{"Affiliation":[{"label":"Affiliation","value":"Applied Science, Faculty of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."},{"label":"Affiliation","value":"Chemical and Biological Engineering, Department of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."}],"AggregatedSourceRepository":[{"label":"AggregatedSourceRepository","value":"DSpace","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","classmap":"ore:Aggregation","property":"edm:dataProvider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","explain":"A Europeana Data Model Property; The name or identifier of the organization who contributes data indirectly to an aggregation service (e.g. Europeana)"}],"Campus":[{"label":"Campus","value":"UBCV","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","classmap":"oc:ThesisDescription","property":"oc:degreeCampus"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the name of the campus from which the graduate completed their degree."}],"Creator":[{"label":"Creator","value":"Ein-Mozaffari, Farhad","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/creator","classmap":"dpla:SourceResource","property":"dcterms:creator"},"iri":"http:\/\/purl.org\/dc\/terms\/creator","explain":"A Dublin Core Terms Property; An entity primarily responsible for making the resource.; Examples of a Contributor include a person, an organization, or a service."}],"DateAvailable":[{"label":"DateAvailable","value":"2009-11-11T13:42:09Z","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"edm:WebResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"DateIssued":[{"label":"DateIssued","value":"2002","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"oc:SourceResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"Degree":[{"label":"Degree","value":"Doctor of Philosophy - PhD","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","classmap":"vivo:ThesisDegree","property":"vivo:relatedDegree"},"iri":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","explain":"VIVO-ISF Ontology V1.6 Property; The thesis degree; Extended Property specified by UBC, as per https:\/\/wiki.duraspace.org\/display\/VIVO\/Ontology+Editor%27s+Guide"}],"DegreeGrantor":[{"label":"DegreeGrantor","value":"University of British Columbia","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","classmap":"oc:ThesisDescription","property":"oc:degreeGrantor"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the institution where thesis was granted."}],"Description":[{"label":"Description","value":"Agitated pulp chests provide attenuation of high-frequency disturbances in fibre mass concentration, freeness and other quality factors. This contrasts with process control loops, which attenuate low-frequency variations. Dynamic tests made on industrial stock chests show the existence of non-ideal flows such as channeling, recirculation and dead zones. Since these non-ideal flows reduce the degree of disturbance attenuation from the chests, they have been considered in the dynamic modeling of the chest. This model allows for two parallel suspension flow paths: a mixing zone consisting of a first order plus delay transfer function with a positive feedback for recirculation, and a channeling zone consisting of a first order plus delay transfer function. A new identification method was developed for estimation of dynamic model parameters. A scale-model stock chest was designed and built to study macroscale mixing and disturbance attenuation in a laboratory setting. Fully bleached kraft pulp (FBK) was used for preparation of the pulp suspensions. First preliminary batch studies were made on the scalemodel chest to characterize its behavior and to develop test protocols for use in dynamic tests. Initial tests in batch-mode confirmed established trends for the power required for chest behavior, although existing literature correlations underpredict the power and momentum flux requirements need for complete motion inside the chest. Our visual observation with the aid of a digital video camera showed that the power recommended by existing design criteria is not sufficient to eliminate stagnant zones and even when the whole suspension is in motion, poor mixing regions, where pulp flows significantly slower than in the bulk motion zone, still exist inside the chest. Dynamic response of liquid and solid phase tracers showed that a liquid phase tracer (saline solution) can be used to trace the fibre phase provided the fibre mass concentration is > 2%. It was found that mixing-time for the laboratory chest is both a function of impeller momentum flux and fibre mass concentration. The extent of non-ideal flow in the scale-model chest was evaluated by exciting the system. The process of model identification required two experiments. In the first experiment, the input signal was a rectangular pulse, which allowed the estimation of an approximate model for designing the excitation for the second experiment. The excitation energy for the