Dried Timber how to specify correctly

Transcription

1 Dried Timber how to specify correctly 2010 Version 1.0 1

2 Contents Preface... 3 Wood quality and drying quality... 4 Selection of a suitable target moisture content... 9 Moisture content measurements EN Shrinkage and swelling of wood Measuring drying stresses, ENV Drying quality according to EN and ENV Take care of the timber European Drying Group (EDG) and COST E based on previous rights provided by SP Trätek Editor: Johannes Welling Illustrator: Claes Folkeson and Lea Knaack Download sites: 2

3 Preface COST E53 focuses on Quality of Wood and Wood Products. Besides techniques for scanning round and sawn timber for wood properties and grading of sawn timber into different use classes methods and strategies for assessment of drying quality are important topics. In order to enable the European wood industries and the users of sawn timber to specify and control drying quality of sawn timber products a number of CEN standards have been edited and published recently. Up to now these standards have not yet been fully implemented, understood and accepted by the various players. Under the leadership of J. Welling, Johann Heinrich von Thuenen-Insitute (vti), WG 2 in COST E53 has taken over to task to inform the sector by reviewing and modifying a brochure on drying quality, which had been published in 1996 by SP Trätek in Swedish language. For achieving the goal of producing a document which can be used all over Europe the Swedish text was updated by B. Esping, translated by J.-G. Salin, Y. Steiner, H. Forsen, V. Tarvainen and V. Möttönen. The English mother version of the document was reviewed by a small task group consisting of J. Welling, J.-G. Salin, J. de Corte, K.-M. Sandland, R. Nemeth, G. Milic, G. Knaggs. Pictures were redrawn by L. Knaack, graphs were up-dated by A. Behrens. Sincere thank is given to European Science Foundation for providing the framework for COST Action E53 and bringing together all persons and institutions that have supported this work without having received any financial compensation. SP Trätek has granted the copyright to COST E53 and to EDG to enable translation into various European languages and to secure a fast spreading of this brochure at lowest possible cost. The pdf-version of this document in the different European languages is provided free of cost. It can be downloaded from the COST E53 web-page ( or from the EDG web-page ( Copying and circulation of this information is welcome as long as origin and source is correctly cited and the content is not changed. The editors of this leaflet have done their best in referring to and in interpreting of recent literature and standards. Nevertheless the document may contain errors. The editors refrain from taking any kind of responsibility or grant warranty for losses arising from using this document. Hamburg, March 2010 Johannes Welling (for the editing team) 3

4 Wood quality and drying quality The drying quality concept actually covers many factors. Some of these depend on the quality of the wood itself, others on how the drying process has been performed. Some factors are easy to describe and measure, some are more difficult but often more important for the producer and for the user of the wooden product. Wrong mean moisture content, moisture content variation or drying induced stresses belong to this group they are not visible but represent a clear disadvantage. Examples of factors related to: Wood quality Resin pockets Reaction wood / juvenile wood Knots Mechanical properties Grain direction Density Shrinkage coefficients Frost shake False heart Drying Deviation of mean MC from MC targ MC variation within a timber batch MC variation in the timber cross section Drying induced stresses Drying checks (surface, internal, and end-checks) Others Wood mould Discolouration Resin flow Knot with cracks Twist, bow, spring Size deviation Difference between quality and quality Wood is a biological material and each piece of wood is unique, in the same way as each person is a unique individual. Timber properties such as shrinkage, straightness, tendency to check, moisture content, visual appearance etc. reflect the character of the timber. If one characteristic feature is too pronounced, a piece of wood may be unsuitable for a certain purpose. One of the most important properties, as far as timber is concerned, is the moisture content. Wrong moisture content is the most severe drying defect. Wrong moisture content in combination with certain properties or characteristic features, may lead to drying defects. As an example, the timber deforms more or less if it continues to dry in finished beams or in a piece of joinery. The greater the slope of grain or reaction wood, the greater the distortion (twist, spring/crook, bow) that will occur as the moisture content changes. If the timber moisture content ordered deviates too much from the moisture content which the timber attains in the final product, a quality defect will be created. The deviation is due to two drying errors: (1) The mean moisture content of the timber batch deviates too much from the target (required) moisture content and (2) the variation in moisture content in the batch is too high. Wrong moisture content does not need to be due to the fact that the delivered batch differs from the specified requirements, but the requirements may be incorrect or even missing. It is too common that the buyer orders 50 x 100 mm studs without specifying the moisture content according to the standard EN Timber deformations may be reduced, but not eliminated by actions in the drying process. An increased drying temperature, top loading and pre-twisting of the boards during drying can be mentioned as examples. 4

5 We are concentrating in this booklet on moisture content and drying stresses as it is more complicated to assess the quality regarding these factors than the other quality factors. Other drying defects as for instance resin flow, changes in knot appearance, discolouration and mould are thus not discussed in this booklet. Target moisture content The target moisture content, i.e. the required moisture content, indicates the suitable moisture content in the wooden material when the product is used or produced. When ordering timber it is thus advisable to give a target moisture content for the batch. It is, however, in practice impossible to exactly hit the target moisture content, due to the fact that the present moisture measurement techniques lack precision, and the wood itself exhibits a certain variability. According to a rule of thumb the mean moisture content of a batch will approximately deviate from the target up to 1/10 of the target value, i.e. with target values 12% and 18%, the batch mean value will be within about 12±1,2% and 18±1,8% respectively, see Figure % 14 Fig.2. All measured moisture content values for a timber batch are put in the columns above. For instance the column for 7% contains all 260 pieces that fall in the range 7,0-7,9%. If all the column values are connected a so-called normal distribution curve is obtained and then an ordinary pocket calculator with statistical functions may be used for simple calculation of the standard deviation, s. Fig.1. Variation of average MC around the target of 12 and 18% MC; left: narrow, right: wider variation The moisture content variation within a timber batch A further important characterization of the moisture content is the variation between individual boards in a batch. The variation arises due to natural differences in the material and depending on how uniform the climate in the kiln cross section is. The variation can be considerably reduced by an equalising phase at the end of the drying process. The moisture content variation in a batch, i.e. the spread in the moisture content, is close to normally distributed and thus symmetric in relation to the mean value and it is then correct to give it using a statistical measure of spread, i.e. the so called standard deviation, s. The standard deviation is best described by the graphs in Fig. 2 and % 14 s 2 s Fig.3. If the moisture content values are normally distributed, it holds that the orange area A (mean ±s) covers 68% of the moisture content values in the batch. In practice the yellow area B is used, i.e. mean value ±2s. Then the limits within which 95% of the moisture contents are located are given. In the graph this corresponds to 8%±3% (2s=±2 1,5) with 8 being the mean value. A B 5

6 Variation within the timber cross section The moisture content variation in a cross section of a piece of timber is due to the fact that drying process moves inwards from the surface, i.e. the surface is drier and the internal parts are wetter. In practice it may be equally important to specify these variations in moisture content, as well as the spread in moisture content within the whole batch. But these quality factors have so far not been used in product specifications. The moisture content variation in the timber cross section is often called moisture content gradient. Re-sawing timber with a moisture content gradient in the cross section gives pieces where one side of the piece has a higher moisture content and the other a lower. It is important to avoid this drawback, for instance when gluing components or plates. The most common example is a joint in glulam where the gluing performance starts to deteriorate for differences higher than 2% moisture content between the lamellas. Fig.4. Recently dried timber with a mean moisture content of 16%. The variation in the cross section illustrates the moisture content gradient. Re-sawing a piece of timber with a gradient in the moisture content gives pieces with differing surface moisture contents. Fig.5. In gluing, for instance panels for furniture, the result is at risk if the difference in moisture content between two adjacent lamellas is higher than 2%. For glulam beams the maximum allowable moisture content difference between lamellas is 4 % (EN 386). 6

