
0.001
0.01
0.1 1 . γ, [1/s]
10
100
-3 1000
FIG. 4: Shear viscosity as a function of shear rate with and without presence of vibration. Normal force signal shows no differences between pure shear and shear with vibration. From data shown in FIG. 4 and FIG. 5 follows that no difference were found between normal force measured in shear and shear with vibration. This evidence confirms that no wall slippage is present during experiment of shear flow with vibration. Thus decrease of shear viscosity under influence of vibration can only be related to disaggregation of suspension structure. Paper 1: 5 / 10
1 ∆Fs, shear with vibration ∆Fv, shear with wall slip
0.8 0.6
∆F , [N]
0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 0.001
0.01
0.1
1 . γ, [1/s]
10
100
1000
FIG. 5: Comparison between normal force difference ∆F of suspension sheared under vibration and without vibration and suspension sheared with and without wall slip. Results Reference rheological investigation of concentrated chocolate suspension was done prior to experiments with orthogonal vibration. As shown in FIG. 6 curve (1), typical for molten chocolate strongly shear-thinning flow behaviour was recorded. We select typical for our chocolate processing shear rate window and apply Power law fit. Selected part of fitted flow curve is shown in FIG. 6 curve (3). The result of the fit is taken as a reference Power law index, n = 0.265. Under applied vibration in shear flow, chocolate suspension approaches Newtonian flow behaviour, see FIG. 6, curve (2). Index n is also calculated from the fit into flow curve (4), FIG. 6. The resulting n = 0.895 in reference shear rates window equal. For Newtonian fluid n = 1. These results will be used later for calculation of the flow velocity profile of chocolate suspension in pressure driven shear flow. 1000
1000 1 Strongly NonNewtonian (shear thinning)
100 n = 0.2651
3
10
τ, [Pa]
η, [Pa·s]
100
4
10
n = 0.8953
1
2 Newtonian
1 0.001
0.1 0.01
0.1
. [1/s] γ,
1
10
100
FIG. 6: Shear viscosity, η (γD ) , as a function of shear rate, result of experiment with and without orthogonal
vibration: (1) η (γD ) , no vibration applied; (2) η (γD ) , vibration applied, f = 50 Hz, a ≈ 0.5 mm; (3) τ (γD ) , correlated with curve (1) and (4) τ (γD ) , correlated with curve (2).
Magnitude k of shear viscosity decrease under influence of vibration is shown in FIG. 7. The factor k highlights difference between reference shear viscosity and the one measured under vibration. Two slopes of k were found if plotted against shear rate. These slopes cross each other at -1 approximately 1 s , as shown with solid lines in FIG. 7. The magnitude of viscosity decrease varies -1 -1 between k = 50 at shear rate 0.04 s and k = 1.07 at shear rate 100 s . Paper 1: 6 / 10
From shown in FIG. 7 result we conclude that efficiency of vibration is strongly related to shear rate. Strong viscosity decrease will lead to low energy processing. Also one of possible advantages for chocolate processing will be in reduced shear heating effect in flow of vibrated suspension.
k,n ,[.] [.]
100
10
y = 6.3447x-0.6427 R2 = 0.9983 y = 6.0434x-0.3795 R2 = 0.999
1 0.01
0.1
. 1
10
100
γ, [1/s]
FIG. 7: Difference k between shear viscosity curves (1) and (2), shown in Fig. 4. Factor k is plotted as a function of shear rate.
Fast response of material structure on vibration can be concluded from rapid viscosity decrease right after vibration source is activated. As shown in FIG. 7, strong influence of vibration on -1 -1 flow properties of chocolate suspension was measured below 1 s . From shear rates above 1 s , efficiency of vibration decreases. Factor k = 1 show no differences between shear viscosity of -1 suspension sheared with and without vibration. Such condition is reached already at 100 s shear rate. Shear viscosity and yield value as a function of vibration frequency, amplitude and shear rate. Here we discuss two effects; influence of vibration on yield value and vibration induced shearthickening effect. In FIG. 8 result of shear viscosity measurement under vibration of different frequencies and amplitudes is shown. Parameters of mechanical vibration are summarised in figure legend and correspond with given in Table 1 values.
◊ [Pas]
10
η, [Pas] 60Hz, A = 0.988 [mm] 60Hz, A = 0.5 [mm] 30Hz, A = 1 [mm] 30Hz, A = 0.53 [mm]
13.1 30.8
1 0.01
0.1
1
10 100 1000 . γ, [s ] FIG. 8: Shear viscosity measurements of highly concentrated suspension under influence of vibration. . ◊ [1/s] -1
-1
At shear rates above 5 s a slight thickening effect in flow of chocolate suspension become visible, FIG. 8. Mentioned thickening effect appears to be a function of applied frequency and deformation velocity. While increasing frequency of vibration from 30 Hz to 60 Hz the offset of viscosity curves decreases. As shown in FIG. 8 two amplitudes were tested at constant vibration frequency. While increasing amplitude at constant frequency the shear viscosity offset increases. Such increase is also shear rate depended. A crossover point between two viscosity curves at constant frequency is highlighted in FIG. 8. While increasing frequency of vibration the crossover point shifts towards higher shear rates. As Paper 1: 7 / 10
illustrated in FIG. 8 an increase of frequency by a factor of 2 shifts crossover point to 2.3 times higher shear rate. An interesting observation is confirmed with results shown in FIG. 9. The yield value of chocolate suspension is eliminated by means of mechanical vibration. Furthermore, as followed from data shown in FIG. 9, this effect found to be independent on tested vibration frequencies or amplitudes. Table 1: Vibration parameters: frequency against vibration amplitude.
a, [mm]
30 [Hz]
1
•
60 [Hz]
•
0.988 •
0.53
•
0.5 10000 1000
τ , [Pa]
100
10.4 Pa
10
60Hz, a = 0.98 [mm] 60Hz, a = 0.5 [mm] 30Hz, a = 1 [mm] 30Hz, a = 0.53 [mm] No vibration
0.05 1/s
1 0.1
No yield value
0.01 0.01
0.1
1
. γ [1/s]
10
100
1000
FIG. 9: Flow curves of highly concentrated suspension measured with and without influence of vibration.
