r...“ u . WV;- .4 "F u .1 ud '-‘ 1d“. 44.‘ . ‘. _.I a' ”(v "F?!” .. a‘.‘ chi. . $10391): 9. :I "#53. , "is ’ "CJM' 3’ y ..: r P M- "X W . . V‘ k ’ a. 1‘ l‘ ’5 v. w . “Wt-.139. ' ’3 5h ’ 5 w‘g AR“. 5; J.- K’éi 1"“? - Vagréq‘m A: *2 w- Am”,- 9334'.- 1 $3 u {‘1 A “35%;? -i. ‘43?" ’Jro .71‘ 8.41 ‘ V. . A , . .. L " wp ». 333%flz‘32911wg} ; » ..-. “ . by V 49‘53’ :5: ‘W- « , 9%}! ’5‘? . P . ..e2?‘~1, .‘Lfi‘w $W‘gfl‘x " “‘71?" .. a ( , .r‘. ‘ .f é (V‘IU' if: aid/‘3 \" “4‘ ‘ ‘v a. a x . #nl'u’n mu. . -1) »'r' ' :41! f. - l ’- 7 m3“? <~ m . 9: 4'3! 5 ,m, Ar..— 9 “‘45:: v 'fi J.- ,.,-,c. a v 5 hw-ay- RU: :‘n-V‘ 5%. ‘7 .- v-u a‘ , ‘4‘? I‘M! ‘l 7‘" 1:15;. ' * WFS‘E lllllllllllllllllllllllllllllllllllllll 3 1293 0140 7420 This is to certify that the thesis entitled Yield Stress Measurement of Semi-solid Foods Using the Back Extrusion (Annular Pumping) Method presented by Aunur 'Rofiq Badi has been accepted towards fulfillment of the requirements for M.S. degree in Food Science 33mg F 5/054. Major professor Date QSNO-H i? J O-7639 MS U is an Affirmative Action/Equal Opportunity Institution —_¥ ’ LIBRARY MiChigan State University PLACE ll RETURN BOX to mow thb chockout from your ncord. To AVOID FINES Man on or More data duo. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative MloNEqunl Opportunlty Institution Wanna Yi Yield Stress Measurement of Semi-solid Foods Using the Back Extrusion (Annular Pumping) Method by Aunur Rofiq Hadi A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1989 (anr rati pent yie' non rec err ach Spe dif the -44 ”‘2 Ci» 1.- of) ”.7 LyJ ABSTRACT by Aunur Rofiq Hadi Three major factors in yield stress measurement using back extrusion (annular pumping) were considered: friction factor on plunger surface, ratio of plunger to sample tube radius, plunger velocity and depth of penetration. A lubricated plunger compared to an unlubricated plunger, in evaluating yield stress revealed that friction on the plunger surface resulted in nonreproducible:yie1d.stress measurements. Smooth, lubricated.plungers are recommended for yield stress determination. As the ratio of plunger to sample tube radius approached one, a more erratic curve of force reduction during relaxation was produced. By using a wider annular gap, the erratic curve disappeared. Good results were achieved with plunger to container radius ratios less than 0.67. Yield stress determined after subjecting,a fluid to different plunger speeds or different the depths of penetration did not show significant differences. High yield stress foods needed. a very long observation time (as long as four hours in this study) to reach equilibrium states, otherwise the yield stress determined using this method was still time-dependent. my re: Ofc for Fe: dur ACKNOWLEDGMENT I would like to express sincere appreciation to Dr. James F. Steffe, my major Professor, for his unending guidance and advice concerning the research as well as the many hours spent reviewing this manuscript. Sincere appreciation is extended to my committee members: Dr. Denise M. Smith for her advice and support during the research, Dr. Robert Y. Ofoli and Dr. John A. Partridge for their helpful comments. Special appreciation is also extended to Kevin Rose and Joseph Esch for their help to build up the equipments as well as to Kevin Mackey and Fernando Osorio for their expertise. Most of all, I thank my wife and best friend, Tiwi, for her patience during my study in the USA. iii TABLE OF CONTENTS 1 INTRODUCTION 2 Literature Review . 2.1 Definition and Importance of Yield Stress. 2.2 Measurement of Yield Stress . . . . 2.2.1 Indirect Measurement 2.2.2 Direct Measurement . . 2.2.2.1 Force to Initiate Flow 2. 2. 2. 2 Stress Relaxation 2.3 Time-Dependent Yield Stress 3 Theoretical Considerations in Back Extrusion 3.1 Basic Principles of Back Extrusion 3. 2 Yield Stress Calculation . . 3. 3 Influence of Different Velocity and Depth of Penetration . . . 3. 4 Influence of Surface Friction 4. Materials and Methods . 4.1 Experimental Materials Experimental Equipment . .2.1 Back Extrusion Device. .2. 2 Gun Rheometer. .2. 3 Vane Device Experimental Procedures . . . .3.1 Operation of Back Extruder . .3. 2 Operation of Gun Rheometer . .3.3 Operation of Vane Device Experimental Design. Back Extrusion . . .1. 1 Smooth Versus Grooved Plunger. .1. 2 Lubricated Versus Unlubricated Plunger .1. 3 Ratio of Plunger to Tube Diameter. .1. 4 Different Plunger Velocity and Depth of Penetration . . . . .1. 5. Time- -Dependent Yield Stress . Comparison of Back Extrusion to Other Methods .2.1 Comparison to Gun Rheometer Data . .2.2 Comparison to Vane Device Data . J.‘ N ebbbbwbpb 15>be J-‘J-‘J-‘J—‘D bbwb 499be 5 Results and discussion 5.1 Back Extrusion . .1. 1 Smooth Versus Grooved Plunger . .1. 2 Lubricated Versus Unlubricated Plunger .1. 3 Ratio of Plunger to Tube Diameter .1. 4 Different Plunger Velocity and Depth of Penetration . . . . .1. 5 Time- -Dependent Yield Stress . omparison of Back Extrusion to Other Methods .2.1 Comparison to Gun Rheometer Data . .2 2 Comparison to Vane Device Data . 5.2 UIUOU'I U1U1U1U'I 6 Conclusions 7 Future Research iv oomww H 10 17 18 20 20 22 24 25 28 28 29 29 29 32 34 38 39 4O 4O 40 41 42 43 43 43 43 44 45 45 45 47 49 54 61 64 64 7O 74 76 8 Bibliography Appendix A Appendix Appendix Appendix Appendix B C D E Schematic of back extrusion experiment . Results using back extrusion . Stress relaxation curves Results using gun rheometer Results using vane device 77 81 82 92 103 106 Table 10. LIST OF TABLES List of some yield stress measurement methods Mathematical models used to express rheological behavior of fluids with yield stress Yield stress of semi-solid foods measured using differ- ent plunger surfaces with K-0.80 Yield stress measured using lubricated and unlubricated aluminum plunger Yield stress of tomato paste measured using lubricated aluminum plungers, with different ratios (K values) of plunger to inner graduate cylinder radius after various relaxation period Yield stress of chicken batter measured using different plunger velocities with a lubricated aluminum plunger and K-0.52 Yield stress of tomato paste measured with different depth of penetration using a lubricated aluminum plunger with K-0.62 Yield stress of tomato paste measured using lubricated and unlubricated plungers with K-0.52 Yield stress of peanut butter, tomato paste and chicken batter measured using gun rheometer and back extrusion (lubricated plunger and K-0.52) Yield stress of mixed cereal measured with back extru- sion, vane viscometer and gun rheometer vi Page 46 48 52 56 58 63 65 72 Figure 10. ll. 12. 13. 14. 15. l6. 17. LIST OF FIGURES Cone penetrometer Profile of fluid velocity during plunger movement in Back extrusion Typical instron data obtained from back extrusion experiment Profile of sample deformation in lubricated and unlubricated compression Schematic diagram of back extrusion device using an Instron Universal Testing Machine to collect experimental data Schematic diagram of gun rheometer system Schematic diagram of vane device system Stuffer as a loader for tomato paste and chicken batter Position of plunger before and after testing Typical shape of fluids in relaxed state for smooth, lubricated and unlubricated plungers Stress relaxation curve of tomato paste in back extrusion using a lubricated plunger and differ- ent annular gap Stress relaxation curve of chicken batter in back extrusion usin a lubricated plunger and differ- ent plunger ve ocities Stress relaxation curve of tomato paste in back extrusion using a lubricated plunger and differ- ent depth of penetrations Stress relaxation curve of tomato paste in back extrusion using a lubricated plunger and differ- ent end distance Stress relaxation curve of tomato paste in back extrusion using a lubricated and an unlubricated plunger with K-0.52 Plot of gun rheometer data for tomato paste Plot of gun rheometer data for peanut butter vii Page 12 21 23 27 30 31 33 35 37 50 53 55 59 60 62 67 68 18. 19. Stress relaxation curve of tomato aste, peanut butter and chicken batter using lu ricated plunger with K-0.52 Stress relaxation curve of mixed cereal in back extrusion using lubricated and unlubricated plungers viii 69 73 as. Q n ‘l O 3 S Y‘lflkkw'fivwkw ‘\ u 5. Q cp NOMENCLATURE chart speed of recorderJn/s diameter of a sample tube (gun rheometer),n1 diameter of a vane,n1 force,N buoyancy force,N total forces (back extrusion)yN total forces at equilibrium,N height of a vane,n1 ratio of plunger to cylinder radius, dimensionless consistency index (Herschel-Bulkley),Pa. 5" length of immersed plunger,n1 length of a sample (gun rheometer),n1 length of a spindle,n1 mass of a dropping device (cone penetrometer),kg air pressure,N/m2 critical radius,n1 radius of sample tube (gun rheometer),n1 radius of plunger,n1 radius of graduate cylinder,n1 radius of plates (squeezing flow),n1 radius of a spindles,n1 torque, N m maximum torque,Al n1 volume of a sample (squeezing flow),n1 gravity acceleration, 0.981171/52 the depth of penetration (cone penetration),n1 limiting height (squeezing flow),n1 ix chart length,n1 flow behavior index, dimensionless velocity of plungerJn/s Greek Symbols one half cone angle, degrees shear rate,s'I shear rate on plunger wall (back extrusion),s'I apparent viscosity,Pa s Newtonian viscosity,Pa 5 high shear limiting viscosity (Bingham plastic),Pa .3 high shear limiting viscosity (Ofoli et al.),Pa 5' density of a fluid, kg/m3 shear stress,Pa yield stress,Pa shear stress on plunger wall (back extrusion),Pa logi move flui Bacl test firs Morg flui the Stei prac (201] the to take for and Yie] ValL ofs Chapter 1 INTRODUCTION Back extrusion or annular pumping has been used for measuring rheo- logical properties of various food materials. In this method two physical movements are involved : (l) a cylindrical plunger is forced down into a fluid and (2) the fluid flows upward through a concentric annular space. Back extrusion terminology comes from the fact that fluid motion during testing, is in the opposite direction to plunger movement. This method was first proposed to determine the characteristics of Newtonian fluids by Morgan et a1. (1979). Later, a mathematical analysis of non-Newtonian fluids was developed by Osorio-Lira (1985). However, due to the fact that the calculation procedures for Herschel-Bulkley fluids are very complicated, Steffe and Osorio (1987) did not recommend these methods for industrial practice. The back extrusion method has also been proposed as one of the techniques to measure yield stress directly (Osorio-Lira, 1985). The technique uses the stress relaxation technique in which the fluid, after being subjected to a constant shear rate, is allowed to rest and the remaining stress is taken as a function of the yield stress. The reliability of this method for measuring yield stress has been demonstrated for baby food by Steffe and Osorio (1987). However, its capacity to evaluate fluids with a high yield stress has not been explored. It is important that yield stress values calculated from back.extrusion data be reproducible and independent of strain history. cap fro mea of mea gec exp The objectives of this study, therefore, were: 1. To evaluate the capability of back extrusion testing to evaluate yield stress in fluids from "free-flow" to "semi-solid" types; 2. To determine if the yield stress measured by back extrusion is independent of plunger velocity and depth of penetration; 3. To improve experimental techniques in yield stress measurement using the back extrusion method by evaluating the effect of geometrical factors (plunger radius/container radius), lubrication and experimental surfaces (material