7 Drying induced stresses When green wood is dried, the outer parts of the wood piece start to shrink at the same time as the inner parts retain its moisture and volume. The result is that the outer parts either strain in an elastic or plastic manner (the latter is called casehardening) or crack. In properly controlled drying, there will be strain and creep in the outer parts with no cracking, cf. Fig.6. When the inner parts later in the drying process start to shrink, the outer overstretched parts become too big for the internal volume. Normally this happens when the average moisture content in the cross section has reached 18-14%. Then the outer parts try to stretch the inner parts and the inner parts try to compress the outer parts. The outer parts resist the shrinkage in the inner parts. This creates drying stresses in the timber. Compression stress will thus occur in the surface layer and tension in the inner parts, see Figure 7. If the timber later on is stored in a warehouse by the supplier or the customer, the moisture gradient in the timber cross section will decrease. A smaller moisture gradient means in reality that the moisture content in the inner parts will decrease further and normally the surface layer will absorb more moisture from the ambient air. The inner parts try thus to shrink even more but is resisted by the surface layer, i.e. the drying stress level increases. The compression stress in the surface layer increases further, due to the overstretched outer parts. Checks generated earlier in the surface may be pressed together and are no longer visible. The level of drying stresses can be indirectly obtained by measuring the stress induced deformations in a so called slicing test according to ENV More details are found in section Case-hardening measurement. If the timber has been conditioned in the kiln at the sawmill in order to reduce the drying stresses, then the moisture content gradient is reduced as well which also is a prerequisite for a stable timber. The higher the temperature during the timber drying, the more stretching occurs and the higher residual stress in the dried timber is the result. If the timber is not conditioned correctly the stress increases as the moisture gradients level out after the kiln. The timber will more or less deform during a subsequent cutting process in the fibre direction (board direction). Because of that a so called slicing test should be done as a quality control measure on delivery, cf. the section Case-hardening measurement and the standard ENV Fig.6. As green timber dries, tension stress (blue arrows) develops gradually in the surface layer, at the same time as compression stress (green arrows) develops in the inner parts. If the tension stress exceeds the strength of the wood a drying check will occur. Fig.7. After drying at the sawmill to, for instance to a cross section moisture content average of 12%, the inner parts of the timber have reached about 14% while the surface layer is about 6%. As the timber surface layer has been overstretched in an earlier part of the drying cycle, then tension stress develops in the inner parts as these start to shrink. The more the moisture content evens out in the cross section, the higher stresses arise. 7

8 Drying checks As mentioned above, checks can develop as the timber starts to dry in the outer parts surface checks. End checks develop because green timber dries much faster from the end surface than perpendicular to the grain. Checks can occur both in living trees (usually in big logs) and due to inappropriate drying conditions. Checking is not further considered in this booklet, because checks are defined and measurements well described in the normal timber grading rules. Distortions In addition to the previously mentioned distortions of spring, bow and twist, cupping of the timber may also occur. Cup develops because shrinkage in the direction of the annual ring is about double compared to shrinkage across the annual ring. Cup is hardly apparent at about 20% moisture content but increases more and more at lower moisture contents. Cup is more pronounced in wide timber than in narrow timber. After re-sawing the amount of cup is dependent on the level of casehardening. This is because the stress-induced cup is combined with the normal shrinkage-related cup, see Fig.8. Timber for boarding is often re-sawn from thicker dimensions. Timber prone to twist, such as timber sawn close to the pith, can be taken care of by top loading or by reversed pre-twisting during the drying process at elevated temperatures. Fig.8. Cup Bow Twist Crook /Spring 8

9 Selection of a suitable target moisture content In addition to the fact that there is a maximum allowable moisture content level for avoiding mould or rot in the wood, it is obvious that the moisture content from the beginning has to be adapted to the climate where the wood will be used, in order to avoid problems from shrinkage, deformation/distortion, gaps, checks etc. Consequently some target moisture content values have been selected that are considered suitable for use in practice for different purposes. Equilibrium Moisture Content The moisture content in wood depends on the relative humidity (RH) and the temperature (T) in the surrounding air. If a piece of wood is placed in a room with a constant climate, then the moisture in the wood will gradually attain a moisture content that is in equilibrium with the climate (T and RH), the so-called Equilibrium Moisture Content (EMC). If the wood moisture content is lower than the value corresponding to the surrounding air RH, then moisture is absorbed by the wood and it swells. If the wood moisture content is higher than the equilibrium moisture content, then moisture is released and the wood shrinks. The wood moisture content is adjusting towards the equilibrium moisture content. The red curve in Fig. 9 shows the equilibrium moisture content for different RH in the temperature range 5-22 o C. The time required for attaining equilibrium depends on the temperature, initial moisture content, difference between initial and EMC, thickness, species, density and whether there is a moisture protecting surface treatment. The climate is in practice seldom constant but changes in time and the wood is all the time adapting to the prevailing climate. In order to avoid too much shrinkage, swelling, cup or twist, it is necessary to dry the timber to a target moisture content that approximately corresponds to the annual mean equilibrium moisture content. Wood tends to approximately reach the same equilibrium moisture content for a given RH. It is thus possible to present a relation between RH and timber equilibrium moisture content, seen as the red curve in Figure 9. EMC % Fig. 9. Wood attains an equilibrium moisture content that is governed by the surrounding relative humidity and temperature. Some species may deviate from the average curve (e.g. Robinia and Teak show lower, Larch and Poplar show higher EMC). rh % 9

10 EMC % rh % 30 Fig.10. If the RH is kept constant (red lines) then the wood equilibrium moisture content (EMC) changes only slightly when the temperature changes. Example: if an air dehumidifier is put into a cold storage/timber storehouse and set at RH 60 %, then the equilibrium moisture content will change only 1 %, even if the temperature v aries between 0 C and 30 C. Different climates The climate surrounding a piece of wood thus determines its moisture content. As the climate changes the wood moisture content will also gradually change. The most important climate parameters are temperature and RH, i.e. the amount of vapour in the air. Although it is the RH that primarily determines the wood moisture content, the temperature has indirectly a great importance. This is due to the fact that a change in temperature will strongly influence the RH. If the temperature is changed, but the RH is kept constant, then the change in wood equilibrium moisture content is however small, as seen in Figure 10. For a given temperature in the surrounding air the wood equilibrium moisture content is determined by the air RH. If the temperature is changed, the RH will normally be changed too and thus the equilibrium moisture content. A temperature change from 20 o C to 40 o C will decrease the equilibrium moisture content about 1 %-unit if the RH is kept constant. But in practice the equilibrium moisture content will decrease much more as the RH in a room at the same time will strongly decrease due to a temperature increase. This can be seen in a Mollier-chart for humid air. As an example; in a sealed room with RH 80% and temperature 10 o C the equilibrium moisture content is 16 %. If the temperature is increased to 30 o C, the RH will decrease to 23% and the equilibrium moisture content to 5 %. The equilibrium moisture content is thus decreased by 16-5 = 11 %. Another example: If the temperature is increased from 0 o C to 30 o C in a body of air that at 0 o C has a RH of 60 % then the RH at 30 o C will be 9 %. For a higher air temperature the more vapour can be held by the air. (RH is defined as the ratio of actual vapour content to the maximum possible vapour content at each temperature.) The indoor climate in heated buildings is relatively dry in the winter, i.e. the RH is low, while the outdoor climate at the same time is rather humid, i.e. the RH is high. It is important to find out the correct RH in each individual case; there may for instance be humidifying devices. In swimming baths, in un-heated summer cottages in winter time and in similar premises, the RH is relatively high. The outdoor climate varies a lot during the year, but also during one day. In summertime, in a warm and dry afternoon, the RH can be about 50 %. As the temperature decreases in the evening, the RH may rise to 90 %. It is thus important to find out the RH level in each individual case and the outdoor RH varies geographically too. 10