second experiment was chosen at frequencies where the magnitude of the Bode plot is sensitive to parameter variations. A frequency-modulated random binary input signal was designed for this purpose. Dynamic test results showed that the extent of non-ideal flow and the degree of disturbance attenuation are significantly affected by the location of the input and output in the chest, the fibre mass concentration, the impeller speed and diameter, and the pulp flow rate through the chest. At higher pulp flow rates and fibre mass concentration greater than 3% the system is prone to a high percentage of channeling and dead volume, and a low degree of upset attenuation even at impeller speeds above the criteria of complete motion used to size the chest. Under these circumstances, the degree of disturbance attenuation could be improved by reducing the pulp flow rate through the chest, increasing impeller speed, or decreasing fibre mass concentration. It was found that the degree of upset attenuation is a function of the impeller momentum flux, rather than the power input. Dynamic tests made on scale-model and industrial chests showed that the power calculated based on smooth surface motion and even the onset of complete motion inside the chest does not completely eliminate dead volume and channeling. Additional power is required to have a desired dynamic response from the chest.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/description","classmap":"dpla:SourceResource","property":"dcterms:description"},"iri":"http:\/\/purl.org\/dc\/terms\/description","explain":"A Dublin Core Terms Property; An account of the resource.; Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."}],"DigitalResourceOriginalRecord":[{"label":"DigitalResourceOriginalRecord","value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/14780?expand=metadata","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","classmap":"ore:Aggregation","property":"edm:aggregatedCHO"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","explain":"A Europeana Data Model Property; The identifier of the source object, e.g. the Mona Lisa itself. This could be a full linked open date URI or an internal identifier"}],"Extent":[{"label":"Extent","value":"12902922 bytes","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/extent","classmap":"dpla:SourceResource","property":"dcterms:extent"},"iri":"http:\/\/purl.org\/dc\/terms\/extent","explain":"A Dublin Core Terms Property; The size or duration of the resource."}],"FileFormat":[{"label":"FileFormat","value":"application\/pdf","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/elements\/1.1\/format","classmap":"edm:WebResource","property":"dc:format"},"iri":"http:\/\/purl.org\/dc\/elements\/1.1\/format","explain":"A Dublin Core Elements Property; The file format, physical medium, or dimensions of the resource.; Examples of dimensions include size and duration. Recommended best practice is to use a controlled vocabulary such as the list of Internet Media Types [MIME]."}],"FullText":[{"label":"FullText","value":"MACROSCALE MIXING AND DYNAMIC BEHAVIOR OF AGITATED PULP STOCK CHESTS by F A R H A D EIN-MOZAFFARI B.Sc , Amir Kabir University of Technology, Tehran, Iran, 1986 M . S c , Amir Kabir University of Technology, Tehran, Iran, 1989 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES D E P A R T M E N T OF C H E M I C A L A N D BIOLOGICAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A December 2002 \u00a9 Farhad Ein-Mozaffari, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Cke.micaA O l d Bfolo^tCQ^ t^ ' n c e r i iVJ The University of British Columbia Vancouver, Canada Date DE-6 (2\/88) 11 A B S T R A C T Agitated pulp chests provide attenuation of high-frequency disturbances in fibre mass concentration, freeness and other quality factors. This contrasts with process control loops, which attenuate low-frequency variations. Dynamic tests made on industrial stock chests show the existence of non-ideal flows such as channeling, recirculation and dead zones. Since these non-ideal flows reduce the degree of disturbance attenuation from the chests, they have been considered in the dynamic modeling of the chest. This model allows for two parallel suspension flow paths: a mixing zone consisting of a first order plus delay transfer function with a positive feedback for recirculation, and a channeling zone consisting of a first order plus delay transfer function. A new identification method was developed for estimation of dynamic model parameters. A scale-model stock chest was designed and built to study macroscale mixing and disturbance attenuation in a laboratory setting. Fully bleached kraft pulp (FBK) was used for preparation of the pulp suspensions. First preliminary batch studies were made on the scale-model chest to characterize its behavior and to develop test protocols for use in dynamic tests. Initial tests in batch-mode confirmed established trends for the power required