WO2008110017A1 - Systems and methods for monitoring wood product characteristics - Google Patents

Abstract

Characteristics of wood and similar materials such as the bulk average orientation of fibres in composite fibrous materials, moisture content, density, microfibril orientation, wood species, and strength, can be evaluated based on measurements made with terahertz (THz) electromagnetic signals.

Description

SYSTEMS AND METHODS FOR MONITORING WOOD PRODUCT CHARACTERISTICS

Reference to Related Applications [0001] This application claims the benefit of U.S. Provisional Application Serial No. 60/895,073 filed 15 March 2007 and entitled SYSTEMS AND METHODS FOR MONITORING WOOD PRODUCT CHARACTERISTICS, which is hereby incorporated by reference in its entirety.

Technical Field

[0002] The invention relates to the wood processing industry. The invention relates particularly to methods for measuring characteristics of products made of wood or similar materials. Aspects of the invention provide methods and apparatus for determining characteristics such as: fibre orientation, density, moisture content, wood species, whether or not wood is "compression wood", and microfibril orientation.

Background

[0003] Engineered wood products such as oriented strand board (OSB) and particle board are made by adhering together fragments of wood. The properties of the resulting materials depend upon a range of factors including:

• the way in which the fragments are aligned with one another; the orientations of fibres within the fragments of wood;

• the species of wood;

• the water content of the fragments of wood; • whether or not the wood is "compression wood" which has mechanical characteristics that are different from other wood; and,

• so on.

[0004] The strength and other mechanical properties of many composite fibrous materials are sub-optimal in cases where fibres in the materials are arranged so that the fibres are preferentially aligned in one direction. For example, in OSB, if adjacent layers of wood chips are not aligned with the grain extending perpendicularly to one another then the mechanical strength of the OSB will be reduced. Similarly, the orientation of fibres in paper can affect mechanical characteristics of the paper. [0005] Wood product industries use various tests to measure a range of wood characteristics. Some of these tests are performed in laboratories and require interpretation by a skilled technicians. Such tests are not useful for real-time control of industrial processes.

[0006] There are various ways to measure fibre orientations. Some of these are as follows:

• One can dissect samples of a material under a microscope. This is not a practical method for process control. • One can measure DC or low-frequency AC dielectric constants of the material.

The dielectric constant perpendicular to the wood grain is known to be different from the dielectric constant parallel to the wood grain. It is not practical to use measurements of dielectric constant for process control because measuring the dielectric constant of a material is typically too slow and also requires contact with the material in most cases.

• Optical methods can be used to determine fibre orientation but these methods can only determine the orientation of fibres at the surface of the material.

• X-ray analysis may be performed, however the equipment required for X-ray analysis is expensive and there are safety issues with X-rays. P.H. Friedlander 'The measurement of fibre orientation in newsprint with respect to the machine direction by X-ray diffraction" Pulp and Paper Magazine of Canada, June 1958, pp. 102-103 and H. Ruck et al., "The determination of the fibre orientation in paper" Pulp and Paper Magazine of Canada, June 1958, pp. 183- 190 disclose the use of X-ray diffraction to monitor fibre orientation in paper. • US 4,654,529 entitled "Method for measuring the fiber orientation anisotropy in a fibrous structure" discloses a method for evaluating the fiber alignment anisotropy in a fibrous structure by measuring attenuation as a function of polarization.

[0007] In commonly-owned PCT patent application No. PCT/CA2007/000062, the present inventors describe systems and methods that apply terahertz (THz) radiation for measuring fibre orientation in products made of fibrous materials such as wood. [0008] There is a need for technology that can be used to determine additional characteristics of wood, wood products and other fibrous materials. There is a particular need for such technology that can be applied in industrial settings.

Summary of the Invention

[0009] This invention provides methods and apparatus for characterizing wood, wood products, or similar materials. Some non-limiting examples of materials to which the techniques according to the invention may be applied are OSB, particle board, plywood, wood chips, paper, cardboard, fibreglass, carbon fibre composites and solid wood.

[0010] Embodiments of the invention measure a range of qualities of materials such as wood or wood products by measuring how samples of the materials interact with THz radiation. Systems according to some embodiments of the invention can measure one or more of:

• average fibre orientation;

• moisture content;

• wood species;

• microfibril orientation; • density;

• whether or not wood is "compression wood";

• strength (modulus of elasticity); and,

• strength of composite materials.

[0011] In some embodiments, such systems are stand-alone systems. In other embodiments, such systems are integrated with a manufacturing process and provide one or more of:

• measurement of product characteristics and/or raw material characteristics for process control; • material rejection, material sorting, material grading, and/or quality control;

• imaging of materials; - A -

• reports regarding material characteristics and/or statistical trends in material characteristics.

[0012] Various aspects of the invention and features of specific embodiments of the invention are described below.

Brief Description of the Drawings [0013] In the accompanying drawings:

Figure 1 is a block diagram of an apparatus for measuring qualities such as bulk average fibre orientation in a sample of a material;

Figure 2 shows plots of received signal as a function of time for various angles between the polarization axis of a THz signal and a wood sample;

Figure 3 shows results of measurements of index of refraction, n_m and absorption coefficient, α_w, as a function of frequency for a sample of spruce wood; Figure 4 illustrates variations in the time taken for a THz signal to propagate through a sample as a function of angle, where the sample is made up of 27 sheets of lens paper stacked so that the directions in which fibres in each sheet are preferentially oriented are aligned;

Figure 5 is a set of plots illustrating the expected variation in the time taken for a THz signal to propagate through samples in which fibres are oriented with differing degrees of anisotropy;

Figures 6, 6A, 7, and 7A are schematic views showing alternative apparatus for evaluating the degree to which the transmission time of a polarized signal varies with the angle of the polarization axis of the signal relative to a sample; Figure 8 shows the time evolution of a THz waveform and indicates a possible operating point for the apparatus of Figure 7;

Figure 9 shows schematically a plant for making a composite fibrous material that includes apparatus for evaluating the anisotropy of fibres in the fibrous material;

Figures 1OA and 1OB are schematic views of example mechanisms for measuring characteristics of the distributions of fibres at points within a moving web of material; Figure 11 is a flow chart illustrating a method according to one example embodiment of the invention;

Figures 12A and 12B are plots of sample calibration curves for the attenuation of THz radiation (field amplitude) transmitted through a sample of wood as a function of moisture content;

Figures 13A and 13B are plots of sample calibration curves for the time delay of polarized THz radiation transmitted through a sample of wood as a function of moisture content;

Figure 14 is a plot of a sample calibration curve for the birefringence of THz radiation transmitted through samples of wood, as a function of moisture content;

Figures 15A and 15B are plots of absorption coefficients as a function of frequency for THz radiation transmitted at the two orthogonal polarization states through a sample of wood at different levels of moisture content;

Figure 15C is a plot of a sample calibration curve for the differences in the average absorption coefficient between 0.2 and 0.4 Hz (measured between the two orthogonal polarization states) as a function of moisture content;

Figure 15D is a plot of a sample calibration curve for the differences in the slope of the absorption curves between the two orthogonal polarization states, as a function of moisture content; Figure 16 A, 16B and 16C are density profile transmission images generated by detecting the intensity of THz radiation transmitted through sheets of material;

Figures 17A and 17B are density profile transmission images generated by detecting the time of delay of THz radiation transmitted through sheets of material;

Figure 18 is a schematic view of a system for determining and monitoring various characteristics of wood, wood products and other fibrous materials in accordance with an embodiment of the invention;

Figures 19A, 19B and 19C are sample calibration curves for the average index of refraction, average absorption coefficient and time delay, respectively, of THz radiation polarized parallel to the machine direction transmitted through a composite wood product (OSB), as a function of density; Figure 20 is a sample curve showing a correlation of physical density with breaking load, as determined from samples of a composite wood product (OSB) tested on an industrial quality control unit for measuring breaking strength, MOE and MOR; Figures 21 A, 21B and 21 C are sample calibration curves for the average index of refraction, average absorption coefficient, and average time delay, respectively, of polarized THz radiation transmitted through a composite wood product (OSB), as a function of breaking load (where the average index of refraction is measured prior to physically breaking the samples);

Figures 22A(i) through 22A(x) (collectively referred to as Figure 22A) illustrate the profiles of THz waveforms, with corresponding reference scans in the absence of a sample, for 10 different species of wood and using two different polarization states (parallel and perpendicular to the visible grain);

Figures 22B(i) through 22B(x) (collectively referred to as Figure 22B) and Figures 22C(i) through 22C(x) (collectively referred to as Figure 22C) respectively illustrate graphs of absorption coefficients and index of refraction for THz radiation (transmitted at the two orthogonal polarization states) for 10 different species of wood;

Figure 23 A is an optical picture of an OSB sheet;

Figure 23B is a density profile transmission image generated by detecting the field amplitude of transmitted THz radiation, polarized parallel to the machine direction, transmitted through the OSB sample shown in Figure 23 A; and

Figure 23 C is a density profile transmission image generated by detecting the field amplitude of transmitted THz radiation, polarized perpendicular to the machine direction, transmitted through the OSB sample shown in Figure 23 A.

Description

[0014] Throughout the following description, specific details are set forth in order to provide a more thorough understanding of the invention. However, the invention may be practiced without these particulars. In other instances, well known elements have not been shown or described in detail to avoid unnecessarily obscuring the invention. Accordingly, the specification and drawings are to be regarded in an illustrative, rather than a restrictive, sense. [0015] Figure 1 shows a stand-alone apparatus 10 for determining qualities such as bulk average fibre orientation in a sample of material. Apparatus 10 may also be used to determine other qualities of the sample such as moisture content, wood density, microfibril angle, and the like, as will be described herein. In this example, the sample of material is a piece of wood. Apparatus 10 exploits the fact that wood is essentially transparent to electromagnetic radiation in a broad frequency range extending up to the terahertz (THz) region of the spectrum and that wood is birefringent in this frequency range (i.e. the propagation speed of a polarized THz signal through wood depends upon the direction of polarization of the signal relative to the direction of the grain of the wood through which the signal is propagating). The term "THz radiation" as used herein means radiation having frequencies in the range of 0.01 to 3 THz. Frequencies in the range of 0.2 to 1.4 THz are particularly useful for studying fibre orientation in wood products.

[0016] Apparatus 10 comprises a source 12 of polarized THz radiation on one side of a sample S and an analyser 14 on an opposite side of sample S. Polarized THz radiation from source 12 passes through sample S and is detected at analyser 14. Analyzer 14 detects directly or indirectly the average speed of propagation of the THz radiation as it passes through the sample (or, equivalently, the degree to which the sample delays the THz radiation).

