Improvements of pulp mill efficiency of disk based mill machines in papermaking industry

MINISTRY OF EDUCATION AND TRAINING THAI NGUYEN UNIVERSITY TRUONG THI THU HUONG IMPROVEMENTS OF PULP MILL EFFICIENCY OF DISK BASED MILL MACHINES IN PAPERMAKING INDUSTRY Specialty: Manufacturing Engineering Code : 62.52.01.03 INFORMATION OF DOCTORAL DISSERTATION Thai Nguyen - 2014 The research is carried out at Thai Nguyen University of Technology, Thai Nguyen University Scientific supervisors: Assoc Prof Nguyen Dang Hoe Prof Pham Vn Lang, Dr of Sci Examiner 1: Examiner 2: Examiner 3: The doctoral dissertation is defended at the Doctoral Committee of Thai Nguyen University in at …… hour, date … month … year ……… The dissertation is available at following libraries: PUBLICATIONS OF THE DISSERTATION Trương Thị Thu Hương, Nguyễn Đăng Hoè (2012) “Influences of tooth angle on specific loads of disk based pulp mill machines”, Tạp chí Cơng nghiệp nông thôn - ISSN 1859 - 4026, 5/2012 Trương Thị Thu Hương, Đỗ Thị Tám (2013) “Applications of modeling, similarity and dimensional analysis theorem in pulp mill process of disk based pulp mill machines”, Journal of Mechanics - Vietnam, 1/2013 Trương Thị Thu Hương (2013) “Influences of structural geometries of mill tools on pulp quality”, Journal of Science and Technology – Thai Nguyen University, 12 (Vol 112), 2013 Truong Thi Thu Huong, Do Thi Tam (2013) “Application of the modeling, similitude and dimensional analysis to study paper refiner models”, International Workshop on Agricultural Engineering and Post Harvest Technology for Asia Sustainability - AEPAS INTRODUCTION Necessity of the research Paper plays a very important role in almost all aspects for any country In 2013, Vietnam consumed approximately 655.000 ton of papers; however, domestic paper industry has just satisfied approximately 50% of need Among difficulties in paper production, satisfied quality of pulp is critical Pulp is produced after a fine mill process The fine mill process (lowconcentrated grinding) makes a change of wood fiber to obtain required properties of paper products This process critically influences on the paper quality of the final products Pulp quality mainly depends on several parameters, such as mill system, mill disks (grinders) and types of raw fibers Among of them, geometries of tools and their cutting teeth play main influences on the pulp quality as well as power consumption of the process Beside the grinding quality, the energy consumption during the mill process is the main factor on the market competition Grinders’ geometries and process parameters highly influence on this issue, and then directly impact on the competition of the paper products on the market In Viet Nam, raw materials using in the fine grinding process are usually mixed between short fibers (domestic made) and long fibers (imported) Besides, the grinders are completely ordered from abroad companies, in which design of disk geometries are computed in the case that they can be used for various types of pulp This common way may decrease the pulp quality, and increases time, energy consumption and production cost Till now, there has been not much research on studying geometries of structural grinders and the tool teeth with regards to pulp types produced from local sources This research provides an overall solution for multi-parameter optimization, compromising between the benefit of increasing pulp quality and decreasing energy consumption during the process with using domestic sources Objectives Based on the interrelation of major geometric parameters of the grinders, process parameters on the pulp quality and specific energy consumption during the fine grinding process with using commonly domestic raw materials in Viet Nam, a set of relevant parameters to increase pulp quality and decreasing energy consumption on the disk-based pulp mill tools will be proposed Fields and range of research 3.1 Research fields This study mainly focuses on the influences of the geometries of the mill tools and process parameters during the fine mill process with using common pulp materials in Viet Nam 3.2 Research range - The research just only concentrates on fine pulp mill process with using mixed pulp types commonly used in many paper company in Viet Nam; - Also, geometries of the mill