Introduction

The functionality of paper and paper board in packaging applications is greatly improved when its flat two-dimensional structure can be formed into three-dimensional shapes. Especially when deeper shapes can be created, the paper and board can replace plastics in certain packaging applications. In fixed-blank forming processes, the limiting factor is the extensibility of the paper material. The extensibility, i.e. the material ability to elongate under external load, is low for both fibres and papers dried under restraint. Recently, extensibility has been studied widely on several structural levels (fibres, fibre bonds and network structure) for improved paperboard performance in packaging applications (Zeng et al. 2013; Vishtal 2015; Khakalo 2017a, b; Östlund 2017; Kouko and Retulainen 2018; Ketola et al. 2018; Kouko et al. 2020; Laukala et al. 2023; Marin 2023). The combined effect of extensibility at the different structural levels can be additive (Svensson et al. 2013).

In addition, extensibility of cellulose paper web is affected by the loading conditions and deformation mechanics of the structure, properties of structural components, and the factors that affect the shape of the load-deformation behaviour under applied 3D forming conditions (temperature, moisture, strain rate, etc.). Elongation of many commercial produced paperboard grades are typically between 2.5 and 8%, depending on the orientation direction (Niskanen and Kärenlampi 1998; Kouko et al. 2023). However, elongations above 20% in machine direction (MD) have been achieved by creping tissue papers (Park et al. 2020). In-plane compaction (mechanical shrinking) operations can produce elongations above 10% in MD for commercial paper grades. There are at least two technologies for in-plane compaction that are applied commercially: Expanda® and Clupak® processes (Ihrman and Öhrn 1965; Cluett 1947).

The mechanical treatment of pulp has been shown to be an efficient way to improve the elongation of paper (Vishtal and Retulainen 2014). Improving the strength of the fibre-fibre bonding also increases the extensibility and strength of the paper (Ketola et al. 2018). Poorly bonded fibre networks fail before the full elongation potential of fibres is reached, whereas in well-bonded networks the fibres are under greater stress and are more strained before network failure (Kouko et al. 2020). The treatment of fibres with increased curl has also been shown to improve the 3D elongation potential in fixed blank forming (Laukala et al. 2023). In this work, the effect of mechanical treatments on the tensile behaviour of commercial bleached softwood kraft (BSK) pulps and the corresponding papers was studied in order to improve the extensibility and 3D forming performance of this type of material.

In papermaking, dry strength is typically enhanced with polymers. The most common ones are starch and synthetic dry-strength additives, e.g. polyacrylamide (PAM) (Ketola and Andersson 1999). When the purpose is to enhance 3D formability, the bonds should be such that they become flexible during press forming, but are rigid and strong afterwards. Typically, this is achieved using thermoplastic polymers like polyurethane, latex or PVA (Juvonen et al. 2015).

Long man-made fibres can be longer than pulp fibres. Their role is to improve the local strength of the structure. Paper rupture starts from the weakest spot, where it spreads through the structure. A long fibre, which is bonded to the stronger areas on both side of the weak spot, enhances the resistance against rupture and can prevent the breakage locally. The usage of long fibres has been seen to improve elongation of paper (Torniainen et al. 2014).

The majority of 3D forming investigations for paperboards in recent years have been focused on processes such as press forming (Tanninen et al. 2020), deep drawing (Hofmann et al. 2019) and hydroforming (Franke et al. 2021). Dry thermoforming of paper using a method developed for plastics has recently been investigated using commercial paperboards, and it has been observed that the CD strain of breakage of paperboards correlates well with quality parameters, such as forming angle, forming depth, and thickness changes of materials (Afshariantorghabeh et al. 2022). It has also been found that in the thermoforming of paperboards, in addition to the material properties, the effect of forming tooling and parameters can have a significant impact on the final formability of a thermoformed product (Afshariantorghabeh et al. 2023). The listed forming methods differ from each other significantly – some of them are mainly based on the stretching of the materials, while in others the controlled folding of the materials plays the main role. There are also essential differences in the magnitude of the forming forces and temperatures used.

The objective of the work was to increase the extensibility of foam-laid formed material by using binder chemicals and binder fibre applications with selected raw material composition. The research questions were: (1) Can certain thermoplastic binder chemicals or fibres improve extensibility? (2) Are they effective for extensibility? (3) Can they be applied using foam-laid forming? This study introduced a novel concept for producing 3D formable paperboard materials with competitive extensibility for many industrial applications. In addition to demanding performance requirements of barrier properties on paperboard, the extensibility of paperboard has so far been the most significant limiting factor for an industrial breakthrough of the paperboard-based dry 3D formable material concept. Large unit quantity for packages is one of the key benefits of dry 3D forming production concept compared to fibre moulding production concept. This study fills the gap between the current paperboards and advanced materials with higher extensibility requirement.

