Content uploaded by Damien Mahis
Author content
All content in this area was uploaded by Damien Mahis on Feb 04, 2019
Content may be subject to copyright.
Research paper
Thermal characterization of bio-based phase changing materials in
decorative wood-based panels for thermal energy storage
Damien Mathis
a,
*, Pierre Blanchet
a
,V
eronic Landry
a
, Philippe Lagi
ere
b
a
NSERC Industrial Research Chair on Ecoresponsible Wood Construction (CIRCERB), Laval University, 2425 Rue de La Terrasse, Quebec City, QC G1V 0A6,
Canada
b
Engineering and Mechanics Institute, University of Bordeaux, Avenue D'Aquitaine, 33170 Gradignan, France
Received 26 January 2018; revised 1 May 2018; accepted 31 May 2018
Available online 8 June 2018
Abstract
Decorative wood panels containing pouches of bio-based phase changing materials (PCMs) were prepared. Three different PCM mixtures
were used: a blend of capric and lauric acids as well as two commercial products, Puretemp®20 and Puretemp®23 (Puretemp). The panels
consist of engraved Medium Density Fiberboard (MDF) filled with a plastic pouch filled with PCM. High density fiberboard (HDF) was used on
top of the panels to enclose the PCM pouches. PCM mixtures were first tested by differential scanning calorimetry (DSC). Phase change
temperature and total heat storage of the panels were measured for both fusion and solidification with a Dynamic Heat-Flow Meter Apparatus
(DHFMA). DSC and DHFMA results were compared, allowing a better understanding of results gathered from these two techniques. DSC
calibration has been revealed important when assessing PCMs. The panels present a phase change temperature and a latent heat storage suitable
for buildings applications. The panel made with Puretemp®23 presented the highest energy, with 57.1 J g
1
. Thermal cycling was conducted on
the panels to investigate thermal reliability, which revealed small modifications of thermal properties for two products. For all cases, latent heat
was found stable. Hygro-mechanical behavior of the panels was also evaluated as these where designed to be esthetic decorative panels. This
study exposes the potential of a new type of wood-based panels loaded with PCM for thermal energy storage and brings overall knowledge about
PCM products thermal characterization.
©2018, Institute of Process Engineering, Chinese Academy of Sciences. Publishing services by Elsevier B.V. on behalf of KeAi Communi-
cations Co., Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: PCM; Wood; Biosourced; Fatty acids; DHFMA
1. Introduction
Phase change materials (PCMs) are recognized as an
effective way to enhance buildings thermal mass [1].As
melting occurs, a significant amount of energy is stored. This
energy is then released in the building by solidification as
interior temperature decreases. If implemented carefully, PCM
products can permit to achieve energy savings in the heating
period [2,3] and reduce buildings overheating in summer [4,5].
A wide range of solutions, either active or passive, have been
experimented [6] with PCMs.
Potential benefits of PCM wallboards have already been
demonstrated across multiple full-scale experimental studies
[7–9]. An advantage of wallboards is their proximity with the
interior air of buildings, compared to solutions that implement
PCMs within the walls. As thermal comfort depends on oper-
ative temperature, panels containing PCMs could have a great
contribution by keeping wall temperatures within the comfort
zone [10]. They can store energy by receiving direct sunlight or
according to the convection inside the building [11].
There are only a few trials of PCM composites using wood
as a substrate. Although, wood is a renewable material that
*Corresponding author. 680 avenue Calixa Lavall
ee, G1S 3G6 Qu
ebec,
Qu
ebec, Canada.
E-mail address: damien.mathis.1@ulaval.ca (D. Mathis).
Available online at www.sciencedirect.com
ScienceDirect
Green Energy &Environment 4 (2019) 56e65
www.keaipublishing.com/gee
https://doi.org/10.1016/j.gee.2018.05.004
2468-0257/©2018, Institute of Process Engineering, Chinese Academy of Sciences. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co.,
Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
could lead to lower embodied energy products [12]. Xi Guo
et al. (2016) prepared wood plastic composites (WPC) by
integrating microencapsulated dodecanol into wood flour with
high-density polyethylene (HDPE). They obtained a compos-
ite with a modulus of elasticity between 0.9 and 2.2 GPa, and
20.3 J g
1
of latent heat [13]. Mechanical properties were
decreasing with the increase of PCM microcapsules ratio.
Jeong et al. (2012) proposed to add microencapsulated PCMs
(MPCMs) in wood adhesives to enhance wood flooring ther-
mal mass. Bonding strength remained good, but thermal mass
enhancement was low [14]. Barreneche et al. (2017) impreg-
nated wood with paraffinic PCMs RT-21 and RT-27 from
Rubitherm®[15]. They obtained a composite with a latent
heat of 20.62 J g
1
for a PCM mass proportion of 29.9%. To
avoid any leakage, they encapsulated the wood with a poly-
styrene solution in organic solvent. However, wood is a ma-
terial that can age and undergo dimensional variations.
