6 Microscopic study of sheet densification by wet pressing

Transcription

1 6 Microscopic study o sheet densiication by wet pressing 6.1 Introduction Fibre bonded area and sheet density are strongly aected by wet pressing. A better understanding o the echaniss o the changes o paper structure in wet pressing is o critical iportance or the understanding o the ibre bonding and sheet densiication. This inoration is also iportant or network odelling. In a sense wet pressing will bring ibres closer together and thus proote bonding. Wet pressing will also copress the ibres in the network and thus increase the degree o ibre collapse. However, how uch these changes contribute to the sheet densiication is still not clear. Paulapuro (Paulapuro 2001) recently established that the observed eects o wet pressing on paper properties are related to changes o sheet density, the z-direction density distribution created by the pressing and surace evenness (topography). Szikla and Paulapuro (Szikla 1986; Szikla 1989) ound that increasing ibre bonding is the doinant echanis in the densiying eect o wet pressing (Szikla 1989). Gorres et al (Gorres 1993) believed that ibre collapse and delection o ree ibre segents are the predoinant echaniss at low pressing pressure, while at high pressure all spaces within the sheet which can be illed by such delections have been illed, and thereore copletion o ibre collapse and perhaps other eects then predoinate. Fro these studies, it appears that the echaniss in the densiication eect o wet pressing on paper structure are still not copletely understood. Previous studies o ibre cross-sections easured the transverse diensions o individual pulp ibres deposited on glass slides or ebedded individually in resin (Kibblewhite 1991; Gorres 1993; Jang 1998). Gorres et al pressed very thin ibre networks (less than 0.1g/ 2 ) on glass slides and used a stylus proiloeter to evaluate the transverse diensions o ibres in the networks (Gorres 1993). However, no quantitative studies have been reported on the behaviour o ibre cross-sections in true paper in wet pressing, priarily because beore the work o this thesis there was no technique to easure ibre cross-sectional diensions directly in a sheet. 91

2 The collapse degree o ibres has been characterized in dierent ways, such as by the ibre aspect ratio (Kibblewhite 1991) and the ibre collapse index, which is deined either as the reduction o ibre thickness (Gorres 1993) or the reduction o ibre luen area (Jang 1995), copared to an un-collapsed ibre. To calculate the ibre collapse index, the original ibre shape ust be assued to be either circular or rectangular in shape (Gorres 1993; Jang 1995). The collapse index cannot be easured directly and it only describes the shape o the ibre luen area, copletely neglecting the shape o ibre wall area. One actor that has not been considered in studies o pressing is the twisting o the ibres so that their cross-sectional ajor axis rotates to be closer to parallel to the paper plane. The orientation o the ibre with respect to the paper plane could strongly aect the structure o the paper. Decreasing the nuber o twists in ibres will increase sheet strength (Mohlin, Dahlbo et al. 1996). In this chapter, a ibre shape actor and a twist angle o the ibre cross-section are deined. These paraeters are easured directly in handsheets using the cross-section preparation and analysis technique describe in Chapter 4. The results are used to quantiy the changes o ibres in the handsheets in wet pressing. The echaniss behind densiication eects o wet pressing on paper structure are discussed. 6.2 Theory Degree o ibre collapse in paper For the purpose o quantiying the degree o collapse o ibres in sheets we deined the new ter o shape actor,, as the ratio o the ibre wall area, sallest rectangular bounding box, A, to the area o the A s, that can copletely enclose the irregular shape o the ibre. (Reer to subsection 4.4.5). In this section, we will exaine the relationship between the shape actor and the degree o collapse o the ibre. To do this we begin by assuing that the ibre cross section beore wet pressing is circular with diaeter, D, and wall thickness, t, and that the ibre wall area, A, is constant with 92