Experimental

Raw materials

The cellulose pulp raw material was a dried bleached softwood kraft (BSK) pulp, a mixture of spruce and pine, from a Nordic pulp mill. The pulp composition (according to the supplier) is mainly pine (Pinus sylvestris) (50–85%), while spruce (Picea abies) content is in the range of 15–50%. Low consistency (LC) refining of the BSK pulp was done at KCL at 4% consistency using conical plates in a Voith refiner. The applied specific refining energy was 100 kWh/t and the specific edge load 2.5 J/m. The Schopper-Riegler value of the 2 kg batch that was only LC-refined was increased from 13.5 to 21.5. High-consistency (HC) refining of the BSK pulp was done at VTT. A batch of about 3 kg was disintegrated in self-made dynamic drainage jar SDDJ, tumbled and homogenised. Refining was performed at 3000 rpm using the ‘R + S D2C505_uusi’ plate. The SEC was 0.42 MWh/t.

Fibre properties of the unrefined, LC-refined and HC + LC-refined pulps were measured using the L&W Fiber Tester Plus (the fibres were measured in swollen state). The length weighted (LW) average fibre lengths and widths, as well as the shape factors, are presented in Table 1.

Table 1 Average length weighted (LW) fibre properties with standard deviation of two repeats and SR number of the studied BSK pulps
Full size table

PVA (Polyvinyl alcohol) fibres were used as a thermoplastic fibre. The PVA fibres were VPB105-2 (dtex 1.2) with a length of 4 mm, from Kuraray. Two kinds of foaming agents were used. Anionic sodium dodecyl sulphate SDS (Merck Chemicals Ltd) was used only as a surfactant for enabling foaming, with a dosage of 0.3 g/L. The molecular weight of SDS (CH3(CH2)11SO4Na) is 288 g/mol. SDS is mentioned in the recommendation of BfR in the list of materials intended to come into contact with food and it is also biodegradable. CH-Polymers CHP 689 was a self-cross-linking aqueous dispersion based on acrylic and methacrylic esters, with a particle size of 0.2 μm and solids content 41%. The chemical was relatively hard polymer (Tg 50 °C) and it had maximal thermoset properties. The contact angle of the polymer film and water was 45 °, i.e. not very hydrophobic, but still water-durable. CHP 689 was a foamable latex, and was thus used as a foaming agent in foam-laid forming with a dosage of 40 g/L of suspension. Air content of 62% was achieved with it in foaming phase and it was able to increase bonding between the cellulose fibres.

Preparation of the sheets

The foam-laid forming of the sheets was made according to the methods presented by Hjelt et al. (2021). The used tailored methods were based on the instructions from an industrial partner (other than Billerud) and were not based on any standard method. The foam-laid formed sheets were 350 mm × 220 mm and the target basis weight of all the sheets was 140 g/m2. When the fibre-foam suspension was prepared, the mixing was stopped and wet fibre foam was poured into a modified hand sheet mould, where the flow was directed onto a forming fabric. The foam volume was 3 L with fibre consistency 3.6 g/L and air content between 40% and 67% depending on the fibre raw material and used surfactant. The used forming fabrics were similar to those used in water-laid forming. Foam was decanted into the hand sheet mould using a tilted plate, which oriented the fibres along the flow direction. The usual ratio of tensile indices in a machine and cross-machine direction has been about 2. If foam was poured into the centre of the mould directly from a container, the foam would spread radially along the fabric, causing circular patterns in the formed sheet. Vacuum-assisted dewatering was used in the mould.

The water-laid forming of the sheets was made using a vacuum-assisted sheet former introduced by Lehmonen et al. (2017). In this study, a vacuum level of 30 kPa was applied. The area of the formed sheet was approximately 320 mm × 220 mm, and similar forming fabrics as in paper machine surroundings was used. Forming consistency of the sheets was approximately 0.68 g/L. In contrary to the foam-laid laboratory sheets, all the water-laid laboratory sheets had random fibre orientation.

The sheet forming was usually followed by wet pressing at 350 kPa for 5 min + 2 min with blotting papers. The used wet pressing method followed ISO 5269-1 with the modified sheet size. However, in some cases wet pressing was omitted. The sheets were dried unrestrained, allowing for drying shrinkage at a temperature of 23 °C and at 50% relative humidity in accordance with ISO 187:1990. Some of the sheets that were supposed to be compacted in-plane were marked with red and blue lines with a 30 mm distance in order to detect dimensional changes due to in-plane compaction and 3D forming (Fig. 1).

Fig. 1
figure 1

A foam-formed sheet (size 350 mm x 220 mm) marked with red and blue lines with 30 mm distance before compaction

Full size image

The unrestrained drying of the sheets at standard climate was performed between two forming wires that were supported by a rigid backing and separated by 5 mm-thick rods on the edges, in order to allow for shrinkage and prevent cockling and curling of the sheets. The drying rigs are presented in (Kouko and Retulainen 2018).

Pilot machine formed samples were made using the VTT SUORA pilot line, consisting of an approach system, a forming section (hybrid former), and of a press section (one shoe nip) (Lehmonen et al. 2019, Kiiskinen et al. 2019). Detailed forming conditions of the water-laid pilot paper samples used in this study is presented in (Kouko and Hjelt under review).

Before the elongation test and 3D forming, the latex containing laboratory sheets was heat-treated. The heat treatment was done in an oven at 140 ºC for individual sheets without surrounding plates. The activation time was 5 min starting from the moment the oven reached the target temperature. The time for increasing the temperature was a couple of minutes. The heat-treatment procedure was obtained from the chemical supplier.

