Measuring the optical and acoustic properties of turbid materials by scattering photoacoustic method

Field of the invention

The invention relates to a method and measurement arrangement for measuring optical parameters and composition of turbid materials by pulsed laser radiation.

Background of the invention

Measuring optical and acoustic parameters of materials is important because it is closely relative to the material composition and property. The techniques based on the integrating sphere (measuring transmission and reflectance), optical diffusion with added a small quality of absorbing material, time-of-flight of photons, and time-resolved stress detection (TRSD) with measuring total diffusive reflectance have been developed to measure optical parameters of various media. Although these techniques provide accurate results, their use in particle applications has some drawbacks.

The integrating sphere method is used for measuring a thin layer sample, which method can not apply for thick object. The optical diffusive method with added absorbing material can not be used in non-invasive or on-line measurement. The devices required by time-of-flight technique include an ultra short pulse laser and a single-photon avalanche photodiode which are very expensive. Moreover, these optical methods can not give any acoustic information of material.

Conventional photoacoustic techniques are based on the mechanism of optical absorption of studied medium, in which a photoacoustic source (PA) is produced. They are divided into two categories according to their measuring parameters: photoacoustic spectroscopy (PAS) which detects the photoacoustic amplitude produced by the studied object, and time-resolved photoacoustic technique (or time-resolved stress detection TRSD) which determines both stress profile and amplitude.

However, in some cases, both PAS and TRSD techniques will be ineffective be- cause of the photoacoustic effect produced in a studied object is too weak to be detected; for example, when the absorption coefficient of a highly turbid material is

very weak or near to zero at study wavelengths, or if a low energy laser is used as the excitation source for low cost and portable purposes.

One measurement configuration is known from US 2005/0105095. The reference discloses a system for measuring optical absorption properties of materials. In the described method the unabsorbed light energy, either directly traversed through the material or back scattered from the material, is first detected. The material absorbs a part of the light which the material afterwards transmits as ultrasonic waves. Also this ultrasonic energy is detected. The detected energies are utilized when calculating the absorption coefficient of the studied material.

In a publication of Optica Applicata, VoI XXXIV, No 4, 2004 pages 647-656 is described a method where an approximation of diffusion theory is used to estimate the reduced scattering coefficient. The reduced scattering coefficient is utilized to calculate the fines content of pulp.

An example of a measuring application in an industrial process is a measurement of the fibre and fines consistencies in paper pulp. In paper industry, the pulp consists of wood cells called fibres, fibre fragments called fines and fillers are added to improve the optical and solidity properties of the final paper product. On-line measuring the pulp consistency during the processes is a requirement for better paper quality, more effective use of raw materials, lower production costs and faster paper machine speeds. The prior art measuring methods are based on detecting the attenuation and scattering of optical signal, attenuation and retardation of microwaves, and changes in mechanical viscosity of pulp slurry. In all these methods, the calibration of the measurement devices is difficult and in a key position. Furthermore, they can not simultaneously give rise to the optical and acoustic properties of materials.

Recently, the inventors have used laser radar and ultrasonic attenuation techniques to measure the fibre and fines consistencies in paper pulp. The results show that the optical propagation delay is more sensitive to the fines content and acoustic attenuation is more correlative to the fibre consistency. However, the devices of both techniques should be used simultaneously during measurement. The main drawbacks are inconvenient install of both apparatus in one location of online measurement and higher devices cost.

It is well known that optical propagation delay in turbid material is closely related to the optical scattering property of the material. The problem is how to detect the optical scattering and acoustic attenuation simultaneously in convenience and by low cost. By measuring both properties it would be possible to determine the fines and fibre consistencies in pulp suspensions.

Summary of the invention

An object of the invention is to avoid above-described problems of the prior art and to provide a new, cost effective measuring method and apparatus for measuring optical scattering and absorption and acoustic attenuation properties of a weakly- absorbing material simultaneously, by utilizing photoacoustic phenomenon.

The objects of the invention are achieved by a method where a low-energy laser pulse is directed to a weakly-absorbing material, which is to be analyzed. The Ia- ser pulse is partly absorbed in the material causing a photoacoustic signal and partly scattered. A part of the scattered photons is received by an absorber which is in contact with the studied material producing another photoacoustic signal in the absorber. From amplitudes of the photoacoustic signals can be studied what are the scattering and absorption properties of the material. The measuring method according to the invention is called scattering photoacoustic (SPA) technique in this application.

