Docstoc

Using poplar timber in light constructions The problem of non

Document Sample
Using poplar timber in light constructions The problem of non Powered By Docstoc
					Laboratory of Wood Technology - Faculty of Bioscience Engineering - Ghent University
                        Department of Forest and Water Management
                         Coupure Links 653, 9000 GENT (Belgium)




    Using poplar timber in light constructions:
     The problem of non-uniform moisture
            distributions after drying

      Lieven De Boever, Dries Vansteenkiste and Joris Van Acker



             “Broad Spectrum Utilisation of Wood” , Vienna, June 14-16, 2005
                  Outline

•   Introduction
•   Objectives
•   Experimental methods
•   Results and discussion
•   Conclusions and Prospectives
                       Introduction

• 380 000 m³ round wood of poplar yearly harvested in
  Belgium
• Actually covering 35 000 ha
• short rotation --> rapid available raw material
• Monoclonal plantations are under discussion
• Solutions in multi to polyclonal stands with a broader
  spectrum of clones.
• Introduction of higher variability in the quality of the raw
  material
• On the other hand an evolution towards more value
  added products
                         Introduction

•   Poplar is light and has a considerable specific strength
•   Use in light construction ?
•   When using sawn timber --> step of drying !
•   Starting with:
    – Very wide range of mean starting moisture content
    – Presence of wetwood
    – Also large variations whithin stems
• Going to:
    – Uniform dried material at 12%
    – Low amount of quality losses
                         Introduction

• Drying is needed:
   – improving resistance for biological attack
   – improving strength and stiffness
   – improving gluing and finishing properties
• Specific attention for poplar lumber:
   – tension wood
   – juvenile wood
   – occurrence of wet pockets after drying
                Objectives

• Evaluate the variability in end moisture
  content for several poplar clones from a
  multi-clonal stand
• Consequences for end-uses ?
• How quantify variability ?
• Objective definition of a wet pocket ?
                Experimental methods
• Selection of 8 clones was made (4 presented here)
  (Ghoy - Gaver - Primo - Robusta )
• Of all clones 3 trees were sampled.
• Per clone a minimum of 4 m³ round wood was used
• Optimal sawn pattern --> beams of 2000x50x95 mm
• Conventional kiln drying was applied (40 °C)
• Moisture content was measured every 25 cm
   – directly after drying
   – after a two week homogenizing period
• Evaluation of
   – end moisture content and variation
   – presence of wet pockets
               Experimental methods

• Evaluation of 4 different definitions for a “wet pocket”
• All the resulting parameters of wet pockets incidence
  were compared and cross-correlated.
                         Results and discussion

                                  End moisture      End MC per longitudinal distance to beam end [%]
   clone   Number of     moment    content [%]     12.5 cm      37.5 cm         62.5 cm       87.5 cm
            batons
Ghoy         109       0 weeks     17.4   (4.3)   15.8   (4.4)   17.8   (5.2)   17.9   (5.5)   18.0   (5.5)
                       2 weeks     14.3   (3.8)   12.7   (4.1)   14.5   (5.0)   15.1   (5.2)   14.9   (5.1)
Primo        151       0 weeks     16.6   (2.7)   14.8   (3.1)   16.8   (3.7)   17.1   (3.9)   17.7   (4.3)
                       2 weeks     14.5   (3.7)   12.7   (4.1)   14.6   (4.6)   15.2   (5.2)   15.5   (5.5)
Gaver        134       0 weeks     18.5   (3.9)   17.1   (4.0)   19.0   (5.1)   18.9   (4.7)   19.0   (5.2)
                       2 weeks     17.5   (4.5)   15.8   (4.4)   18.0   (5.2)   18.0   (5.6)   18.4   (6.0)
Robusta      108       0 weeks     23.3   (5.1)   19.2   (5.3)   23.3   (6.7)   24.2   (7.1)   26.4   (7.6)
                       2 weeks     19.5   (3.3)   17.7   (4.1)   19.5   (4.1)   19.8   (4.6)   20.9   (4.5)
0.16                                                                         0.12

0.14
                                                                              0.1
                                                                                            Gaver
0.12


 0.1
           Primo                                                             0.08


0.08                                                                         0.06

0.06
                                                                             0.04
0.04

