MANUFACTURING & PRODUCTS PROJECT NUMBER: PN04.2001 Wood/water relationships during kiln drying and reconditioning of softwoods This report can also be viewed on the FWPRDC website www.fwprdc.org.au FWPRDC PO Box 69, World Trade Centre Melbourne 8005, Victoria T +61 3 9614 7544 F +61 3 9614 6822 MAY 2007
68
Embed
Forest and Wood Products Australia Limited (FWPA) - n Products … wood... · 2018. 8. 2. · Manufacturing & Products Project number: P n 04.2001 Wood/water relationships during
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Ma
nu
fac
turi
ng
&
Pro
du
cts
Pro
jec
t n
um
ber
: Pn
04.2
001 Wood/water
relationships during kiln drying and reconditioning of softwoods
this
rep
ort
can
als
o b
e vi
ewed
on
th
e FW
PrD
c w
ebsi
te
ww
w.f
wp
rdc.
org
.au
fWPr
dc
Po B
ox 6
9, W
orld
tra
de c
entr
eM
elbo
urne
800
5, V
icto
riat
+61
3 9
614
7544
f +
61 3
961
4 68
22
may 2007
Wood/water relationships during kiln drying and reconditioning
of softwoods
Prepared for the
Forest and Wood Products Research and Development Corporation
by
R. Northway and I. Burgar
The FWPRDC is jointly funded by the Australian forest and wood products industry and the Australian Government.
Publication: Wood/water relationships during kiln drying and reconditioning of softwoods
The Forest and Wood Products Research and Development Corporation (“FWPRDC”) makes no warranties or assurances with respect to this publication including merchantability, fitness for purpose or otherwise. FWPRDC and all persons associated with it exclude all liability (including liability for negligence) in relation to any opinion, advice or information contained in this publication or for any consequences arising from the use of such opinion, advice or information.
This work is copyright under the Copyright Act 1968 (Cth). All material except the FWPRDC logo may be reproduced in whole or in part, provided that it is not sold or used for commercial benefit and its source (Forest and Wood Products Research and Development Corporation) is acknowledged. Reproduction or copying for other purposes, which is strictly reserved only for the owner or licensee of copyright under the Copyright Act, is prohibited without the prior written consent of the Forest and Wood Products Research and Development Corporation.
Project no: PN04.2001 Researchers: R. Northway Ensis
Private Bag 10, Clayton South, Vic 3169 I. Burgar
CSIRO
Private Bag 33, Clayton South MDC, Vic 3169
Final report received by the FWPRDC in May 2007.
Forest and Wood Products Research and Development Corporation PO Box 69, World Trade Centre, Victoria 8005 T +61 3 9614 7544 F +61 3 9614 6822 E [email protected] W www.fwprdc.org.au
T in T out T surf T 3mm T 9mm T 15mm T 21mm T - 9mm
(c) Air and wood temp. during Run 6 (d) Air and wood temp. during Run 7
Figure 11 Drying plots for Run 6 and Run 7
For Run 6, DBT was held constant for the 4 hour drying period; for Run 7 DBT was ramped
down to 120C after 2 hours and 20 minutes for 5:40 drying time. WBT was reached much
more quickly in Run 7 than in Run 6. For both runs 3 hours steaming was given.
MC Profiles Run Rh6
0
2
4
6
8
10
12
0 1 5 8 9
Slice
MC
, %
RH1Ir RH1Gd RH1Gr RH1Cr RH1Fr
MC Profiles Run Rh7
0
2
4
6
8
10
12
0 1 5 8 9
Slice
MC
, %
RH1Hd RH1Hr RH1Ed RH1Er RH1Dr
(a) MC profiles for Run 6 (b) MC profiles for Run 7
Figure 12 Profiles of moisture content through the thickness of specimens, generally radial, determined by oven drying, for Run 6 and Run 7.
Slices 0, 1, 8 & 9 are 6mm thick, slice 0 from the outer face and slice 9 from the inner face
(near pith). Labels ending d = at the end of drying; r = after steam reconditioning and cooling
overnight.
Figure 12 shows that moisture content is more uniform through the thickness after steaming.
18
Run 6 NMR measurements
0
1
2
3
4
5
6
7
8
9
10
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
plitu
de
Rh1IR* Rh1GD Rh1GR Rh1CR Rh1FR
Run 6 NMR measurements
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
1/T
2 m
s-1
Rh1IR* Rh1GD Rh1GR Rh1CR Rh1FR
Run 7 NMR measurements
0
2
4
6
8
10
12
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
pli
tu
de
Rh1HD* Rh1HR* Rh1ED Rh1ER Rh1DR
Run 7 NMR measurements
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice1
/T2
m
s-1
Rh1HD* Rh1HR* Rh1ED Rh1ER Rh1DR
(a) NMR profiles for Run 6 (b) NMR profiles for Run 7
Figure 13 Low field MOUSE NMR measurements on slices from cores through the thickness
of specimens, generally radial, for Run 6 and Run 7.
Slices 0, 1, 8 & 9 are 5mm thick (sawn at 6mm spacing).
Comparison of Figure 12 and Figure 13 shows that amplitude generally mirrors the moisture
content determined by oven drying.
Steaming results in more mobile water near the wood surfaces, indicated by the elevated 1/T2
in surface slices.
MC Run 6
0
5
10
15
20
25
30
35
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh1I Rh1G Rh1F Rh1C
MC Run 7
0
5
10
15
20
25
30
35
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh1H Rh1E Rh1D
(a) MC records for Run 6 (b) MC records for Run 7
Figure 14 Average specimen moisture content of specimens of Run 6 and Run 7 during drying
and later environmental cycling.
G = Green; D = Dry; F = final (Reconditioned and cooled or cooled); T3 = Condition before humidity cycling; 12% = after 1 week at 12% EMC; 17% = after 1 week at 17% EMC; 5% = after 1 week at 5% EMC; 12% = after 1 week at 12% EMC;
19
Twist Run 6
-5.5
-5
-4.5
-4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Rh1I Restrained Rh1G Weighed Rh1F Rh1C T/Cs
Twist Run 7
-5.5
-5
-4.5
-4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Rh1H Restrained Rh1E Weighed Rh1D T/Cs
(a) Twist records for Run 6 (b) Twist records for Run 7
Figure 15 Twist of specimens of Run 6 and Run 7 during drying and later environmental
cycling.
G = Green; D = Dry; F = final (Reconditioned and cooled or cooled); T3 = Condition before humidity cycling; 12% = after 1 week at 12% EMC; 17% = after 1 week at 17% EMC; 5% = after 1 week at 5% EMC; 12% = after 1 week at 12% EMC;
Effect of drying conditions and restraint
• The amount of twist during drying varied between specimens. As can be seen in the
photos below, the specimens with the greatest twist, Rh1C and D has pith at one face and
similar knots. Specimen Rh1G twisted much more than E, possibly because it included a
loop of pith.
• Specimen Rh1H increased in twist by almost 50% over the cycling whereas Rh1I did not.
Rh1H had slightly lower surface moisture content after drying and increased in mass a
little more up to the 17% condition. This would be expected to have reduced twist.
• All other specimens showed little change in twist over the cycling, showing no difference
between the two runs.
• Comparing the specimens restrained against twist, I and H, with the most closely
matching unrestrained specimens,Rh1E and F, show no apparent effect of the restraint.
Figure 16 Specimens of Run 6 and Run 7 after tests with ends trimmed. Specimens are in a
sequence A, B…I, cut from Radiata Pine heartwood Board 1
20
Detailed drying run 15 – radiata pine near-pith sapwood
For Run 15 kiln temperature ramped down in two stages and no steam reconditioning was
(a) Dryer conditions for Run 15 (b) Air and wood temperatures during Run 15
Figure 17 Drying records for Run 15
MC Profiles Run Rsh15
0.0
5.0
10.0
15.0
20.0
25.0
0 1 5 8 9
Slice
MC
, %
Rsh3Df Rsh3Gf Rsh3Ff Rsh3Cf
Figure 18 Profiles of moisture content from sliced cores through the thickness of specimens,
generally radial, determined by oven drying, for Run 15.
Slices 0, 1, 8 & 9 are 6mm thick, slice 0 from the outer face and slice 9 from the inner face (near pith). Labels ending d = at the end of drying; r = after steam reconditioning and cooling overnight; f = final where no reconditioning applied
21
g
Run 15 Restrained vs Non-restrained
0
2
4
6
8
10
12
14
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
plitu
de
Rsh3DF* Rsh3CF
Run 15 Restrained vs Non-restrained
0
2
4
6
8
10
12
14
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
plitu
de
Rsh3DF* Rsh3CF Rsh3DF* 12% Rsh3CF 12%
Run 15 Restrained vs Non-restrained
0
1
1
2
2
3
3
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice1/T
2 -
m
s-1
Rsh3DF* Rsh3CF Rsh3DF* - 12% Rsh3CF - 12%
Run 15 Restrained vs Non-restrained
0
1
1
2
2
3
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
1/T
2 -
m
s-1
Rsh3DF* Rsh3CF
Figure 19 Low field MOUSE NMR measurements on slices from cores through the thickness
of specimens, generally radial, for Run 15.
