Dynamic mechanical properties of decorative papers impregnated with resin Andreas Kandelbauer, Alfred Teischinger

To cite this version:

Andreas Kandelbauer, Alfred Teischinger. Dynamic mechanical properties of decorative papers im- pregnated with melamine formaldehyde resin. European Journal of and Wood Products, Springer Verlag, 2009, 68 (2), pp.179-187. ￿10.1007/s00107-009-0356-7￿. ￿hal-00568243￿

HAL Id: hal-00568243 https://hal.archives-ouvertes.fr/hal-00568243 Submitted on 23 Feb 2011

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Holz als Roh- und Werkstoff

Draft Manuscript for Review

Dynamic mechanical properties of decorative papers impregnated with melamine formaldehyde resin

ForJournal: Holz Peer als Roh- und WerkstoffReview

Manuscript ID: HRW-08-0139.R1

Manuscript Type: ORIGINALARBEITEN / ORIGINALS

Date Submitted by the 28-Jan-2009 Author:

Complete List of Authors: Kandelbauer, Andreas; University of Natural Resources and Applied Life Sciences, Wood Science and Technology; Kompetenzzentrum Holz GmbH, WOOD Carinthian Competence Centre (W3C) Teischinger, Alfred; University of Natural Resources and Applied Life Sciences, Wood Science and Technology

Decorative laminates, Impregnated paper, Melamine formaldehyde Keywords: resin, Wood based panel industry

Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 1 of 14 Holz als Roh- und Werkstoff

1 2 3 4 Dynamic mechanical properties of decorative 5 6 papers impregnated with melamine 7 formaldehyde resin 8 9 10 11 Dynamisch mechanische Eigenschaften von 12 13 mit Melamin-Formaldehyd-Harz imprägnierten 14 Dekorpapieren 15 16 17 18 1, 2 1 19 Andreas Kandelbauer *, Alfred Teischinger 20 For Peer Review 21 22 1 23 University of Natural Resources and Applied Life Sciences, Dept. of Wood Science and 24 Technology, Peter Jordan Strasse 82, A-1190 Vienna, Austria 25 2Kompetenzzentrum Holz GmbH, WOOD Carinthian Competence Centre (W3C), 26 Klagenfurterstrasse 87 – 89, A-9300 St. Veit an der Glan, [email protected] , 27 28 29 30 * Corresponding author 31 32 33 34 35 Abstract 36 Papers impregnated with melamine formaldehyde based resins are widely used in 37 decorative surface finishing of engineered wood based panels for indoor and outdoor 38 applications. For cost-effective production of high-quality impregnated papers it is of great 39 importance to understand the complex interplay between manufacturing conditions and 40 technological property profile. In the present study, three raw papers from different 41 suppliers were impregnated with melamine formaldehyde resin in an industrial scale 42 experiment to study the influence of some important manufacturing variables on the 43 processability of impregnated papers. As numerical factors the resin loading, the final 44 moisture content and the amount of curing catalyst were systematically varied according 45 to a statistical central composite design. The model papers were analyzed for their 46 rheological and thermal properties using the dynamic mechanical method developed by 47 Golombek. As target values flow time, cure time, curing rate and flexibility were used to 48 calculate quantitative models for the processability of the impregnated papers using 49 response surface methodology. It is shown that the relevant rheological and thermal 50 paper parameters are significantly influenced by the supplier of the raw paper as well as 51 the manufacturing variables. 52 53 54 55 Zusammenfassung 56 Melaminharz imprägnierte Dekorpapiere werden häufig in der Oberflächenveredelung 57 von plattenförmigen Holzwerkstoffen für Innen- und Außenanwendungen verwendet. Für 58 die kosteneffiziente Produktion von qualitativ hochwertigen imprägnierten Papieren ist es 59 nötig, das komplexe Zusammenspiel zwischen den industriellen 60 Herstellungsbedingungen und dem technologischen Eigenschaftsprofil der Materialien zu verstehen. In der vorliegenden Studie wurden drei vergleichbare Rohpapiere von unterschiedlichen Herstellern im Rahmen eines industriellen Experiments unter verschiedenen Bedingungen mit Melaminharz imprägniert, um den Einfluss einiger

