Estimating dimensional change

There is a difference between shrinkage, which occurs as the change in dimensions of wood on initial drying from the green condition, and movement, the dimensional change which takes place when wood that has been dried is subjected to changes in atmospheric conditions below the fibre saturation point (Farmer, 1972). It is possible that a wood may shrink quite appreciably in drying from the green condition yet it may undergo comparatively small dimensional changes when subjected to a given range of atmospheric conditions in service. The reason is that the fibre saturation point, the moisture content value at which appreciable shrinkage begins to take place, varies between different species. In addition, the moisture con-

Figure 2.16 Roughly speaking, tangential shrinkage is double radial shrinkage. This difference in dimensional change when wood is seasoned causes the distortion seen in representative shapes cut from logs (a) and commonly results in radial cracks in logs (b), or squared timbers (c) containing the pith of the tree are not commonly available, as are data for perpendicular shrinkage (i.e. shrinkage across the grain). It should be cautioned, however, that abnormal wood tissue, such as juvenile wood, reaction wood, or pieces with cross-grain, may exhibit longitudinal shrinkage of up to 10—20 times normal. In addition, it should be expected that abnormal wood will occur unevenly in severity and in distribution, and that the resulting uneven longitudinal shrinkage may produce warping.

Figure 2.17 Typical relationships of tangential, radial and longitudinal shrinkage to equilibrium moisture content, shown from oven dry (OD) to fibre saturation point (FSP). Note that shrinkage along the grain (longitudinal) is negligible. Tangential shrinkage is roughly double radial shrinkage, thus the historical preference for quartersawn boards when maximum dimensional stability was required

Figure 2.17 Typical relationships of tangential, radial and longitudinal shrinkage to equilibrium moisture content, shown from oven dry (OD) to fibre saturation point (FSP). Note that shrinkage along the grain (longitudinal) is negligible. Tangential shrinkage is roughly double radial shrinkage, thus the historical preference for quartersawn boards when maximum dimensional stability was required

10 20

/

/

%

/

C

/

2 I

lUVUVv

/

LU

1 VVw\

ftm

imvv\

^¿r -----

JJLI\\V\\

l' \ fe \ \ \ \ \

/ i !

11 \ l\\ \ \ \ \ \ t\'w\\\\\

X ! !

/ > 1 / . . i . , . i

40 50 60 Relative humidity (%)

90 100

40 50 60 Relative humidity (%)

90 100

Change from 70% RH to 30% RH

1.2% shrinkage

Figure 2.18 The relationship between relative humidity, equilibrium moisture content and shrinkage. Choose shrinkage according to species and growth ring orientation, for example, 5.1% for a tangentially sawn mahogany board (shown as a dotted line on the graph on the right). In this case a change from 70% RH to 30% RH will cause 1.2% shrinkage. In a board one metre wide, this produces 12 mm of shrinkage tent change of one timber corresponding to any given range of atmospheric conditions often varies considerably from that of another. However, in a given direction, dimensional change is roughly proportional to moisture content over the range of bound water loss, as shown in Figure 2.17.

Because the shrinkage percentages are based upon shrinkage from the green condition, the formula given in section 2.4.3 is accurate only for shrinkage starting from the green condition. For dimensional change of wood starting at a partially dry condition this formula will introduce an average error of about 5% of the calculated change in dimension. For most purposes, such error is insignificant in view of other inherent sources of error. However, where a more refined estimate is desirable the following formula should be used: D\ (MC - MCf)

FSP S

where

D = dimensional change, in linear units Di = initial dimension MCi = initial moisture content, per cent MCf = final moisture content, per cent FSP = fibre saturation point, per cent S = published value for shrinkage, as percentage (St for tangential, Sr for radial etc.) of green dimension from green to oven-dry moisture content.

In using this formula, the shrinkage percentages are taken as average published values for the species, which are listed as either radial or tangential. In specific cases involving a piece of wood with intermediate growth ring placement, an approximate value would have to be deduced by rough interpolation between the radial and tangential values. It should be noted that since shrinkage takes place only below fibre saturation, neither MCi nor MCf can be greater than the FSP. Positive values of dimensional change indicate shrinkage; negative values indicate swelling.

