![[BioMatNet Database - FAIR Program]](../images/Fair.gif) |
FAIR-CT97-3781
Improving the performance and adhesion properties of wood-coatings with the use of flame ionisation technology (FLAME) |
 |
Contract No: | FAIR-CT97-3781 |
| Date Prepared: | April 2000 | |
| Source: | Progress Report |
Introduction
The aim of this project programme is to improve the long-term performance of factory finished joinery components for exterior use using flame ionisation technology. Within this overall aim there are a number of specific objectives related various stages within the chain from "raw-material to market", as follows:
To develop a technique optimised for specific timber species using existing flame- ionisation technology in order to improve the wetability and adhesion properties of wood-coatings to softwood and hardwood species of commercial importance.
To assess the influence of substrate variables in order to establish optimised "process templates" for individual species.
To relate changes in the surface properties of solid wood substrates after flame ionisation, with their effects on long-term (exposure) performance of wood-coatings.
To verify and quantify the beneficial effects of the treatment on a range of commercially important softwood and hardwood species of timber.
To assess the commercial viability of the technique as part of an in-line factory process.
Activities
The main experimental programme consists of five work tasks as follows:
Task 1 Establishment of Main Processing variables
Task 2: Species specific surface activation
Task 3: Assessment of long and short-term adhesion properties of coated flame activated surfaces
Task 4: Long term weathering and QUV trials
Task 5 Feasibility study of the use of flame ionisation as part of an industrial in-line process
Results
Task 1.1 examined the influence of processing and substrate variables on the effectiveness of the flame plasma treatment process with the different commercial timber species selected for examination. The investigations centred on morphological and anatomical properties common to all species of wood. No attempt was made to assess individual species variables from the standpoint of chemical constitution as this aspect was addressed in later tasks. In addition processing influences as a result of primary and secondary conversion as well as storage conditions were evaluated. Wood variables common to all species were examined, to establish the influence of the following:
The, work that was carried out on oak, pine and meranti, was divided amongst the participants. The selection of suitable species provoked a great deal of debate amongst the project consortium and differ from those originally scheduled within the Technical Annex to the project. The reasons for these changes related to the specific variables that were examined. The partners considered that, wherever possible it was important to use the same species throughout this task in order to avoid species effects. Eventually the partnership agreed on the use of meranti and pine as the most useful species in view of their respective differences from the standpoint of surface texture, porosity and density. Furthermore both were commercially important in the context of window and door joinery throughout Europe. The only exception to this was the selection of oak and pine in determining the influence of heartwood/sapwood disposition on the plasma treatment process. This was done due to overcome the recognised difficulty in visually discriminating between heartwood and sapwood in meranti. Oak was therefore substituted only for this analysis.
In addition to the above changes in the test protocol, the partners agreed that their seemed little merit in including pH as one of the variables influencing the response of timbers to flame ionisation treatment. This was done on the grounds that no known scientific mechanism or argument attributable to pH could be identified which could interfere with the process of plasma treatment. The parameter would also have been exceptionally difficult to investigate and the group did not consider the disproportionate expenditure of additional manpower required to assess this parameter as justified. A number of minor changes to the schedule were made in order streamline the research activities.
Further work covered UV-vis measurements of surface colour changes associated with flame ionisation treatment. This work provided evidence of chemical changes on the surface of the test timbers, detected using IR spectroscopy. Colour changes were assessed on flame treated and untreated samples using a model which considers the respective contributions of three components namely; L(black/white), a (green/red) and b (blue/yellow). The colour change results showed only marginal colour differences at the surface of all test samples occurred and that these were in the greater part not visible to the naked eye. The greatest differences occurred in component L
which decreased (i.e. got darker) after flame treatment. Components a and b remained relatively unchanged subsequent to flame treatment for all species examined. The colour difference changed, subsequent to flame treatment, with all species due to a change in the L component but stabilised within a few days after treatment. Stabilisation took between one and ten days after treatment depending on the species tested.
