Content uploaded by Morwenna Spear
Author content
All content in this area was uploaded by Morwenna Spear on Oct 12, 2015
Content may be subject to copyright.
98
PHYSICAL PROPERTIES OF UK GROWN LARCH
SUBJECTED TO MILD AND MODERATE THERMAL
MODIFICATION PROCESSES
Spear M.J.1, Binding T.2, Nicholls J.1, Jenkins D.2, Ormondroyd G.A.1
ABSTRACT
Home grown Japanese larch (Larix kaempferi) is a strong attractive species with potential
for diversifying softwood timber consumption in the UK. The fast growth rate due to the
mild British climate presents some challenges to its use in high value applications, however
mild and moderate thermal modification systems have been devised which have been
shown to improve machinability and working properties of the timber. The thermal
modification process developed in Wales has been tested in the production of cladding and
joinery products. The timber from treatment runs with durations at the target treatment
temperature ranging from 1 hours to 5 hours have been compared. The density, colour and
surface quality of planed surfaces are presented in this paper, and the influence of the
different thermal regimes on the level of modification are discussed. The mild thermal
modification shows potential for interior applications, and eradicates many issues
commonly associated with UK grown larch such as resin pockets and springback of the
latewood after machining.
Key words: Mild thermal modification; treatment conditions, density, colour.
INTRODUCTION
A thermal modification using conditions which are considerably milder than conventional
commercial processes has been developed and trialled in Wales. The primary objective is
the altered machining properties of this mild thermally treated timber. The system has been
tested on two fast growing softwood species, Japanese larch (Larix kaempferi) and Western
hemlock (Tsuga heterophylla). When grown in the western side of the UK, the mild
maritime climate favours high yield classes, however these timbers are less favoured in the
market. Wales has a large number of small and medium sized enterprises working with
timber, both at the sawmilling, processing and production stages. The thermal modification
system was designed to suit working practices within this sector, and to add value to timbers
1 Research Scientist, The BioComposites Centre, Bangor University, Deiniol Road, Bangor, LL57 2UW,
Tel: +44 1248 382029, E-mail: m.j.spear@bangor.ac.uk
2 Coed Cymru, The Old Sawmill, Tregynon, Newtown, Powys, SY16 3PL, Tel: +44 1686 640777, E-mail:
tabithab@coedcymru.org.uk
99
by increasing their suitability for joinery or furniture applications. This paper reports
experiments conducted on Japanese larch.
MATERIAL AND METHODS
Thermal modification of timber
Planks of Japanese larch were sourced from two sawmills within Wales using Welsh grown
timber. Plank cross sections were 110 x 30mm and 155 x 38mm for the small (pilot scale)
kiln and 155 x 25mm for full scale production. Planks were supplied at lengths of 3m for
the full scale kiln, and cross cut to lengths of 90cm for use in the pilot scale kiln.
Thermal modifications were conducted in three stages: a drying day, a treatment day and a
conditioning day. This was selected to allow small businesses to operate kilns on a single
shift working pattern, rather than requiring a larger workforce to operate a three shift pattern
across 24 hours. The conditions for the drying day and the conditioning day were consistent
throughout the study, and utilised temperatures of 120°C and 80°C respectively, with
continuous humidity. The maximum temperature and duration at this temperature for the
treatment day are detailed in the results section, again steam was supplied to elevate
humidity throughout the high temperature stages.
Both treatment ovens are single direction air flow chambers with a baffle above the timber
stack to deflect air through the stack during operation. Conditions within the oven were
controlled by a Eurotherm programmable controller, which determined the set point
throughout the kiln run. The controller received data from a thermocouple located near the
roof of the kiln, in the returning air flow. The conditions within the timber in the stack were
recorded using K-type thermocouples drilled into the core of the planks, or located on the
surface of the planks as a location adjacent to a core thermocouple.
Evaluation of modified timber
The thermally modified planks produced were evaluated on leaving the kiln to observe
uniformity within the stack. A sampling matrix was used to cut samples for moisture
content determination from locations at each side of the kiln, at the centre and from the
upper and lower regions of the stack. The mass of moisture content blocks was recorded
immediately after cutting, and after drying for 24h at 50°C then for 24h at 105°C. The
dimensions of the moisture content samples was also recorded before and after drying.