Slices 0, 1, 8 & 9 are 5mm thick (sawn at 6mm spacing).
• The moisture content profiles from Run 15 show that ramping down without steaming
after drying resulted in significant gradients; final moisture contents were high, however.
MC Run 15
0
10
20
30
40
50
60
70
80
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rsh3D Rsh3B Rsh3C
MC Run 15
0
10
20
30
40
50
60
70
80
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rsh3G Rsh3E Rsh3F
Twist Run 15
-1
-0.5
0
0.5
1
1.5
2
2.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Rsh3D Restrained Rsh3B Cored Rsh3C T/Cs
Twist Run 15
-0.5
0
0.5
1
1.5
2
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Rsh3G Restrained Rsh3E Cored Rsh3F Weighed
Figure 20 Average specimen moisture content and twist of specimens of Run 15 during
drying and later environmental cycling.
22
• Although most specimens dried further between stages F and T3, twist did not change to
any extent.
• Only Specimen Rsh3D increased significantly in twist during cycling.
Figure 21 Specimens of Run 15 after drying and cycling.
Note: Ends trimmed to show grain patterns.
23
Detailed drying runs 16 & 17: radiata pine heartwood
For Run 16 kiln conditions were ramped down in two stages and no steam reconditioning was
used; Run 17 followed the full Radiata test schedule.
SWW Run Rh16
0
20
40
60
80
100
120
140
160
0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00
Drying Time, hrs
Tem
pera
ture
, C
0
5
10
15
20
25
30
35
40
MC
%
DBT WBT Corr. Stack MC Specimen AvMC
SWW Run Rh17
0
20
40
60
80
100
120
140
160
0:00 1:00 2:00 3:00 4:00 5:00
Drying Time, hrs
Te
mp
era
ture
, C
0
5
10
15
20
25
30
35
40
MC
, %
DBT WBT Corr. Stack MC Specimen AvMC
(a) Dryer conditions for Run 16 (b) Dryer conditions for Run 17
(c) Air and wood temperatures during Run 16 (d) Air and wood temperatures during Run 17
Figure 22 Drying records for Run 16 and Run 17
MC Profiles Run Rh16
0
5
10
15
0 1 5 8 9
Slice
MC
, %
Rh2Ff Rh2Af Rh6Cf Rh6Gf
MC Profiles Run Rh17
0
5
10
15
20
0 1 5 8 9
Slice
MC
, %
Rh6Dd Rh6Dr Rh6Hr
MC Profiles Run Rh17
0
5
10
15
20
0 1 5 8 9
Slice
MC
, %
Rh2GD Rh2Gr Rh2Bd Rh2Br
(a) MC profiles for Run 16 (b) MC profiles for Run 17
Figure 23 Profiles of moisture content through the thickness of specimens, generally radial,
determined by oven drying, for Run 16 and Run 17.
Slices 0, 1, 8 & 9 are 6mm thick, slice 0 from the outer face and slice 9 from the inner face (near pith). Labels ending d = at the end of drying; r = after steam reconditioning and cooling overnight; f = final where no reconditioning applied.
24
Run 16 Restrained vs Non-restrained
0
2
4
6
8
10
Slice 0 Slice 1 Slice 7/5 Slice 8 Slice 9
Slice number
Am
pli
tud
eRh6GR Rh6CR*
Run 16 Restrained vs Non-restrained
0
2
4
6
8
10
Slice 0 Slice 1 Slice 7/5 Slice 8 Slice 9
Slice number
Am
pli
tud
e
Rh2AR Rh2FR*
Run 16 Restrained vs Non-restrained
0
1
2
3
4
Slice 0 Slice 1 Slice 7/5 Slice 8 Slice 9
Slice number
1/T
2 -
ms
-1
Rh6GR Rh6CR*
Run 16 Restrained vs Non-restrained
0
1
2
3
4
5
6
Slice 0 Slice 1 Slice 7/5 Slice 8 Slice 9
Slice number
1/T
2 -
ms
-1
Rh2AR Rh2FR*
(a) NMR measurements for Run 16
Run 17 Restrained vs Non-restrained
0
2
4
6
8
10
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
plitu
de
Rh6DR* Rh6HR Rh6DD*
Run 17 Restrained vs Non-restrained
0
1
2
3
4
5
6
7
8
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
Am
plitu
de
Rh2GR* Rh2BR Rh2GD* Rh2BD
Run 17 Restrained vs Non-restrained
0.00
1.00
2.00
3.00
4.00
5.00
Slice 0 Slice 1 Slice 7/5 Slice 8 Slice 9
Slice
1/T
2 m
s-1
Rh2GR* Rh2BR Rh2GD* Rh2BD
Run 17 Restrained vs Non-restrained
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice
1/T
2 m
s-1
Rh6DR* Rh6HR Rh6DD*
(b) NMR measurements for Run 17
Figure 24 Low field MOUSE NMR measurements on slices from cores through the thickness
of specimens, generally radial, for Run 16 and Run 17.
Slices 0, 1, 8 & 9 are 5mm thick (sawn at 6mm spacing).
• Run 17 shows that moisture content becomes more uniform through the thickness from
steaming.
• Steaming results in more mobile water near the wood surfaces, indicated by the elevated
1/T2 in surface slices, slices 0 and 9.
• The moisture content profiles from Run 16 were not excessive; ramping down can
minimise gradients.
25
MC Run 16
0
5
10
15
20
25
30
35
40
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh6C Rh6E Rh6G
MC Run 17
0
5
10
15
20
25
30
35
40
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh6D Rh6F Rh6H
MC Run 16
0
5
10
15
20
25
30
35
40
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh2F Rh2D Rh2A
MC Run 17
0
5
10
15
20
25
30
35
40
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Rh2G Rh2E Rh2B
Figure 25 Average specimen moisture content of specimens of Run 16 and Run 17 during
drying and later environmental cycling
Twist Run 16
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees/4
50m
m
Rh6C Restrained Rh6E Cored Rh6G T/Cs
Twist Run 17
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees/4
50m
m
Rh6D Restrained Rh6F Cored Rh6H T/Cs
Twist Run 16
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees/4
50m
m
Rh2F Restrained Rh2D Cored Rh2A Weighed
Twist Run 17
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees/4
50m
m
Rh2G Restrained Rh2E Cored Rh2B Weighed
Figure 26 Twist of specimens of Run 16 and Run 17 during drying and later environmental
cycling
26
• Moisture content estimates were particularly difficult for cored specimens; the actual MC
of Rh6E and Rh2D were higher than plotted. The environment in the laboratory is close
to 10% EMC.
Effect of drying conditions and restraint
• Twist changes generally followed moisture changes with environmental changes.
• The restrained specimens in both runs showed increased twist over the cycling process.
• There is apparently more distortion in Run 17 specimens, but this is not consistent and
may be due to variation in wood characteristics along the boards.
Figure 27 Specimens of Run 16 and Run 17 after tests with ends trimmed.
Specimens are in a sequence A, B…I, cut from Radiata Pine heartwood Boards 2 and 6
27
Detailed drying test runs 10, 11 and 12: slash pine heartwood
Runs 10 and 11 differ only in the shorter steaming time for Run 11; Run 12 had no steaming after
two stages of ramp-down of temperature, and with humidity increased for the final stage (Figure 28).
T in T out T surf T 3mm T 9mm T 15mm T 21mm T -9mm
(c) Air and wood temp. during Run 13 (d) Air and wood temp. during Run 14
Figure 39 Drying records for Run 13 and Run 14
For Run 14 temperature was ramped down in two stages and no steam reconditioning was
used; Run 13 followed the Slash test schedule.
MC profiles Run Sh13
0.0
5.0
10.0
15.0
20.0
25.0
0 1 5 8 9
Slice
MC
, %
Sh5Cd Sh5Cr Sh5Ed Sh5Er
MC Profiles Run Sh14
0
5
10
15
20
25
0 1 5 8 9
Slice
MC
, %
Sh5Dd Sh5Dr Sh5Fd Sh5Fr
(a) MC profiles for Run 13 (b) MC profiles for Run 14
Figure 40 Profiles of moisture content from sliced cores through the thickness of specimens
5, generally radial, determined by oven drying, for Run 13 and Run 14
Slices 0, 1, 8 & 9 are 6mm thick, slice 0 from the outer face and slice 9 from the inner face (near pith). Labels ending d = at the end of drying; r = after steam reconditioning and cooling overnight; f = final where no reconditioning applied.
