Technical Field

The invention relates to the measurement of moisture content of timber by thermalizing fast neutrons which are attenuated by the hydrogen content of the timber. Then, by means of calibration data, the moisture content of the timber can be calculated from the number of thermalized neutrons measured per unit time, and be used for adjusting the parameters of drying during the drying process.

Background Art

Until some 30 - 35 years ago drying of timber was carried out in the open air. Today nearly all timbering is dried in drying chambers or ducts. To establish an appropriate drying geometry in such a drying chamber, the boards are arranged in piles, the spacing between the layers of boards being approx. 22 mm. Wooden laths are used as spacers. In the space between the layers of boards warm air is circulated as a carrier of energy and water flown off. As for transport considerations the drying arrangement is formed of clusters of boards, each made up of 100 - 200 boards depending on the board dimen¬ sioning. In a drying chamber, there may be 16 clusters stacked to form four columns, for example. Large circulation fans cause the air to flow through the clusters of boards, and a set of heaters supplies the heat required. To achieve the desired climate during the drying process some of the air may be vented to the outside and be replaced by dryer outdoor air, or if it is too dry in the drying chamber vapour or water may be added by means of nozzles (scalding/wetting).

In fresh timber, water is present in not bound and bound states. The uncombined water is present in the capillary veins and is easy to fly off by evaporation. However, the combined water makes up a part of the cellular structure of the wood and must be removed by diffusion. When the uncombined water is removed a moisture content of the wood is achieved which is defined as the fiber saturation point of the wood. Any change in the moisture content in the range below fiber saturation causes a change in the volume of the timber: by drying it shrinks, and by absorbing water it swells. From the point of fiber saturation to the point of absolutely dry timber, the shrinking and swelling is approximately proportional to the change in the moisture content. Changes in the moisture content above fiber saturation causes no changes in the volume.

To maintain the initial quality of the timber when drying below the point of fiber satura¬ tion, the parameters of drying must be controlled such that the water transport from the interior of the timber towards the surface is balanced against the proper tension in the timber, such tension occuring as a result of the water transport. This can be achieved by knowing at all times the moisture content of the timber and adapting the parameters of the drying process correspondingly. In other words, the parameters of drying may be synchronized with the moisture content of the timber, with the desired drying rate and the intended final moisture content.

Today the control of timber drying usually is based on a drying scheme indicating nominal values of dry bulb and wet bulb temperatures throughout the drying process. Such a drying scheme is based on values of experience and all the time it tries to produce a drying climate causing minimum drying damage to the timber. If one does not know the moisture content of the timber it is not possible to achieve normal control of the drying climate, as such a control must be based on knowledge of the moisture level at all times.

Continuous measurement of the moisture content of the timber during the drying process gives the possibility of improving the control of the drying climate such that damages due to the drying are reduced. Then it is also possible to terminate the drying process when the timber has obtained the desired final moisture content, and thus predetermined specifications regarding the amount of moisture content can easily and reproducibly be met. In addition, there is no need to monitor the material, thereby saving both time and energy to dry. Other application areas also exist, in which it is usefull to measure the moisture content, such as by rough grading of materials and quality control.

Various methods of measuring the moisture content of timber are known. The most common are the method of drying and weighing, and one whereby handheld instruments are used to measure the resistance or capacitance of the timber. It is referred to publication 1 of the bibliography at the end of the description. Such methods cannot be utilized for measuring moisture content during the drying of timber.

To continuously measure the moisture content of the timber, the moisture level may be measured by means of electrodes which are pierced into the wood to measure electrical resistance. However, in this way, only a minute portion of the timber can be measured with respect of moisture content, and therefore it may be difficult to get an impression of

the mean moisture content of the drying chamber as a whole. Also, this measurement method is not suitable for measuring moisture contents above 25 - 30 %, and therefore, by using measuring equipment of this kind, it is not possible to control the course of the drying above the fiber saturation point. In practice, most moisture meters which are commercially available exhibit a highly limited accuracy due to the lack of proper temperature compensation, and/or because they are not able to measure the moisture content continuously in the timber kiln. In this kind of measurements, the measurement accuracy is very uncertain due to the complex interaction between resistance, tempera¬ ture and moisture (see publications 2, 3 and 4 of the bibliography).

The measurement of moisture content of timber by means of nuclear radiation is previously examined by Loos (see publication 5 of the bibliography), for example. At that time, one concluded that two different measurement methods had to be used to be able to determine the moisture content. A nuclear radiation technique has been employed also for measuring the moisture content of wooden chips but then such a measurement method had to be combined with another measurement means for the measurement of density (see publication 6 of the bibliography).