To illustrate, suppose a flat-sawn mahogany board had been conditioned to an outdoor

environment of 70% RH (from Figure 2.14 assume an EMC = 13%). The board is then moved to an indoor environment of 30% RH (EMC = 6%). From Figure 2.14 assume an FSP of 30%. Published data for mahogany indicate a 5.1% tangential shrinkage (Hoadley, 1980). Therefore, the estimated change in the width of the board is calculated as:

0.051

Thus the board would shrink by 0.178 inches.

Where direct measurement of moisture content is impossible and where precise numerical values are unrealistic, the above formula may be of more academic than practical use. An equally useful and reasonable approach is a graphic method of estimating dimensional change. Combining the oscillating curve of Figure 2.14 with the idea of Figure 2.17, a composite working graph can be devised, as shown in Figure 2.18.

Based on the species and growth-ring orientation of a piece of wood in question, the appropriate shrinkage percentage (St, Sr, or interpolated estimate) is taken from published data. On the right-hand side of Figure 2.18, choose the EMC/S line that most closely matches the shrinkage percentage of the subject. Estimates of changes in RH can now be translated into percentage dimensional change by following RH values up and over to corresponding EMC values, then over and down to corresponding S values.

The graphic relationship between relative humidity, moisture content and shrinkage draws attention to the important point that relative humidity is the important controlling parameter and dimensional change is the eventual consequence. Too often, relative humidity is not given the serious attention it deserves. Although moisture content is usually not of direct concern, it can be important indirectly, if we remember that the weight of wood reflects the moisture content. A furniture object probably loses or gains weight primarily as a response to changes of moisture content of its wooden components. This change takes place before any significant dimensional change

(where wood is restrained, it may not actually be possible to observe dimensional change). Placing of furniture on a weighing device such as a load cell could be used as a quite accurate and relatively inexpensive means of monitoring the weight of an object. It could therefore be an excellent way to detect and prevent changing conditions which might eventually result in dimensional change problems, especially when wooden objects are being transported or relocated in a new environment. Such a device could actually be connected to humidification equipment thereby allowing the object to control its own environment.

Although dimensional change alone may be a serious consequence of moisture variation, shrinkage and swelling that is uneven, even though in minor amounts, can cause distortion of a piece from its desired or intended shape. Various forms of distortion include cup (deviation from flatness across the width of a board), bow (deviation from lengthwise flatness of a board), crook or spring (departure in end-to-end straightness along the edge of a board), and twist (where four corners of a flat face do not lie in the same plane).

A common source of uneven dimensional change is simply the greater tangential than radial shrinkage percentage, a routine cause of cup in boards, as shown in Figure 2.16. Note that flat sawn boards located closest to the pith have most severe cup, concentrated near the centre. Unequal radial and tangential shrinkage causes round turnings to dry to ovals, squares with diagonal ring placement to become diamond-shaped (as a rule of thumb, growth ring lines tend to straighten). In log sections, or timbers containing the pith of the tree, the greater tangential shrinkage develops circumferential stresses which may exceed the strength of the wood, resulting in radial cracks.

The fact that dimensional change is greater at right-angles to the grain than parallel to it also produces problems. The opening of mitre joints and the loosening of mortise and tenon joints are classic examples. Other causes of uneven shrinkage are uneven drying, or material with abnormal wood (such as reaction wood, juvenile wood etc.) This topic is discussed in more detail in Chapter 7.

With time, the dimensional response of wood may lessen, in part because hygroscop-icity of the wood may decrease, or because of mechanical effects of repeated shrinkage/ swelling cycles, or stress-setting of the wood. However, experiments with wood taken from artefacts thousands of years old have shown the wood to have retained its hygroscopicity and its capacity to dimensionally respond to changes in moisture content. The assumption should therefore prevail that wooden objects, regardless of age, can demonstrate dimensional movement when subjected to variable relative humidity conditions.

Wood Working for Amateur Craftsman

Wood Working for Amateur Craftsman

THIS book is one of the series of Handbooks on industrial subjects being published by the Popular Mechanics Company. Like Popular Mechanics Magazine, and like the other books in this series, it is written so you can understand it. The purpose of Popular Mechanics Handbooks is to supply a growing demand for high-class, up-to-date and accurate text-books, suitable for home study as well as for class use, on all mechanical subjects. The textand illustrations, in each instance, have been prepared expressly for this series by well known experts, and revised by the editor of Popular Mechanics.

Get My Free Ebook


Post a comment