Now we can transfer this knowledge on pressure driven flow situation. Based on presented data we expect reduced plug flow under influence of vibration in pressure driven shear flow of chocolate suspension. Time dependent behaviour of concentrated suspension under vibration and shear Now we discuss a constant shear rate experiment. The measurement of shear viscosity is performed at -1 constant shear rate 1 s . During the measurement vibration source was activated several times for 14 sec. As shown in FIG. 10, at applied 60 Hz and 0.5 mm amplitude the difference between viscosity level with and without vibration approaches factor k = 10.
η, [Pa·s]
100
t = 2 sec
t = 6 sec
12 18 24
30
60 Hz, 0.5 mm, 1 1/s
10
1 0
6
36
42 48
54
60 66
72
78
84
t, [sec] FIG. 10: Constant shear rate experiment, shear viscosity at shear rate
γD
-1
≈ 1 [s ] shown as a function of time.
Vibration: f = 60 Hz, a ≈ 0.5 mm.
Paper 1: 8 / 10
Analyses of the data in FIG. 10 shows that shear viscosity decreases from highest to it lowest level within 2 sec after vibration source is activated. Retardation of the suspension structure needs approximately 6 sec after vibration is removed. Thus, structure recovery of concentrated chocolate suspension needs considerably longer time in pure shear flow in contrast to disaggregation coursed by mechanical vibration in shear. Such viscosity decrease under vibration and increase after vibration source is deactivated is precisely reproducible in time. Therefore time can be used as reliable process parameter beside frequency and amplitude. Also vibration can be applied for preconditioning of the chocolate suspension prior or during any of the production process steps. Expected influence of vibration on pressure driven shear flow of chocolate suspension On FIG. 11 visualized velocity profile of chocolate suspension is shown. Visualization was performed in-line using ultrasound Doppler anemometer (UVP X2, Met-Flow SA, [7]). Time averaged velocity profile is shown with symbols. Theoretical fit is calculated and shown with solid line. In order to see potential influence of vibration on pressure driven flow we calculate velocity profile from vibration experiment viscosity data, see FIG. 6. Basically we assume that volumetric flow rate is constant and equal Q = 294 l/h. This value of volumetric flow rate is integrated from the fit steady flow velocity profile shown in FIG. 11, highlighted with Herschel-Bulkley fit. Also magnitude of Q meets typical for chocolate processing industrial requirements. Following our discussion about elimination of yield value under influence of vibration we found no need in use of Herschel-Bulkley model for calculation of velocity profile shape. Thus, we reduce calculation of velocity profile under vibration to Power law model. Using already evaluated in FIG. 6 power law exponent n = 0.89 and constant Q, we plot fit velocity profile of shear flow with vibration in FIG. 11. According to shown in FIG. 11 data we conclude that central flow plug region, shape of the velocity profile and maximum flow velocity are strongly affected if vibration applied in pressure driven shear flow at constant volumetric flow rate Q. 210 No yield value
V x (r) [mm/sec]
180 150
Plug Flow
120 90 Raw Data Herschel-Bulkley Fit Under vibration
60 30 0 0
4
8
12
16
20
24
28
32
D d [mm]
FIG. 11: Velocity profiles Vx(r) of chocolate suspension visualized in a process pipe of Dd = 32 mm in diameter and fitted on-line using UVP-PD technique. Symbols represent measured data, solid lines theoretically calculated velocity profiles.
In order to complete interpretation of flow velocity data we consider rheological information as shear stress in the flow pipe channel. Wall shear stress is straight proportional to pressure gradient in steady pressure driven shear flow of pipe section. In our case wall shear rate in steady shear without -1 -1 vibration approaches 30 s at constant Q. Wall shear rate decreases down to ≈ 20 s with increase of power law exponent n under influence of vibration. Consequently, wall shear stress in shear with 60 -1 Hz and 0.5 mm amplitude vibration will decrease from 182 Pa at shear rate 30 s down to 60 Pa at ≈ -1 20 s . As a result pressure gradient will decrease under influence of vibration by the factor of 3 while volumetric flow rate remains constant. This result could be a good argument for development of low energy fluid treatment technology which is based on simultaneous shear and mechanical vibration [814, 18]. Conclusions According to presented results the following conclusions can be drawn. At first, orthogonal to flow direction mechanical vibration modify molten chocolate from strongly shear thinning to almost Newtonian fluid. Slight thickening effect could be found in suspension under influence of shear and vibration. Increase of shear viscosity at high shear rates cannot be related to Taylor effect due to low Ta number at measured shear rates. An analysis of wall slippage in Couette geometry using normal force procedure was performed. No wall slippage could be detected during the measurement of shear viscosity of chocolate Paper 1: 9 / 10
suspension under shear and vibration. Thus decrease of shear viscosity under influence of vibration shown in FIG. 6 and FIG. 4 is related to disaggregation of suspension structure. This is in a good agreement with results of Uriev [18, 19]. Efficiency of mechanical vibration k is a function of shear rate, FIG. 7. Two shear rate ranges can be selected. In first range influence of mechanical vibration on suspension structure rapidly -1 decreases while shear rate approaches 1 s . Within second range efficiency of vibration slowly decreases until no differences between reference shear viscosity and the one measured under -1 vibration is detected. The magnitude of viscosity decrease varies between k = 50 at shear rate 0.04 s -1 and k = 1.07 at shear rate 100 s . Analyses of constant shear rate experiment in FIG. 10 shows that shear viscosity decreases from highest to its lowest level within 2 sec after vibration