32
Runs 13 NMR measurements
0.0
0.5
1.0
1.5
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
1/T
2 -
ms
-1
Sh5CD* Sh5CR* Sh5ED Sh5ER
Runs 13 NMR measurements
0
2
4
6
8
10
12
14
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
Am
plitu
de
Sh5CD* Sh5CR* Sh5ED Sh5ER
Runs 14 NMR measurements
0
2
4
6
8
10
12
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
Am
plitu
de
Sh5DD* Sh5DR* Sh5FD Sh5FR
Runs 14 NMR measurements
0.0
0.5
1.0
1.5
2.0
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
1/T
2 -
ms
-1
Sh5DD* Sh5DR* Sh5FD Sh5FR
(a) NMR profiles for Run 13 (b) NMR profiles for Run 14
Figure 41 Low field MOUSE NMR measurements on slices from cores through the thickness
of specimens 5, generally radial, for Run 13 and Run 14.
Slices 0, 1, 8 & 9 are 5mm thick (sawn at 6mm spacing).
MC profiles Run Sh13
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 1 5 8 9
Slice
MC
, %
Sh6Bd Sh6Br Sh6Dr
MC Profiles Run Sh14
0
2
4
6
8
10
12
14
16
0 1 5 8 9
Slice
MC
, %
Sh6Cd Sh6Cr Sh6Er
(a) Oven-dry MC profiles for Run 13 (b) Oven-dry MC profiles for Run 14
Figure 42 Profiles of moisture content from sliced cores through the thickness of specimens
6, generally radial, determined by oven drying, for Run 13 and Run 14
Slices 0, 1, 8 & 9 are 6mm thick, slice 0 from the outer face and slice 9 from the inner face (near pith). Labels ending d = at the end of drying; r = after steam reconditioning and cooling overnight; f = final where no reconditioning applied.
33
Runs 13 NMR measurements
0
2
4
6
8
10
12
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
Am
plitu
de
Sh6BD* Sh6BR* Sh6DR
Runs 13 NMR measurements
0.0
0.5
1.0
1.5
2.0
2.5
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
1/T
2 -
ms
-1
Sh6BD* Sh6BR* Sh6DR
Runs 14 NMR measurements
0
2
4
6
8
10
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
Am
plitu
de
Sh6CD* Sh6CR* Sh6ER
Runs 14 NMR measurements
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Slice 0 Slice 1 Slice 5/7 Slice 8 Slice 9
Slice number
1/T
2 -
ms-1
Sh6CD* Sh6CR* Sh6ER
(a) NMR profiles for Run 13 (b) NMR profiles for Run 14
Figure 43 Low field MOUSE NMR measurements on slices from cores through the thickness
of specimens 6, generally radial, for Run 13 and Run 14.
Slices 0, 1, 8 & 9 are 5mm thick (sawn at 6mm spacing).
MC Run 13
0
10
20
30
40
50
60
70
80
90
100
110
120
130
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Sh5C Sh5A Sh5E
MC Run 14
0
10
20
30
40
50
60
70
80
90
100
110
120
130
G D F T3 12% 17% 5% 12%Condition
MC
, %
Sh5D Sh5B Sh5F
(a) Av. Specimen MC for Run 13 (b) Av. Specimen MC for Run 14
Figure 44 Average MC of specimens from board 5 in Run 13 and Run 14 during drying and
later environmental cycling.
Twist Run 13
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Sh5C Restrained Sh5A Cored Sh5E Weighed
Twist Run 14
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
de
gre
es
Sh5D Restrained Sh5B Cored Sh5F Weighed
(a) Specimen twist for Run 13 (b) Specimen twist for Run 14
Figure 45 Twist of specimens from board 5 in Run 13 and Run 14 during drying and later
environmental cycling.
34
MC Run 13
0
10
20
30
40
50
60
70
80
90
100
110
120
130
G D F T3 12% 17% 5% 12%Condition
MC
, %
Sh6B Sh6F Sh6D
MC Run 14
0
10
20
30
40
50
60
70
80
90
100
110
120
130
G D F T3 12% 17% 5% 12%
Condition
MC
, %
Sh6C Sh6G Sh6E
(a) Av. Specimen MC for Run 13 (b) Av. Specimen MC for Run 14
Figure 46 Average MC of specimens from board 6 in Run 13 and Run 14 during drying and
later environmental cycling
Twist Run 13
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
deg
rees
Sh6B Restrained Sh6F Cored Sh6D T/Cs
Twist Run 14
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
G D F T3 12% 17% 5% 12%
Condition
Tw
ist,
de
gre
es
Sh6C Restrained Sh6G Cored Sh6E T/Cs
(a) Specimen twist for Run 13 (b) Specimen twist for Run 14
Figure 47 Twist of specimens from board 6 in Run 13 and Run 14 during drying and later
environmental cycling
Figure 48 Specimens of Run 13 and Run 14 after tests with ends trimmed.
Specimens are in a sequence A, B…I, cut from Slash Pine heartwood Boards 5 and 6
35
NMR measurements and analysis
NMR Surface Analyser
The NMR MOUSE (MObile Universal Surface Explorer) is a novel NMR device designed for
relaxation measurements on surfaces of arbitrary shaped samples. The design of the mobile
probe with the permanent magnets and transmitter and receiver coil system allows measurements
of the proton NMR signals (15.9MHz) from various samples.
The NMR surface analyzer, MOUSE, detects the signal from water via the special imaging
sequence, spin echo, which gives the NMR signal predominantly of water close to the surface.
The observation volume is about 1mm in depth over a 0.5cm2. The observed signal is detected
only from the moisture with a sufficient mobility or a long enough relaxation time, but not the
signal from the protons of solid organic molecules, whose relaxation time is in general shorter.
The data are presented as probability 1/T2 for water to be bound and the amplitude reflects the
overall number of protons (water molecules) detected. The scale is given in relative terms and is
proportional to total moisture content.
High field Solid-state NMR spectroscopy
The solid wood samples were investigated by solid-state proton NMR spectroscopy - Varian
Unity Plus 300 MHz NMR spectrometer. The method used was the proton magic angle spinning
NMR spectroscopy, which in a typical spectrum of a wood sample would normally give the
broad water resonance band (bound water) and small but sharp signals from the organic
molecules. By using a standard spin-echo pulse sequence (CPMG – sequence) we found that the
water signal disappears faster with the increased spin-echo time than the organic signal. The
standard CPMG sequence was used in order to determine the relaxation time T2 of these two
components, and the inversion recovery sequence, 180°-90°, was used to determine the spin
lattice relaxation time T1.
Model for wood drying NMR parameters
The general concept reported in 1999 (PN008.96 Softwood Drying Research Project) for NMR
signal of water in wood was adopted as follows:
“Studies of T2 relaxation of water suggest:
essentially unbound water in cell lumens,
water associated with walls and smaller reservoirs – droplets & cracks
tightly bound cell wall water.
Short T2 ranges from 0.2 ms to 1 ms, medium T2 ranges from 5 ms to 40 ms, and long T2
ranges from 50 ms to 800 ms.”
In the current investigation of the drying process and moisture effects on wood structural
changes this model and water categories were investigated in detail in order to establish the
specific moisture deformation correlations where previous data was not conclusive. In such a
study we found it essential to investigate first the limitations of the above definition and to
suggest a more comprehensive description of the drying process yielding potentially better
control of this process for the industry. The investigation was conducted in two directions: first
the definitions and physical differentiation (especially by NMR parameters) between the
proposed categories of water in wood, and second the more comprehensive analysis of NMR
spectral data in the view of the above definitions.
1) The general definition of water in the hydrophilic polymers is correctly stated in the previous
report and it can be found in general literature on this subject where water content, Wc, is a sum
36
of non-freezing water (structural water or tightly bound to cell walls), Wnf, freezing bound water
or water associated with walls and smaller reservoirs, Wfb, and free water or essentially unbound
water in cell lumens, Wf.10,11
The relationship between these categories was studied closely and
the characteristic finding by H & T Hatakeyama can be graphically summarised as shown here11
.