tools and major process parameters on the pulp quality and energy consumption are studied Scientific contributions and range of applications 4.1 Scientific contributions - Successful implementation of the modeling and similarity theory to figure out experiments describing the process conditions so that cost for experiments can be much reduced - Problem solving of the multi-factor optimization consisting of objective function with various pro and contra parameters by using the design of experiment tools 4.2 Range of applications - The proposed grinder geometries and process parameters determined experimentally can be implemented in various paper companies in Viet Nam; - The dilemma of increasing pulp quality but less energy can be solved satisfactorily - The research outcomes can be flexibly applied into different manufacturing processes by using the theory of modeling and similarity Methodology - Evaluation, expert asking - Modeling and similarity models - Design of Experiments Thesis content Chapter Overview of pulp processes and machines Chapter Fundamentals of pulp mill process Chapter Models and experiments Chapter Results and discussions Chapter OVERVIEW OF PULP PROCESSES AND MACHINES 1.1 Introduction 1.2 Criteria for pulp quality 1.2.1 Length of fibers Length of fibers is critical for pulp strength The mill processes increase quality of pulp but decrease the length of fibers Thus, it is important to figure out relevant tools and process parameters so that fiber length can be satisfied together with maintenance of good pulp properties 1.2.2 Freeness of pulp Freeness (or dewatering or drainage) of pulp can be measured by Schopper Riegler (SR) Higher SR increases the pulp quality However, higher SR increases time and cost This dilemma can be solved by selecting relevant tools and process parameters to both increase pulp quality but less energy consumption 1.2.3 Strength of finished papers Strength (including tensile and shear strength) of finished papers is much closed to the length of fibers and freeness of pulp More material fibers are tear down when increasing time resulting in higher tensile strength Nevertheless, fibers are shorter and thus decrease shear strength This problem can be solved by selecting relevant tools and process parameters to both increase pulp quality but less energy consumption 1.3 Introduction to pulp mill processes 1.3.1 Pulp mill process chain Papermaking from wood-fiber raw materials is often carried out by two processes: pre-mill and fine mill The fine process works on rough bleaching pulp to refine the wood-fiber properties with regards to satisfied finished papers The fine process is highly critical on finished papers 1.3.2 Pulp mill tools Many tools are available currently, for example: beater roll, taper beater roll and grinding disks This research introduces different pulp mill tools and evaluates energy consumption and quality of pulp for each type of tool Among of above tools, disk mill is typically used in many paper companies It is highly necessary for Vietnam paper industry to increase efficiency and paper quality in disk mill based paper production 1.4 Mechanism in fine disk mill-based process 1.4.1 Fundamentals of fine disk mill-based process The most fundamentals of the fine disk mill-based process is that the mechanical forces impact on cellulose fibers in a mixing chemical liquid to break down the bulk structure of the fiber source, be it chip into the constituent fibers with regards to required quality of finished papers 1.4.2 Motion of mixing wood liquid The vortex flow of mixing wood liquid provides bonding conditions for wood and then to be milled in the milling area 1.4.3 Force components on disk mill teeth Forces on the disk teeth provide impacts, scratches, tensile and compress of the fibers and thus increasing bonding area among fibers These forces are the major factor to change the geometry, structure and properties of fibers 1.5 Fiber structure and pulp quality 1.5.1 Horizontal structure of wood fiber The horizontal structure of wood cell includes parts: heartwood (W), lumen (S1,2,3), primary wall (P) and middle lamella (M) Disk mill plays an improtant function to remove the primary wall (P) and the outer layer of lumen (S1) so that increasing bonding betwwen wood fibers, shorter cut and