source is activated. Retardation of the suspension structure needs approximately 6 sec after vibration is removed. According to presented data, chocolate suspension needs considerably longer time in pure shear flow to recover after vibration removed as to disaggregate right after vibration is activated. Viscosity decreases under vibration and increases after vibration removed in precisely reproducible manner. Characterization of chocolate suspension flow velocity profiles in steady pressure driven shear flow is introduced. Calculation of the velocity profiles for steady shear flow with and without vibration were done under assumption of constant flow rate Q = 294 l/h. According to shown in FIG. 11 data we conclude that central flow plug region, shape of the velocity profile and maximum flow velocity are strongly affected if vibration applied in pressure driven shear flow at constant Q. Beside velocity information also wall sear stress and pressure difference were analyzed. According to our expectations, under influence of vibration pressure loss will decrease by a factor of 3 at constant flow rate Q. This is a good argument for development of low energy fluid treatment technology which is based on shear and mechanical vibration [8-14, 18]. The subject of vibration modulation, orientation to flow and magnitude of vibration parameters and their influence on rheology of concentrated suspensions, emulsions and polymers in drag and pressure driven flows is selected for our future research activity. Introduced results stipulated development of fundamental research and development of Process - Structure - Rheology - Control concept which was presented in FIG. 1 [8-14, 18] and patented, [8 - 14]. Such development opens an opportunity for precise and controllable justification of multiphase fluid structural properties beside an opportunity to design low energy fluid processing. References 1. Ajdari, A., ‘‘Slippage at a polymer–polymer interface-Entanglements and Associated Friction,’’ C. R. Acad. Sci., Ser. II: Mec., Phys., Chim., Sci. Terre Univers 317, 1159–1163 (1993). 2. Barnes, H.A., 1995, J. Non-Newt. Fluid Mech. 3. Braun, P.; Windhab, E., Proc. 1st Int. Symposium on Food Rheology and Structure, ETH Zürich, 16-21 March 1997; Vincentz Verlag, Hannover, 133-138. 4. Gilmour, J.A., Dlugogorski, B.Z., Kennedy, E.M. and Eager, S., Proceedings of th 26th Australasian Chemical Engineering Conference (Chemeca 98), Port Douglas, Australia (1998). 5. Money, M., (1931), J. Rheol. 2(2): p. 210-222. 6. Oldroyd, J.G. Non-Newtonian flow of solids and liquids. In: Eirich, F.R. (editor). Rheology: Theory and Applications, Vol. 1. Academic Press Inc., New York. Pg 653-682, (1956) 7. Ouriev, B.N., (2000) Ultrasound Doppler Based In-line Rheometry of Highly Concentrated Suspensions. PhD Thesis, ISBN: 3-905609-11-8, Zurich. 8. Ouriev, B., WO PCT/CH 200200103; publication date: 16. January 2003 9. Ouriev, B., DE PCT/101 32 069.8; publication date: 16. January 2003 10. Ouriev, B., WO PCT/CH 200200678; publication date: 17. July 2003 11. Ouriev, B., WO PCT/CH 200300033; publication date: 24. July 2003 12. Ouriev, B., DE PCT/102 022 48, publication date: 31. July 2003 13. Ouriev, B., DE PCT/102 02 238.0; publication date: 31. July 2003 14. Ouriev, B.N., Method for influencing the rheological properties of a liquid by means of oscillations, Bühler PCT application, WO 03/004882 A1, Uzwil, Switzerland 15. Qiu, C-G. and Rao. M.A., J. Texture Stud. 20: 57-70, (1989) 16. Steffe, J.F. 1984. Yield stress: phenomena and measurement. In: Singh, R.P. and A. Wirakartakusumah (editors); Advances in Food Engineering. CRC Press, Boca Raton, FL. Pg. 363-376, (1992) 17. Steffe, J.F., Rheological methods in food process engineering, Freeman Press, Michigan (1998). 18. Uriev, N.B., (1988) Physico-chemical Fundamentals of the Technology of Disperse Systems and Materials, Khimiya, Moscow, p. 256 19. Uriev, N.B., (1992) J. of Colloids and Surfaces, 87, pages 1-4. Paper 1: 10 / 10
Paper 1
Influence of vibration on structure-rheological properties of highly concentrated suspension in steady shear flow Dr. B. Ouriev, Prof. N. Ur'ev, Dr. P. Braun This is a part of industrial research project for development of material characterization and implementation of Rheology in material process development.
Problem statement and research goal Characterization of flow effects and their correlation with flow properties of multiphase fluid systems is a key solution for know-how of the flow process development. How to generate and purposeful use flow effects? This is a core question in our research with the goal to accelerate, optimize and control structure formation in flow. Wide range of industrial fluid treatment technologies such as milling, dispersing and disagglomeration, extrusion, casting, calendaring and mixing processes are often have a variety of hidden flow effects. Such flow effects are: slip-stick, memory effect, shear heating, dynamic sedimentation, separation and forced aggregation, fracture and etc. Such uncontrollable flow effects strongly influence energy consumption of the fluid treatment and as most important they disapprovingly contribute to characteristic quality parameters of the fluid. As alternative to actual development procedure of fluid processing we introduce Process - Structure - Rheology - Control concept. As illustrated in FIG. 1 this concept is based on application of the controllable flow effects in combination with structure characterization technique. Actual S ituation
Process
Hidden Effects + Controllable Effects
Product Quality Technological Problem
S olution
Product structural- properties Flow
Hidden Effects
+
Controllable Effects
On-line monitoring Flow control Flow effect control Flow behaviour control
Novel Technology
In-Line Rheometry
UVP-PD and Reference
characterization
FIG. 1: Schematics of conventional development procedure for fluid processing and proposed alternative solution which is based on combination of rheology, rheometry, controllable flow effects and on-line characterization.