Figure 49 from Hatakeyama11
This diagram indicates that in the dry wood samples where MC is around 10% one can expect
mainly bound water (freezing and non-freezing). Such interpretation was given by R Guzenda et
al.3 at the 12th International symposium on non-destructive testing of wood, Hungary 2000, who
concluded that lost of free water in the drying process (moisture going below the fibre saturation
point) is evident from NMR data where the relaxation spin-lattice time T1, which reveals two
components in green wood, becomes a single component parameter below the fibre saturation
point. In our investigation we quickly confirmed the above conclusion (milestone report) but
investigating further the spin-spin relaxation time T2 we also expected to see some difference
between the bound waters (freezing and non-freezing), especially in correlation to the drying
regime and to conditioning after drying. But again as in the previous project report we reached a
similar conclusion:
“There are large and inconsistent variations in the proportions of water in each bonding state for
different samples inconsistent with respect to other parameters or drying conditions. Lesser
variability was observed after further conditioning of the wood and proportion of bound water
was higher than for freshly dried samples - moisture content after drying exhibited large
variation from matched samples, masking any consistent relationship”.
This fact by itself was leading us to the assumption that in the proton NMR spectra we are
observing too large variations due to the additional species such as resin and other wood
components, which possess the protons in their molecular structure (as already indicated in the
previous milestone report). The results of a detailed investigation have now been incorporated
into the improved model for NMR data interpretation as outlined below.
2) The assumption that the proton NMR signal analysed in the wood is mainly coming from water
in different states of interaction with the solid matrix needs to be revised. The water definition
10
S.L. Maunu, NMR studies of wood and wood products, Progress in Nuclear Magnetic Resonance Spectroscopy, 40 (2002) 151-174 11
H. Hatakeyama and T. Hatakeyama, Interaction between water and hydrophilic polymers, Thermometrica Acta, 308 (1998) 3-22
37
statement in the first report assumed that the whole proton NMR signal, as analysed by the T2
measurements using FID (free induction decay) method, is coming from water in the wood. This
simplified picture is not always correct because the protons from other components of wood such
as cellulose, hemicellulose and lignin (major ones) may appear in the signal as well, especially
when their molecular mobility is comparable to the mobility of water molecules. Proton NMR
polarisation, which was investigated by NMR methods in wood samples, is therefore from both
water molecules and organic components (cellulose, hemicellulose and lignin). These protons can
be detected as separate species but they also interact at the interfaces yielding the additional NMR
polarisation exchange, which further changes the proton NMR signal.
The dried sample 2AR (Run 16) is used here as an example where the difference between two
slices is evident in the signal obtained by the NMR MOUSE instrument in around 21/2 minutes.
The dried wood samples all have a short T2 and only one T1 component but in these two spectra,
Figure 50 below, it is obvious that we have a longer and shorter component in signal decay. The
MOUSE detection utilizes the spin-echo sequence at a certain time where all the signal from
materials with the shorter T2 component (for solid wood) is already gone (relaxed) and only the
mobile part of the organic components can be detected. It was assumed that the longer T2
component belongs to the water molecules alone, but it is evident from our investigation that
such organic components do exist and they are likely to be organic extractable compounds or
mobile parts of cellulose/hemicellulose polymers.
(a)
20
15
10
5
0
Sig
na
l
160140120100806040200
Time(ms)
Run 16 sample 2AR slice 5
(b)
20
15
10
5
0
Sig
na
l
30252015105
Time(ms)
Run 16 sample 2AR slice 8
Figure 50 The NMR signal, MOUSE, from two slices (a) with pith and (b) without.
In order to separate different molecular species in the proton NMR spectrum one needs to use the
high field solid-state NMR methods. Fast magic angle spinning is one of them12,13
and it can
differentiate the resonances of water and other species. These methods are most effective when
the magnetic environment of protons is sufficiently different (chemical shift difference) and they
are not fast exchanging NMR polarisation. Recent work of A.M. Gill and co-workers13
shows
that proton high-resolution magic angle spinning spectra of natural polymeric materials like
wood yield sufficient information to identify the proton resonances in these materials. Their
conclusion, spectral assignment, is presented in Table 3 below:
12
R. Guzenda, W.Olek, H.M. Baranowska, Identification of free and bound water content in wood by means of NMR relaxometry, 12
th International symposium on non-destructuve testing of wood (2000),
Sopron, Hungary 13
A.M. Gill, M.H. Lopes, C. Pascoal Neto, and J. Rocha, Very high-resolution 1H MAS NMRE spectra of
natural polymeric material, Solid State Nuclear Magnetic Resonance, 15 (1999) 59-67
38
Table 3 From Gill et al13
Their spectral interpretation and assignment did help us to compare our solid-state NMR spectra
with theirs and to identify the additional species detected by NMR method in the proton
spectrum. 1H MAS NMR spectra of lignin, cellulose and hemicellulose, as shown below, show
the different spectral appearance of these species13
.
Figure 51 From Gill et al13
39
Solid-state NMR analysis
The carbon CPMAS (cross-polarisation magic angle spinning) spectrum of wood, sample 1BR from
Radiata Run RH7, is presented in Figure 52. This spectrum is a characteristic spectrum of wood
samples revealing cellulose, hemicellulose and lignin components.1,5 The central resonances from
50 ppm to 110 ppm are clearly from cellulose, crystalline and amorphous, as well as hemicellulose.
The broad resonance peaks from 110 ppm to 160 ppm are mainly from the lignin aromatic carbons.
The standard proton MAS (magic angle spinning) spectrum, Figure 52, reveals mainly a large
and broad water signal – bound water (3 ppm to 8ppm) – and small additional peaks due to the
organics components in wood (probably cellulose, hemicellulose, and lignin). These resonances,
sharp peaks at 1.2 ppm and 1.6 ppm, and smaller at 1.9 ppm and 2.3 ppm, in comparison with the
spectra from reference 13 are most likely due to the cellulose/hemicellulose species.
In order to further identify these resonances in the proton MAS spectrum it is proper to recall the
general wood structure model. The general model for the cell walls in the wood predicts that it is
comprised of cellulose polymers, which form microfibrils, that are bound together by
hemicellulose. A low, or reduced, hemicellulose content has been associated with the high
dimensional stability of wood structure. It may then be postulated that hemicellulose in particular
in the hydration state (plasticised by water) will be the main contributor in relation to the wood
distortion. Therefore, if in our 1H NMR MAS spectra we are able to observe and quantify the
hemicellulose that is hydrated (in physical contact with the bound water), we may have the
detection tool to estimate the potential distortion in the wood after drying.
The 1H MAS NMR spectra were obtained at 300MHz and a magic angle spinning speed of
around 9000 Hz. According to the literature13
only very high spinning speeds can provide the
sufficient line narrowing (30 kHz) for such resolution. It may be quickly assumed that we are
observing a broad line of water and narrow lines from cellulose due to the fact that the bound
water and hemicellulose interact (fast exchanging the magnetic polarisation) due to direct
contact. Therefore further relaxation studies of T1 and T2 for both species is required to confirm
their contact.
The spin-lattice relaxation time, T1, determined for the water proton peak, resonance at 4.8 ppm, and
the peak at 1.6 ppm (-CH2- in cellulose and hemicellulose) are nearly the same at 300 MHz; 324 ms
and 334 ms respectively. This may be interpreted as confirmation that water and ‘hemicellulose’
ends are closely associated – exchanging the magnetic polarisation in the NMR time scale.
The spin-spin relaxation time, T2, measured by the spin echo technique using a CPMG pulse
sequence, is on the other hand, quite different; around 1ms for water and 20 ms for cellulose.
This indicates that the bound water is in fast exchanging rate between immobile state (“non-
freezing” - attached to cell wall surface) and semi-free state in the pores (”freezing bound”),
whilst the organic protons (cellulose, hemicellulose and lignin) are different from the water
relaxation time. The solid crystalline component is expected to have a very short relaxation time,
but an amorphous more flexible structure may have quite a long T2 in comparison to the water.
So far the hemicellulose model of hydration is not in contradiction with the obtained results.
Using the spin-echo method with variable spin-echo time will further provide the separation of
these two components, bound water and organic signal, making a final better identification of
these organics. If we are using spin-echo detection with variable detection time, the NMR signal
from protons with a shorter T2 will disappear from the spin-echo signal before the proton signal
from the components with a long T2. Therefore, the presence of the organic protons in the signal
can be enhanced by making the abundant, but short-lived, water signal disappear.
In Figure 54, one can easily see the relative larger intensities from the organic components in
relation to the water. By using the spin-echo time, which is a few times longer than the water
relaxation time T2, completely removes the water signal from the NMR spectrum, Figure 55.
40
We used the spin echo technique with variable delays to define the relaxation time T2 of the
residual water (bound) signal and it was found to be always short, smaller than 1 ms. Therefore
one can conclude that we have a bound water signal that can not be differentiated further into
separate categories at room temperature by these NMR measurements. On the other hand, the
measurement of water T1 also reveals only one T1 component. The bound water, which may
come from two physical states, non-freezing bound water bonded to cell walls and freezing
bound water which is H-bonded to non-freezing bound water, seems to appear as one under the
room temperature and dry wood sample conditions. Of course the so-called “free” water was
evidently removed in the drying cycle.