characteristics) on results. Chapter 2 Literature Review 2.1 Definition and Importance of Yield Stress A yield stress may be defined as a minimum shear stress required to initiate flow. The yield stress marks the transition from elastic to viscous behavior; below the yield stress the substance is considered an.elastic-solid and expected to follow'Hooke's law; and.above the yield stress, it is considered a fluid and may follow one of the fluid flow models such as Bingham plastic if linear, Herschel-Bulkley or Casson Eqs. if nonlinear. Beyond the yield stress, the fluid flows with a shear rate that depends on the excess stress 03-00) If a fluid exhibits yield stress under the action of an applied constant shear rate, the shear stress will increase steadily until the yield stress is reached and then the stress will hold constant (Dzuy and Boger, 1983). The yield stress in this case is independent of the shear rate and can be shown as the interception of shear stress-shear rate flow curve at zero shear rate. A yield stress is commonly found in highly concentrated emulsions under conditions where interparticle interactions bring about mutual attractions among individual particles (Dzuy and Boger, 1983) . Princen' 5 study (1985) with foams and highly concentrated emulsions found that the magnitude of yield stress was given by an interfacial tension, a volume fraction of the dispersed phase and the surface-volume mean drop radius. The yield stress may also be a result of physical entanglement of molecules or particles due to high degree of branching or irregular .r ii shape, a network formation, a covalent or a secondary interparticles interaction, or an increased non-specific interaction between molecules or particles (Rha, 1978). The experiment done by Barnes and Walters (1985) showed that a Bingham plastic fluid measured at very low shear rates (10'3-1CY55”) did not exhibit a yield stress. That experiment has raised the question of the existence of'a yield stress. However, the concept of yield stress is still generally accepted by experimental rheologists. Cheng (1986) addressed the question using practical reasoning by noting that, when measuring the yield stress, it is not necessary to go to the lowest or zero shear rate, because yield stress is more important when related to process design or industrial purpose which has a limited residence time. Ofoli et al. (1987) also included a yield stress parameter in their model due to the fact that food manufacturing processes have strict time limitations. The exact value of the yield stress should be known when food companies want to increase the efficiency of their operation or obtain better quality products. The qualities of food products which have a yield stress include shape retaining ability (cheese or gelatin), coating characteristics (chocolate products) and spreadability (margarine or butter). The thickness of salad dressing or tomato ketchup and the softness of some dairy products are also altered by the value of the yield stress. In the beverage and drug industry, the yield stress prevents the small particles from settling during storage. Yield stress must be taken into account in process design and equipment specification due to possible adherence on vessels or pipes, the occurrence of dead regions during mixing operations , and the alteration of F-values in aseptic process due to unique velocity profiles in heat exchangers. 2.2 Measurement of Yield Stress Yield stress may be obtained by an extrapolation of the rheogram to zero shear rate. If the rheogram is linear, such as found with the Bingham plastic model, the yield stress can be determined accurately. Unfortunately, many fluids exhibit nonlinear behavior at shear stresses above the yield value. As a result, one should go to as low a shear rate as possible to obtain a true yield stress. In practice, it is difficult to measure the shear stress at a very low shear rate due to the limited capability of many rheometers. Direct measurement of yield stress without measuring shear rate, therefore, is very attractive. Numerous methods have been established, but many are tedious and need very specific conditions. It is not unusual that a yield stress obtained by one technique is non-reproducible and com- parable to that obtained using a different technique. Therefore, some researchers believe that the yield stress may be time-dependent (Lang and Rha, 1981 and Cheng, 1986). It is difficult to define the yield stress as an absolute rheological parameter. Cheng (1986) also recom- mended that determination of yield stress must be made relevant to practical application. To minimize sedimentation in suspending fluid, for example, one should measure a yield stress at an extremely low shear rate. In contrast, to determine start-up power requirement in pipe line transportation or mixing operation, one should consider higher shear rates and a short measurement time. Table 1 shows yield stresses of food and nonfood materials which have been measured using different methods. Table 1. Lists of some yield stress measurement methods. Heinz-Casson 23 Method Material 00 Reference (Pa) Indirect methods (extrapolation) Herschel-Bulkley fish paste 1 800 Nakayama et a1. (1980) Herschel-Bulkley meat batter Burge and Acton ll % fat 587 (1984) 18 % fat 239 26 % fat 148 graphically Titanium dioxide 128 Dzuy and Boger (1983) Bingham plastic (37.3 % solid) 234 Herschel-Bukley 125 Casson 128 graphically miracle whip 26 Ofoli et a1. Bingham plastic 54 (1987) Herschel-Bukley 30 Casson 39 Method Material 00 Reference (Pa) Direct measurement Food cone penetrometer margarine 10 000 Tanaka et a1. @ 20 C cheese 6 800 (1971) peanut butter 600 squeezing flow tomato paste 120 Campanella and Peleg ketchup 26 (1987) mustard 56 mayonnaise 84 balance plate condensed milk 2 De Kee et a1. @ room temp. corn syrup 2 (1980) molasses 3 ketchup 16 mayonnaise 25 tomato paste 84 stress decay guar gum 2 % 4 Lang and Rha stress tn) initiate cornstarch 4.3% 20 (19 1) flow cornstarch 4.3% 12 cone and plate mayonnaise 70 Elliot and Ganz (1977) back extrusion fruit-dessert 20 Steffe and Osorio baby food (1987) Nonfood squeezing flow toothpaste 4 Covey and Stanmore Titanium dioxide 128 (1981) (37.3 % solid) concentric cylinder kaolin 24 Vocadlo and Charles Titanium dioxide 40 (1971) stress relaxation Titanium dioxide 106 Dzuy and Boger (1983) (37.3 % solid) vane device red mud 168 Dzuy and Boger (1983) (66 % solid) 2.2.1 Indirect Measurement Determining the yield stress by extrapolation of. shear stress-shear rate data is an indirect method since it requires rheological data from a previous measurement. There are two techniques which can be used in extrapolation: (l) to extrapolate raw data of the relationship between shear stress and shear rate directly, graphically; and (2) to extrapolate the data using a mathematical model determined via linear or nonlinear regression. Most common rheological instruments may be used to collect data to determine yield stress using indirect methods: capillary, concentric cylinder, cone-plate or parallel plate viscometers, etc. The advantage of direct extrapolation of raw data is that it does not depend on any specific mathematical model whose validity must be established separately. A yield stress is simply determined from an extrapolation of the shear stress to zero shear rate or the shear stress where the apparent viscosity goes to infinity on a plot of apparent viscosity versus shear stress. To get a more accurate result, Kaletung-Gencer and Peleg (1984) used a digitizer aided determination. However, the value of yield stress obtained from this technique can not be confirmed using any statistical method. In addition, the accuracy of the yield value depends on the reliability of the rheological data at low shear rates. The other method is the extrapolation of a mathematical model which has been established separately based on experimental data. The most common Eqs. used to express behavior of fluids with a yield stress are the Bingham plastic, Herschel-Bulkley, Casson and Heinz-Casson models. The recent model proposed by Ofoli et a1. (1987) is a generalization of those traditional models (Table 2). Table 2. Mathematical models used to express rheological behavior of fluids with a yield stress. Model Name Shear stress Apparent viscosity Bingham plastic 0'00*H1Y ,1za Y “*Ho (Bingham, 1922) Herschel-Bulkley o _ 00+ Kn)?“ “=00-Y—1 + KN”..- (Herschel and Bulkley, 1926) Casson 0'53005+(u_9)5 n_((O°y-l)5+(u.)5)2 (Casson, 1959) l/n Heinz-Casson on-03+(uoy)n "=((00Y)R+CHOYO (Heinz, 1959) General1zed model a"-o;'+p-y"2 00 M ,%_M (Ofoli et a1. 1987) n (*‘) ‘*H.Y llul 10 The value of yield stress predicted by each model (Table 2) from the same rheological data may be significantly different (Ofoli et a1. 1987). In this case, the yield stress is strongly determined by a selected model and curve fitting technique rather than physical characteristics of the fluid itself. Although the mathematical model is very important for process design, a fundamental characteristic of the fluid like yield stress should be determined independent of those models for the most accurate result. 2.2.2 Direct Measurement It is important to measure yield stress directly, independent of a particular mathematical model. Ofoli et a1. (1987) showed that a specified yield stress can alter the value of the other rheological parameters for every model. Much attention has been given to direct measurement and many methods have been proposed. Unfortunately, most methods are still not reliable due to their limited applications and non-reproducible results. The methods are only consistent for specific materials or experimental conditions. Two major methods applied in direct measurement involve the use of a force to generate flow and stress relaxation after an applied force. Another proposed method which has a very limited application is the oscillatory test (James et al. 1987). A yield stress may be obtained from response waves, where, below the yield stress, the response is a sine wave and above it a flat wave. 2.2.2.1 Force to Initiate Flow The cone penetration test has been used to measure shearing stress since 1949 (Tanaka et al., 1971), and more recently was used to measure yield stress of margarine or butter (Dixon and ll Parekh, 1979). The principle of this method is based on the assumption that, by applying a constant weight, the cone ( Fig. 1) will penetrate into the material until an equilibrium point is reached. The shear force is given as Algcosa and the conical area of penetration is nhfptana/cosa. The yield stress, then, can be calculated from the equilibrium point: gAdcosza (1) 0 ° nhfptana where: g - gravitation acceleration, 0J98l rn/sz 1M'- mass of a dropping device, kg h“; the depth of penetration, fit a a one-half cone angle, degrees Tanaka et al. (1971) found that the exponent of hq,was not exactly two, but fell between 1.4 and 2.0 depending on the type and the temperature of the food. Dixon and Parekh (1979) used different cone angles for measuring the firmness of butter, and found that the expression of the yield stress was =M(Za)-l.65 (2) 0 2 h”. In addition, Keentok (1981) mentioned that the equilibrium might require penetration time of 12 hours or longer. 12 Figure 1. Cone penetrometer. 13 Squeezing film viscometers, which consist of parallel plates used to force substance movement with a constant force, have been called by several different names: parallel plate plastometer (Dienes and Klemm, 1946), parallel plate viscometer (Gent, 1960), transverse flow viscometer (Van Wazer et a1. 