Paper 1: 1 / 10
We focus our aim on selection and characterization of the way to generate flow effects. Subsequently, control of desired flow properties narrow selection of flow effects. Finally we introduce selected flow effects as controllable fundamental bases for development of fluid treatment technology. Shown above conceptual idea consider precise characterization of flow properties priory to layout of process flow parameters. Most common way of fluid structure characterization under defined boundary conditions is application of Rotational Rheometry. In fact, while measuring integral value of torque or speed in rotational experiment we impose shear forces on fluid structure and often lead to uncontrollable flow effects whether on micro or macro scale. Thus we influence and observe structure formation simultaneously while varying flow velocity, sample temperature or pressure. Similar situation can be often found in industrial fluid processing. All of the mentioned common sources of energy, eider temperature, pressure or velocity can be hardly considered as individually controllable parameters. Increase of flow velocity is followed with pressure increase and usually followed with shear-heating effect in fluid sample or can also lead to uncontrollable slip-stick effect and unpredictable velocity or stress pulsations and finally to quality problems of treated fluid. There are many examples of such chain reaction between conventional process parameters, flow effects and fluid flow properties can be found in Industry. Our goal is to design Process - Structure - Rheology - Control concept with following content: Rheology: To generate and investigate different flow effects and their influence on time-dependent and time-independent rheological properties. Structure: To study structure formation in combination with particularly selected flow effect and to validate how flow effect influences fluid structure formation. At this step we define fluid structure related control parameters and physical quantities which mostly reflect structural state of processing fluid. Process: Based on rheology-structure background we design flow process geometry and define art of processing with individual process parameters. Control: We select, develop and test on-line characterization technique. Such on-line technique is prepared for process feedback control. In this paper we introduce one of example of Process - Structure - Rheology - Control concept application to highly concentrated multiphase system. To generate desired flow effect we select mechanical vibration and impose it to shear flow. Source of mechanical vibration can be controlled independently on process flow rate, pressure or temperature. Also vibration can be applied at the surface of flow channel or in volume of processed fluid and can be of drag or pressure driven nature [8-14, 18, 19]. Combination of flow type and parameters of oscillations can be selected and adjusted upon requirements of particular application case. We propose to use combination of mechanical vibration, shear and non-invasive on-line characterization technique for design and development of low energy flow processing with controllable fluid structure treatment. Introduction In presented work influence of mechanical vibration on flow properties of highly concentrated multiphase food system under conditions of drag shear flow is investigated. Experiment was performed on chocolate suspension with the help of modified conventional rotational rheometer. Rheometer was equipped with self-designed high intensity vibration source and vibration isolation system. Time dependent and time independent rheological tests were performed in combination with orthogonal to shear direction vibration. It was found that imposed superposition of mechanical vibration and shear flow dramatically decreases shear viscosity level. Comparison between shear viscosities at specified shear rates with and without vibration shows differences in viscosity level up to 80 times beside conversion of chocolate properties from strongly shear-thinning to Newtonian type of flow behavior. It was also found that slight shear-thickening effect induced at high shear rates under influence of vibration in flow of shear-thinning suspension. Taylor analyses shows that such viscosity increase at high shear rates could not be explained by Taylor effect which is typical for low viscous systems in flow between coaxial cylinders. Also wall slip effect under influence of vibration was investigated. The authors introduce simplified procedure for wall slip detection. This procedure based on analyses of normal force acting at rotating cylinder in Couette geometry. Paper 1: 2 / 10
The analysis of boundary conditions shows that no side effects as wall slippage or Taylor effect were present during shear experiment with vibration. Thus we exclude side effects from discussion of our results and consider strong influence of shear and vibration on structure of chocolate suspension. Introduced results show reach technological potential of non-conventional influence on flow of concentrated suspensions. Results of shear experiment under vibration were analysed using power law model. A shape of velocity profile was estimated from result of the fit into rotational rheometer data. Calculated velocity profile was compared to results of on-line fit into measured stationary shear flow velocity profile [7]. Comparison shows that vibration can considerably change shape of velocity profile without alternating flow rate, temperature or composition of the suspension. Thus forces acting on structure of suspension in flow can be steered independently of typical process parameters as flow velocity, process temperature, suspension composition or texture. Method A vibration system was adapted to commercial rotational rheometer (MCR 500, Physica), see FIG. 2. It provides focused vibration energy transfer into investigated fluid probe. Measuring geometry consist of two coaxial cylinders. As described in work of Uriev [18, 19], the outer cylinder is mounted at the electromagnetic shaker axis [14]. As discussed above, measurements of shear viscosity function is performed under influence of vibration in the range between 1 Hz and 6 kHz frequency and amplitudes in the range between 0.05mm and 2mm peak-to-peak. For Rheometer protection a broad frequency band damping system with incorporated vibration isolation was invented to prevent transfer of 2 kN vibration force onto rheometer housing and electronics. Two different test sequences are introduced. During first test sequence, shear viscosity measured in a wide shear rate range. During the second test sequence suspension was investigated under constant shear rate conditions. In both cases liquid was exposed to orthogonal vibration during the measurement. Rheology
F M Ω
ω/Α
Geometry
Vibration
Source of vibration
FIG. 2: Schematics of rotational cylinder - cylinder geometry with orthogonal vibration source.