for chest behavior, although existing literature correlations underpredict the power and momentum flux requirements need for complete motion inside the chest. Our visual observation with the aid of a digital video camera showed that the power recommended by existing design criteria is not sufficient to eliminate stagnant zones and even when the whole suspension is in motion, poor mixing regions, where pulp flows significantly slower than in the bulk motion zone, still exist inside the chest. Dynamic response of liquid and solid phase tracers showed that a liquid phase I l l tracer (saline solution) can be used to trace the fibre phase provided the fibre mass concentration is > 2%. It was found that mixing-time for the laboratory chest is both a function of impeller momentum flux and fibre mass concentration. The extent of non-ideal flow in the scale-model chest was evaluated by exciting the system. The process of model identification required two experiments. In the first experiment, the input signal was a rectangular pulse, which allowed the estimation of an approximate model for designing the excitation for the second experiment. The excitation energy for the second experiment was chosen at frequencies where the magnitude of the Bode plot is sensitive to parameter variations. A frequency-modulated random binary input signal was designed for this purpose. Dynamic test results showed that the extent of non-ideal flow and the degree of disturbance attenuation are significantly affected by the location of the input and output in the chest, the fibre mass concentration, the impeller speed and diameter, and the pulp flow rate through the chest. At higher pulp flow rates and fibre mass concentration greater than 3 % the system is prone to a high percentage of channeling and dead volume, and a low degree of upset attenuation even at impeller speeds above the criteria of complete motion used to size the chest. Under these circumstances, the degree of disturbance attenuation could be improved by reducing the pulp flow rate through the chest, increasing impeller speed, or decreasing fibre mass concentration. It was found that the degree of upset attenuation is a function of the impeller momentum flux, rather than the power input. Dynamic tests made on scale-model and industrial chests showed that the power calculated based on smooth surface motion and even the onset of complete motion inside the chest does not completely eliminate dead volume and channeling. Additional power is required to have a desired dynamic response from the chest. iv T A B L E O F C O N T E N T S ABSTRACT ii List of Tables viii List of Figures x Acknowledgements xviii 1 INTRODUCTION 1 2 LITERATURE REVIEW 4 2.1 Introduction 4 2.2 Pulp suspension rheology 5 2.3 Mixing scales 9 2.4 General equation 10 2.4.1 Power consumption for Newtonian fluid 10 2.4.2 Power consumption for non-Newtonian fluids 12 2.4.3 Power consumption for pulp suspension 12 2.4.4 Yackel's method 15 2.5 Chest shape 17 2.6 Impeller types 20 2.7 Dynamic behavior of agitated pulp stock chest 21 2.8 System identification 24 2.8.1 Non-parametric methods 25 V 2.8.2 Parametric methods 26 2.9 Research objectives 27 3 EXPERIMENTAL 28 3.1 Experimental setup 28 3.2 Impeller specifications 30 3.3 Power measurement ' 32 3.4 Materials 33 3.5 Test procedure 34 3.5.1 Dynamic response of liquid and solid phase tracers 34 3.5.2 Mixing time 35 3.5.3 Dynamic testing procedure 37 3.6 Exp erimental conditions 38 4 MACROSCALE MIXING IN AGITATED PULP STOCK CHESTS 39 4.1 Introduction 39 4.2 Power number 39 4.3 Required power for complete motion 40 4.4 Evaluation of existing correlation and design methods 43 4.5 Dynamic response of liquid and solid phase tracers 48 4.6 Mixing time 51 4.7 Flow pattern 54 4.8 Summary 57 5 DYNAMIC MODELING OF AGITATED PULP STOCK CHESTS 59 5.1 Introduction 59 5.2 An example of step response for an industrial stock chest 61 - vi 5.3 Dynamic model 64 5.4 Estimation of dynamic model parameters from input-output data 65 5.4.1 First stage 68 5.4.2 Second stage 69 5.5 Summary 71 6 EXCITATION PROCEDURE AND INPUT SIGNAL DESIGN 73 6.1 Introduction 73 6.2 Excitation procedure .' 75 6.2.1 First experiment: system excitation by a rectangular pulse 76 6.2.2 Second experiment: system excitation by a frequency-modulated random binary signal 80 6.3 Input signal design for the scale-model chest 83 6.4 Model validation 88 6.5 Summary 93 7 RESULTS OF DYNAMIC TESTS MADE ON SCALE-MODEL CHEST 95 7.1 Introduction 95 7.2 Fully mixed volume calculation 96 7.3 Dynamic results and discussion 96 7.3.1 The effect of impeller speed and pulp flow rate 96 7.3.2 The effect of fibre mass concentration 100 7.3.3 The effect of pulp feed and exit location 103 7.3.4 Impeller momentum flux and dynamic response 107 7.3.5 Design criteria and dynamic response 112 7.4 Summary 117 8 RESULTS OF DYNAMIC TESTS MADE ON INDUSTRIAL STOCK CHESTS... 