An advantage of the invention is that the optical absorption and scattering coefficients of a material can be deduced simultaneously by a one-detector arrange- ment.

Another advantage of the invention is that when analyzing weakly-absorbing, highly turbid medium there are two photoacoustic pulses produced by medium, i.e. absorption and scattering, which can be analyzed utilizing detected signal amplitudes.

A further advantage of the invention is that only a low-energy laser and an acoustic detector is needed in the measurement.

A further advantage of the invention is that it reduces transducer bandwidth requirements when compared with TRSD.

A further advantage of the invention is that it can simultaneously study optical and acoustic properties of the analyzed material, simply by applying another same absorber. An example is to determine fibre and fines consistencies of pulp.

Yet a further advantage of the invention is that the measurement arrangement is simple and low cost for on-line apparatus in pulp measurement.

The characterising features of the invention are disclosed in the independent claims.

The idea of the invention is as follows. Huge amounts of photons scatter out from the laser illuminated region in the highly turbid material. These scattering photons also record optical properties of the studied material. These photons can advantageously be received by at least one disk absorber with high absorption coefficient at study wavelengths. Therefore, a photoacoustic source is produced also at the reception surface of the absorber. This is called scattering photoacoustic source.

In a normal case, when a pulse laser illuminates a turbid material, above- mentioned photoacoustic source PA will be produced in the studied material and scattering photoacoustic source SPA at one disk absorber respectively. An acoustic detector is used to receive the signals from both sources, successively. These signals strongly relate to optical absorption and scattering proprieties of the material extracting its absorption and reduced scattering coefficients.

If two absorbers of same kind are used and they are symmetrically arranged with incident light beam, they will generate two photoacoustic sources with same intensities. If an acoustic detector contacting with one of the absorbers receives both acoustic waves, the ratio of both signal amplitudes is proportional to the acoustic attenuation of the material; because one wave goes through the material but an- other does not before they are received. Therefore, SPA technique is capable of monitoring the optical scattering, absorption, and acoustic attenuation of materials simultaneously. It is a combination of optical and acoustic methods.

Brief description of the drawings

The invention is described in detail below. Reference is made to the accompanying drawings in which

Fig. 1 a shows a schematical representation of a measurement arrangement according to the first embodiment of the invention;

Fig. 1 b shows a schematical representation of a measurement arrangement according to the second embodiment of the invention;

Fig. 1 c shows schematically received photoacoustic signals according to the second embodiment of the invention;

Fig. 2 shows an exemplary measurement arrangement according to the invention;

Fig. 3a shows an example of received signals when a PVDF transducer is utilized;

Fig. 3b shows an example of received signals when a PZT transducer is util- ized;

Fig. 4 shows the correlation between V _S p _A and intralipid concentration;

Fig. 5 shows a graph of the equation (6), where μ _a = 0.01 15 mm ^"1 ;

Fig. 6 shows ln(rV)-r lines measured in intralipid samples by the PZT transducer;

Fig. 7 shows ln(rV)-r lines measured in intralipid samples by the PVDF transducer;

Fig. 8a shows measured relationship between scattering and V _PA in ink- intralipid mixes;

Fig. 8b shows measured relationship between scattering and V _SPA in ink- intralipid mixes;

Fig. 9a shows measured relationship between absorption and V _PA in ink- intralipid mixes;

Fig. 9b shows measured relationship between absorption and V _SPA in ink- intralipid mixes;

Fig. 10 shows measured signals produced in three pulp samples;

Fig. 1 1 a shows measured amplitudes of V _spa i, versus fines concentration;

Fig. 1 1 b shows measured amplitudes of V _pa , versus fines concentration;

Fig. 1 1 c shows measured amplitude ratios of V _S p _a2 /V _S p _a i, versus fines concentration;

Fig. 12a shows measured amplitudes of V _spa i, versus fiber consistency;

Fig. 12b shows measured amplitudes of V _pa , versus fiber consistency;

Fig. 12c shows measured amplitude ratios of V _spa 2/V _spa i, versus fibre consistency;

Detailed description of advantageous embodiments of the invention

Figure 1 a depicts schematically an advantageous embodiment of the present invention. A turbid sample 12 is loaded in a cuvette 10, which is illuminated by a pulsed laser beam 1 1. A part of optical energy is absorbed by the sample 12. This produces a photoacoustic source 16a in the turbid sample, which photoacoustic source 16a produces acoustic waves 13.