However, without knowing the density of the timber it is still possible according to the present invention to measure the moisture content when the measurement of the moisture content is carried out in such a manner that it can distinguish between hydro¬ gen bound to water and to wood, respectively. Measurements carried out at Institutt for Energiteknikk show that the nuclear moisture meter developed according to the invention is well suited for continuous measuring the moisture of timber without additional measurements, as it has appeared that the thermalization process of fast neutrons in wet timber is highly dependent on whether the hydrogen is bound to water or to wood.

Hence, an object of the invention is to provide a measuring device to measure the moisture content of timber, and a method of calibrating such a device in such a manner that the moisture content can be measured with a satisfactory accuracy independently of the density of the timber.

Another object of the invention is to specify a method of adjusted the drying of timber controlled by such a calibrated measuring device.

Disclosure of Invention

The invention relates to a method of calibrating a device for the measurement of moisture content of timber with respect of varying dry density of the different types of timber to be measured, whereby the timber is irradiated by fast neutrons from a neutron source and the number of resulting thermal neutrons which per unit time has been attenuated by hydrogen atoms of the moisture content of the timber is recorded by at least one detector as an expression of the moisture content of the timber and indicated on a measuring instrument, the method being characterized by the following step: a) for the timber to be measured recording graphs of the readings of the measuring instrument as a function of a known moisture content at maximum and minimum dry density, respectively, and determining a mean value graph between the maximum graph and the minimum graph, b) determining a first alteration in the instrument reading between the mean value graph and the graphs of maximum and minimum dry density, respectively, at a given constant moisture value for which the greatest difference in dry density exists between the graphs, c) determining a second alteration as the change in instrument reading per moisture content percentage measured along the mean value graph, and d) dividing said first alteration with the second alteration and using the deduced quotient as an expression of the measurement accuracy of the device within the defined measurement range between the graphs of maximum and minimum dry density, respectively.

The invention also relates to a device for measuring the moisture content of timber, comprising a neutron source for irradiating the timber by fast neutrons and at least one detector to count the number of resulting thermal neutrons which per unit time has been attenuated by hydrogen atoms of the water content of the timber, and a measuring instrument to indicate a metered value as a function of the counted number of neutrons as a measure of the moisture content of the timber. The device of the invention is characterized in that the neutron source and the detector are arranged side by side in a wall, floor or ceiling adjacent the timber, and that the device comprises a data register for the storage of parameters of drying and calibrated processing of neutron measure¬ ment data from the detector to be displayed on the measuring instrument.

Preferably, the measuring instrument comprises a computer screen for the display of the moisture content of the timber as a function of time.

Furthermore, the invention relates to a method of drying wet timber by means of a controlled forced air current and adjusted wet bulb and dry bulb temperatures of the air current by making use of the measuring device and calibration specified above.

Then, the characterizing features of the method lie the steps of: a) reading into a drying control module, the wood species and dimensions of the timber to be measured, as well as a recorded outside temperature, the initial mean moisture content of the timber and the desired final moisture content thereof, b) for the complete desired course of drying, by use of the control module, calculating empirically proper parameters of drying (that is, wet and dry bulb temperatures, and fan operating program) on the basis of the data stored in step a), c) under control of the control module, tracing the course of drying on a computer screen of the measuring device in the form of a graph having moisture contents indicated along the ordinate axis and drying period along the abscissa axis, and d) during the course of drying, measuring the moisture content of the timber by means of the measuring device and, by altering the parameters of drying, correcting a possible difference between a measured value of drying and the intended and traced course of drying.

Preferably, the moisture content of the timber is measured continuously by means of the measuring device and the average moisture content during a selected time period, such as the last three hours, is calculated and displayed on a computer screen as an actual control value, and if this actual value, at the time of display, differs from a corresponding nominal value on the graph of the drying course, a quick correction of the drying climate is made by altering the parameters of drying. Then the calculated moisture content can be displayed on the computer screen at fixed intervals, such as once every hour.

The metering principle being the basis of the invention is founded on the ability of the hydrogen to modify fast neutrons to slow, socalled thermalized neutrons. Therefore, this thermalization of neutrons can be utilized to measure the moisture content of a mixture of water and dry matter to record the relationship between the number of fast neutrons being transmitted into the medium to be measured and the number of thermal neutrons being reflected. The number of neutrons being thermalized and detectable, is deter¬ mined by, inter alia, the composition of materials, the dry density, water content, and the dimensions of the object being measured (measurement geometry).