119 8.1 Introduction 119 8.2 Performance of industrial blend chest #1 119 8.3 Performance of industrial machine chest #2 122 8.4 Performance of industrial chest #3 124 8.5 Performance of industrial blend chest #4 127 8.6 Performance of industrial machine chest #5 130 8.7 Performance of latency removal chest #6 132 8.8 Summary 135 9 OVERALL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK 137 9.1 Overall conclusions 137 9.2 Recommendations for future work 140 NOMENCLATURE 141 BIBLIOGRAPHY 145 APPENDIX A: COMPUTER PROGRAM 153 APPENDIX B: DYNAMIC DATA ANALYSIS 164 APPENDIX C: DATA TABLES 168 viii L i s t o f T a b l e s Table 3.1: Maxflo impeller specifications 32 Table 3.2: Fibre Quality Analyzer results for F B K fibre 34 Table 4.1: The effect of fibre mass concentration on A and n 54 Table 6.1: Real parameters compared with estimated parameters from the excitation performed by a rectangular pulse 78 Table 6.2: Real parameters compared with estimated parameters from the excitation performed by a frequency-modulated random binary signal 82 Table 6.3: Estimated parameters from the first set of data. These parameters were used for simulation in model validation 89 Table 7.1: Experimental conditions 95 Table 7.2: Power consumption for two impellers having different diameters but providing same momentum flux I l l Table 8.1: Specifications of blend chest #1 121 Table 8.2: Specifications of machine chest #2 123 Table 8.3: Specifications of industrial chest #3 126 Table 8.4: Specifications of industrial blend chest #4 128 Table 8.5: Specifications of industrial machine chest #5 131 Table 8.6: Specifications of industrial latency removal chest #6 134 Table 8.7: Industrial results 136 ix Table C l : Dynamic data (Q = 37.1 L\/min , Cm = 3.3% , Config.l) 168 Table C.2: Dynamic data (Q = 31.2 L\/min , Cm = 3.3% , Config.l) 168 Table C.3: Dynamic data (Q = 28.6 L\/min , Cm = 3.3% , Config.l) 169 Table C.4: Dynamic data (Q = 21.1 L\/min , Cm = 3.3% , Config.l) 169 Table C.5: Dynamic data (Q = 7.9 L\/min , Cm = 3.3% , Config.l) 170 Table C.6: Dynamic data (Q = 37.1 L\/min , Cm = 3.3% , Config.2) 170 Table C.7: Dynamic data (Q = 7.9 L\/min , Cm = 3.3% , Config.2) 171 Table C.8: Dynamic data (Q - 37.1 L\/min , Cm = 2.7% , Config.l) 171 Table C.9: Dynamic data (Q = 37.1 L\/min , Cm = 2.7% , Config.2) 172 Table CIO: Dynamic data (Q = 37.1 L\/min , Cm = 2.1% , Config.l) 172 Table C . l l : Dynamic data (Q = 37.1 L\/min , C m = 2.1% , Config.2) 172 X L i s t o f F i g u r e s Figure 2.1: Torque vs. rotational speed response of pulp in dynamic shear test. Results are for a bleached pine kraft tested by Gullichsen and Harkonen( 1981) 7 Figure 2.2: Velocity profile for an axial-flow impeller (Nienow, 1997) 16 Figure 2.3: Plan view of a chest with L\/W> 1.5 equipped with two impellers 18 Figure 2.4: Schematic of side-entering impeller in a controlled zone agitation system 19 Figure 2.5: Flow pattern for a radial-flow impeller: (a) top-entry (b) side-entry 20 Figure 2.6: Flow pattern for an axial-flow impeller: (a) top-entry (b) side-entry. 21 Figure 2.7: A perfectly mixed chest 23 Figure 3.1: Schematic of experimental setup 29 Figure 3.2: Scale-model stock chest dimensions 30 Figure 3.3: Experimental setup 31 Figure 3.4: Maxflo impeller 31 Figure 3.5: Relationship between conductivity and salt concentration in aqueous phase 36 Figure 3.6: Relationship between brightness and blue fibre concentration in hand sheets 36 Figure 3.7: Method of determining mixing time 37 xi Figure 3.8: Input-output configurations studied 38 Figure 4.1: Power versus the cube of impeller speed for Maxflo impeller 41 Figure 4.2: Power required for complete stock motion versus the fibre mass concentration. Impeller diameter is given as a parameter 42 Figure 4.3: Power required for complete motion versus stock height to width ratio (Z\/W) 43 Figure 4.4: Power required for complete motion versus chest length to width ratio (L\/W) 44 Figure 4.5: Power number versus ND\/m for tests conducted in the scale-model chest 45 Figure 4.6: Comparison between power needed for complete motion in the scale-model chest and that calculated using Yackel's method 46 Figure 4.7: Comparison between impeller momentum number needed for complete motion in the scale-model chest and that calculated using Yackel's method 47 Figure 4.8: Dynamic response of liquid and solid tracers at Cm = 1.2% and N=549rpm 49 Figure 4.9: Dynamic response of liquid and solid tracers at Cm =2.4% and N=94S rpm 50 Figure 4.10: Dynamic response of liquid and solid tracers at Cm = 4.4% and N = 1531 rpm 50 Figure 4.11: Mixing time versus impeller speed (L\/W= 1.3 , Z\/W= 1.1) 52 Figure 4.12: Mixing time versus power (L\/W= 1.3 , Z\/W= 1.1) 53 xii Figure 4.13: Mixing time versus N2D4. Lines are given by the correlation, Equation 4.6 55 Figure 4.14: Flow pattern obtained in scale-model chest for TV = 1320 rpm, D = 16.5 cm and Cm - 3.3% views from (a) side, (b) top, (c) impeller wall and (d) opposite impeller wall 56 Figure 5.1: Typical control loop frequency response (Bialkowski, 1992). 