Part of photons of the laser beam are scattered out in the sample, reference 19a. In a first advantageous embodiment of the invention the scattered photons 19a are absorbed by an absorber 14. This produces photoacoustic source 16b on a surface of the absorber 14. By an arrow 19c is depicted that an acoustic wave propagates from the acoustic source 16b through the absorber 14 to a detector 17.

For weakly-absorbing highly scattering medium, the shape of photoacoustic source 16a is approximately spherical-like shape. Following thermo-elastic stress generation in the spherical model, measured signal amplitude can be described by

where k is a constant, r is the measurement distance, β is the expansion coefficient, v \s the acoustic velocity and C _P is specific heat of the studied medium. R _a is the source size in the direction of signal reception. E _a is the absorbed light energy in source, which is proportional to the absorption coefficient μ _a .

In highly turbid materials, optical transport follows the diffusion theory. If the energy of a spot source in an infinite medium is E and the pulse duration is longer than the order of 1 ns, the energy fluence can be determined by steady-state diffusion theory

F(r) = -ζ- ■ exp(-rJμ _α /D) (2)

where D is the diffusion constant, r is the distance from light source to observer and μ _a is the absorption coefficient of the turbid medium. If a laser beam 1 1 irradi- ates on a highly turbid sample 12, it approximately can be seen as a spot source by the detector 17. On the other hand, a part of incident optical energy will be diffusely reflected from the sample surface, references 19a and 19b, causing the effective optical energy of the spot source in sample approximately to be

E = (I - R^ )E ₁ . (3)

Here E/ is the laser pulse energy and Rj _∞ is the total diffuse reflectance which could be calculated as

where μ _s ' is the reduced scattering coefficient. In the case of highly turbid sample D»(3μ _s ') ^~ι is true because of μ _s ' » μ _a .

The absorber 14 has very high absorption coefficient in study wavelengths and the shape of photoacoustic source 16b is plane-like one. Therefore, the PA signal amplitude generated by photoacoustic source 16b is described by

where a is the absorption coefficient, β is the expansion coefficient, v' is the acoustic velocity and C _P ' is the specific heat of the absorber material and k' is a constant. F is the energy fluence (energy per unit area) at the surface of absorber 14. In equation (5), it is assumed that all scattering photons hitting on the absorber 14 contribute to produce PA signal.

Substituting above equations (2)-(4) into (5), V _SPA becomes

where K is a new constant:

Multiplying (6) by rand then taking the logarithm, it has

\n(rV _SPA ) = -pμ _a μ _s ' ^■ r + ln{4 - exp(- iJμjξμTfy, '} ( ₈ )

Hence, if K is known, linearly fitting the measurement points, μ _a and μ _s ' can be de- duced from the slope and intercept of the \r\(rV _S p _A )-r relationship. K can be calibrated out by a sample with known absorption and scattering coefficient.

In turbid suspension measurement, there are two interested special cases. If absorption coefficient μ _a of a suspension is near equal to zero at some wavelengths, equation (6) becomes

meaning that reduced scattering coefficient μ _s ' can be determined by scattering photoacoustic signal.

Another case is that if the scattering particles have little effect on the matrix absorption, μ _s ' can be calculated out directly from the slope of line if the absorption of matrix is known before.

In Fig 1 b, which is a second advantageous embodiment of the invention, there are two absorbers, references 14 and 15, which are located on the opposite side walls of the cuvette 10. In the example of Fig 1 b the absorbers 14 and 15 are equally located compared to the laser beam 1 1. Part of photons of the laser beam 1 1 are scattered out in the sample, references 19a and 19b. Therefore, three photo- acoustic sources are produced. The first one is in the turbid sample, reference 16a. The second source, reference 16b, is in the first absorber 14. And the third source, reference 16c, is in the second absorber 15, which is located on the other side wall of the cuvette 10. The third acoustic source 16c produces an acoustic wave 18.