Infrared absorption spectra were obtained for four species using an ATM microscope. This technique was adopted on account of its non-destructive nature. However, the absorption levels returned were low and the quality of the spectra was deemed poor. An alternative Traditional IR technique using KBr pellets was used instead. This technique assumes that the chemical differences imparted to wood surfaces by the flame treatment are homogenous and that the prepared pellets were similarly homogenous. IR spectra were measured at 0, 1, 2 3, 6, 9,13 and 21 days for all four test species considered.
The spectra from the KBr pellets showed that oak and pine were more affected by the flame process than meranti and spruce. This supports previous work carried out in other tasks which suggests that oak and pine are relatively easily treated species and that spruce and meranti were species where it was far more difficult to promote increases in surface wetting greater surface wetability using the flame ionisation process. Furthermore, the spectra were found to be unstable in that after a day they declined rapidly and thereafter progressively increased to levels comparable to post-treatment levels after approximately 21 days. This is again in keeping with some of the trends observed for pine with the long-term persistence data obtained using contact angle studies. These suggested that a certain level of improvement in surface wetting may occur several weeks after flame treatment and suggested an oxidative mechanism may be at work with one or more components present in some species.
For this task it was concluded that it is likely that a combined oxidative and/or degradative process may be occurring on the surface of flame treated wood. However, more specific information on the precise nature of these reactions is not possible without further work.
The objective of task 2.2 was to determine the "time window" during which surface coatings can be applied for each test species and examined before the beneficial effects of surface activation begins to decline. This approach was based on the assumption that by increasing the surface activity of the wood surface activation begin to decline, better wetability of applied surface coatings would result, which would ultimately promote better adhesion and long-term performance. To achieve it was necessary to establish a time window after flame treatment where the surface activity of the treated substrate was still at a level which promoted good wetability before a decline in the effect took place, for each of the test species of commercial importance. The work schedule of task 2.2 also incorporated IR studies in order to provide a fuller understanding of the mechanisms responsible for the enhancement in surface activity of individual wood species, and its subsequent decrease. Additionally, UV-vis studies were carried out to monitor changes in colour as a result of the treatment.
The trial was set up in order to asses the wetability of the wood species prior to and after flame ionisation treatment. The experimental protocol required that assessment and monitoring of the treated sample material be carried out for a period of three weeks after initial treatment during which time all samples were stored in controlled atmospheric conditions. The results returned from the work were extremely variable despite care being taken to ensure that all material for each species was directly cross-matched. The high variability in the findings prompted the entire trial to be repeated in an effort to increase the data and to reduce standard deviation in the results. The repeat trial again gave results which were considered to be inconsistent when viewed against the body of knowledge gained for these species in earlier work. In particular meranti showed relatively little or no improvements in surface wetability after treatment whereas pine exhibited only slight improvements. Spruce actually demonstrated a decrease in wetability after treatment. The only species to behave in a manner "consistent with expectation" was oak, which exhibited a strongly positive wetting response after flame treatment. Furthermore, the oak results showed good stability over a period of between 4 and 10 days, whereafter a strong decline in wetting occurred. The results confirmed previous experiences with oak, in that this is a species that reacts well to flame treatment in producing a strongly positive increase in wetting. The results for pine and meranti were inconclusive insofar as the initial responses to the flame treatment were inconsistent with those obtained from previous trials and underscored the "unpredictability" of these species with respect to their wetting behaviour after treatment.
Chemical identification and quantification of wood extractives contained in different wood species was considered important to the project objectives. This was relevant in so far as the proposed mechanisms for the decrease in wetability of wood species subsequent to machining cited by much of the earlier literature was attributed to the presumed movement of extractives to the freshly cut surface shortly after the machining process. The results obtained did, to a certain extent, verify these suggestions and support a possible correlation with the wetability results obtained both before and after flame ionisation treatment.