60 Figure 5.2: Step response of the industrial stock chest shown in Figure 5.3 61 Figure 5.3: Industrial stock chest (Provided by EnTech Control Engineering Inc) 62 Figure 5.4: The frequency response measured for the industrial chest compared with its ideal response 63 Figure 5.5: Continuous-time dynamic model of the stock chest 65 Figure 5.6: A grid of 5 by 5 points linearly spaced by any power of 2. Square shows the point with the minimum cost function 69 Figure 5.7: A new grid with half of the previous spacing is re-centred at the point (square) with the minimum cost function 70 Figure 5.8: A line search in the direction of minimum cost function (black circles). Cross is the final point 71 Figure 6.1: System identification loop 74 Figure 6.2: Simulated chest input and output signals in black lines and model output in gray line 77 Figure 6.3: Dynamic model of agitated pulp stock chest in SEVIULINK '.79 Figure 6.4: Magnitude Bode plot of partial derivatives of the model 80 Figure 6.5: Procedure for designing a frequency-modulated random binary input signal 81 xiii Figure 6.6: Simulated chest input and output signals in black lines and model output in gray line 82 Figure 6.7: Scale-model chest input and output signals in black lines and model output in gray line 83 Figure 6.8: Frequency response of the approximate model 85 Figure 6.9: Magnitude Bode plot of partial derivatives of the model based on the excitation made on the scale-model chest by a rectangular pulse 85 Figure 6.10: Dynamic test made on the scale-model chest by exciting the system with a fast frequency-modulated random binary signal. Input signal is in black line and output signal is in gray line 86 Figure 6.11: Input signal designed for scale-model chest: (a) frequency -modulated random binary signal, (b) its spectrum and (c) its periodogram 87 Figure 6.12: Scale-model chest input and output signals in black lines and model output in gray line 88 Figure 6.13: Model validation procedure 89 Figure 6.14: Scale-model chest input and output signals (first set of data) in black lines. Model output in gray line 90 Figure 6.15: Validation data: scale-model chest input and output signals (second set of data) in solid black lines. Estimated model from the first set of data in gray line 91 xiv Figure 6.16: Auto-correlation function of the residuals (e) associated with the validation data and the model. Dotted lines denote 99% confidence intervals 92 Figure 6.17: Cross-correlation function between the residuals (e) and the input (u) for the validation data. Dotted lines denote 99% confidence intervals 92 Figure 7.1: The effect of impeller speed on the upset attenuation. Input signal is in black line and output signal is in gray line (Q = 37.1 L\/min , Cm = 3.3% , Config.l) 97 Figure 7.2: The effect of impeller speed and pulp flow rate on channeling 98 Figure 7.3: The effect of impeller speed and pulp flow rate on fully mixed volume 99 Figure 7.4: The effect of impeller speed on the degree of disturbance attenuation ; 100 Figure 7.5: The effect of pulp flow rate through the chest on the degree of disturbance attenuation 101 Figure 7.6: The effect of impeller speed and fibre mass concentration on channeling 102 Figure 7.7: The effect of impeller speed and fibre mass concentration on fully mixed volume 103 Figure 7.8: The effect of fibre mass concentration on the degree of upset attenuation 104 Figure 7.9: The effect of impeller speed and input-output locations on channeling 105 XV Figure 7.10: The effect of impeller speed and input-output locations on fully mixed volume 106 Figure 7.11: The effect of input-output locations on the degree of disturbance attenuation 107 Figure 7.12: The effect of impeller speed and pulp flow rate through the chest on channeling for Config.2 108 Figure 7.13: The effect of impeller speed and pulp flow rate through the chest on fully mixed volume for Config.2 109 Figure 7.14: The effect of impeller speed and fibre mass concentration on channeling for Config.2 110 Figure 7.15: The effect of impeller speed and fibre mass concentration on fully mixed volume for Config.2 110 Figure 7.16: Frequency response measured for two impellers having different diameters but providing same momentum flux (a case where no channeling exists) I l l Figure 7.17: Frequency response measured for two impellers having different diameters but providing same momentum flux (a case where channeling exists) 112 Figure 7.18: The effect of chest performance on power consumption (Q = 7.9 L\/min , Cm = 3.3% , Config.l) 113 Figure 7.19: The effect of chest performance on power consumption (Q = 7.9 L\/min , Cm = 2.7% , Config.l) 114 Figure 7.20: The effect of chest performance on power consumption (0 = 21.1 L\/min, Cm = 3.3%, Config.l) 115 xvi Figure 7.21: The effect of chest performance on power consumption (Q = 7.9 L\/min , Cm = 3.3% , Config.2) 116 Figure 8.1: Configuration of the blend chest #1 120 Figure 8.2: Step response