Laboratory in-plane compaction

The samples presented in 3.2. experienced in-plane compaction in the MD and in the CD directions with various ways that are presented in detailed manner. The water-laid pilot sheets presented in 3.3. were two-way in-plane compacted at first in the MD and then in the CD and the one-way compaction was performed in the MD. Contrary to the pilot paper samples, all the rest laboratory sheets presented in 3.4 and after were two-way in-plane compacted at first in the CD and then in the MD and the one-way compaction was performed in the CD. Before the in-plane compaction the foam-laid laboratory sheets were cut in two equal size pieces width 175 mm (equal to the original MD) and length 220 mm (equal to the original CD).

Before the compaction of the sheets, the material was re-wetted to reduce stiffness of the fibre network. According to Vishtal and Retulainen (2014), the most suitable solids content of the sheet for in-plane compaction is 60–75%. The sheets were weighed (Wdry) and their equilibrium solids content was estimated to be 93%. The wet sheet weight, Wwet, corresponding to the target solids content (x) that allowed functioning compaction, was calculated using Eq. 1.

$${W}_{wet}=\frac{{W}_{dry} \cdot 0.93}{x}$$
(1)

The target solids content was determined individually for the different materials by compaction trials in order to find the operating window without wrinkling or breaks. Water was then sprayed evenly onto both sides of the sheets by means of a sprayer (Fig. 2b) until the target weight was reached. The re-moistened sheets were placed in plastic bags, which were placed on a planar surface and sealed. To even out the moisture gradient (due to spraying) within the sheets, the sheets were kept in the plastic bag for a minimum of 5 h.

Fig. 2
figure 2

a Schematic presentation of the Expanda process used for in-plane compaction of paper sheets at Billerud. b Re-moisturisation of sheets. c Drying phase after compaction was conducted using an L&W Rapid Dryer

Full size image

In-plane compaction was performed using the Billerud (Gruvön) laboratory’s Expanda compactor (see Fig. 2a) in order to increase extensibility of the sheets. The Expanda unit consisted of a double roll press. One of the rolls was covered with rubber and the other was a hard roll (e.g. steel). The speed of the rubber-covered roll (ω1) was reduced in comparison to the hard roll (ω2) and in the ideal case, the compaction of the paper web can be approximated by the speed difference (%) between the two rolls. In the Expanda unit the speed, speed difference and pressure could be adjusted. The unit could be run with a speed difference of as much as 50% between the two rolls and a speed of up to 20 m/min. Directly after the compaction step, the sheets were placed inside a dryer. In this case the drying was performed using a L&W Rapid Dryer (Fig. 2c). In the L&W Rapid Dryer the sheets were placed between two heated plates held at 180 °C with low pressure. Even though the sheets were held between the plates, restrained drying was not achieved, and a certain shrinkage was able to occur. The sheets were placed inside the dryer until fully dry.

Press forming method

The formability of the sample materials was evaluated with a series of press-forming tests using the MiniMould tool (see Tanninen et al. 2017a for a more detailed description of the method). The basic principle of the press forming process was to place a pre-cut (and optionally pre-creased) tray blank between male and female moulds (see Fig. 3). The tray blank is clamped between the rim tool and the female mould, and sliding of the blank is controlled by a blank-holding force.

Fig. 3
figure 3

The press-forming mould device at LUT (Tanninen et al. 2017a). The shape was created by male (1) and female moulds (2). Blank holding force was applied by a rim tool (3). Heating unit (4)

Full size image

The blank-holding force adjustment can be used as a means of preventing or reducing material breakage during the forming cycle (Tanninen et al. 2020). In order to maximise the rigidity of the trays, the blank is usually fed into the press so that the longer side of the blank is parallel to the machine direction (MD) (Tanninen et al. 2017a).

A MiniMould tool was set for testing laboratory hand sheets into a Packer 2020 forming device (see Fig. 4). The size of the MiniMould test tray was 90 mm × 80 mm × 35 mm (depth can be varied) and, accordingly, the dimensions of the tray blank, i.e. the area compressed between the mould surfaces, was 140 mm × 150 mm. Tray blanks were die-cut from the test sheets. The temperature of the mould parts was monitored with three sensors. The forming tool allowed both a fixed and a sliding blank converting procedure by adjusting the blank holding force. In the fixed blank procedure, the blank was not allowed to move during the forming process, and in the sliding blank procedure the blank was allowed to move during the forming process. In the latter method, some planar movement and folding also occurred in the material sheet during forming. The forming performance in the fixed blank was based mainly on the elongation of the material, and the forming depth was therefore limited. In the sliding blank method, corners of the blank were creased, and it was possible for the blank to slide. The creasing pattern was optimised to improve 3D-forming and package quality in the earlier development work of the used forming tool set. Forming was based on the folding of the blank, in addition to the stretching of the material, and this made the forming of deeper trays possible. This study employed both a fixed blank process and a sliding blank process with a high blank holding force, which combines the straining and sliding of the blank.