1963), compression plastometer (Mooney, 1958) and squeezing flow viscometer (Leider and Bird, 1974). In squeezing flow, a yield stress can be determined from the limiting height or plate separation. The method was developed to characterize polymer melts, asphalt and other nonfood materials. Recently, Campanella and Peleg (1987) used it for measuring yield stress of semi-solid foods like tomato paste, ketchup, mustard and mayonnaise. The theory of squeezing flow viscometer has been described in detail by Leider and Bird (1974). Measuring yield stress, in particular, using this method was described by Convey and Stanmore (1981). There are two geometrical versions called.1constant-volume in which the diameter of a sample is always less than the diameter of the plates and constant-radius in which the diameter of a sample is larger than the diameter of the plates. The limiting separation is expressed by following Eqs.: for constant-volume 5 (3) h (ZVLSOOY L 3n°5F for constant-radius, substitute nRiflzto Vs“ so that 2nR2,oo (4) h" 3F 14 where: F - force applied to plates, N R,, - radius of plates, m If“ - volume of a sample,m3 These Eqs. are good for both Bingham plastic and Herschel- Bulkley fluids since there is no motion at the limiting height situation (Covey and Stanmore, 1981). The disadvantage of the squeezing flow method is its inability to measure a fluid with a small yield stress or a free-flowing fluid. One of the most popular viscometers is the concentric cylinder type. This instrument consists of an inner cylinder or spindle (bob) and an outer cylinder (cup). Besides measuring rheological properties under flow conditions, the concentric cylinder vis- cometer is also used in yield stress measurement. Three techniques may be used: stress to initiate flow; plug flow radius; and stress relaxation. The last technique will be described later in another section. The torque on the bob of a viscometer when fluid starts to flow is a function of the yield stress. The technique was described by Lang and Rha (1981) as follows: the bob of a wide-gap viscometer was lowered into a sample, the sample was allowed to relax for ten minutes, before the outer cup was manually rotated. The stress at the onset of flow was designated as a yield stress: ° 2nR3L, where: T'- torque on bob when the fluid begins to flow, Al N: R, - radius of the bob, n1 Lo - length of the bob, n1 15 In the plug flow radius technique, a wide-gap viscometer is required, so there is a dead region of fluid when the bob is rotated. The yield stress, then, is calculated using the critical radius located between the flow and the dead region: 0 g T (6) ° 2anL, where: Rc- critical radius, n1 When the presence of slip is suspected, a grooved cylinder may be used along with a smooth cylinder as a comparison (Vocadlo and Charles, 1971). To eliminate end effects, Princen (1985) put mercury at the bottom of bob instead of an air bubble trap; however, this practice is not recommended for food products. The vane device, used widely in soil mechanics, has received much attention for measuring yield stress in various chemical and food fluids. The advantages of using a vane device, instead of a concentric cylinder include no wall slip, no end effect and a minimum disturbance when the vane is introduced into the sample. The theory and procedures of this device for measuring yield stress have been described elsewhere (Keentok, 1982, Dzuy and Boger, 1983,1985, and Keentok et a1. 1985). In principle, a vane with 4 to 8 blades is immersed into a sample and then rotated very slowly at a constant speed. The torque on the vane shaft is recorded as a function of time. It reaches a maximum value when the stress applied to the sample is equal to yield stress and the vane starts to rotate. By considering the geometry of the yield surface, the dis- tribution of shear rate on the surface and the assumption that the 16 shear stress is uniformly distributed around the tip of the vane and equal to yield stress, the simple relationship between1 ‘_ + + plunger stop Force (N) /l Time (minutes) Figure 3. Typical Instron data obtained from back extrusion experiment. 24 At the equilibrium state, the shear rate and Af’are equal to zero, and Eqs. (12) and (13) become: F},'99Lflki (14) for a Newtonian fluid, and F,.-2nR.La.+poI-nkf (15) for the Herschel-Bulkley (or Bingham plastic) fluid where: Fr.- equilibrium force after cessation of plunger movement, N Rearrangement of Eq. (15) gives the direct expression for the yield stress: a .. Fn-poLnR? ° ZnRJ.‘ (16) The reliability of the back extrusion method to measure yield stress has been demonstrated for baby food by Steffe and Osorio (1987). However, its capacity to measure fluids with high yield stress has not been fully explored. Due to the lack of information on stress relaxation in back extrusion, some suspected factors that contribute significant effects during testing should be examined. These are plunger velocity, depth of plunger penetration” a friction factor on plunger surface and annular gap size. 3.3 Influence of Different Velocity and Depth of Penetration As seen in Eqs. (12) and (13), a yield stress may be determined when the shear rate is zero. At this condition there is no force at 25 the bottom of plunger due to upward flow of the fluid. The buoyancy force and the force due to yield stress must be distributed uniformly along the vertical surface of the immersed plunger. This condition will be fulfilled if therelis no shear force remain.after the fluid is allowed to rest for sufficient time. There must also be good contact between the fluid and immersed plunger surface. To determine if a yield stress measured with back extrusion is independent of plunger movement, some different treatments have to be given to the variables related to plunger movement such as different plunger velocities and depths of penetration. 3.4 Influence of Surface Friction Based on the mathematical expression described by Osorio-Lira (1985), the shear stress generated during plunger movement in back extrusion is function of'a ratio of plunger to tube radius, the velocity of plunger movement, the depth of plunger penetration and a friction factor on the plunger surface. The significant role of a frictional effect in uniaxial compression has been investigated by some researchers ('Chatraei et a1. 1981, Christianson et al. 1985 and Bagley and Christianson, 1988). They found that a frictional effect not only altered the magnitude of shear stress, but also contributed to non-reproducible results. That condition was due to the fact that the fluid deformation in unlubricated plates did not have uniform (and reproducible) contact with the plates. A similar response may be expected in back extrusion when the surface of the plunger has a variable friction factor. This friction factor is very important in measuring rheological properties of fluids because a no slip condition at annular wall is required in mathematical analysis (Osorio-Lira, 1985). The yield stress measured using the stress 26 relaxation method may depend on this friction factor. In addition, small annular gap size is suspected as a factor which contributes significant effects in force distribution along immersed.p1unger during relaxation. Unfortunately, no study'has evaluated these factors with regard.to yield stress measurement using back extrusion. 27 Lubricated Unlubricated Figure 4. Profile of sample deformation in lubricated and unlubricated compression (Christianson et al. 1985). Chapter 4 Materials and Methods 4.1 Experimental Materials Foods with high yield stress were chosen for experimentation: tomato paste (Beatrice/Hunt-Wesson, Inc.), peanut butter (Groeb Farms, Inc.), mixed-cereal baby food (Gerber Products Co.) and chicken frankfurter prepared from mechanically deboned chicken in the Food Science Building at Michigan State University. Tomato paste (6 1b. cans) , peanut butter (5 lb. buckets) and mixed-cereal (5 oz. jars) were purchased from the Michigan State University Food Store. To prepare chicken frankfurter batter, 15 lbs block of frozen mechanically deboned chicken was cut into 2500 g blocks, wrapped in polyethylene , and stored at -30 ° C. A frozen mechanically deboned chicken block was stored at 4-° C before being used the next day. Ingredients consisted of 2200 g mechanically deboned chicken, 33.5 g sugar, 9.2 g pepper, 54.85 g sodium chloride, 0.34 g sodium nitrite and 1.37 ascorbic acid, were weighed out. Frankfurters were manufactured at room temperature (approximately 25"C. MDC, spice mix, salt and 500 g ice were placed in a cutter and chopped for 5 min. The batter, having a temperature approximately 4—7 ° C, was stuffed into graduate cylinders immediately after chopping. In the experiments with a lubricated plunger, Vegalene pan coating (Tryson Company) was used as a lubricant. 28 29 4.2 Experimental Equipment 4.2.1 Back Extrusion Device The Instron Universal Testing Machine (Instron Corporation), Model 4202, with a 50 N compression load cell was used as the main device in back extrusion (Fig. 5). This machine was connected to a Hewlet-Packard 86B computer as an input program device, a recorder and a printer as output devices. A plunger rod, used to force the fluid during testing, was screwed to the load cell on cross head. The Hewlet-Packard system was also used to control the Instron. Three different diameters of smooth aluminum plunger rods were used: 20 mm, 24 mm and 28.58 mm. Two plunger rods (6 total) with the same diameter (28.58 cm), but different surface characteristics, smooth and grooved, were made from plexiglass, teflon and aluminum. Graduate cylinders of 100 ml, 250 m1 and 500 ml capacity. with diameters of 25.62 mm, 35.68 mm and 46.46 mm, respectively, were used as sample holders. 4.2.2 Gun Rheometer A gun rheometer (Fig. 6) based on principles described by Cheng (1986) was built at.Michigan.State University. As shown.schematica11y (Fig. 6) the wall of a pressure vessel was made of PVC (8.16 mm thickness) with a diameter of 15 cm and height of 38 cm. Top and bottom were also covered with plexiglass (19.18 mm thickness). 30 load ceH r—— “H\ plunger rod . C] \ o o o / graduate cyl. control panel Figure 5. Schematic diagram of back extrusion device - -—. _..__.. w -_ -..._-__.._—-_.-. using an lnstron Universal Testing Machine to collect experimental data. 31 Air Supply :6 r {it Monometer j, t II F_-m__g Pressure Gauge Pressure Vessel T acts: a 1f Release Valve Sample Tube Figure 6. Schematic diagram system. a: of gun rheometer 32 To prevent the pressure vessel from leaking when subjected to high air pressure the top and bottom covers were sealed with rubber and screwed together with threaded steel bars and wingnuts. A safety valve was installed to release air if the pressure in the vessel because too high. The system was connected to a laboratory air supply. To measure air pressure in the vessel, a water manometer along with a dial pressure gauge were used. The dial pressure gauge was used when the air pressure was higher than 5 kPa. Sample tubes, connected to vessel using a rubber tube, were made of plexiglass with 10 mm inside diameter, 2.75 mm wall thickness and 300 mm length. 4.2.3 Vane Device The vane device (Fig. 7) used to measure yield stress in this study was similar to the device used by Dzuy and Boger (1985). The vane consisted of four blades centered around a small shaft. The diameter of the vane was 24.5 mm and the length of blades was 50.8 mm giving a H/D equal to 2.06. This ratio was still less than 3.5 as recommended by Dzuy and Boger (1985). A Brookfield HBTD viscometer with full scale of 54,496 dyne cm was used as a rotational motor. The torque generated during testing was registered using a strip chart recorder and the maximum torque was calculated from the peak of the curve. 