chipping inner and outer of the fibers, then changes shapes and propeties of wood structure to obtain required paper quality 1.5.2 Vertical structure of wood fiber Paper strength is relevant to the length of fibers, which is determined by different tree sources, soil types, environment, and age of resource tree and production of pulp To increase the strength of paper, it is necessary to enhance chipping of fibers and limit the decrease of fiber length To it end, it is required to propose process parameters with regards to raw materials thus enhance pulp quality and less energy consumption 1.6 Energy consumption Pulp mill process requires quite lot energy Averagely, it is required approximately 50 to 100kW for ton of pulp material Energy consumption is an important factor to evaluate efficiency of the pulp process The process is called more efficient if it requires less energy under the same working conditions of the same material and producing the same pulp quality 1.7 Influences of structural geometries of mill tools and process parameters on pulp quality and power consumption 1.7.1 Process velocity Increasing velocity provides benefits for fiber deformation, fibrillation, and decreasing fiber breaks, but increases energy 1.7.2 Gap between opposite disks Gap between two opposite cutting disks influences on load, specific energy and pulp quality 1.7.3 Pulp flows Pulp flows influence on energy penetrating on pulp movement and pulp properties 1.7.4 Pulp concentration Pulp concentration is an important parameter in defining the boundary velocity of mill disks in order to ensure the ability of pulp flow in the mill area and energy required The boundary velocity of disks in the refining mill process at low concentration requires 15 to 25 m/s 1.7.5 Influences of structural geometries of mill disks In the current market, there are various types of mill disks; however, there are not tools which can satisfy all required mill processes To determine the optimized mill disks requires experimental research on this point Therefore, this research on designing relevant tools which take the local raw material supply in Viet Nam into account is highly necessary Conclusions of chapter Pulp quality is determined by fiber length, freeness °SR of pulp and strength of finished paper During papermaking process, pulp mill is a preliminary chain which is critical influence on the final paper quality An efficient mill process depends on several factors, including raw material, tool geometries and process parameters It is highly required that the freeness of pulp is in the range of 35 to 38 °SR in the refining mill process to obtain high finished paper quality In most common papermaking companies, °SR usually attains from 34 to 36°SR, but they consume average energy from 50 to 100kWh for each ton of paper Under this condition, the quality of paper can not be satisfied, resulting in inconvenient conditions during printing Therefore, it is extremely necessary to carry out a research on designing mill tool geometries and relevant process parameters in order to obtain optimized tool specifications and les energy consumption, which are currently the big problems to the paper industry in Viet Nam Chapter FUNDAMENTALS OF PULP MILL PROCESS 2.1 Introduction 2.2 Kinematics of wood fibers in the mixing fluids 2.2.1 Flow properties of mixing fluids Pulp suspension is charged into the center of mill disk grinders Under the centrifugal force, pulp is pushed out radially The rotational speed of rotor disk changes the velocity of pulp suspension in the gap between the stator and rotor of dish grinders 2.2.2 Identity of fluids Pulp suspension is usually classified as non-Newtonian fluid However, it can be seen as Newtonian fluid at 4% suspension In that case, the Navier-Stockes can be applied for such that Newtonian fluid as following: v v v v   (r rz )   z  zz  vr z   z  vz z      (2.3) r r  z z   r r r  z  Also, energy equation can be written as: T     T   2T  2Tzz   T v T Cv  vr   vz   k   r   2     r r   z   r r  r  r  z    vr vz      v  vr   v   vz v   ,,,     v     r     rr r  r   r  zz z   r  r  r  r   rz  r   z    q            (2.4) It can be recognized that the application of the Navier-Stockes is not often used due to its complexity Alternatively, the application of modeling, similarity