                                                                             0.02
0.02


  0                                                                            0
       0        5        10        15        20        25        30     35          0        5       10   15   20   25   30    35




0.12                                                                          0.12


 0.1                                                                            0.1
                                                                                            Ghoy
0.08
           Robusta                                                            0.08


0.06                                                                          0.06



0.04
                                                                              0.04


                                                                              0.02
0.02


                                                                                    0
   0
                                                                                        0        5   10   15   20   25   30   35
       0    5       10        15   20   25        30        35    40   45
               clone   Moment        Wet          Wet          Wet
                                 pockets by pockets by pockets by
                                 definition 1 definition 2 definition 3
                                     [%]          [%]          [%]
             Ghoy      0 weeks        9.0         52.0         2.9
                       2 weeks        5.4         47.0         3.9
             Primo     0 weeks        2.9         47.1         7.8
                       2 weeks        4.7         39.0         3.9
             Gaver     0 weeks        8.0         40.0         4.1
                       2 weeks        8.8         44.0         2.8
             Robusta   0 weeks       49.1         34.0         6.6
                       2 weeks       10.9         49.1         6.6



1. Definition based on absolute value of 30%
2. Definition based on underlying normal frequency
   distribution. Average increased with 1.5 times the standard
   deviation
3. Theoretical surface between normal and two-parameter
   Weibull distributions
                clone   Moment        Wet          Wet          Wet
                                  pockets by pockets by pockets by
                                  definition 1 definition 2 definition 3
                                      [%]          [%]          [%]
              Ghoy      0 weeks        9.0         52.0         2.9
                        2 weeks        5.4         47.0         3.9
              Primo     0 weeks        2.9         47.1         7.8
                        2 weeks        4.7         39.0         3.9
              Gaver     0 weeks        8.0         40.0         4.1
                        2 weeks        8.8         44.0         2.8
              Robusta   0 weeks       49.1         34.0         6.6
                        2 weeks       10.9         49.1         6.6



• First definition:
   – Primo < Ghoy and Gaver < Robusta
   – Robusta --> high end moisture content
   – not suitable for comparing the clones considered nor for comparing
     drying schedules
                 clone   Moment        Wet          Wet          Wet
                                   pockets by pockets by pockets by
                                   definition 1 definition 2 definition 3
                                       [%]          [%]          [%]
               Ghoy      0 weeks        9.0         52.0         2.9
                         2 weeks        5.4         47.0         3.9
               Primo     0 weeks        2.9         47.1         7.8
                         2 weeks        4.7         39.0         3.9
               Gaver     0 weeks        8.0         40.0         4.1
                         2 weeks        8.8         44.0         2.8
               Robusta   0 weeks       49.1         34.0         6.6
                         2 weeks       10.9         49.1         6.6



• Second definition
   – “Assumed” normal distribution (including the points you want to
     detect)
   – Robusta < Primo and Gaver < Ghoy !
   – High standard deviation --> low detection
   – low standard deviation --> high detection
0,16



0,14



0,12



 0,1



0,08



0,06
                                                                                             Surface of 7,78 %



0,04

                                                    Surface of 3,91 %

0,02



  0
       0   5                        10                  15                20                        25           30
               weibull distribution measurement 1                       weibull distribution measurement 2
               normal distribution measurement 1                        normal distribution measurement 2
                  clone   Moment        Wet          Wet          Wet
                                    pockets by pockets by pockets by
                                    definition 1 definition 2 definition 3
                                        [%]          [%]          [%]
                Ghoy      0 weeks        9.0         52.0         2.9
                          2 weeks        5.4         47.0         3.9
                Primo     0 weeks        2.9         47.1         7.8
                          2 weeks        4.7         39.0         3.9
                Gaver     0 weeks        8.0         40.0         4.1
                          2 weeks        8.8         44.0         2.8
                Robusta   0 weeks       49.1         34.0         6.6
                          2 weeks       10.9         49.1         6.6



