Modelling air-drying of wooden poles Jarl-Gunnar Salin 1 Jarl Gunnar - PowerPoint PPT Presentation

Modelling air-drying of wooden poles Jarl-Gunnar Salin 1 Jarl Gunnar Salin Peder Gjerdrum 2 1 Romensvägen 12 A, Esbo, Finland jarlgunnar.salin@welho.com 2 The Norwegian Forest and Landscape Institute Aas Norway The Norwegian Forest and Landscape Institute, Aas, Norway peder.gjerdrum@skogoglandskap.no

Subject of investigation: Preserved wooden poles used for power and tele- used for power and tele communication lines, etc. Before the creosote preservation these poles preservation these poles have to be dried to a MC below the FSP every below the FSP every- where in the pole.

Air-drying Air drying In the present case air-drying outdoors is used. This method has both benefits and drawbacks. One drawback is One drawback is the long drying ti time and a good d d way to determine the drying end point is thus p important.

End point determination End point determination • Sampling possible only from the pole ends, but p g p y p , these are not reliable and a sample shortens the pole. pole • Resistance meters are difficult to use as the sapwood/heartwood borderline depth is not accurately known. y • Advanced methods like X-ray, CT-scanning etc. are too expensive in this small scale operation are too expensive in this small scale operation. • One possibility is to use simulation models and this alternative has now been investigated.

Simulation model Simulation model The model consists of two parts: 1 1. A model for moisture migration in a cylindrical A model for moisture migration in a cylindrical solid. Sapwood and heartwood have to be considered as two different materials considered as two different materials. 2. A model for the interaction with the surround- ing climate. The climate is defined by data from a nearby weather station (temperature, from a nearby weather station (temperature, RH, wind speed, rain).

Internal moisture migration Internal moisture migration 35% MC 130% MC

Internal moisture migration Internal moisture migration ∂ ∂ ∂ ⎛ ⎛ ⎞ ⎞ u 1 u = ⎜ ⎟ Dr ∂ ∂ ∂ ∂ ∂ ∂ ⎝ ⎝ ⎠ ⎠ t t r r r r r r Fick’s equation in cylindrical coordinates MC is not an adequate potential for describing MC is not an adequate potential for describing flow across the heartwood/sapwood border. The equilibrium MC pairs in these two materials have equilibrium MC-pairs in these two materials have to be determined in order to define a replacing potential potential.

Heartwood and sapwood mutual equilibrium MC values . 40 38 38 % MC, % 36 34 34 artwood 32 30 30 Hea 28 26 26 25 50 75 100 125 150 Sapwood MC % Sapwood MC, %

Internal model part Internal model part Moisture migration potential used in the model: 1 1. In sapwood: MC (as normally) In sapwood: MC (as normally) 2. In heartwood: The MC in sapwood that is in equilibrium with the actual MC in heartwood ilib i i h h l MC i h d (linear relationship assumed). Th fi The final internal model is an updated version l i t l d l i d t d i of an old model for kiln drying of logs and contains thus no further adjustable parameters.

External model part External model part • Daily average climate data are obtained from a y g nearby weather station (temperature, RH, wind speed and rain) speed and rain). • The main problem is to determine the relation between the meteorological wind speed and the b h l i l i d d d h external heat (and mass) transfer coefficient, h, in the stack of poles . = ⋅ Attempt: Attempt: 0 0 , 67 67 h h a w a = adjustable parameter, w = wind speed dj t bl t i d d

Tuning of the model Tuning of the model • The model contains only one adjustable para- meter, i.e. the factor a in the wind speed equation • The factor was determined by weighing 31 poles The factor was determined by weighing 31 poles in a test stack during one summer period. In this way the MC development could be followed and a the MC de elopment co ld be follo ed and compared to model predictions for different a - values Test stack

Test poles Test poles Pole Diameter Butt Colour Stack length, m class diameter, code layer mm 12 Medium 261 Yellow Top 12 12 S Stout 366 366 G Green U d Under 11 Medium 305 Red Under 9 Light 239 Blue Under

Results Results • In the first analysis the influence of rain was neglected. The rain is assumed to flow off the poles without absorption. p p • It turned out that the uppermost pole layer has a ~27% higher a -value than lower layers. This is 27% higher al e than lo er la ers This is probably due to the influence of sunshine and a higher air velocity above the top layer. Thus the top layer and the lower layers are simulated p y y separately. • The results are presented for the different classes: • The results are presented for the different classes:

Simulation results Simulation results – best fit best fit Blue class Blue class 120 % ntent, % Model MC M d l MC 100 100 Obs. MC 80 ure con 60 40 40 Moistu 20 0 2008- 2008- 2008- 2008- 2008- 2008- 2008- 02 22 02-22 04 12 04-12 06 01 06-01 07-21 07 21 09-09 09 09 10-29 10 29 12 18 12-18

Simulation results Simulation results – worst fit worst fit Yello Yellow class class 120 ntent, % % 100 100 Model MC Obs. MC 80 ure con 60 40 40 Moistu 20 0 2008- 2008- 2008- 2008- 2008- 2008- 2008- 02 22 02-22 04 12 04-12 06 01 06-01 07-21 07 21 09-09 09 09 10-29 10 29 12 18 12-18

Further analysis Further analysis • In the next step it was assumed that a certain p amount of the rain hitting a pole is absorbed and has to be evaporated in the drying process. This p y g p model has thus two adjustable parameters. Further a 24 hour sine-variation was super- p imposed on the daily average temperature. • Again it turned out that the top layer differs from • Again it turned out that the top layer differs from the rest as a higher amount of rain is absorbed. • In the model tuning process the last weighing in I th d l t i th l t i hi i December with already frozen poles was now i included. l d d

New simulation New simulation – best fit best fit Green class 120 % ntent, 100 80 ure con 60 40 40 Moistu 20 M 0 0 2008- 2008- 2008- 2008- 2008- 2008- 2008- 02 22 02-22 04 12 04-12 06-01 06 01 07-21 07 21 09-09 09 09 10-29 10 29 12-18 12 18

New simulation New simulation – worst fit worst fit Yellow class 120 % ontent, 100 100 80 ure co 60 40 Moist 20 0 0 2008- 2008- 2008- 2008- 2008- 2008- 2008- 02-22 04-12 06-01 07-21 09-09 10-29 12-18

Conclusions Conclusions • The model seems to capture the main features of the pole drying process, despite only one (or two) p y g p , p y ( ) adjustable parameters. It is thus believed that the model can be used as an additional valuable tool model can be used as an additional valuable tool in the determination of the drying end point. • The factory is interested in the MC development in the innermost part of the sapwood, i.e. the p p point when all free water has been removed.

Example of model use Example of model use 140 Sapwood max MC Butt 120 Sapwood max MC 1/4 Sapwood max MC 1/4 ent, % Sapwood max MC Half 100 Sapwood max MC 3/4 re conte Sapwood max MC Top 80 Pole average MC 60 60 Moistur 40 M 20 0 0 2008-03-25 2008-05-25 2008-07-25 2008-09-24 2008-11-24