1 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 2 of 14

1 2 3 wichtiger Prozessvariablen auf die Herstellung von imprägniertem Papier zu analysieren. 4 Die Modellpapiere wurden auf ihre rheologischen und dynamisch-mechanischen 5 Eigenschaften nach der Methode von Golombek untersucht. Als Zielgrößen wurden 6 Fließzeit, Härtungszeit, Härtungsgeschwindigkeit und Flexibilität bestimmt und der 7 Einfluss der Produktionsvariablen wurde nach Response Surface-Methoden quantifiziert. 8 Es wird gezeigt, dass die relevanten rheologischen und dynamisch-mechanischen 9 Papierparameter signifikant vom Rohpapierhersteller und den Produktionsbedingungen 10 abhängen. 11 12 13 14 15 16 1. Introduction 17 Paper sheets impregnated with an aminoplastic thermosetting resin are frequently used 18 for the surface protection and decoration of medium density fibreboards (MDF) and 19 particleboards in the furniture (Kandelbauer et al. 2008, Nemli and Colakoglu 2005, Soiné 20 1995) and theFor laminate Peerflooring (Kalaycioglu Review and Hiziroglu 2006, Bauer and Kandelbauer 21 2004) industries. Such paper sheets are typically core impregnated with 22 formaldehyde (UF) or melamine formaldehyde (MF) resin in a first step and surface 23 coated with MF resin in a second step (Bader et al. 2000, Ruhdorfer 1980) using a paper 24 impregnation machine such as the one designed by Vits Systems GmbH (Vits 2007, 25 2001), (Figure 1). After impregnation, the paper is dried to a final moisture content of 6 – 26 9 % and subsequently pressed onto the carrier board at temperatures around 180 °C. 27 Since the impregnation resin is not completely cured during paper impregnation there is 28 no additional adhesive required for the surface finishing of the boards. 29 30 31 Figure 1: Schematic representation of the impregnation process of papers for decorative 32 laminates 33 34 The technological properties of impregnated papers must not only fulfil numerous 35 requirements with respect to the final product such as durable surface films, hardness, 36 temperature and chemical resistance etc (cf. for example EN 13 329). They must also fit 37 the demands of the laminates manufacturer in terms of processability such as sufficiently 38 long shelf-life, complete curing during rapid pressing, no blocking when stored in a stack, 39 homogenous film formation etc. The properties of the finished panel product as well as of 40 the intermediate paper sheet are governed by the manufacturing conditions of 41 impregnated papers. Important process parameters that need to be carefully adjusted 42 during the manufacture of such papers are the composition of the resin solution, the 43 numerous roller parameters, the drier conditions and the type of paper used as raw 44 material. 45 One of the most important components of the impregnation solution is the curing catalyst (Becker and Braun 1988). The reactivity of the catalyst governs the speed at which 46 impregnated papers can be processed in the laminating step (Barash 2008). Very 47 reactive systems may cause the resin to pre-cure during impregnated paper manufacture, 48 leading to bad gluability. The reactivity of impregnated paper needs therefore to be 49 carefully adjusted. The drier conditions are important since they govern the cross-linking 50 of the impregnation resin. If set too harsh the required self-gluing property of the paper 51 may be lost rendering the product unusable. 52 The amount of resin is another relevant factor. To reduce the costs, low resin loads are 53 desired. However, if too low amounts of resin are applied the paper is not saturated 54 during core impregnation, the pores are not completely filled and subsequently applied 55 coating resin may sink in leaving a defective surface film (Roberts and Evans 2004). 56 Finally, the paper properties of the raw paper such as density, wet strength or ash 57 content may strongly influence the production process. For instance, paper density 58 affects resin penetration and indirectly determines production speed. Although testing 59 methods have been devised to determine the penetration behaviour of paper/resin 60 systems (see for example http://www.emtec-papertest.com/deutsch/index.html), little is known about the interrelation of raw paper and the properties of impregnated paper. The rheological and thermal properties of impregnated paper such as paper flexibility, resin flow and curing characteristics of MF are very important for evaluating its quality and

2 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 3 of 14 Holz als Roh- und Werkstoff

1 2 3 processability. For instance, the resin flow at the surface of impregnated paper governs 4 film formation and the final surface properties of the coated board. Too low resin flow 5 causes an inhomogeneous film leading to optical defects like low gloss, stains, flow 6 marks or bad cleanability. Resin flow, paper flexibility as well as reactivity of impregnated 7 papers can be quantitatively determined using dynamic mechanical measurements, and 8 thermal analysis of impregnated papers may serve as a rich source of information on 9 both the processability and the final film properties of impregnated paper. 10 In the present contribution the rheological and dynamic mechanical properties of 11 impregnated papers and their relevance for the impregnation process are investigated 12 and discussed. It is aimed at a quantitative mathematical model for the prediction of the 13 processability of impregnated paper sheets under varying manufacturing conditions. In a 14 large-scale industrial experiment at an Austrian impregnation factory the four major 15 process variables raw paper type, resin loading, catalyst level and final moisture content 16 were varied. As response the dynamic mechanical behaviour of the impregnated papers 17 was studied and the effects of process variables on several rheological and thermal 18 paper parameters were quantified using response surface methodology. The used 19 dynamic mechanical descriptors are routinely used for evaluation of the processability of 20 impregnatedFor papers. Hence, Peer the current Review study gives insights in the predictive modelling of 21 the processability of impregnated papers in dependence of the process conditions. 22 23 24 2. Materials and methods 25 26 27 Chemicals 28 29 For paper impregnation, industrial melamine formaldehyde (MF) resin containing various 30 levels of a commercial curing catalyst was used. The impregnation resin was freshly 31 prepared at the Impress Décor Austria GmbH impregnation plant in St. Veit/Glan, Austria 32 and used directly in the impregnation experiments at the industrial machinery. 33 34 35 Paper 36 37 The raw papers used for the present study were all kraft papers of the same type but 38 obtained from three different suppliers, namely Hoffsümmer (raw paper 1), Smurfit (raw 39 paper 2) and Technocell (raw paper 3). They are typical for the use in decorative 40 laminates and are often interchangeably applied for the same type of product. Table 1 41 summarizes their technological properties. 42 43 44 Preparation of impregnated paper samples 45 All impregnation experiments were performed with the industrial paper treatment plant at 46 Impress Décor Austria GmbH in St. Veit/Glan, Austria. As factors, the amount of curing 47 catalyst in the impregnation resin mixture, the final moisture content of impregnated 48 paper, the raw paper type and the resin loading were systematically varied. The amount 49 of curing catalyst and the remaining moisture content each were varied on 5 levels 50 according to a regular central composite design for all three papers at an intermediate 51 resin loading of 135 gm -2. The limits for catalyst concentration were set to 0.5 and 1.5 % 52 (w/w) with respect to the resin dry mass. The drier conditions were adjusted such that the 53 final moisture content of the impregnated paper was regularly varied between the limits of 54 6 and 9 %. Additionally, supplemental papers were impregnated at centre and vertices 55 points under variation of the resin load between 230 and 255 gm -2 final paper weights to 56 account for effects of resin loading. Samples of 30 x 30 cm² were cut from the industrial 57 formats, wrapped in plastic and stored at room temperature and 50 % humidity until 58 further use. 59 60