and dimensional theorem is highly recommended when studying on different fluid problems 2.3 Mechanical aspects of the pulp mill process 2.3.1 Relations of forces during the pulp mill process Directions and values of force vectors mainly influence on almost all aspects of mill process, including: cutting short, flatting, chipping, etc on the wood fibers 2.3.2 Specific load on the milling edge teeth Specific energy (SRE) can be computed as: P SRE  hd (2.9) qxC Specific load on the edge teeth (SEL) is: Phd P P SEL   hd  hd (2.10) nr ns l.n L.n Ls 2.3.3 Specific load on the milling teeth (SSL) Specific energy of pulp mill is calculated as: (2.12) SRE  IN  SSL  IL SEL SEL (2.13) SSL   cos rs IL ab So, SEL and SSL are main components of total energy consumption during the mill process, depending on tool geometries and process parameters Moreover, tool geometries such as tooth width, groove width, teeth height, teeth angle, wood fiber discharge, gap and mill speed play critical roles on mill area, probability of fiber bundle pushing into mill area, thus directly impact on quality of the mill process Therefore, structure of disk grinders and process parameters hold important functions of inputs on studying the quality and energy consumption during the pulp mill process 2.4 Influences of tool geometries and process parameters The tooth profile of the grinders mainly influences on cut shorter (pregrinding) or chipping (refined grinding) Positions of tooth on the mill disks highly impact on pulp movement, consuming time of treated pulp in the working area, thus determine the pulp quality and energy consumption Refining mill process is carried out at low pulp concentration, narrow gap, low air and temperature so that the disks not require a rough mill part Besides, in order to increase working length on the mill tool, simple manufacturing and benefits for pulp movement, it highly suggests a mill tool design with straight and parallel teeth Tooth width is crucial on the number of teeth and the overlapping cut between two opposite teeth Narrow tooth width can be usually found in short fiber mill or cut shorter processes On the contrary, long fiber mill or chipping processes require wider tooth width Gap width and gap height also determine the pulp flow into the working area The less width and depth of the tooth grooves increase the pulp mill process but decrease pulp flow into the working area and vice versa Teeth angle () influences on the number of teeth and tooth length as well as the pulp flow process between mill disks Increasing angle contributes higher milled length, thus enhance the mill process and mill fibers Considering two opposite mill teeth in working condition, velocity used for fiber cut and chipping during mill process as following: 10 Applying dimensional analysis, similarity numbers can be defined as: Q p  1    ;    E ;    ; Q u v.D  v  v.D. Re L h a (3.2)    ; 5  ; 6  s ; 7  ; 8  ; v D D b c g D 9  ; 10  ; 11   D D v Fr where Π2 is Eulerian (Eu) number; Π3 is Reynolds (Re) number and Π11 is Froude number According to Pi theorem Π, the Π can be written as:  p Ls gD h a b c  Q  1    , ,  ,  , , , , , ,  (3.3)  vD  v  vD  v v D D D D   During experiments, similarity numbers of the mill machine model are carried out at the same conditions (D, a, b, c, μ, ρ, g, p are constant) p However, n, α, φ, and h are varied and characterized for: Π 3= = ; Π v  Re h = α, Π = φ and Π = and considered the ’’input parameter’’ D experiments To sum up, based on 14 input parameters, the application of the theorem of modeling, similarity and dimensional analysis reduces the number of input parameters for study up to only This benefit provides fewer experiments but remain the same input parameters as designed which ensures the quality evaluation of all influences on pulp freeness and energy consumption of the machine tool 3.3 Design of experimental models 3.3.1 Mill system 3.3.1.1 Mill disks a Disk structure This research proposes the straight parallel teeth with following geometries as shown in Table 3.3 11 Table 3.3 Geometries of mill disks for experiments Experi mental disks Tooth angle α(0) BĐ1 BĐ2 BĐ3 12 18 24 Diameter (d) Mill Groov Tooth Tooth insert e Inner Outer width height angle diameter diameter width a(mm) c(mm) θ(0) b(mm) (mm) (mm) 22.5 80 240 22.5 80 