When the dry matter of the object to be measured mainly consists of molecules having "heavy" atoms, in practice, only the hydrogen in the water contributes to the thermaliza¬ tion, and the number of thermal neutrons detected is then a measure of the moisture content of that measurement object. This allows the instrument to be calibrated on the basis of moisture samples taken from the medium to be measured.

If the dry matter of the object to be measured is an organic material, such as wood, the dry matter also contributes to some extent to the thermalization. Then the number of thermal neutrons which are detectable is a function not only of water density and geometry but also of the dry density. To permit one measurement only based on the thermalization of the fast neutrons while there are two uknown quantities, i.e. dry matter and water content, the contribution of the dry matter to the thermalization must be mapped by the calibration. The dry matter of wood contains about 6,2 % of hydrogen (H), 50 % of carbon (C) and about 43 % of oxygen (O), and the contents vary from one wood species to another. Additionally the wood species exhibit large differences in dry density.

The mass of a hydrogen atom and that of a neutron are approximately the same (1 amu). Hydrogen atoms and neutrons collide at a varying degree, all the way from sideslips with nearly no neutron energy loss, to head-on collisions whereby the neutron looses neariy all of its energy. On average, about 18 collisions with hydrogen atoms are required to slow down a fast neutron from a source to thermal energy, which means that it becomes thermally balanced with the medium being measured. In comparison, on average, 114 collisions are necessary to thermalize a neutron by carbon atoms, and correspondingly 150 collisions, if it is to be thermalized by oxygen atoms. Therefore, the sensitivity of the neutrons to hydrogen atoms is substantially higher than for other atoms in wood.

It is known that the effective cross-section of hydrogen in water is larger than that of hydrogen in paraffin. No known value of the effective cross-section of hydrogen in wood is available. It was, however, desirable to find out whether the latter was so different from the effective cross-section of hydrogen in water that it would be possible to measure to what extent the hydrogen atoms were bound in water or wood. If the effective cross-sections were substantially different, the number of thermalized neutrons would be variable, even though the number of hydrogen atoms is constant.

Brief Description of Drawings

Viewed against this background the invention is now explained in greater detail by describing the development work leading to the present invention, and by reference to the appended drawings, on which: Figure 1 is a schematic view of a drying chamber, figure 2 schematically shows a measuring device according to the invention for measuring moisture contents during the drying of timber in a chamber, figure 3 is a diagram showing graphs of instrument readings on a moisture meter for completely dry wood having varying dry density, and wood having constant dry density and varying moisture content, respectively, as a function of the density, figure 4 is a diagram indicating graphs for comparison of two different drying courses of one and the same wooden material but having different initial moisture content, figure 5 shows a calibration graph indicating instrument readings as a function of moisture content of 50 mm spruce, figure 6 is a diagram of a system arrangement utilizing nuclear moisture meters for the control of a timber kiln, figure 7 is a graphic diagram indicating instrument readings as a function of moisture content of timber having varying dry density, and which can be used to calculate the measurement accuracy, and figure 8 is a diagram of calibration graphs indicating instrument readings as a function of hydrogen density at low, mean and high density, respectively.

Best Mode for Carrying Out the Invention

Initial tests were made by means of a nuclear moisture meter developed by Noratom/- Norcontrol for measuring, inter alia, the water content of coke. Primarily the purpose was to find out whether this moisture meter could be used to determine the moisture content in clusters of boards. Firstly the apparatus was used in the usual way, the measurement geometry being indefinite and not indefinite, respectively. It appeared that existing apparatus could be used to determine the moisture of an indefinite measure¬ ment geometry, if the dry density was constant. Then the apparatus was used in a different manner, that is, with one neutron channel only. The results of these tests were so interesting that furthering the work was desirable.

Therefore, a nuclear moisture meter was developed which was adapted to be used when drying timber in a timber kiln. The metering equipment consists of the following main components:

- a container including a source 7, detector 4, reflector 3 and an elevation mechanism 5, 8, 10 for the neutron source, collectively denoted as a probe and shown in figure 2,

- a junction box having tube connections,

- an apparatus and control cabinet,

- a hydraulic cylinder and pressure regulator, and

- a calculation unit for reading out the moisture percentage of the timber.

The neutron source 7 employed in the measuring probe consists of an americium- beryllium source. Americium emits α-particles reacting with beryllium cores which then emit neutrons. The detector 4 consists of two ³ He proportionality counters. The reflector is a water container 3 covered by a cadmium plate 18.