These three acoustic sources according to the second embodiment of the invention generate three acoustic wave pulses, depicted in Fig 1 c with V _pa , V _spa i and V _spa 2, which are received by the acoustic detector 17, which is in contact with the absorber 14. According to the time order detected by absorber 17, V _spa r can be called a first acoustic signal, V _pa can be called a second acoustic signal and V _spa2 can be called a third acoustic signal.

In Fig 1 c V _spa1 , depicts a signal generated by the photoacoustic source 16b on the first absorber 14. V _spa2 , depicts an acoustic signal 18 generated by the photo- acoustic source 16c on the second absorber 15. Signal V _pa depicts an acoustic signal 13 generated by the photoacoustic source 16a in the turbid media 12.

If the photoacoustic source 16a in the turbid media is cylindrical-like, V _pa can be expressed as

(1 0)

where μ _a is absorption coefficient of the sample, R is the radius of the cylindrical source, E is the pulse laser energy in the source, and r ₀ is the distance between the source 16a and the detector 17.

On the other hand, if the photoacoustic source 16a is spherical-like, V _pa is described by equation (1 ), rewritten below:

(1 1 )

Equations (10) and (11 ) can be used to qualitative describe the photoacoustic signal amplitude produced in scattering suspensions with weak absorption. If the scattering concentration is very low, the photoacoustic source 16a is cylindrical- like and its radius is near equal to incident laser beam. When the concentration in- creases, more and more scattering photons are scattered out from the source beam, causing an increase of source radius and a decrease of energy density in the beam. Therefore, based on equation (10), V _pa will decline accordingly. Along with the concentration increases, the source radius continues to increase but the length of source is shortened, forming the spherical-like source. If the concentra- tion further increases, such that strong multi-scattering happens, many photons will be scattered back into the source and absorbed at there, avoiding R _a further increasing. This effect may cause V _pa decreasing slowly or almost unchanged, according to equation (11 ).

The photoacoustic sources 16b and 16c produced on two absorbers' reception surfaces are plane-like, because the absorbers 14 and 15 are highly absorbing material. The signal amplitude can be described by equation (6). If μ _s ' is not too high to satisfy

rj3μ _α μ _s ' < L26 , (12)

V _spa increases monotonously with μ _s '\ otherwise, V _spa will be decreased with higher μ _s '.

In Fig 1 b, if the laser beam 11 is incident along the path with same distance from the two absorbers 14 and 15, both acoustic sources 16b and 16c produced by the absorbers should have same intensity. However, if the sample is acoustic attenuating, the acoustic wave 18 emitted by the second absorber 15 will be attenuated by the sample before accepted by the detector 17, whereas the acoustic wave pro- duced by absorber 14 is not attenuated. Hence, the ratio of signal amplitudes V _spa 2/V _spa i can be used to evaluate also the acoustic attenuation of the sample.

Therefore, based on measuring V _pa , V _spa1 , V _spa2 and calculating V _spa2 IV _spa1 , it is possible to evaluate the optical scattering, optical absorption and acoustic attenua- tion of the sample and therefore to measure its composition.

Figure 2 shows an advantageous embodiment of a measurement arrangement according to the first embodiment of the invention.

In Fig 2, a diode pumped Q-switched Nd:YAG laser (model LCS- DTL- 1 12QT, LASER-COMPACT Co. Ltd) is used as exciting source, reference 21. The duration of each light pulse is about 14 ns. The output pulse energy is 2 μJ at 1064nm and 1 μJ at 532 nm wavelengths, respectively. The laser beam 1 1 is projected on the incident window of a cuvette 10 where the diameter of the light spot is focused about 0.6 mm. The cuvette has thickness of 30 mm, width of 30 mm and height of 40 mm. A small hole and a valve located on the bottom of cuvette are used to replace the samples without moving the cuvette.

A prism 23 after lens 22 separates the spots with different wavelength. A filter 28 chooses the illuminated wavelength and a stop 25 is used to keep away the remnant pumping light. A detector 27, comprising a piezoelectric ultrasonic transducer connected directly with a preamplifier, is acoustically contacted with an absorber 24 which is embedded in a side wall of the cuvette 10. The thickness of the ab- sorber 24 is 3 mm and its diameter is 8 mm. On the surface of the absorber 24 the scattered photons produces an acoustic source 26.