3 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 4 of 14

1 2 3 Dynamic mechanical measurements 4 5 Dynamic mechanic parameters of the impregnated papers were determined using the 6 paper sheet tester developed by Golombek (Golombek 1991; Figure 2a). This tester is 7 widely used in the laminate industry and consists of an oil bath and a vertically adjustable 8 sample holder coupled to a reversible rotary drive. The motor is connected to a 9 galvanometer which allows continuous monitoring of the motor current required to move 10 the paper sample. 11 12 Figure 2a: Schematic representation of the measurement principle for dynamic 13 mechanical properties of impregnated paper. a … Thermostat; b … Inert silicone oil; c … 14 Paper sample; d … Sample holder; e … Rotary drive; f … Data processor 15 16 Small sample sheets (96 x 60 mm²) of impregnated paper were cut from the stored 17 sheets and immersed in a hot silicone oil bath at 140 °C and the paper was cured. 18 Oscillatory movements were performed at a frequency of 1 rpm and the consumed 19 electrical current for performing the oscillations was recorded in dependence of time. 20 While the impregnatedFor paperPeer was exposed Review to the hot silicone oil, resin curing took place 21 and the stiffness of the paper increased due to cross-linking of MF resin. This increase 22 was recorded during the test based on electrical current measurement. Figure 2b shows 23 the time course of motor current versus time during a typical Golombek experiment. 24 25 26 Figure 2b: Typical response curve of dynamic mechanical measurement using the max 27 Golombek-apparatus. IP … Inflection point; t flow … Flow time; t cure … Time point of 28 maximum curing rate; t cure … Time for 95 % cross-linking of MF resin; r cure … Maximum 29 curing rate 30 31 The initial current value reflects the force required to move the not yet fully warmed, 32 rather stiff paper sample. As the resin liquefies after being exposed to 140 °C the paper 33 sample softens and its resistance against circular motion declines and correspondingly 34 the energy consumption of the motor decreases. After passing through a minimum the 35 current consumption increases since curing of the impregnation resin causes the paper 36 stiffness to rise. At the inflection point (IP) the curing rate reaches its maximum and slows 37 down again until the resin is completely cross-linked. The tests were finished when the 38 impregnation resin was completely cured as indicated by no further increase in electric 39 current. This typically was the case after approx. 5 minutes. 40 From the curve shown in Figure 2b several rheological and thermal parameters were obtained such as the flow time tflow , the curing rate rcure , the time of maximum cure rate 41 max 42 tcure , the curing time tcure and the flexibility f. Flow time, tflow , is the time at 140 °C until 43 the curing of the liquefied resin in the impregnated paper starts to significantly increase 44 the current response. It is defined in relation to the minimum response as illustrated in Figure 2b. The curing rate, rcure , is defined as the slope of the curve at the inflection point 45 max 46 (IP, Figure 2b). Correspondingly, tcure , represents the time where the curing process 47 proceeds at the highest rate and is defined as the intercept with the x-axis at the inflection 48 point. The curing time, tcure , is defined as the time at which 95 % of the cross-linking has 49 occurred. Flexibility, f, is the percentage of cross-linking that can be reached with the 50 liquefied resin. For its definition, also refer to Figure 2b. All parameters describe impregnated papers well and allow drawing conclusions on their technological 51 performance. 52 All dynamic mechanical measurements were repeated twice for each sample. The 53 average values of each response were used for response surface analysis (Myers and 54 Montgomery 2002). In the current paper, the statistical analysis was focussed on tflow , 55 max t , r and f. Since the remaining parameter t yields no additional information it 56 cure cure cure was considered redundant for the presented data material. 57 For the statistical evaluation and response surface modelling of the results the computer 58 software Design Expert (Stat-Ease Inc., Minneapolis, MN, USA) was used. For all three 59 categorical levels studied a quantitative mathematical model was calculated and 60 validated with randomly chosen experimental conditions within the limits of the design which were produced subsequently to the industrial experiment at the paper impregnation plant.

4 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 5 of 14 Holz als Roh- und Werkstoff

1 2 3 4 3. Results and discussion 5 6 7 Raw paper properties 8 Prior to analyzing the dynamic mechanical properties of impregnated papers, the three 9 investigated raw papers were characterized. Although their average paper weight was 10 always the same, the papers supplied from different manufacturers varied significantly in 11 some of their technological properties. Some important characteristics of the various raw 12 papers are summarized in Table 1. 13 Wet strength (WS) is one of the most important parameters of raw papers intended for 14 impregnation with aqueous melamine resins for decorative laminates. The wet strength 15 strongly influences the production speed at which a certain paper may be processed 16 without paper breaks. During impregnation, the papers are wetted and soaked with 17 aqueous dispersions of resin. In a typical impregnation plant, papers are processed with 18 speeds between 50 and 100 mmin -1 and they must withstand the forces of numerous 19 rapidly spinning roller coating cylinders in a wet state. Thus, in principle a higher wet 20 strength especiallyFor in productionPeer direction, Review WS , allows for safer production at higher ║ 21 impregnation speeds. 22 In the present case, RP 1 had clearly the highest values for wet strength whereas both 23 WS ║ and WS ╦ of RP 2 and 3 were about 50 % smaller. With respect to paper anisotropy, 24 all papers displayed a higher wet strength in the production direction than perpendicular 25 to it. RP 3 showed the highest ratio of the values for paper strength in parallel and 26 perpendicular to the direction of paper production for both the wet and the dry strengths 27 of the raw paper. RP 1 was the most anisotropic paper with the same ratio of 1.36 for 28 both the dry and the wet strengths. The other papers were more symmetric in dry