240 22.5 80 240 Figure 3.3 illustrates geometries of one disk sample Fig 3.3 Geometries of one mill disk sample b Material Martensite stainless steel 2X13 is chosen for mill disks during experiments c Manufacturing chain of mill disks 3.3.1.2 Experimental mill machine Specifications of mill machine used in experiments are shown in Table 3.4 Table 3.4 Specifications of the experimental mill machine No Specifications Outer disk diameter Inner disk diameter Productivity Power Values 240mm 80mm Qmax = 466 (kg/h) and Qmin = 216 (kg/h) 7kW 12 Speed range Pulley diameters Transmission Shaft diameter Bearings 1600, 1200, 1000, 800, 700 (v/ph) 290 - 216-180-145-126 (mm) V-belt type, diameter D = 260 (mm), speed v =23.9 (m/s) and speed ratio z = 1.94 d = 34.14 (mm) Part inherent with V-belt pulley d1 = 35 mm Part inherent with bearing d2 = 40 mm Part from bearing to mill disk d3 = 50 mm + Inner diameter d = 40 mm; + Outer diameter D = 90 mm + Width B = 23 + Dynamic load capacity C = 61,0 (kN) + Static load capacity C0 = 46,0 (kN) - Back bearing: + Inner diameter d = 40 mm + Outer diameter D = 90 mm + Width B = 23 + Dynamic load capacity C = 31,9 (kN) + Static load capacity C0 = 21,7 (kN) Figure 3.8 illustrates the pulp mill machine carried out in this research Fig 3.8 Pulp disk mill machine illustration 3.3.2 Raw pulp material Table 3.5 presents two types of raw material for pulp mill used in this research Table 3.5 Properties of pulp materials No Pulp type BKHP BKSP Fibrillation length, mm 1,05 0,64 Content of less than 0,2 mm fibrillation in length 17,7 13,4 Specifications 13 Tensile load capacity, Nm/g Shear load capacity, mN.m2/g Crack resistance, kPa.m2/g Pulp freeness after stirring, SR 60,7 6,8 3,6 13 80,5 9,7 5,5 12 3.3.3 Measurements of outputs 3.3.3.1 Consumed power N Measurement process of power consumption was carried out at the University of Transport and Communications Figure 3.11 presents the measuring process EXPERIMENTAL LAYOUT PHASE POWER SUPPLY ELECTROMETER PHASE MOTOR COMPUTER DISK BASED MILL MACHINE Fig 3.11 Experimental setup of power consumption measurement 3.3.3.2 Pulp quality evaluation a Measuring tools Figure 3.12 – 3.16 illustrates evaluations of pulp quality with variant of specifications Fig 3.12 Freeness measurement Fig 3.13 Mill machine test PFI Fig 3.14 Rapid - Kothen 14 Fig 3.15 Tensil measurement (Hounfield) Fig 3.16 Paper shear strength test (Frank) b Measurement of pulp freeness 3.4 System operation 3.5 Design of experiments 3.5.1 Experimental parameters Design of experiment is presented as shown in Figure 3.17: x1 = 4 = α x2= 7 = h x3 = 3 = n x4 = 5 = q  p L gD h a b c  Q  1     , , , , s , , , ,  v v D D D D vD   v  vD yN: Freeness (0SR) yK: Specific power consumption (ws/kg) Fig 3.17 Influences and outputs of the experiments 3.5.2 Matrix experiments and regression (Box – Behnken with n=4) 3.6 Principles of data processing h thực nghiệm 3.6.1 Regressive model selection 3.6.2 Evaluation of model relevance 3.6.3 Solution for multi-variable optimization Conclusions of chapter The experimental model is built based on the observation of commonly current industrial machines and previous experiments By applying the theorem of modeling, similarity and dimensional analysis, author has concluded major influences as input parameters during the experiments, including: mill speed (n), gap between two opposite disks (h), tooth angle (α) and pulp flow (q) A successful sample mill machine has been built which flexibly allows changes of input parameters and satisfies almost all requirements during experiments Outputs has also been collected and tested with relevant and reliable instruments 15 Design of experiments has been based on the Box – Behnken with n=4 method Accordingly, 27 experiments have been carried out, allowing the fewest work but ensuring reliable results (Table 3.7) Principles of data processing after experiments have been carried out based on probability science which guarantees a relevant regressive model reflecting good relations between objective function and experimental variables and the relevance of the experimental model Chapter RESULTS AND DISCUSSIONS 4.1 Introduction 4.2 Experimental results Table 4.2 presents results after experiments Table 4.2 Table of experimental results Code u Ex1 Ex2 Ex3 Ex4 Ex5 Ex6 Ex7 Ex8 