11 The wood tends all the time to adjust to these variations in the climate. But the timber dimension and type of surface treatment may strongly influence how fast the wood moisture content adapts to the actual climate. Thicker dimensions react more slowly to sudden changes in the climate. A paint/varnish coating will slow down the timber moisture absorption/desorption. In some constructions, such as outer walls, outer doors etc., the climate may vary a lot between the inner and outer sides of the object, depending on the time of the year. EMC % rh % Month Fig.11. Monthly average values for relative humidity RH (red) and equilibrium moisture content EMC (blue) for wood in Stockholm during a whole year. The dotted line represents the annual average equilibrium moisture content As guidance when drying timber, it can be said that for indoor use in the Nordic countries with heated ambient air the target moisture content should be 8-9 %, i.e. slightly below the annual mean value according to Figure 12. A wooden product for indoor use should thus be produced with a target moisture content between 8 % and 9 %. The product should retain its quality for all monthly average moisture contents during the year in the country in question. If the timber is intended for outdoor use, sheltered from rain, the target moisture content should be %. This corresponds to the normal climate variations in the Nordic countries, from spring-summer with the driest climate to autumn-winter with the wettest climate, as seen in Figure 11. For some wooden indoor products, such as construction timber, glulam, inner lining etc., a higher shrinkage/ swelling movement than for joinery may be allowed. Such products should be as straight as possible and a higher target moisture content than the annual mean equilibrium moisture content is allowed, i.e. a target moisture content %, c.f. the chapter Target moisture content selection by the wood industry. At lower moisture contents the risk of greater twist in the timber increases. There are product standards for many common wood products. 11

12 Many times the target moisture content is given in an unclear way in the product standards often as an absolute demand, which is not correct. As mentioned earlier, it is not possible for the sawmill to dry every piece of timber to the same moisture content. There will always be a deviation in the mean moisture content from the target moisture content (ordered moisture content). There will in addition always be a variation in moisture within the batch, even if the sawmill has performed a moisture equalisation phase. It is thus not possible to buy timber for which the seller to 100% can guarantee that the moisture content is within given limits. In Europe a standard ENV has been developed that makes it simpler for the buyer to select a certain quality level and to verify it, in a statistically correct way without complicated statistical calculations, as will be discussed later. Selection of suitable moisture contents when using wood indoors and outdoors in Sweden To avoid checks and distortion a wooden product or construction should not shrink or swell too much due to the climate. Wood should because of that, when installed, have a moisture content that corresponds to the annual mean equilibrium moisture content. Such a target value can be determined from the graph to the left. An example for southern Sweden (Malmö) in November Outdoors: RH = 88% (average for Nov.) Equilibrium moisture content is about 21% Indoors: RH = 32% (average for Nov.) Equilibrium moisture content is about 6,5% Fig. 12. Alignment chart for the determination of equilibrium moisture content from climate data. Red line = Luleå, blue line = Malmö; full lines = RH outdoors, dotted lines = RK indoors These values are only valid if no additional moisture occurs, such as from cooking, evaporation from people or animals, bath, washing etc. This additional moisture may in Malmö increase the RH by 18 %, i.e. from 32 to 50%. This corresponds to an increase in equilibrium moisture content from 6,5 to 9,5% according to the green curve. In Malmö during summer the corresponding equilibrium moisture content change is 12 -> 15% with windows and doors shut. In northern Sweden (Luleå) the equilibrium moisture content is slightly lower indoors in November 3,5% and increases to 6,5% due to additional moisture and in summer the corresponding values are 10,5% -> 14,5%. 12

13 Conclusions When producing wood products in Sweden for indoor purposes, the moisture content in the timber should correspond to the average annual equilibrium moisture content including additional moisture sources, i.e.11 % in southern Sweden and 9 % in northern Sweden or 10 % for Sweden in general. From a practical point of view, it is often an advantage if the moisture content is slightly lower than this theoretical value, as a joint will open from shrinkage but can withstand some compression from swelling. It is thus probably better to instead of a target of 10 % to use a target moisture content of 8-9 %. According to the same reasoning should timber in unheated rooms have a target moisture content of %. Example of suitable target moisture contents in European countries The aim is that the target moisture content for a wood product should correspond to the annual average equilibrium moisture content (EMC) in the place (indoors or outdoors) where the product will be used (rounded off to the next whole number). This annual average varies between countries but it is quite similar in many countries. Table 1 gives a simplified overview of EMC-values for some European countries. The average EMC is the average of EMCs in January and July. A better alternative is to find the EMC s for each month and average them. Table 1. Rough target moisture contents for wood products outdoors in some European countries. Country EMC % January EMC % July EMC % annual average EMC % practical annual average Iceland Great Britain The Netherlands Poland , Germany Norway Spain Greece Turkey To determine the correct target moisture content for indoor wood products tha local/regional heating and air conditioning habits have to be taken into account. Please note that within countries and regions substantial regional differences may occur. 13

14 Target moisture content selection by the wood industry At the moment (2008) the wood using industry has developed a number of EN-standards that, inter alia, describe the moisture requirements for wood products, Table 2. Table 2. Today s moisture content requirements from some product standards for wood products (2008). The MC expression used in the standard is shown in bold. In some cases MC requirements have to be recalculated to match the MC specification system described in EN (first column of MC specification. Sawn and planed timber components or end Moisture content products EN MC range Timber in joinery (EN 942) 1 - internal use, heated buildings > 21 C 8 ± 2% 6 10% - internal use, heated buildings C 11 ± 2% 9 13% - internal use, unheated buildings - external use 14 ± 2% 16 ± 4% 12 16% 12 19% 2 Wood flooring elements (EN 13228) - individual elements - individual chestnut elements 9 ± 2% 10 ± 3% 7 11% 7 13% Timber planks and semi-finished profiles 3 (EN 13307) - internal use (e.g. doors, stairs) 9 ± 3% 6 12% - external use (e.g. doors, windows) - weather exposure (e.g. fences, stairs) 12 ± 3% 15 ± 3% 9 15% 12 18% 4 Hardwood floor boards (EN 13629) - individual element 9 ±3 % 6 12% 5 Softwood floor boards (EN 13990) - for internal use, heated buildings 9 ± 2% 7 11% - for other uses 17 ± 2% 15 19% 6 External windows, door leaves and door 16% frames (pren 14220) Internal windows, door leaves and door frames (pren 14221) - heated buildings - unheated buildings Machined softwood profiles with tongue and groove (EN 14519) - internal use - (MaritimePine) - external use Machined hardwood profiles (EN 14951) - individual panelling element - individual cladding element Machined softwood profiles without tongue and groove (pren 15146) - internal use - external use 12 ± 2% (11 ± 3%) 17 ± 2% 10 ± 3% 15 ± 3% 12 ± 2% 17 ± 2% 13% 16% 10 14% (8 14%) 15 19% 7 13% 12 18% 10 14% 15 19% 14