Measured quantities are: force and amplitude of oscillation at the cup and bob respectively. All sensors were synchronized and signals were triggered using high sampling rate data acquisition. Simultaneously with shear viscosity measurements, the force transfer from vibrating cup to the rotating bob was determined. Also phase shift and force damping between stationary cup and rotary bob were recorded and analyzed. Material A milk chocolate was used for rheological tests under vibration. Temperature during the test -1 -1 was controlled at the level of 28°C, shear rates were adjusted between 0.01s ÷ 100 s . The sample was loaded in cylinder-cylinder geometry of Ri = 26.6 mm and Ro = 28.92 mm, where Ri and Ro are inner and outer cylinder radii respectively. Analyses of boundary conditions Simplified procedure to detect wall slippage in rotational rheometer Various techniques are available for recognition and characterization of wall slip [1, 2, 4, 5, 15 – 17, 18] especially if measured with Couette geometry. One of examples of such technique is Mooney method that was used by Qiu and Rao [15], also described by Steffe [17], to evaluate slip in apple souse. The investigators found that wall slip correction increases consistency coefficient of the fluid. In earlier work Paper 1: 3 / 10
Oldroyd [6] suggested to consider wall slip in terms of general expression for angular velocity Ω that is a function of wall shear stress at the bob and the cup In our work detection of wall slippage was simplified in order to reduce number of tests and to remain experiments with single cylinder geometry [7, 14]. At first wall slippage was experimentally simulated. The walls of cup and bob were lubricated with silicon coating. Chosen lubricant is Silicon spray (GE Bayer Silicones). Shear viscosity of lubricating layer varies slightly around 0.4 ± 0.01 Pas -1 -1 and fluid exhibits Newtonian flow behaviour between 1 s and 1000 s . Lubricant is slightly sensitive to temperature increase in the range between 25 and 40 °C. At 25 °C η = 0.51 Pas and linearly decreases to η = 0.32 Pas at 40 °C. For investigation of the wall slippage we choose strongly shear-thinning fluid as bulk material and low viscous Newtonian fluid as boundary layer. We do not consider subject of particle migration within low viscous boundary layer. The only usage of the layer is to experimentally simulate wall slip effect and it influence on bulk rheological properties of the concentrated suspension. This idea reflects widely used procedure in industrial polymer extrusion applications [1]. For wire coating processes extrusion dies are lubricated with low viscous Newtonian fluid with the purpose of pressure loss reduction. As already mentioned, we generate a wall slippage using modified boundary layer at the bob and cup geometry. For that silicon lubricant was sprayed at the bob and cup surfaces prior measurement of chocolate suspension. Using normal force controlled positioning of measuring geometry the inner rotary bob was introduce into the chocolate sample without damage of the boundary layer. Result of thixotropic loop experiment is introduced in FIG. 3. Shear rate gradually increases -1 -1 -1 from 0.01 s to 1000 s (marked with "Shear flow, up") and decreases again down to 0.01 s (marked with "Shear flow, down"). Shear viscosity is recorded simultaneously with normal force measurement. Pushing normal force F acts axially on rotary shaft and shows positive values and pulling normal force shows negative values, see FIG. 3. Following shown in FIG. 3 data we conclude sudden decrease of normal force which is -1 followed with abrupt shear viscosity decrease at shear rate 500 s . The authors consider this shear rate level as a starting point for wall slippage between suspension sample and boundary layer. From -1 shown data we also conclude that induced wall slippage remains irreversible unless 2 s shear rate is reached. For consideration of wall slippage we calculate ∆F from data shown in FIG. 3 and FIG. 5. For data shown in FIG. 3 ∆Fslip is calculated as ∆Fs = Fup - Fdown. Where Fup and Fdown, slip are normal forces measured during thixotropic loop experiment during increase and decrease of shear rates, respectively. Slip
1000
1 Shear flow, up Shear, wall slip, down Fup, shear flow, up Fdown , wall slip, down
100
0.8 0.6
0.2
F , [N]
η, [Pas]
0.4
0
10
-0.2 -0.4
1 0.01
-0.6
0.1
1
.
γ, [1/s]
10
100
1000
FIG. 3: Shear viscosity as a function of shear rate with and without presence of wall slip effect. Normal force signal indicate appearance of wall slippage and flow fracture at critical sear rates.
Below our conclusions regarding experiment with lubricant layer are summarized: -1 -1 Wall slippage begins at 500 s and remains irreversible unless 2 s limiting shear rate is reached.
Paper 1: 4 / 10
Shear viscosity at downstream shear rates decreases stronger than viscosity values measured at upstream shear rates. Sudden decrease of normal force together with abrupt decrease of shear viscosity reflects appearance of wall slippage.
