29.08.2013 Views

Proceedings e report - Firenze University Press

Proceedings e report - Firenze University Press

Proceedings e report - Firenze University Press

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

DIMENSIONAL CHANGE OF UNRESTRICTED WOOD AT DIFFERENT LEVELS OF EQUILIBRIUM MOISTURE CONTENT<br />

S = published value for shrinkage (%)<br />

Linear Dimensional Change<br />

Considering that all dimensions in every orientation would undergo radial contraction/expansion, the<br />

extent of tangential contraction/expansion needs to be found. Formula (3) giving the linear<br />

dimensional change [4]was used to arrive at the final increase/decrease for every dimension recorded.<br />

D<br />

f<br />

D ⎡<br />

⎤<br />

i ⎛⎛<br />

Vt<br />

−Vr<br />

0 ⎞ ⎞<br />

× ⎢100<br />

− ⎜⎜<br />

× O ⎟ + V ⎟⎥<br />

100 ⎣ ⎝⎝<br />

90 ⎠ ⎠⎦<br />

= r<br />

where: Df = final dimension<br />

Di = initial dimension<br />

Vt = calculated tangential shrinkage (%)<br />

Vr = calculated radial shrinkage (%)<br />

O O = orientation (in degrees)<br />

2.3. Reconstruction of the Panel End Grain presenting Contraction/Expansion and<br />

Warping on AutoCAD<br />

The results obtained as outlined above, by using a Microsoft Excel spreadsheet, were then transferred<br />

back to AutoCAD. The resulting outline (based on the new set of data) led to the indications of<br />

dimensional change as well as any cupping the panel would undergo as a result of changes in the T<br />

and the RH. Longitudinal contraction/shrinkage is deemed as insignificant, especially over short<br />

lengths, but it can still be predicted by using the same computer routine.<br />

3. Limitations of the computer analysis<br />

The computer analysis only leads to approximate results. These results are only reliable under given<br />

conditions while certain other conditions affect the behaviour of wood. Such instances outlined<br />

hereafter.<br />

The wood panels should be free from natural defects such as: cross grains, knots, compression/tension<br />

wood, juvenile wood, etc.<br />

The wood should not be restricted by crossbars, frames and/or mounting hardware.<br />

Preparation and painting layers (as in panel paintings) lower the accuracy of the results obtained from<br />

this computer analysis, especially when cupping is evaluated.<br />

Published literature data for the coefficient of shrinkage (tangential, radial and longitudinal) as well as<br />

the FSP are all average values, which might result to be quite different in reality. Dimensional<br />

response of aged wood may result to be slightly lower due to decrease in hygroscopicity and/or the<br />

mechanical effects of repeated shrinkage/swelling cycles. [5] In the case of panel paintings,<br />

unpredictable cupping may take place even on radially cut boards due to rapid moisture adsorption by<br />

the untreated back. [6]<br />

4 . Possibility of errors<br />

As the grain orientation may change, adjacent sides of sections may have slightly different<br />

dimensions. Such discrepancies are negligible and are not visible to the naked eye. They may be<br />

eliminated by narrowing the sections and, consequently, any changes in grain orientation result to be<br />

more gradual.<br />

5. Testing the computer procedure<br />

5.1. Aim<br />

This test was carried out in order to simulate situations of contraction/expansion as well as warping.<br />

Literature data for Shellbark Hickory (Carya laciniosa) were chosen for this test, due to the high<br />

tangential and radial shrinkage values, i.e. 12.6% and 7.6% respectively. [7] The test was also carried<br />

out on actual pine samples.<br />

38<br />

(3)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!