• Third definition
   –   theoretical approach (need of correlation)
   –   no direct practical significance
   –   Ghoy and Gaver < Primo < Robusta
   –   decrease in surface --> easy diffusion
   –   Robusta surface remains the same --> not reached end moisture
       content (evaporation)
                   5% above percentile   10 % above percentile value   15% above percentile value
                        value [%]                   [%]                          [%]
clone   timing     Without      With   Without     With         With  Without   With         With
                  neighbour neighbour neighbour neighbour neighbour neighbour neighbour neighbour
                                 5%                  5%         10%              5%          10%
Ghoy    0 weeks      57.7       27.9     16.2       14.4          6.3   2.7      2.7          1.8
        2 weeks      55.0       28.8     20.7       16.2          4.5   5.4      5.4          3.6
Primo   0 weeks      52.8       29.6      9.2         8.5         3.5   2.1      2.1          0.7
        2 weeks      53.0       31.5     17.4       13.4          6.7   4.0      3.4          2.7
Gaver   0 weeks      42.6       25.0      9.6         9.6         4.4   3.7      2.9          1.5
        2 weeks      48.5       25.0     12.5       12.5          5.1   2.9      2.9          2.2
Robusta 0 weeks      83.6       52.7     42.7       38.2        20.9   14.5      1.6          6.4
        2 weeks      62.7       38.2     11.8       11.8          6.4   3.6      3.6          2.7



   • Fourth definition
        – based on 25% percentile + absolute moisture %
        – with or without neighbour restrictions
        – with neighbours turned out to be more reliable --> “local”
                Parameter                 Percentile+5%        Percentile+10%      Percentile+15%
                                         (neighbour >5% )      (neighbour >5%)    (neighbour >10% )
Surface                Pearson’s         0.575(0.136)*       0.207(0.623)         0.301(0.469)
(definition3)          Kendall’sT au     0.519(0.079)**     -0.255(0.283)        -0.360(0.901)
                              an’s ho
                       Spearm R          0.639(0.088)**     -0.371(0.365)        -0.480(0.910)
M ean                  Pearson’s         0.745(0.034)***     0.672(0.068)**       0.776(0.023)***
     al
(norm distribution)    Kendall’sT au     0.109(0.708)       -0.671(0.805)         0.286(0.322)
                              an’s ho
                       Spearm R          0.252(0.548)       -0.240(0.955)         0.781(0.352)
Standarddeviation      Pearson’s         0.383(0.349)        0.729(0.040)***      0.687(0.060)**
     al
(norm distribution)    Kendall’sT au    -0.255(0.383)        0.500(0.083)**       0.286(0.322)
                              an’s ho
                       Spearm R         -0.216(0.608)        0.619(0.102)*        0.781(0.352)
M ean                  Pearson’s         0.781(0.022)***     0.776(0.024)***      0.791(0.019)***
  eibull
(w distribution)       Kendall’sT au     0.182(0.533)        0.000(1.000)         0.214(0.458)
                              an’s ho
                       Spearm R          0.323(0.435)        0.048(0.911)         0.238(0.570)
Standarddeviation      Pearson’s         0.771(0.025)***     0.741(0.035)***      0.762(0.028)***
   eibull
(W distribution)       Kendall’sT au     0.182(0.533)        0.000(1.000)         0.214(0.458)
                              an’s ho
                       Spearm R          0.323(0.435)        0.048(0.911)         0.238(0.570)
• Two important groups of correlation
   – surface parameter out of definition 3 With definition 4 based on a
     reference of 5% with neighbouring restriction.
   – Standard deviation of the “assumed” underlying normal
     distribution with definition 4 based on a nreference of 10% with
     neighbouring restrictions
• INDEPENDENT OF TIME AND CLONE !!
• 6.3 times the surface (def 3) gives the number of wet
  pockets
• 4.2 times the standard deviation gives the number of wet
  pockets
                  Conclusions
•   For use in light construction drying is needed
•   Clonal character --> more research
•   Not all clones reached set end moisture content
•   Adaptation of drying time and/or temperature
•   Impact on company logistics
•   Objective definition for detection of wet pockets
    – based on surface parameter
    – based on 5% above percentile value with
      neighbouring restrictions
                Prospectives
• Further research towards drying of poplar clones
  or groups of clones.
• Testing the usefulness of the new definition in
  repetitive experiments and for other drying
  schedules and methods
• Possibility of pre-sorting ?
• Linkage to drying deformations
• Some future chalenges:
  – multiclonal populiculture
  – poplar forestry
Thank you !


Questions ?

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:2
posted:9/16/2011
language:English
pages:22