29 strength and showed higher ratios of WS║ and WS ╦. 30 The ash content of a paper is also very important for decorative papers. Since papers 31 with high filler content contain comparatively lower amounts of swellable fibre material at 32 the same paper weights, ash content may influence the speed of resin penetration during 33 impregnation and the total amount of resin load a paper can take up. RP3 showed the 34 highest content of inorganic fillers. The ash content is obviously not directly related to the 35 mechanical strength since although the ash content of RP 1 was not significantly lower 36 than that of RP 3, its absolute values in wet and dry strength were much higher. The pH 37 of all papers was neutral. 38 39 40 Flow time 41 42 The flow time tflow is a very important descriptor of impregnated paper since it indicates 43 the time frame for the MF film formation on the paper surface during the curing in the hot 44 pressing of the laminates. Papers with low tflow are already pre-cured to a high degree. 45 They might not allow the formation of a continuous and homogenous surface and hence 46 yield laminates of poor surface quality. Thus, tflow is desired to be as high as possible. 47 Figure 3 summarizes the influences of the amount of curing catalyst and resin load on the 48 flow time of impregnated paper as measured with the Golombek method for the raw 49 papers from three different suppliers studied (Figure 3 a–c). Table 2 summarizes the 50 statistical parameters of the response surface analysis. For all three papers, the flow time responded linearly to changes in catalyst amount and was also significantly dependent on 51 the paper supplier. Resin load and final moisture content were not significant. The linear 52 equations describing the response surfaces for t constitute three parallel planes as the 53 flow linear combination of catalyst amount (– 9.37113[A]), final moisture content (0.028597[B]) 54 and resin loading (0.048257[D]) and differed only in the off-set of the planes for the three 55 papers (9.71344 for RP1, 11.45718 for RP2 and 10.31303 for RP3, respectively; see also 56 2 Figure 3abc). The accuracy of the linear model was R =0.7700. 57 The flow time was higher with lower concentrations of catalyst. Since small percentages 58 of catalyst means a slower resin cure the resin stays in a liquefied state for a longer time 59 and consequently the flow time increases. Higher resin loadings also caused a slight 60 increase in flow time. This can be explained by the fact that an overall higher amount of resin will take more time to cure and thus will also display slightly longer flow times in the Golombek tester. However, within the experimental setup the variations in resin loadings

5 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 6 of 14

1 2 3 were obviously not wide enough to statistically significantly prove this expected 4 observation. 5 The final moisture content of the paper after drying the impregnated paper in the second 6 drier line does not have any significant effect on the flow time of the material (see Table 7 2). 8 On the other hand, it is interesting that the paper supplier has a significant influence on 9 the flow time (Table 2). Although RP 1 and 3 have practically identical response surfaces, 10 the flow times obtained with RP 2 are comparatively higher and the response surface is 11 shifted for about two seconds towards higher values. Since raw papers 1 and 2 have the 12 same pH of 6.9, this observation is most probably not caused by a direct catalytic effect 13 of the paper on the curing reaction of the reaction through acid catalysis. It is more likely 14 that the mechanical properties of the paper influence the measurement and the findings 15 are related to differences in the stiffness of the papers (see below). 16 17 18 Cure time and curing rate 19 20 The cure timeFor tcure and curingPeer rate rcure areReview important target values since they allow 21 drawing conclusions on the possible manufacturing speed in laminate pressing and can 22 be used to derive suitable press programs. Since cost efficient mass production relies on 23 short cycle times, tcure and rcure should be as high as possible. 24 The results for both tcure and rcure are very similar and the amount of curing catalyst, raw 25 paper supplier and final moisture content were the important factors. With rcure the effect 26 of paper supplier was less pronounced. Figure 4 (a, b, c) summarizes the influences of 27 the amount of curing catalyst and the final moisture content on the cure time of 28 impregnated paper as measured with the Golombek method for the raw papers from the three different suppliers studied. For the response surface analysis a quadratic model 29 2 2 30 was applied. The accuracy of the quadratic model was R =0.9364 ( tcure ) and R =0.9772 31 (rcure ). Table 2 summarizes the corresponding statistical parameters. For raw papers 1 to 32 3, the final equations in terms of actual factors for tcure after impregnation are: 33 2 tcure, RP1 = – 22.33365 + 0.52308 [A] – 0.55043[B] + 0.20787[D] +0.87688[A] – 34 2 -4 2 -3 35 0.011021[B] – 4.56582*10 [D] – 0.024994[A][B] – 0.010630[A][D] +3.16365*10 [B][D]

36 2 tcure, RP2 = – 0.81333 + 0.53862[A] – 0.53591[B] + 0.20130[D] + 0.87688[A] – 37 2 -4 2 -3 38 0.011021[B] – 4.56582*10 [D] – 0.024994[A][B] – 0.010630[A][D] +3.16365*10 [B][D]

39 2 tcure, RP3 = – 20.33254 + 0.33982[A] – 0.53282[B] + 0.19965[D] + 0.87688[A] – 40 2 -4 2 -3 41 0.011021[B] – 4.56582*10 [D] – 0.024994[A][B] – 0.010630[A][D] + 3.16365*10 [B][D]