[0017] In the illustrated embodiment, source 12 comprises an emitter 16 and a polarizer 18. Emitter 16 emits very short pulses of radiation (for example, pulses on the order of 0.5 ps in duration). Such pulses have a bandwidth on the order of 1 THz.

[0018] Analyser 14 comprises a polarizer 19 and a detector 20. Polarizers 18 and 19 can be set to a desired angle with respect to sample S. When source 12 emits a THz pulse, the pulse is detected by detector 20. The time elapsed between the emission and detection of the pulse can be determined. This can be repeated at different angular settings of polarizers 18 and 19 relative to sample S. [0019] Apparatus 10 may be used to evaluate the degree to which a sample S consisting of a solid piece of wood exhibits birefringence. Apparatus 10 can also be applied to evaluate the degree to which the fibres are aligned in a preferred direction in a sample of a material made of multiple smaller pieces or wood or other fibrous birefringent material.

[0020] Apparatus 10 may be used by first transmitting a reference pulse in the absence of sample S. The reference pulse provides an indication of incident power. The transmission spectra of the sample S can be determined by computing the ratio of transmitted power to incident power.

[0021] If the ratio of the product of the thickness of sample S and the birefringence of sample S to the speed of light in vacuum is at least as great as the duration of the pulse emitted by source 12 then the birefringence of a sample S can readily be measured directly in the time domain. Figure 2 shows plots of received signal as a function of time for various angles of polarizers 18 and 19 relative to the grain of a sample of poplar 6.54 mm thick. In Figure 2, zero degrees corresponds to alignment of the axis of polarization with the grain of the sample and ninety degrees corresponds to the axis of polarization being perpendicular to the grain of the sample.

[0022] It can be seen in Figure 2 that, at some angles, two time-separated and orthogonally polarized pulses can be seen in the detected signal. For example, with the polarizers 18 and 19 set at an angle of 22 degrees to the grain of sample S, a first pulse 22 A is detected at a time delay of approximately 15 ps and a second pulse 22B is detected at a time delay of approximately 17 ¹A ps.

[0023] An estimate of the frequency-averaged birefringence Jn of sample S can be obtained directly from this time domain data. IfJr is the difference in time between pulses 22A and 22B then An is given by: Δ« = — - (1)

where L is the thickness of sample S.

[0024] Another way to obtain a measure of the birefringence of sample S is to obtain a measure of the index of refraction of sample S. Where detection is coherent, the frequency-resolved, complex, index of refraction can be determined directly from the measured data. The ratio of the strength of the transmitted signal to the incident signal as a function of frequency can be approximated for a thick sample by:

Z7 ^ . Λ hSAMPlA^V) _ 7 f _ —⁽"^»-"^■<

AW¹WA" = Re lθ

(2)

^E REh- [Y)

where:

E _SAMPLE v ^) *^{s me} complex-valued field strength of the component of the transmitted signal at the frequency v, obtained through Fourier transformation of the time dependent electric field;

E_REF(v) is the complex-valued field strength of the component of the incident signal

at the frequency v, obtained through Fourier transformation of the time dependent electric field; c is the speed of light in vacuum; L is the thickness of the sample; t _AW and t_WA are respectively the Fresnel transmission coefficients of an air-to-sample and sample-to-air interface; and, n_w and n_A are the complex indices of refraction in the sample and in air respectively

and ή_A = 1 .

[0025] Since: K = ⁿw - ^ikw (3) where k is related to the absorption coefficient and represents the imaginary component of the complex index of refraction, it can be shown that:

n_w = - + 1 (4)

2/rvL

and also that:

where α_w is the absorption coefficient of the sample. Equations (4) and (5) are based on the assumption that k<n. For more general values of A, the values of n_w and a_^ can be evaluated but typically this requires numerically solving two coupled nonlinear equations in order to extract n_w and a_w.

[0026] Figure 3 shows measurements of n _w and α_was a function of frequency for a sample of spruce wood having a thickness of 3.025 mm. The values for n_wand α_w presented in solid lines are based upon a reference signal received with no wood sample present and signals transmitted through the wood with the polarization parallel to the wood grain. The values for n_w and α_w presented in dashed lines are based upon the reference signal and signals transmitted through the wood with the polarization perpendicular to the wood grain. It can be seen that both n_w and α_w differ significantly for signals polarized parallel to the grain and for signals polarized perpendicular to the grain of the wood sample.

[0027] Since n _w varies with the angle between the direction of the grain of a sample and the direction of polarization of the signal, and n_w is related to the speed at which the signal propagates through the sample, one can obtain a measure of the degree to which fibres in the sample extend preferentially in some direction (i.e. the degree of anisotropy of the arrangement of fibres in the sample) by monitoring the time taken for signals to propagate through the sample as the angle of polarization of the signal relative to the sample is varied. [0028] Figure 4 illustrates variations in the time taken for a THz signal to propagate through a sample made up of 27 sheets of lens paper stacked so that the directions in which fibres are preferentially oriented in each sheet are aligned. The signal was polarized in the horizontal direction (i.e. at an angle of 90 degrees) while the sample was rotated. It can be seen that there is a significant variation in the signal as a function of angle. It can also be seen that all of the information in the plot of Figure 4 can be obtained by scanning the angle through 90 degrees.

[0029] In the examples above, the samples for which data has been presented have been samples of solid wood having the grain running in a direction perpendicular to the direction of propagation of the signal. Consider the case where the signal passes through different regions in which the grain runs in different directions. In such cases the signal will propagate through the sample in a time that depends upon the thicknesses of the different regions and on the orientations of the grain in the different regions relative to a direction of polarization of the signal. If there are a lot of different regions and the grain directions of the different regions are random then there will be no preferred grain direction and the time taken for the signals to propagate through the sample will not vary significantly as the direction of polarization of the signals is rotated relative to the sample. On the other hand, if there is one preferred grain direction then the signal propagation time will vary in the general manner shown in curve 24A of Figure 5. If there are two or more preferred grain directions (as, for example, in the case where, in some regions, the grain is oriented in a first direction and in other regions the grain is oriented in a second direction perpendicular to the first direction) but, on average, the fibres are still oriented in a preferred direction then one would expect the signal propagation time to vary by a smaller amount, as indicated by curves 24B, 24C and 24D of Figure 5. If, in the volume through which the radiation passes, the same number of fibres are oriented in each of the two directions then the signal propagation time will not vary significantly with angle, as indicated by line 24E. [0030] One can obtain an indication of the degree to which fibres in the sample are aligned preferentially in one or more directions by obtaining a measure of how the signal propagation time varies for different angles between a direction of polarization of the signal and a sample. The speed of propagation may be measured directly or indirectly.

[0031] One way to make such measurements is through use of a THz detector capable of making time-resolved measurements of the time taken for a THz signal to pass through a sample.

[0032] Various other apparatus may be used to evaluate the degree to which the transmission time of a polarized signal varies with angle. Figure 6 shows one example apparatus 30. Apparatus 30 has a source 32 of polarized far-infrared radiation 33. In the illustrated example, source 32 comprises a continuous wave (cw) far-infrared laser 34 and a polarizer 36. Polarizer 36 may comprise a wire-grid polarizer, for example. Radiation 33 from source 32 passes through sample S to a detection system 38 comprising a quarter-wave plate 40, a polarizing beam splitter 42 and a pair of detectors 44A and 44B. Quarter-wave plate 40 is adjusted so that, in the absence of sample S, the signals detected at detectors 44A and 44B are equal.

[0033] If sample S exhibits birefringence, the polarization state of radiation 33 changes as radiation 33 passes through sample S. This change results in the signals at detectors 44A and 44B being different. The angle between the direction of polarization of incident radiation 33 and the grain orientation in sample S can be determined, for example, from the relationship:

= sin(2<9)sin[r] (6)

where:

S_44A and S_44B are the signals output by detectors 44A and 44B respectively; θ is the angle between the signal polarization direction and the grain of sample S; and, r is the phase retardation introduced by sample S.

F depends both on the thickness of sample S and the magnitude of the birefringence of sample S at the wavelength of radiation 33.

[0034] Apparatus 30 may be used to determine the residual volume-averaged fibre orientation within a composite sample. Since Equation (6) has two unknowns (θ and F) it is necessary to obtain values of S_44A and S_44B for each of at least two different polarization states of the input signal to solve for both θ and F.

[0035] Consider, for example, the case where a first THz input signal is linearly polarized and has an axis of polarization that makes an unknown angle θ with a preferred axis of sample S. A second THz input signal is also linearly polarized but has an axis of polarization that makes an angle θ+π/4 with the preferred axis of sample S. If the value of the left-hand side of Equation (6) is designated as a for the first THz input signal and as b for the second THz input signal then it can be shown that:

and also that:

or, equivalently,

[0036] Apparatus 30 of Figure 6 may be used to obtain values for a and b. During the measurement of a, polarizer 36 and polarizing beam splitter 42 are oriented appropriately for the first THz input signal. During the measurement of b, polarizer 36 and polarizing beam splitter 42 are oriented appropriately for the second THz input signal. This can be achieved, for example, by rotating apparatus 30 relative to sample S about the optical axis of radiation 33 between measurements of the first and second signals.

[0037] As noted above, a and b may be measured essentially simultaneously. Figure 6A shows an apparatus 3OA that can be used for this purpose. Apparatus 3OA has a source 32A of polarized THz radiation 33A. Source 32 A may be controlled to selectively generate THz radiation in one of at least two polarization states. For example, source 32 A may generate THz radiation in one of a first polarization state Ll and a second polarization state L2. The first and second polarization states may be linear polarization states although, in general this is not necessary. One or both of the first and second polarization states could be a circular, elliptical, or other distinct polarization state. Ll and L2 are not necessarily orthogonal to one another. For example, where Ll and L2 are linear polarization states, the polarization vectors of Ll and L2 may be rotated through an angle in the range of 30 degrees to 60 degrees(e.g. approximately 45 degrees) relative to one another.

