EFFECT OF CYCLIC HUMIDITY ON CREEP BEHAVIOUR OF WOOD-BASED FURNITURE PANELS

Barbara Ożarska, Gerry Harris
School of Forest and Ecosystem Science, The University of Melbourne, Australia

ABSTRACT

A study was undertaken to determine creep characteristics of furniture panels, both unlaminated and laminated, in cyclic humidity conditions. Test specimens were prepared from laminated and unlaminated moisture resistant MDF. Two types of laminating materials were used: melamine laminate and hardwood (mountain ash Eucalyptus regnans) veneer. Specimens were subjected to four-point bending loads with a single stress level of 15% using vertical creep testing rigs. Relative humidity was set to be cycled between 35% and 85%, and the temperature was set at a constant level of 23°C. Centre-point deflection of the specimens was measured using Linear Variable Displacement Transducers.
The study revealed that the creep behaviour of MDF panels subjected to cyclic humidity can be significantly reduced by using surface laminations. The greatest reductions in relative creep values in bending of MDF panels (up to 3.3 times), were observed in the panels laminated with melamine surface overlays. The total relative creep values of the panels laminated with hardwood veneer were half those without any laminations. This finding will be useful in the design of wood-based furniture panels used in shelving and bottoms of cupboards.
Analysis of the creep deflection during the adsorption and desorption stages revealed that there was the difference in the performance of laminated and unlaminated panels. The creep deformation in the laminated panels increased during the adsorption phase and decreased during the desorption phase. In the unlaminated panels, reversal of the creep deflection during the cyclic humidity test was observed. A significant loss of both MOE and internal bond strength in evaluated furniture panels was observed after the creep test and the recovery period. It was evident that the use of laminations reduced these losses, with the highest reduction observed in the panels with melamine overlays followed by the hardwood veneer laminations.

Key words: creep, cyclic humidity, furniture panels, surface laminations.

INTRODUCTION

The growing use of wood-based materials for the construction of furniture panels is increasing the importance of a sound knowledge of their serviceability and life-time performance. Furniture panels are often used in applications where they are likely to be subjected to sustained loading for considerable periods of time, for example as furniture shelves and case bottoms. It is essential that the appearance and structural integrity of the panels is maintained in service.

The long term performance of furniture panels to a high extent depends on their creep characteristics. Creep data is essential in engineering design of furniture members in which excessive deflection is considered as a failure, and moisture content has a profound effect on the amount and rate of creep. Special allowances should be made for creep, therefore, when it is known that the shelves are to be used in humid environments or environments subjected to conditions of widely fluctuating humidity. Without performing actual tests, creep can only be estimated. A study was therefore undertaken to determine creep characteristics of furniture panels, both unlaminated and laminated, in cyclic humidity conditions.

A large number of studies have been carried out in various countries to assess the creep behaviour of wood-based panels, particularly particleboard, in changing environmental conditions. However, little information is available on creep and stability of these panels laminated with various overlays as used in furniture applications.

The creep characteristics of particleboard were studied by Dinwoodie et al. [11] at the Building Research Establishment, UK. The studies revealed that changing relative humidity above 65% results in higher creep deformation. However, an even greater effect was found if the relative humidity was cycled between two levels. It was observed that for moisture resistant particleboard, increasing the humidity from 30% to 90% caused an increase in deflection of four times, but that cycling the humidity between these two levels, caused a larger increase of over ten times after 20 weeks [10].

Comparative studies were carried out on various types of wood-based panels. It was observed that relative creep was greatest in MDF, followed by particleboard, plywood, glued laminated wood and solid wood. Relative creep in MDF was some five times that of solid wood, under relatively low loads and dry environmental conditions. However, Dinwoodie et al. [11] demonstrated that the relative order of creep changes for different materials as load levels or environmental conditions are altered.

The studies described above related to the creep behaviour of raw (unlaminated) wood-based panels, but only limited data has been published on the long-term performance of composite furniture panels. Chow [6] investigated the elastic and inelastic behaviour of veneered particleboard using a commercial brand of particleboard and walnut veneers of various thicknesses. He observed that a regular ¾ inch particleboard is inferior in creep resistance to solid walnut, being one-quarter as resistant as solid walnut. During a very damp summer on rainy days, the creep of a shelf, bookcase, or tabletop will be about two to four times more than would generally be observed in a normal humidity of 65%.