Ex9 Ex10 Ex11 Ex12 Ex13 Ex14 Ex15 Ex16 Ex17 Ex18 Ex19 Ex20 x1() x2(h) x3(n) x4(q) -1 +1 -1 +1 -1 +1 -1 +1 0 0 -1 +1 -1 +1 0 0 -1 -1 +1 +1 0 0 -1 +1 -1 +1 0 0 -1 +1 -1 +1 0 0 -1 -1 +1 +1 -1 -1 +1 +1 0 0 0 0 0 0 0 0 0 0 -1 -1 +1 +1 -1 -1 +1 +1 x1 (0 ) 12 24 12 24 12 24 12 24 18 18 18 18 12 24 12 24 18 18 18 18 Value YN YK x2 x3 x4 (mm) (rpm) (l/min) 0.1 1200 12 15413.2 34.7 0.1 1200 12 15106.4 36.2 0.1 1200 12 15574.2 27.5 0.5 1200 12 15094.4 30.2 0.3 800 12 14679.4 30.7 0.3 800 12 15155.2 31.2 0.3 1600 12 14796.8 32.5 0.3 1600 12 15821.4 30.6 0.1 800 12 15042 37.3 0.5 800 12 14673.2 29.8 0.1 1600 12 15522.2 38.2 0.5 1600 12 15082.2 31.5 0.3 1200 15083.8 33.1 0.3 1200 15139.4 31.8 0.3 1200 20 15865.8 30.8 0.3 1200 20 15965.6 33.6 0.1 1200 15340.8 36.6 0.5 1200 14890.6 31.5 0.1 1200 20 15942.4 37.1 0.5 1200 20 15661.8 32.5 16 Ex21 Ex22 Ex23 Ex24 Ex25 Ex26 Ex27 0 0 0 0 0 0 0 -1 +1 -1 +1 0 -1 -1 +1 +1 0 18 18 18 18 18 18 18 0.3 0.3 0.2 0.3 0.3 0.3 0.3 800 1600 800 1600 1200 1200 1200 4 20 20 12 12 12 15049.2 15207 15455.2 15604.8 15088.2 15090 15086.4 4.2 Regressive model from objective function Table 4.3 Analysis of experiments Box-Behnken Design Factors: Replicates: Base runs: 27 Total runs: 81 Base blocks: Total blocks: Center points: Response Surface Regression: YN versus x1, x2, x3, x4 The analysis was done using coded units Estimated Regression Coefficients for YN Term Coef SE Coef T P Constant 150146.10 15.117 169.304 0.000 x1 -93.33 7.558 -12.348 0.000 x2 20.77 7.558 2.748 0.008 x3 182.99 7.558 24.210 0.000 x4 100.06 7.558 13.239 0.000 x1*x1 31.29 11.337 2.760 0.007 x2*x2 99.11 11.337 8.742 0.000 x3*x3 89.47 11.337 7.891 0.000 x4*x4 -48.38 11.337 -4.268 0.000 x1*x3 22.01 13.091 1.681 0.097 x3*x4 72.48 13.091 5.537 0.000 S = 45.3497 PRESS = 198899 R-Sq = 94.24% R-Sq(pred) = 92.04% R-Sq(adj) = 93.41% Analysis of Variance for YN Source DF Seq SS Adj SS Adj MS F Regression 10 2353781 2353781 235378 114.45 Linear 1895015 1895015 473754 230.36 Square 389908 389908 97477 47.40 Interaction 68858 68858 34429 16.74 Residual Error 70 143962 143962 2057 Lack-of-Fit 14 122375 122375 8741 22.68 Pure Error 56 21587 21587 385 Total 80 2497743 P 0.000 0.000 0.000 0.000 0.080 29.3 31.8 35.4 38.3 36.1 35.4 36.3 17 Values larger than α should be ignored For example, the product x1  x in the regressive function should be neglected because the correspondent value P is equal to 0.097 larger than 0.05 Referring to values in the Coef column, combining with referred values in the P column, the regressive function YN with regards to variables x1, x2, x3, x4 as shown: YN =150146.10  93.33x1  20.77x  182.99x  100.06x  (4.1)  31.29x1  x1  99.11x  x  89.47x  x   48.38x  x  72.48x  x Similarly, regressive function for pulp quality YK with regards to x1, x2, x3, x4 can be found: YK =38.4352  2.486x1  1.081x  2.175x  1.286x  (4.2)  3.810x1  x1  4.910x  x  3.085x  x   4.235x  x  1.158x1  x  2.058x1  x  1.242x  x 4.3 Multi-variable optimization process 4.3.1 Optimizing specific energy consumption YN The following shows the optimizing result with applying Minitab software: Response Optimization Parameters Goal lower Target Upper Weight Import YN Minimum 144502 152137 154800 1 Global Solution x1 = x2 = - 0.111111 x3 = - 0.737374 x4 = - Predicted Responses YN = 7498.65, desirability = 0.876683 Composite Desirability = 0.876683 The result show the predicted response of the objective function YNmin  144502 , with desirability d = 0.87668 Accordingly, Figure 4.1 shows the optimizing plot of the function YN 18 Fig 4.1 Optimizing plot YN 4.3.2 Optimizing pulp quality YK Similarly as YN, result shows the optimizing value of objective function YK is equal to YKmax = 38.218 with desirability d = 0.904761 Fig 4.2 Optimizing plot YK 4.3.3 Optimizing YN and YK Using Minitab, following result shows the response optimization between YN and YK: Response Optimization Parameters Goal Lower Target Upper Weight Import YK Maximum 36 37 38 1 YN Minimum 144502 152137 154800 1 Global Solution x1 = - 0.431220 x2 = - 0.0303030 x3 = - 0.162994 x4 = - 0.756556 Predicted Responses 19 YK = 37.1052, desirability = 0.999942 YN = 76320, desirability = 0.955651 Composite Desirability = 0.977546 The response optimizing value for YK is 37.1052, desirability d = 0.999942 Also, the response optimizing value for YN is 