Fig. 4
figure 4

Packer 2020 forming device at LUT Packaging Technology

Full size image

Prior to press forming, the sample materials were preconditioned for 24 h at 80% relative humidity at 23 °C. For one forming tool (female mould), temperature 160 ºC was applied. The temperature for the male mould was 50 ºC. Blank holding forces of 4.7 kN and 0.8 kN were applied for the fixed blank and the sliding blank, respectively. The applied pressing force was 100 kN, the pressing velocity 40 mm/s and dwell time 500 ms.

Measurement of mechanical properties

Testing of the laboratory sheet samples took place at a temperature of 23 °C and at 50% relative humidity in accordance with ISO 187:1990. The thickness and apparent sheet density of the dry paper were measured using ISO 534:2011, and air-dry basis weight using ISO 536:2019. The thickness and density of the compacted papers were most likely influenced by the minor wrinkles at surfaces, which were not taken into account as density was determined. The samples with the different shrinkage were also strained to failure, using a strain rate of 100 mm/min in a Lloyd tensile tester (Ametek, Berwyn PA, USA) in accordance with ISO 1924-2:2008.

Measurement of shrinkage and relation to elongation

The amount of total shrinkage of the sheets was varied by using restrained and unrestrained drying procedures and in-plane compaction combined with unrestrained drying. Shrinkage of restrained and unrestrained dried sheets was measured by applying marker lines with known separations in the sheets. The elongation of shrunken paper in percentage units is presented according to Eq. 2, which is slightly modified from the earlier proposed relationship (Kouko and Retulainen 2018). The difference in elongation between the previous and present equations at 0–30% shrinkage increases from 0 to 1.5%, respectively.

$$Elongation=\frac{\varDelta S+{\varepsilon }_{Restrained}}{100-\varDelta S}$$
(2)

where ΔS is the shrinkage, i.e. the percentage length reduction of the sample, ‘100’ refers to the length of the sample before the shrinkage and εRestrained is the percentage elongation of a corresponding completely restrained dried sheet. It is noteworthy that Eq. 2 is not an empirical equation. It is not based on any specific data, but it can be used as a reference for any data provided that ΔS and εRestrained are known. The elongation obtained from Eq. 2 has a different reference length (‘100-ΔS’) than the shrinkage percentage ΔS, although the shrinkage and subsequent elongation could be exactly the same on an absolute length scale (e.g. a mm-scale). (Kouko and Retulainen 2018)

In case the shrinkage process of paper consists of several phases, an easy, reliable and straightforward method to determine total shrinkage for a sheet would be to measure the dimensions of the sheet edges before and after all the events that generate shrinkage. However, it is usually not possible to follow the edges, because sheets are usually trimmed or cut to smaller pieces. One method for measuring the development of shrinkage is to use marker lines or marker dots with known separations on the sheets. Separate shrinkage events can be connected for calculating (total) shrinkage by using Eq. 3.

$$\varDelta S=100\times \prod\nolimits_{n=1}^{p}\frac{{(100-S}_{n})}{100}=100\times \frac{\left(100-{S}_{1}\right)}{100}\times \frac{\left(100-{S}_{2}\right)}{100}\times \dots \times \frac{\left(100-{S}_{p}\right)}{100}$$
(3)

where Sn is the shrinkage of the sample in a single event that creates a certain length reduction, i.e. the percentage length reduction of the sample. Equation 3 means that shrinkages of separate events cannot be added together (one can confirm this by trying to add, e.g., three subsequent shrinkages of 50%), but the changed sheet lengths in percentages need to be multiplied in order to obtain total shrinkage. At low shrinkage levels, such as less than 10%, adding two subsequent shrinkages typically causes minor error compared to multiplied results, and that is very difficult to detect.

Results and discussion

Compaction conditions

The two most important parameters for in-plane compaction (in-plane forced shrinking) of bleached softwood kraft (BSK) pulp-based material are the moisture content of the sheet and the extent (intensity) of the in-plane compaction process. It was known beforehand that a moisture content of approximately 30% is close to the optimal condition for BSK-based sheet in-plane compaction (Vishtal and Retulainen 2014).

The following observations were summarised during the in-plane compactions at the Expanda laboratory compactor. An overly high solids content of sheets entering the compaction nip caused problems with wrinkles forming on the sheets. It resulted in poor compaction results due to the greater strength/stiffness of the fibre network. An overly low solids content resulted in slippage between the sheets and the rolls, which in turn resulted in low compaction. The operating window of solids contents for compaction was not significantly influenced between the different materials. Therefore, similar compaction results were obtained with the operating parameters (solids content, pressure and speed difference). The sheets were moisturized to about 65% solids content by spraying water on each sheet. An uneven moisture distribution from the spraying was evened out by stacking all the moist sheets on each other and letting them homogenize overnight. Latex was not homogeneously distributed in the samples containing latex and locally melted patches were formed during drying. However, influence of the melted polymer patches on the compaction result was minor. During start-up of the Expanda nip, the rolls were in contact with each other without a sheet in between. This resulted in wear and roughening of the rubber. The imprint of the rougher rubber surface was transferred to sheets leading to a rough surface. However, this did not influence compaction and the overall mechanical properties of the compacted material.