Chow [6,7] found that a combination of some hardwood-veneered particleboard or hardboard with a given level of shelling ratio (the ratio between the total thickness of veneer to the total thickness of the panel) and core density, can make an ideal creep-resistant structural panel.

In another study, Chow [8] reported that the bending strength of MDF overlaid with red oak veneers was greatly improved and the increase was higher than in overlaid veneered particleboard or hardboard panels. However, the tests were performed only at a constant relative humidity of 50% and at about 21°C temperature. A further study by Chow (1982) showed that relative humidity had a significant effect on the creep deformation of MDF boards laminated with sugar maple veneers.

Fernandez-Golfin and Diez Barra [13] studied creep behaviour of UF particleboards overlaid with melamine resin-impregnated paper (80 g/m²). The authors observed that the application of this type of overlay reduced relative creep and increased time to failure. Their further study (1998) investigated commercial MDF panels, raw and overlaid with melamine resin-impregnated paper, under alternating humidity conditions (30%-90%) and constant temperature 20°C, loaded at three different levels (20%, 30% and 40%). The relative creep of the MDF panels was considerably higher than in particleboards with similar characteristics. Creep behaviour of the raw panels was strongly influenced by melamine coating of the board’s surfaces but not by the edge coating, except when both were combined.

In Australia it is required that moisture resistant (MR) boards, as opposed to standard (UF) boards, are used for furniture which is likely to be used in high humidity conditions or in an environment with fluctuating humidity (e.g. kitchen and bathroom). The vast majority of moisture resistant furniture panels are made with melamine urea-formaldehyde (MUF) resin. For the purpose of this study the moisture resistant MUF MDF boards were used.

MATERIALS AND METHODS

Test specimens were prepared from laminated and unlaminated moisture resistant (MR) medium density fibreboard (MDF). Two types of laminating materials were used: melamine laminate and hardwood veneer. The wood veneer used was 0.4 mm thick mountain ash (Eucalyptus regnans F.Muell.). The samples were cut from commercially manufactured furniture panels purchased from a local supplier. The panels were typical of material produced for the furniture and joinery industry.

Two matching sets of samples were cut from three panels of each type. Set 1 consisted of samples of pairs of side-matched specimens which were used to determine the short term bending strength of the creep specimens. This allowed calculation of the load used in the creep test. Test specimens were cut to 50 mm x 390 mm, according to proposed European Test Method Pr EN(112) [12] for assessment of relative creep. Additional 50 mm x 50 mm specimens were prepared to determine internal bond strength (IB) of tested materials prior to the cyclic humidity test, as well as 20 mm x 200 mm stability specimens to monitor thickness swelling during the cyclic test. The specimens were prepared according to Australian/New Zealand Standard AS/NZS 4266.14:1996. Standard 100 mm x 100 mm specimens were used for testing moisture content during the test.

Set 2 consisted of samples prepared for testing internal bond and stability after the completion of the humidity cyclic test, to determine how the properties of the products were affected during the humidity cycling process. The samples were matched with equivalent samples of Set 1.

Before the commencement of the testing all specimens were placed in a humidity chamber at 65% relative humidity and 23°C until they reached equilibrium moisture content.

Three specimens of each type of panel were subjected to four-point bending loads with a loading span of 350 mm, which was approximately 19 x panel thickness. Vertical creep testing rigs were used for the loading, each capable of holding up to four samples (Fig.1). Preliminary creep tests were carried out to determine a suitable stress level to apply to the creep specimens. The draft standard EN 112 (1993) suggested that a single stress level of 25% of the short-term strength of side-matched specimens is used in testing creep of wood-based panels. However, the preliminary tests revealed that the deflections of MDF unlaminated specimens loaded at 25% stress level, as well as at 20%, were very high after 15 days of loading. For this reason, a single stress level of 15% was applied to the specimens in the study.