Whether shear-thickening behaviour under shear and vibration is a result of Taylor effect As we have already discussed, shown in FIG. 6 curve (2), increase of shear viscosity above -1 10 s shear rate can be referred to Taylor instabilities. For low viscous fluids this effect is postulated at Ta numbers Ta > 41. Therefore taking lowest shear viscosity level shown in FIG. 6 curve (2) and using Taylor analyses of flow instabilities in Couette geometry we can calculate Taylor number, Ta, as:
ΩRi 2 ρ (Ro − Ri ) Ta = η∞ 1
3
2
< 41 ,
(1)
where ρ denotes density, Ri and Ro are radii of inner and out cylinders, respectively, and η ∞ is a Newtonian shear viscosity. To calculate rotational velocity for known cylinder geometry and shear rate, the following expression should be used:
γD =
π ⋅ Ω 1+ δ 2 , ⋅ 30 δ 2 − 1
(2)
where, δ = Ro/Ri = 1.0872. Using above equations we conclude that at calculated Ta = 1.77 [.] no flow instabilities can be generated (Ta < 41). We can also conclude with high probability that increase of shear viscosity (See FIG. 6, curve 2) coursed by hydrodynamic forces leading to forced aggregation under influence of shear flow and reduced viscosity. This effect can be presented as vibration induced thickening under shear and vibration. Remaining explanation of slight viscosity increase can be in development of wall slippage. Wall slip effect could be considered as a function of flow velocity and vibration frequency. Therefore an analysis of possible slippage at the wall of vibrating and rotating cylinders was performed. Wall slippage in rotational rheometer under influence of vibration We already considered evidence of wall slip effect and it influence on normal force measurement in Couette geometry. For consideration of wall slippage we calculate ∆F from data shown in FIG. 3 and FIG. 5. For data shown in FIG. 5 ∆F is calculated as ∆F = Fnv - Fv. Where Fv and Fnv are normal forces measured during shear viscosity measurements with and without applied mechanical vibration, respectively. The comparison between calculated ∆F values is introduced in FIG. 5. Vibration start
1'000
η, [Pas]
-2 Shear without vibration Shear with vibration: 60 Hz, 0.5 mm Fnv : no vibration Fv : 60 Hz, 0.5 mm
-2.2 -2.4
100 -2.6
10
1 0.0001
F, [N]
10'000
-2.8
0.001
0.01
0.1 1 . γ, [1/s]
10
100
-3 1000
FIG. 4: Shear viscosity as a function of shear rate with and without presence of vibration. Normal force signal shows no differences between pure shear and shear with vibration. From data shown in FIG. 4 and FIG. 5 follows that no difference were found between normal force measured in shear and shear with vibration. This evidence confirms that no wall slippage is present during experiment of shear flow with vibration. Thus decrease of shear viscosity under influence of vibration can only be related to disaggregation of suspension structure. Paper 1: 5 / 10
1 ∆Fs, shear with vibration ∆Fv, shear with wall slip
0.8 0.6
∆F , [N]
0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 0.001
0.01
0.1
1 . γ, [1/s]
10
100
1000
FIG. 5: Comparison between normal force difference ∆F of suspension sheared under vibration and without vibration and suspension sheared with and without wall slip. Results Reference rheological investigation of concentrated chocolate suspension was done prior to experiments with orthogonal vibration. As shown in FIG. 6 curve (1), typical for molten chocolate strongly shear-thinning flow behaviour was recorded. We select typical for our chocolate processing shear rate window and apply Power law fit. Selected part of fitted flow curve is shown in FIG. 6 curve (3). The result of the fit is taken as a reference Power law index, n = 0.265. Under applied vibration in shear flow, chocolate suspension approaches Newtonian flow behaviour, see FIG. 6, curve (2). Index n is also calculated from the fit into flow curve (4), FIG. 6. The resulting n = 0.895 in reference shear rates window equal. For Newtonian fluid n = 1. These results will be used later for calculation of the flow velocity profile of chocolate suspension in pressure driven shear flow. 1000
1000 1 Strongly NonNewtonian (shear thinning)
100 n = 0.2651
3
10
τ, [Pa]
η, [Pa·s]
100
4
10
n = 0.8953
1
2 Newtonian
1 0.001
0.1 0.01
0.1
. [1/s] γ,
1
10
100
FIG. 6: Shear viscosity, η (γD ) , as a function of shear rate, result of experiment with and without orthogonal
vibration: (1) η (γD ) , no vibration applied; (2) η (γD ) , vibration applied, f = 50 Hz, a ≈ 0.5 mm; (3) τ (γD ) , correlated with curve (1) and (4) τ (γD ) , correlated with curve (2).
Magnitude k of shear viscosity decrease under influence of vibration is shown in FIG. 7. The factor k highlights difference between reference shear viscosity and the one measured under vibration. Two slopes of k were found if plotted against shear rate. These slopes cross each other at -1 approximately 1 s , as shown with solid lines in FIG. 7. The magnitude of viscosity decrease varies -1 -1 between k = 50 at shear rate 0.04 s and k = 1.07 at shear rate 100 s . Paper 1: 6 / 10
From shown in FIG. 7 result we conclude that efficiency of vibration is strongly related to shear rate. Strong viscosity decrease will lead to low energy processing. Also one of possible advantages for chocolate processing will be in reduced shear heating effect in flow of vibrated suspension.
k,n ,[.] [.]
100
10
y = 6.3447x-0.6427 R2 = 0.9983 y = 6.0434x-0.3795 R2 = 0.999
1 0.01
0.1
. 1
10
100
γ, [1/s]
FIG. 7: Difference k between shear viscosity curves (1) and (2), shown in Fig. 4. Factor k is plotted as a function of shear rate.
Fast response of material structure on vibration can be concluded from rapid viscosity decrease right after vibration source is activated. As shown in FIG. 7, strong influence of vibration on -1 -1 flow properties of chocolate suspension was measured below 1 s . From shear rates above 1 s , efficiency of vibration decreases. Factor k = 1 show no differences between shear viscosity of -1 suspension sheared with and without vibration. Such condition is reached already at 100 s shear rate. Shear viscosity and yield value as a function of vibration frequency, amplitude and shear rate. Here we discuss two effects; influence of vibration on yield value and vibration induced shearthickening effect. In FIG. 8 result of shear viscosity measurement under vibration of different frequencies and amplitudes is shown. Parameters of mechanical vibration are summarised in figure legend and correspond with given in Table 1 values.