Comparison between foam- and water-laid formed materials

The influence of water-laid forming and foam-laid forming on the extensibility and tensile strength of paper was studied using 100% HC and LC refined BSK. In previous studies, HC refining enhanced the elongation in the case of water forming compared to LC refining (Zeng et al. 2013; Vishtal 2015). Thus, in this study, the comparison was made with material that has been shown to work well with water-laid forming. The main difference between water- and foam-laid fibre structures is in formation (Hjelt et al. 2021). With foam-laid forming more homogeneous distribution of fibres and excellent formation can be achieved compared to water-laid forming (Lehmonen et al. 2019). Another difference in structure is on pore size distribution. Foam-formed structure features many more large pores. The difference increases when stiffer fibres are used. The foam also has an effect at the fibre level. It can make fibres curlier because they are wrapped around foam bubbles.

Figure 5 presents elongation and tensile index (in this study this is calculated using the basis weight of the compacted samples) for the foam-formed and water-formed in-plane compacted samples. The pilot materials were only water-laid formed and made from a different BSK pulp batch using different HC and LC-refiners compared to the laboratory samples, thus those pilot materials were used as a reference (Kouko and Hjelt under review). The laboratory samples were prepared 100% HC and LC-refined BSK using both water and foam-laid forming. A full comparison of in-plane compacted samples’ tensile properties can be made using only the CD as the direction of the in-plane compaction and tensile test, because in-plane compaction of the water-laid sample in the MD direction was unexpectedly not successful. Reason for the unsuccessful result may have been a mistake in the direction of in-plane compaction or tensile test, that could not be fixed afterwards due to the limited number of samples.

Fig. 5
figure 5

Average elongation and tensile index with 95% confidence intervals of the foam- and water-formed in-plane compacted sheets

Full size image

The shrinkages for the four dried foam- and water-laid laboratory samples presented in Fig. 5 in the order of appearance were 21.4%, 21.7%, 19.2% and 18.7%. The foam-laid samples had shrinkages of 2–3% points greater. The shrinkage of foam- and water-laid CD laboratory samples after the in-plane compaction using the Expanda and drying were 21.4% and 19.2%, respectively. The foam-formed sample had greater elongation at 27.4% (compared to 22.1%), whereas the water-laid sheet had a higher tensile index of 33.5 Nm/g (compared to 25.4 Nm/g). The observed 2.2% percentage-point difference in the CD shrinkage was clearly smaller than the difference in the CD elongation between the foam- and water-laid laboratory samples. Elongation of the water-laid pilot sample in the CD after compaction in the same direction was 22.3%, i.e., similar to the laboratory sample, but the tensile index was lower, at 16.3 Nm/g compared to 33.5 Nm/g. These results indicate that in this study the foam-forming produced slightly greater elongation compared to water-laid forming, whereas the greater tensile strength was reached with the water-laid fibre networks. The shrinkages of the pilot prepared samples were not measured.

Comparison of one-way and two-way compaction

Materials need to elongate (expand) in press forming events, and not only in the machine direction (MD). Therefore, one would assume that the maximal performance of a BSK material in 3D forming could be reached after in-plane compaction in both MD and CD directions. However, performance of industrial in-plane compaction technologies for board web production are currently more advanced in the MD. In this study, the influence of consecutive CD and MD (two-way) in-plane compactions was investigated and compared to an MD (one-way) in-plane compaction. The samples were two-way in-plane compacted at first in the MD and then in the CD and the one-way compaction was performed in the MD.

The Billerud laboratory compactor had sufficiently large dimensions for a sheet of 120 mm x 120 mm, which enabled a reasonable sample size for laboratory tensile and press-forming tests. The sheets were water-laid pilot paper samples using 100% HC and LC-refined BSK as a raw material, as shown in Fig. 5. Figure 6a and b presents elongation and tensile index for the two-way and one-way in-plane compacted pilot samples. Figures 5 and 6 present the same water laid pilot material from two different in-plane compaction, which explain the difference in results. The in-plane compaction in the CD after the MD compaction increased the maximum elongation in the CD from 5.7 to 15.0% and decreased elongation in the MD from 22.3 to 16.3%. However, the (arithmetic) means of the MD and CD elongations for the two-way and one-way compacted cases were almost equal, at 15.6% and 14.0%, respectively. Furthermore, the averages of the MD and CD tensile indexes for the two-way and one-way compacted cases were almost equal at 25.6 Nm/g and 28.5 Nm/g, respectively.

Fig. 6
figure 6

a Average elongation and tensile index of the two-way and one-way in-plane compacted sheets made of 100% BSK with 95% confidence intervals. b Tensile-elongation curves of the one-way and two-way in-plane compacted samples with the references

Full size image

The one-way compacted tensile curve in the MD and the two-way compacted in the CD showed a concave upward shape, shown in Fig. 6b, which is characteristic of compacted materials tested in the direction of compaction (Welsh 1966). Also, compacted individual pulp fibres usually have similar characteristic shapes of tensile curves (Dumbleton 1971). However, the double compaction seemed to remove the inflection point in the MD tensile curve. The light blue curve in Fig. 6b looks like ‘normal’ paper with high elongation. Compaction decreased the tensile index despite the compaction direction, but MD one-way compaction did not influence the CD elongation (Fig. 6b).