Applying a coating that will stabilize his structure may cause
later damages and thus leakage.
The composites previously cited are made with a large
proportion of non-renewable resources, such as HDPE, poly-
styrene and paraffinic PCMs. Furthermore, the mass of
MPCMs that can be added into WPC is limited because of its
negative effect on mechanical properties [13]. This leads to a
composite with a low mass fraction of PCM, thus a limited
latent heat. In this context, wood decorative panels containing
pouches of bio-based organic PCMs were designed. A simple
design would lead to an ease for recycling.
Various types of PCMs have been used over decades [16].
Organic PCMs are mainly constituted by paraffinic and bio-
based compounds. Inorganic PCMs includes salt hydrates
and metallic compounds. Fatty acids are widely used organic
PCMs. They are used as binary eutectic mixtures in order to
have suitable phase change temperatures. Main advantages of
these eutectic mixtures are a low supercooling, chemical sta-
bility and possibly low environmental impact [17,18].Asif
binary mixtures are made of two compounds, there is no
segregation over phase change cycles [19].
Thermal characterization of PCMs can be achieved with
several techniques [20]. Differential Scanning Calorimetry
(DSC) is a fast and efficient technique that has been widely
used, however it is associated with some disadvantages or
uncertainties regarding PCM evaluation. There is still no
consensus about how DSC test and calibration should be
performed for PCM testing [21,22]. However, the calibration
will have an impact on the melting and solidifications tem-
peratures. Any error will lead to a wrong estimation of the
PCM behavior and thus modify the strategies required for a
successful integration in a building. In addition, DSC only
allows to test small samples. In 2014, ASTM introduced the
new test standard C1784 to measure thermal storage properties
of PCM-enhanced products of a larger scale based on the
Dynamic Heat-Flow Meter Apparatus (DHFMA) principle and
method. This methodology requires a longer time of charac-
terization than DSC but allows to avoid some of its constraints
[23]. With DSC, thermal inertia of the samples affects their
phase change temperature, which is not the case with
DHFMA. In 2015, Shukla and Kosny characterized five PCM-
integrated products to fill the lack of test data and describe a
reliable procedure for the DHFMA method [23]. However,
scientific literature about this innovative and efficient new
methodology is still lacking. A comparison between DSC and
DHFMA results would allow a better understanding of PCM
thermal characterization. Determining accurately the thermal
properties of PCM products is critical to achieve successful
implementations within buildings [24].
For this study, a eutectic mixture of capric and lauric acids
was prepared. In addition, commercial products from Pure-
temp®were used. PCMs were first tested by DSC. Then,
wood-based panels were formed and characterized. The DSC
results for PCMs mixtures were compared to the DHFMA
results for panels. This comparison will allow a better un-
derstanding of thermal characterization of PCM products.
Finally, thermal and hygro-mechanical behavior of the panels
was assessed.
This study aims to determine the thermal storage capacity
and hygro-mechanical behavior of decorative wooden panels
loaded with different bio-based PCMs. An innovative com-
parison between DSC and DHMFA tests is presented. Results
of the present study will allow to assess about the potential of
such panels for thermal energy storage and contribute to the
overall knowledge of the thermal characterization of PCM
products.
2. Material and methods
2.1. Materials
2.1.1. Phase change materials
Capric (CA) and lauric (LA) acids were purchased from
Sigma Aldrich (Canada). In order to simulate an industrial use
with reasonable costs, a purity of 95% was selected. Accord-
ing to Sigma Aldrich, melting range of CA is 27–32 C and its
density is 0.893 g cm
3
at 25 C; melting point of LA is
43.8 C and its density is 0.49 g cm
3
at 20 C.
The two other PCMs used in this study were commercial
products. They were bought from Puretemp
®
(USA). Pure-
temp®products are USDA certified 100% bio-based products.
Exact composition is not known as these products are under
patent protection. Two products used were: PT20 and PT23,
respectively with phase change temperatures of 20 C and
23 C.
2.1.2. Wood panels
Medium Density Fiberboard (MDF) was obtained from
Uniboard®(Canada). They were used as the main component
of the decorative panels developed. MDF is a widely used
wood panel prepared from wood fibers combined and a urea
formaldehyde resin by applying high temperature and
pressure.
On the upper face of the panels (Fig. 1b), High Density
Fiberboard (HDF) was chosen, for several reasons. As it
constitutes the face of the panels that will face the interior of
the building, this part needs to have a good mechanical and
57D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
physical properties. In addition, the layer facing the interior
needs to be thin. The thinner this layer is, the better the
thermal exchange between the PCM and the interior air is.
Furthermore, HDF can be found in low thickness on a com-
mercial scale. HDF Fibrex®from the company Goodfellow®
(Canada) was used.