[0038] The THz radiation passes through sample S and is then separated into two parts, preferably equal parts. In apparatus 3OA this is achieved by passing the radiation through a non-polarizing 50/50 beam splitter 35. Each part of the radiation is delivered to a detection system 38. Detector systems 38-1 and 38-2 are substantially similar to one another. A first detection system 38-1 comprises a quarter-wave plate 40-1, a polarizing beam splitter 42-1 and a pair of detectors 44A-1 and 44B-1. Polarizing beam splitter 42-1 is oriented to pass radiation of the first polarization state. A second detection system 38-2 comprises a quarter-wave plate 40-2, a polarizing beam splitter 42-2 and a pair of detectors 44A-2 and 44B-2. Polarizing beam splitter 42-2 is oriented to pass radiation of the second polarization state. Quarter-wave plate 40-1 is adjusted so that, in the absence of sample S the signals detected at detectors 44A-1 and 44B-1 are equal for the first polarization state. Quarter-wave plate 40-2 is adjusted so that, in the absence of sample S the signals detected at detectors 44A-2 and 44B-2 are equal for the second polarization state.

[0039] Apparatus 3OA can be operated by causing radiation source 32A to issue radiation in the first polarization state during a first time interval tl. In interval tl the outputs of detectors 44A-1 and 44B-1 are captured. A value for a can be determined from the captured values as described above. Radiation source 32B is then caused to issue radiation in the second polarization state during a second time interval t2. In interval t2 the outputs of detectors 44A-2 and 44B-2 are captured. A value for b can be determined from the captured values as described above.

[0040] One way to exploit the invention is to pass THz radiation of two polarization states through sample S, essentially simultaneously (i.e. within a time period short enough that the sample does not move significantly between passing the radiation of the different polarization states through the sample). After transmission, for each of the input polarization states, two orthogonal polarization states may be measured differentially. This may be done, for example, as outlined above in the discussion of Figure 6, with polarizing beam splitter 42 oriented differently for each of the input polarizations. The information so obtained may be used to extract information about both the preferred fibre-orientation direction and the magnitude of the birefringence of the sample.

[0041] To illustrate this, consider the case where the two input polarizations are linear and are not perpendicular to one another. For simplicity, consider the case where:

• the first input polarization makes an arbitrary angle, θ, with a preferred axis along which fibres in the sample are preferentially oriented; and,

• the second input polarization makes an angle (θ+π/4) with the preferred axis. [0042] Consider the case where the differential output of apparatus 30 (i.e. the left- hand side of Equation (6) above) has a value a for the first input polarization and has a value b for the second input polarization. It can be shown that:

and also that the birefringence, F, is given by:

and also by:

[0043] In an industrial setting where it is desired to make measurements quickly, it may be desirable to make measurements essentially simultaneously for both input polarizations. This may be done in a number of ways including through use of an apparatus 3OA as shown in Figure 6A. Controller 37 causes source 32A to produce THz radiation having a first polarization state Ll for a first time period tl and a second polarization state L2 for a second time period t2. Ll and L2 are not orthogonal to one another. The radiation is incident on and passes through sample S.

[0044] During time interval tl the output of detectors 44A-1 and 44B-1 of detector assembly 38-1 are monitored and used to provide a value for the differential output a. During time interval t2 the output of detectors 44A-2 and 44B-2 of detector assembly 38-2 are monitored and used to provide a value for the differential output b. These differential output values can then be used to determine θ and F through the use of equation (7) and one of equations (8) and (9), for example. [0045] Figure 7 shows another apparatus 50 that can be used to directly determine the variation in signal transit time through a sample with angle. Apparatus 50 comprises a source 42 of optical radiation. Source 42 may comprise, for example, a Ti-sapphire laser. A beam 53 of radiation is split by a beam-splitter 54 into a pump beam 53A and a probe beam 53B. Pump beam 53A is focused onto a semi-large aperture photo- conductive switch 55 to generate THz radiation in the manner described, for example, in G. Zhao et al. Design and performance of THz emission and detection setup based on a semi-insulating GaAs emitter, Rev. Sci. Inst. v. 73, pp. 1715-1719, 2002, or Zhang et al. US 5,420,595 entitled Microwave Radiation Source.

[0046] In a prototype embodiment of the invention, source 42 comprises a Ti:Sapphire oscillator providing 300 mW of output power in 100 fs pulses delivered at 80 MHz. In the prototype, switch 55 comprised a Si-GaAs substrate having electrodes formed with silver and having a minimum separation of 470 μm. Switch 55 was biased with a 250V peak 40 kHz sine wave.

[0047] In the apparatus of Figure 7, pump beam 53A passes through a half-wave plate 58 and a polarizer 59 before it is focused on switch 55 by lens 60. THz radiation 56 is generated at switch 55. Parabolic mirrors 62 A and 62B direct THz radiation 56 from switch 55 onto an intermediate focal plane 64 on which a sample S can be located. Parabolic mirrors 62C and 62D direct radiation 56 arriving from focal plane 64 onto a detector 66.

[0048] Probe beam 53B passes through an optical delay line 70 and is focused onto detector 66 by a lens 72.

[0049] In the prototype embodiment, detector 66 comprises a ZnTe crystal 68 that detects incident radiation by way of the linear electro-optic effect. The THz radiation from switch 55 interacts with probe beam 53B at crystal 68. Probe beam 53B initially has a known polarization, for example, it may be linearly horizontally polarized. The electric fields of the THz radiation alter the polarization of the probe beam. After interacting with the THz radiation at crystal 68, radiation from the probe beam passes through a quarter-wave plate 74 to a Wollaston prism 76.

[0050] Wollaston prism 76 splits the light from probe beam 53B into beams 53B- 1 and 53B-2 that have mutually perpendicular polarization vectors. Beams 53B-1 and 53B-2 are respectively detected by light detectors 78-1 and 78-2. Light detectors 78-1 and 78-2 may be matched photodiodes, for example. The outputs from light detectors 78-1 and 78-2 are provided as inputs to a lock-in amplifier 80.

[0051] When THz radiation 56 changes the polarization of probe beam 53B, the balance between beams 53B-1 and 53B-2 is altered and the output of lock-in amplifier 80 changes.

[0052] Since the duration of each pulse of light in probe beam 53B is much shorter than the period of THz radiation 56, the way in which THz radiation 56 varies in time can be determined by sweeping optical delay line 70 to provide delays of different lengths. Optical delay line 70 may be computer-controlled to facilitate this. The output of lock-in amplifier 80 as a function of the setting of optical delay-line 70 provides a trace that indicates the time-variation of THz radiation 56.

[0053] As noted above, the degree of anisotropy of fibres in a sample can be determined by monitoring the degree to which the sample delays the propagation of THz signal 56 for different relative angles between the polarization axis of signal 56 and the sample. The variation in this time delay can be conveniently measured in relation to a portion of the waveform of THz signal 56 that is rapidly varying. For example, where signal 56 exhibits a time variation as shown in the waveform 82 of Figure 8, an operating point 84 may be selected on steeply-varying portion 86. Operating point 84 corresponds to a selected output level of lock-in amplifier 80. [0054] As the sample is rotated relative to the polarization axis of signal 56, it is not necessary to scan optical delay line 70 to obtain a full time-domain waveform. It is only necessary to identify the delay that corresponds to the point on the waveform where the output signal from lock-in amplifier is at the operating point.

[0055] The way in which the delay at the operating point varies as the angle of the sample is changed provides information regarding anisotropies in the arrangements of fibres within the sample. The delay can be measured for several different angles. Preferably, the delay is determined for a selection of angles that spans approximately 90 degrees of rotation. In some embodiments, the selection of angles includes, for example, 4 to 15 different angles that span approximately 90 degrees.

[0056] The rate at which delays can be determined using the apparatus of Figure 7 is limited by the time constant of the lock-in amplifier. In a prototype apparatus a lock-in time constant of 100 μs was used. This time constant permits an angular range of 90 degrees to be scanned at a sufficient number of angles to obtain a good indication of the degree of anisotropy of the fibres within a sample in significantly less than one second.

[0057] Figure 7A is a schematic view of an alternative apparatus 88 useful for determining the degree of anisotropy of the fibres within a sample by making direct measurements of variations in the propagation times of THz radiation through the sample. Apparatus 88 has a source 90 of THz radiation that is directed to pass through a sample S. Source 90 can be controlled to selectively generate THz radiation of any of four polarization different states. The four polarization states may be linear polarization states although this is not mandatory. It is convenient for the four polarization states to include first and second pairs of orthogonal polarization states. For example, in some embodiments, the four polarization states include first and second pairs of orthogonal linear polarization states. In the example described in detail herein the four polarization states are linear polarization states identified as Ll, L2, L3 and L4 with Ll horizontally polarized, L2 vertically polarized, L3 polarized at 45° above the horizontal and L4 polarized at 45° below the horizontal. This is a convenient choice for the four polarization states.

[0058] Apparatus 88 has a mechanism 92 that divides radiation that passes through sample S into 4 equal parts. In the illustrated embodiment, mechanism 92 comprises three non-polarizing beam splitters 93A, 93B, and 93C. Each of the four parts of radiation is detected at a coherent THz detector 94 (the detectors are identified individually as 94A, 94B, 94C and 94D). Associated with each detector 94 is a mechanism that permits determination of a time delay caused by the presence of sample S. This mechanism may be, for example, the mechanism illustrated in detail in Figure 7.

[0059] Source 90 is controlled to generate THz radiation of each of the plurality of polarization states during a corresponding time interval with Ll being generated in a time interval tl, L2 being generated in a time interval t2, L3 being generated in a time interval t3 and L4 being generated in a time interval t4. Detectors 94 are gated so that detector 94 A is active during tl, detector 94B is active during t2, detector 94C is active during t3, and detector 94D is active during t4. For each detector 94, an operating point is chosen on a portion of the corresponding THz waveform that is rapidly varying in the absence of a sample. Any deviations in time delay caused by the presence of sample S are manifested as changes in the amplitude of the signal detected at the detector 94 as described in the discussion of Figure 7. Detectors 94 are zeroed in the absence of sample S (equivalently the outputs of detectors 94 are measured in the absence of sample S and the measured values are saved for use in determining the effect of sample S on the output signal).

[0060] Apparatus 88 may be operated to obtain information about the bulk average angle of fibre orientation in sample S and the degree of anisotropy in the arrangement of fibres within sample S by obtaining measurements 95 (identified individually as 95A, 95B, 95C and 95D) corresponding to each of detectors 94 during the time interval when the corresponding detector is active. The information may be obtained by processing measurements 95 in a processor 96.

[0061] A first value ml may be obtained by subtracting measurements 95 A and 95B for the first pair of orthogonal polarization states. A second value m2 may be obtained by subtracting measurements 95C and 95D for the second pair of orthogonal polarization states. Values indicating the degree of anisotropy of the fibres within sample S and the angle made by the preferred fibre orientation (if there is one) can be obtained from ml and m2.