33 Vone Brookfield Viscometer '- l P .1 I l l . . / l l: ' ' Strip Chart : : 5.06 cm g l ‘ l l? n I T ’ 1 2.5 cm I l T Li); I 5.06 cm \/ k J \ ’ / J 2.5 cm H H 2.45 cm 6 cm Figure 7. Schematic diagram of vane system. 34 4.3 Experimental Procedures 4.3.1 Operation of the Back Extruder Two major problems in back extrusion testing are nonhomogeneous fluids and entrapped air bubbles during sample loading. Although the samples for each experimental series were taken from the same can- tainer, it was not uncommon to find that a sample taken from the top part was different from one taken from the bottom part of the can due to storage changes. To avoid inconsistency, the material in each container was stirred slowly and thoroughly before being loaded into sample holders. A container of sample might consist of several jars of baby food, a can of tomato paste, a bucket of peanut butter or chicken batter prepared at the same time. To prevent entrapping air bubbles during sample loading, two techniques were applied. The first technique, applied to peanut butter, involved melting the sample at 50 °C and pouring it into a graduate cylinder. With the other technique, applied to tomato paste and chicken batter, the sample was loaded using a stuffer (Fig. 8). Mixed cereal did not need a special treatment. It was just poured slowly into a graduate cylinder. It is important to describe the essentials of sample loading using stuffer. The stirred sample was put into the cylinder and forced down. During this action, the entrapped air left through the gap between the plunger and the cylinder wall. The fluid forced through the tube was expected to have no entrapped air bubbles. It was, then, loaded into a graduate cylinder starting from the bottom. The stuffer tube was slowly removed during loading to ensure uniform fill. 35 l a <——— Plunger 1 Sample _, le‘YSi‘é‘é‘é‘é'éi'm Sample Tube Graduate Cylinder Figure 8. Stuffer as a loader for tomato paste and chicken batter. 36 The loaded graduate cylinders were allowed to rest for 24 hours before testing. Chicken batter was stored at 5 °C and the other fluids were stored at the room temperature, approximately 25 °C. During storage, samples were covered.with aluminum foil to eliminate surface drying. By allowing a 24 hour rest period, it was expected that strain history due to loading would not influence subsequent tests. The density of a f1uid.was calculated from the weight per unit volume of the loaded samples. Yield stress measurement procedures using back.extrusion in this study was the same as procedures described by Steffe and Osorio (1987). The sample was subjected to a constant shear rate (constant plunger velocity) using the Instron Universal Testing Machine. Testing was conducted at room temperature and the force on the plunger versus time was recorded on a strip-chart recorder. In addition, every two minutes the magnitude of the force was printed out. After one minute of downward motion, the plunger was stopped and the fluid was allowed to relax. The force on the plunger was recorded until an equilibrium state was reached. Before the yield stress may be calculated (Eq. 16) a number of variables must be evaluated. The equilibrium force Frocan be read from the recorder. Density of the fluid, gravitational acceleration and the radius of plunger are known. The length of immersed plunger must be determined. The length of plunger (Fig. 9) that penetrated the fluid 15235 and the volume of fluid displaced by the rod is _ 13.1 to anO_B. Since the displaced fluid is forced up, around the annulus, the following relationship must be valid (Osorio and Steffe, 1985): 37 l r“"—_"| l O 1 l O 0 T i L l I .‘ Bl ! ‘ l RI ' i l Lift; L l l Before After Figure 9. Position of plunger before and after testing. 38 anOF-nkiT-nkfr (17) or .__ RE .__ (18) .4 -R2_R20 and L-T+_§ (19) by putting Eq. (18) into Eq. (19), the value of L will be L=-—§§—- (20) l-K2 Where: K-R/Ro OE can be determined from chart length as 53-5—30 (21) where: (m - chart length, from recorder, m (3” - chart speed of the recorder, m/s up - velocity of the plunger, m/s By knowing the value of L.and the other relevant variables, the yield stress of fluid may be calculated using Eq. (16). 4.3.2 Operation of the Gun Rheometer Samples for gun rheometer testing (Fig. 6) were prepared from the same material used in back extrusion testing. It was loaded into a sample tube using a Model 1142 grease gun made by Lincoln-St. Louis. The length of sample, as recommended, was more than ten times its diameter. The entrapped air in the loaded sample was removed with a syringe. As in back extrusion testing, the samples were covered with 39 aluminum foil and allowed to rest for 24 hours before testing. Except for chicken batter that was stored at 5 °C, all samples were stored at room temperature. The technique of excess pressure as described in Section 2.2.2.1 was used in this study. For satisfactory measurements, a variety of velocities had to be obtained; thus, a different level of excess pressure had to be applied to each sample. Before a sample was tested with the gun rheometer, the air pressure in the vessel was set to a pressure higher than the pressure needed to initiate flow. By opening the air supply valve, the air pressure in the vessel was increased to line pressure. Then the valve was closed and if the air pressure in the vessel was too high a pressure adjustment was made using the release valve. The water manometer was only used if the air pressure in the vessel was below 5 kPa. Excess pressure was applied to each sample for two minutes and the discharged sample was measured. The data collected from all samples were plotted as pressure versus mass average velocity (volumetric flow rate divided by cross-sectional area). By extrap- olating to zero velocity, the minimum pressure to initiate flow was obtained. The yield stress, then, was calculated from this value using Eq. 9. 4.3.3 Operation of the Vane Device Samples were prepared in the same manner as that described for back extrusion. The only important difference was the sample holders for the vane device: glass containers with diameters of 6 cm and heights of 10 cm. 40 The vane spindle (Fig. 7) connected to the Brookfield was introduced smoothly into a sample until the top of the vane was immersed to a depth approximately equal to the vane diameter (2.45 cm). After waiting for approximately 10 minutes, the vane was rotated as slowly as possible with a constant speed (0.5 rpm). The torque on the shaft was recorded as function of time. Maximum torque was determined from a maximum value on the digital window of Brookfield viscometer multiplied by full scale torque value of the device and divided by 100. The yield stress of the fluid was calculated from the maximum torque (Eq. 7). 4.4 Experimental Design To fulfill the objectives of this study (as mention in the Introduction), yield stress of a fluid was measured using back extrusion with different plunger surface characteristics and geometrical factors. Each treatment was tested with three replications. The average yield stress of one treatment was compared statistically to that of another treatment using t-student distribution at 0.05 confidence level. If the two treatments did not have a significantly means difference, the variances of the yield stress were compared to analyze their uniformities. The following experimental design fulfilled the objectives. 4.4.1 Back Extrusion 4.4.1.1 Smooth Versus Grooved Plunger Three smooth plunger rods made of aluminum, teflon. and plexiglass respectively were used in this experiment. The selection of the three materials was to assess whether or not the yield stress measured with plungers made of different materials would alter the calculated value. These plungers had the same diameter 41 (28.56 mm) and all samples were held in 250 m1 graduate cylinders. Thus, K was held at a constant value of 0.80. This follows the recommendation by Marte-Guzman (1987) for measuring rheological properties during back extrusion testing. To examine the no slip condition required in the back extrusion method, as described by Osorio-Lira (1985), three ribbed plunger rods were tested. These grooved plunger rods were made of the same material and diameter as the smooth plunger rods. The plunger rods were ribbed horizontally with 1.25 mm width and 0.72 mm depth grooves, so that there were 8 grooves per cm of plunger length. To measure yield stress of the fluids, a sample was subjected to a constant downward plunger velocity of 100 mm/min. After one minute, the plunger was stopped and the fluid was allowed to relax. The forces on the plunger surface during relaxation were recorded on the strip chart paper. The equilibrium force was determined one hour after the plunger was stopped. Testing was repeated three times for every plunger; thus (using 6 plungers) there were a total of 18 experiments. The yield stress values measured with different plungers were compared and examined. 4.4.1.2 Lubricated Versus Unlubricated Plunger This experiment referred to the study done by Christianson et al. (1985), which revealed that friction factors on the plates during compression testing made a large contribution to nonre- producibility of results. The same problem may be found in back extrusion testing when measuring,yield stresses with unlubricated plungers. 42 To evaluate whether a friction factor on the plunger surface had an effect on yield stress measurement, the results from a lubricated and an unlubricated.plunger were compared” The friction factor on the plunger surface of a lubricated plunger was reduced by lubricating the plunger with vegetable oil (pan coating oil from Vegalene). In the lubricated plunger experiment, the shear rate reduced to a small value. The total forces on the plunger when measuring fluids with yield stress, therefore, equal to the total of the force due to yield stress, buoyancy force and.the force responsible for fluid motion. The Eq. 13 may be expressed as F,-2nR,Loo+anAP+ngan (22) Tomato paste and peanut butter were tested using lubricated and unlubricated aluminum plungers with diameter of 28. 56 mm, three replications for each plunger. The 250 ml graduate cylinders were used as sample holders and the yield stress was determined one hour after the plunger was stopped. 4.4.1.3 Ratio of Plunger to Tube Diameter The distance between solid boundaries in back extrusion represented by annular gap may influence the yield stress cal- culation. A narrow annular gap may retard the relaxation. To observe the effect of solid boundaries on yield stress measurement, three different annular gaps, expressed as the ratio of plunger to graduate cylinder diameter, were used: K.values of 0.78, 0.62 and 0.52. The equilibrium force measured one hour after the plunger was stopped and the curve of force reduction on the plunger versus time were recorded for a three hour period. 43 4.4.1.4 Different Plunger Velocities and Depths of Penetration The yield stress was determined from the force at equilibrium. At that state, no plunger movement effect was expected. To observe whether plunger movement affected the equilibrium force, a fluid was tested.using different plunger speeds and.different the depths of penetration. The characteristics of the plunger and the ratio of plunger to tube diameter were chosen from earlier experiments. 4.4.1.5 Time-Dependent Yield Stress Due to the fact that most yield stress value measured using stress relaxation depend upon the observation time , it was important to examine whether or not the yield stress measured at one hour after the plunger stopped was time-dependent. Therefore, data of force on the plunger versus time was plotted for three hour periods. 