Main properties of wood components used are listed in
Table 1.
2.2. Samples preparation
2.2.1. Phase change materials
A blend of capric acid and lauric acid was prepared by
mixing them at 40 C for 10 min. Eutectic point is reached
when the two components melt simultaneously. On the phase
diagram and therefore on the DSC enthalpy curve, this cor-
responds to a single peak of fusion and solidification. Eutectic
is the composition for which the melting point is the lowest. In
the literature, eutectic composition was ranging from 61.5-
38.5% to 67-33% for CA/LA ratio [18].
The commercial Puretemp®mixtures were used as
received, no preparation was required.
2.2.2. Panels preparation
In order to prepare the wood panels, PCM mixtures were
encapsulated first. This was achieved using a vacuum machine
and 0.08 mm thick polyethylene bags.
MDF panels 13 mm thick were grooved to contain the PCM
pouches in it. 5.5 mm thick HDF was glued with a PVA
adhesive to the MDF to close the panel. Then, a maple veneer
was manually applied using a pressure-sensitive adhesive.
Finally, panels had an overall thickness of 19.5 ±0.5 mm in
the Fig. 1a), a sawn panel so the PCM pouch is visible. A
sketch of the assembly is given in Fig. 1b).
2.3. Characterization of the panels
2.3.1. DSC
Measurements of PCMs melting, freezing temperatures and
latent heats were performed by DSC (METTLER Toledo
822E, Canada). The DSC instrument was calibrated with in-
dium at 10 C min
1
. Measurements were performed at
10 C min
1
constant heating rate, with a temperature range
from 0 Cto50C under a constant stream of nitrogen of
80 ml min
1
. For each compound, three samples were tested.
Results are the average of three measurements given with the
standard deviation. Aluminum crucibles were filled with a
mass of 4 mg–6 mg.
There is no consensus about the DSC heating rate when
testing PCM samples [21,22], as the result of transition tem-
peratures are dependent on the heating rate [25]. A rate of
10 C min
1
was selected. Some studies report that lower
heating rates or a step-by-step method could enhance DSC test
accuracy [22,26,27]. An explanation for this is that PCMs
embody a large amount of latent heat. This can cause lag if the
test is performed too fast, as it generates a temperature
gradient in the sample [28]. However, preliminary DSC tests
were conducted on distilled water with hermetic aluminum
Fig. 1. (a) Panel with the PCM pouch (b) sketch of a wood/PCM panel.
Table 1
Properties of wood panels as provided by the suppliers.
Density (kg m
3
) Module of rupture (N mm
2
) Resin Thickness (mm) Reference
MDF 525 15 Urea-formaldehyde 13 mm Uniboard®
HDF 900 42,7 Urea-formaldehyde 5.5 mm Goodfellow®
58 D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
crucibles. These tests revealed that that lowering the heating
rate was giving wrong ice melting temperature values. A
1C min
1
heating rate was giving a melting point of 2.78 C
while distilled water melts at 0 C. This could be explained by
the large difference of melting point between the material used
for calibration and the PCMs [29]. Indium melts at 156.60 C
[30] while the PCMs used melts around 20 C. The approxi-
mation made by the DSC could bring shifted results.
In addition, calibrating with a metal to assess organic ma-
terials properties could bias the accuracy. The main causes
could be that a liquid metal will have a higher thermal con-
ductivity than an organic liquid [31]. Thus, the DSC sensors
would receive lately the heat from the liquid PCM. This delay
could lead to a higher melting point, such as measured for
distilled water in the calibration.
Distilled water has a melting point and latent heat closer to
the PCMs than Indium. Then, considering that the
10 C min
1
rate was measuring a melting temperature of
water at 0.13 C, this rate was selected for the tests.
For both freezing and melting points, onset temperatures
were considered. The latent heat was determined by numerical
integration of the peak. Results are the average of three
measurements. For each test, cooling was conducted first so
the samples would have the same solidification rate.
2.3.2. Heat storage measurements
Total heat storage measurements were carried out on the
panels with a DHFMA method, using a Lasercomp Fox FX314
(TA instruments, USA). Tests were conducted in according
with ASTM C1784 “Standard Test Method for Using a Heat
Flow Meter Apparatus for Measuring Thermal Storage
Properties of Phase Change Materials and Products”(ASTM,
2016). This experimental method allows to characterize larger
samples than DSC, such as panels or flooring solutions. Total
heat storage is the addition of sensible heat storage of both
wood and PCM plus the latent heat storage of the PCM.
For each panel, cooling was conducted first so the PCMs
can have the same solidification rate. 30 by 30 cm panels were
placed in the apparatus. Upper and lower plate were main-
tained at the same temperature. A ramp of temperature was
applied from 10above the estimated melting temperature,
until 10below this temperature. According to ASTM C1784,
temperature steps must be 1.5 ±0.5 C. For these tests, 1.2 C
step was chosen, so the test could have sufficient precision to
avoid saturation. The enthalpy measured is the total heat
storage. It includes wood specific heat, PCM specific heat in
both liquid and solid state, plus the phase change enthalpy.