◊ [Pas]
10
η, [Pas] 60Hz, A = 0.988 [mm] 60Hz, A = 0.5 [mm] 30Hz, A = 1 [mm] 30Hz, A = 0.53 [mm]
13.1 30.8
1 0.01
0.1
1
10 100 1000 . γ, [s ] FIG. 8: Shear viscosity measurements of highly concentrated suspension under influence of vibration. . ◊ [1/s] -1
-1
At shear rates above 5 s a slight thickening effect in flow of chocolate suspension become visible, FIG. 8. Mentioned thickening effect appears to be a function of applied frequency and deformation velocity. While increasing frequency of vibration from 30 Hz to 60 Hz the offset of viscosity curves decreases. As shown in FIG. 8 two amplitudes were tested at constant vibration frequency. While increasing amplitude at constant frequency the shear viscosity offset increases. Such increase is also shear rate depended. A crossover point between two viscosity curves at constant frequency is highlighted in FIG. 8. While increasing frequency of vibration the crossover point shifts towards higher shear rates. As Paper 1: 7 / 10
illustrated in FIG. 8 an increase of frequency by a factor of 2 shifts crossover point to 2.3 times higher shear rate. An interesting observation is confirmed with results shown in FIG. 9. The yield value of chocolate suspension is eliminated by means of mechanical vibration. Furthermore, as followed from data shown in FIG. 9, this effect found to be independent on tested vibration frequencies or amplitudes. Table 1: Vibration parameters: frequency against vibration amplitude.
a, [mm]
30 [Hz]
1
•
60 [Hz]
•
0.988 •
0.53
•
0.5 10000 1000
τ , [Pa]
100
10.4 Pa
10
60Hz, a = 0.98 [mm] 60Hz, a = 0.5 [mm] 30Hz, a = 1 [mm] 30Hz, a = 0.53 [mm] No vibration
0.05 1/s
1 0.1
No yield value
0.01 0.01
0.1
1
. γ [1/s]
10
100
1000
FIG. 9: Flow curves of highly concentrated suspension measured with and without influence of vibration.
Now we can transfer this knowledge on pressure driven flow situation. Based on presented data we expect reduced plug flow under influence of vibration in pressure driven shear flow of chocolate suspension. Time dependent behaviour of concentrated suspension under vibration and shear Now we discuss a constant shear rate experiment. The measurement of shear viscosity is performed at -1 constant shear rate 1 s . During the measurement vibration source was activated several times for 14 sec. As shown in FIG. 10, at applied 60 Hz and 0.5 mm amplitude the difference between viscosity level with and without vibration approaches factor k = 10.
η, [Pa·s]
100
t = 2 sec
t = 6 sec
12 18 24
30
60 Hz, 0.5 mm, 1 1/s
10
1 0
6
36
42 48
54
60 66
72
78
84
t, [sec] FIG. 10: Constant shear rate experiment, shear viscosity at shear rate
γD
-1
≈ 1 [s ] shown as a function of time.
Vibration: f = 60 Hz, a ≈ 0.5 mm.
Paper 1: 8 / 10
Analyses of the data in FIG. 10 shows that shear viscosity decreases from highest to it lowest level within 2 sec after vibration source is activated. Retardation of the suspension structure needs approximately 6 sec after vibration is removed. Thus, structure recovery of concentrated chocolate suspension needs considerably longer time in pure shear flow in contrast to disaggregation coursed by mechanical vibration in shear. Such viscosity decrease under vibration and increase after vibration source is deactivated is precisely reproducible in time. Therefore time can be used as reliable process parameter beside frequency and amplitude. Also vibration can be applied for preconditioning of the chocolate suspension prior or during any of the production process steps. Expected influence of vibration on pressure driven shear flow of chocolate suspension On FIG. 11 visualized velocity profile of chocolate suspension is shown. Visualization was performed in-line using ultrasound Doppler anemometer (UVP X2, Met-Flow SA, [7]). Time averaged velocity profile is shown with symbols. Theoretical fit is calculated and shown with solid line. In order to see potential influence of vibration on pressure driven flow we calculate velocity profile from vibration experiment viscosity data, see FIG. 6. Basically we assume that volumetric flow rate is constant and equal Q = 294 l/h. This value of volumetric flow rate is integrated from the fit steady flow velocity profile shown in FIG. 11, highlighted with Herschel-Bulkley fit. Also magnitude of Q meets typical for chocolate processing industrial requirements. Following our discussion about elimination of yield value under influence of vibration we found no need in use of Herschel-Bulkley model for calculation of velocity profile shape. Thus, we reduce calculation of velocity profile under vibration to Power law model. Using already evaluated in FIG. 6 power law exponent n = 0.89 and constant Q, we plot fit velocity profile of shear flow with vibration in FIG. 11. According to shown in FIG. 11 data we conclude that central flow plug region, shape of the velocity profile and maximum flow velocity are strongly affected if vibration applied in pressure driven shear flow at constant volumetric flow rate Q. 210 No yield value
V x (r) [mm/sec]
180 150
Plug Flow
120 90 Raw Data Herschel-Bulkley Fit Under vibration
60 30 0 0
4
8
12
16
20
24
28
32
D d [mm]
FIG. 11: Velocity profiles Vx(r) of chocolate suspension visualized in a process pipe of Dd = 32 mm in diameter and fitted on-line using UVP-PD technique. Symbols represent measured data, solid lines theoretically calculated velocity profiles.