4.4.2 Comparison of Back Extrusion to Other Methods The best experimental techniques for back extrusion were established from_ the result of the previous experiments. This technique, then, was used to measure yield stress of four fluids: chicken batter, tomato paste, peanut butter and mixed cereal baby food. To validate these results, they were compared to other yield stress measurement methods. 4.4.2.1 Comparison to Gun RheOmeter Data The same fluids considered in back extrusion testing were measured using a gun rheometer. A yield stress was determined using an excess pressure technique and linear regression. The yield 44 stress of each fluid obtained from the regression method was compared to the average yield stress of each fluid obtained from the back extrusion technique. 4.4.2.2 Comparison to Vane Device Data The lowest rotational speed of the Brookfield equipment was 0.5 rpm; however, the fluids were tested using three rotational speeds (0.5,l.0 and 2.5 rpm). If the yield stresses measured using these three speeds had different values, the yield stress at zero rpm (determined using extrapolation method) was taken as the actual value. Chapter 5 Results and Discussion 5.1 Back Extrusion 5.1.1 Smooth Versus Grooved Plunger Table 3 shows the average yield stress values and standard deviations obtained from back extrusion using smooth and grooved plungers. The values of the yield stress are very high in deviation. Thus, it is impossible to determine the true value of yield stress from those data, no single plunger produced consistent results. For example, the average yield stress of tomato paste measured with smooth plexiglass is 138.1 Pa and standard deviation is equal to 121.9 Pa. It means as Chebyshev's rule guarantees (Bhattacharyya and Johnson, 1977), at least 75 percent (2 std dev.) of the yield stress measured with the same condition will be varied from minus 105.7 Pa to 381.9 Pa. The smallest standard deviation is for the yield stress of chicken batter when measured with smooth plunger made of teflon, but this plunger yielded very high standard deviations when measuring other foods. Therefore, the method used by Steffe and Osorio (1987) can not be applied to measure a yield stress of semi-solid food which have a high yield value. The reason why the method did not perform well with these materials is difficult to explain. 45 46 Table 3. Yield stress of semi-solid foods measured using different unlubricated plunger surfaces with K-0.80. Tomato paste Peanut butter Chicken batter Plunger sur~ .Average Std.deV' Average Std.dev .Average Std.dev face (Pa) . (Pa) (Pa) Plexiglass smooth 138.1 121.9 55.8 25.8 556.7 54.7 grooved 204.9 53.9 190.0 53.9 747.3 49.9 Teflon smooth 202.1 143.1 88.8 131.9 759.7 38.0 grooved 126.3 25.7 130.8 121.7 732.7 44.0 Aluminum smooth 132.6 80.1 111.1 48.7 570.7 144.9 grooved 109.6 96.9 137.1 68.9 842.4 102.2 47 The possibility of friction on the surface of plunger or entrapped air bubbles at grooved plunger resulted in nonuniform distribution of forces along the plunger. The poor reproducibility was also found by Dzuy and.Boger (1983) when measuring red mud suspensions using stress relaxation method. However, their conclusion, that the problem was due to a slip effect, can not be proved in this experiment. 5.1.2 Lubricated Versus Unlubricated Plunger A lubricated plunger was used to reduce errors due to surface friction has been applied in back extrusion. The results of yield stress measurement can be seen in Table 4. It shows that yield stress values obtained using a lubricated plunger are more consistent than those found using an unlubricated plunger. Although the average yield stresses are different when measured using a lubricated or an unlubricated plunger, the difference is not statistically significant at 0.05 confidence level. However, the variance between the lubricated and the unlubricated plungers are significantly different at 0.05 confidence level. Thus, the values of yield stress measured using a lubricated plunger are more uniform than using an unlubricated plunger. It may be concluded that surface friction has a strong influence on yield stress measurement. The best reproducibility in experimental data will be obtained using a lubricated plunger when measuring yield stress using the back extrusion technique. 48 Table 4. Yield stress measured using lubricated and unlubricated aluminum plunger. Plunger surface Yield stress (Pa) K-0.80 Peanut butter Tomato paste Lubricated 61.9 145.5 58.8 152.0 71.9 175.1 Average 64.1 157.6 Std. dev. 6.9 15.6 Unlubricated 58.4 224.1 154.6 98.1 120.4 230.7 Average 111.1 184.3 Std. dev. 48.8 74.7 49 Typical shape of fluid after the completion of a back extrusion test is illustrated in Fig. 10. For the lubricated.plunger, the fluid filled the annular space at the same level but for the unlubricated plunger it did not. It is possible that the friction factor on the plunger surface may inhibit upward fluid movement and the development of the velocity profile with zero slip. The magnitude of buoyancy force may also be altered by this shape. In addition, it may create a nonuniform distribution of force along the immersed plunger surface bringing about an error in the calculation of the yield stress. This response is similar to the result found by Christianson et a1. (1985) in uniaxial compression experiments where an unlubricated plate altered the distribution of force (Fig. 4). 5.1.3 Ratio of Plunger to Tube Diameter The effects of annular gap during stress relaxation can be seen in Table 5 and Fig. 11. Yield stress determined after one hour observation reveals that higher K values produced smaller calculated yield stresses. The differences are statistically significant at 0.05 confidence level compared to either yield stress measured with K-0.67 or K-0.52. Yield stresses measured with K- 0.67 had lower values than yield stresses measured with K- 0.52. However, the differences are not statistically significant at a confidence level of 0.05. For longer observation time (until three hours), yield stress decreased faster for the higher R value. 50 l l l 3 3R 3 3A 3 2 . l l l ' 3 °llllTL°i ll! l l "383 ._l_l_ .3; l l l L l. l Lubricated Unlubricated Figure 10. Typical shape of fluid in relaxed state for smooth, lubricated and unlubricated plungers. 51 The consistency of result are also affected by annular gap. The smaller K value, the more uniform the yield stresses obtained. Good reproducibility of experimental data is very important in yield stress measurement because most methods have failed to produce uniform results. Thus, it may be concluded that reproducible data will be obtain in back extrusion testing,to determine yield stress when.using a lubricated plunger and K - 0.52. The effect of solid boundaries may be explained on the basis of the force balance along the immersed plunger during stress relaxation. Two possibilities may affect the distribution: shear rate history and solid boundaries, because a narrower annular gap also produce higher shear rate. However, due to the purpose of this experiment only to investigate solid boundaries effect, it is assumed no strain history effect was present at equilibrium. The experiments show that a narrower annular gap (higher'K value) results in more rapid relaxation and sometimes a force with opposite direction to buoyancy force occurs on the plunger (negative balance at three hour period) . The negative yield stress may not be interpreted in physical meaning, but it may due to a mistake of mathematical expression. This phenomenon, which can not be understood, was also found when polyisobutylene in primal was subjected to a high shear rate (Bird et al., 1987). The stress, then, recovered to the same equilibrium as found with lower shear rates for longer observation times. It also appeared in back extrusion tests with high K values when observed for 24 hours, but the data is not reliable due to dehydration during the observation period. 52 Table 5. Yield stress of tomato paste measured using lubricated aluminum plungers, with different ratios (K values) of plunger to inner graduate cylinder radius after various relaxation period. Ratio plunger 00 (Pa) 00 (Pa) 00 (Pa) to tube radius 1 hour 2 hours 3 hours K-0.78 62.3 47.7 -l4.1 73.7 8.2 -l4.l 11.4 -4.4 -1l.4 Average 49.1 17.2 -13.2 Std.dev. 33.2 27.2 1.5 K-0.67 69.2 57.4 40.7 100.0 66.6 37.6 106.0 86.7 72.5 Average 91.7 70.2 50.3 Std dev. 19.7 15.0 19.3 K-0.52 94.6 73.0 63.2 105.4 101.6 87.2 110.1 105.4 91.8 Average 103.4 93.3 80.7 Std.dev. 7.9 17.7 15.4 53 dam to_:cco “actuate use 5933 cousins. a 05m: co_mabxo soon 5 ounce 39:3 to ciao cozoxgot mmobm .: 059... AmoSEEV arc: CON om? on? 0: ON? 09 cm ow ow ON 0 o m Pr 1 .fl \_ 1 Fl Fl. E E rl .ll E Fl I: N f. m 3 a l m ) Boone. A. E /N\ Psoux m. l m nnndnx .®l - r -1....l..-l..,l Ell- lie 2 54 5.1.4 Different Plunger Velocities and Depths of Penetration The early observations revealed that friction on the plunger surface and a narrow annular gap had strong effects on yield stress evaluation. of .semi-solid foods. To investigate 'whether plunger movement affected the measurement of yield stress, other effects should be eliminated or constant, so a lubricated aluminum plunger and a wide annular gap (K-0.52) were used. There are two ways to obtained different plunger movement in back extrusion at the same annular gap with the lubricated plunger. First, the foods are measured with the same depth of penetrations but different plunger velocities. Second, the foods are measured with the same plunger velocities but different depths of penetration as in Fig. 12 and 13, respectively. Fig. 12 demonstrates that different plunger velocities do not affect yield stress measurement even though they result in different shear stresses. In three hours of observation, the yield stress of chicken frankfurter batter is still time-dependent (Fig. 12 or Table 6). The yield stresses measured every hour do not have significant differences between 50 mm per minute velocity and 100 mm per minute velocity. 55 .mo_:oo_o> 80:33 ecototfi pco toocga negates a open: co_m:bxo soon E teflon 50.35 to o>t3o cozoxsot mmobm .NF 333m AmoyzEEv act: com 09 Om: 0: ON_ COP om 00 CV ON 0 4 d 4 .— 1 1 4| 1 ‘ A3 ‘3 .— l1 q ‘ d J O A .sE\EE 09 Al m .EE\EE on .3 l .l ll-l.l -lllllll-l .-.ll .lllllll- .l. . - l OF 56 Table 6. Yield stress of chicken batter measured using different plun er velocities with a lubricated aluminum plunger and K-0. 2. Plunger velocity 00 (Pa) 00 (Pa) 00 (Pa) 1 hour 2 hours 3 hours 50 mm/minute 177.5 140.0 119.2 190.8 144.3 119.5 188.8 138.2 115.0 Average 185.7 140.8 117.9 Std. dev. 7.2 3.1 2.5 100 mm/minute 171.0 139.2 119.8 202.7 150.9 122.8 178.5 133.5 111.2 Average 184.1 141.2 117.9 Std. dev. 16.6 8.9 6.0 57 On the other hand, Fig. 13 shows that different depths of penetrationlhave produced different equilibrium forces. These forces were observed only for one hour period. Since yield stress is calculated from equilibrium forces , with consideration of the depth of penetration or the length of immersed plunger, we are not able to see the value of those yield stresses in Fig. 13. However, Table 7, shows that the yield stresses calculated from Fig. 13 have almost the same magnitude. The equilibrium forces per unit length of immersed plunger also can be seen in Table 7. However, different depths of penetration (Fig. 14) have different results. When plunger penetration is close to the bottom of sample tube. The end effects produce erratic curves even to a negative force when operated 1.5 cm from the bottom of sample tube. Therefore, to eliminate this effect, the plunger has to be operated far enough (at least 3 cm for the current study or 1.5 of plunger diameter) from the bottom of the sample tube. 58 Table 7. Yield stress of tomato paste measured with different depth of penetrations using lubricated aluminum plunger with K-0.62. Length of immersed Fr. 00 Full plunger (m) (N) (Pa) (N/m) 0.110 2.4 166.3 22.1 0.129 2.7 152.3 20.9 0.173 3.5 145.4 20.3 0.194 4.0 152.0 20.8 0.227 5.2 175.1 22.9 0.270 5.6 149.2 20.6 Average 156.7 Std. dev. 11.4 59 .cozobocoa to mfiaop 296:6 pco toocsa negates. a 0:6: co_m:.axo soon E Smog 30:3“ *o 023 cozoxsot mmobm .n— 9.sz APSEEEV actr— 60 .moucoumfi pco «c3035 pco 50:33 95035:. a co_m:bxo xuon E Bmoa 8083 to o>t3u canoxsot mmobm .3. 