Every total heat curve is the result of two replicate runs on
a panel. Two tests are conducted with a 0.6 C temperature
shift. Then, the two tests are merged to obtain the final total
heat curve. This methodology, described in ASTM C1784,
allows to enhance DHFMA test accuracy.
For both freezing and melting points, onset temperatures
were considered. They were obtained considering initial
baselines and their intersections with the maximum slope of
the total heat curve, as shown in Fig. 2. Tm and Ts respec-
tively refer to melting and solidification temperatures. Heat
storage for both fusion and solidification were calculated using
the data provided by the FOX FX314 considering a transition
range of 3. For each PCM mixture, two panels were tested.
Results are the average of the two panels.
A correction must be applied on the results given by the
Fox FX314. Indeed, the heat flux meter are on the center of
each plate. It sees and takes measurements in the central part
of the panel that is a laminated composite of 3 materials:
MDF, PCM pouch and HDF. However, the panel is not hori-
zontally homogenous. Indeed, the panels are made with a
wood edge allowing MDF and HDF to be glued, ensuring
mechanical stability. In order to have real total heat storage,
results must be corrected using the surface percentage of the
panel that does not embody any PCM. To achieve this, MDF
and HDF specific heats were measured and a correction was
applied.
2.3.3. Thermal cycling
PCM panels need to be energetically efficient over years. In
order to achieve this efficiency, their properties need to remain
constant over time. Indeed, fusion temperature and fusion
enthalpy are chosen according to the climatic zone and the
house architecture. Any variation in the material properties
would make the energetic calculus and previsions obsolete.
In the literature, thermal reliability of PCM products is
tested by cycling the materials. It starts from a few dozens to
several thousands [32]. In this study, each panel was thermally
cycled 200 times. One cycle includes one fusion and one so-
lidification. Cycling was achieved using the aging module of
the Lasercomp Fox FX314. To ensure that phase transitions
were completed, heat flux was measured using the aging
module of the Lasercomp Fox FX314.
2.3.4. Moisture behavior
The wood-based PCM panels are prepared to cover interior
walls of buildings. Thus, they need to be aesthetics. In this
context, stability in their design over time must be kept to
ensure the decorative function. However, the entrapped pouch
of PCM and its effects on moisture management needed to be
known. Moisture could degrade the panels and drive surface
defects. In order to characterize this, moisture behavior tests
were conducted with two samples: PCM panels and control
panels. PCM panels were those described previously, filled
with CA and LA acids. As a control, panels of the same size
but without the hole and the PCM pouch were used.
Tests were conducted in three phases: stabilization under an
atmosphere at 20 C/80% relative humidity (RH), then 20 C/
20% RH and then again 20 C/80% RH. Each phase lasted 21
days. Measurements were done for each cycle after 4, 7, 14
and 21 days. Similar tests were conducted by Blanchet et al. to
assess moisture behavior of engineered wood flooring [33].
This methodology allows to assess the moisture behavior of
similar products under high (summer) and low (winter)
59D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
relative humidity in North America. All sides of the panel
were exposed to the water.
Nine surface points were considered on the panels,
distributed as shown in Fig. 3. A 3D portable coordinate
measuring arm, Fusion FaroArm (FARO, Canada) was used to
measure the deformation each week. Measurements were done
using the probing mode with a probe diameter of 6 mm.
In order to assess panels global deformation, surface points
were grouped. Group A includes points 1-6-7, group B is for
points 2-5-8 and group C is for points 3-4-9. The average
position of points was calculated for each group at each day of
measure. This allows to trace curves representing the average
shape of the panels over cycles.
In addition, the amplitude of deformation for each cycle
was assessed. Amplitude is the highest measure minus the
lowest measure within all the points of the panel. This was
calculated for each panel at each cycle. Then, the average
amplitude for both control and PCM panels were calculated.
3. Results and discussion
3.1. DSC of PCM mixtures
DSC results performed on the acids used in this study was
confirmed to be eutectic at 64% of capric acid and 36% of
lauric acid. Thermal properties of PCM measured with DSC
tests are presented in Table 2.
Results for the melting point of the capric/lauric (CA/LA)
mixture are in accordance with literature, as presented in Table
3. However, solidifying temperatures are higher than excepted.
Indeed, organic PCMs usually show a little supercooling
[34,35]. This higher solidifying temperature could result from
differences in the methodology used with DSC tests.
For commercial products, PT20 and PT23, latent heats
were close to the values of the technical data sheets but the
fusion temperature found were lower. Results are shown in
Table 4. As these mixtures are made from agricultural waste,
their properties may vary a little over batches. Indeed, two
different batches were tested and had different melting tem-
peratures. The results presented here are from the batch that
was used to prepare the panels.