15 Unfortunately the MC requirements expressed in these standards do not always comply with the requirements described in EN Sawn timber Assessment of drying quality. This standard should be used and referred to when buying timber for further processing and the building industry. By applying the standard specification of drying quality is facilitated. Correspondingly, wooden products can be manufactured from timber with greater dimensional stability. The cost of such timber with better drying quality is only slightly more expensive, but the higher price is justified by the higher quality. The dimensional stability is improved as the drying induced stresses and the moisture gradient are minimized at the same time as the moisture content spread in the batch gets smaller. Then it also becomes easier to hit the correct target moisture content. An additional benefit of using EN is that the timber buyer does not necessarily need to perform a full moisture content check. Both types of moisture meters (resistive and capacitive type, online and hand held) have a lower measurement accuracy as compared to the oven drying method. Nevertheless both types are widely used in industry because they are non-destructive, easy to use and rapid. The order of magnitude the measurement inaccuracy (see the next chapter) is about 2 % MC. Many sawmills are today able to dry timber with a high precision provided that they have got precise requirements regarding the moisture content specified, i.e. expressed according to the new European standard EN Distinct requirements will lead to increased competition between suppliers because it becomes clear which sawmills can fulfil a certain quality level. A sawmill/supplier may also more easily find out where the shortcomings are. Established techniques are available for achieving a high drying quality. Using modern equipment and techniques the sawmills are able to hit the target moisture content within the ranges given in table 3 for the different moisture content classes. The standard was put together by industrial specialists and wood moisture content researchers in Europe. It was also formulated according to the drying technology available today for sawmills. This technology can produce considerably better dried timber than was possible earlier. The standard EN refers also to the method of selection as described in TS (Criteria for the assessment of conformity a lot of sawn timber). This standard for conformity stipulates how to statistically confirm that a delivered lot fulfils the drying quality the buyer has ordered. The seller and the buyer of timber do thus not need an arbitrator in a dispute. Both these standards are now included in the standard contract Softwood Contract Form that is used in the Nordic countries, UK and Ireland. EN helps the joinery, flooring, panelling industry etc. to order timber that suits their requirements with respect to moisture content. The standard is put together in such a way that it is easy to perform a reliable and unambiguous inspection of the drying quality delivered. That is why EN and ENV use the so called standardized quality control system based on attribute control (Yes/No decision) instead of the alternative variable control (statistic calculation with measured values). Values from the attribute control can be converted into variable control values, that in some cases may better suit sawmills/suppliers. Many sawmills are already using the parameters of variable control in delivery assessments, i.e. mean values and standard deviation. 15

16 According to the new standard EN the customer may choose a target moisture content between 7% and 18%. However, in order to decrease the number of MC levels, the business sector should select a few important target moisture content values that can be used for the most common environment (see Table 3). The business sector should as a suggestion prioritize the six target moisture contents given in the following table (expressed according to EN Standard drying ). Table 3. Suggested target moisture content values for Europe Target moisture content % 7±1 10±1,5 12±1,5 15-2/+1,5 16-2,5/ ,5/+2 Standard deviation (s) when MC targ is hit/max deviation from MC targ, % 1,1/0,7 1,6/1,0 2,0/1,4 2,4/1,7 2,6/1,5 2,9/1.9 Examples of the environment where the product is going to be used Indoors, heated buildings, internal joinery, floors in northern Europe. Indoors, heated buildings, internal joinery, floors in central Europe Outdoor joinery in southern Europe Indoor/outdoor for certain products in Europe Indoor/outdoor construction timber in Europe Indoor/outdoor construction timber in Europe (however, risk of mould for a surface moisture content 18 at 20oC during 4 weeks (pine)) When selecting target moisture content levels for wooden products the annual average equilibrium moisture content should be the starting point. It is recommended that this annual average value is reduced by 1-2 %-units, as discussed earlier. The sawmills do not normally produce all moisture content levels, unless a very big order (ca 100 m 3 ) is obtained. It is thus convenient to group the most common target moisture content levels according to Table 3 and adapt the order to these levels. Fig. 13: Average equilibrium moisture contents in European regions 16

17 Moisture content measurements EN In this chapter the method for determining the average moisture content of a piece of timber by three different methods is described. These methods have been standardized in three European standards and are included in standard contracts, for instance in NORSO F (for the Nordic countries, UK and Ireland). In the chapter Drying quality the methods for determining the average moisture content and the moisture content variation/standard deviation for a batch of timber using these measurement methods are described. Three measurement methods There are several methods for the measurement of the moisture content in wood. The three European standards for moisture content measurement are as follows: EN Moisture content of a piece of sawn timber. Part 1: Determination by oven dry method, EN Part 2: Estimation by electrical resistance method, EN Part 3: Estimation by capacitance method. The standards are not only valid for sawn timber, but also for planed or timber machined in another way. One of the goals with these standards is that the measurements should be done in a similar way in both the delivery and acceptance controls. Differences between measurement methods The oven dry method is the reference method. With that method the MC of the whole cross section of the piece of timber is determined. The other two methods are indirect methods that measure a part of the cross section. This results in an estimate of the MC. Measurements using the reference method require a rather long time, while the indirect methods are faster but have a lower accuracy. The capacitance method is generally inferior to the resistance method. If a batch of timber is controlled, it is recommended to use an indirect method first preferably the resistance method. If the moisture content seems to deviate too much from the target, then the checking is continued with the oven dry method. When the indirect methods are used, the meters should always be correctly calibrated according to the wood species in question and the wood temperature, in order not to have excessive systematic deviations from the reference method. Resistance meters should be calibrated for each wood species against a standard resistance block approved by research or certification institutes. Some of the capacitance meters have a kind of a calibration unit (normally not controlled or certified for the actual wood species by any independent body). 17

18 Moisture content determination with the oven dry method/reference method according to EN The measurement should be performed in the following way: Cut a sample of at least 20 mm in the fibre direction, at a distance of at least 0,3 m from either timber end. If a defect such as knot, resinous wood, bark etc. occurs, then a sample further in from the end is taken, until a sample free from defects is obtained. For a piece of timber shorter than 0,6 m the sample is taken in the middle of the piece. The sample is immediately weighed using a scale with a resolution of 0,01 g to determine the mass of moist wood G 1. For samples above 100 g a scale with a resolution of 0,1 g may be used. The samples should be put into a ventilated oven. Each sample is then dried in the oven with a temperature of 103±2 o C until the weight does not change more than 0,1% between two successive weighings two hours apart (in practice after 8-20 hours depending on oven size, sample size, heating up time, heating power, number of samples in oven, etc.). In order to achieve a high measurement accuracy for timber that contains a high amount of volatile compounds (resin), the sample can be dried in vacuum (pressure < 100 Pa) at low temperatures (max 50 o C) or in a desiccator containing a hygroscopic substance. The sample is then immediately weighed to determine the oven-dry weight, G 0. The moisture content MC expressed as a percentage is then MC = (G 1 - G 0 )/G 0 x 100% or easier to calculate MC = (G 1 /G 0 1) x 100%. The result is rounded off to the nearest 0,1 % MC. (The standard states that the moisture content should be denoted by the Greek letter ω, but in practice most often the old notation u or the abbreviation MC is used.) Fig. 14: Follow up of action for conduction the oven dry method for determination of moisture content 18

19 Estimation of the moisture content by the electrical resistance method according to EN About the measuring instrument The resistance moisture content meters found in the market are of various designs. The measured electrical resistance is transformed into moisture content. The relation used for a species may be differ depending on the brand of the instrument. For Scots pine (Pinus silvestris) and Norway spruce (Picea abies) a type of resistance box has been developed by SP Trätek and European Drying Group (EDG). With such a box all types of resistance meters can be checked. The electrical resistance method is suitable for the measurement of moisture contents between about 7 % and 30 %. For moisture contents above 20 % the measured values become increasingly more unreliable. The electrode contact area will slightly influence the moisture content reading. The moisture gradients in the timber influence also the results of the measurement. Electrodes are thus insulated, except for the slightly thicker tip. The insulation prevents wet surfaces from influencing the measurement. A resistance meter should be equipped with settings for wood species and with wood temperature compensation. Where no species and temperature correction switches are avaible, correction tables provided by the meter manufacturer have to be used. The electrodes of the meter should be of a type that can be hammered into the wood and be long enough to penetrate to a depth of 0.3 times the timber thickness. A length of about 25 mm is generally enough. Wood treated with salts-based preservatives will generally give meter readings that are too high. The measuring accuracy has been specified by the Technical Research Centre of Finland, VTT, in an extensive survey. Measurements were carried out with 17 different meters. In the research study it was shown that in the moisture content range % the meters with long measuring tips (8 mm) on the electrodes gave up to 3 %-units higher values than those with short measuring tips (3 mm). This means that for measurements with different measuring tips, specific calibrating values in the calibration box have to be used. Resistance (MOhm) model % conf. interval 97.5 conf. interval data Actual moisture content of wood (%) Fig 16. Example of the relationship between moisture content and resistance for Nordic Pine with 95 % confidence limits. The test material had been kept in a constant climate during a long time. The measuring error is greater at higher moisture contents. 19