of the industrial blend chest #1. The output signal has been shifted by -0.2 so as not to overlap with the input signal 121 Figure 8.3: Configuration of the machine chest #2 122 Figure 8.4: Step response of the industrial machine chest #2 (chest input and output signals in black lines and model output in gray line) 124 Figure 8.5: Configuration of the industrial chest #3 125 Figure 8.6: Step response of the industrial stock chest #3 (chest input and output signals in black lines and model output in gray line) 126 Figure 8.7: Configuration of the industrial blend chest #4 127 Figure 8.8: Step response of the industrial blend chest #4 129 Figure 8.9: Configuration of the industrial machine chest #5 130 Figure 8.10: Step response of the industrial machine chest #5 (model output is in gray line) 132 Figure 8.11: Configuration of the industrial latency removal chest #6 133 Figure 8.12: Step response of the industrial latency removal chest #6 (model output is in gray line) 135 Figure B . l :Scale-model chest input and output signals (Q = 7.9 L\/min , Cm = 3.3% , N= 1363 rpm , Config.l) 164 Figure B.2 :Scale-model chest input and output signals in black lines and model output in gray line (Q = 7.9L\/min , Cm = 3.3% , N = 1363 rpm, Config.l) 166 XVII Figure B.3: Frequency response of the scale-model chest (Q = 7.9 L\/min , Cm = 3.3% , N= 1363 rpm , Config.l) 167 X V 1 1 1 A c k n o w l e d g e m e n t s I would like to express my profound appreciation to my research supervisors Dr. Chad P.J. Bennington and Dr. Guy A. Dumont for their guidance, encouragement, and support provided throughout this project. I also gratefully acknowledge the advice and helpful suggestions of the members of my thesis committee, Dr. Richard M.R. Branion, Dr. Richard J. Kerekes, and Dr. Ezra Kwok. I am deeply indebted to Dr. Leonardo C. Kammer of P A P R I C A N for his innumerable assistance, especially in system identification. I am grateful to Mr. B i l l Bialkowski - EnTech Control Engineering Inc. - for providing industrial data. I also thank Chemineer Inc. for supplying the impellers. I acknowledge the assistance of all the staff in Pulp and Paper center and Chemical and Biological Engineering Department at UBC. I am particularly grateful to Peter Taylor and Tim Patterson for their valuable help in-the fabrication and installation of the experimental setup. Technical assistance from Ken Wong, computer support from Brian McMillan, and administrative assistance from Brenda Dutka, Lisa Brandly, Helsa Leong, and Lori Tanaka are highly appreciated. Sincere thanks go to Dave L. Pouw of PAPRICAN for his help in the instrumentation of the experimental setup. Financial support from the Mechanical Wood-Pulps Network and the Pulp and Paper Research Institute of Canada (PAPRICAN) is gratefully acknowledged. xix To Laleh, For her love, encouragement, and support. Chapter 1: Introduction 1 C H A P T E R 1 1 I N T R O D U C T I O N Agitated pulp stock chests in pulp and paper mills perform a number of functions. Their chief purpose is to reduce high-frequency disturbances in pulp properties (mixture composition, fibre mass concentration, freeness, etc.) ahead of many pulping and papermaking unit operations. In essence, those chests behave as low-pass filters. This complements the action of control loops which can only attenuate low-frequency variability below the loop cut-off frequency (Bialkowski, 1992). It is important to ensure these chests are properly designed to achieve the desired overall degree of upset attenuation. Dynamic tests made on industrial stock chests show that non-ideal flows such as channeling, recirculation, and stagnant zones exist in agitated pulp stock chests. These non-ideal flows reduce the degree of upset attenuation at frequencies higher than the cut-off frequencies of paper machine control loops (Ein-Mozaffari et a l , 2001). These disturbances are not fully attenuated, reducing paper quality and machine run-ability. Little information is available on the effects of non-ideal flow on the dynamic behavior of stock chests, although studies have been made on ideal chests (Walker and Cholette, 1958; Reynolds et a l , 1964; Brown, 1968). Since ignoring non-ideal flows can lead to errors in system design (Levenspiel, 1998), it is necessary to study the dynamic behavior of stock mixing under realistic mixing conditions. Current stock chest design is largely empirical and mostly proprietary. Therefore, the objectives of this study are: to enhance understanding of the dynamic behavior of stock chests, to improve upset attenuation by reducing the effects of non-ideal flows, and to incorporate process dynamics in the design criteria so that good mixing with a predictable dynamic response profile can be attained. Chapter 1: Introduction Chapter two reviews our current understanding of pulp suspension rheology, the mixing of Newtonian and non-Newtonian fluids, the existing design criteria of agitated pulp stock chests, and