[0062] It can be seen that the absolute time delay Δτ caused by sample S on a THz pulse is given by:

LL_

A τ = -In₁ cos(#) + m sin(#)l (13) c

where: L is the thickness of sample S, c is the speed of light in a vacuum, θ is the angle between the preferred axis of sample S and the angle of polarization of the THz radiation and n_± and /i₍ are the indices of refraction for the birefringent sample S. It can be seen that: ml cc Δrø(sin θ + cos θ) (14)

where An-K₁-H₁.

[0063] For the particular example being described herein, where the polarization states are linear polarizations and the second pair of orthogonal polarization states are rotated by 45 degrees relative to the first pair of polarization states, it can be shown that:

(15)

and

where: β is a constant calibration factor.

[0064] Any of the embodiments described above may be used to monitor the fibre arrangements during the manufacture of products such as OSB, particle board, papers or the like. Figure 9 shows schematically an example application. In Figure 9, a moving web 102 of a material 104 emerges from a manufacturing line 100. Material 104 is a fibrous composite material (meaning that material 104 is made by bonding together pieces of one or more starting materials and that material 104 includes fibres of some kind). Material 104 may be, for example, a composite wood product, for example, particle board, press board or OSB, a cardboard, or paper.

[0065] A source 110 of polarized THz radiation is located on a first side of web 102 and a detector 112 of the polarized THz radiation is located on a second side of web 102. The THz radiation is polarized along a polarization axis that is substantially in the plane of web 102. Source 110 and detector 112 are arranged so that the angle made by the polarization axis of the THz radiation to the direction of motion of web 102 can be varied through a range large enough to provide information regarding the degree of anisotropy in the arrangement of fibres within web 102.

[0066] Source 110 and detector 112 operate to obtain a measure of how the propagation speed of THz signals through web 102 varies with angle. Source 110 and detector 112 could, for example, be the source and detector in a system as shown in any one of Figures 1 , 6, 6 A, 7 or 7 A. Apparatus for quality control in the manufacture of a composite fibrous material may comprise a controller connected to control the angle of polarization of the radiation emitted by the source.

[0067] The controller may comprise a data processor connected to receive output signals from detector 112 and to control the operation of source 110 and detector 112. For example, the controller may comprise a suitable industrial programmable controller, a programmed personal computer equipped with interfaces that allow it to read data from detector 112 and to write control signals to source 110 and detector 112 or the like. The controller is configured to operate the detector to obtain a measure of the degree of birefringence of the sample at THz wavelengths. This may be done directly by measuring a propagation speed of THz radiation in the material for each of a plurality of angles or indirectly, for example, by monitoring changes caused by the sample in the polarization states of two or more THz signals.

[0068] The controller may process the data from detector 110 to obtain information characterizing a distribution of fibres in the sample. A controller 114 is indicated schematically in Figure 9. Similar controllers may be present in the apparatus shown in Figures 1, 6, 6 A, 7 and 7 A but are not shown in those Figures.

[0069] In some embodiments controller 114 may automatically control an orientation of fibres in one or more layers of the composite material and/or control a thickness of a layer of a component of the composite material in response to the processed information. The control may be maintained by delivering control signals 115 to manufacturing line 100. For those composite fibrous products in which it is desirable to have fibres distributed equally in orthogonal directions or for composite fibrous products in which it is desirable to have fibres randomly arranged with no bulk average preferred fibre orientation the controller may monitor deviations from zero birefringence and may optionally automatically control one or more process conditions to maintain the measured birefringence of the composite product to be zero or near to zero.

[0070] In some alternative embodiments, a THz radiation source and detector are mounted on an assembly that moves at, or approximately at, the speed of web 102 while measurements are being made so that variations in the propagation speed of the THz signals (or, equivalently, variations in the degree to which passage through the sample delays the THz signals) can be made at the same location on web 102 for a range of different polarization states while web 102 is moving. Figure 1OA shows an example embodiment wherein a source 110 and detector 112 are each mounted on a reciprocating shuttle 119.

[0071] In some other alternative embodiments, two or more pairs 120 of source 110 and detector 112 are arranged at locations that are aligned with one another and spaced apart in a direction of motion of web 102. Different pairs 120 of source 110 and detector 112 are controlled to take measurements at times that coincide with a sample location 122 on web 102 being present adjacent the pair 120. Successive source-detector pairs 120 can measure propagation speed of THz radiation through the sample at the same sample location 122 for different polarization angles relative to web 102. An example of such an embodiment is shown schematically in Figure 1OB.

[0072] As described above, the arrangements of fibres in a sample of a composite product can be studied by observing how the speed of propagation of a polarized THz signal varies with the angle between the polarization axis of the signal and the sample. Figure 11 is a flow chart illustrating a method 200 according to an example embodiment of the invention. The execution of method 200 may be coordinated by a programmed data processor. All calculations performed in method 200 may be executed automatically in a programmed data processor. Method 200 begins in block 204 by providing a sample of a composite product. The sample is preferably in the form of a sheet having substantially parallel faces. The composite product includes fibres that are distributed in some way in the plane of the sheet.

[0073] In block 206 method 200 obtains a measure of a speed of propagation of polarized THz radiation through the sheet for a given angle between a polarization axis of the THz radiation and an arbitrary reference direction in the plane of the sheet. The measure does not necessarily provide a numerical value for the speed. It is sufficient if the measure results in a value that varies as a function of the propagation speed of THz radiation through the sample. [0074] Loop 208 indicates that block 206 is performed for a plurality of different polarization states (e.g. a plurality of different angles of the polarization axis of the THz radiation to the reference direction). For example, block 206 may be performed for a sequence of 4 to 25 angles. In some embodiments the angles are equally-spaced, for example by an angular step in the range of 1 to 13 degrees. The angles for which block 206 is performed preferably span at least 90 degrees. However, in some embodiments the angles span 45 degrees or even less.

[0075] In block 210 the data from block 206 is analyzed to detect patterns in the data. For example, the data from block 206 may be analyzed to obtain one or more of:

• a bulk volume-averaged fibre orientation;

• an angle between two axes along which fibres in the sample tend to be aligned;

• a relative amount of fibres in a sample which are generally aligned with each of two non-parallel axes in the sample; » a measure of the degree to which fibres of the sample are arranged anisotropically in the plane of the sample; and/or

• the like.

[0076] The determination of average fibre orientation by any of the methods described above can be improved by knowledge of the moisture content of the wood. Further, it can be valuable to have knowledge of the moisture content for other purposes. Moisture content of wood can be expressed relative to the dry weight of wood or the wet weight of the wood. Wood science publications and forest products industry typically express moisture content by the oven dry weight, which is given by:

MC_Dιy _ = (^Ww - ^Wo U)) x 100%

W D₁

where MC_Dry is the moisture content by the oven-dry method; W_w is the wet weight of the wood; and W_D is the dry weight of the wood. MC_Dry can exceed 100%. In the pulp and paper industry, moisture content is typically expressed by the green weight method in which: MC_amι, J^K - ^W°K_m%

where MC_Green is the moisture content by the green weight method.

[0077] THz radiation can be used to measure moisture content in a range of ways which are described below. Some embodiments permit measurement of moisture contents exceeding 30% (this is a range in which it is difficult or impossible to measure moisture contents by many prior methods).

[0078] Moisture content can be measured in solid wood or in wood products including derivatives such as OSB, veneer products, manufactured fibre products and the like. Moisture content can also be measured in precursors to wood products (for example, THz radiation may be used to determine moisture content of a layer of chips being prepared for processing into OSB).

[0079] As moisture content increases, birefrigence is expected to increase. The increased water in the cell walls causes the dielectric contrast between the air in the porous structure and the surrounding walls to increase as a result of the large dielectric constant of water at THz frequencies. Also, wood tends to swell with increasing moisture content, which increases the microfibril angle, and thereby increases the observed birefrigence.

[0080] The large absorption coefficient of water at THz frequencies also causes the absorption coefficient of a material to increase with increasing moisture content. This may be observed as a reduction in field amplitude of transmitted THz radiation relative to a sample with no moisture content.

[0081] Apparatus 10 as shown in Figure 1 may be used to determine moisture content, by transmitting THz radiation emitted from emitter 16 through sample S, detecting the THz radiation received at detector 20, and exploiting a known correlation between moisture content and one or more of the following:

• THz transmission (attenuation),

• THz index of refraction (time delay), • THz birefringence, and

• differences in THz absorption coefficients.

For each of the foregoing observed characteristics (transmission, index of refraction, birefringence, and absorption coefficients), a calibration curve may be determined for a given species and/or wood product type which correlates the observed characteristic to moisture content. This calibration curve may be experimentally derived by transmitting THz radiation at a given incident polarization state through a sample at varying levels of moisture content, and plotting the observed characteristic on a graph. A curve is then fitted to the experimental data to provide a calibration curve. The calibration curve may be linear in the simplest cases, or polynomial, parabolic, hyperbolic, asymptotic, logarithmic, exponential, Gaussian or the like. To determine moisture content, THz radiation is emitted by emitter 16 through a sample S of unknown moisture content, and the value of the observed characteristic detected by detector 20 is compared to the applicable calibration curve in order to extract the value for moisture content. The methods and apparatus for detecting time delay (phase shift) which are discussed above for determining average bulk fibre orientation may also be used for determining moisture content.

[0082] For example, Figure 12A shows a sample calibration curve for the attenuation of THz radiation (field amplitude) transmitted through a sample of aspen wood of approximately 0.9 mm thickness, relative to a sample with no moisture content, where the incident THz radiation is polarized parallel to the visible grain. Figure 12B shows a sample calibration curve for the attenuation of THz radiation (field amplitude) transmitted through a sample of aspen wood, of approximately 0.9 mm thickness, relative to a sample with no moisture content, where the incident THz radiation is polarized perpendicular to the visible grain. Calibration curves for other incident polarization states may also be developed. If field amplitude is measured, the calibration curve tends to be linear (like the curves of Figures 12A and 12B). If intensity is measured, the calibration curve tends to be quadratic. However, other calibration curves may be possible, depending on the wood species and/or wood product evaluated. To obtain the calibration curves of Figures 12A and 12B, the moisture content was determined based on the oven-dry method. To determine the moisture content of a sample S (of unknown moisture content), THz radiation is emitted by emitter 16 at a fixed incident polarization state through sample S, and the attenuation in THz radiation detected by detector 20 is compared to the calibration curve (for the incident polarization state) to extract the value for moisture content.