The specimens were exposed to changing conditions of relative humidity (rh) in a large “walk in” type computer controlled environmental chamber. Relative humidity in the chamber was set to be cycled between 35% and 85%, with the cycle started in the drier state (35%). A timer was fitted to the conditioning equipment to allow the conditions to be cycled between upper and lower set points on a predetermined time interval (i.e. 48 h). The temperature was set at a constant level of 23°C.

Centre-point deflection of the specimens was measured using Linear Variable Displacement Transducers (LVDT’S) with a resolution of +/-0.01 mm. The output from each transducer was monitored via a signal conditioning amplifier and recorded by a Data Taker DT505 data logger. A constant temperature instrumentation room was constructed to house the data logger and signal conditioning unit to obviate the effect of ambient temperature variation on the instrumentation. The deflection was monitored after 5, 10, 50, 100 and 500 min, and thereafter at 24 h intervals.

Constant measurements of thickness swelling were also taken using the LVDTs mounted in a specially designed rig (Fig. 2).

The duration of the creep test was 560 days. After removing the load, the creep specimens were reconditioned to equilibrium moisture content at a constant 65% rh at 23°C for four weeks and then destructively tested in bending to evaluate their residual MOE and MOR. The samples of Set 2 were tested for internal bond strength to determine the board deterioration during the cyclic test.

RESULTS AND DISCUSSION

Effect of cyclic humidity on creep characteristics
The results of the creep deflection measurements of the three furniture panels are presented in Figure 3. The curves for creep deflection-time and calculated average relative creep-time for each type of the panel are plotted in Figures 4 and 5. The relative creep R_c(t) (also known as the creep factor) is defined as the ratio of difference between the deflection (a_t) measured at time t and the instantaneous deflection (a_o): R_c(t) = (a_t - a_o)/ a_o)). The density of the MDF boards was 757 kg·m^-3.

From Figures 3 and 4, it is evident that creep resistance of unlaminated specimens exposed to cyclic humidity was much higher than of the specimens laminated with veneer and melamine surface layers. The maximum deflection of unlaminated MDF specimens was 19.26 mm, which was 137% higher than for veneered MDF specimens and 180% higher than for melamine laminated specimens. After creep recovery, the unlaminated specimens had the highest permanent deflection (15.17 mm) followed by the specimens laminated with veneer and melamine overlays (Table 1). Researchers had previously observed that changing relative humidity above 65% resulted in a higher creep deformation [10,11]. However, an even greater effect was found if the relative humidity was cycled between two levels. Dinwoodie et al. [10] observed that for moisture resistant particleboard, increasing the humidity from 30% to 90% caused an increase in deflection of four times, but the cycling humidity between these two levels, caused a larger increase of over 10 times after 20 weeks.

Analysis of the relative creep results, summarized in Table 1, showed a similar ranking order to that of creep deflection. The highest average value of relative creep (8.53) was observed in unlaminated MDF specimens followed by the veneered specimens (4.41) and the melamine overlaid specimens (2.56). Previous studies revealed that, depending on the load and environment history, relative creep values for wood-based panels subjected to changing conditions may be as high as 12 [17].

This study revealed that the relative creep of MDF panels can be reduced 3.3 times with the use of melamine overlays and about twice with the use of hardwood veneer laminations. Fernandez-Golfin and Diez Bara [13] observed that the relative creep of MDF panels loaded at 20% stress levels subjected to cyclic humidity (30%-90%) was radically reduced when melamine overlays were applied only on surfaces or on surfaces and edges. After 4368 h of loading the relative creep of MDF panels laminated with melamine surface overlays was half that of the raw panels. No comparable data was available on MDF panels laminated with hardwood veneers.

The improvement in the creep characteristics of the laminated panels can be explained by their slower response to the humidity changes due to the presence of the surface overlays. A previous authors study indicated that the use of laminations significantly slowed down the changes in the moisture content of MDF panels.