The measurement probe was placed in a hole in the floor underneath a cluster of boards. In a laboratory at Institutt for Energiteknikk (IFE) calibration of the measuring device was carried out by means of materials having known moisture content and variable density. These tests confirm the assumption that the effective cross-section of hydrogen bound to water and to wood, respectively, mutually are so different that it is possible to measure the moisture content in timber by means of the developed measur¬ ing device. As indicated by the solid line in figure 3, as a result of the measurement of completely dry wood having variable density, a straight line was achieved having a rise gradient a for instrument readings indicating the number of thermal neutrons per unit time as a function of density. Measurements of materials having constant high density and variable moisture content produced a straight line having a rise gradient b in the same co-ordinate system, as shown by the broken line in figure 3. Rise gradient b is obviously several times greater than gradient a.

After the laboratory tests at IFE the apparatus was tested for one year in a timber kiln at Fossum Bruk of Røa, Oslo. From the measurements performed one can clearly conclude that the thermalization process is dependent on whether the hydrogen involved is bound to water or to wood.

The measuring device was developed further to include, inter alia, a counter circuit to be installed in a computer unit (PC) as a replacement of the calculation unit. Then the

device was installed in a chamber dryer at Frognerseteren Bruk, Oslo. Here the purpose of the tests was to produce reference values of sufficient quality to be able to translate with a sufficient degree of certainty, a pulse rate being read into a moisture percentage by means of an experimentally determined calibration curve. Furthermore, experiments were carried out, inter alia, to judge the reproducibility of the measured values and to check whether deviations were present in the measuring device.

Measurements made on one single pile of timber demonstrated acceptable reproduc¬ ibility for one and the same wood species but with a distinct difference between spruce and pine. The operation of the device was controlled by placing an aluminium container of a known volume of water above the measuring instrument. Initially it was found that the deviations of the apparatus were too large but this was repaired by changing the amplifier and high voltage source. Then no instrument deviation could be detected for the measuring device.

Then the measuring device was installed in a kiln at Ormstad Sag og Høvleri, and some further improvements were made. At Ormstad the measuring device has exhibited very good reproducibility of the measurement results. Most clearly this can be seen by comparing the fiber saturation point of different courses of drying. In figure 4 the moisture, dry bulb temperature and the temperature difference between the dry bulb and wet bulb temperatures are shown for two courses of drying A and B. Roughly, the temperature difference is the same in the two drying courses, and the difference in dry bulb temperature is relatively small. In drying courses A and B, the initial moisture content was 63 %, and 46 %, respectively. In the beginning the moisture content increased (heating phase) and the timber started to dry as the moisture content was reduced. When the temperature difference is constant the rate of drying will also be constant. In drying course A the temperature difference increased after 60 hours of drying, and this is also reflected by the drying rate which increases significantly. Even though the temperature difference is constant after drying for 120 hours, the drying rate, however, is reduced, which appears from a "break" in the moisture graph of drying course A. This "break" demonstrates that the fiber saturation point is reached and that now it is more difficult to force the water out, as this is more strongly bound to the cellular structure of the wood. In drying course B the moisture content of the timber passes through the fiber saturation point after a drying period of approx. 80 hours, and here also that temperature difference is constant. In both drying courses the "breaking point" occurs at the same moisture content in the timber, namely 27 %.

Based on the measurement data produced a calibration formula is computed, the associated calibration curve of which being indicated in figure 5. At present, tests are made to find calibration data for some more dimensions of both spruce and pine.

Some of the practical advantages of the measuring device are that it does not disturb the process and does not involve additional labour when loading the timber in and out of the kiln. The design and location of the measuring device render it physically protected. It is very important that a well defined measurement geometry is measured, and usually this is achieved by locating the measuring probe underneath the innermost cluster of boards.

In practice the measuring device comprises a detector unit which, during the drying process, reads the moisture content in one or two of said clusters of boards, a computer assembly (control module) having input means for loading data concerning wood species, wood dimensions, outside temperature and desired moisture content, and a PC for processing and displaying data, see figure 6. A plurality of detector units may be connected to the same PC, such as shown in the figure.

When the timber is stacked in a drying chamber the operator of the drying process gives a message to the control module through his computer keyboard concerning type of wood, board dimensioning and desired final moisture content. Simultaneously the mean moisture content of the timber and the outside temperature are measured. When this data is acquired and stored, the correct parameters of drying based on experience are calculated for the complete course of drying (that is, wet and dry bulb temperatures as well as a program of the operation of fans). Then the process of drying can be started.

The intended progress of drying is traced on the screen of the PC as a graph with moisture contents along the y-axis and drying period along the x-axis. This curve then defines the nominal value of the course of drying. To take control of the drying, the moisture content of the timber is measured continuously, and the average moisture content during the last three hours is displayed on the screen once an hour as the actual value. If it appears that there is a difference between the actual and nominal values one can quickly make corrections to the drying climate by changing the parameters of drying such that the intended optimum course of drying is maximally followed. This may be accomplished automatically or manually.