[0083] Where detection is coherent, the index of refraction of a sample is related to the time delay (phase shift) of a pulse of THz radiation transmitted through the sample, relative to a reference THz pulse that is transmitted in the absence of a sample. Thus, a calibration curve correlating index of refraction to moisture content for a given species of wood product type may be determined by measuring a time delay (phase shift) of a pulse of THz radiation transmitted through the sample at a given incident polarization state and at varying levels of moisture content. Figure 13A shows a sample linear calibration curve for the time delay of THz radiation transmitted through a sample of aspen wood, of approximately 0.9 mm thickness, where the incident THz radiation is polarized parallel to the visible grain. Figure 13B shows a sample linear calibration curve for the time delay of THz radiation transmitted through a sample of aspen wood, of approximately 0.9 mm thickness, where the incident THz radiation is polarized perpendicular to the visible grain. Calibration curves for other incident polarization states may also be developed. To obtain the calibration curves of Figures 13A and 13B, the moisture content was determined based on the oven-dry method. The moisture content of a sample S (of unknown moisture content) may be determined by detecting the time delay in the transmitted THz radiation at a fixed incident polarization state, and comparing the value to the calibration curve (for the incident polarization state) to extract the value for moisture content. [0084] Where detection is not coherent, other standard techniques for measuring index of refraction may be used, such as a standard refractometer or interferometer.

[0085] As the index of refraction is related to the time delay, the birefringence of a sample (difference in index of refraction) is related to a difference in time delay

(phase shift) measurements for two different states of polarization. Birefringence may be determined by taking a first measurement of time delay of a THz radiation pulse transmitted through sample S, for a first incident polarization state, and taking a second measurement of time delay of a THz radiation pulse transmitted through sample S, for a second incident polarization state. The first and second incident polarization states are different (for example, the first incident polarization state may be parallel to the visible grain while the second incident polarization state may be perpendicular to the visible grain).

[0086] A calibration curve may be determined by measuring THz birefringence of a sample, or samples have varying moisture contents. Figure 14 shows a sample linear calibration curve for birefringence of a sample of aspen wood, of approximately 0.9 mm thickness, as a function of moisture content. To determine moisture content, THz radiation is emitted by emitter 16 through a sample S of unknown moisture content, and the birefringence of the sample as determined from signals detected by detector 20 is compared to the calibration curve to extract the value for moisture content. This method of determining moisture content based on birefrigence may be used in situations where it is difficult to measure index of refraction, or where the exact orientation of the visible grain is not known, for example, in situations where the orientation of the visible grain is a random variable of the manufacturing process, as is the case in processes for making wood products such as OSB, from wood chips.

[0087] A calibration curve may also be determined relating moisture content to differences in absorption coefficients, determined from frequency-resolved absorption information. Absorption coefficients may be evaluated using the relations described herein (see, for example, equation (5) above). Figure 15A shows a plot of absorption coefficients as a function of frequency for THz radiation transmitted at the two orthogonal polarization states through aspen at 0% moisture content. Figure 15B shows a plot of absorption coefficients as a function of frequency for THz radiation transmitted at the two orthogonal polarization states through aspen at 8% moisture content. As seen in Figures 15A and Figure 15B, the slope of the absorption curve varies according to the moisture content and the polarization state. Thus, moisture content may be determined based on its correlation to measured differences in absorption coefficients.

[0088] To avoid any difficulties associated with diattenuation in the wood

(dependence of intensity transmittance on the incident polarization state), for example, where the orientation of the visible grain is not known, or is a random variable of the manufacturing process, differences in absorption coefficients may be measured between two different polarization states, such as the two orthogonal polarization states. Figure 15C shows a sample linear calibration curve fitted to a plot of differences in the average absorption coefficient between 0.2 and 0.4 Hz (measured between the two orthogonal polarization states) as a function of moisture content. Figure 15D shows a sample linear calibration curve fitted to a plot of differences in the slope of the absorption curves between the two orthogonal polarization states, as a function of moisture content. In Figure 15D, the correlation between the evaluated absorption coefficient characteristics and moisture content is not as strong as that shown in Figure 15C. In either case, the differences in absorption coefficients at multiple frequencies are measured, either as an average or to generate a slope.

[0089] Calibration curves may also be determined without using two different polarization states. For example, absorption coefficients may be determined for two or more different frequencies at a fixed incident polarization state for varying levels of moisture content, and the average absorption coefficient may be correlated with moisture content. [0090] Moisture content in a sheet of material may be mapped by taking measurements of moisture content at spaced-apart locations over the sheet of material. The determination of proper and uniform moisture content of wood products can be important in the manufacturing of the wood products. For example, quality or manufacturing problems can result where wood elements, such as wood flakes or veneers, are not at an optimal moisture content.

[0091] Some of the additional characteristics described below can be best determined with knowledge of the moisture content of the wood being studied. For measuring these characteristics the moisture content can be determined as described above or may be measured using any suitable alternative method. Known methods for measuring moisture content include:

• gravimetric methods;

• chemical methods such as distillation and Karl Fischer titration methods; • NMR (nuclear magnetic resonance) methods; and,

• methods which measure electrical properties such as resistance or dielectric constant.

[0092] Another characteristic that can be measured through the use of THz radiation is wood density. Wood density can be expressed in various ways including:

• the specific gravity or ratio of the mass of a piece of the wood to its volume (at the current moisture content).

• the ratio of the weight of the wood after oven drying to the volume of the wood after oven drying. • the ratio of the weight of the wood after oven drying to the volume of the wood when green (also known as the "basic density").

[0093] The specific gravity of wood can be measured using THz radiation in a variety of ways. Density measurements may be based on moisture content and one or more of: • absorption of THz radiation;

• THz birefringence; and, • THz index of refraction.

Such density measurements are explained in further detail below.

[0094] The intensity of the THz radiation transmitted through a sample may be related to the absorption properties of the sample using Beer's law as follows:

I = I(0)e-^{a(v) d}, where:

/ is the intensity of the radiation which has been transmitted through the sample, 1(0) is the intensity of the radiation incident on the sample, a(v) is the absorption coefficient at frequency v, and d is the thickness of the material.

[0095] The absorption coefficient α is proportional to the density of the sample. Therefore, through Beer's law, the intensity of the transmitted THz radiation (or the amplitude of the transmitted field) has an exponential dependence on density. By detecting the intensity of the transmitted radiation (or the amplitude of the transmitted field) at spaced-apart locations over a sheet of material, a density profile transmission image may be generated. Three sample density profile transmission images produced using such a technique are shown in Figures 16A, B and C. Darker pixels in the images correspond to areas of higher density.

[0096] The optical properties of most materials can be described as a microscopic electromagnetic excitation of dipole moments in the constituents of the materials. The optical response depends on the number of oscillating dipoles, and hence, the density of the material. This interaction can be described in the dipole approximation as:

where n is the index of refraction, N is the density of oscillators, m_e is the mass of an electron, ω_o is the resonant frequency of the oscillator, ω is the optical frequency of the light, and ε₀ is the permittivity of free space. Assuming that the second term in the foregoing relationship is small relative to 1, then to a reasonable approximation:

[0097] In this case, n (ω) will depend linearly on the density of the material. This is a reasonable approximation for materials having indices of refraction close to 1 , such as wood (where n = 1 to 1.5). When examining the complex index of refraction, both the real (related to the index of refraction) and imaginary (related to absorption coefficient) components of n² will depend on density.

[0098] As noted above, the index of refraction of a material is linearly proportional to the density of the material. Given that the time delay (phase shift) of a transmitted THz pulse is linearly proportional to the index of refraction, it is expected that detection of the absolute time delay (phase shift) of a transmitted pulse through the sample (relative to a reference pulse transmitted in the absence of a sample) may be used to produce a transmission image of time delay (phase shift) that has a linear dependence on density. Sample density profile transmission images of OSB samples generated by taking measurements of time delay (phase shift) at spaced-apart locations over the sheet of material are shown in Figures 17A and 17B. Lighter pixels in the images correspond to areas of higher density. The OSB sample imaged in Figure 17A therefore has a higher average density than the OSB sample imaged in Figure 17B.

[0099] To obtain a correlation or calibration curve between a THz transmission image and density, the sample may be cut into sections, and the actual measured density of each section (determined by weighing the section and measuring its dimensions) is correlated with the corresponding pixel value(s) in the image. In investigating the optical properties at THz frequencies of OSB samples, and measuring the physical density of the samples, a correlation was observed between the physical density and the average index of refraction, absorption coefficient and time delay, respectively, as shown in Figures 19A, 19B, and 19C. Once physical density is calibrated in each case, an estimate of density can be obtained from images of THz properties such as intensity of transmitted THz radiation or time delay (phase shift) of transmitted THz pulses.

[0100] The rate at which wood dries is related to the density of wood. In many cases, denser wood dries more slowly than wood that is less dense. Density measurements and/or moisture content measurements, as described herein, may be used to sort wood by density and/or initial moisture content to permit better scheduling of drying. Species of similar density or drying characteristics, including initial moisture content, may be dried together. The basic density may be estimated from the current moisture content, and specific gravity from knowledge of the wood species and the typical moisture content of that wood species when green.

[0101] Further, the density of material having a preferred fibre orientation (fibre density) can be measured using THz transmission images. In the process of manufacturing OSB, pieces of wood are cut into regularly sized chips, called flake, which are formed into layers by a former and pressed together with resins to produce the final OSB product. Typically, a first layer of ftake (surface flake) is created with the direction of the visible grain of the flake oriented parallel to the machine direction. A second layer of flake (core flake) is created with the direction of the visible grain of the flake oriented perpendicular to the machine direction. A third layer of flake (surface flake) is created with the direction of the visible grain of the flake oriented parallel to the machine direction. The typical ratio of core to surface flake is approximately 3:7. Other orientations and combinations of layers of flake may be used to create a mat of OSB.

[0102] The index of refraction for THz transmission at a given incident state of polarization depends on the orientation of the fibres relative to the polarization direction. Thus, for a fibre composite product such as OSB, the fibre density in a direction of polarization may be measured by transmission of pulses of polarized THz radiation through the sample, and measuring the time delay (phase shift) of the pulse, relative to transmission of THz radiation in the absence of the sample.

[0103] The fibrous structure of the OSB results in birefringence, so that there is a difference in the index of refraction for THz radiation depending on the polarization axis relative to the direction of the visible grain. Thus, fibre density may be determined by measuring a difference in time delays between transmitted THz pulses having different polarization states (e.g. polarization parallel to and perpendicular to the machine direction).