Single component in T2 analysis indicates that the NMR signal from water molecules in the
“non-freezing” state and in the “freezing bound” state is averaged out due to fast exchange rate
between these molecules in the NMR time-scale (at room temperature) and strong spin-spin
interactions. The detected NMR signals (NMR is a bulk detection method of atomic magnetic
properties) are averaged in space and time in the NMR timescale through all different positions
or sites. This makes interpretation of bound and free water from NMR data alone more complex
due to the fact that relaxation times of the individual molecules may be different at different
positions in the wood structure.
It can also be assumed that there is a continuous distribution of the relaxation times T2 in the
‘certain’ range of values, but yielding only the average value in the final analysis as stated above.
ppm20406080100120140160180
Figure 52 The carbon CPMAS spectrum of wood – sample 1BR.
ppm-1-012345678910111213
Figure 53 The proton MAS spectrum of wood – sample 1BR.
41
ppm-1-0123456789101112
Figure 54 The proton MAS spectrum of sample 1BR using spin echo at 1 ms.
ppm-1-0123456789101112
Figure 55 The proton MAS spectrum of sample 1BR using spin echo at 4ms.
Figure 56 The proton MAS spectra of sample 1BR using different spin-echo times,
from 0.4ms to 2.4ms.
42
Dimensional changes in wood
Wood is an anisotropic material, that is, its dimensions change differently in three dimensions:
tangentially, radially and longitudinally. Tangential dimensional change has the highest rate of
change due to parallel orientation of microfibrils along the axis of the cell wall. Radial change is
the second largest and longitudinal change is negligible for most practical applications. In
general, dimensional change is expressed as a percentage as a function of initial dimension and
ratio of moisture content over fibre saturation point. This simple model denotes that the change
of moisture content in drying below fibre saturation point is proportional to the variation in the
shape.
Dimensional changes in wood could be calculated using the following formula14
100**
FSP
CMCOD
SV
DC=
where DC is dimensional change due to change in moisture content (CMC), OD is original
dimension, SV is shrinkage value from green to oven dry moisture content, and FSP is fibre
saturation point. Assuming that FSP is around 28% and it does not change much between the
softwood samples the rate of distortion change, DC/SV, is proportional to the rate of moisture
change, CMC/FSP.
On the other hand there is also a direct relationship between the density of wood and shrinkage values14
. Species with higher density shrink more than those with lower density. There have been many studies aimed at stabilizing the cell wall (resin treatment and alike) so that shrinkage of wood can be controlled, however none of these methods has been put into practical use due to economical and technical considerations
14.
Water dynamics in wood
Results of measuring the relaxation time T1 ( NMR solid-state methods)
The relaxation time T1 of water was measured in dry samples with high field solid state NMR
method in order to observe any differences which may arise from the difference in treatment (run
16 no steaming versus run 17 with 3 hour steaming) or in restrained versus non-restrained boards
during drying.
0.00
100.00
200.00
300.00
400.00
500.00
600.00
700.00
800.00
2A,2B 6G,6H 2F,2G 6C,6D
Run16 Run17
Figure 57 Effect of steaming: T1
14
Source: http://www.agweb.okstate.edu, “Dimensions Changes in Wood”, S. Hiziroglu
43
0.00
100.00
200.00
300.00
400.00
500.00
600.00
700.00
800.00
2A,2F 6G,6C 2B,2G 6H,6D
unrestrained restrained
Figure 58 Effect of restraint: T1
A similar comparison is given on the T2 data collected at high field solid-state NMR which
isolate the water resonance from the organic protons.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
2A,2B 6G,6H 2F,2G 6C,6D
Run16 Run17
Figure 59 Effect of steaming: T2
44
0.00
0.20
0.40
0.60
0.80
1.00
1.20
2A,2F 6G,6C 2B,2G 6H,6D
unrestrained restrained
Figure 60 Effect of restraint: T2
Discussion *(solid-state data):
T1 data of water resonance at 300 MHz (MAS NMR spectroscopy) exhibit some kind of trend
between the runs 16 and 17, one without steaming and other with steaming, at the end of the
drying cycle. After steaming, the T1 becomes shorter, even more so for the boards having a
longer initial T1. It seems that steaming enhanced the exchange between different types of bound
water making relaxation T1 shorter. The restrain on the other hand is making the similar
enhancement in exchange and shorten T1.
T2 data, collected at 300 MHz, presented here are given here for water resonance by using spin
echo decay (CPMG method). Changes in these T2’s don’t show any significant correlation as
function of steaming or restrain. There may be some shortening trend in T2 with the restrain but
the body of data is too small to make this observation certain. More data were collected by
Mouse that can be analysed further in order to identify the correlations to drying conditions:
steaming or restrain.
MOUSE relaxation data
Mouse relaxation data on slices 0 and 1 have been averaged into one set (see Table 4 below).
They reveal some relationship to the steaming and non-steaming drying regime as well as to the
restraint and non-restraint condition during drying.
45
Table 4 The average amplitude and T2 for slices 0 and 1 of samples (run16 and 17).
Samples Condition Comparison Amplitude T2(ms)
2AR-2BR R Steaming -/+ 4.90 ; 6.11 0.457 ; 0.547
12% 4.79 ; 5.24 0.452 ; 0.510
17% 6.58 ; 7.27 0.642 ; 0.680
2FR-2GR R Steaming -/+ 5.26 ; 5.19 0.345 ; 0.538
12% 5.98 ; 5.08 0.360 ; 0.478
17% 5.53 ; 6.63 0.578 ; 0.608
2AR-2FR R Restrained -/+ 4.90 ; 5.26 0.457 ; 0.345
12% 4.79 ; 5.98 0.452 ; 0.360
17% 6.58 ; 5.53 0.642 ; 0.578
2BR-2GR R Restrained -/+ 6.11 ; 5.19 0.547 ; 0.538
12% 5.24 ; 5.08 0.510 ; 0.478
17% 7.27 ; 6.63 0.680 ; 0.608
6GR-6HR R Steamed -/+ 3.77 ; 5.97 0.375 ; 0.508
12% 5.13 ; 5.95 0.412 ; 0.540
17% 6.13 ; 7.12 0.622 ; 0.720
6CR-6DR R Steamed -/+ 5.36 ; 4.68 0.330 ; 0.778
12% 5.19 ; 4.97 0.378 ; 0.635
17% 5.36 ; 7.28 0.648 ; 0.748
6GR-6CR R Restrained -/+ 3.76 ; 5.36 0.375 ; 0.330
12% 5.13 ; 5.19 0.412 ; 0.378
17% 6.13 ; 5.36 0.622 ; 0.648
6HR-6DR R Restrained -/+ 5.97 ; 4.68 0.508 ; 0.778
12% 5.95 ; 4.97 0.540 ; 0.635
17% 7.12 ; 7.28 0.720 ; 0.748
Discussion (MOUSE data):
The data from un-steamed, Run 16, versus steamed, Run 17 of the same board always gives the
longer T2 component for steamed samples. The expected higher amplitude for the steam samples
is not always detected as such, because there are other factors (structural) that influence the
amount of re-adsorbed moisture by steaming.
The data or non-restrained and restrained boards shows less correlation, but it still can be said
that on average, the relaxation time T2 is generally shorter for restrained board samples in
comparison to the same board non-restrained sample. However, on the other hand, the signal
amplitude shows no correlation to the changes between the restrained and non-restrained regime
indicating that other (structural) factors prevail in determining the total amount of bound water in
the samples.
Correlation between high field solid-state NMR data and low field Mouse data
Assuming that water in our dried wood samples is mainly bound water that plays a structural role
(non-freeze bound water) and small pores role (freeze bound water), it can be accepted that first
type of water will have a much shorter T2 than the other one due to the difference in the
molecular mobility. The measured relaxation time is therefore the average between these two as
defined by the following relationship:
fbnfT
pT
pT
+=222
1)1(
11
Steaming does increase the contribution of the second term, which results in the overall longer
T2. If a similar relationship is used for the T1, the steaming should increase the second part as
well, but the overall effect, as experimentally determined, is a shorter T1 after steaming. The
46
possible explanation for this discrepancy between the two relaxation times behaviour can be
found in the theory of the relaxation time and phase of the materials.
T1
T2
log (T1,2)
log ( c)o c = 1
Figure 61 Relaxation times T1 and T2 as a function of mobility correlation time c
It can be seen that the difference between the measured T1 and T2 values indicate already that we
are at the bottom of the T1 curve or even to the right side of it, where small changes in T2 (larger
correlation time tc) means larger changes in T1. Or in other words, the water correlation time is
longer than 10-8
s, indicating bound water.