152640, desirability d = 0.955651 Thus, the composite desirability D = 0.977546 All those values are close to 1, meaning that the very good result of each objective with regards to composite desirability Figure 4.3 illustrates the optimizing plot of the multi-variable optimization process: Fig 4.3 Optimizing plot of the multi-variable optimization Based on the above plot, optimized parameters of mill disk and machine can be obtained: Table 4.4 Optimized parameters of mill disk and machine x1() 16.27 Values x3(n) x2() 0.25 950.16 x4(q) 6.4 YK ( SR) YN (ws/kg) 37.1052 152640 Therefore, relevant design parameters for the experimental machine layout can be as follows: Table 4.5 Design parameters for the experimental disk and machine Specifications Teeth angle Gap Spindle speed Pulp flow Value 16.27 0.25 mm 950 rpm 6.4 l/min 20 37.1052 0SR 42.4 Kwh/ton Quality Energy consumption Conclusions It can be seen that with angle of 16.270 and gap 2.5mm, spindle speed 950 rpm and pulp flow 6.4 l/min, the pulp freeness can be obtained as 37.1 0SR and energy consumption 42.4 kWh/ton Thus, comparing with the current process in industry which has been shown in Chapter 1, this design machine allows better pulp quality with freeness higher than the current machines from 1-30SR while energy can be saved approximately 7.6 kWh/ton Comparisons of pulp quality on standard and experimental machines * Influences of mill on pulp freeness and fibrillation length Ảnh hưởngofcủa độ time nghiềnonđến chiều dài xơlength sợi Influence mill fibrillation Ảnh hưởngofcủa gian đến độ nghiền Influence millthời time onnghiền pulp freeness 45 40 40 0.8 42 0.7 35 30 24 25 20 15 16 Freeness Độ Nghiền SR SR 18 13 Chiều dài xơ sợi 36 0.6 0.5 0.4 0.1 0 0.7 1.4 2.1 2.8 3.5 4.2 Thời gian, Time, phút Test sample Mẫu nghiềnPFI PFI 0.2 10 Experiment Mẫu TN sample 0.3 13 20 30 35 40 45 50 Freeness, Độ nghiền,SR SR Fig 4.4 Mill time and pulp freeness (a) and fibrillation length (b) Pulp freeness increases with increase of mill time It slowly increases at the beginning and end of the process On the other hand, pulp freeness increases and fibrillation length decreases with the mill time * Influences of mill process on the fiber properties 100 50 Mẫu TN sample Experiment Test sample Mẫu nghiềnPFI PFI 13 20 30 35 40 45 50 Freeness, SR Freeness influences on shear strength Shear strength, mN.m2/g Tensile strength, Nm/g Freeness influences on tensile strength 10 Mẫu TN sample Experiment Mẫu nghiềnPFI PFI Test sample 13 20 30 35 40 45 50 Freeness, SR Fig 4.6 Influences of freeness on tensile strength (a) and shear strength (b) Freeness of pulp increases from 13 to 40 0SR It can be seen that the freeness of both standard and experimental machines are quite equivalent This result illustrates the desirability of the experimental machine Figure 4.7 presents the structural changes of pulp before and after mill process: 21 a) b) Fig 4.7 Pulp before (a) and after (b) process Due to the refining process, the fiber structure becomes smoother, better horizontal and vertical crossing between fibers, thus more durable finished paper They are all required properties of pulp for finished product quality 4.4 Implementation on built-in machines 4.4.1 Determination of similarity terms with regards to mill power The mill power can be computed according to [44]: N  f (D, v,r,  ,g) Then, following terms can be determined: N N p 1     Eu ; v D  vD v  v   g.D 2    Re1 ;     Fr1  D.v Re v Fr (4.6) Required power for the mill process then can be followed by equation: m  vD   N (4.10)  f  v3 D5     This turns out Eulerian number (Eu) for pulp mill process on the mill disk based machines 4.4.2 Determination of similarity terms with regards to productivity The general form to define productivity is followed [44]: Q  f  D, h, n,  ,   (4.11) Then, following terms can be determined: Q h   1 1  ;   ;     R (4.12) e  n.D3 D n.D  v.D. where Re is Reynolds number 22 c nc Dc2  nD From Re  another similarity number can be found Lc1  1 c  Under the same condition, viscosity and density of pulp can be assumed constant with 𝜇𝑐 = and 𝜌𝑐 = Then, relation between D and n is defined as: D0 n (4.18)  M  DM n0 Q h a From   , 1  , which have been defined in Chapter    n.D3 D D 3, other relations, including gap and disk diameter; productivity, speed and