Two kinds of piezoelectric transducers are advantageously used. One is made of PVDF polymer film (Polyvinylidene fluoride) with thickness of 52 μm and active di- ameter of 5 mm. Another is made of PZT ceramic disk (Lead Zirconate Titanate) with both thickness and diameter of 4 mm. The device diameters of both transducers are 6 mm and 10 mm, respectively.

A preamplifier included in the detector 27 has a gain of 40 dB and a bandwidth from 150 kHz to 3 MHz. A translation stage (not shown in Fig 2) is used to change the transversal distance between the exciting light spot 11 and the detector 27. The output from the detector is connected to a digital oscilloscope 29 to display and store the data. Because PVDF transducer has lower response than PZT one, another amplifier with 60 dB gain (not shown in Fig.2) is added if necessary.

The measured and stored photoacoustic data can advantageously be conveyed from the digital oscilloscope 29 to a computer (not shown in Figure 2) for further data processing.

An example of the first embodiment:

In an example of the first advantageous embodiment of the invention, the scattered photons 19a are absorbed by one absorber 14. 10%-lntralipid ^® suspension (Fresenius Kab. AB., Uppsala, Sweden) is used as a scattering material 12. The diameter of 99% scattering particles in the suspension is in the range of 25 nm to 500 nm. The absorption coefficient μ _a of suspension is low, approximately equal to that of pure water in near infrared wavelength range. It is a popular tissue scattering phantom in biomedical research. The absorption material used in this experi- ment is a kind of Chinese ink, which is added into the diluted suspensions to adjust the absorption property of the samples. The scattering effect of the ink is too small to be observed at the study wavelength.

10%-lntralipid ^® suspension is diluted by different amount of pure water to make the samples. These samples are loaded successively, from low concentration to high, into the cuvette 10. The light with 1064 nm wavelength is chosen as exciting source. Typical received signals by PVDF and PZT transducers, which are displayed on the digital oscilloscope, are shown in Fig 3a and Fig 3b. Signal V _PA is produced in the material (Fig 1 a reference 16a) and signal V _SPA is in the acoustic source (Fig 1 a reference 16b) connected to the absorber 14.

The measured relationship between V _SPA and intralipid concentration is shown in Figure 4, whereas the theoretical forecast is illustrated in Figure 5. It can be seen that they are identical very well, because the reduced scattering coefficient μ _s ' is proportional to intralipid concentration (when the concentration is lower than 2%). It can be seen that V _SPA quickly rises at first, and then gradually falls with concentration increasing. This is due to that water matrix has a definite absorption (0.0115 mm ^"1 ) and more multiple scattering events happen in the higher concentration samples. At a definite range of concentration from 0.05% to 0.3% in Fig 4, there is an approximately linear relationship between V _SPA and the reduced scattering coefficient μ _s '.

In near-infrared wavelengths, the absorption coefficient of the intralipid samples is approximately equal to that of water. Therefore, in agreement with Equation (8), their reduced scattering coefficients can be deduced by measuring the slope of the ln(rV _S pA)-r line. Figs. 6 and 7 present the results of such measurements obtained with the PZT and PVDF transducers. The estimated error at any measurement point is no larger than the marks shown in the figures. As the Figures 6 and 7 illus-

trate, all samples exhibit a linear relationship between ln(rV _S pA) and r. Table 1 shows the deduced μ _s ' of the samples.

Table 1. Measured values of the reduced scattering coefficient of intralipid sus- pensions with different concentrations at 1064 nm

Concentration of intralipid 0.1 % 0. 2% 0. 5% 1 % 2 % 5 % ............ __ _ __ __ .............. __ _ __ μ _s ' (mm ¹ ) by PZT transducer ^"""" 67TT ^"""" 67 ^"""" μ _s ' (mm ¹ ) by PVDF transducer 0.1 0 .16 0 .36 0 78 1. 39 μ _s ' (mm ¹ ) by PIN diode 0 .12 0 .37 0 73 1. 40 2 .4 μs' (mm ¹ ) by 0.07 0 .13 0 .34 0 67 1. 34 3. 35 van Staveren et al.