Given the complexity of organic compounds present within wood and the overall differences in the respective quantities of these compounds between different species, the task of chemically identifying individual or specific compounds was considered to be far too labour intensive to justify. Given the prohibitive costs, which would have inevitably resulted from such a programme of analysis, it was agreed that a simpler more cost-effective programme of indicative analyses would be just as useful to the project. As wetability is considered by many to be directly connected to the quantity and the polarity of the substances present on wood surfaces, the study was therefore principally aimed at directly revealing the quantity of polar and non-polar substances present in different wood species.
In general work under Task 2.1 demonstrated that process optimisation for individual species to bring about improvements in wetability are possible. However, the drawbacks are:
a) that the number of passes involved (approx. 24 or more) may be non-viable in a commercial context requiring an excessive amount of handling operations.
b) each timber species has a specific responsiveness to flame treatment which prevents "cross-species" generalisations from being made.
However, despite this the current status of knowledge concerning the relationships between the energy output of the burner and the pass number suggest that it may yet be possible to use this as a basis for a technique to predict optimum treatment. This conclusion is drawn from indications that in species exhibiting good early responsiveness to treatment (e.g. oak) the relationships of the set-up parameters required to effect good surface wetting are not as complex as with other, more demanding species. One of the main concerns from the research work carried out so far is that optimisation of the flame treatment parameters may involve high numbers of passes which ultimately may discourage take-up by industry. A possible way around this problem may be to develop burner head arrangements for timber that involve multiple burner heads spaced at regular intervals. This may decrease the handling requirements of components in a commercial situation although the solution will not be without a cost penalty in terms of initial capital outlay.
In general, whilst some of the variables examined produced demonstrable differences in wetability values, most of these effects were either relatively insubstantial or strongly dependent on wood species. With respect to heartwood / sapwood differences, opposite trends were found for the two wood species studied, one being the hardwood oak and one the softwood pine. Furthermore, it was concluded that wood density alone had little effect on the outcome of flame ionisation treatment. The hardwood meranti showed better results at the higher density, but for the softwood pine no consistent pattern was revealed. The effect of moisture content was found to vary according to the moisture regime. It was found that by storing samples in a mid-range condition of atmospheric humidity (66% RH), greater improvements in wetability as a result of flame ionisation treatment could be obtained.
Conclusions
Briefly stated, the influence of the substrate properties studied on flame treatment effect is variable and in some cases seems difficult to predict. However, it is much too early to conclude that the treatment does not have the robustness needed for industrial applications. Treatment schedules can be further optimised, and other factors like annual ring orientation and extractives taken into account.
It was concluded that storage temperature had greatest effect on the polar dispersive energy of the samples and that surface texture has the effect of increasing wetability, which is made more substantial by treatment with flame plasma. Wood density had little effect on the outcome of flame ionisation treatment. However, the effects of moisture content were found to vary according to the moisture regime. It was found that by storing samples in a mid-range condition of atmospheric humidity (66%), greater improvements in wetability as a result of flame ionisation treatment could be obtained.
Investigations into the effects of drying regime (i.e. kiln dried versus air-dried) on the effectiveness of flame ionisation on wood surfaces were hampered by the difficulty in obtaining air-dried samples of meranti and pine. The only species available in both kiln-dried forms was oak, which therefore constituted the bulk of the data generated for work involving this variable. The research concluded that the wetability of kiln- dried timber was better than that of air-dried stock. Furthermore, there exists an apparent depth gradient with kiln-dried timber, which does not occur with air-dried samples. Kiln dried material showed gradients where the contact angles decreased up to approximately 2mm depth after which no further differences were apparent.
The work therefore suggested that whereas surface preparation and storage temperature did not have a significant effect on the effectiveness of the flame plasma treatment, drying regime i.e. the manner in which timber is dried, did have an influence on the process. Some anomalous results were found and possible explanations for these have been suggested.
© Copyright 2006 Policy Statements
Updated
by CPL Press:
03/07/2007
- biomatnet@biomatnet.org
 |
 |
 |
News |
 |
Events |
|
|