3.2. Panels thermal properties
Total heat storage curves for one of each panel are shown in
Fig. 4.
Fig. 2. Determination of Ts and Tm.
Fig. 3. Distribution of measure points on the panel.
Table 2
DSC results.
Capric (64%)/
Lauric (36%)
PT20 PT23
Melting point onset (C) 19.7 ±0.1 18.6 ±0.1 22.3 ±0.2
Latent heat of fusion (J g
1
) 128.7 ±1.7 174.6 ±2.2 197.9 ±13.2
Solidifying point onset (C) 23.5 ±0.3 23.8 ±0.1 25.6 ±0.2
Latent heat of solidification (J g
1
) 124.9 ±1.5 172.5 ±1.3 200.1 ±12.4
60 D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
For each panel, total heat storages over a transition range of
3C, fusion temperature and solidification temperature were
calculated. Results are given in Table 5.
Panel PT23 exhibits the highest heat storage. DSCs tests
also indicated that PT23 mixture had the highest heat storage.
For each panel, solidification temperature is lower than fusion
temperature. It is concordant with the supercooling often
encountered with organic PCMs [30,31].
Solidification total heat is higher than fusion total heat. This
can come from the fact that fusion takes place within a larger
temperature range. As the total heat is here calculated over
3C, it does not allow to consider the whole fusion heat.
Extending this range of calculus would have included more
specific heat. Therefore, the 3 C range was kept.
Repeatability of the FOX FX314 measurements was high.
Three tests on the same panels gave the exact same solidifi-
cation temperature. This was possible by the fact that the tests
were conducted step-by-step every 0.6 C. This methodology
leads to an incertitude that depends on the steps width. Smaller
steps would have led to a lower incertitude.
For the PT20 panel, the two samples had different solidi-
fication temperatures with a 1.2 C shift. The two samples
were tested with an interval of 7 months. A reason for this shift
could be that the mixture PT20 aged. Indeed, the second panel
tested has the same solidification temperature than the panel
that was thermally cycled. Thermal cycling performance will
be presented in the next section. As the PT23 product was
stored in a laboratory where temperature can fluctuate, it is
possible that cycles of fusion and solidification occurred
naturally.
The Table 6 presents a comparison of several phase-change
wallboards in literature. Wood-based panels from the present
study exhibit characteristics comparable to existing products
and embody the highest fusion heat for a wood-based
composite.
Wood-based products such as MDF and HDF are materials
with a low thermal conductivity. This can limit the thermal
exchange and thus the PCM melting and solidification rate. In
order to assess this thermal exchange, an additional experiment
was achieved. Three panels containing PT23 were manufac-
tured and characterized. The first panel was a standard wood-
based panel as presented above. For the second and third
panel, HDF was replaced by a 2 mm thick aluminum plate and
by a standard 9.5 mm thick gypsum board, respectively. The
panels were placed in the Fox FX314, stabilized at 18 C and
then the temperature was fixed at 26 C. The time required to
melt the PCM with the HDF, the aluminum and the plasterboard
were recorded and were respectively 3.6 h, 2.8 h and 4.8 h.
It is important to consider that the panels from the present
study are designed to constitute the most inner layer of a wall.
Therefore, the 5.5 mm HDF sheet (covered with 1 mm thick
maple veneer) is the only layer separating the PCM from the
interior air. In comparison with an implementation of PCMs
behind a plasterboard, the thermal exchange could be higher
for the wood-based panels, as our experiment suggests. A full
scale-experiment is required to conclude about the efficiency
of such panels as a thermal management solution.
3.3. Thermal cycling
Fusion, solidification temperatures and total heat storages
of the panels, before and after cycling, are shown in Table 7.
Thermal cycling did not seem to affect fusion tempera-
tures. Solidification temperatures of Puretemp®products
were affected. PT20 had a solidifying point 1.2higher,
which is something positive, as its supercooling was reduced.
For PT23, it was 0.6 lower. Explaining these shifts is
complicated as Puretemp®mixtures composition is unknown.
A comparison of the panel's thermal behavior before and after
aging can be seen in Fig. 5. Except for the solidification shift,
panels had overall the same thermal behavior before and after
aging.
Table 3
Comparison of CA/LA mixtures properties.
Eutectic composition Purity Melting point
onset (C)
Latent heat of
melting (C)
Solidification
point onset (C)
Latent heat of
solidification
Ref Methodology
CA/LA (64-36 w%) 95e95% 19.7 128.7 23,5 124.9 Present study DSC Mettler Toledo
Calibrated Indium 10min
1
Test ramp 10min
1
CA/LA (64-36 w%) 98e98% 19.6 150.0 19.4 149.0 [36] DSC PerkinElmer Jade
Calibrated Indium
Test ramp 5min
1
CA/LA (62-38) e18.6 121.0 ee[37] DSC, Q20
Test ramp 2 C min
1
Table 4
Puretemp®products properties.