42 The most significant factor for tcure was the amount of curing catalyst which was expected 43 since the catalyst concentration will directly affect the curing kinetics and high levels of 44 catalyst speed up the condensation. 45 The paper supplier was also very significant which was less expected. It is difficult to 46 understand how papers of different provenience should exhibit any catalytic effect on the 47 curing behaviour of MF resin since they are all of the same pH and are generally 48 considered as an inert resin carrier matrix. The most logic explanation again is that the 49 measurement is influenced by the different mechanical properties of the various papers 50 (see below). 51 Eventually also the final moisture content does effect the curing time. Higher moisture 52 content was associated with lower curing times which can be explained by different pre- 53 condensation degrees of differently dried papers. Higher final moisture content means 54 that the paper is treated more mildly during drying than is paper with a low final moisture 55 content. Hence MF in moist paper is less pre-condensed and will take longer for the 56 cross-linking to complete during the Golombek test. 57 Resin loading was not considered a statistically significant factor although higher resin 58 loads take slightly longer to cure since the total amount of resin present in the reaction 59 system is higher. However, within the experimental error of the design the variations in 60 final paper weights of resin impregnated papers obviously were not pronounced enough to see the effect as statistically significant.

6 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 7 of 14 Holz als Roh- und Werkstoff

1 2 3 Flexibility 4 5 Flexibility is another important descriptor for impregnated papers and gives an indication 6 of its cross-linking status. In highly flexible papers the degree of curing during lamination 7 is high and hence a large portion of the MF remains yet uncured. On the other hand if f is 8 too low, the remaining activity of MF may not be sufficient for proper lamination. Values of 9 f < 80 typically result in a deficient surface finish. 10 Figure 5 summarizes the influences of resin loading and the final moisture content on the 11 flexibility f of impregnated paper at low (Figures 5 a, b and c) and high (Figures 5 d, e and 12 f) amounts of curing catalyst as measured with the Golombek method for the raw papers 13 from the three different suppliers studied. For the response surface analysis a quadratic 2 14 model was applied. The accuracy of the quadratic model was R =0.9658. Table 2 15 summarizes the corresponding statistical parameters. For raw papers 1 to 3, the final 16 equations in terms of actual factors for f after impregnation are: 17 2 fRP1 = –1868.08355 +110.92874[A] +10.68553[B] +15.27254[D] – 19.67073[A] – 18 1.02168[B] 2 – 0.031080[D] 2 + 3.35986[A][B] – 0.42632[A][D] + 0.025223[B][D] 19 20 For Peer Review 2 fRP2 = –1838.39170 + 106.23929[A] + 9.77641[B] + 15.20582[D] – 19.67073[A] – 21 1.02168[B] 2 – 0.031080[D] 2 + 3.35986[A][B] – 0.42632[A][D] + 0.025223[B][D] 22 23 2 fRP3 = –1701.65691 + 102.69526[A] + 10.80965[B] + 14.61338[D] – 19.67073[A] – 24 1.02168[B] 2 – 0.031080[D] 2 + 3.35986[A][B] – 0.42632[A][D] + 0.025223[B][D] 25 26 The most important process variable for flexibility of impregnated paper was the final 27 moisture content with a P>F much lower than 0.0001. With increasing moisture content 28 the flexibility increased. This is explained by the fact that water molecules act as 29 flexibilizers in the three-dimensional resin network. Flexibility also varied among the 30 tested papers and different shapes of response surfaces were calculated for RP1, RP2 31 and RP3. The effect of raw paper is illustrated by the significant P>F-value of factor [C] in 32 Table 2. The values for flexibility were always lower at lower moisture content. While the 33 shapes of the response surfaces for RP1 and RP2 (Figures 5a/5d and 5b/5e) were very 34 similar, the behaviour of RP 3 (Figures 5c/5f) differed much. The most significant 35 influence of the catalyst amount was observed with RP 3 (5c / 5f), where the slope of the 36 response surface became steeper when the catalyst amount was increased. In contrast, 37 with RP 1 and 2, flexibility was practically not influenced at all by the amount of curing 38 catalyst while for RP 3 a moderate effect was observed. Although obvious in the 39 response surfaces, the effect of resin load was statistically much less significant than that 40 for the catalyst amount. When comparing the P>F values for factor [D] (=resin load) it is 41 seen that the only term of real statistical significance is for the interaction term between 42 [C] (=the raw paper) and [D]. Neither [D] (P>F = 0.3726) nor [D2] (P>F = 0.1387) contribute much to the overall model. In principle, [D] and [D 2] could even be omitted from 43 the model calculations without loosing much information or changing the response 44 surface much. However, it is difficult to meaningfully discuss effects of interactions 45 without considering at least some contribution of the corresponding main effects of the 46 factor. 47

48 49 Influence of raw paper supplier 50 51 For all responses it was found that the characteristic shape of the response surface was 52 the same for all three raw papers studied. The mathematical form of the response surface 53 was equivalent and the numerical factors final moisture content, catalyst concentration 54 and resin loading influenced the technological properties of all papers in a similar way. 55 However, the papers varied significantly in the exact quantitative values and even more 56 important in the gradient of susceptibility of a response in dependence of the factor level. 57 All measurement results were strongly influenced by the raw paper (see also the 58 significance level for raw paper in Table 2) and the factor “raw paper supplier” played an 59 important role. 60 In general, if papers of different densities are measured with the Golombek tester, no comparable testing results are obtained since their mechanical stiffness is co-measured with the apparatus. Papers of a density of 105 gm -2 evidently exhibit much greater resistance against torsion in the hot oil bath than papers of a density of 80 gm -2 and thus