For comparison, Table 1 also lists results from our optical measurements as well as values deduced from van Staveren's results (Appl. Opt. 30, 4507-4514, 1991 ). It reveals that the results are in good agreement, except for 5% intralipid. This is because van Staveren's measurements used lower intralipid concentrations. As for high density samples (>2%), μ _s ' does not increase linearly with intralipid concentration (Bondani, J. Opt. Soc. Am. B20, 2383-2388, 2003); therefore, the result deduced from van Staveren is not valid. Since the transducer's diameter is larger than the PIN diode used in our measurement, the misalignment between the transducer and the light spot is greater for high density intralipids. As a result, the transducer records smaller values than the PIN diode.

To investigate how V _PA and V _SPA vary with the absorption and scattering properties of samples, an absorption material, namely, blue Chinese ink, was added to in- tralipid suspensions of different concentrations. However, because the ink may produce a scattering effect at 1064 nm, this part of the study used light at 532 nm as the excited source. At this wavelength, an aqueous solution of ink exhibited no scattering, and its absorption coefficient, measured by the optical transmission method, was found to equal 0.018 mm ^"1 at the 0.04% concentration level.

Figure 2 illustrates the measurement, in which the incident laser beam was at a distance of 16 mm from the absorber's reception surface. Firstly, a 0.04% ink solution was loaded into the cuvette, followed by successive injections of different amounts of 10%-intralipid. As the intralipid concentration increased, the reduced scattering coefficient of the ink-intralipid suspension changed accordingly. Figures 8a and 8b record the signal amplitudes of V _PA and V _SPA - TO obvert the absorption of the signals, the sample in the cuvette was replaced by a 1 % intralipid suspen-

sion, and different amounts of ink solution were successively added into the cuvette. This caused an absorption increase in the sample. The corresponding changes in signal amplitudes of V _PA and V _SPA are shown in Figures 9a and 9b.

It can be seen from Figure 8a that V _PA decreases steeply when the reduced scattering coefficient is less than 0.2 mm ^"1 . At higher values (>0.5 mm ^"1 ), V _PA plateaus as the reduced scattering coefficient varies, which is a very useful characteristic, as shall be demonstrated below.

The variation in V _SPA in Figure 8b is similar to that shown in Figure 4, measured in intralipid suspensions at 1064 nm. Moreover, Figures 9a and 9b show that, when the concentration or absorption coefficient of ink increases, V _PA increases linearly, while V _SPA decreases very quickly and approaches zero.

Thus, in weakly absorbing and highly turbid samples, μ _a and μ _s ' can be measured using both V _PA and V _SPA - TO illustrate this, we used two samples. Sample 1 contained 35 ml of 1.5%-intralipid mixed with 20 μl of ink; thus, the ink concentration of the sample was 0.057%. Sample 2, on the other hand, consisted of 35 ml of 0.5%-intralipid with an ink concentration of 0.067%. The effective attenuation coef- ficients μ _eff (same as the slope of ln(rV _S pA)-r line) of the two samples were 0.3655 and 0.2527, respectively, measured by the method described above. As for the samples' absorption coefficient, they can be deduced as below.

According to Figure 9a, V _PA has a linear relationship with μ _a (or ink concentration). Although measured in 1 % intralipid-ink mixes, it can be assumed that the relationship holds for suspensions with intralipid content higher than 0.3%. In this case, V _PA remains unchanged even if the intralipid content varies, as shown in Figure 8a. Hence, by measuring the samples' V _PA , we may deduce their absorption coefficients from Figure 9a. In the end, μ _s ' can be calculated from μ _eff and μ _a . The re- suits are shown in Table 2. For comparison, table 2 lists in brackets the predicted absorption coefficient for ink and the reduced scattering coefficient for intralipid. It can be seen that the measured values are almost identical with the predicted ones. So in conclusion, the experiments described above demonstrate that the SPA method allows the direct deduction of the absorption and reduced scattering coefficient of highly turbid materials.

Table 2. Measured optical coefficients of two samples at 532 nm; the values in the brackets indicate predicted values.

An example of the second embodiment:

In an example of the second advantageous embodiment of the invention the scattering photoacoustic technique is utilized in pulp measurement. Under the aid of two absorbers (references 14 and 15 in Fig 1 b) and one acoustic transducer 17 we can measure three photoacoustic signals V _spa1 , V _pa and V _spa2 simultaneously. With these three signals it is possible to simultaneously study the forward optical scattering, transverse optical scattering, and acoustic attenuation of pulp suspension. In this example the scattered photons 19a and 19b are absorbed by two absorbers 14 and 15.