The 3D forming was done at LUT using the fixed blank method with the MiniMould tool (Tanninen et al. 2017a). Figure 7a and b present the 3D formed samples of the two-way and one-way in-plane compacted sheets. Despite the two-way and one-way compacted samples having clearly different elongation in MD and CD, the samples had similar average elongation, and both performed similarly in tray forming, having 15.5 mm and 16 mm depths, respectively. This result indicates that materials that have similar average MD and CD elongation can perform similarly in press forming. On the other hand, the results indicate that tensile strength has a role in 3D forming performance, due to the lower average elongation and greater tensile strength of one-way in-plane compacted sample leading to the greater tray depth. Under industrial production conditions, it is currently technically significantly easier to create MD in-plane compaction compared to CD in-plane compaction. This result indicates that reasonable performance for BSK-based materials for 3D forming applications can be reached by applying only one-way in-plane compaction.

Fig. 7
figure 7

The formed trays for (a) two-way (depth 15.5 mm) and (b) one-way (depth 16.0 mm) in-plane compacted samples. The sheets were water-laid applying 100% BSK using the LUT MiniMould tool

Full size image

Role of HC refining in foam-forming

HC refining followed by LC refining has been found to produce relatively good elongation of paper with reasonable water retention when water forming is used (Zeng et al. 2013). In this study, the combination of HC and LC refining of BSK was compared to LC refining in foam-laid forming, additional chemicals and fibres and in-plane sheet compaction. The samples were two-way in-plane compacted at first in the CD and then in the MD. Elongation for the investigated recipes is shown in Fig. 8; Tables 2 and 3. Foamable latex was used due to its foaming and fibre network bonding ability. PVA fibres of 4 mm in length were used due to the hypothesis of long fibre ability to improve stress distribution. The role of the additives will be studied in the next subsection.

Fig. 8
figure 8

Average elongation with 95% confidence intervals for three different formulations using foam-laid LC-refined and HC + LC-refined BSK laboratory sheets. Tensile properties of unrestrained dried samples were measured in the MD. In-plane compaction of the materials was done in first in the CD then in the MD. Tensile properties were measured in the original CD of the sheets. The used additives are presented in the labels

Full size image
Table 2 Shrinkage, average elongation, tensile index and tensile energy absorption (TEA) index with 95% confidence intervals in the MD and, density of the foam-laid BSK samples after unrestrained drying
Full size table
Table 3 Shrinkage, average elongation, tensile index and tensile energy absorption (TEA) index with 95% confidence intervals, density, and 3D form depth for the materials
Full size table

The results show that HC refining improved elongation of the foam-laid formed unrestrained dried sheets (Fig. 8a), but it did not increase elongation of the foam-laid formed in-plane compacted sheets (Fig. 8b). The tensile index of both LC and HC- and LC-refined sheets were similar, despite the fact that LC-refined foam-laid sheets typically had remarkably lower densities compared to HC and LC-refined foam-laid sheets (see Tables 2 and 3). In addition, the water-laid HC and LC-refined reference had clearly lower density compared to the foam-laid sheets (see Table 3). The results indicate that relatively high values of elongation can be obtained without HC refining. Additionally, this means that rather high-cost HC refining is not needed in foam-laid forming.

The in-plane compaction phase was carried out at the Billerud laboratory, employing the laboratory Expanda compaction device. The best performing samples presented in Table 3 reached elongation greater than 20% after the in-plane compaction. The influence of PVA fibres and latex on the shrinkage and elongation was not consistent, but in general the results seem to indicate that they improved the chance of success in in-plane compaction.

Role of PVA fibres and latex on extensibility

The hypothesis before the study was that the combination of thermoplastic binder chemical and certain long man-made fibres would increase elongation. In this study, foamable latex was used as a thermoplastic component with a relatively high dosage 40 g/L. Our previous studies have shown that in water-laid forming, poor retention is reached for a latex mixed with cellulose. However, reasonable retention for latex can be reached as it is used as a foaming agent in foam-forming with cellulose fibres. In this study, latex retention was not directly measured, but based on the increase in sheet weight due to latex addition, the amount of latex in dried sheets was around the targeted 10%.

The increased retention of latex is due to the location of latex polymers at foam. The foaming agent (latex) is shown to locate at the surface of the foam bubbles, because it stabilises at the water-air interfaces. The average foam bubble diameter (~ 200 μm) is larger than the pore throats in the wet cardboard structure. For that reason, foam bubbles will have to squeeze through the pore throats. When this happens, the latex polymers positioned at the bubble surface are forced to contact fibres and have the possibility to attach to the fibre surface. Thus, foam-forming enables the usage of the latex as a thermoplastic component.