Mixture Melting point
onset (C)
Latent heat of
melting (C)
Experimental method Ref
PT20 18.7 ±0.2 179 ±4 DSC Mettler Toledo
Calibrated
Indium 10min
1
Test ramp 10min
1
Current study
PT20 19.9 ±0.3 171 DSC Q2000
Test ramp 1min
1
Puretemp®
PT23 22.1 ±0.1 193 ±20 DSC Mettler Toledo
Calibrated
Indium 10min
1
Test ramp 10min
1
Current study
PT23 23.4 ±0.2 201 DSC Q2000
Test ramp 1min
1
Puretemp®
61D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
3.4. Comparison between DSC and DHFMA methods
Table 8 presents a comparison between DSC and DHFMA
tests. Product PT20 was removed as it’s repeatability among
tests wasn't sufficient, as detailed above. The heat storage
values were not compared because DHFMA tests are run on
panels, while DSC tests are run on pure products. Thus,
DHFMA measurements include the sensible heat of wood.
Comparison will be done with the mixtures before aging.
For product PT23, melting points were found almost
identical with the two methods, with a 0.1 C difference. For
CA/LA mixture, melting point was found 0.9 C higher with
DHFMA test method. As DHFMA tests are driven according
to a step-by-step method, the incertitude over the measure is
dictated by the setpoints width. A shifted third experiment on
the panels would have reduced this incertitude. It would also
have enhanced the number of points available to determine the
onset fusion temperature. Indeed, it is done by drawing a line
at the maximum slope of the peak and precision is higher with
Fig. 4. Panels heat storage properties a) PT20, b) PT23 and c) CA/LA.
Table 5
Decorative wood-based panels properties.
Panel CA/LA Panel PT20 Panel PT23
Melting point onset (C) 20.6 ±0.1 20.2 ±0.1 22.2 ±0.2
Heat of fusion (J g
1
) 47.5 ±1.6 56.9 ±1.8 57.1 ±0.5
Solidifying point (C) 20.5 ±0.1 19.8 ±0.6 22.2 ±0.1
Heat of solidification (J g
1
) 49.8 ±4.3 60.7 ±3.3 65.9 ±3.4
62 D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
a larger number of points. Beside these considerations, the
DHFMA step-by-step method has the advantage to get rid of
the thermal inertia of samples and surface effects that can
affect the results with the DSC [28].
In DSC tests, solidifying points were respectively found
3C and 3.4 C higher for CA/LA and product PT23. DHFMA
results are concordant with literature, as organic PCM's usu-
ally show a little supercooling [34]. Reasons for the higher
values with DSC tests were already discussed previously.
Table 6
Comparison of wallboards properties.
Material Melting
temperature (C)
Solidifying
temperature (C)
Melting
heat (J g
1
)
Reference Measurement method
Capric-myristic acid/Expanded
perlite (55/45 w%)
21.7 20.7 85.4 [38] DSC
PEG (50 w%)/diatomite 27.7 32.2 87.1 [39] DSC
WPC- Microencapsulated
dodecanol/wood flour/HDPE
22.2e28.7 22.2e28.7 20.3 [13] DSC (Q2000, USA)
Wood/RT27 26.2 e20.6 [15] DSC Mettler Toledo
Wood-based PT20 20.2 19.8 56.9 Present study DHFMA Lasercomp FX314
Wood-based PT23 22.2 22.2 57.1 Present study DHFMA
Lasercomp FX314
Wood-based CA/LA 20.8 20.5 47.5 Present study DHFMA Lasercomp FX314
Table 7
Aging results for the wood-based panels.
Panel filled
with
Melting
point (C)
Total fusion
heat over
3C(Jg
1
)
Solidifying
point (C)
Total
solidification
heat over 3 C(Jg
1
)
CA/LA 20.6 48.6 20.4 52.8
CA/LA after
cycling
20.5 48.1 20.4 48.4
PT20 20.2 58.7 19.6 63.9
PT20 after
cycling
20.1 60 20.4 60.3
PT23 22.1 54.1 22.2 63.7
PT23 after
cycling
22.3 54.7 21.6 63.6
Fig. 5. Comparison between control and aged curves.
Table 8
Comparison of DSC and DHFMA panels results.
Mixture Test method Melting point
onset (C)
Solidifying point
onset (C)
CA/LA DSC 19,7 ±0.1 23.5 ±0.3
CA/LA DHFMA 20.6 ±0.1 20.5 ±0.1
PT23 DSC 22,3 ±0.2 25.6 ±0.2
PT23 DHFMA 22.2 ±0.2 22.2 ±0.1
63D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
3.5. Moisture behavior
Deformation of panels at different steps are shown in Fig. 6.