3 wet pressing. We assue that i the thickness decreases by a distanceδ D, then the width will increase by the sae aount (see Figure 6-1). Accordingly, is: A A πt( D t) = = 6. 1 A s = 2 2 ( D 2δ D )( D + 2δ D ) D 4δ D Equation 6.1 can also be written as: 2 π ( t / P)(1 t / P) = ( δ / r) D where P is the ibre perieter, and r is the ibre radius. Following Jang (Jang 2001), the collapse index, CI, is deined as CI thin walled ibre, t << P, and ( 1 t / P) 1, and Equation 6.2 becoes: 2 = δd / ). For a ( r 2 π ( t / P) = 1 CI 6. 3 Obviously, uncollapsed ibres and ully collapsed ibres will have the iniu and the axiu values o respectively. These axiu and iniu values o can be used to create an approxiate scale to separate ibres in a sheet into three collapsestates, viz. uncollapsed ibres, partially collapsed ibres and ully collapsed ibres (Figure 6-2). The validity o using the boundary values o in describing the collapse-state o a ibre will be veriied in this study. Figure 6-1 Model ibre cross-section 93

4 A = B = C = D = E = F = Figure 6-2 Dierent collapse states o ibre: uncollapsed ibres (A, B), partially collapsed ibres(c, D), and ully collapsed ibres (E, F) Reduction in paper thickness by ibre twist and ibre collapse Assuing the paper has layered structure, ibres in each layer are twisted and collapsed in wet pressing, reducing the thickness o each layer. The total reduction in the paper thickness is the su o the thickness reduction o each layer. Figure 6-3 shows scheatically the i th ibre in a layer o a sheet beore and ater it is pressed. Beore the ibre is pressed, the ibre has width, D wi, thickness, D hi, and twist angle, β i. Ater it is pressed, the width, thickness and twist angle are D wi, D hi and β i respectively. I the paper has independent layers and each layer has n ibres, the total reductions in paper thickness ro the ibre twist, T twist, and ro ibre collapse, collapse T, are given by Equations 6.4 and 6.5 respectively: T twist = n n [ D wi sin i Dwi sin β i ] i= 1 β 6. 4 T collapse = n i= D hi Dhi 1 cos β i cos βi n

5 Z Paper thickness D wi D hi Fibre ajor axes ' D hi β i ' β i ' D wi Paper plane Figure 6-3 Scheatic showing the reduction in paper thickness contributed by ibre twist and ibre collapse in any layer o a sheet 6.3 Saple used in this study The low kappa pulp was used in this study (reer to Appendix A). This is the pulp labelled L 0 in subsection Five sets o 60g/ 2 square handsheets were ade on the Moving Belt Sheet Forer. Each set o handsheets was pressed statically at one o 5 pressures, viz 100kPa, 200kPa, 500kPa, 2000kPa and 4000kPa (or details reer to subsection 5.2.1). Cross-sections o handsheets pressed at 100kPa, 500kPa and 4000kPa, which are denoted as L 0 P 1 L 0 P 3, and L 0 P 5 respectively, were exained in a conocal icroscope. The ethod or saple preparation or the conocal icroscope has been given in subsection The ethods or easuring the shape actor,, and the twist angle, β, have been given in Subsection Results and discussion Changes in paper cross-section Figure 6-4 shows typical cross-sectional iages o the three saples L 0 P 1, L 0 P 3 and L 0 P 5. The saple thickness and the void space are reduced with increasing pressing pressure. In saple L 0 P 1, it can be seen that ost ibres are either uncollapsed or partially collapsed, but there are also soe ully collapsed ibres. In saple L 0 P 3, the nuber o uncollapsed ibres is draatically reduced, and around hal the ibres are ully collapsed. In saple L 0 P 5, ost o the ibres have been ully collapsed, but a ew uncollapsed ibres still can be seen. 95

6 Figure 6-4 also shows that ibres in the sheet have been rearranged in wet pressing. Soe o the ibres are pressed into gaps or void space. The type o oveent is called gap closure in this project. In suary three types o oveent o ibres in wet pressing have been observed in this study including collapse o ibre, twist o the ibre and gap closure. These three types o oveents, which control the change in paper thickness, have been quantiied in this study and the results are discussed in the ollowing paragraphs. A B C * The size bar is 10µ Figure 6-4 Iages o cross-sections o saple L 0 P 1 (A), L 0 P 3 (B) and L 0 P 5 (C)Table 6-1 suarizes the ibre diensions and the sheet properties easured in this work. The ibre wall area was not changed by wet pressing and the ean value o increases with increasing wet pressing pressure. The twist angle alls ro 9 o to 6.8 o with increasing wet pressing. 96