7 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 8 of 14

1 2 3 they greatly vary in their rheological response (data not shown). Eventually, papers of 4 much lower grammage such as overlay papers of 20–30 gm -2 do not give any useful 5 response at all. Since their stiffness is much too low, they simply fold and agglutinate 6 when they are subjected to rotation in the hot oil bath. 7 However, in the present case papers of the same grammage (~105 g) were compared to 8 each other and the actual quantitative values for the target responses did still not match 9 exactly, although the overall trends as reflected by the form of their response surface 10 profiles were similar. To some extent this may be explained by differences in paper 11 stiffness of the raw papers as reflected by the variety in mechanical strength parameters 12 given in Table 1. Still, this finding is rather unexpected and interesting since in industrial 13 practice papers of the same type (décor paper, balance sheets or overlays) and the same 14 grammage are often used interchangeably for the same purpose. Papers are often 15 exchanged from one supplier to another depending on the current cost structure but the 16 corresponding recipes and procedures for their impregnation are treated as compatible. 17 Typically, the Golombek tester is used for determining the impregnation properties of a 18 new lot of raw paper based on laboratory impregnations in comparison to an earlier 19 reference paper from another lot; if the measured values match, the tested impregnated 20 paper is assumedFor to result Peer in the expected Review technological profile. The results of the current 21 study however imply that the absolute reference values used for comparison with new 22 production lots of raw paper purchased from different suppliers do not necessarily match 23 perfectly since the absolute values of rheological and thermal responses differ 24 significantly. The findings indicate that models based on data obtained with one paper 25 may not be simply transferred when the raw paper supplier is changed but will have to be 26 adapted. 27 It is difficult to relate the raw paper properties from Table 1 to the results of the response 28 factor analysis. It seems that values for flow time, cure time and flexibility are all 29 especially low with the raw paper that displays the highest values for wet strength 30 whereas the reason for this might not be simply deduced. Based on the results collected 31 so far it is also not yet possible to quantify the relation between selected raw paper 32 quality parameters and the technological property profiles of papers in their impregnated 33 state. Such studies will be the focus of future work. 34 35 36 4. Conclusion 37 38 In the present study, the mathematical modelling of practically rheological and thermal 39 parameters for impregnated paper characterization has been accomplished. The 40 calculations were based on the response surface analysis of an industrial experiment 41 where the four production parameters raw material (raw paper supplier), impregnation 42 resin mixture (catalyst concentration), roller coater (resin loading) and dryer (final 43 moisture content) were systematically varied according to a central composite statistical 44 design. It was found that the numerical factors catalyst amount, moisture content and 45 paper supplier were statistically significant and strongly influenced the further 46 processability of impregnated papers which was quantitatively expressed in terms of flow 47 time, cure time, curing rate and flexibility of the impregnated paper sheet. A quantitative 48 model suitable for predicting the paper processability in dependence of chosen 49 production conditions was established. Moreover, the provenience of raw paper 50 influenced strongly the rheological and thermal properties of the impregnated papers as 51 measured with the Golombek apparatus. Paper wet strength seems to be correlated with 52 the Golombek parameters as it is associated with the mechanical stiffness of a paper. 53 54 55 56 5. Acknowledgements 57 58 This work was funded by the Austrian Research Promotion Agency FFG within the 59 COMET programme line. The authors would like to thank the representatives of Impress 60 Décor Austria GmbH for their cooperation in performing the experimental work and permitting publication of the presented data.

8 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 9 of 14 Holz als Roh- und Werkstoff

1 2 3 4 6. References 5 6 Bader K, Wippel W, Kummer M (2000) Concentrated stable aqueous melamine formaldehyde resin composition. PCT/EP00/09749 7

8 Bauer K, Kandelbauer A (2004) Transparenzverbesserungen im 9 Fußbodenlaminatbereich. In: Bachofner, R., Nanocoating Days, Nanofair: Nanocoating 10 Days, 14.–15. Sept. 2004, St. Gallen, Switzerland, 1-13 11

12 Barash L (2008) Laminates, decorative surfaces. Conference Proceedings of the 13 European Laminates Conference and Workshop, Paper 5: 35–54, European Laminates 14 Conference and Workshop, 8.–10. April 2008, Vienna 15 16 Becker GW, Braun D (1988) Duroplaste – Kunststoff Handbuch Vol. 10, 2 nd Edition, Carl 17 Hanser Verlag, München – Wien 18 19 EN 13 329 European Standard (2001) Laminate floor coverings – specifications, 20 requirements,For and testing Peer methods. Österreichisches Review Normungsinstitut, Wien 21 22 Golombek J (1991) Method for determining solidification degree of carrier impregnated 23 with reaction resin. US Patent US 5,001,068 24 25 Kalaycioglu H, Hiziroglu S (2006) Evaluation of surface characteristics of laminated 26 flooring. Building and Environment 41, 756–762 27 28 Kandelbauer A, Burger-Scheidlin H, Hölbling B (2008) Rapid manufacturing of high gloss 29 melamine films. Conference Proceedings of the European Laminates Conference and 30 Workshop, Paper 3: 17–22, European Laminates Conference and Workshop, 8.–10. April 31 2008, Vienna 32 33 Myers RH, Montgomery DC (2002) Response Surface Methodology – Process and 34 Product Optimization Using Designed Experiments. 2 nd edition, John Wiley & Sons, New 35 York, USA 36 37 Nemli G, Colakoglu G (2005) The influence of lamination technique on the properties of 38 particleboard. Building and Environment 40, 83–87 39 40 Roberts RJ, Evans PD (2004) Effects of manufacturing variables on surface quality and 41 distribution of melamine formaldehyde resin in paper laminates. Composites Part A: 42 Applied Science and Manufacturing 36: 95–104 43 44 Ruhdorfer H (1980) Verfahren zum Herstellen einer Tränkharzmischung für 45 Trägerbahnen zur Oberflächenveredelung von Holzwerkstoffplatten. AT 374 811 46 47 Soiné HG (1995) Holzwerkstoffe – Herstellung und Verarbeitung. Platten, 48 Beschichtungsstoffe, Formteile, Türen, Möbel. DRW-Verlag, Leinfelden-Echterdingen, 49 Germany 50 51 Vits systems GmbH (2007) Vorrichtung zum Imprägnieren einer durchlaufenden Bahn. 52 Ger. Offen. DE 202007011437U 53 54 Vits systems GmbH (2001) Vorrichtung zum Auftragen einer Beschichtungsmasse auf 55 eine durchlaufende Papierbahn. Ger. Offen. DE 20020255U 56 57 58 59 60