[0104] For a typical OSB mat as described above, having a ratio of core to surface flake of 3:7, it is expected that the THz transmission images will show a higher fibre density for incident THz radiation polarized in a direction parallel to the machine direction and a lower fibre density for incident THz radiation polarized in a direction perpendicular to the machine direction. In order to extract values for fibre density from the THz transmission images, calibration curves may be determined, in much the same way as calibration curves described above for measuring the specific gravity of a sample. Fibre density may be measured for the calibration curve by cutting a sample into sections and further cutting each section into pieces each formed of a layer of flake oriented in a preferential direction, and comparing the weight and dimension of each piece with the weight and dimension of the section from which the piece was obtained.

[0105] The variability in fibre density in the polarization direction, as the polarization direction is varied, is a function of the anistropy of the fibres. No variation with polarization angle indicates a completely isotropic distribution of fibres, whereas a distribution that is highly centered at specific orientation(s) indicates a very large degree of anistropy. For OSB manufacturing, a very large degree of anisotropy is desired for each individual layer of flake, to provide mechanical strength in the two orthogonal directions. Also, there will be residual anistropy in the manufactured OSB due to the 3:7 ratio between core to surface flake. [0106] Figure 23 A is an optical picture of an OSB sample. The OSB sample has two layers of surface flake (with the visible grain of the flake oriented substantially parallel to the machine direction) and one layer of core flake (with the visible grain of the flake oriented substantially perpendicular to the machine direction). Figure 23B is a density profile transmission image generated by detecting the field amplitude of the transmitted THz radiation, polarized parallel to the machine direction, transmitted through the OSB sample shown in Figure 23A. Figure 23C is a density profile transmission image generated by detecting the field amplitude of the transmitted THz radiation, polarized perpendicular to the machine direction, transmitted through the OSB sample shown in Figure 23 A. Darker areas of each image indicate areas of lesser density than the brighter areas of the image. As expected, the density profile images show substantially less density for transmitted THz radiation polarized perpendicular to the machine direction (Figure 23C), than for transmitted THz radiation polarized parallel to the machine direction (Figure 23B). Therefore, fibre density can be preferentially probed in oriented composite wood products. The core flake is preferentially probed in Figure 23B, whereas the surface flake density is being preferentially probed in Figure 23C.

[0107] Another characteristic of wood that may be measured using THz radiation is the microfibril angle. Microfibrils are bundles of cellulose chains that are bound together in the cell walls of wood by a sheath of hemicellulose and lignin. The cell wall is made up of layers that include sub-layers of microfibrils. The microfibril angle is defined between the axis of the tree and the microfibrils within the sub-layers in the cell walls. The microfibril angle affects a number of mechanical properties of wood such as the longitudinal modulus of elasticity (MOE) (strength properties) and shrinkage.

[0108] Compression wood forms in the underside of a softwood tree that is growing at an incline, as there is an increased amount of auxin, a growth-controlling substance, in the underside of the wood. Compression wood tends to have a higher microfibrillar angle and density compared to normal wood. By measuring one or both of microfibril angle and density in a piece of wood, areas in which the wood is made up of compression wood can be distinguished from other areas. This information can be applied in various ways including as an input to a lumber optimization system which will control cutting of the wood to maximize value by making some pieces of lumber substantially entirely in the compression wood. Generally, boards and planks containing some compression wood are susceptible to bow in seasoning because of the increased shrinkage of the wood. As well, compression wood has relatively low bending strength and lacks toughness compared to normal wood.

[0109] Juvenile or first- formed wood, which is laid down by the cambium near the center of a tree, may have different properties from more mature wood formed at a greater number of rings from the pith. Some types of juvenile wood have increased shrinkage over more mature wood, due to an increased microfibril angle and/or density in the juvenile wood. Thus, areas of juvenile wood can be distinguished from more mature wood, by measuring one or both of microfibril angle and density in a piece of wood. Information regarding the locations in a log of juvenile wood may be input into a lumber optimization system in a sawmill to selectively cut boards that do or do not include juvenile wood.

[0110] Microfibril angle affects the birefringence of wood in the THz spectrum. Thus, birefrigence may be correlated with microfibril angle for a species of wood to obtain a calibration curve for estimating the microfibril angle of a wood sample based on a measurement of birefrigence. As the microfibril angle is generally larger in compression and juvenile wood than in normal wood, a determination of microfibril angle can then be used to determine if the wood is compression or juvenile wood.

[0111] Since moisture content also affects THz birefringence, methods for determining microfibril angle according to the invention may include measuring moisture content, as described above, and also measuring the birefringence (or a property that varies in a known manner with the birefringence). The microfibril angle can then be determined. This may be done, for example, by experimentally obtaining a function 5(MC, ψ) relating the observed birefringence to the moisture content (MC) and the microfibril angle, ψ. This function can then be inverted to provide a function ψ(MC, B) that provides the microfibril angle for a known combination of moisture content and birefringence.

[0112] The strength of wood or wood products can be measured using THz radiation in a variety of ways. Indicators of the mechanical strength of an OSB board include, for example, one or more of the following observed characteristics: specific gravity, the degree of anistropy of the board, the magnitude of fibre density (in a preferred orientation) of the overall board, and the difference between fibre density parallel to the machine direction and fibre density perpendicular to the machine direction. The breaking strength of an OSB board may be correlated to any one of such characteristics. In some embodiments, strength is determined based on fibre orientation and density, as discussed above. In some embodiments, strength is determined based on microfibril angle of the wood and density. Other measures of strength are also possible.

[0113] Mechanical strength of OSB may be evaluated by scanning samples using a THz imaging system, and testing the strength of the samples with a mechanical device which measures the amount of stress that can be applied before breaking. The mechanical strength of materials depends on density, and so there is a correlation of breaking strength with density, as shown in Figure 20.

[0114] As seen in Figures 21A, 21B and 22C, there is also a correlation of the breaking strength of OSB with index of refraction, absorption coefficient, and time delay of THz pulse propagation respectively. Therefore, these THz properties can be used as online monitors of breaking strength. A likely source of the correlation is that these properties have a correlation with physical density, as discussed above, and that physical density is related to the breaking strength. [0115] A sheet of OSB or other wood-based material can be imaged using THz radiation in a manner that indicates strength of the sheet at different locations. Such images can be used to grade sheets of such material. The material can then be marked with the applicable grade and/or sorted according to grade using automated printing / sorting equipment that receives the grade as an input. Grade may be based upon one or more of minimum strength; average strength; variation in strength; either over the entire sheet or in selected areas within the sheet.

[0116] In addition to grading, images of strength, density or the like, acquired according to methods as described herein can be applied to the automatic identification and repair of defects such as knots, splits, rot, juvenile or compression wood. Areas having undesirably low strength (or density) can be identified in the image, and coordinates identifying the low-strength identified areas can be provided to a patching system which can automatically cut out and replace a section of material with a suitable patch. Suitable patching systems are known in the art.

[0117] THz properties may also be used as an indicator of the species of wood. As seen in Figure 22A, the profile of transmitted THz waveforms varies for wood from different tree species. To determine the species of a sample of wood, signal analysis may be performed on a transmitted THz waveform through the sample. For example, by analysing the transmission profiles for the three samples of spruce wood shown in the last row of profiles of Figure 22 A, it is clear that the unidentified sample (lorne) is white spruce, and not the same spruce species as in the 2x4 sample.

[0118] Figures 22B and 22C show the absorption coefficients and index of refraction for THz radiation transmitted at the two orthogonal polarization states through wood samples from 10 different species. It can be seen that the absorption coefficient and index of refraction vary according to the species. By quantifying these and other properties for different moisture contents and species, as well as determining the variability in these quantities within a given species, a look-up table can be constructed to determine the species. [0119] Microfibril angle also varies among tree species. A measurement of the microfibril angle (or more generally the birefringence of a wood sample to THz radiation) in a sample of wood can be used together with other characteristics of the wood sample, such as density, to determine what species of tree produced the sample of wood.

[0120] Figure 18 is a schematic view of a system for determining and monitoring various characteristics of wood, wood products and other fibrous materials in accordance with an embodiment of the invention. Wood monitoring system 300 comprises a THz transmission and detection subsystem 302 for transmission of THz radiation through a sample which interacts with the radiation, and detection of the transmitted THz radiation. THz transmission and detection subsystem 302 may comprise an apparatus such as that described for apparatus 10. Because moisture content of a sample affects the way THz radiation interacts with the sample, predetermined moisture content information may be input into THz transmission and detection subsystem 302 in order to obtain more accurate measurements of the characteristics of the material. Moisture content information may be determined either through a THz transmission and detection system (using the methods and apparatus described herein) as shown at block 303, through other moisture content systems, as illustrated by block 305, or any suitable combination thereof. Optionally, species information of the sample may also be input to THz transmission and detection subsystem 302 to enhance the measurements of the characteristics of the material.

[0121] Using measurements obtained by THz transmission and detection subsystem 302, bulk volume-averaged fibre orientation, density (such as specific gravity, basic density or fibre density), and microfibril orientation may be determined respectively by fibre orientation subsystem 306, density measurement subsystem 308 and microfibril orientation subsystem 310. Fibre orientation subsystem 306 may determine bulk volume-averaged fibre orientation by executing a function which relates fibre orientation to observed THz time delay (phase shift) or birefrigence. Density measurement subsystem 308 may determine density measurements by executing a function which relates density to observed THz time delay (phase shift), birefrigence or intensity (absorption properties). Microfibril orientation subsystem 310 may determine microfibril orientation by executing a function which relates microfibril orientation to observed THz birefrigence and moisture content.

[0122] Further characteristics may be derived or determined from the characteristics determined by the above subsystems. For example, strength of a composite material such as OSB may be determined by a strength (composites) subsystem 311 based on information obtained from THz transmission and detection subsystem 302 or density measurement subsystem 308 or both. Optionally, strength (composites) subsystem 311 may also use microfibril orientation information from microfibril orientation subsystem 310 in order to evaluate strength.

[0123] The modulus of elasticity of a material may be determined by a strength (modulus of elasticity) subsystem 313 based on information obtained from density measurement subsystem 308 and microfibril orientation subsystem 310. The presence of compression wood in a softwood sample may be detected by a compression wood (softwood) subsystem 315 based on information obtained from density measurement subsystem 308 or microfibril orientation subsystem 310 or both. The species of a sample of wood may be determined by a species subsystem 317 based on density information obtained from density measurement subsystem 308 and microfibril orientation information obtained from microfibril orientation subsystem 310, and optionally, moisture content information.