20 Measurement according to EN The following measurement procedure shall be used: Calibrate the instument with an approved resistance box. Set the species and wood temperature correction or correct the measured value with the aid of the correction tables delivered with the instrument. Drive the electrodes into the face side of the piece, normally along the fibres, at a distance of at least 0,3 m from either end of the piece (or at the mid point of pieces less than 0,6 m long) and at a distance of 0,3 times the width from one edge, so that the tips of the electrodes penetrate to a depth of 0,3 times the thickness of the piece (see Figure 15). The measurement area shall be free from resinous wood and features such as bark, knots and resin pockets. If such features exist, the measurement shall be taken at the nearest clear area towards the centre of the piece. Record the reading after it has been displayed for 2-3 seconds. The moisture content reading shall be expressed at least to the nearest 1 percentage-point moisture content. Increasing the number of measurements in individual test pieces does not increase significantly the accuracy of the result when estimating the moisture content of a lot or consignment. If necessary, for example when estimating the moisture content of a single test piece, or estimating the moisture content of a very small lot, sampling and testing frequencies should comply with those given in Table 4. The individual test results should be recorded together with at least the following: specification of the lot of timber (number, internal coding, supplier, customer... etc), species, dimensions, date, type of instrument used, species setting, temperature setting, penetration depth. Number of tested pieces Number of measurements per test piece >5 1 Measurements should be taken at random along the length excluding 0,3 m at each end (or at the mid point of pieces less than 0,6 m long). All measurement results should be noted. Fig. 15. Measurement (estimation) of the average moisture content of a piece of timber according to EN Depth of electrode penetration is 0,3 times the timber thickness (c), distance from edge is 0,3 times width (b) of the board, distance from end (a) is 0,3. Table 4. Number of measurements (according to Annex A in EN ) if a lot contains less than five pieces 20

21 Estimation of the moisture content by the capacitance method according to EN About the measuring instruments Moisture meters using the capacitance method as the measuring principle can be classified into hand-held and in-line moisture meters. The measuring principle is based on the fact that wood and water have very different dielectric constants (є water 80, є wood 4). When wood which contains moisture is placed between the plates of a capacitor, the capacitance of the capacitor is changed. In principle, capacitance type moisture meters are able to detect water in wood over a much wider range of moisture content than is possible with electrical resistance type moisture meters. As the capacitance method reacts to the amount of water in the vicinity of the capacitor plates of the instrument, the method is heavily affected by the density of the wood to be measured. Let s assume timber with an average density ρ 0 = 400 kg/m3 and a moisture content of 12%. This timber contains 48 kg of water per m³. Density in the lot of timber varies from ρ 0 = 320 kg/m³ to ρ 0 = 550 kg/m³ which corresponds to 38 kg and 66 kg water Measurement according to EN In the case of hand-held capacitance type moisture meters good contact between the measuring plate and the surface of the timber is crucial. An air gap strongly affects the reading. Wet surface also has a negative effect on the reading even though the overall effect of surface moisture content on average moisture content may be very small. Positions where to measure moisture content on the timber are similar to those described in EN per m³ respectively. As the measuring result of the capacitance type moisture meters is linearly related to the amount of water (and not to the moisture content percentage) the very strong impact of density variation becomes obvious. Capacitance type moisture meters must be calibrated for density and not for species and temperature. The inherent natural density variations in assortments of timber are the major reason for the apparently lower accuracy of the capacitance type moisture meters. Capacitance meters are also affected by the moisture content distribution in the piece of timber to be measured. Nevertheless, the reading of a capacitance type moisture meters is estimating the average moisture content of a piece of timber whereas the electrical resistance type meters estimate the local moisture content between the tips of insulated electrodes or the wettest region in contact with the non-insulated electrodes. It is more difficult to check or calibrate capacitance meters. In case of in-line moisture meters, also the geometry of the measuring set-up strongly influences the measuring result. This is the reason why capacitance type in-line moisture meters have to be calibrated at the site where they are used and for each species and dimension of sawn timber. Annex a in EN describes a procedure how to determine and express the accuracy of an capacitance-type moisture meter. 21

22 Shrinkage and swelling of wood Wood shrinks and swells different amounts in different directions. Changes are smallest in the fibre direction and largest tangential to the annual rings. Different species move different amounts when the moisture content changes. This has to be taken into account when moisture related movements are critical for the product. Orientation and moisture movement Fig.16. Cupping is positive (concave) on the outside face A and negative (convex) on the inside face B. The negative cupping at B is larger than the positive cupping at A. Maximum shrinkage in each direction occurs when the moisture content decreases from ca. 30 % down to 0 %, i.e. absolutely dry wood. 30 % moisture content represents the so called fibre saturation. Down to this level the moisture leaving the wood is free water located in voids of the cell structure. Shrinkage will not occur until the water located in the cell walls in the fibres is removed. Axial radial tangential The Table 5 shows average values for the maximum shrinkage as a percentage when the moisture content changes from 30 % down to 0 %. Table 5. Values for total shrinkage in % (from green to oven dry) Species Axial Radial Tangential Pine 0,4 4,0 7,7 Spruce 0,3 3,9 8,3 Birch 0,6 5,3 7,8 Beech 0,3 5,2 10,9 Oak 0,4 4,5 8,9 Fig. 17. Standard sawing pattern producing main product and side boards Dimensional changes depend on the amount of the MC change and the location of the board in the log. Movements are smallest in the axial direction and largest in the tangential direction. Cup is the most common manifestation that wood has different deformations in different directions. Minimum cupping is in boards with vertical annual rings. The wider the board is and the closer to the pith it is cut the larger is the cupping. Such boards are 22

23 sawn radially from a log. This requires a special sawing system. Within a species the wood density clearly affects the amount of shrinkage and swelling. This is seen in Fig. 22. The same dependency may be used for spruce also. Because shrinkage and swelling are proportional to the moisture content change it is possible to calculate dimensional changes in any moisture content interval between 0 and 30 %. Please note that the values presented are averages and the deviation can be large even between pieces taken from the same tree. Calculation of moisture related movements To be able to calculate the moisture related movement of wood, the variation in climatic conditions must be known. Because the moisture related movement of wood is proportional to changes in moisture content of wood, it must be calculated from variation in climate conditions by using the changes in the corresponding equilibrium moisture content of the wood. Sawn timber has a big variation in wood density and therefore a big variation in shrinkage/swelling as well, as can be seen from Figure 19. Example how to calculate shrinkage: Assume a table-top of pine wood, width 1000 mm, glued from boards with annual rings parallel to the flat side. How much will the table-top shrink if the relative humidity changes from 60% in autumn to 30% in winter? We are interested in the moisture related movement in the tangential direction; the direction where the moisture movement is the largest. According to Figure 12, the equilibrium moisture content is 11% and 6% for the relative humidity of 60% and 30%, respectively if the indoor temperature is 20 C. The corresponding change of moisture content is 11-6 = 5%. If the moisture content of pine wood changes from 30% (fibre saturation) to 0%, the shrinkage in the tangential direction is 7,7% (see Table 5). The shrinkage of the tabletop will be: 7,7/100 5/ mm = 12,8 mm It should be pointed out that the calculations like this are inexact, because the variation between individual sawn timber pieces can be remarkable. However, these rough calculations give a good conception of the actual moisture movement. Fig. 18. No cupping will take place in radially/quarter sawn timber (left hand side). The closer to the pith and the wider of the board, the larger is the cupping (piece on the right). A wide and thin centre piece will show high cupping. Fig.19. Shrinkage of pine in the tangential (green curve) and in the radial (red curve) direction for different basic densities (oven dry weight/green volume) according to F.E.Siimes, Finland