In order to complete interpretation of flow velocity data we consider rheological information as shear stress in the flow pipe channel. Wall shear stress is straight proportional to pressure gradient in steady pressure driven shear flow of pipe section. In our case wall shear rate in steady shear without -1 -1 vibration approaches 30 s at constant Q. Wall shear rate decreases down to ≈ 20 s with increase of power law exponent n under influence of vibration. Consequently, wall shear stress in shear with 60 -1 Hz and 0.5 mm amplitude vibration will decrease from 182 Pa at shear rate 30 s down to 60 Pa at ≈ -1 20 s . As a result pressure gradient will decrease under influence of vibration by the factor of 3 while volumetric flow rate remains constant. This result could be a good argument for development of low energy fluid treatment technology which is based on simultaneous shear and mechanical vibration [814, 18]. Conclusions According to presented results the following conclusions can be drawn. At first, orthogonal to flow direction mechanical vibration modify molten chocolate from strongly shear thinning to almost Newtonian fluid. Slight thickening effect could be found in suspension under influence of shear and vibration. Increase of shear viscosity at high shear rates cannot be related to Taylor effect due to low Ta number at measured shear rates. An analysis of wall slippage in Couette geometry using normal force procedure was performed. No wall slippage could be detected during the measurement of shear viscosity of chocolate Paper 1: 9 / 10
suspension under shear and vibration. Thus decrease of shear viscosity under influence of vibration shown in FIG. 6 and FIG. 4 is related to disaggregation of suspension structure. This is in a good agreement with results of Uriev [18, 19]. Efficiency of mechanical vibration k is a function of shear rate, FIG. 7. Two shear rate ranges can be selected. In first range influence of mechanical vibration on suspension structure rapidly -1 decreases while shear rate approaches 1 s . Within second range efficiency of vibration slowly decreases until no differences between reference shear viscosity and the one measured under -1 vibration is detected. The magnitude of viscosity decrease varies between k = 50 at shear rate 0.04 s -1 and k = 1.07 at shear rate 100 s . Analyses of constant shear rate experiment in FIG. 10 shows that shear viscosity decreases from highest to its lowest level within 2 sec after vibration source is activated. Retardation of the suspension structure needs approximately 6 sec after vibration is removed. According to presented data, chocolate suspension needs considerably longer time in pure shear flow to recover after vibration removed as to disaggregate right after vibration is activated. Viscosity decreases under vibration and increases after vibration removed in precisely reproducible manner. Characterization of chocolate suspension flow velocity profiles in steady pressure driven shear flow is introduced. Calculation of the velocity profiles for steady shear flow with and without vibration were done under assumption of constant flow rate Q = 294 l/h. According to shown in FIG. 11 data we conclude that central flow plug region, shape of the velocity profile and maximum flow velocity are strongly affected if vibration applied in pressure driven shear flow at constant Q. Beside velocity information also wall sear stress and pressure difference were analyzed. According to our expectations, under influence of vibration pressure loss will decrease by a factor of 3 at constant flow rate Q. This is a good argument for development of low energy fluid treatment technology which is based on shear and mechanical vibration [8-14, 18]. The subject of vibration modulation, orientation to flow and magnitude of vibration parameters and their influence on rheology of concentrated suspensions, emulsions and polymers in drag and pressure driven flows is selected for our future research activity. Introduced results stipulated development of fundamental research and development of Process - Structure - Rheology - Control concept which was presented in FIG. 1 [8-14, 18] and patented, [8 - 14]. Such development opens an opportunity for precise and controllable justification of multiphase fluid structural properties beside an opportunity to design low energy fluid processing. References 1. Ajdari, A., ‘‘Slippage at a polymer–polymer interface-Entanglements and Associated Friction,’’ C. R. Acad. Sci., Ser. II: Mec., Phys., Chim., Sci. Terre Univers 317, 1159–1163 (1993). 2. Barnes, H.A., 1995, J. Non-Newt. Fluid Mech. 3. Braun, P.; Windhab, E., Proc. 1st Int. Symposium on Food Rheology and Structure, ETH Zürich, 16-21 March 1997; Vincentz Verlag, Hannover, 133-138. 4. Gilmour, J.A., Dlugogorski, B.Z., Kennedy, E.M. and Eager, S., Proceedings of th 26th Australasian Chemical Engineering Conference (Chemeca 98), Port Douglas, Australia (1998). 5. Money, M., (1931), J. Rheol. 2(2): p. 210-222. 6. Oldroyd, J.G. Non-Newtonian flow of solids and liquids. In: Eirich, F.R. (editor). Rheology: Theory and Applications, Vol. 1. Academic Press Inc., New York. Pg 653-682, (1956) 7. Ouriev, B.N., (2000) Ultrasound Doppler Based In-line Rheometry of Highly Concentrated Suspensions. PhD Thesis, ISBN: 3-905609-11-8, Zurich. 8. Ouriev, B., WO PCT/CH 200200103; publication date: 16. January 2003 9. Ouriev, B., DE PCT/101 32 069.8; publication date: 16. January 2003 10. Ouriev, B., WO PCT/CH 200200678; publication date: 17. July 2003 11. Ouriev, B., WO PCT/CH 200300033; publication date: 24. July 2003 12. Ouriev, B., DE PCT/102 022 48, publication date: 31. July 2003 13. Ouriev, B., DE PCT/102 02 238.0; publication date: 31. July 2003 14. Ouriev, B.N., Method for influencing the rheological properties of a liquid by means of oscillations, Bühler PCT application, WO 03/004882 A1, Uzwil, Switzerland 15. Qiu, C-G. and Rao. M.A., J. Texture Stud. 20: 57-70, (1989) 16. Steffe, J.F. 1984. Yield stress: phenomena and measurement. In: Singh, R.P. and A. Wirakartakusumah (editors); Advances in Food Engineering. CRC Press, Boca Raton, FL. Pg. 363-376, (1992) 17. Steffe, J.F., Rheological methods in food process engineering, Freeman Press, Michigan (1998). 18. Uriev, N.B., (1988) Physico-chemical Fundamentals of the Technology of Disperse Systems and Materials, Khimiya, Moscow, p. 256 19. Uriev, N.B., (1992) J. of Colloids and Surfaces, 87, pages 1-4. Paper 1: 10 / 10