Here the measuring device comprises two detectors 4, one neutron source 7, one pulse amplifier, and a high voltage unit (not shown). Normally the device is installed in a 100 x 120 mm acid-proof channel underneath the inner column of boards transversely to the direction of the boards. The same channel is to serve as one or more supports of the column. A power supply and registers "R" which are updated with measurement data, are located in a box outside the drying chamber. The data is acquired by the PC unit every 60 seconds and stored in a data base. Once every hour the average of the recordings during the last three hours is calculated expressed in moisture percentage. The result is displayed on the computer screen and stored in a data base. The reason for such a long period of integration is that the neutron recording pulses from the detector which represents the moisture content, are statistically distributed. Therefore, a large number of pulses is required to obtain accurate measurements. In stead of having high source power, time is used. This is advantageous both for economical and technical radiation reasons.

The detector unit may also be located between two clusters of boards. This may especially be the case in relation with experimental work and research.

When calibrating the measuring device the number of thermal neutrons is recorded as a function of the total hydrogen density, using hydrogen density in water p _H . _{wa er} and hydrogen density in wood PH- _WOO d ^as variables. Since the neutron cross-sections σ _tota| of ρ _H . _water and of p _H . _wood are different, one obtains graphs as shown in figure 7. From these graphs the measuring device can be calibrated and the resolution of the apparatus be determined.

The dry density of timber varies naturally between p _d (max) and p _dry (min). In the calibration process, an average dry density ρ _d (min) is deduced. To determine the accuracy of the measuring device as viewed against these variations in dry density, the change in the number of thermal neutrons as a function of the maximum difference from p _d (mean) for a given moisture content (change 1) must be compared with the change in the number of thermalized neutrons per percentage change in moisture content by a given value of ρ _dry (mean), i.e. change 2. The measuring accuracy of the apparatus is then found as the difference, + or -, from the true moisture content, by dividing change 1 by change 2.

Consequently the measuring accuracy can be determined by comparing the alterations in the number of thermal neutrons as a function of:

1. The maximum change in dry density (at constant moisture content).

2. The change per percentage moisture (at average dry density).

The measurement accuracy, + or -, is then given by the following equation:

X1 (max/min density) - X2 (mean density) X3 (moisture B) - X4 (moisture B +/- 1 %)

where:

X1 = the number of thermalized neutrons when the wood has maximum or minimum dry density at moisture value A (as indicated in figure 7), X2 = the number of thermalized neutrons when the wood has average density and moisture A, X3 = the number of thermalized neutrons when the wood has moisture value B (as indicated in figure 7) and mean density, and X4 = the number of thermalized neutrons when the wood has moisture B +/- 1 % and mean density .

In figure 8 three calibration graphs a, b and c are shown for low, mean and high dry density p _d , respectively, the instrument reading of the number of thermal neutrons per unit time being indicated as a function of the total hydrogen density p _H . On each of these calibration graphs there are indicated three calibration points of different moisture values. Points marked with the same symbol represent the same moisture content. If a line rises through the same symbolic marks from the y-axis towards the righthand side of the illustration, there is an increasing under-compensation in the calibration for a decreasing dry density ρ _d . If the inclination of such a line is opposite, there is an increasing over-compensation for falling ρ _d . When the Iines through the marked points of the same moisture content are horizontal, the apparatus compensates completely the variations in dry density.

The measurement accuracy lies within the accuracy of the references used in the calibration process, and in practice it will be approx. ± 1 ,5.

Bibliography:

1. Esping, B., "Tratorking la - Grunder i torkning", Tratek. Gothenburg 1992, ISBN 91- 88170-06-3.

2. Forrer, J.B. and Vermaas, H.F., "Development of an improved moisture meter for wood", Forest Products Journal, 1987, Vol 37(2), pp 67 - 71.

3. Breiner, T.A. and Quarles, S.L., "Performance of an In-kiln Moisture Meter - Preliminary Results", Proceedings, Westem Dry Kiln Association 1990, pp 104 - 109.

4. Breiner, T.A., Arganbright, D.G. and Pong, W. Y., "Performance of In-line Moisture Meters", Forest Products Journal, 1987, Vol 37(4), pp 9 - 16.

5. Loos, W.E., "Determining Moisture Content of Wood by Nuclear Radiation Techni¬ ques", Forest Products Journal, 1965, Vol 15(3), p 102.

6. Stenstrόm, L, "Vi vill ha analoga signaler", Elteknik med aktuell elektronik 1980:9.