8:9... AmoSEEV oEfi OON om? Om: of: GNP 00— om om ow ON 0 WHEN—Na «l4 . O v m J... N— m or D 1. ON\8} .H com/NK Eu 0.4 rm. WWW Eu OH. rm. in mm . l 0% rl 2»? All mg: 61 5.1.5 Time-Dependent Yield Stress The magnitudes of equilibrium forces and.yield stress calculated from them were determined for data one hour after the plunger was stopped. At that time it is expected that the force has already decayed to equilibrium state and strain history has no effect on the force on the plunger surface. However, in a longer observation time (Table 8), the value of yield stress calculated during stress relaxation is time-dependent, even after three hours. The longer the observation time, the smaller the calculated yield stress. For a lubricated plunger, the residual force is reduced at a constant rate; however, the residual force does not follow a specific pattern with the an unlubricated plunger. For example, a higher yield stress could be calculated from the data (Fig. 15) when determined at 180 ndnutes than determined at 160 minutes after plunger was stopped The average yield stress measured using a lubricated plunger seems higher than those found using an unlubricated plunger but the differences are not statistically significant (Table 8). Thus, this experiment shows that lubricated plunger (with K-0.52) increases uniformity, but does not alter the determined yield stress. 62 .mmdux 53 .6333 poSEBaE: co pco payouts... o mEm: 363:6 xoon E 383 03:13 to 633 5388.9. mmobm .9 65¢: AmoSEEV oEc. com 09 cm: oi ONF om: om om 04 cm 0 m; o.— :3 N m 3 em a . .. m N /N\ NM. 33373333 .®l 9m. E *loououtnj A: 63 Table 8. Yield stress of tomato paste measured using lubricated and unlubricated plungers with K-O.52. Plunger 00 (Pa) 00 (Pa) 00 (Pa) surface 1 hour 2 hours 3 hours Lubricated 86. 65. 56.3 96. 92. 79.0 100. 96. 83.4 Average 94. 84. 72.9 Std. dev. 3. 5. 5.3 Unlubricated 86. 112. 87.5 105. 70. 48.4 76. 55. 34.6 Average 89. 79. 56.8 Std. dev. 14. 12. 11.2 64 5.2 Comparison of Back Extrusion to Other Methods 5.2.1 Comparison to Gun Rheometer Data The gun rheometer was used for comparison because it is a reliable device, not relying on a spring like the rotary viscometer method and able to measure any type of fluid. Table 9 shows the results of yield stress measurement using back extrusion and gun rheometry for peanut butter, tomato paste and chicken batter. Before comparing the two methods, it is important to explain why the gun rheometer failed to measure the yield stress of the chicken batter. As described in a previous section, the way the gun rheometer measures yield stress is to determine a minimum pressure required to initiate flow. In the case of chicken batter, the sample did not flow, but simply slid out of the sample tube. A combination of high yield stress and high fat content may have caused this problem. The minimum pressure needed to create sliding varied from 3.5 kPa to 30 kPa for the same 20 cm sample length. It is impossible to accept those values for yield stress calculations. The yield stress of peanut butter obtained from the gun rheometer is higher than that determined from back extrusion after three hours relaxation time. In contrast, the yield stress of tomato paste is lower. There are many possible reasons why those values are not comparable. 65 Table 9. Yield stress of peanut butter, tomato paste and chicken batter measured using gun rheometer and back extrusion (lubricated plunger and K-0.52). Method 00 (Pa) 00 (Pa) 00 (Pa) 1 hour 2 hours 3 hours Back extrusion Peanut butter 103.2 45.2 21.7 124.6 66.8 29.7 136.7 68.8 30.0 Average 121.5 80.2 27.2 Std. dev. 17.0 13.1 4.7 Chicken batter 177.5 139.9 119.2 190.8 144.3 119.5 188.8 138.2 115.0 Average 185.7 140.8 117.9 Std. dev. 7.2 3.1 2.5 Tomato paste 94.6 73.0 63.2 105.4 101.6 87.2 110.1 105.4 91.8 Average 103.4 93.3 80.7 Std. dev. 7.9 17.7 15.4 Gun rheometer Yield stress Coeff. corr. Peanut butter 40.2 0.703 Tomato paste 30.6 0.984 Chicken batter failed failed 66 First, the relaxation for samples in back extrusion may require a longer time than three hours, as seen in Fig. 18. It is true, for all three samples, that their curves of reducing force have not reached equilibrium states after three hours relaxation time. The curve for chicken frankfurter batter and tomato paste exhibit constant decreasing force with time. But the curve of peanut butter reduces with an inconsistent pattern. The pattern is similar to the curve of tomato paste (Fig. 11) when it measured in the narrower annular gap (K-0.78), but in the wider gap the erratic curve disappears and a higher yield stress is obtained after three hours relaxation. In conclusion, peanut butter may need a wider annular gap, greater than K=0.52, to obtain the best curve. It may be true that each fluid measured using back extrusion needs a specific condition to obtain a true yield stress. For example, measuring the yield stress of Mixed Cereal can be satisfactory accomplished with K-0.67 and a relaxation time less than one hour. However, chicken frankfurter batter, tomato paste and peanut butter require a relaxation time longer than three hours. Also a K value less than 0.52 is needed for peanut butter to achieve optimum results. In the gun rheometer method, the difficult aspect of measuring yield stress is to determine the minimum pressure (to cause flow) due to the fact that the fluid may not flow but only slide out (debond from solid surface). A low coefficient for correlation of 0.703 for peanut butter, when the data of pressure drop were extrapolated to zero velocity with linear regression method, may be caused by slip (Fig. 16 and 17). To prevent slip, Goodrich et a1. (1989) placed internal "fins" in the sample tube when measured yield stress of gels. The same technique may be applied to the gun rheometer to eliminate the slip. 67 mm .Bmoa BoEo“ t2 Boo touoEoof :30 do «2n. .3 659... A.c_E\EEV b_uo_o> -u_-‘-..*___—.MM- ON or o— m o * ‘1 d 1 — d A 4 o L om - -- O J Ow (Dd) ssang L om & 1 oo— 68 9 NF .cmtan Scooa cow Son 33:52.. cam do “oi .5 0.59... A.c_E\EEV 3_uo.o> m 41 fl '— 1 1 J on ov on on Do. (od) SSSJIS 69 .deux £3, tomc:_a.E:c_E:_o 6033.53. 0 oEm: Bmoa BoEou oco tozan Scooa .tozon coquu do 033 cozoxsoc mmobm .9 0.39... AmoSEEV oEfi OON owe 00— 0: ON— 00? om om o¢ ON 0 4 4111 1 «14 J . a 4 u 41. q d _ 1 O N} y l C C C AA [A N m. u a w 3 ) o (Nx , 323 22.2 .m: w . _ .6345 Stood 10: a w _ A r .532. 5573 1&1 .. 70 5.2.2 Comparison to Vane Device Data It was very difficult to compare back extrusion results to the results of the vane method because of the high yield stress samples. A popular method, that is believed by some researcher as the most reliable method, the vane device could not be applied to peanut butter, tomato paste and chicken frankfurter batter using the HBTD Brookfield viscometer due to the yield stress of the samples being too high for the torque capability of the viscometer. Since back extrusion has capability of measuring yield stress of any type of fluid, from fluid like to solid like, Mixed Cereal baby food was used as a sample in compaang the vane rheometer to back extrusion method. Mixed Cereal from Gerber was measured using three methods: back extrusion, vane device and gun rheometer. The results of these methods can be seen at Table 10. Yield stresses obtained by the vane device have small standard deviations, but depend upon the rotational velocity of the vane. Although Dzuy and Boger (1983) found that rotational velocity from 0.1 rpm to 8 rpm produced consistent results when they measured 66 percent of Red mud, yield stresses of mixed cereal measured using 0.5 to 2.5 rpm gave different results. The theory of the vane device described.by Dzuy and Boger (1983) requires a very low rotation of the vane. Unfortunately, the Brookfield viscometer used in this experiment did not have a rotational speed lower than 0.5 rpm. The effect of rotational speeds was not expected for high yield stress materials (Dzuy and Boger, 1983), but for low yield stress materials, the viscous resistance together with instrument inertia as described by Dzuy and Boger (1985) has already occured at 0.5 rpm vane speeds. If the value of yield stress at zero 71 rpm was obtained using the extrapolation method, the yield stress of mixed cereal measured by vane device was 12.58 Pa with correlation coefficient is 0.978. This magnitude is still higher when compares to the result of that measured by either the gun rheometer or the back extruder. These results agreed with previous experiments indicating that stress relaxation methods give results lower than those of the stress to initiate flow methods (Dzuy and Boger, 1985). Back extrusion with a lubricated plunger has a higher yield stress than that found with an unlubricated plunger even though is not statistically significant at 0.05 confidence level. This clearly shows, once again, that the lubricated plunger will produce more uniform result than the unlubricated plunger. The gun rheometer has the lowest yield stress value: 4.47 Pa. The distinct results among three methods make it difficult to conclude which value of them is the actual yield stress. The results of the vane device used in this experiment are not perfect due to the lack of low rotational speed even though the data show small standard deviations. The result of the gun rheometer is also obtained from extrapolation to zero velocity with correlation coefficient is 0.903. The back extrusion method also has a weakness because the fluids may have a time-dependent yield stress when measured using stress relaxation methods. 72 Table 10. Yield stress of Mixed cereal measured with back extrusion, vane device and gun rheometer. Methods Specific 00 (Pa) Std. dev. treatment (Pa) Vane device 2.5 rpm 22.9 0.3 1.0 rpm 17.6 0.4 0.5 rpm 13.8 0.8 Back extrusion Lubricated 7.8 1.1 (K—0.52) Unlubricated 4.4 1.7 Gun rheometer 4.5 r-0.903 73 .tomcga 6303525 co oco “.8329: 0 05m: co_m:bxm xuon E 60.8 ooze: ._o 0?.3 cozoxoxt mmobw .9 050E AmoSEEV oEfi om. ow on cm 9 o ooSfiBEc: mel l m _ 88253 A: .. o as - nil-1:11, 13.1.1 1!!!!- -.!Ila 15:13:11.1! .