Panels undergo cupping when submitted to humidity as
shown Fig. 6b and d. They also tend to crown during the dry
cycle as shown Fig. 6c.
The panels seem to have similar behaviors. However,
amplitude measurements from Table 9 reveals a difference in
behaviors between PCM and control panels.
PCM panels seem to have more withdrawal during the dry
cycle, as their amplitude of measure is higher. This could be
explained by the fact that PCM panels are grooved and thus
have less mechanical strength. They would resist less to the
bending forced caused by the withdrawal. For an industrial
use, this behavior should be carefully mastered to ensure that
the wood-based PCM panels meet customer expectations.
4. Conclusions
Phase change materials can enhance buildings thermal
mass and thus lead to energy savings. This study aimed to
assess properties of bio-based PCM and of decorative wood-
based panels made with those PCMs.
These experiments show that the wood-based interior
panels have a suitable phase change temperature and latent
heat storage to achieve thermal energy storage in buildings.
They can embody a large amount of energy, as high as
57.1 J g
1
for a melting temperature of 22.2 C, which is
comparable to existing PCM wallboards. Thermal melting
temperature over aging of CA/LA mixture and Puretemp®
products was found stable. Solidification temperatures of
Puretemp®products were found a slightly unstable. The
comparison between DSC and DHFMA tests on PCM prod-
ucts revealed small differences for melting temperature and
higher differences for solidification temperature. Reasons for
these shifts were discussed and impact of DSC calibration was
highlighted. The results presented in this article enhance the
global knowledge on PCM thermal testing.
Hygro-mechanical tests revealed that the panels filled with
PCMs undergo to higher dimensional fluctuations during a dry
cycle. In case of an integration in a building, mechanical
stability of such panels should be controlled.
The results presented above suggest that such decorative
wood-based panels loaded with bio-based PCM have a great
potential for thermal energy storage. They could be added in
existing buildings with further developments. However, in order
to assess the efficiency of such panels as a thermal management
solution for buildings, full-scale experiments are required.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgments
The authors are grateful to Natural Sciences and Engi-
neering Research Council of Canada for the financial support
through its ICP and CRD programs (IRCPJ 461745-12 and
RDCPJ 445200-12) as well as the industrial partners of the
NSERC industrial chair on eco-responsible wood construction
(CIRCERB).
Initial
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
AABBCC
Cycle 1-20°C/80 RH
Cycle 2-20°C/20 RH Cycle 3-20°C/80 RH
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
Reference
PCM panel
panel
Reference
PCM panel
panel
(a)
(c)
(d) ABC
Cycle 1-20°C/80 RH
Cycle 3-20°C/80 RH
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
0,600
0,500
0,400
0,200
0,300
0,100
0,000
-0,100
-0,200
-0,300
Measure (mm)
Reference
PCM panel
panel
Reference
PCM panel
panel
(d)
(b)
Reference
PCM panel
panel
Reference
PCM panel
panel
Fig. 6. Panels deformation over moisture.
Table 9
Amplitude of measure for PCM and control panels.
1 2 3 Average
Control panels
Amplitude Cycle 1 (mm) 0,513 0,67 0,5035 0,562
Amplitude Cycle 2 (mm) 0,666 0644 1,30825 0,873
Amplitude Cycle 3 (mm) 0,669 0,65025 0,6195 0,646
PCM panels
Amplitude Cycle 1 (mm) 0,581 0510 0,594 0562
Amplitude Cycle 2 (mm) 1391 1538 1383 1437
Amplitude Cycle 3 (mm) 0,665 0694 0,689 0682
64 D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
References
[1] R. Zeng, X. Wang, H. Di, F. Jiang, Y. Zhang, Energy Build. 43 (2011)
1081–1090.
[2] K. Peippo, P. Kauranen, P.D. Lund, Energy Build. 17 (1991) 259–270.
[3] Z. Zhou, Z. Zhang, J. Zuo, K. Huang, L. Zhang, Renew. Sustain. Energy
Rev. 48 (2015) 692–703.
[4] A. Castell, M.M. Farid, Energy Build. 81 (2014) 59–71.
[5] B. Zalba, J.M. Marın, L.F. Cabeza, H. Mehling, Int. J. Refrig. 27 (2004)
839–849.
[6] M. Pomianowski, P. Heiselberg, Y. Zhang, Energy Build. 67 (2013) 56–
69.
[7] F. Kuznik, J. Virgone, J.J. Roux, Energy Build. 40 (2008) 148–156.
[8] A.K. Athienitis, C. Liu, D. Hawes, D. Banu, D. Feldman, Build. Environ.
32 (1997) 405–410.
[9] F. Kuznik, J. Virgone, K. Johannes, Renew. Energy 36 (2011) 1458–
1462.