47
Part 3 EXPERIMENTAL KILN TRIALS
Trials were conducted in an experimental kiln with Radiata Pine corewood to determine the
extent to which NMR could explain the differences in wood behaviour from different drying
schedules, particularly those identified in the detailed tests as showing equivalent stability, i.e.
using a longer kiln schedule with ramped-down temperature and no final steaming compared to a
standard high temperature schedule with steaming after drying. Boards were monitored for three
weeks after processing, with regular measurements of moisture content, shape and stiffness and
sampling for NMR measurements. These experimental kiln trials are described in Ensis Client
Report No. 1678.
MATERIALS AND METHODS
Procedure
One pair of kiln runs was conducted:
Material
Radiata pine boards from typical resources were provided by Green Triangle Forest Products:
Approximately (120) “heart in” (HI) 100 x 40mm boards, 6.0m long, were provided, plastic
wrapped and trucked to Clayton. As these were from production HI material there was
considerable variation in board characteristics; many boards were partly sapwood, often at one
end.
Preparation
Boards were selected by grain pattern, near-pith or with pith included, and with uniform
grain pattern along the length.
(72) boards were cut to each produce (2) 2.8m long end-matched specimens.
Cross-sections were cut from each end and at the centre of each board, and moisture
content (MC) and basic density (BD) were determined by oven drying.
The 2.8m long specimens were end-coated and weighed.
Acoustic wave velocity (AWV) was measured on the green specimens.
The paired specimens were allocated alternately to (2) kiln loads to ensure equal numbers
of butt and top specimens in each stack.
Stack
6 boards wide; 12 boards high; full length in kiln
25mm thick stickers
2.8m long x 1.2m wide x 200mm thick concrete weight – equivalent to 400mm thick
weight (800kg/m3 stack top area)
top baffle to weight
Kiln runs
Conditions
As for detailed test runs 16 & 17. (Part 2)
48
Table 5 Drying conditions for Run 1 and Run 2 of the kiln trials
Drying Cooling Steaming Cooling
Run 1 140/90 for 4+ hours outside kiln 1 hour
4 hrs outside kiln with weight
Run 2 140/90 for 3 hours; ramp to 120/90 over 1 hour; 120/90 for 1 hours; 110/95 for 1 hour; total drying 6 hours
none none outside kiln
Monitoring
DBT and WBT in the kiln were monitored during drying; wood temperature near the core
of several boards was monitored in each run during drying, steaming and cooling. Boards
for temperature monitoring were selected from near the left, centre and right sides of the
stack, looking from the door.
Restraint
The stack weight was left on the stack overnight (at least 12 hours after drying).
Measurements after drying
Mass and AWV of 2.8m length specimens were measured.
All specimens were cut to 2.4m length and sections taken for MC determination; the
sections removed were from the end cut to separate the specimen from its pair.
Mass & AWV of the 2.4m length specimens were measured. Distortion was measured
over 1.2m gauge length for each half-length and the centre half-length; measurements
were combined to give total distortion.
Boards from each run were re-stacked with stickers.
Stability during storage
Conditions
The two stacks of 2.4m long boards were stored for 21 days with stickers, without
restraint in a well-ventilated building.
Monitoring
Temperature and humidity were logged during storage
Weekly measurements
Mass and distortion were measured. Average moisture content for the specimens was
calculated from the mass, assuming the average green MC was the average of the MC at
each end. Where the section MCs were different, the calculation was done assuming that
the kiln-dried MC was that of the dry section cut after drying.
Sample cores for MC distribution and NMR measurements were taken from 12 boards
from each run (the two middle layers of the test stack).
Dynamic stiffness of core samples – ultrasound along and across grain – were not
measured as equipment could not be acquired.
AWV were measured,
Stiffness on flat, 3-point bending
49
RESULTS AND DISCUSSION
Kiln runs
The planned schedules were followed as detailed in Table 6. However, due to an error in
programming the kiln controller, the conditions were not logged for either Run 1 or Run 2. The
screen image at the end of Run 2 is reproduced in Figure 62.
Table 6 Kiln trial conditions and times
Drying Cooling Steaming Cooling Process time
Run 1 140/90 for 4+ hours outside kiln 1 hour
4 hrs outside kiln with weight
9 hours + cooling
Run 2 140/90 for 3 hours; ramp to
120/90 over 1 hour; 120/90 for 1 hours; 110/95 for 1 hour; total drying 6 hours
none none outside kiln 6 hours + cooling
Figure 62 Kiln controller screen for Run 2
Wood temperatures were logged for Run 1 from approximately halfway through drying (Figure
63) and for all of Run 2 (Figure 64).
50
Core wood temperature in 3 specimens during Kiln Trial 1
Figure 64 Core temperatures of three specimens during drying and cooling – Run 2
Environmental conditions
The conditions during the three weeks of monitoring are shown in Figure 65.
Environment in Factory
0
10
20
30
40
50
60
70
80
90
2-Sep-05 9-Sep-05 16-Sep-05 23-Sep-05
Tem
peratu
re, C
; R
H%
0
5
10
15
20
25
30
35
40
45
EM
C%
Temperature
RH%
EMC%
96 per. Mov. Avg. (EMC%)
A B C D
Figure 65 Environmental conditions for period of storage and monitoring
51
Moisture content
These runs followed quite closely the schedules used in the laboratory dryer tests. In each run
there were specimens from near sapwood boards with higher moisture content which skewed the
result. Specimens in Run 1 were slightly higher in MC than specimens in Run 2 (Figure 66).
Figure 66 and Figure 67 show the distribution of average moisture content and the change of average
moisture content during storage for each run. From Run 1 moisture content was higher than intended
and it reduced during storage to stage B and again to Stage D, as might be expected from the
environmental conditions (Figure 65) when starting with moist surfaces after steaming. Specimens in
Run 2 generally gained throughout storage, the initially drier surfaces taking up moisture.
The average moisture contents of the runs were significantly different throughout the period of
observation.
MC Distribution after drying - Run 1
0
10
20
30
40
50
60
5 10 15 20 25 More
MC% Category upper limit
No
. S
pe
cim
en
s Av. MC 14.2%
MC Distribution after drying - Run 2
0
10
20
30
40
50
60
5 10 15 20 25 More
MC% Category upper limit
No
. s
pe
cim
en
sAv. MC 12.8%
(a) Run 1 (b) Run 2
Figure 66 Moisture content of sections cut from all specimens after drying
Average MC of Specimens during storage
11.0
11.5
12.0
12.5
13.0
13.5
14.0
14.5
15.0
15.5
A B C D
Stage, weekly intervals
MC
%
Run 1
Run 2
#
# ##
$$
$ $
Figure 67 Average moisture content of all specimens in each run, weekly during storage. Points
which do not share the same symbol are significantly different (Scheffe Test).
The moisture content of the specimens from which NMR cores and MC cores were taken is shown in
Figure 68. Most specimens initially gained moisture, particularly those from Run 2 which were not
steamed after drying. Core MC is from 19mm cores cut adjacent to the 35mm cores cut for NMR, and
cut at the same time. The moisture content of cores was quite variable; some of the variation may be
52
from varying degrees of wetness in the centre of specimens. Some variation may be along the boards;
this variation may influence interpretation of the NMR measurements over storage time.
The paired specimens are presented here in a set labelled Set A. This has five pairs which were
well matched in initial moisture content and grain pattern.
(a) Av. MC of specimens
from mass
Average MC of Specimens during storage - Set A
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date; Stage
Av. M
C %
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(b) Av. MC of cores
Av. MC of cores taken during storage - Set A
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
DDate; Stage
Av.
MC
%
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(c) Av. Surface MC of cores
Surf. MC of cores taken during storage - Set A
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date; Stage
Av. M
C %
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
Figure 68 Average Moisture Content of a set of five paired specimens, estimated from specimen
mass and MC of cores taken from specimens.
Solid symbols/lines = Specimens from Run 1; Open symbols/dotted lines = Specimens from Run 2.
53
Distortion
Figure 69 shows the change in average twist, spring and bow for all specimens of each run
during the storage period. Run 2 had more twist than Run 1 but the behaviour during the storage
period is similar. Spring and Bow behaviour after the initial period is also similar.
(a) Twist over 2.4m length
Av. Twist of specimens during storage
-7.5
-7.0
-6.5
-6.0
-5.5
-5.0
-4.5
-4.0
-3.5
-3.0
A B C D
Stage, weekly intervals
Tw
ist
ov
er 2
.4m
, d
eg
re
es
Run 1
Run 2
#
$$
$
&
&,%
%%
(b) Spring at centre of 2.4m length
Av. Spring of specimens during storage
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A B C D
Stage, weekly intervals
Sp
rin
g a
t cen
tre o
f 2.4
m, m
m
Run 1
Run 2
#
##
#
#
##
#
(c) Bow at centre of 2.4m length
Av. Bow of specimens during storage
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
A B C D
Stage, weekly intervals
Bo
w a
t cen
tre o
f 2.4
m, m
m
Run 1
Run 2
#
%,@%,@
[]
#,@
$
$
#,%
Figure 69 Average distortion of all specimens in each run during storage.