- O (N) awoi Chapter 6 Conclusion The experiments showed that the back extrusion method can be used to measure yield stress of semi-solid foods. However, some specific conditions may be required to obtain reproducible and reliable values. The conditions are related to selection of plunger surface, annular gap and observation time. Back extrusion using grooved and unlubricated smooth plungers revealed that the assumption of no slip on the annular wall for measuring other rheological parameters will be a critical factor which result in non- reproducible values in yield stress measurement. A lubricated plunger applied in yield stress measurement to avoid an uncontrollable friction factor produce more consistent yield stress value than using an unlubricated plunger. In principle, a wide annular gap is needed to measure yield stress. Every fluid has a wide limit for producing uniform curve, below that limit the curve will be inconsistent K-0.52 for chicken frankfurter batter and tomato paste, and smaller K is needed for peanut butter). Each fluid also needs a different relaxation time to reach an equilibrium state. Some fluid may achieve an equilibrium state within one hour (Mixed Cereal baby food) , but other fluid may need more than three hours observation time. Otherwise the measured yield stress is still time-dependent (chicken frankfurter batter, tomato paste and peanut butter). Yield stress measurement should be established separately from mea- suring other rheological parameters in the case of back extrusion testing. The value of yield stress obtained using a stress relaxation method, such 74 75 as back extrusion, is lower than the vane device method. The true value of yield stress can not be determined during this study due to equipment difficulties and material time-dependency. Chapter 7 Future Research A major disadvantage found in measuring the yield stress of semi-solid foods using back extrusion was a very long observation time. The equilibrium state was not reached even over three hours of observation and the yield stress determined during that observation time was still time-dependent. Furthermore , to keep the sample from environmental effects was very difficult and dehydration or temperature changes may be a problem. Therefore, future research is needed to find a method of estimating the yield stress from shorter observation times which minimize environmental effects. The stress relaxation method produced different yield stress values compared to the stress to initiate flow methods. It is important to observe, in more detail, the behavior of the fluid during stress relaxation. However, the yield stress determined with back extrusion may be pueferable if compared to other stress relaxation methods, involving other instruments such as a concentric cylinder viscometer. The experimental evaluation of the strain history effect is also important because the negative value of yield-stress encountered.in testing with K-0.78 (after three hours observation) may be caused by a high preceding shear rate rather than narrow solid boundaries. 76 Chapter 8 Bibliography Bagley, E.B., Christianson, D.D. and Wolf, W.J. 1985. Frictional effects in compressional deformation of gelatin and starch gels and comparison of material response in simple shear, torsion, and lubricated uniaxial compression. J. Rheology, 29(1): 103-108. Bagley, E.B., Wolf, W.J. and Christianson, D.D. 1985. Effect of sample dimensions, lubrication and deformation rate on uniaxial compression of gelatin gels. Rheologica Acta, 24(3): 265-271. Bagley, 8.8. and Christianson, D.D. 1988. Uniaxial compression of viscoelastic rings-effect of friction at the platen/sample interface for gels and doughs. J. Rheology, 32(6): 555-573. Balmaceda, E., Rha, C. and Huang, F. 1973. Rheological properties of hydrocolloids. J. Food Science 38(7): 1169-1173. Barbosa Canovas, G.V. and Peleg, M. 1983. Flow parameters of selected commercial semi-liquid food products. J. Texture Studies 14(3): 213-234. Barnes, H.A. and Walters, K. 1985. The yield stress myth?. Rheologica Acta 24(4): 323-326. Bhatatacharyya, G.K. and-Johnson, R.A. 1977. Statistical Concepts and Methods, John Wiley & Sons, New York. Bingham, E.C. 1922. Fludity and Plasticity, McGraw-Hill, New York Bird, R.B., Armstrong, R.C. and Hassager, 0. 1987. Dynamics of Polymeric Liquids Vol. 1 (Fluid Mechanics) 2nd ed. p. 123, John Wiley & Sons, New York. Campanella, O.H. and Peleg, M. 1987. Determination of yield stress of semi-liquid foods from squeezing flow data (research note) . J. Food Science 52(1): 214-217. Campanella, O.H. and Peleg, M. 1988. On food compression by soft machines. J. Texture Studies 19(1): 39-50. Casson, N. 1959. A flow equation for pigmented-oil suspension of the printing ink type. In Rheology of Disperse Systems, (C.C. Mill, ed.) pp. 82-104, Pergamon Press, New York. Chatraei, Sh., Macosko, C.W. and Winter, H.H. 1981. Lubricated squeezing flow: a new biaxial extensional rheometer. J. Rheology 25(4): 433-443. 77 78 Cheng, D.C-H. 1986. Yield stress: a time-dependent property and how to measure it. Rheologica Acta 25(5): 542-554. Christianson, D.D., Casiraghi, E.M. and Bagley, 8.8. 1985. Uniaxial compression of bonded and lubricated gels . J. Rheology, 29(6): 671-684. Covey, G.H. and Stanmore, B.R. 1981. Use of the parallel-plate plastometer for the characterization of viscous fluids with a ield stress. J. Non-Newtonian Fluid Mechanics 8(3+4): 24 -260. De Kee, D., Turcotte, G., Fildey, K. and Harrison, B. 1980. New method for the determination of ield stress (research note). J. Texture Studies 10(3): 281- 88. Dienes, C.J. and Klemm, H.F. 1946. Theory and application of the parallel plate plastometer. J. Applied Physics 17(6): 458-471. Dixon, B.D. and Parekh, J.V. 1979. Use of the cone penetrometer for testing the firmness of butter. J. Texture Studies 10(4): 421-434. Dzuy, N.Q. and Boger, D.V. 1983. Yield stress measurement for concentrated suspensions. J. Rheology 27(4): 321-349. Dzuy, N.Q. and Boger, D.V. 1985. Direct yield stress measurement with the vane method. J. Rheology 29(3): 335-347. Elliott, J.H. and Ganz, A.J. 1977. Salad dressings - preliminary rheolo ical characteristics. J. Texture Studies 8(3): 359-37 . Gent, A.N. 1960. Theory of the parallel plate viscometer. British J. Applied Physics 11(2): 85-87. Goodrich, K., Yoshimura, A. and Prud'homme, R.K. 1989. Measurement of the modulus and yield stren th of soft els: experiments and numerical simulation. J. eology 33(2%: 317-327. Heinz (1959) quoted in Fincke, A. 1961. Beitrage zur Losung rheologisher probleme in der Schokoladentechnologie, Dis- sertation, TH Karlsruhe. Herschel, W.H. and Bulkley, R. 1926. Konziztenzmessungen von gummi-bensollosugen. Kolloid-Zeitzchr. 39, 291-300; Proc. Am. Soc. Test. Matls. 26, 621-633. James, A.E., Williams, D.J.A. and Williams, P.R. 1987. Direct measurement of static yield pro erties of cohesive suspen- sions. Rheologica Acta 26(5): 4 7-446. Kaletunc-Gencer, G. and Peleg, M. 1984. Di itizer aided deter- mination of yield stress in semi-liquid oods. J. Food Science 49(6): 1620-1621. Keentok, M. 1982. The measurement of the yield stress of liquids. Rheologica Acta 21(3): 325-332. 79 Keentok, M. , Milthorpe, J.F. and O’Donovan, E. 1985. On the shearing zone around rotating vanes in plastic liquids: theory and experiment. J. Non-Newtonian Fluid Mechanics 17(1): 23-35. Lang, E.B. and Rha, C. 1981. Determination of the yield stress of hydrocolloid dispersions. J. Texture Studies 12(1): 47-62. Leider, P.J. and Bird, R.B. 1974. Squeezing flow“between parallel disks. i. Theoretical analysis. Ind. Eng. Chem., Fundam. 13(4): 336-341. Leider, P.J. and Bird, R.B. 1974. Squeezing flow“between.parallel disks. ii. experimental results. Ind. Eng. Chem., Fundam. 13(4): 342-346. Marte-Guzman, N.J. 1987. Non-Newtonian shear rate approximation and optimal experimental conditions in back extrusion testing. M.S. Thesis, Department of Food Science and Human Nutrition, Michigan State University. Mooney, FL 1958. Rheology (Elrich, P.R. ed.) 2(5):195-196. Academic Press, New York. Nagase, Y. and Okada, K. 1986. Heterogeneous behavior after yielding of solid suspensions. J. Rheology 30(6): 1123-1142. Nakayama, T., Niwa, E. and Hamada, I. 1980. Pipe transportation of minced fish paste. J. Food Science 45(4): 844-847. Nguyen, Q.D. and Boger, D.V. 1985. Thixotropic behavior of concentrated bauxite residue suspensions. Rheologica Acta 24(4): 427-437. Nguyen, Q.D. and Boger, D.V. 1987. Characterization of yield stress fluids with concentric cylinder viscometers. Rheo- logica Acta 26(6): 508-515. Ofoli, R.Y., Morgan, R.C. and Steffe, J.F. 1987. A generalized rheological model for inelastic fluid foods. J. Texture Studies 18(3): 213-230. Osorio-Lira, F.A. 1985. Back extrusion of power law, Bingham plastics and Herschel-Bulkley fluids. M.S. Thesis, Department of Food Science and Human Nutrition, Michigan State Uni- versity. Osorio, F.A. and Steffe, J.F. 1987. Back extrusion of power law fluids. J. Texture Studies 18(1): 43-63. Princen, H.H. 1985. Rheology of foams and highly concentrated emulsions. J. Colloid and Interface Science 15(1): 150-171. Princen, H.M. 1986. A novel design to eliminate end effects in a concentric cylinder viscometer. J. Rheology 30(2): 271-283. Rha, C. 1978. Rheology of fluid foods. Food Technology 32(6): 77-82. 80 Steffe, J.F. and Osorio, F.A. 1987. Back extrusion of non-Newtonian fluids. Food Technology 41(3): 72-77. Tanaka, M., De Man, J.M. and Voisey, P.W. 1971. Measurement of textural properties of foods with constant speed cone penetrometer. J. Texture Studies 2(3): 306-315. Van Wazer, J.R., Lyons, J.W., Kim, K.Y. and Colwell, R.E. 1963. Viscosity and Flow Measurement. pp. 292-295. Wiley, New York. Vocadlo, J.J. and Charles, M.E. 1971. Measurement of yield stress of fluid-like viscoplastic substances. The Canadian J. of Chemical Engineering 49(10): 576-582. Yoshimura, A.S., Prud'Homme, R.K., Princen, H.M. and Kiss, A.D. 1987 . A comparison of techniques for measuring yield stresses. J. Rheology 31(8): 699-710. Yoshimura, A.S. and Prud'Homme, R.K. 1988. Wall slip corrections for couette and parallel disk‘viscometers. J. Rheology 32(1): 53-67. Appendix A Schematic diagram of back extrusion experiments Sample prepcrotion loading / With Iii-4"“ volume & weight deneity calculation recordc cha't paper printl' 24 hr storage chicken 5 C other: 25 C l teeting 100 mm/min. 100 mm l absewation l to 3 Iva. immreed pl. bolonnce farce calculation yield stren —-—————> calculation 81 plot of relaxation cu've 82 Appendix B Results of yield stress measurement using back extrusion method. Table Bl. Data of chicken batter obtained from back extrusion testing using plunger radius of 0.0142 m, graduate cylinder radius of 0.0178 m and plunger speed of 100 mm/min. Plunger FT FT. [m Fo L do (N) (N) (In) (N) (In) (Pa) Plexiglass smooth 35.69 10.49 0.116 1.1 0.17 615.3 32.52 8.98 0.118 1.1 0.17 506.8 34.97 9.14 0.112 1.1 0.16 548.4 grooved 40.74 11.15 0.111 14) 0.16 691.4 44.39 13.34 0.118 1.1 0.17 787.0 36.97 11.99 0.109 1.0 0.16 763.9 Teflon smooth 36.75 12.28 0.116 1.1 0.17 732.4 41.46 12.90 0.112 1.1 0.16 803.2 37.40 11.27 0.105 1.0 0.15 743.6 grooved 34.91 11.93 0.113 1.1 0.17 730.3 40.28 11.97 0.107 1.0 0.16 778.1 36.75 10.63 0.106 1.0 0.15 690.1 Aluminum smooth 30.66 11.53 0 113 1.1 0.17 703.4 37.93 9.79 0.112 1.1 0.16 593.0 35.93 7.19 0.112 1.1 0.16 416.1 grooved 40.63 14.44 0.108 1.0 0.16 943.8 45.05 14.11 0.117 1.1 0.17 844.3 45.26 11.96 0.112 1.1 0.16 739.5 83 Table BZ. Data of peanut butter obtained from back extrusion testing using plunger radius of 0.0142 m, graduate cylinder radius of 0 0178 m and plunger speed of 100 mm/min. Plunger FT FT. [’Ch F0 L 00 (N) (N) (In) (N) (111) (Pa) Plexiglass smooth. 36.51 1.98 0.102 1.0 0.15 66.8 44.54 2.04 0.104 1.1 0.15 71.4 38.31 1.38 0.103 1.0 0.15 24.5 grooved 45.02 3.10 0.102 1.0 0.15 153.5 45.14 2.70 0.085 0.9 0.13 161.2 45.15 2.55 0.059 0.6 0.09 250.5 Teflon smooth 45.11 0.81 0.102 1.0 0.15 -l7.3 45.05 1.11 0.071 0.7 0.10 43.4 45.10 3.79 0.092 0.9 0.13 235.4 grooved. 33.14 2.87 0.103 1.0 0.15 134.1 45.02 4.29 0.100 1.0 0.15 248.4 41.56 1.14 0.105 1.1 0.15 5.2 Aluminum smooth 44.58 1.82 0.103 1.0 0.15 56.8 45.01 2.61 0.086 0.9 0.13 153.0 45.15 2.64 0.102 1.0 0.15 118.8 grooved. 33.81 3.74 0.104 1.6 0.15 195.4 36.83 1.65 0.091 0.9 0.13 60.2 33.25 2.71 0.090 0.9 0.13 150.9 84 Table B3. Data of tomato paste obtained from back extrusion testing gusing plunger radius of 0. 