[10] B. Nghana, F. Tariku, Build. Environ. 99 (2016) 221–238.
[11] V.V. Tyagi, D. Buddhi, Renew. Sustain. Energy Rev. 11 (2007) 1146–
1166.
[12] J. Perez-Garcia, B. Lippke, J. Comnick, C. Manriquez, Wood Fiber Sci.
37 (2007) 140–148.
[13] X. Guo, J. Cao, Y. Peng, R. Liu, Mater. Des. 89 (2016) 1325–1334.
[14] S.G. Jeong, J. Jeon, J. Seo, J.H. Lee, S. Kim, Energy Convers. Manag. 64
(2012) 516–521.
[15] C. Barreneche, J. Vecstaudza, D. Bajare, A.I. Fernandez, IOP Conf. Ser.:
Mater. Sci. Eng. 251 (2017) 012111.
[16] G.A. Lane, in: C. Guyer (Ed.), Phase Change Thermal Storage Materials,
McGraw Hill Book Co, USA, 1989, pp. 720.
[17] M.M. Farid, A.M. Khudhair, S.A.K. Razack, S. Al-Hallaj, Energy
Convers. Manag. 45 (2004) 1597–1615.
[18] Y. Yuan, N. Zhang, W. Tao, X. Cao, Y. He, Renew. Sustain. Energy Rev.
29 (2014) 482–498.
[19] J.A. No€
el, P.M. Allred, M.A. White, Int. J. Life Cycle Assess. 20 (2015)
367–376.
[20] J. Ko
sny, in: J. Ko
sny (Ed.), PCM-Enhanced Building Components,
Springer, UK, 2015, pp. 109–137.
[21] R.M.R. Saeed, Thermal characterization of Phase Change Materials for
Thermal Energy Storage. Missouri University of Science and Technol-
ogy. ProQuest, USA, 2016.
[22] C. Castell
on, E. Gu
¨nther, H. Mehling, S. Hiebler, L.F. Cabeza, Int. J.
Energy Res. 32 (2008) 1258–1265.
[23] N. Shukla, J. Kosny, Curr. Sustain. Renew. Energy Rep. 2 (2015) 41–46.
[24] G. Feng, K. Huang, H. Xie, H. Li, X. Liu, S. Liu, C. Cao, Renew. Energy
87 (2016) 1148–1153.
[25] X. Jin, X. Xu, X. Zhang, Y. Yin, Thermochim. Acta 595 (2014) 17–21.
[26] C. Barreneche, A. Sol
e, L. Mir
o, I. Martorell, A.I. Fern
andez,
L.F. Cabeza, Thermochim. Acta 553 (2013) 23–26.
[27] E. Gu
¨nther, S. Hiebler, H. Mehling, Determination of the Heat Storage
Capacity of PCM and PCM-objects as a Function of Temperature, in:
Proc of ECOSTOCK, 10th International Conference on Thermal Energy
Storage, 2006. USA.
[28] E. Gu
¨nther, S. Hiebler, H. Mehling, R. Redlich, Int. J. Thermophys. 30
(2009) 1257–1269.
[29] G.W.H. H€
ohne, H.K. Cammenga, W. Eysel, E. Gmelin, W. Hemminger,
Thermochim. Acta 160 (1990) 1–12.
[30] H.K. Cammenga, W. Eysel, E. Gmelin, W. Hemminger, G.W. H€
ohne,
S.M. Sarge, Thermochim. Acta 219 (1993) 333–342.
[31] D. Price, J. Therm. Anal. Calorim. 45 (1995) 1285–1296.
[32] G. Ferrer, A. Sol
e, C. Barreneche, I. Martorell, L.F. Cabeza, Renew.
Sustain. Energy Rev. 50 (2015) 665–685.
[33] P. Blanchet, R. Beauregard, A. Cloutier, G. Gendron, M. Lefebvre, For.
Prod. J. 53 (2003) 30–37.
[34] A. Karaipekli, A. Sarı, J. Ind. Eng. Chem. 16 (2010) 767–773.
[35] R.K. Sharma, P. Ganesan, V.V. Tyagi, H.S.C. Metselaar, S.C. Sandaran,
Energy Convers. Manag. 95 (2015) 193–228.
[36] Z. Zhang, Y. Yuan, N. Zhang, X. Cao, Energy 90 (2015) 359–368.
[37] F. Kuznik, D. David, K. Johannes, J.J. Roux, Renew. Sustain. Energy
Rev. 15 (2011) 379–391.
[38] A. Karaipekli, A. Sarı, Renew. Energy 33 (2008) 2599–2605.
[39] S. Karaman, A. Karaipekli, A. Sarı, A. Bicer, Sol. Energy Mater. Sol.
Cell. 95 (2011) 1647–1653.
65D. Mathis et al. / Green Energy &Environment 4 (2019) 56–65