9 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 10 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Legends of Figures and Tables 16 17 18 Tables 19 20 Table 1: TechnologicalFor Peerproperties of raw Review papers tested in the present study 21

22 23 Table 2: ANOVA for the response surface analysis of the response „flow time of 24 impregnated paper”, „cure time of impregnated paper” and „flexibility of impregnated 25 paper” 26 27 28 29 Figures 30 31 Figure 1: Schematic representation of the impregnation process of papers for decorative 32 laminates 33

34 35 Figure 2a: Schematic representation of the measurement principle for dynamic 36 mechanical properties of impregnated paper. a … Thermostat; b … Inert silicone oil; c … 37 Paper sample; d … Sample holder; e … Rotary drive; f … Data processor 38 39 40 Figure 2b: Typical response curve of dynamic mechanical measurement using the max 41 Golombek-apparatus. IP … Inflection point; t flow … Flow time; t cure … Time point of 42 maximum curing rate; t cure … Time for 95 % cross-linking of MF resin; r cure … Maximum 43 curing rate 44 45 46 Figure 3 (a–c): Influence of the amount of catalyst and the resin load on the flow time of 47 the impregnated paper for the three raw papers tested 48 49 50 Figure 4 (a–c): Influence of the final moisture content and the amount of catalyst on the 51 cure time of impregnated papers 1 – 3 at high resin loads 52 53 54 Figure 5 (a–f): Influence of final moisture content and resin load on the flexibility of 55 impregnated papers 1 – 3 for low (a–c) and high (d–f) amounts of curing catalyst in the 56 resin recipe 57 58 59 Tabelle 1: Technologische Eigenschaften der in der Studie getesteten 60 Rohpapiere

Tabelle 2: ANOVA für die Response Surface Analyse der Zeitgrößen "Fließzeit", "Härtungszeit" und "Flexibilität" von imprägniertem Papier

10 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 11 of 14 Holz als Roh- und Werkstoff

1 2 3 4 Abb. 1: Schematische Darstellung des Imprägnierprozesses von Papieren 5 für dekorative Schichtstoffe 6 7 Abb. 2a: Schematische Darstellung des Messprinzips der 8 dynamisch-mechanischen Eigenschaften von imprägniertem Papier. a) 9 Thermostat; b) inertes Silikonöl; c) Papierprobe; d) Probenhalterung; e) Motor mit Antriebswelle; 10 f) Datenverarbeitung 11 mit dem Golombek-Apparat. IP: Wendepunkt; tflow: Fließzeit; tcuremax:Zeitpunkt der maximalen 12 Aushärtungsgeschwindigkeit; tcure: Zeit bis zum Erreichen von 95 % Vernetzung von 13 Melaminharz; rcure: Maximale Aushärtungsgeschwindigkeit 14 15 16 Abb. 3(a-c): Einfluss der Härtermenge und der Harzbeladung auf die 17 Fließzeit von imprägniertem Papier für die drei getesteten Rohpapiere 18 19 Abb. 4 (a-c): Einfluss des Feuchtigkeitsgehaltes des imprägnierten 20 Papiers sowieFor der Härtermenge Peer im Imprägnierharz Review auf die Härtungszeit der 21 imprägnierten Papiere 1 - 3 bei hohen Harzbeladungen 22 23 Abb. 5 (a-f): Einfluss des Feuchtigkeitsgehaltes des imprägnierten 24 Papiers und der Harzbeladung auf die Flexibilität der imprägnierten 25 Papiere 1 - 3 bei niedrigen (a-c) und hohen (d-f) Konzentrationen von 26 Härtungskatalysator in der Imprägnierharzmischung 27 28 29 30 31 32 33 Tables 34 35 36 37 Table 1: Technological properties of raw papers tested in the present study 38 39 RP D WS ║ /WS ╦ WS ║/WS ╦ DS ║ DS ╦ DS ║/DS ╦ Ash pH [gm -2] [g/15 [g/15 [g/15 mm] [g/15 mm] [%] 40 mm] mm] 41 1 106 1700 1250 1.36 4140 3040 1.36 26.5 6.9 42 2 105 910 605 1.50 3300 2610 1.26 22.4 6.9 43 3 105 825 480 1.72 3100 2050 1.51 26.8 6.6 44 RP … raw paper, D … density, WS ║ … wet strength parallel to the direction of paper production, WS ╦ … 45 wet strength perpendicular to the direction of paper production, DS … dry strength parallel to the direction 46 ║ of paper production, DS ╦… dry strength perpendicular to the direction of paper production 47 48 49 50 Table 2: ANOVA results (partial sum of squares) for the response surface analysis of the responses flow time, 51 cure time and flexibility of impregnated paper for the three different raw papers 1 to 3 52 53 tflow rcure tcure f 54 F-value P > F F-value P > F F-value P > F F-value P > F