[0124] Wood monitoring system 300 may comprise an interface 320 permitting an operator to request and access information determined by wood monitoring system 300. For example, interface 320 may provide reports and trending information 322 regarding characteristics of the wood sample, quality control or material rejection information 324, and imaging 326 shown on a display (or printed) to show the materials and characteristics that are being monitored. Based on certain information obtained from wood monitoring system 300 (for example, the presence of compression or low-strength wood, or less than optimally aligned layers of flake in an OSB), interface 320 may also send feedback 328 to a wood processing system, which is not shown, so that appropriate corrections or changes can be made in the wood processing to enhance the quality of the final product or to optimize the cutting of a piece of wood, and the like.

[0125] Information about the characteristics of the materials may be applied by wood monitoring system 300 to grade the materials (e.g. based on strength, density, etc.) Wood monitoring system 300 may identify materials which have a grade below a certain threshold grade, and/or sort the materials by grade.

[0126] In some embodiments, apart from subsystem 302, the subsystems of wood monitoring system 300 are provided by software modules that perform appropriate computations based on data from THz transmission and detection subsystem 302.

[0127] Where a component (e.g. a software module, processor, assembly, device, circuit, etc.) is referred to above, unless otherwise indicated, reference to that component (including a reference to a "means") should be interpreted as including as equivalents of that component any component which performs the function of the described component (i.e., that is functionally equivalent), including components which are not structurally equivalent to the disclosed structure which performs the function in the illustrated exemplary embodiments of the invention.

[0128] The invention has a broad range of aspects including, without limitation:

• determining moisture content using THz radiation as described herein;

• determining average fibre orientation using THz radiation and a measure of moisture content (which is obtained by using THz radiation in some embodiments and is obtained in alternative ways which do not employ THz radiation in other embodiments); • determining density of a sample using THz radiation as described herein;

• determining microfibril orientation using THz radiation as described herein;

• determining wood species based upon one or more of the above;

• imaging based upon one or more of the above; • controlling a process for making a wood product according to one or more of the above;

• grading and/or sorting wood-based materials according to one or more of the above.

[0129] As will be apparent to those skilled in the art in the light of the foregoing disclosure, many alterations and modifications are possible in the practice of this invention without departing from the spirit or scope thereof. For example, the invention may be applied to determining characteristics of fibrous plant-based materials other than wood from trees such as grasses (including bamboos), and products made from such materials.

Claims

WHAT IS CLAIMED IS:

1. A method of determining density of a fibrous material, the method comprising: directing polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the material; detecting THz radiation originating from the source of THz radiation and emerging on a second face of the material; obtaining a value for a THz transmission characteristic of the material based at least in part on the detected THz radiation; and applying a first predetermined correlation function to the THz transmission characteristic value to obtain a measure of density of the material, wherein the first predetermined correlation function relates density values with values of the THz transmission characteristic.

2. A method according to claim 1 , wherein obtaining the THz transmission characteristic value comprises measuring intensity of the detected THz radiation.

3. A method according to claim 1 , wherein obtaining the THz transmission characteristic value comprises measuring field amplitude of the detected THz radiation.

4. A method according to claim 1 , comprising: directing a first pulse of THz radiation from the source of THz radiation in a selected incident polarization state at the first face of the material; and detecting the first pulse of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of index of refraction of the material.

5. A method according to claim 4, wherein the measure of index of refraction of the material is based at least in part on a time delay at which the first pulse of THz radiation is detected relative to a reference pulse of polarized THz radiation originating from the source of THz radiation.

6. A method according to claim 1, comprising: directing a first pulse of THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of birefringence.

7. A method according to claim 6, wherein the measure of birefrigence is based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected.

8. A method according to claim 7, wherein the first and second incident polarization states are linear polarization states wherein the first pulse of THz radiation is polarized along a polarization axis which is substantially orthogonal to a polarization axis along which the second pulse of THz radiation is polarized.

9. A method according to claim 1, comprising: determining moisture content of the material; and applying the first predetermined correlation function to obtain the measure of density of the material, wherein inputs to the first predetermined correlation function comprise the THz transmission characteristic value and the moisture content.

10. A method according to any one of claims 1 to 8, comprising: applying a second predetermined correlation function to the THz transmission characteristic value to obtain a measure of moisture content of the material, wherein the second predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

11. A method according to claim 10, wherein obtaining the THz transmission characteristic value comprises evaluating a transmission attenuation of THz radiation for the material relative to a transmission attenuation of THz radiation directed from the source of THz radiation in the selected incident polarization state at a reference material having substantially no moisture content.

12. A method according to claim 11, wherein evaluating the transmission attenuation of THz radiation for the material comprises measuring a field amplitude of the detected THz radiation.

13. A method according to claim 11 , wherein evaluating the transmission attenuation of THz radiation for the material comprises measuring an intensity of the detected THz radiation.

_* 14. A method according to any one of claims 1 to 7. comprising: directing first THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing second THz radiation from the source of THz radiation in a second incident polarization state different from the second incident polarization state at the first face of the material; detecting the first and second THz radiation originating from the source of THz radiation and emerging on the second face of the material; determining a difference in average absorption coefficients of the detected first and second THz radiation; and applying a second predetermined correlation function to the difference in average absorption coefficients to obtain a measure of moisture content of the material, wherein the second predetermined correlation function relates moisture content values with differences in average absorption coefficients for the first and second incident polarization states.

15. A method according to any one of claims 1 to 14 comprising: applying a third predetermined correlation function to the THz transmission characteristic value to obtain a measure of strength of the material, wherein the third predetermined correlation function relates strength values with values of the THz transmission characteristic.

16. A method according to claim 14, comprising: applying a third predetermined correlation function to the difference in average absorption coefficients to obtain a measure of strength of the material, wherein the third predetermined correlation function relates strength values with differences in average absorption coefficients for the first and second incident polarization states.

17. A method according to either one of claims 6 and 7, comprising: applying a second predetermined correlation function to the measure of birefrigence to obtain a measure of microfibril orientation of the material, wherein the second predetermined correlation function relates microfibril orientation values with birefrigence values.

18. A method according to either one of claims 6 and 7, comprising: determining moisture content of the material; applying a second predetermined correlation function to the measure of birefrigence and the moisture content to obtain a measure of microfibril orientation of the material, wherein the second predetermined correlation function relates microfibril orientation values with birefrigence values and moisture content values.

19. A method according to claim 18, wherein determining moisture content comprises applying a third predetermined correlation function to the

THz transmission characteristic value to obtain a measure of moisture content of the material, wherein the third predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

20. A method according to either one of claims 17 or 18, comprising detecting compression wood in the material by comparing the measured microfibril orientation with a threshold value indicating a presence of compression wood in the material.

21. A method according to any one of claims 1 to 20, wherein the material forms a web of material moving in a plane of the material and the method is performed while the material is moving.

22. A method according to any one of claims 1 to 16, wherein the material comprises composite fibrous material.

23. A method according to any one of claims 1 to 16, wherein the material comprises oriented strand board.

24. A method according to any one of claims 1 to 21 , wherein the material comprises wood.

25. An apparatus for determining density of a fibrous material, the apparatus comprising: a source of THz radiation located to direct polarized THz radiation at a first face of the material in a selected incident polarization state; a detector of THz radiation located to detect the THz radiation originating from the source of THz radiation emerging on a second face of the material; a THz transmission measurement subsystem configured to measure at least one THz transmission characteristic value of the material based at least in part on the detected THz radiation; and a density measurement subsystem configured to apply a first predetermined correlation function to the at least one THz transmission characteristic value to obtain a measure of density of the material, wherein the first predetermined correlation function relates density values with values of the THz transmission characteristic.

26. An apparatus according to claim 25, comprising a strength measurement subsystem configured to apply a second predetermined correlation function to the at least one THz transmission characteristic value to obtain a measure of strength of the material, wherein the second predetermined correlation function relates strength values with values of the THz transmission characteristic.

27. An apparatus according to either one of claims 25 or 26, comprising a moisture measurement subsystem configured to apply a third predetermined correlation function to the at least one THz transmission characteristic value to obtain a measure of moisture content of the material, wherein the third predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

28. An apparatus according to either one of claims 25 or 26, comprising a moisture measurement subsystem configured to determine moisture content of the material by gravimetric analysis, chemical analysis, nuclear magnetic resonance analysis or electrical property analysis.

29. An apparatus according to either one of claims 27 or 28, comprising a microfibril orientation subsystem configured to apply a fourth predetermined correlation function to the at least one THz transmission characteristic value to obtain a measure of microfibril orientation of the material, wherein the fourth predetermined correlation function relates microfibril orientation values with values of the THz transmission characteristic.

30. An apparatus according to claim 29, comprising a compression wood detection subsystem configured to detect compression wood in the material by comparing the measured microfibril orientation with a threshold value indicating a presence of compression wood in the material.

31. An apparatus according to claim 30, comprising a species identification subsystem configured to identify species of wood of the material by comparing the at least one THz transmission characteristic value to each one of a plurality of predetermined ranges of THz transmission characteristic values, wherein each one of the predetermined ranges of THz transmission characteristic values corresponds to a species of wood.

32. An apparatus according to any one of claims 25 to 31 , comprising a display configured to display a density profile image map for the material.

33. An apparatus according to any one of claims 26 to 31 , comprising a display configured to display a strength profile image map for the material.

34. An apparatus according to any one of claims 27 to 31, comprising a display configured to display a moisture profile image map for the material.

35. An apparatus according to either one of claims 30 and 31 , comprising a display configured to display an image of the material indicating regions of the material which have a probability of compression wood above a predetermined threshold value.

36. An apparatus according to any one of claims 26 to 31 , comprising a grader configured to assign a grade to the material according to the measured strength of the material.

37. An apparatus according to claim 36, comprising a sorter configured to direct the material to a first location if the material has an assigned grade below a predetermined threshold grade.

38. An apparatus according to claim 36, comprising a feedback subsystem configured to output an alert if the material has an assigned grade below a predetermined threshold value.

39. A method of determining density of a composite fibrous material, the method comprising: directing polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the material; detecting THz radiation originating from the source of THz radiation and emerging on a second face of the material; obtaining a value for a THz transmission characteristic of the material based at least in part on the detected THz radiation; and applying a predetermined correlation function to the THz transmission characteristic value to obtain a measure of the density of the material, wherein the predetermined correlation function relates density values with values of the THz transmission characteristic.