The colour of the timber was monitored using the moisture content samples after drying for
1 day at 50°C but prior to drying at 105°C for oven dry density determination (this was to
avoid detecting any additional colour change induced due to the higher temperature drying
step). Colour was measured in CIE Lab space using a Datacolor Check II Plus. ΔL, Δa, Δb,
ΔE, ΔC and Δh were calculated against a reference sample of untreated larch heartwood
measured while recently cut, i.e. before any ageing or UV induced processes altered the
colour. Each reading was the average of four measurements, located within a small square
100
pattern on the freshly cut end grain of the timber. This method has been shown to offer a
robust mechanism to avoid colour differences due to proportions of exposed earlywood and
latewood on the transverse faces (Spear et al. 2014). Four such spectral readings were taken
for each sample to calculate the L* a* and b* values, and samples from at least 15 samples
were used for the colorimetric data of each kiln run.
The remaining timber sections cut from between moisture content samples were then used
for sample preparation, including mechanical properties, dimensional stability and
weathering (reported elsewhere). Prior to this, the timber sections were planed, and
observations were made about the quality of the finished surface.
RESULTS AND DISCUSSION
Kiln conditions and dwell time above target temperature
The maximum temperature, and the duration of time spent at this temperature, will be
referred to as the treatment phase for the treatment day. Observation of the thermocouple
data however reveals that the temperature within the timber lagged behind the oven
temperature, and that the peak temperature within the wood varied to a small extent with
location in the stack, and with the level of drying which had occurred in that stack during
the drying day. It is therefore also necessary to define a treatment zone – which for this
paper will be considered the duration for which the temperature of the wood was greater
than 160°C. Typically this zone began at a similar time to the treatment phase, but extended
for a period beyond the end of the treatment phase, due to thermal lag within the stack on
cooling.
Fig.1 Schematic of the treatment phase and treatment zone – direct control of the
treatment phase via kiln programme gives indirect control of treatment zone, i.e. duration
the timber is above the threshold temperature.
101
By altering the kiln set point and duration for the treatment phase, the duration for which
the timber was in the treatment zone can be controlled (Figure 1). Conditions were selected
for kiln runs Regular 1 to 4 to achieve 1 to 5 hours at a temperature of 160°C or greater.
The initial moisture content of the timber prior to the treatment step contributes to the
stability of the relationship between treatment phase parameters and the treatment zone that
results. Similarly, plank thickness, the size of the timber stack, and the thermal output of
the oven contribute to variations in this relationship when increasing load size or moving
between ovens. Within a set of experiments from a single oven, replicability of this
treatment zone conditions relies on efficient pre-conditioning of the timber on day 1, which
for the process used here generated a moisture content of 9 to 11.5% in the planks (the
initial moisture content prior to the day 1 step was 18% and above). Representative data
from the pilot oven and from the full scale oven (Run thin 6 and Run thin 9) are given in
Table 1. This demonstrates the relatively small difference between ovens, and relatively
larger influence of plank thickness – compare for example Run regular 4 and Run large 2,
which were both conducted in the pilot scale oven.
Table 1 Examples of treatment phase with the observed duration of the treatment
zone
Treatment phase
Treatment zone
Temperature (°C)
Duration (min)
Threshold
temperature (°C)
Duration
Run regular 1
180°C
3h
160°C
1h or more
Run regular 2
190°C
3h
160°C
3h or more
Run regular 4
190°C
5h
160°C
4 to 5h
Run large 1
180°C
5h
160°C
1 ½ h
Run large 2
190°C
6h 40
160°C
4 to 5 h
Run thin 6
190°C
2 ½ h
160°C
3 h
Run thin 9
190°C
2 ½ h
160°C
3 ½ h
Density
The bulk density of the planks was calculated using the mass and dimensions of the
moisture content samples. Density between planks is known to vary considerably within
UK grown larch, primarily relating to the proportion of juvenile or mature wood present in
the plank. The number of growth rings per inch (rpi) of the planks also showed wide
variation, with the lowest recorded rpi of 2.8, and the highest 9.0. These values are lower
than would typically be expected for Alpine grown European larch (Larix decidua) or
Scandinavian grown Siberian larch (Larix siberica), and form part of the reason for seeking
to modify this timber to increase uniformity in machining.
102
Table 2 Air dry density data after treatment and conditioning procedure. Data has
been segregated into juvenile and mature wood
Density (g/cm3)
Run large 1
Run large 2
Run large 3
Run large 4
Juvenile
0.501
0.469
0.429
0.444
Mature
0.509
0.517
0.527
0.487
All samples
0.504
0.507
0.451
0.453
As a result of this variability the density data was separated into two groups according to
the growth rings per inch (rpi). An arbitrary value of 6 rpi was used to segregate juvenile
and mature wood, and density data for all samples containing any single rpi measurement
below 6 has been grouped as juvenile in Table 2.