7 Table 6-1 Handsheet properties and ean ibre properties Wet press Average Shape actor, * Fibre Out-oplane angle o ibres density thickness index Nuber Sheet Paper Tensile Saple pressure twist wall area (%) (kpa) angle (µ 2 ) (degree) easured (kg/ 3 ) (µ) (kn/kg) L 0P ±0.014 ** 9.0± ±9 ** 5.13± / 200 / / / / / L 0P ± ± ±9 4.49± / 2000 / / / / / L 0P ± ± ±9 4.64± *Fibre wall area and ibre axes have been corrected by its angle to the surace o the paper cross-section. ** ± is 95% conidence interval. Figure 6-5 shows the twist angle distribution o ibres in the three saples L 0 P 1, L 0 P 3 and L 0 P 5. As the pressing pressure is increased, the nuber o ibres with twist angle o o less than 10 increases, while the nuber o ibres with twist angle ore than 10 decreases. It can be seen that 67% o the ibres in saple L 0 P 1 have a twist angle less o than 10 and this percentage is increased by wet pressing to 71% or saple P 3 and to 78% or saple L 0 P 5. This reduction in ibre twist will reduce the aount o space taken by the ibre in the paper structure, thereore reducing the void space and increasing the density o the paper. The reduction in ibre twist could also increase the potential bonding surace area o the ibres, especially or collapsed ibres. 50 Percentage ( %) P1 P3 P Twist angle (degree) Figure 6-5 Frequency distribution o twist angle o ibre cross-section o saple L 0 P 1, L 0 P 3 and L 0 P 5 97

8 50 Percentage ( %) P1 P3 P Figure 6-6 Frequency distribution o the ibre shape actor o saple L 0 P 1, L 0 P 3 and L 0 P 5 The requency distributions o are shown in Figure 6-6. For saple L 0 P 1, has an alost syetrical distribution. When the pressing pressure is increased, the distribution o skews to a higher range and becoes narrower. Since the three saples were ade ro the sae pulp, ibres in the three saples should have the sae wall thickness and perieter. In that case, according to Equation 6.3, the value o should be ainly aected by the degree o collapse o the ibres. The trends in are thereore ost likely indicative o ibre collapse. As discussed in subsection 6.2.1, the axiu and the iniu values o can be used to assign the ibres as collapsed, partially collapsed or uncollapsed. The boundary values o used to do this were set soewhat arbitrarily. Fibres with less than 0.50 were treated as uncollapsed ibres, ibres with greater than 0.60 were treated as ully collapsed ibres and ibres with in between 0.50 and 0.60 were treated as partially collapsed ibres. Using this classiication 35% o ibres in the slightly pressed saple P 1 were uncollapsed and 33% were ully collapsed (Figure 6-6). Ater the saple was pressed at 500kPa, the percentage o uncollapsed ibres dropped to 24% and the percentage o ully collapsed ibres rose to 49%. Ater the saple was heavily pressed, the percentage o ully collapsed ibres increased to 54% and the percentage o uncollapsed ibres ell to 13%. The percentage o partially collapsed ibres stayed 98

9 relatively stable at around 30% in all the three saples. These quantitative results are consistent with the observations shown in Figure 6-4, but quite dierent ro those reported by Gorres and his coworkers (Gorres 1993) who ound ibres o a sotwood krat pulp can be totally collapsed at a pressing pressure o 2240kPa. This indicates that the collapse behaviour o ibres on glass slides and in paper is dierent. There are two ajor actors that ay cause the observed non-unior collapse o ibres. Firstly, ibre collapsibility will vary between ibres and at dierent points along ibres. Secondly, and perhaps ore iportantly, the non-unior structure o paper causes an uneven pressure transer in the network that results in the non-unior collapse o ibre Relationship between apparent density and the changes in ibre cross-section The apparent density o the handsheets increased ro 286 kg/ 3 to 522 kg/ 3 as the pressure increased ro 100 to 500 kpa. However, pressing at 4,000 kpa increased the sheet density only slightly, to 596 kg/ 3, copared to pressing at 500kPa (Table 6-1). Gorres et al (Gorres 1993) ound a siilar relationship between density o paper and the wet pressing pressure. They ailed to explain the continuing increase in density at high pressure where they believed that ibres are totally collapsed. One reason is that they used the easured degree o collapse o ibres pressed on glass slides to represent the degree o collapse o ibres in paper, which as discussed beore is a dierent situation to collapse in the pressing o paper. Another reason is that they did not take account o the contribution o ibre twist to the density o paper. 99