Samples to be tested are produced from Thermo-mechanical pulp (TMP). The TMP was fractionated by Bauer-McNett fractionator according to SCAN-standard 6:69. The finest fraction, passing through the 200 wire mesh was then subsequently filtered through 400 wire mesh. This results the fine fraction consisting of particles with sizes varying between 30 and 74 μm. The long fiber fraction used in this study consists of the fibers that passed through the 48 wire mesh from TMP. The length of fibers is about 1-3 mm, diameter is in the order of 10s micron. After separating of fibers and particles to different classes, the fractions were mixed in different proportions, resulting in a range of fines and fiber contents in samples. For convenient description, the pulp sample is named as x%_y%, meaning it consists of x% fibers and y% fines.

The measurement system is depicted in Fig 2 with an exception of the second absorber 15 of Fig 1 b, which is added to the arrangement of Fig 2.

The typically recorded signals are shown in Figure 10, where there are three signal curves produced by samples 0.5%_0, 0.5%_0.125% and 0.5%_0.25%, respectively. For every signal, there are three pulses with amplitudes of V _spa1 , V _pa and V _spa2 , produced in absorber 14, the sample 16a, and absorber 15, respectively.

The time delay between the acoustic pulses is depended upon the distance between them and acoustic speed in the sample 12. It can be seen that, when the fines content increases, V _spa i, V _pa and V _spa 2 will be changed.

Because fines and fibre have different physical properties, they will affect on the signal amplitudes with different mechanism. Figures 1 1 a-c give raise to the values of V _S p _a i, V _pa , and to an amplitude ratio V _spa2 IV _S p _a1 produced in pure fines. Correspondingly Figures 12a-c give raise to the values of V _spa1 , V _pa , and to an amplitude ratio V _spa2 IV _spa1 produced in pure fibers.

In Figures 1 1 a-c, it can be seen that V _spa i increases, whereas V _pa decreases with concentration increase. However, in higher concentration, V _spa i decreases, but V _pa almost keeps unchanged when fines concentration rises. In all study concentrations, ratio Vs _pa2 /V _spa i is approximately unchanged with fines concentration, meaning that the acoustic attenuation produced by fines can be neglected.

On the other hand, in pure fibre samples (Figures 12a-c), V _spa1 increases mono- tonically while both V _pa and V _spa2 IV _spa1 approximately linearly falls with fibre consistency increase. These mean that fibres produce acoustic attenuation apparently, but they cause smaller transversely scattering comparing with fines.

Considering that the pulp samples are not homogeneous, especially for the samples with higher fibre consistency, every sample is therefore measured a few times and the average value and standard deviation are calculated to distinguish the sample with different fines and fibre content. Tables 3-5 give rise to these values of Vsp _a i, V _pa and V _spa2 IV _spa1 for all pulp samples.

Table 3. Average values (standard deviations) of V _spa i

Table 4. Avera e values standard deviations of V.a

Based on Tables 3-5, it is possible to determine the fines concentration and fibre consistency in a pulp sample. For instance, if one sample is measured and the average values of its V _spa1 , V _pa , and V _spa2 IV _spa1 are 0.45 V, 0.44 V, and 0.57 V, respectively, its fibre consistency should be in the range of 0.5% to 0.75% deduced from Table 5. Its fines concentration should be about 0.125%, according to Table 3 and Table 4. However, in the case of high fibre consistency (>0.75%), the standard deviations in Tables 3-5 are so big that they may affect on distinguishing samples. The main reason is the worse homogeneity due to fibre flocks in samples. The effect of fibre flocks can be reduced by measuring the pulp samples in the flowing case.

It is possible that fibre flocks may lead different optical and acoustic properties in different location of a sample. The processes of fibre flocks forming and fines attaching or depositing may cause a continuous variation of the sample property. These factors result an apparent measurement deviation and uncertainty. To alleviate these effects, it is also possible to reduce the fibre flocks by flowing or suitably stirring the pulp suspension during measurement. Moreover, increasing the measurement times and averaging the results will improve the ability of sample identification.

Some advantageous embodiments according to the invention were described above. The invention is not limited to the embodiments described. The inventional idea can be applied in numerous ways within the scope defined by the claims attached hereto.