then dynamic behavior and the application of parametric and non-parametric methods in dynamic modeling. In chapter three the specifications of the scale-model chest designed for this study, the experimental setup and procedures are described. In chapter four macroscale mixing characteristics are examined in terms of required power for complete motion, mixing time, and the resulting flow pattern observed in the scale-model chest. Existing correlations and design methods of agitated pulp stock chests are evaluated and the dynamic response of liquid and solid phase tracers are compared. Therefore, this chapter concentrates on preliminary batch studies made on the scale-model chest to characterize its behavior and to develop test protocols for use in later dynamic tests. In chapter five the effect of non-ideal flow (channeling, recirculation and dead zone) on the performance of agitated pulp stock chests as low-pass filters is discussed and a dynamic model of an agitated pulp stock chest, incorporating these non-ideal flows, is developed. A numerical method for the estimation of dynamic model parameters from input-output data is presented. To estimate the parameters of the dynamic model, the system must be excited by a proper input signal. Chapter six concentrates on excitation design and input-output data collection. The choice of input signals has a very substantial influence on the parameter estimation. In fact all modes of the system should be excited during the identification experiment. Therefore, in this chapter a frequency-modulated random binary input signal is designed for this purpose. Chapter 1: Introduction 3 To improve the upset attenuation in the chest, the effect of non-ideal flows such as channeling and dead volumes must be reduced. In chapter seven the effect of impeller speed and diameter, fibre mass concentration, pulp flow rate through the chest, and pulp feed and exit locations on channeling, fully mixed volume, and disturbance attenuation is explored. In this chapter the effect of existing design methods using the criterion of complete surface motion (Yackel's method) or the onset of complete motion on dynamic response is investigated. In chapter eight the performances of six industrial stock chests are evaluated based on dynamic tests made on these chests. Chapter nine summarizes the overall conclusions of this study and gives recommendations for future work. Chapter 2: Literature Review C H A P T E R 2 2 L I T E R A T U R E R E V I E W 4 2.1 Introduction Mixing is important in all the chemical process industries, and the pulp and paper industry is no exception. Mixing and agitation are an integral part of pulp and paper manufacturing process. For stock blending, consistency control, bleaching, chemical generation, and deinking, effective mixing is vital to successful process results. In pulp and paper industry -as in all other industries- mixing may be defined as an operation to reduce non-uniformities in composition and\/or properties or to prevent the occurrence of the formation of non-uniformities. The agitated pulp stock chest is a subject of great importance as it impacts almost all facets of pulp and paper manufacture. Stock chests perform a number of functions. They act as a buffer between processes to permit continuous operation in the event of a breakdown, and reduce variability in fibre mass concentration, freeness and other quality factors. From the standpoint of variability reduction, stock chests provide a mean of reducing fast or high frequency variability. In essence those chests behave as a low-pass filters. This complements the action of control loops which can only control low-frequency variability below the loop cut-off frequency (Bialkowski, 1992). It is important to ensure that such chests are designed properly to achieve the desired degree of upset attenuation. Although much work has been done on fundamental understanding of mixing (Uhl and Gray, 1966; Holland and Chapman, 1966; Brodkey, 1975; Nagata, 1975; Nauman and Buffman, 1983; Oldshue, 1983; Ulbrecht and Patterson, 1985; Harnby et a l , 1985; McDonough, 1992; Zlokarnik, 2001), pulp suspension rheology (Gullichsen and Harkonen, Chapter 2: Literature Review 5 1981; Bennington, 1988; Bennington et al., 1990 and 1995) pulp suspension mixing (Oldshue and Gretton, 1956, 1958; Walker and Cholette, 1958; Attwood and Gibbon, 1963; Reynolds et al., 1964; Brown, 1968; Blasinski and Rzyski, 1972; Oldshue and Devries, 1985;Yackel, 1990; Bakker and Fasano, 1993; Bennington and Kerekes, 1996; Bennington, 1996; Wikstrom and Rasmuson, 1998), there still remains uncertainties in the interpretation of mixing in more complex hydrodynamic situations such as the one posed by pulp fibre suspension. This chapter reviews our current understanding of pulp suspension rheology, general equations for the mixing of Newtonian and non-Newtonian fluid, existing design criteria for agitated pulp stock chests, and then dynamic behavior and the application of parametric and non-parametric methods in the dynamic modeling. 