55 Model 22.09 < 0.0001 53.06 < 0.0001 18.20 < 0.0001 34.84 < 0.0001 56 [A] 86.27 < 0.0001 569.78 < 0.0001 168.96 < 0.0001 7.15 0.0142 57 [B] 0.068 0.7958 7.87 0.0106 11.17 0.0031 105.52 < 0.0001 58 [C] 12.32 0.0001 15.54 < 0.0001 10.84 0.0006 6.79 0.0053 59 [D] 2.10 0.1568 4.81 0.0397 2.49 0.1293 0.83 0.3726 60 [A] 2 11.06 0.0032 6.38 0.0196 1.55 0.2262 [B] 2 15.32 0.0008 8.67 0.0078 36.03 < 0.0001 [D] 2 0.69 0.4165 1.06 0.3158 2.37 0.1387 [AB] 0.021 0.8850 0.14 0.7149 1.20 0.2859 [AC] 0.72 0.4993 1.11 0.3473 0.98 0.3904

11 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 12 of 14

1 2 3 [AD] 0.024 0.8790 0.16 0.6887 0.13 0.7236 4 [BC] 0.44 0.6527 0.55 0.5826 1.78 0.1939 5 [BD] 1.39 0.2509 2.86 0.1053 0.088 0.7695 6 [CD] 1.08 0.3586 0.93 0.4092 3.80 0.0391 7

8 tflow … flow time; rcure … curing rate; tcure … cure time; f … flexibility; [A] … amount of curing catalyst in 9 the impregnation resin mixture; [B] … final moisture content of impregnated paper; [C] … raw paper 10 supplier; [D] … resin load of impregnated paper 11 12 aValues of “P>F” less than 0.05 (5 %) indicate that model terms are significant 13 14 15 16 Figure 1 17 18 Aeration Edge 19 Unwind Streak Air Drier 1 Air Drier 2 Trimming Cutting 20 Station For Peer Review 21 22 23 Stacking 24 Coating 25 26 Prewetting Core 27 Impregnation 28 29 30 31 32 Figure 2 a 33 34 35 36 37 e f 38 39 40 41 42 43 44 45 46 47 a b c d 48 49 50 51 52 53 54 55 56 57 58 59 60

12 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Page 13 of 14 Holz als Roh- und Werkstoff

1 2 3 Figure 2 b 4 5 6 7 8 9 tC95% 10 11 12 rcure 13

14 Electric Current IP 15 16 17 18 Flexibility f [%] 95 % Curing 95 % 19 4 % Curing 100 % 20 For Peer Review 21 tflow 22 Time max 23 tcure 24 t 25 cure 26 27 28 Fig. 3 abc 29 30 Paper 1 Paper 2 Paper 3

1.15 10.5 31 11.0 12.5 32 11.0 11.5 33 11.5 13.0 1.05 12.0 13.5 34 12.0 12.5 35 12.5 14.0 36 0.95 13.0 13.0 14.5 37 13.5 13.5 15.0 14.0 38 Catalyst [%] 15.5 0.85 14.0 39 14.5 14.5 16.0 40 15.0 16.5 41 0.75 42 230 235 240 245 250 230 235 240 245 250 230 235 240 245 250 43 -2 -2 -2 44 Resin load [gm ] Resin load [gm ] Resin load [gm ] 45 46 Fig. 4 abc 47 48 49 Paper 1 Paper 2 Paper 3 50 9.00 1.05 1.10 1.00 51 1.00 1.05 0.95 0.90 52 8.25 0.95 1.00 0.85 53 0.90 0.95 0.80 54 0.85 0.90 7.50 0.75 55 0.85 56 0.80 0.70 0.80 57 6.75 0.75 0.65 0.75 0.60 58 0.70 Moisture content Moisture [%] 59 6.00 0.70 0.55 60 0.75 0.85 0.95 1.05 1.15 0.75 0.85 0.95 1.05 1.15 0.75 0.85 0.95 1.05 1.15

Catalyst [%] Catalyst [%] Catalyst [%]

13 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany Holz als Roh- und Werkstoff Page 14 of 14

1 2 3 Fig. 5abc 4 5 Paper 1 Paper 2 Paper 3 6 250 7 77 8 82 ]

-2 245 9 93 87 10 95

11 240 92 12 13 88 92 14 235 89 83 Resin load [gm 15 86

16 78 83 97 230 17 6.00 6.75 7.50 8.25 9.00 6.00 6.75 7.50 8.25 9.00 6.00 6.75 7.50 8.25 9.00 18 19 Moisture content [%] Moisture content [%] Moisture content [%] 20 For Peer Review 21 22 Fig. 5 def 23 24 25 26 Paper 1 Paper 2 Paper 3 250 27 80 72 28 ] 78 -2 29 245 82 30 84 31 93 84 86 88 90 92 94 240 88 32 90 33

34 235 83 35 Resin load [gm

36 78 96 37 230 38 6.00 6.75 7.50 8.25 9.00 6.00 6.75 7.50 8.25 9.00 6.00 6.75 7.50 8.25 9.00 39 Moisture content [%] Moisture content [%] Moisture content [%] 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

14 Editorial Office, TU Mˆ…nchen, Holzforschung Mˆ…nchen, Winzererstr. 45, 80797 Mˆ…nchen, Germany