40. A method according to claim 39. wherein obtaining the THz transmission characteristic value comprises measuring intensity of the detected THz radiation.

41. A method according to claim 39, wherein obtaining the THz transmission characteristic value comprises measuring field amplitude of the detected THz radiation.

42. A method according to claim 39, comprising: directing a first pulse of THz radiation from the source of THz radiation in a selected incident polarization state at the first face of the material; and detecting the first pulse of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of index of refraction.

43. A method according to claim 42, wherein the measure of index of refraction of the material is based at least in part on a time delay at which the first pulse of THz radiation is detected relative to a reference pulse of polarized THz radiation originating from the source of THz radiation.

44. A method according to claim 39, comprising: directing a first pulse of THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of birefringence of the material.

45. A method according to claim 44, wherein the measure of birefringence of the material is based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected.

46. A method according to claim 45, wherein the first and second incident polarization states are linear polarization states wherein the first pulse of THz radiation is polarized along a polarization axis which is substantially orthogonal to a polarization axis along which the second pulse of THz radiation is polarized.

47. A method according to claim 39, comprising: determining moisture content of the material; and applying the predetermined correlation function to obtain the measure of the density of the material, wherein inputs to the predetermined correlation function comprise the THz transmission characteristic value and the moisture content.

48. A method of generating a map of density of a fibrous material, the method comprising repeating the steps of claim 39 for each of a plurality of spaced apart regions of the material to determine a density of the region of the material.

49. A method of determining moisture content of a composite fibrous material, the method comprising: directing polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the material; detecting THz radiation originating from the source of THz radiation and emerging on a second face of the material; obtaining a value for a THz transmission characteristic of the material based at least in part on the detected THz radiation; and applying a predetermined correlation function to the THz transmission characteristic value to obtain a measure of the moisture content of the material, wherein the predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

50. A method according to claim 47, wherein determining moisture content comprises: applying a second predetermined correlation function to the THz transmission characteristic value to obtain a measure of the moisture content of the material, wherein the second predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

51. A method according to claim 49, wherein obtaining the THz transmission characteristic value comprises evaluating a transmission attenuation of THz radiation for the material relative to a transmission attenuation of THz radiation directed from the source of THz radiation in the selected incident polarization state at a reference material having substantially no moisture content.

52. A method according to claim 51 , wherein evaluating the transmission attenuation of THz radiation for the material comprises measuring a field amplitude of the detected THz radiation.

53. A method according to claim 51 , wherein evaluating the transmission attenuation of THz radiation for the material comprises measuring an intensity of the detected THz radiation.

54. A method according to claim 49, comprising: directing a first pulse of THz radiation from the source of THz radiation in a selected incident polarization state at the first face of the material; and detecting the first pulse of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of index of refraction of the material.

55. A method according to claim 54, wherein the measure of index of refraction of the material is based at least in part on a time delay at which the first pulse of THz radiation is detected relative to a reference pulse of polarized THz radiation originating from the source of THz radiation.

56. A method according to claim 49, comprising: directing a first pulse of THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of birefringence of the material.

57. A method according to claim 56, wherein determining the measure of the birefrigence of the material is based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected.

58. A method according to claim 57, wherein the first and second incident polarization states are linear polarization states wherein the first and second pulses of THz radiation are each polarized along a different polarization axis and wherein the polarization axes of the first and second pulses of THz radiation each make a different angle with a reference axis in the plane of the material.

59. A method according to claim 49, comprising: directing first THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing second THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a difference in average absorption coefficients of the detected first and second THz radiation.

60. A method according to claim 59, wherein the first and second incident polarization states are linear polarization states wherein the first THz radiation is polarized along a polarization axis which is substantially orthogonal to a polarization axis along which the second THz radiation is polarized.

61. A method of generating a map of moisture content of a fibrous material, the method comprising repeating the steps of claim 49 for each of a plurality of spaced apart regions of the material to determine a moisture content of the region of the material.

62. A method of determining microfibril orientation of a composite fibrous material, the method comprising: directing a first pulse of THz radiation from a source of THz radiation in a first incident polarization state at a first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on a second face of the material; obtaining a value for birefringence of the material based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected; and applying a predetermined correlation function to the birefrigence value to obtain a measure of the microfibril orientation of the material, wherein the predetermined correlation function relates microfibril orientation values with birefrigence values.

63. A method of determining microfibril orientation of a composite fibrous material, the method comprising: directing a first pulse of THz radiation from a source of THz radiation in a first incident polarization state at a first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on a second face of the material; determining moisture content of the material; obtaining a value for birefringence of the material based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected; and applying a predetermined correlation function to the birefrigence value and the moisture content to obtain a measure of the microfibril orientation of the material, wherein the predetermined correlation function relates microfibril orientation values with birefrigence values and moisture content values.

64. A method according to claim 63, wherein determining moisture content comprises: applying a second predetermined correlation function to the THz transmission characteristic value to obtain a measure of the moisture content of the material, wherein the second predetermined correlation function relates moisture content values with values of the THz transmission characteristic.

65. A method of detecting compression wood in a composite fibrous material, the material comprising wood or pieces of wood, the method comprising, for each of a plurality of spaced apart regions of the material, determining microfibril orientation of the region according to the method of claim 62 or 63, and comparing the determined microfibril orientation of the region with a threshold value indicating a presence of compression wood in the material.

66. A method of detecting compression wood in a composite fibrous material, the material comprising wood or pieces of wood, the method comprising, for each of a plurality of spaced apart regions of the material, determining microfibril orientation of the region according to the method of claim 62 or 63 and determining density of the region, and comparing the determined microfibril orientation and density of the region with threshold values indicating a presence of compression wood in the material.

67. A method of determining strength of a composite fibrous material, the method comprising: directing polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the material; detecting THz radiation originating from the source of THz radiation and emerging on a second face of the material; obtaining a value for a THz transmission characteristic of the material based at least in part on the detected THz radiation; and applying a predetermined correlation function to the THz transmission characteristic value to obtain a measure of the strength of the material, wherein the predetermined correlation function relates strength values with values of the THz transmission characteristic.

68. A method according to claim 67, comprising: directing a first pulse of THz radiation from the source of THz radiation in a selected incident polarization state at the first face of the material; and detecting the first pulse of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of index of refraction of the material.

69. A method according to claim 68, wherein the measure of index of refraction of the material is based at least in part on a time delay at which the first pulse of THz radiation is detected relative to a reference pulse of polarized THz radiation originating from the source of THz radiation.

70. A method according to claim 67, comprising: directing a first pulse of THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a measure of birefringence of the material.

71. A method according to claim 70, wherein the measure of birefrigence of the material is based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected.

72. A method according to claim 71 , wherein the first and second incident polarization states are linear polarization states wherein the first and second pulses of THz radiation are each polarized along a different polarization axis and wherein the polarization axes of the first and second pulses of THz radiation each make a different angle with a reference axis in the plane of the material.

73. A method according to claim 67, comprising: directing first THz radiation from the source of THz radiation in a first incident polarization state at the first face of the material; directing second THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the material; and detecting the first and second THz radiation originating from the source of THz radiation and emerging on the second face of the material; wherein obtaining the THz transmission characteristic value comprises determining a difference in average absorption coefficients of the detected first and second THz radiation.

74. A method according to claim 73, wherein the first and second incident polarization states are linear polarization states wherein the first THz radiation is polarized along a polarization axis which is substantially orthogonal to a polarization axis along which the second THz radiation is polarized.

75. A method of generating a map of strength of a fibrous material, the method comprising repeating the steps of claim 67 for each of a plurality of spaced apart regions of the material to determine a strength of the region of the material.

76. A method of determining species of a piece of wood, the method comprising: directing a pulse of polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the wood; detecting the pulse of THz radiation originating from the source of THz radiation and emerging on a second face of the wood; measuring field amplitude of the detected THz radiation over a predetermined period of time to generate a time domain signal of the detected THz radiation; and comparing the time domain signal of the detected THz radiation with each one of a plurality of predetermined time domain signals, wherein each one of the predetermined time domain signals corresponds to a species of wood.

77. A method of determining species of a piece of wood, the method comprising: directing polarized THz radiation from a source of THz radiation in a selected incident polarization state at a first face of the wood; detecting THz radiation originating from the source of THz radiation and emerging on a second face of the wood; obtaining a value for a THz transmission characteristic of the wood based at least in part on the detected THz radiation; and comparing the THz transmission characteristic value to each one of a plurality of predetermined ranges of THz transmission characteristic values, wherein each one of the predetermined ranges of THz transmission characteristic values corresponds to a species of wood.

78. A method according to claim 77, comprising: directing a first pulse of THz radiation from the source of THz radiation in a selected incident polarization state at the first face of the wood; and detecting the first pulse of THz radiation originating from the source of THz radiation and emerging on the second face of the wood; wherein obtaining the THz transmission characteristic value comprises determining a measure of index of refraction of the wood.

79. A method according to claim 78, wherein the measure of index of refraction of the wood is based at least in part on a time delay at which the first pulse of THz radiation is detected relative to a reference pulse of polarized

THz radiation originating from the source of THz radiation.

80. A method according to claim 77, comprising: directing a first pulse of THz radiation from the source of THz radiation in a first incident polarization state at the first face of the wood; directing a second pulse of THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the wood; and detecting the first and second pulses of THz radiation originating from the source of THz radiation and emerging on the second face of the wood; wherein obtaining the THz transmission characteristic value comprises determining a measure of birefringence of the wood.

81. A method according to claim 80, wherein determining the measure of birefrigence of the wood is based at least in part on a difference in time delays at which the first and second pulses of THz radiation are detected.

82. A method according to claim 81, wherein the first and second incident polarization states are linear polarization states wherein the first and second pulses of THz radiation are each polarized along a different polarization axis and wherein the polarization axes of the first and second pulses of THz radiation each make a different angle with a reference axis in the plane of the wood.

83. A method according to claim 77, comprising: directing first THz radiation from the source of THz radiation in a first incident polarization state at the first face of the wood; directing second THz radiation from the source of THz radiation in a second incident polarization state different from the first incident polarization state at the first face of the wood; and detecting the first and second THz radiation originating from the source of THz radiation and emerging on the second face of the wood; wherein obtaining the THz transmission characteristic value comprises determining a difference in average absorption coefficients of the detected first and second THz radiation.

84. A method according to claim 83, wherein the first and second incident polarization states are linear polarization states wherein the first THz radiation is polarized along a polarization axis which is substantially orthogonal to a polarization axis along which the second THz radiation is polarized.