The analysis of the creep results showed that cyclic changes of relative humidity resulting in changes in the moisture content of furniture panels greatly increased their creep. Moisture adsorption and desorption, which resulted in both dimensional changes (linear and thickness swelling and shrinkage) and changes of stiffness, strength and internal bond, significantly affected the creep behaviour of the panels. The creep deflection during the adsorption and desorption stages was analyzed separately for each type of the furniture panels and presented in Figure 5. It was observed that there was a difference in the performance of laminated and unlaminated panels. In both types of laminated panels the creep deformation increased during the adsorption phase and decreased during the desorption phase (Fig. 5 d and Fig. 5 e). This pattern was observed during the whole cyclic humidity process. These results are in agreement with some previous studies of cyclic humidity on the creep deflection of wood-based panels. Bryan and Schniewind [5] and Halligan [15] found that, unlike wood, the increased deflection of particleboard panels occurred during adsorption of moisture. However, this finding differs from observations made by other researchers. For example, Armstrong and Kingston [2] found that the behaviour of particleboard exposed to cyclic humidity was similar to that of solid wood (the creep deformation increases during desorption of moisture). Similar results were obtained for hardboard beams by Armstrong and Grossman [1]. Lack of agreement on the findings by various researchers suggests that a further detailed study should be undertaken to clarify this matter.

The analysis of the creep deflection of the unlaminated panels unexpectedly revealed that, although the increasing deformation of the specimens was monitored during the first desorption (35% rh for 48 h), the largest increase in deflection was observed during the first adsorption (48-96 h from the commencement of the test) (Fig. 5 a). During the subsequent cycles, further increases in deflection occurred during adsorption of moisture with only small increases or even decreases in deflection during desorption of moisture. However, a reversal of this pattern was observed during the fifth week of the test when an increased creep deflection was recorded during the desorption of moisture (Fig. 5 b). This reversed behaviour was monitored until the final stage of the experiment (Fig. 5 c). It is presumed that the reversal pattern of the creep deflection is related to the reduction in the panel internal structure which caused significant losses in the stiffness and internal bond during the humidity cyclic process. A further detailed study was undertaken by the researchers to explain the mechanism of this process.

The reversal of the creep deflection during the cyclic humidity test was also observed by Fernandez-Golfin and Diez Barra [14] who found that in MDF boards the deformation increased during the adsorption phase until week 18 from the commencement of the experiments. From that time onwards the pattern was reversed, following one similar to that of solid wood (i.e. deformation increased during desorption phase).

Effect of cyclic humidity on MOE/MOR and internal bond
Comparisons of static bending MOE and MOR and internal bond of the panels before and after the cyclic humidity test are presented in Table 2 and Figures 6(a-c).

Australian Standard requirements (AS/NZS 1859.2: 1997) for MOE, MOR and IB for moisture resistant MDF in 65% rh are:

MOE = 2.8 GPa
MOR = 30 MPa
IB = 600 KPa

The comparison of the stiffness data of unlaminated and laminated MDF panels (before the commencement of the cyclic humidity test) revealed that the stiffness increased about 28% for the panels laminated with hardwood veneer, and about 7% for melamine laminations. No improvement of bending strength (MOR) was observed in MDF panels with the use of surface laminations. These results are consistent with a previous authors study). This may be explained by the mode of failure in bending, which for MDF is almost always in shear.

A significant reduction in bending MOE and internal bond was observed in all types of specimens subjected to the creep test and the recovery period. The highest loss of MOE was observed in unlaminated specimens (36%), followed by the veneered specimens (23%) and the specimens with melamine overlays (18%). No reduction in bending MOR was observed.

The analysis of the results revealed a significant decrease in internal bond (IB) values for all types of panels after the exposure to cyclic changes of relative humidity which suggests a permanent destruction of the bond from exposure to high humidity. These values were significantly below the required values specified by Australian/New Zealand standard (AS/NZS 1859.2:1997).

The results of this study showed that the loss of MOE and internal bond of furniture MDF panels exposed in cyclic humidity conditions is greater than in panels exposed in various constant humidities for long period of time.

CONCLUSIONS

The study revealed that the creep behaviour of MDF panels subjected to cyclic humidity can be significantly reduced by using surface laminations. The greatest reductions in relative creep values in bending of MDF panels (up to 3.3 times) were observed in the panels laminated with melamine surface overlays. The total relative creep values of the panels laminated with hardwood veneer were half those without any laminations. This finding will be useful in the design of wood-based furniture panels used in shelving and bottoms of cupboards.