0142 m, graduate cylinder radius of 0 017 m and plunger speed of 100 mm/min. Plunger F, Fr. L” F, L do (N) (N) (In) (N) (111) (Pa) Plexiglass smooth 19.23 1.18 0.112 1.2 0.16 -1.1 21.32 4.84 0.139 1.5 0.20 183.1 21.54 5.85 0.143 1.5 0.21 229.3 grooved. 18.45 3.69 0.106 1.3 0.15 183.7 19.15 5.93 0.130 1.4 0.19 265.0 20.85 4.50 0.140 1.5 0.20 163.0 Teflon smooth 19.76 1.89 0.113 1.2 0.17 45.8 16.89 4.50 0.110 1.2 0.16 229.6 18.96 6.62 0.123 1.3 0.18 327.7 grooved 20.17 3.03 0.107 1.1 0.16 133.7 16.56 3.94 0.132 1.4 0.19 145.7 16.03 3.02 0.129 1.4 0.19 96.4 Aluminum smooth i17.32 4.89 0.122 1.3 0.18 224.1 21.35 3.26 0.139 1.5 0.20 98.1 26.17 6.75 0.165 1.7 0.24 230.7 grooved 20.88 1.99 0.124 1.3 0.18 40.8 20.26 5.15 0.130 1.4 0.19 219.6 23.05 2.57 0.133 1.4 0.19 65.5 85 Table B4. Data obtained from back extrusion testing using a lubricated aluminum plunger with radius of 0.0142 m, graduate cylinder radius of 0.0178 m and plunger speed of 100 mm/min. Plunger F, FT. Lm Fo L do (N) (N) (m) (N) (In) (Pa) Chicken batter 23.99 6.88 0.110 1.1 0.16 403.8 24.30 5.57 0.112 1.1 0.16 306.5 21.97 5.40 0.109 1.0 0.16 304.9 Peanut butter 33.81 1.90 0.105 1.1 0.15 60.2 ‘ 36.83 1.81 0.102 1.0 0.15 57.2 33.25 2.00 0.103 1.0 0.15 70.3 Tomato paste 15.48 3.51 0.118 1.2 0.17 145.5 19.44 4.04 0.132 1.4 0.19 152.0 24.81 5.21 0.155 1.6 0.23 175.1 86 Table B5. Data of tomato paste obtained from back extrusion testing using a lubricated aluminum plunger with different ratios of plunger to graduate cylinder diameter (Vp - 100 mm/min.) Plunger F} Fr. L” Fo L do (N) (N) (In) (N) (111) (Pa) 1 hour K - 0.78 9.04 1.58 0.120 0.7 0.20 68.9 (R; - 0.0128 an 9.53 1.71 0.119 0.7 0.20 81.0 (1?,- 0.01 m) 10.36 0.86 0.114 0.7 0.19 15.2 K - 0.672 6.23 1.68 0.128 0.8 0.15 76.8 (Roz- 0.0178 m) 6.84 1.93 0.120 0.7 0.14 109.3 (R,- 0.012 m) 6.87 2.08 0.125 0.8 0.15 115.6 K - 0.52 3.37 1.44 0.131 0.6 0.12 94.7 (R; - 0.0232 m) 3.55 1.54- 0.131 0.6 0.12 105.4 (R,-— 0.012 m) 3.45 1.55 0.129 0.6 0.12 110.4 2 hours K - 0.78 9.04 1.38 0.120 0.7 0.20 53.5 9.53 0.85 0.119 0.7 0.20 11.8 10.36 0.66 0.114 0.7 0.19 -1.5 K - 0.67 6.23 1.54 0.128 0.8 0.15 64.3 6.84 1.55 0.120 0.7 0.14 74.1 6.87 1.85 0.125 0.8 0.15 95.3 K - 0.52 3.37 1.25 0.131 0.6 0.12 74.0 3.55 1.50 0.131 0.6 0.12 101.6 3.45 1.51 0.129 0.6 0.12 105.4 3 hours K - 0.78 9.04 0.56 0.120 0.7 0.20 -11.7 9.53 0.56 0.119 0.7 0.20 -11.7 10.36 0.57 0.114 0.7 0.19 -8.9 K - 0.67 6.23 1.33 0.128 0.8 0.15 46.7 6.84 1.21 0.120 0.7 0.14 43.4 6.87 1.68 0.125 0.8 0.15 80.3 K - 0.52 3.37 1.16 0.131 0.6 0.12 63.2 3.55 1.38 0.131 0.6 0.12 87.2 3.45 1.40 0.129 0.6 0.12 91.8 87 Table B6. Data of chicken batter obtained from back extrusion testing using a lubricated aluminum plunger with different plunger speeds (K - 0.52). Plunger F, FT. L." F, L do (N) (N) (In) (N) (111) (Pa) 1 hour 50 mm/min. 8.36 2.20 0.137 0.6 0.12 177.5 8.75 2.15 0.127 0.5 0.11 190.8 8.66 2.10 0.125 0.5 0.11 188.8 100 mm/min. 9.70 1.94 0.062 0.5 0.11 171.0 9.90 2.20 0.062 0.5 0.11 202.7 10.45 2.03 0.062 0.5 0.11 178.5 2 hours 50 mm/min. 8.36 1.85 0.137 0.6 0.12 140.0 8.75 1.75 0.127 0.5 0.11 144.3 8.66 1.67 0.125 0.5 0.11 138.2 100 mm/min. 9.70 1.67 0.062 0.5 0.11 139.2 9.90 1.77 0.062 0.5 0.11 150.9 10.45 1.65 0.062 0.5 0.11 133.5 3 hours 50 mm/min. 8.36 1.66 0.137 0.6 0.12 119.2 8.75 1.54 0.127 0.5 0.11 119.5 8.66 1.48 0.125 0.5 0.11 115.0 100 mm/min. 9.70 1.51 0.062 0.5 0.11 119.8 9.90 1.53 0.062 0.5 0.11 122.8 10.45 1.46 0.062 0.5 0.11 111.2 88 Table B7. Data of tomato paste obtained from back extrusion testing a lubricated and an unlubricated aluminum plunger with di f the depth of penetrations (Vp - 100 mm/min. and K - 0.67). using erent L“ F, Fr. F, L do (In) (N) (N) (N) (m) (Pa) Lubricated 0.075 11.73 2.43 0.8 0.11 166.3 0.088 12.58 2.69 0.9 0.13 152.3 0.118 15.48 3.51 1.2 0.17 145.5 0.132 19.44 4.04 1.4 0.19 152.0 0.155 24.81 5.21 1.6 0.23 175.1 0.184 25.46 5.58 1.9 0.27 149.2 Unlubricated 0.088 14.41 2.79 0.9 0.13 160.7 0.104 16.70 1.80 1.1 0.15 51.1 0.122 17.32 4.89 1.3 0.18 224.1 0.139 21.35 3.26 1.5 0.20 98.1 0.165 26.17 6.75 1.7 0.24 230.7 0.180 29.92 6.88 1.9 0.26 210.2 89 Table B8. Data of tomato paste obtained from back extrusion testing using a unlubricated aluminum plunger (Vp - 100 mm/min. and K - 0.52). Plunger Fr Fr. Len Fb L do (N) (N) (In) (N) (In) (Pa) 1 hour 4.30 1.49 0.135 0.6 0.12 95.5 3.56 1.52 0.123 0.6 0.11 114.6 3.65 1.33 0.129 0.6 0.12 84.7 2 hours 4.30 1.74 0.135 0.6 0.12 122.8 3.56 1.22 0 123 0.6 0.11 78.0 3.65 1.14 0 129 0.6 0.12 62.5 3 hours 4.30 1.50 0.135 0.6 0.12 96.2 3.56 1.02 0.123 0.6 0.11 54.9 3.65 0.95 0.129 0.6 0.12 40.3 90 Table B9. Data of chicken batter, peanut butter and tomato paste obtained from back extrusion testin using a lubricated aluminum plunger (Vp - 50 mm/min. and K - .52). Plunger F, F,. Leo FD L do (N) (N) (m) (N) (m) (Pa) 1 hour Chicken batter 8.36 2.20 0.137 0.6 0.12 177.5 8.75 2.15 0.127 0.5 0.11 190.8 8.66 2.10 0.125 0.5 0.11 188 8 Peanut butter 8.98 1.33 0.117 0.5 0.10 103.2 4.47 1.46 0.114 0.5 0.10 124.6 6.09 1.76 0.129 0.6 0.12 ' 136.7 Tomato paste 3.38 1.44 0.131 0.6 0.12 94.6 3.55 1.54 0.131 0.6 0.12 105.4 3.46 1.56 0.129 0.6 0.12 110 l 2 hours Chicken batter 8.36 1.85 0.137 0.6 0.12 139.9 8.75 1.75 0.127 0.5 0.11 144 3 8.66 1.67 0.125 0.5 0.11 138.2 Peanut butter 8.98 0.90 0.117 0.5 0.10 48.9 4.47 1.00 0.114 0.5 0.10 64.5 6.09 1.17 0.129 0.6 0.12 68.8 Tomato paste 3.38 1.25 0.131 0.6 0.12 73.0 3.55 1.50 0.131 0.6 0.12 101.6 3.46 1.51 0.129 0.6 0.12 105.4 3 hours Chicken batter 8.36 1.66 0.137 0.6 0.12 119.2 8.75 1.54 0.127 0.5 0.11 119.5 8.66 1.48 0.125 0.5 0.11 115.0 Peanut butter 8.98 0.71 0.117 0.5 0.10 24.7 4.47 0.72 0.114 0.5 0.10 28.1 6.09 0.83 0.129 0.6 0.12 30.0 Tomato paste 3.38 1.16 0.131 0.6 0.12 63.2 3.55 1.38 0.131 0.6 0.12 87.2 3.46 1.20 0.129 0.6 0.12 91.8 91 Table B10. Data of mixed cereal baby food obtained from back extrusion testing using an aluminum {lunger with radius of 0 .0142 111, graduate cylinder radius of 0.017 m and plunger speed of 100 mm/min. Plunger F, F,. Leo Fo L do (N) (N) (m) (N) (111) (Pa) Lubricated 5.31 1.42 0.126 1.3 0.18 8.6 5.09 1.43 0.128 1.3 0.19 8.3 4.50 1.33 0.121 1.2 0.18 6.6 Unlubricated 5.77 1.47 0.136 1.4 0.20 5.3 4.99 1.42 0.131 1.3 0.19 5.5 5.55 1.36 0.130 1.3 0.19 2.4 92 Appendix C Stress relaxation curves .mhdnx 53 .895... 6038.53. E:c_E:_o a 55m: 57.3.28 xuon E 325 39:3 do o>.5u c0386.“: mmobm AuouafiEv 0E2. . CON cup 00— 0.: our cop on on o¢ ON a o 950E (N) 93.10;, 93 .hmdux 53, $933 nouou_._n3_ E3c_E3_o o 32.3 co_m3bxo xuon c_ Boon 39:3 do o>c3u cozoxEoc mmobm .No 333E $33535 oEE. OON om— ODF O¢F ONP 00F om on 0% ON 0 . 4| 4 J a 4 41 4 Id q A 1 — la! 4 .- — q d 1 O l ll. u L 3 J..— m .V \J 3 ) omega Iii...- N “ A. m ( n .30.. IIATI .- N do. -lD-l m F .30.. Ilmwll Fill-II sill-I! sill-ill .3.]: all 1 willllil- ill ill. . - ll !l1 0— 94 .denx 53 53:33 330353. E3c_E3_o a 9:3 cox-33:6 xuon c_ Smog 39:3 :0 o>c3u cozoxsoc mmobm .no 339... Am33£Ev 9:; com 09 00— OS. GNP 00.. cm om 0* . ON 0 mooco>o Iii-l n do: l-ATI. L N do: imll L F do: ll®ll 1 r1: -il..- Ii. ..l-.--.l.l-.l-ll--. ii 5:- Eli. .-.--. all OF (N) 33103 94 9:3 co_m3.3xo 503 E ogmoa 33:3 do o>c3u cozoxsoc mmobm oom omF .deflx £3 tomcaa “633753. E35520 o 00— O¢F $333.5 oEfi GNP cap on on O¢. ON 0 omoto>o n .aot N do: F .aot 71"“ 1'11 - II .I 1-!1 0“--.Ill'. 0— .8 85m: (N) 33103 95 .deux 53 .6953 0303.532: E3:_E3_o o mar-.3 co_m3bxo xuon E oumod 39:3 ho o>c3u cozoxEoc mmobm .30 939... com cm? 00— o: $32.35 oE:. ON— 09 cm 8 o.- om o ovato>o ii. _.._.-.I n do: IIAlI N do: IlUl-l P .QQL IIIQI' ‘40-- 1",.1. OF (N) 93103 96 .3506) c_E\EE on :23 8953 EDEEBO “93035:. 0 mafia coasts 503 E .6303 :3??? B «.23 cox—36.0.. mmobm .m0 339... AmmSEEV mEfi 00m owe 03 0: ON—‘ 09 cm 0m 9» ON 0 .J4_4fiT4‘ 1 +q4 1 O N :3 m ¢ 3 a @0085! ) to W 32¢ . NQSAI w Faexmwr -.i--..‘- i -. -. if: 2 97 .3623 .EE\EE on: 5;, .6953 E3£E20 353752 0 9:2. :23:me xuon E Stop 55:: ,6 o>5u cozoxozt 385 .m0 v.50: AmmSEEV mEE com cm? of 03 ON? OOP ow om O¢ ON 00 L l N I... O l w M 3 moo..m>olll 1 \../ L o W MQGLIDI NQSIL... L w P am: IIP. VIE- as! - E r 98 .denx 5:5 50:33 E:£E3_o 333.53. 0 9:3 sewsbxo xuon E .0309 c335 .6 033 5598.9. mmwbm No 930E AmmSEEV mEfi OQN ow— Om: 0: ON— 00— ow Om Ov. ON 00 N :3 m 4» 3 1 9 mmoLw>o III 1 m N n 9: JD: N am: val w _. a8 A! r---i;1-1..--331-..iii-.. llaii- . 9 99 .deux £3 Loocga 835620 8335:. a mafia co_m:.axo xuon 5 8:3 3:03 B 023 cozoxo_o._ mmmbm .mo «.39... AmmSEEV mEfi OON GE 00— o: ONF 09 cm 00 ow ON 0 o C 0 0‘W’fl‘0/IIJN. N . A, j t o 3 .V J 3 a 3305 II. I. \l : o INK m cm: .ml . N Q0.— l®l 4 m P am: 1A! .__..- E I‘ II l.l.-l' I. - I-O I I'l‘f‘-" . '1'.‘||la|!t I v I. I'll...‘ Il‘ltlv'l ' . ".l|-llll.'l. l"l'll! o— 100 .deux 53, $933 EsEEEo 250353 c mEm: cEmabxm xuon E Smog 39:3 *0 m>.5u cozoxEoL mmmbm .00 0.59... AmQSEEV mE: QON owe 09 O: QNF 00— cm Om ow ON 0 d d 4 I: O .J D a mmob>o III \./ .1 m m n 9.: 1m. E N a8 101 4 m _. am; AI 4 101 .Nndflx £3 .6ch3 EaEEEo 8335:. o mEm: cEmabxo :39 E .033 next: “3 023 c0396.“: mmobm .90 0.59“. AmmSEEv Etc. 00 ow Om. ON or O 1: O .J D 8 «@955 III \/ .. m /N\ n 98 .m1 a a2 1®: 1 m P a9. AI .. O— 102 .Nndflx 53 80:33 E:EE:.o 33352:: :0 oEm: cEmabxo xuon E Bvbu use: .8 033 con—36.0.. mmobm .F —0 0.59... AmBJEEv 9:: on ow on om E o 4.. O J D 8 omega III \.I o W n 8.. .mT N am: um! m — an: 4 OP 103 Appendix D Results of yield stress measurement using a gun rheometer Table 01. Data of tomato paste measured using gun rheometer. Ratio L/D Pressure Pressure Velocity Stress (nunIIgD) (Pa) (mm/min.) (Pa) 13.0 153 1500.376 0.5 28.853 12.7 156 1529.795 1.0 30.114 12.7 182 1784.761 2.0 35.133 12.5 225 2206.435 3.0 44.129 13.5 285 2794.818 8.0 51.756 13.5 309 3030.171 8.0 56.114 13.5 342 3353.782 10.0 62.107 14.0 380 3726.424 10.0 66.543 15.0 480 4707.062 15.0 78.451 15.0 481 4716.869 15.0 78.614 15.6 525 5148.395 16.0 82.506 15.7 578 5668.088 22.0 90.256 Sample tubes had diameter of 10 mm. Yield stress or stress at zero velocity - 30.621 Pa. Coefficient correlation - 0.984 104 Table D2. Data of peanut butter measured using gun rheometer. Ratio L/D Pressure Pressure Velocity Stress (mmHZO) (Pa) (mm/min.) (Pa) 15.0 265 2598.69 1.0 43.31 17.5 305 2990.95 1.0 42.73 14.8 163 1598.44 1.5 27.00 15.0 342 3353.78 3.0 55.90 18.0 463 4540.35 3.5 63.06 17.2 278 2726.17 4.0 39.62 15.8 336 3294.94 5.0 52.13 17.8 378 3706.81 6.0 52.06 18.6 492 4824.74 8.0 64.85 20.0 658 6452.60 10.0 80.66 17.0 476 4667.84 12.0 68.64 17.0 397 3893.13 13.0 57.25 Sample tubes had diameter of 10 mm. Yield stress or stress at zero velocity - 40.152 Pa. Coefficient correlation - 0.703 105 Table D3. Data of mixed cereal measured using gun rheometer. Ratio L/D Pressure Pressure Velocity Stress (mmHZO) (Pa) (mm/min.) (Pa) 13.4 27 264.8 2.0 4.94 19.0 43 421.7 4.0 5.55 13.9 37 362.8 5.0 6.53 19.5 55 539.4 5.0 6.91 13.8 32 313.8 6.0 5.68 14.3 31 304.0 6.0 5.31 19.9 43 421.7 6.0 5.30 14.5 53 519.7 8.0 8.96 20.9 67 657.0 10.0 7.86 18.6 74 725.7 12.0 9.75 15.7 53 519.7 14.0 8.28 13.2 53 519.7 15.0 9.84 21.0 89 872.9 15.0 10.39 22.9 93 912.0 20.0 9.96 16.0 75 735.5 21.0 11.49 Sample tubes had diameter of 10 mm. Yield stress or stress at zero velocity - 4.471 Pa. Coefficient correlation - 0.904 106 Appendix E Results of yield stress measurement using a vane device Table E1. Data of mixed cereal measured using vane device. RPM Windows Tm 00 oo (dyne cm) (dyne cm) (Pa) 0.5 13.0 7474 142.9 14.3 13.0 7474 142 9 14.3 11.8 6785 129.7 13.0 1.0 16.0 9429 180.2 18.0 16.0 9084 173.6 17.4 16.0 9141 174.7 17.5 2.5 21.0 12189 233.0 23.3 20.0 11729 224.2 22.4 21.0 11959 228.6 22.9 'l‘-‘ 1‘. HICHIGRN STRTE UNIV. LIBRQRIES 1|H|IIIHINlllHll"WINIUIWIIWHIIHIHIHINIWI 31293014057420