In water-laid forming, 4 mm-long artificial fibres start to deteriorate the formation of a dried cardboard sheet (Kerekes and Schell 1995). However, in foam-laid forming up to 20 mm-long artificial fibres have been used without having a significant influence on sheet formation (Hjelt et al. 2021). In this study, the PVA fibre 4 mm in length was chosen as a thermoplastic fibre after a comparison with three other long fibre types, namely Lyocell, alkali-treated Lyocell and EVOH. The greatest elongation after unrestrained drying was reached with the PVA fibre-containing sheets. The PVA fibre content influence on the elongation with and without latex was tested applying for BSK pulp and PVA fibres with mixing ratios of 90:10, 80:20 and 70:30, respectively. The best elongation before compaction was at the mixing ratios 80:20 for the PVA fibres with SDS and 90:10 with the latex.

Figure 9 presents the elongation of foam-laid LC-refined and HC + LC-refined BSK pulps combined with latex and PVA fibres, and the water-laid and SDS foam-laid samples as a reference. The results show that the latex alone and PVA fibres in combination with the latex increased elongation compared to the SDS foam-laid samples. High elongation of a foam-laid sample after unrestrained drying (Fig. 9a) also seemed to indicate high elongation after in-plane compaction (Fig. 9b). The samples were two-way in-plane compacted at first in the CD and then in the MD. Tensile indexes of the latex containing HC + LC-refined samples were clearly higher compared to the water-laid and SDS foam-laid samples.

Fig. 9
figure 9

Average elongation with 95% confidence intervals of LC-refined and HC + LC-refined BSK laboratory sheets. Tensile properties of unrestrained dried samples were measured in the MD. In-plane compaction of the materials was done first in the CD and then in the MD. Tensile properties were measured in the original CD of the sheets. The used additives are presented in the labels. Water-laid samples had random fibre orientation

Full size image

Interestingly, the water-laid samples had similar elongation compared to the best foam-laid samples after unrestrained drying (Fig. 9a), but after the in-plane compaction, elongation of all the foam-laid samples was greater (Fig. 9b). This result indicates that when compared to the water-laid samples, the foam-laid samples were more suitable for in-plane compaction targeting high elongation.

Relationship between in-plane shrinkage and elongation

Elongation of all the laboratory samples presented in Figs. 8 and 9 are presented in Fig. 10 as a function of shrinkage.

Fig. 10
figure 10

Elongation as a function of shrinkage for the studied foam- and water-laid SW BSK laboratory samples after unrestrained drying (red squares) and after in-plane compaction (blue circles)

Full size image

Relation between shrinkage and elongation can be random at relatively low shrinkage levels below 10% as can be seen in Fig. 10. Elongation of paper can be a complex function of several almost equally strong sources of factors that can be created by e.g., pulp refining, chemical additives, wet pressing, shrinkage. However, the higher the shrinkage the stronger influence shrinkage has on the elongation in condition that the material strength is sufficient as can be seen in Fig. 10. Mathematically, the result means that percentagewise, the elongation is greater than the shrinkage due to different reference points. Experimentally this means that the same dimensional contraction brought by shrinkage can be approximately strained out in tensile testing. Variation of the elongation between the samples at about the same shrinkage is complex and cannot be explained based on the presented results. However, based on the measurements that are not included in this work, there are strong indications that samples having a tensile index below 20 Nm/g likely do not reach the theoretical elongation as a function of shrinkage. Additionally, in Fig. 10, next to the group of red squares representing the unrestrained dried samples, four blue circles represent unsuccessfully in-plane compacted samples. That clearly shows the importance of a successful in-plane compaction event, which is a condition for high elongations.

Material performance at press-forming

Samples were press-formed using the MiniMould tool (90 mm × 80 mm × 35 mm). The forming conditions were as follows: forming temperature of the female mould was 160 ºC with pre-conditioning of the samples at 80% RH for 24 h. The samples were press-formed using fixed blank and sliding blank forming. The objective was to evaluate performance of the materials by comparing the maximum depths of the trays after fixed blank forming, as shown in Fig. 11, which is mostly influenced by the material strength properties (elongation and strength), but usually severe wrinkles do not appear and decrease sample quality. Sliding blank forming was performed with a low blank holding force, which caused friction of material between mould surfaces and created wrinkles, which play a significant role in surface smoothness and limit the maximum depth as a result. On the other hand, there may be some important press-forming condition, such as temperature-induced adhesion or friction that was not taken into account in the material measurements. The role of friction can be expected to be small in fixed blank forming, because less sliding occurs between metal-paperboard contact (Tanninen et al. 2017a). Change in blank sliding restraining ability caused by a change in temperature or other factors can be compensated by adjusting the blank holding force (Tanninen et al. 2017b).

Fig. 11
figure 11

Tray depths after fixed blank press forming for the LC-refined and HC + LC-refined BSK pulp sheets. Statistical intervals cannot be presented due to the low number of repeats

Full size image

The uniform surface quality of the mould has been found to harmonize the forming event in terms of forming force magnitude and direction when higher temperatures are used (Tanninen et al. 2017b). The influence of all the parameters in the sliding blank process on the depth of tray was not in the scope the study and cannot be evaluated based on the results in this study.