Points which do not share the same symbol are significantly different (Scheffe Test).
54
The distortion of 10 sets of paired specimens is shown in Figure 70. After the initial period when
some specimens exchanged moisture at the surface there was generally little change in all forms
of distortion.
(a) Twist over 2.4m length
Twist during storage - Set A
-12.0
-11.0
-10.0
-9.0
-8.0
-7.0
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date
Tw
ist,
de
gre
es
ov
er 2
.4m
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(b) Spring at centre of 2.4m length
Spring during storage - Set A
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date
Sp
rin
g,
mm
, a
t c
en
tre
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(c) Bow at centre of 2.4m length
Bow during storage - Set A
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date
Bo
w,
mm
, a
t c
en
tre
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
Figure 70 Twist, Spring and Bow of a set of paired specimens measured at weekly interval
during storage.
Solid symbols/lines = Specimens from Run 1; Open symbols/dotted lines = Specimens from Run 2 Sign of spring has no intrinsic meaning; +ve bow is concave on the inner (near pith) face.
55
Stiffness
Figure 71 shows average stiffness for all specimens in each run calculated from two
measurement methods. For each run stiffness measured by bending on flat appears to increase
during storage; this is less evident for that from acoustic measurements. With the bending tests
which were done without rotating specimen supports, after the first set of measurements a wedge
was used to eliminate the point load from twisted specimens at the non-rotating loading points;
this may explain part of the increase from B to C. Figure 72 shows stiffness for selected pairs of
specimens.
Av. Stiffness from of specimens during storage - Acoustic Meast
8.0E+09
8.2E+09
8.4E+09
8.6E+09
8.8E+09
9.0E+09
9.2E+09
9.4E+09
A B C D
Stage, weekly intervals
E
Run 1
Run 2
#,$#
$
#,$
&#,&
#,$
#,$
Av. Stiffness of specimens during storage - 3-pt bending flat
6.0E+09
7.0E+09
8.0E+09
9.0E+09
1.0E+10
1.1E+10
1.2E+10
B C D
Stage, weekly intervals
E
Run 1
Run 2
#
$ $
#
$ $
(a) Stiffness (acoustic tests) (b) Stiffness (3-point bending on flat
Figure 71 Average Stiffness calculated from (a) acoustic tests and from (b) bending on flat for all
specimens of each run.
Points which do not share the same symbol are significantly different (Scheffe Test).
Stiffness during storage - acoustic - Set A
4.0E+09
6.0E+09
8.0E+09
1.0E+10
1.2E+10
1.4E+10
1.6E+10
1.8E+10
2/09/2005
A
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date
Sti
ffn
ess, E
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
Stiffness during storage - 3-point bending flat - Set A
4.0E+09
6.0E+09
8.0E+09
1.0E+10
1.2E+10
1.4E+10
1.6E+10
1.8E+10
9/09/2005
B
15/09/2005
C
23/09/2005
D
Date
Sti
ffn
ess, E
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(a) (b)
Figure 72 Stiffness calculated from (a) acoustic tests and from (b) bending on flat.
Set A has the five best matched pairs. Solid symbols/lines = Specimens from Run 1; Open symbols/dotted lines = Specimens from Run 2.
56
NMR measurements
Figure 73 Solid echo signals for six of the dried specimens of Run 1.
Plots of the solid echo signals for the initial measurements of cores from six specimens of Run 1
are shown in Figure 73 and for matched specimens of Run 2 in Figure 74.
The solid echo signal is the response from all protons in the sample at the frequency. All
different protons, moisture and organic mobile components, contribute to the signal resulting in
the clearly separate contributions to the free induction decay signal (FID): the first is the solid
part with decay generally described by the Gaussian decay function; the second is the semi-
mobile part (from moisture and small organic molecules – like resins) with longer relaxation
times and with decay described by a common Lorentzian function. Therefore the FID signal
from the solid-echo sequence analysed by the combination of Gaussian and Lorentzian functions
gives the characteristic parameters for solid matrix and hydrolysed semi-mobile phases. The
latter is a combination of “bound” water and softened organic matrix where the protons quickly
exchange the spin magnetization between different molecules.
57
Figure 74 Solid echo signals for six of the dried specimens of Run 2.
As a result of these interactions only one T2 value is detectable as an average of these different
sites. Increasing amounts of water soften more organic molecules and the overall dynamic of this
phase increases (longer T2) and as well as the overall signal intensity. On the other hand the
redistribution of moisture in the sample also enhances the magnetization exchange between the
molecules resulting in a sharper (longer T2) and taller (increased intensity) resonance band. In
the proper equilibrium one can expect to have a proportional increase in amplitude and T2 of
longer component (Lorentzian) at different, increasing moisture content. When moisture
redistribution is not in equilibrium the data can be expected to deviate from this proportionality.
The deviation or scattering of the data amplitude versus T2 therefore can reflect the stage of
moisture distribution as well as its changes with different scattering at different times after the
drying process.
Parameters of NMR measurements of cores taken from the specimens are presented in Table 7.
58
Table 7 NMR Solid Echo Results for Cores on Minispec 10 MHz
Gaussian parameters (Figure 75) relate to the solid wood matrix. The variation of amplitude and T2 seen in Figure 75 can not be taken to be meaningful as it is of the same order as the error in the signal.
T2 Gaussian during storage - Set A
0.0128
0.013
0.0132
0.0134
0.0136
0.0138
0.014
0.0142
0.0144
���� 2/09/2005
A
���� 9/09/2005
B
���� 16/09/2005
C
���� 23/09/2005
D
Date; Stage
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
Amplitude Gaussian during storage - Set A
0
100
200
300
400
500
600
700
800
900
1000
����� 2/09/2005
A
���� 9/09/2005
B
�� 16/09/2005
C
�� 23/09/2005
D
Date; Stage
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(a) Amplitude Gaussian (b) T2 Gaussian
Figure 75 Gaussian parameters of NMR measurements of cores taken during storage.
Set A has the five best matched pairs. Solid symbols/lines = Specimens from Run 1; Open symbols/dotted lines = Specimens from Run 2.
Lorentzian parameters
Amplitude Lorentzian during storage - Set A
0
100
200
300
400
500
600
700
���� 2/09/2005
A
���� 9/09/2005
B
���� 16/09/2005
C
���� 23/09/2005
D
Date
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
T2 Lorentzian during storage - Set A
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
����� 2/09/2005
A
����� 9/09/2005
B
���� 16/09/2005
C
���� 23/09/2005
DDate
31-2
36-1
37-2
39-2
42-1
31-1
36-2
37-1
39-1
42-2
(a) Amplitude Lorentzian (b) T2 Lorentzian
Figure 76 Lorentzian parameters of NMR measurements of cores taken during storage.
Set A has the five best matched pairs. Solid symbols/lines = Specimens from Run 1; Open symbols/dotted lines = Specimens from Run 2.
Figure 76 shows the Lorentzian parameters for the cores taken from specimens at weekly stages
during storage. Amplitude Lorentzian is most strongly influenced by moisture content; T2
Lorentzian is an indicator of moisture mobility.
The peaks of Amplitude generally seem to correspond with high core MC (Figure 68).
60
Figure 77 and Figure 78 show the Amplitude Lorentzian for cores taken from specimens at
stages during storage plotted against the average moisture content of matching smaller cores
taken at the same time, for Run 1 and Run 2 respectively. It can be seen in these figures that
Amplitude Lorentzian is generally linearly related to moisture content. Data from two poorly
matched pairs of specimens with higher moisture content have been omitted as the NMR
analysis was “tuned” to moisture content below about 15%.
Amplitude Lorentzian v MC - Run 1
100
200
300
400
500
600
700
8 9 10 11 12 13 14 15 16
Core Av. MC%
Am
plitu
de L
oren
tzia
n
Av. Core MC A
Av. Core MC B
Av. Core MC C
Av. Core MC D
Fitted to all
Figure 77 Amplitude Lorentzian plotted against Av. MC of matching cores – Run 1.
Amplitude Lorentzian v MC - Run 2
100
200
300
400
500
600
700
8 9 10 11 12 13 14 15 16
Core Av. MC%
Am
plitu
de L
oren
tzia
n
Av. Core MC A
Av. Core MC B
Av. Core MC C
Av. Core MC D
Fitted to all
Figure 78 Amplitude Lorentzian plotted against Av. MC of matching cores – Run 2.