The analysis of the creep deflection during the adsorption and desorption stages showed a difference in the performance of laminated and unlaminated panels. The creep deformation in the laminated panels increased during the adsorption phase and decreased during the desorption phase. This pattern was observed during the whole cyclic humidity process. In the unlaminated panels, the reversal of the creep deflection during the cyclic humidity test was observed. The deformation increased during the adsorption phase until week 5 from the commencement of the experiments. From that moment onwards the pattern was reversed, following one similar to that of solid wood (i.e. deformation increased during the desorption phase). A further detailed study has been undertaken by the researchers to explain the mechanism of this process.

A significant loss of MOE and internal bond of evaluated furniture panels was observed after the creep test and the recovery period. It was evident that the use of laminations reduced the loss with the highest reduction observed in the panels with melamine overlays followed by the hardwood veneer laminations.

REFERENCES

1. Armstrong L. D., Grossman, P.U.A., 1972. The behaviour of particleboard and hardboard beams during moisture cycling. Wood Sci. Technol. 6, 128-137.

2. Armstrong L. D., Kingston R.S.T., 1962. The effect of moisture content changes on the deformation of wood under stress. Aust. J. Appl. Sci. 13(4), 257-276.

3. Australian/New Zealand Standard, 1996. AS/NZS 426,14:1996. Reconstituted wood-based panels. Methods of test – Dimensional change associated with changes in relative humidity.

4. Australian/New Zealand Standard. 1997. AS/NZS 1859.2:1997. Reconstituted wood-based panels. Part 2: Medium density fibreboard (MDF).

5. Bryan E. L., Schniewind A.P., 1965. Strength and rheological properties of particleboard as affected by moisture content and sorption. For. Prod. J. 15(4), 143-148.

6. Chow P., 1970. The deflection of composite furniture panels under constant bending stress. For. Prod. J. 20(12), 44-50.

7. Chow P., 1979. Deflection in bending of birch-veneered wood-base composite shelving panels. For. Prod. J. 29(12), 39-40.

8. Chow P., 1980. Static bending of red oak-veneered medium-density fibreboard furniture panels. Int. J. Furn. Res. 2(3), 25-25.

9. Chow P., 1982. Bending creep behavior of Acer saccharum Marsh veneered medium-density fibreboard composite. Wood Sci. Technol. 16, 203-213.

10. Dinwoodie J. M., Pierce C.B., Paxton B.H., 1990. Creep research on particle board. Holz a. Roh-u. Werkst. 48, 5-10.

11. Dinwoodie J. M., Higgins J-A., Robson D.J., Paxton B.H., 1992. Creep in chipboard. Part 11: The effect of cyclic changes in moisture content and temperature on the creep behaviour of a range of boards at different levels of stressing. Wood Sci. Technol. 26, 429-448.

12. European Test Method (Draft), 1993. Pr EN (112). Determination of duration of load and creep factors for wood based panels. Building Research Establishment UK.

13. Fernandez-Golfin J. I., Diez Barra M. R., 1992. Langzeitverhalten von Spanplatten unter vatiablen Feuchtebedingungen. Untersuchung des Einflusses von Beschichtungen und Klebstoffen. Holzforsch. Holzverweret. 44(6), 106-110.

14. Fernandez-Golfin J. I., Diez Barra M. R., 1998. Long-term deformation of MDF panels under alternating humidity conditions. Wood Sci. Technol. 32, 33-41.

15. Halligan A.F., 1972. Effect of moisture on physical and creep properties of particleboard. For. Prod. J. 22(4), 41-48.

16. Haygren J., Hall H., Yang K. N., Sawicki R., 1975. Studies of flexural creep behavior in particleboard under changing humidity conditions. Wood Fiber. 7(2), 74-90.

17. Morlier P., 1994. Creep in timber structures. Rep. of RILEM Technical Committee 112-TSC No. 8.

Accepted for print: 28.08.2007

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.