Figure 11 presents tray depths after fixed blank press-forming for the same LC-refined and HC + LC-refined BSK pulp sheets samples, as presented in the previous section. The samples were two-way in-plane compacted at first in the CD and then in the MD. The LC-refined pulp seems to perform better compared to HC + LC-refined pulp in case of foam-forming. The combination of the latex and PVA fibres led to the deepest trays. The results indicate that, even though strength was important in fixed blank press forming, it did not always correlate with elongation. The tensile index of HC + LC-refined pulp was higher on average compared to the LC-refined samples (as shown in Table 3). On the other hand, the tray depths shown in Table 3 indicate the opposite order of performance. The results indicate that strength needs to be significant enough for good formability. However, the superior strength of a BSK pulp-based material may not necessarily improve its 3D forming performance.

Figure 12 shows examples of fixed blank-formed (90% HC + LC-refined BSK, 10% PVA fibre and latex as surfactant) and sliding blank-formed (80% HC + LC BSK, 20% PVA fibre and SDS as surfactant) samples. Figure 13 shows the fixed and sliding blank-formed 90% HC + LC BSK, 10% PVA fibre and latex tray samples with depths of 18 mm and 35 mm, respectively. Tray depths for all the 3D formed materials are presented in Table 3. The curved blue and red lines drawn on the sheets before 3D forming indicate biaxial elongation and complexity of the 3D forming event.

Fig. 12
figure 12

On the left, a fixed blank-formed sample (90% HC + LC BSK, 10% PVA fibre and latex) with a depth 18 mm and on the right, a sliding blank-formed sample (80% HC + LC BSK, 20% PVA fibre and SDS) with a depth 35 mm. 30 mm x 30 mm squares were marked after sheet forming (before the in-plane compaction) Blue: in the MD of the original laboratory sheet, Red: In the CD of the original laboratory sheet

Full size image
Fig. 13
figure 13

Fixed and sliding blank-formed 90% HC + LC BSK, 10% PVA fibre and latex tray samples with depths 18 mm and 35 mm, respectively. 30 mm x 30 mm squares were marked after sheet-forming (before in-plane compaction) Blue: in the MD of the original laboratory sheet, Red: In the CD of the original laboratory sheet

Full size image

The results shown in Table 3 indicate that the relationship between material properties and the depth of a press-formed 3D shape (tray) may be sensitive and complex. The depths of the 3D shapes were not fully explained by the measured material properties (elongation, tensile index, tensile energy absorption (TEA) index (energy to break index) and density). In particular, the depths of the HC + LC BSK-based 3D shapes were slightly inconsistent with the elongation and tensile strength results. The presented results also indicate that even with the TEA index taking into account elongation and strength of a material may not always consistently predict press forming performance. On the other hand, there may be some important press-forming condition, such as temperature-induced adhesion or friction that was not taken into account in the material measurements.

In general, the results presented in Table 3 indicated that PVA fibres and latex together increased the maximum depth of 3D-formed shapes. The application of PVA fibres either with or without latex seemed to improve the strength of the material, which has a role in the 3D forming performance. However, the 3D shape depths of the fixed blank formed samples containing LC-refined BSK and LC + HC-refined BSK were not fully in line and did not consistently indicate the benefit of PVA fibre and latex. On the other hand, the depths of the sliding blank-formed HC + LC BSK samples indicated the benefit of application of the PVA fibres and latex in the sheets using foam-forming.

Conclusions

The scope of the work was the applicability of extensible paperboard material on press-forming conditions developed for plastics.

HC + LC-refined water-laid BSK performed very well after unrestrained drying. However, in foam-laid forming, HC refining did not improve the results. Foam-laid formed LC refined BSK with latex and PVA fibres led to high elongation, and additionally strength was improved compared to similar water-laid material. The results indicated that the concept containing refined BSK and thermoplastic binder chemicals and fibres can produce greater extensibility compared to any water-laid forming concepts. On the other hand, in this study no evidence was found for a fibre network with poor bonding being able to perform in an acceptable manner.

In-plane compaction had the greatest increasing influence on elongation. On the other hand, in-plane compaction had a decreasing influence on the strength of the material in the direction of compaction. In-plane compaction increased basis weight and seemed to increase the density of the materials. The results strongly indicated that the foam-laid material composition containing refined BSK and thermoplastic binder chemicals and fibres improved press-forming performance compared to the foam- and water-laid material compositions without the thermoplastic components. The deepest 3D shapes were obtained with the material recipes containing refined BSK and thermoplastic binder chemicals (latex) and fibres (PVA).

The results indicated that materials that have similar average MD and CD elongation can perform similarly in 3D forming. In industrial production conditions it is currently technically significantly easier to create MD in-plane compaction than CD in-plane compaction. This result indicates that reasonable performance for BSK-based materials for 3D forming applications can be reached by applying only one-way in-plane compaction. On the other hand, the results indicate that tensile strength has a role in 3D-forming performance, due to the fact that lower average elongation and greater tensile strength of the one-way in-plane compacted sample led to greater tray depth.

Most of the experimental trial points aligned very well with the theoretical relationship between shrinkage and elongation. The series of subsequent shrinkages must be connected by multiplying rather than adding the shrinkage values, in order to obtain the occurred shrinkage.