24 Mostly, the growth rings curve in the cross-section of sawn timber. A good estimation of the average shrinkage for the width and the thickness of timber sawn is 0.25 % per 1% decrease in moisture content. For this reason this value is given in a number of European standards. The applicability of this rule of thumb for some standard dimensions of sawn timber can be seen in Figure 20. Fig.20. Dimensional change (shrinkage and swelling) in mm of different dimension if moisture content changes for 1% (average for pine and spruce) Measuring drying stresses, ENV The residual drying stresses in the dried sawn timber are perhaps the most important factor for the drying quality. The drying stresses will lead to material loss and rejection after longitudinal cutting and further processing. For example, case-hardening can occur in boards after they are cut parallel to surface and planed from sawn timber. In order to control the material a slicing test method has been developed, and sawmills have made investments on conditioning phase in the drying kilns to reduce the stresses. Case-hardening measurement - How to do a slicing test The degree of deformation, due to drying stresses, can be measured by sawing a 15 mm wide slice, a so-called slicing test, from the sawn timber piece. When a sawn timber lot is controlled for drying stresses, a larger number of specimens have to be taken. Slicing test procedure according to ENV 14464: - Saw a 15 mm thick cross-section with a circular saw, see figure 24. Each sample is taken 300 mm from each end. If the timber is shorter than 600 mm the sample can be taken from the middle. - The case-hardening sample should be free of bark, knots, resin pockets and reaction wood (compression wood/ tension wood). Avoid resinous wood. If such defects exist cut out the casehardening sample at the nearest faultless area towards the sample middle. - If the board is wider than 100 mm there are two possible ways to carry on: a) cut the sample to 100 mm and measure gap directly by putting the slices on top of 24

25 each other, b) use a measuring jig as shown in Fig Each sample is divided in the middle, parallel to the board surface, into two pieces. Use a band saw, circular saw or a knife and hammer. Mark the pieces with 1a and 1b. - Place the two pieces from the case hardening sample in a plastic bag and seal it. Store the sample in room temperature for 24 hours (softwood)/48 hours (hardwood) to equalize the moisture gradient in the sample. The measuring of the case hardening gap is preformed when the equalization is finished. - Use a measurement jig like the one in figure 24. If the sample is between 75 and 100 mm wide use a jig with a distance of 75 mm between the 10 mm thick emplacement pins. - The measuring of the case-hardening should be performed with a calliper, a dial indicator or a wedge with a resolution of 0,1 mm. Fig 21. Procedure for carrying out a slicing test for assessment of case-hardening according to ENV The pieces from the case hardening sample are placed in the measuring jig so that the split surfaces are facing the 10 mm thick alignment pins. The case hardening samples annual rings should be oriented in the same way as when the case hardening sample was split, see figure 24. Measure the gap width. Subtract the alignment pins thickness from the measurement, 25

26 for example if the measurement is 12,3 mm the gap will be 2,3 mm with 10 mm thick alignment pins. If the measurement was done with 75 mm between the emplacement pins the value is multiplied with the factor 1,78, i.e. the case hardening gap is 2,3 x 1,78 = 4,1 mm. Requirement level Up to now limits for case-hardening are not defined for different products. ENV only describes the assessment procedure but does not set any limits. The following casehardening levels suggested: LOW: MODERATE: SEVERE: 80 % of the samples with case hardening gap < 1 mm 80 % of the samples with case hardening gap < 2 mm 80 % of the samples with case hardening gap < 3 mm For example, material for furniture, joinery and floors of high quality should have a low case hardening level. Such wood should also have a target moisture content adapted to the products environment and a small moisture content variation, read under the heading Selection of a suitable target moisture content. Strutural timber which will not be cut lengthwise parallel to the flat side does not require a low level of casehardening. Moisture content requirements are also not so strict because higher MC can be accepted. Reduction of casehardening Sawn timber which is conditioned in a proper way at the end of the drying process is free of stresses. The stresses in sawn timber decrease during conditioning by using an appropriate conditioning temperature and relative humidity regime. By raising the moisture content in the surface layers the compression stress in the surface layers is increase so that the permanent set (plastic deformation) which was generated during initial stages of the drying process is reduced or even reverse. Simple equalising moisture content does not result in a reduction of case-hardening. To achieve a good conditioning a sharp increase of moisture content in the surface is necessary. Conditioning is effective if the EMC in the kiln is raise at least 1-2% above the target moisture content level. It sounds simple, but requires a great deal of technical expertise, time and additional energy. Due to the fact that the conditioning substantially increases the quality of the sawn timber, modern sawmills have made investments which enable them to produce conditioned sawn timber with low level of case-hardening. During storage of timber, moisture content gradients will level out over time. Without a proper conditioning treatment at the end of the drying process stress level will not decrease but increase considerable. Therefore, waiting is not a solution to the case-hardening problem. 26

27 Drying quality according to EN and ENV Principles in the two recent standards for quality control of dried timber EN was developed to facilitate specification and control of drying quality for producers and users of sawn timber. In using EN one starts by determining a relevant requirement for the target moisture content and the maximum deviation of the mean moisture content from the target moisture content in a delivered batch. The target moisture content of a batch has to be distinguished from the mean moisture content of the delivered batch. In producing dried timber there are limits regarding how well a sawmill/supplier can hit the target moisture content. People in the wood trade have to realize that they almost never get timber with the desired moisture content but that the moisture content will always deviate more or less. As an example the allowable interval for a target moisture content of 12% is in EN for standard drying given as ±1,5%, i.e. the average moisture content of a batch should be within 10,5...13,5 %. If a heavier demand than the standard demand is desired, then this can be defined by using the procedures described in EN but it requires a lot more work and efforts from the sawmill/supplier, and of course this will result in a higher cost/price. Drying and conditioning/equalizing have to be performed in high efficiency kilns by competent kiln operators. Such drying quality is called in EN specific end use drying. The moisture content requirements in the EN standards listed in Table 2 have been presented in two comparable ways. The ways of expressing the moisture content in the standards are presented in bold. In EN942, EN and EN requirements are presented as 6-10% and in the others as 8±2 %, which is the same thing. It is important that the industry has a single way to express the moisture content. In addition, the moisture content levels in several product standards are different for products that anyway will have the same surrounding climate in the end use. This indicates that the requirements in the different standards have been determined by different working groups. There should have been a better coordination between groups. The groups have determined the requirements for different products in different and equal environments and expressed these in different ways. In addition, the severity of the requirements is different for similar products and similar end use climates. This means that the supplier has difficulty in delivering a suitable drying quality to the buyer of timber. It will be difficult to find timber that fulfils the extreme quality requirements and, if fulfilled, the price will be high. In addition to the target moisture content and the maximum allowable deviation of the average moisture content from the target, the requirements regarding the moisture content variation in a delivered batch have been standardised in EN Table 2 gives the moisture content requirements as a moisture content range expressed using integer numbers. The way the requirements are written in the product standards it means that 100% of the timber should be within the given range. Normal technical requirements are not expressed in this way if it results in very high costs to fulfil the requirements. Further, a control of each single piece would be required which is possible only using on-line electric moisture content meters. (These have however a low measurement accuracy between ±1 % and ±3 % MC depending on moisture content level. To use the oven 27

