My team and I participated in two factory trials using a bioresin during a project which attempted to reduce formaldehyde emissions during manufacture and use of wood based panels. One was conducted in a particleboard factory in which the bioresin was used as a partial substitute for the existing urea formaldehyde resin; the other was conducted in a plywood factory where the bioresin was used in a pure form to make interior grade plywood (see Fig 1).

The first trial was conducted in the particleboard factory. The bioresin was quite viscous, so there was a delay of about an hour at the start of the trial while the resin was added into one of the dosing tanks of the resin system.

Once in the system, the bioresin mixed well with the standard resin and seemed to pass through the resin blenders without any difficulty. The resultant mattresses coming out of the former looked no different to those made with the standard resin (see Fig 2).

Nothing out of the ordinary seemed to be happening during the pressing cycle, but there was a heart-stopping moment (for me at least) when the first panel blew spectacularly and stuck to the top platen of the hot press (see Fig 3). Fortunately, it fell down of its own accord a few seconds after the press was fully open. The following two panels also blew, but without sticking, and then normal production was achieved and continued without major incident.

Laboratory samples were cut from the panels and were tested by both the company and ourselves. As usual, the properties of the panels made with a commercial production line were much better, by about 50%, than those observed from previously made laboratory panels. This is often observed because:

– the resin distribution achieved in commercial blenders is often better

– the mattress is formed by machine, thus

permitting particle size grading within layers – the press closes much faster than our

laboratory press and this helps improve the density profile

– the panel is much bigger, so the vapour pressure within the panel is higher and this raises the temperature inside the panel. Although the mechanical properties of the panels were good, and their formaldehyde emission was very low, their dimensional stability was slightly impaired. There was not enough time left within the project to see if the dimensional stability could be improved by, for example, increasing the wax content slightly.

The factory trial at the plywood factory passed rather more smoothly and without any worrying moments. The adhesive is, of course, completely inside the panel so there were no press sticking problems. Once again the adhesive was rather thick, which presented us with some difficulty putting the resin into roller-coaters. The high viscosity is unlikely to be a long-term problem because there are both physical and technological solutions to this.

Resin tack is not a particular issue at this factory because the production process does not use a pre-pressing step; the panels are laid up manually, as is often the case in plywood manufacture, and then fed in to the loading cage just in front of the hot press. The daylights of the hot press are loaded simultaneously via the loading cage.

A conventional pressing cycle was used, although the first two press loads were pressed for slightly longer than usual. When these left the press and seemed to conform to standard panels, the press cycle was reduced to the normal length.

Once again the mechanical properties were satisfactory for interior grade panels and, of course, the formaldehyde emissions were negligible.

Sample panels were cut and tested by the plywood factory, a laboratory associated with the resin supplier, and ourselves.

The main test used was the glue bond quality test as described by EN 314, which is a shear test of the glue bond between veneers. When the results were compared, it was apparent that two sets of results had very similar values while the third had much lower ones.

On further investigation it was found that this is caused by the presence of micro-cracks in the central veneers; all veneers have these cracks as they are generated during peeling.

The orientation of the cracks relative to the tension force in the sample has a marked effect on the observed tensile stress. Fig 6 shows an example of a sample in which the micro-cracks are being forced open by the tension forces applied during the test: the micro-cracks in the central veneer of an identical specimen, but with the grooves cut on opposite faces, would not be forced open so much and so the tensile force needed to break the specimen is greater.

The strength difference can be as much as 50%, which is surprising given that the samples can be cut from the same zones of the original panel.

The North American standard, ASTM D906, requires that specimens are cut so that both the best-case and worst-case (Figure 6) scenarios are tested, whereas the European equivalent, EN 314, does not.

This may be because, historically, North America has tended to use thicker veneers than Europe and the presence of micro-cracks is more prominent in thicker veneers.

This experience has caused us, at ESB, to change our test protocol and so now we always cut specimens so that the best- and worst-case specimens are tested.

This factory trial therefore gave an unexpected benefit to us as it improved our understanding of what happens during shear, tensile glue bond testing, which has led to an improvement in our test protocols.

To my knowledge, neither company is currently using the bioresin used in the trials. This is as much to do with the economics of using this resin as to it is to do with the technological implications.

Both companies now have, should they choose to use it, a bioresin-based product line to offer to their clients. I wonder how long it will be before I am allowed out of my laboratory again?