2.2 Pulp suspension rheology The pulp suspension, which is a suspension of wood fibres in water, exhibits a very complex rheology. Pulp suspensions are continuous fibre networks that possess structure and strength resulting from interaction between neighboring fibres. In suspensions having fibre mass concentration greater than 0.5%, cohesive strength occurs from mechanical forces caused by bending and hooking of fibres (Bennington, 1996). As the fibre mass concentration increases, the number of fibre\/fibre interactions increases which in turn increases network strength. However, the distribution of fibres within the network is never uniform and local mass concentrations of fibres give rise to floes within the suspension. Since network strength depends on the number of fibre contacts, floes have a higher strength than the surrounding suspension. In a flowing pulp suspension, floes may behave as independent entities. In mixing it is important to create motion throughout the suspension. This requires the imposition of shear stresses greater than the strength of the pulp network. The initiation of Chapter 2: Literature Review 6 motion in pulp suspension first occurs at the weak zones of suspension, that is, between floes. As the applied shear is increased, the relative motion between floes increases and a reduction in floe size occurs (Kerekes, 1983). A further increase in shear causes more intense movement and leads eventually to turbulent flow. Figure 2.1 shows torque vs. rotational speed for a bleached pine kraft pulp tested in the concentric cylinder device by Gullichsen and Harkonen (1981). These curves show a number of interesting features: 1- A yield stress must be exceeded before motion can be initiated in the suspension. This yield stress increased as the fibre mass concentration increased. 2- Once motion had been initiated, the shear stress increased or remained approximately constant as rotational speed increased. 3- The pulp suspension flow curves exhibited a discontinuity as they neared the flow curve for water. The points on the curves in Figure 2.1 mark the instant at which the vessel contents were observed to come into a complete turbulence. 4- The torque vs. rotational speed curves for the pulp suspensions approximately followed the water curve after attaining this turbulent state. In this figure the curve for water does not pass through the origin. However, water is a fluid without yield stress. The creation of turbulence within a suspension is often referred to as fluidization. Fluidization implies relative motion among fibres, which leads to energy dissipation. Thus one potential method of quantifying fluidization is through power and energy expenditure. This concept was introduced by Wahren (1980) in his estimation of power dissipation per unit volume for the onset of fluidization (sj). He estimated values for the onset of fluidization from pulp suspension yield stress (ry) and obtained the following correlation: Chapter 2: Literature Review 1 REVOLUTIONS, rpm Figure 2.1: Torque vs. rotational speed response of pulp in dynamic shear test. Results are for a bleached pine kraft tested by Gullichsen and Harkonen (1981) e, =I^ = 1 .2x l0 4 C m 5 ' 3 (2-1) where Cm is fibre mass concentration given as a percent and p. is the viscosity of water. Gullichsen and Harkonen (1981) visually determined the onset of fluidization in a rotary device. They defined fluidization as the onset of turbulent motion throughout the vessel. This point was also characterized by a sharp increase in the torque on the rotor (see Figure 2.1). An expression for the s\/was obtained from their data by Bennington (1988) as: Chapter 2: Literature Review 8 ef =3 .4x l0 2 C, 3.4 (2.2) m It can be noted that, for any given fibre mass concentration, the dependence of power dissipation for fluidization on fibre mass concentration measured by Gullichsen and Harkonen is approximately two orders of magnitude less than that estimated by Wahren. This may be due to differing definitions of fluidization (Bennington et al., 1991). However, a more likely explanation lies in Wahren's use of the viscosity of water in Equation 2.1. If a pulp suspension behaves as a fluid, the suspension has an apparent viscosity, which is most certainly larger than that of water (Bennington and Kerekes, 1996). Therefore the use of the viscosity of water to estimate the power required to fluidize a pulp suspension is inappropriate. Bennington and Kerekes (1996) used the criterion chosen by Gullichsen and Harkonen (1981) for the onset of fluidization to develop an expression to estimate the energy dissipation necessary for fluidization: where DT is the diameter of the housing, D is the diameter of the rotor, and Cm is the fibre mass concentration, in percent. This equation can be extrapolated to a zero gap size (D^DT), to obtain: ef =A.5x\\tfCm2\\DTIDy23; 1.3