28 dry method instead would be unreasonably costly and not possible, because it is a destructive way of assessing moisture content.) In order to avoid the drawbacks from 100% requirements, wood moisture content experts have decided to express the requirements as in the drying quality standard EN There the quality assessment standard CEN/TS is used, which has been put together especially for the wood sector, but in general follows normal industrial quality assessment. This means that one, in a statistically correct way, allows a maximum amount of nonconforming pieces. As an example, one can allow that only x % of the random sample moisture content values are within 1,3 times the target and 0,7 times the target moisture content. Acceptance control step by step A quality assessment and sampling system for using AQL 1 principles in accordance to CEN/TS with moisture content requirements according to EN will be described. For standard drying AQL 6,5 shall be used and for specific end use drying with higher quality requirements AQL 4 can be chosen. In some cases, when a lower quality can be accepted, AQL 10 can be chosen. In addition the deviation from the target moisture content may be selected within reasonable limits, i.e. what the sawmill is able to produce. This means that instead of the variation 1,3 x target moisture content 0,7 x target (as in C), another range may be selected, for example 1,2 x target MC 0,6 x target MC. A single sampling is carried out according to steps A to E. Double sampling is carried out from step A to F. Single sampling requires a greater number of pieces to be tested, whereas the double sampling consists of a two stage sampling, the first one with a comparatively small sample and the need for a large sample only if no definitive decision could be made. Table 6. Standard drying according to EN Allowable range of the average moisture content of a lot relative to the target moisture content Target moisture content % Allowable range of average moisture content around target moisture content % 7-1/ / / ,5/+ 1,5 11-1,5/+ 1,5 12-1,5/+ 1,5 13-2,0/+ 1,5 14-2,0/+ 1,5 15-2,0/+ 1,5 16-2,5/+ 2,0 17-2,5/+ 2,0 18-2,5/+ 2,0 1 AQL - Acceptable Quality Level: This is usually defined as the worst case quality level, in percentage non-conforming pieces, that is still considered acceptable. 28

29 Single sampling: Standard drying (standard quality) A. Note the target moisture content you have ordered or which you wanted to achieve in the drying process. (Shall be between 7 % and 18 % according to Table 6). B. Note how much the average moisture content of the lot can deviate from the target moisture content according to Table 6. C. The number of packages to be inspected from the delivered lot is given in Table 7. By dividing the sample size, as given in Table 8 (single sampling), by the number of packages to be opened, the number of test pieces per package is obtained. Round off the result to the nearest lower integer number. If the number of pieces does not agree exactly, select the extra pieces at random. The column for AQL 6,5 in Table 8, shall be used for standard drying. This means that 93,5 % of the pieces in a lot of timber should have a moisture content between 1,3 x the target moisture content and 0,7 x the target. D. The first piece of timber to be measured in each package shall be located in the second layer, because - according to EN the outer layers shall be avoided.. The piece to be measured in this layer is determined randomly (e.g. by throwing a dice). For example, if the dice shows the number three, then it means that the third piece from left shall be measured, see Fig.25. E. To find the next piece to be measured another random number is chosen, e.g. 4. Now every fourth piece is selected for testing, until the necessary number of test pieces per package is reached. The results (number of non-conforming pieces) from all packages are summarized and compared with column A for AQL 6,5 in Table 8 (standard drying). That way the status is obtained for the whole lot, i.e. conformity or non-conformity. If the delivery does not fulfil the requirements above, the lot should be rejected. Double sampling: Standard drying (standard quality) Instead of sampling in one stage according to Table 8, sampling can be made in two stages a so called double sampling, Table 9. This can reduce the number of sample pieces by half if the first sampling fulfils the promised/expected quality requirements according to Table 6. Double sampling is performed first as above (steps A to E) and after that continued according to F below. F. The measuring result after the first sampling can be conformity, non-conformity or uncertain with reference to the agreement. If the result is uncertain, that is when measuring results are between A and R in Table 9, a second sampling is carried out. The result is added to those of the first sampling and compared with the values in the row Total in Table 9. In order for the lot to be accepted, the number of non-conforming pieces of timber has to be equal to or less than A, and it is not accepted if the number of non-conforming pieces is equal to R or more. 29

30 Table 7. Standard and special drying according to CEN/TS and EN Number of packages to be opened. * NB If these packages do not contain the number of pieces required in Table 8 and 9, the necessary additional packages shall be opened Number of packages in the lot Number of packages to be opened or more 4 * Table 8. Standard and special drying Sampling plan by different AQL, single sampling. (S = Number of pieces to be selected for testing; A = Maximum number of non-conforming piece in the samples) Total number of pieces in the lot AQL 4 AQL 6,5 AQL 10 S A S A S A or more For standard drying AQL 6,5 is used and for specific end use drying with higher quality requirements AQL 4 can be chosen. In some cases, when a lower quality can be accepted, AQL 10 can be chosen. 30

31 Table 9. Standard and special drying Sampling plan with different AQL, double sampling. (S = Number of pieces to be taken out; A = Maximum number of nonconforming pieces; R = Minimum number of non-conforming pieces for the lot to be nonconforming; First = sample size in the first measurement round; Total = full sample size including first measuring round). Total number of pieces in the lot First Total First Total First Total First Total First Total First Total or more First Total AQL 4 AQL 6,5 AQL 10 S A / R S A / R S A / R /3 3/4 1/4 4/5 2/5 6/7 3/7 8/9 5/9 12/13 7/11 18/19 11/16 26/ /4 4/5 2/5 6/7 3/7 8/9 5/9 12/13 7/11 18/19 11/16 26/27 11/16 26/ /5 6/7 3/7 8/9 5/9 12/13 7/11 18/19 11/16 26/27 11/16 26/27 11/16 26/27 Fig.22 Selecting pieces in a package (240 studs) by acceptance control. Example: 1) The first piece (no 3) is chosen randomly from one side of the package. The next throw of the dice (here 3) indicates that every third piece is selected for testing. Pieces in the outer layers (red) should not be selected. One extra piece is tested to yield 31 pieces. 31