10 80 Percentage (%) P1 P3 P Out-o-plane angle (degree) Figure 6-7 Frequency distribution o out-o-plane angle o saple L 0 P 1, L 0 P 3 and L 0 P 5 Gorres and Luner developed a odel or the apparent sheet density (Gorres and Luner 1992). The odel assued that increasing ibre delection with increasing wet pressing pressure is a ajor echanis in the densiication o the paper. However, no evidence was given to prove this relationship. In this study, the out-o-plane delection angles o ibre segents in saples L 0 P 1, L 0 P 3 and L 0 P 5 were easured. I the saple is not arranged exactly vertically, an error will arise or the easured value o the out-o-plane delection angle. A correction, siilar to the correction or the twist angle easureents, has been done or each individual ibre to avoid this error. The results (Table 6-1) show that the average out-o-plane delection angle o ibre segents in the o o three saples ranges ro 4.49 to 5.13, and no regular increase with wet pressing pressure can be observed. The distribution, shown in Figure 6-7, o out-o-plane deection angles, also shows no regular trends with pressing level. I the out-o-plane angle o the ibre segent stays constant, the delecting distance o the ibre segent will be reduced as the pressing pressure is increased. This is because wet pressing will reduce the distance between ibre-ibre contacts. To quantiy, using Equations 6.4 and 6.5, the contributions o ibre twist, ibre collapse and gap closure to the easured reduction in paper thickness we assued that the saples have 10 layers and the nuber o layer stays constant as the pressing pressure 100

11 is changed. These data were used to calculate the T twist and T collapse between saples L 0 P 3 and L 0 P 1 and between saples L 0 P 5 and L 0 P 3, and the results are given in Table 6-2. The easured total thickness reduction between saples L 0 P 3 and L 0 P 1 is 114µ, o which the contributions ro the ibre twist and ibre collapse only accounts or 16%. We believe that the rest o the thickness reduction is caused by closing the gap between the layers. The total reduction in paper thickness between saples L 0 P 5 and L 0 P 3 is 18µ which is ostly contributed by ibre twist and ibre collapse. These indings clearly show that ibre twist, ibre collapse and gap closure occur siultaneously at low pressing pressures and the gap closure is the predoinant echanis in paper structure densiication at low pressures. When the pressing pressure was increased ro 500 to 4000kPa, the density increased only slightly. This density increase was ainly due to the additional twist and collapse o the ibres at the high pressure. Table 6-2 Reduction in paper thickness * L 0 P 3 - L 0 P 1 * L 0 P 5 - L 0 P 3 T twist (µ) 7 3 T collapse (µ) ** T gapclosure (µ) 96 5 *** T easured (µ) * L 0P 3- L 0P 1 and L 0P 5- L 0P 3 represents thickness reduction between those saples. ** gapclosure easured twist collaspe T = T T T. *** T easured is the easured total reduction in paper thickness. 6.5 Conclusions A ibre shape actor and twist angle o ibre cross-section in paper have been deined or the purpose o quantiying the changes in the transverse diensions o ibres in paper in wet pressing. It was ound that ibre twist, ibre collapse and gap closure are the ajor types o oveent o ibres in paper in wet pressing. In particular, ibre twist has been ound and quantiied or the irst tie. The results show that the nuber o ibres with o twist angles greater than 10 is reduced by wet pressing and that the average twist angle decreases as the pressing pressure was increased. The ibre shape actor o ibres in the 101

12 lightly pressed handsheets showed an alost syetrical distribution, which indicates that the degree o collapse o ibres in the handsheets is syetrically distributed. The distribution was narrowed and skewed to the high value range when the handsheets were pressed at 500kPa or at 4000kPa. These results show that ibres in the handsheets cannot be totally collapsed by wet pressing even when the very high pressing pressure (4000kPa) is used. The experiental data also suggest that out o plane ibre delection angle is independent o wet pressing pressure. Fibre twist, ibre collapse and gap closure occur siultaneously at low pressing pressures, and the gap closure is the predoinant echanis in paper structure densiication at low pressing pressures (less than 500kPa). Increasing pressing pressure beyond 500kPa only increases the apparent density slightly and the density increase is ainly contributed by the additional twist and collapse o the ibres at the higher pressing pressure. 102