Figure 79 and Figure 80 show T2 Lorentzian for cores taken from specimens (set A of previous
figures) at stages during storage plotted against the average moisture content of matching smaller
cores taken at the same time, for Run 1 and Run 2 respectively. T2 Lorentzian indicates the
mobility of water molecules, as well as other organic molecules associated with water. Water
therefore provides the major part of the signal. For each set of points the steepness of the fitted
line can be taken to indicate increasing water mobility. The fitted lines for specimens of Run 1
(Figure 79), which was steamed after drying are initially steep (Set A; dark blue line) then
61
recline to that at C (orange line) then become slightly steeper at D (plum line). This may indicate
that water in the wood, particularly that added to the surface during steaming, becomes more
strongly “bound” during storage, and during the period C to D the water added to the surface
from the atmosphere is likely to be initially more mobile.
T2 Lorentzian v MC - Run 1
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
8 9 10 11 12 13 14 15 16
Core Av. MC%
T2
Lo
re
ntz
ian
Av. Core MC A
Av. Core MC B
Av. Core MC C
Av. Core MC D
Fitted to all
Linear (Av. Core MC A)
Linear (Av. Core MC B)
Linear (Av. Core MC C)
Linear (Av. Core MC D)
Figure 79 T2 Lorentzian plotted against Av. MC of matching cores – Run 1.
The fitted lines for specimens of Run 2 (Figure 80) which was not steamed after drying follow
quite a different progression; the blue line for state A is shallow, indicating low water mobility
and the lines for states B and D are progressively steeper. State C seems out of step with this
progression. For these specimens the surface moisture content was low at A and the specimens
gained moisture during storage (Figure 68).
T2 Lorentzian v MC - Run 2
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
8 9 10 11 12 13 14 15 16
Core Av. MC%
T2
Lo
re
ntz
ian
Av. Core MC A
Av. Core MC B
Av. Core MC C
Av. Core MC D
Fitted to all
Linear (Av. Core MC A)
Linear (Av. Core MC B)
Linear (Av. Core MC C)
Linear (Av. Core MC D)
Figure 80 T2 Lorentzian plotted against Av. MC of matching cores – Run 2.
These explanations for the observed differences are only the best attempt to interpret these data;
they may not be completely consistent, because the moisture content of the cores seems to show
more variation, perhaps along the (about 300mm) length of the specimens from which successive
cores were taken at each weekly stage (Figure 68 (b)).
62
PROJECT RESULTS AND DISCUSSION
1. Bonding states of water during processing and stabilization
Non-freezing water, Wnf, (structural water or tightly bound to cell walls) and freezing bound
water, Wfb, (water associated with walls and smaller reservoirs) can not be differentiated at
room temperature through NMR spectroscopy detection due to the fast exchange between
them that gives rise to only one average parameter (like chemical shift and relaxation time) in
the NMR signal.
Additional moisture taken in or lost by the wood in conditioning rooms or during storage is a
relatively quick process, in days, and it is not free water but becomes a part of the total bound
water, Wnf + Wfb. This bound water has, as expected, higher NMR signal amplitude when
external humidity is higher but it also has a longer T2 component. Assuming that a fast
exchange mechanism is in place between Wnf and Wfb one can say that the majority of added
moisture becomes firstly Wfb type moisture. This is expected to have a longer T2 (more
mobile) and therefore moves the averaged value for all bound water to a longer T2 at higher
overall moisture content. This is true only when the total moisture content is below the FSP
(around 28%).
The further separation and evaluation of the major NMR parameters (relaxation mechanisms)
has produced an improved physical description of the moisture states in softwood and thus
contribute to the development of a comprehensive softwood drying model incorporating
moisture state as well as location.
The solid echo signal is the response from all protons in the sample at the frequency. All
different protons, moisture and organic mobile components, contribute to the signal resulting
in the clearly separate contributions to the free induction decay signal (FID): the first is the
solid part with decay generally described by the Gaussian decay function; the second is the
semi-mobile part (from moisture and small organic molecules – like resins) with longer
relaxation times and with decay described by a common Lorentzian function. Therefore the
FID signal from the solid-echo sequence analysed by the combination of Gaussian and
Lorentzian functions gives the characteristic parameters for solid matrix and hydrolysed
semi-mobile phases. The latter is a combination of “bound” water and softened organic
matrix where the protons quickly exchange the spin magnetization between different
molecules.
As a result of these interactions only one T2 value is detectable as an average of these
different sites. Increasing amounts of water soften more organic molecules and the overall
dynamic of this phase increases (longer T2) and as well as the overall signal intensity. On the
other hand the redistribution of moisture in the sample also enhances the magnetization
exchange between the molecules resulting in a sharper (longer T2) and taller (increased
intensity) resonance band. In the proper equilibrium one can expect to have a proportional
increase in amplitude and T2 of longer component (Lorentzian) at different, increasing
moisture content. When moisture redistribution is not in equilibrium the data can be expected
to deviate from this proportionality. The deviation or scattering of the data amplitude versus
T2 therefore can reflect the stage of moisture distribution as well as its changes with different
average mobility at different times after the drying process.
2. Bonding states of water, internal stresses and distortion
The variation in structural wood distortion and variations of moisture content seems to be
inconsistent with the expected correlation between moisture and shape change – indicating
that the whole measured moisture is determined by other factors as well. Only the cell wall
63
moisture content, which is only one part of the measured total moisture, is commonly
thought to relate to the structural changes.
The variation between the end of the drying cycle and the later conditioning is not all related
to the cell walls moisture (Wnf) and therefore the moisture content corresponding to the shape
change could not be clearly established from the current data. In order to achieve this
correlation further investigation is needed to find a novel NMR method to quantitatively
identify these two types of bound water. The low field NMR can be used finally after the
high-field solid-state methods prove a satisfactory differentiation and quantification of the
two types of bound water.
Other components of wood may be involved in wood shape changes independent of
moisture. This requires further technique development with solid-state high field NMR
measurements.
3. High temperature kiln drying schedule modifications and/or
treatments that improve stability
A HT kiln schedule modified with ramped reductions in temperature with increasing
humidity instead of steam reconditioning, produced similarly straight and stable timber to the
conventional HT schedule with stream reconditioning, over the three weeks period of
monitoring. The project did not incorporate a longer period.
4. Stabilization treatments to reduce the time to a stable state
No treatments for reducing the time to stable products have been identified.
Solid-state NMR investigation indicates that the process of steaming after drying results in
increased water mobility; the tests of specimen stability during humidity cycling only
showed indications of this. To the extent that moisture changes are related to wood shape
stability it may be best to minimise final steaming. This seems to be one possible direction
for further work and should be investigated in experimental trials.
5. Differences between Heart-in and Free of Heart material
A clear difference in the ‘FID’ signal of green wood was also detected for different wide
faces of the board, when one face of the board consists mainly of sapwood as opposed to
heartwood (Appendix 1).
Knots and bluestain are also distinguishable by analysis of the FID signal.
6. Modification to current commercial practices to minimize kiln
drying time, steaming time and storage periods to produce stable
dried softwood timber
A HT kiln schedule modified with later reductions in temperature and higher humidity and
no steam reconditioning, produced similarly stable timber to the conventional HT schedule
and stream reconditioning. Although the drying time was increased, processing time was the
same. The benefit will need to be clearly established for this to be adopted, as kiln
productivity would be reduced unless initial drying temperatures were increased.
This project has not identified a clear link between storage time and product stability.
64
ACKNOWLEDGEMENTS
This research was undertaken with assistance from the Forests and Wood Products Research and
Development Corporation (www.fwprdc.org.au) which is funded by industry and the Australian
Government.
Timber for experiments was kindly provided by Weyerhaeuser Australia through Green Triangle
Forest Products, Hyne Timber and Wespine Industries. The support and advice of Chris Lafferty
(FWPRDC), Tony Haslett (formerly with Weyerhaeuser Australia, now with Ensis), Stephen
Bolden (formerly with Hyne Timber) and Richard Schaffner (Wespine) is gratefully
acknowledged.
DISCLAIMER
The opinions provided in this Report have been provided in good faith and on the basis that
every endeavour has been made to be accurate and not misleading and to exercise reasonable
care, skill and judgment in providing such opinions. However, CSIRO as project manager, and
the parties to the joint venture known as ensis which carried out the research ('ensis') (CSIRO
and Forest Research NZ) do not guarantee or warrant the accuracy, reliability, completeness or
currency of the information in this report unless contrary to law. Neither ensis nor any of its
staff, contractors, agents or other persons acting on its behalf or under its control accept any
responsibility or liability in respect of any opinion provided in this Report by ensis or any person