32 Example of acceptance control single sampling special drying, low requirements A lot of sawn timber has been delivered to an end user of sawn timber. He had ordered special drying. He had selected a target moisture content 15 %, but will accept -3%/+3% instead of -2/%/+1,5%. His quality expectation are not high, he accepted AQL 10. The lot consists of 3600 pieces (studs) in 15 packages. Each package contains 240 studs. According to Table 7 four packages are to be opened. According to Table 8 for AQL 10 a total of only 125 pieces instead of 200 pieces in standard drying are to be inspected, that means 125/4 = 31 pieces per package. 31 times 4 is 124, therefore one extra piece has to be taken at random from one of the packages. The quality controller selected 3 as the first random number. His second random number also was 3. Starting with stud no 3 in line 2 he now selected every third stud for testing until he reached 30 per package. In one of the packages he selected one extra stud at random as can be seen in Figure 22. If more than 21 out of the selected 125 samples are non-conforming (see Table 11), the lot is non conforming, which means the lot is not accepted. Otherwise it is conforming, which means the lot is accepted. Example of the acceptance control double sampling standard drying A kiln operator has to check conformity of a consignment with the quality requirements fixed in a contract. 10 packages, each comprising 120 oak slats for manufacturing of solid parquet floor have to be delivered to the customer. Target moisture content is set to be 8%; standard quality has been ordered. According to table 6 average moisture content of the lot must be within the range of 7%- 9%. 93,5% of all pieces are expected to be within the limits of ±2,4% (8% x 0,7 / 8% x 1,3) around the target moisture content, this means between 5,6 % and 10,4%. The lot consists of 1200 pieces, in 10 packages. According to table 7, three packages have to be inspected. The quality controller decides the use the double sampling plan in trying to minimize testing efforts. According to table 9 he has to select randomly 50 boards from three packages for the fist testing. He decides to inspect every board in every second row until he reaches 16. From package number 3 he tests 2 additional boards. In the fist sample he calculated a mean of 8,8%, which was close to the allowable limit. In addition he found only 5 boards which showed moisture contents higher than 10,4% (8% x 1,3). None of the inspected boards showed moisture content below 7% (8% x 0,7 = 5,6%) would have been the lower acceptance limit. Therefore, the lot can be considered conforming after the first sampling. Let s assume that in the fist sample the quality controller had found an average moisture content of 9% and/or 8 boards with a moisture content higher than 10,4% (upper allowable limit 8% x 1,3), the sample would have to be classified uncertain. The quality controller would have had to test another 50 boards, bringing to total number of test pieces to 100. After summing up all results, the mean moisture content was 8,8%, with 15 non-conforming boards (MC > 10,4%). Based on this result the lot would have been considered non-conforming. 32

33 To produce right drying quality according to the new EN-standards As mentioned earlier, it is often more suitable for the dried sawn timber producer to do an on-line delivery control in the production. It becomes more and more common that the moisture content is measured by a capacitance meter at the sawmills. The accuracy of the measurement is not so high for each sawn timber piece but the accuracy of the mean value is the same as the calibration accuracy of the measuring instrument. It is important that the mean value is as accurate as possible in order to pass the requirement on the deviation of the moisture content. The base of new standards is that the bigger the deviation is from the target moisture content, the stricter requirement is put on the standard deviation of the moisture content. The sawmill has to do an extensive equalising of the sawn timber in the kiln if the average moisture content deviates too much from the target moisture content. As the dried sawn timber producers measure the average moisture content and the corresponding standard deviation of moisture content, the use of the statistical quality measuring system based on variables (individual measurements) instead of attributes (Yes/No) can be recommended. In Table 10 the requirements in the standards EN14298/ENV are transformed to values for a control scheme based on variables. The requirements in Table 10 are for standard drying. Table 11 shows a proposal for the requirements in case of special drying. If the average MC is lower than the target MC, a higher stdv can be accepted, if the average MC is higher than the target MC the allowable stdv must be reduced. Table 10. Standard drying. Use of the average value and standard deviation (stdv) of moisture content at delivery control instead of the acceptance inspection according to EN and ENV Demand according to EN Ordered MC (target MC) Allowable variation of average MC Lower limit for 93,5 % of the lot Upper limit for 93,5 % of the lot Examples of allowable stdv for MC Max. allowable stdv when target MC is hit exactly 7-1.0/ / / / / / / / / / / /

34 Note: It is common that sawmills in their process evaluate the moisture content variations in timber lots by calculating the standard deviations. Also at an acceptance control by the timber purchaser this method can be used. In case of dispute control according to the AQL-method in ENV is valid. If supplier s value for stdv (standard deviation) is below the value in Table 10 (AQL 6,5), the supplier has apparently fulfilled the MC requirement. Due to limitations in measuring high moisture contents during the drying process deviations between measured and actual results may occur in industrial practice. Here, the well developed calculation programs for kiln drying processes may offer some help. These programs can often estimate the average final moisture content better than the electrical moisture content meters in the kiln. Special drying, preferred target moisture content When selecting the drying quality for wood products like windows, staircases, flooring and special studs it is recommended to use the values in Table 11 which are a bit more strict than those in table 10. The bold target moisture contents 8, 12, and 15 % are preferred MC values. If an old product standard demands that target MC 9 % should be used, it is appropriate to use the preferred MC of 8 % instead as target moisture content. Following this scheme a three MC class system could develop which may help producers and Buyers to reduce the manifold of possible MC specifications to an acceptable limit. Timber with a target moisture content other than those preferred MCs marked in bold, it will be more expensive and more difficult to obtain. Table 11. Special drying. Use of the average value and standard deviation (stdv) of moisture content at delivery control instead of the acceptance inspection according to EN and ENV 12169, preferred MC values are marked in bold. Demand according to EN Ordered MC (target MC) Allowable variation for target MC Lower limit for 93,5 % of the lot Upper limit for 93,5 % of the lot Examples of allowable stdv for MC Max. allowable stdv when target MC is hit exactly 7-1.0/ / / / / / / / / Drying stresses according to ENV In the drying quality standard EN it is stated that requirements may also include the amount of residual drying stresses. How these requirements should be defined is given in the section Measuring drying stresses, ENV

35 Take care of the timber Timber that has been dried correctly in the sawmill deserves to be taken care of at the construction site and in industry. Proper storage is therefore important in order to prevent moist timber being built in, risking mould and rot and to avoid material waste and rejection in the joinery and furniture industries. Here is some short advice on how to avoid quality reduction during storage. Storage in industry Timber that has been dried at the sawmill is often delivered bulk stacked, i.e. the timber is packed without stickers between the different layers of timber. Packages are kept together by steel bands and may be covered by different materials. Wood in bulk stacks incorporates a large amount of water, which means that it takes a very long time for the timber inside the package to adapt to the surrounding climate. The timber in the outer parts of the package can absorb moisture if the wood moisture content is low and the plastic cover does not cover the whole package. Storage of timber packages, with low moisture content, should therefore be done in conditioned storerooms. A unheated storeroom with dehumidifier is suitable. The best guarantee for stable timber moisture content is to keep the package intact and not break it until shortly before use. Fig.23. A dehumidifier makes sure the air moisture content corresponds to the right equilibrium moisture content. The equilibrium moisture content should be the same as the preferred average wood moisture content and the finished products moisture content. It is important to have good control over the climate in the storeroom and in production premises. The conditions should be controlled. 35

36 The reason for this is that, if not controlled, it can be difficult to keep the timber dimension and shape, within the intended tolerances, during production and installation of the products. The surrounding conditions are even more critical if the products consist of different wood species or are combinations of solid wood, board products and veneer or metal parts in combination with wood. In addition the demand increases on a controlled climate through the whole production if materials and components are intermediately stored. Storage at construction Protection against ground moisture The packages should be placed on bearers. The support should allow use of forklift trucks and good ventilation. Floor battens of used railway sleepers are often used. If a timber package is covered up outside with a tarpaulin it must be ventilated, firstly under the package against the ground, but also preferably at the top of the package. Protection against precipitation The timber package should be protected against precipitation. This can be done in different ways: Under roof: In storage packages can, without cover, stand completely unprotected. If there are no walls, the package should have a simple paper- or plastic cover. Covered packages: On the market there are systems for covering packages that are so waterproof that they can remain outdoors without losing their protective effect, on condition that the cover is not damaged so that water is prevented from penetrating into the package. If a stored package is found to be leaking, the package should immediately be placed under roof, opened and stickered. 36

37 Protection against UV-radiation Packages with transparent cover should not be placed in the sun for longer periods. The sun dries the timber on the exposed side, leading to checks, and moistens the timber on the shady side. The risk is that after a while it will start to grow go mouldy. Timber that is left in the sun without cover starts to deteriorate on the surface, which will turn gray. Timber that is to be given a coating must not be exposed to UV-radiation. Only a few weeks exposure reduces the the adhesion of the coating. Protection against dirt The timber should be stored so that it does not get dirty from e.g. splash from roofs or neighbouring roads. 37