diameter; teeth width and disk diameter can be defined: a0 D0 h0 D0 Q0 n0  D0   and   ;      aM DM hM DM QM nM  DM  Finally, from productivity term Eu  N n3 D  and 1  Q , by applying  nD similarity theorem, the relation between N and Q is defined as: N Q0 2    3 N M QM (4.26) From the optimizing results shown in section 4.2, by applying the modeling, similarity and dimensional analysis theorem, the built-in machines can be designed with specifications shown in Table 4.8 Table 4.8 Built-in mill machines No Specifications Required power, kW Productivity Q (t/day) Productivity Q (t/h) Specific energy cost, kWh/ton Spindle speed n, rpm Tooth angle, Disk diameter, mm Disk gap, mm Tooth width, mm Machine MN-1 8.86 0.209 42.4 950 16.27 240 0.2 MN-2 15 0.37 39.6 800 16.27 260 0.3 3.3 MN-3 45 30 1.24 36.2 650 16.27 480 0.35 2.8 23 Conclusions It can be seen that required power for the process is a function of 3rd power with spindle speed and 5th power with diameter; therefore, they are important parameters when set up the mill machine system Based on the optimizing design of the experimental machine, by applying the modeling, similarity and dimensional analysis theorem, it allows to determine a range of machines from to 50 ton/day with required power from to 45 kW The ratio between the disk diameter and width of teeth of the built-in machine and experimental machine is linear with λ, power ratio between the built-in machine and experimental machine is 3th power with λ and productivity ratio is 5th power with λ   nM n0 , where nM is the   spindle speed of experimental machine and n0 is of the built-in machine With the same tooth angle, the increase of the disk diameter and gap with the above ratio allow improve productivity and decrease energy requirement Conclusions of chapter The study has shown results of 27 experiments carried out based on Box – Benkn method and processed with Minitab software The regressive model illustrates equivalent data (shown on equation 4.1 and 4.2).The regressive model helps to solve for the multi-variable optimization (see Figure 4.3) The optimizing solution has completely solved the dilemma between energy consumption and pulp quality The desirability of the optimizing response is approximately 0.977, which is close to This means the confliction between objective functions has been solved reasonably (Figure 4.3) The results from this research help to determine required parameters for the built-in machine range, including machines (Table 4.8) by applying the modeling, similarity and dimensional analysis theorem This opens the potential application of the results from this research 24 CONCLUSIONS AND OUTLOOKS Conclusions Based on the analysis of influences on the pulp mill process on the disk based machine, similarity terms and equivalent equations describing completely physical behaviors during the pulp mill process (set of equation 3.2) have been extracted Also, the study has figured out major input terms (n,h,α,q) from 14 input parameters to carry out the experiments The multi-variable optimization for the pulp mill process on the disk based machine has also solved completely Data from this research have evaluated and measured by reliable methods and tools These results allow to determine specifications for the design process of built-in machine in industry by applying the modeling, similarity theorem (Table 4.8) Outlooks Further research would carry out influences of inclined tooth angle, the intersection angle between cutting teeth This guarantees a full understanding of better influences of tool geometries on the kinematic mill processes It is necessary to continue studying on pulp mill processes for typically local raw material in Viet Nam ... quality in disk mill based paper production 1.4 Mechanism in fine disk mill- based process 1.4.1 Fundamentals of fine disk mill- based process The most fundamentals of the fine disk mill- based process... (pregrinding) or chipping (refined grinding) Positions of tooth on the mill disks highly impact on pulp movement, consuming time of treated pulp in the working area, thus determine the pulp quality... of the pulp mill process 2.3.1 Relations of forces during the pulp mill process Directions and values of force vectors mainly influence on almost all aspects of mill process, including: cutting