For reference, the density of untreated larch timber from the same source was recorded as
0.556 g/cm3 air dry, these samples were predominantly mature wood, but insufficient
sample blocks of untreated larch have been measured to allow for segregation into values
for juvenile and mature wood to be statistically viable. Further work on the fully finished
and conditioned mechanical properties test specimens is anticipated to address density
change in greater depth. For the same reason, a detailed analysis of mass loss has not been
reported, although it is expected to be very low, due to the short duration and relatively low
temperature within the treatment zone.
Colour
The colour of the timber was only affected to a small extent (Table 3). All ΔE values were
low compared to conventional thermal treatments, especially in the ΔL component. For
example Bekhta and Niemz (2003) reported ΔL values ranging from
-29 to -39 at 200 °C for 4 hours under different humidities. Two sets of samples form the
pilot oven recorded a negative ΔL value, and this was small (-6.57). These planks could be
considered to have received a moderate rather than mild thermal treatment.
Table 3 Colorimetric data for mild and moderate thermal treatment kiln runs
Type
Oven
ΔL
Δa
Δb
ΔE
ΔC
Δh
Run regular 2
Mod
Pilot
-6.57
-2.96
1.81
8.02
0.57
7.96
Run regular 4
Mod
Pilot
-6.28
-4.49
4.50
9.87
2.62
12.29
Run large1
Mild
Pilot
4.46
-4.12
-1.63
6.98
-3.10
7.17
Run large 2
Mild
Pilot
1.05
-5.46
1.52
7.02
-0.50
12.49
Run thin 8
Mod
Full
0.34
-5.74
1.92
8.48
-0.13
12.83
Run thin 9
Mod
Full
-0.60
-5.87
2.18
7.55
0.04
13.70
Run thin 12
Mod
Full
-2.63
-6.18
5.68
9.40
3.38
16.12
103
The mild treated planks (Run large 1-2, Table 3) all showed positive rather than negative
ΔL values. To some extent the colour of larch leads to lower anticipated changes in ΔL,
and a different anticipated change in Δa and Δb. Bekhta and Niemz (2003) reported Δa
from 3.2 to 5.7 and Δb 0.5 to 4 (at 4 hours). As the thermal treatment duration increased in
their study, Δb became increasingly negative and Δa diminished. In this current study it is
Δa which alters to a greater extent than Δb. This, and the reduced ΔL value relates to the
pigment of the flavonoids within larch heartwood, unlike the spruce used by Bekhta and
Niemz. The temperature within the timber was also lower than in their study. The L*, a*
and b* values were consistent with those reported by Spear et al. (2014) for pilot studies
on the mild thermal modification system, where a ΔE of 4.54 was reported for low, and
8.02 for high intensity thermal modifications.
During moderate intensity treatments conducted in the full scale oven, a greater degree of
colour change was seen (Table 3). Here the Δa component is the greatest contributor to the
larger ΔE values, with ΔL again being a relatively minor component. These values
correspond well with the ΔC and Δh values of 0.57 and 7.96 respectively, reported for the
pilot study Run Regular 2. It should be noted that the ΔL values for all three full scale runs
in Table 3 were smaller than the ΔL reported for the pilot moderate treatment, although a
limited number of samples did show higher ΔL values which correlated with localised hot
spots within the oven. The moderate treatment level within this timber treatment system
remains considerably lower than standard thermal modification systems.
Surface quality
The timber planed well, with relatively few chip bruising marks, these can be minimised
by improving extraction and blade sharpness. The incidence of post-machining raised
grain was considerably reduced, as was the fuzzy grain (typical of Type III chip
formation in unmodified timber) where grain deviated around knots and the like.
CONCLUSIONS
The larch timber from the mild and moderate thermal modification processes is distinct in
its relatively small level of colour change, yet the working and machining properties were
significantly improved. The density change and mass loss requires further study and will
be reported elsewhere, however appears to be low, as expected for the short duration of the
treatment zone.
ACKNOWLEDGEMENTS
This project was funded by the Welsh Government using European Regional development
Funds under the Collaborative Industrial Research Programme, Academia for Business.
104
Support from Coed Mon and Menter Mon in scaling up the treatment process is greatly
appreciated.
REFERENCES
Bekhta P. and Niemz P. (2003) Effect of high temperature on the change in color,
dimensional stability and mechanical properties of spruce wood. Holzforschung
57(5):539-546.
Spear M., Binding T., Jenkins D., Nicholls J. and Ormondroyd G. (2014) Mild thermal
modification to enhance the machinability of larch. In: Proceedings of the 7th European
Conference on Wood Modification 2014. 10-12th March 2014, Lisbon, Portugal, 4pp
