Industrial Health 2003, 41, 249259

Original Article

Determination of Free Silica in Dust Particles: Effect of Particle Size for the X-Ray Diffraction and Phosphoric Acid Methods
Juji YABUTA* and Hisayosi OHTA
Department of Health Administration, School of Allied Health Science, Kitasato University, 1151 Kitasato, Sagamihara, Kanagawa 228-8555, Japan Received June 6, 2002 and accepted May 9, 2003

Abstract: The X-ray diffraction method and the phosphoric acid method are widely used to determine the fraction of free silica (mainly quartz and other silica polymorphs) in respirable dust sampled in working environments in Japan. In this study, we clarified the size effect of quartz dust for the Xray diffraction method and the phosphoric acid method using size controlled quartz samples. The quartz samples were classified into 6 fractions with different size ranges: 1 m and smaller, 1 to 3 m , 3 to 5 m , 5 to 7 m , 7 to 10 m and 10 m and larger. Both of the determination methods were affected by the particle size, and especially particles smaller than 3 m fairly dissolved in hot phosphoric acid and reduced X-ray diffraction intensity remarkably. If the content of these fine particles in the standard quartz sample is lower than that of the test samples, the fraction of free silica may be underestimated by these methods. For this reason, the standard quartz sample should have a representative size distribution of the field samples. The dust samples containing quartz were collected at a foundry and dissolved by phosphoric acid to remove non-quartz materials. The size fractions of dissolved samples were 50% for 510 m, 25% for 35 m, 20% for 13 m and 5% for 1 m and smaller. As the size distribution is similar to the present standard sample widely used in Japan, we concluded that the standard sample is suitable for these determination methods. Key words: Quartz, Free silica, Phosphoric acid method, X-ray diffraction method, Particle size, Field sample

Introduction
The first step for working environment management of handling harmful substances is to collect information for evaluation of the environment. For this purpose, working environment measurements on specified substances are conducted in Japan1). Airborne dust containing free silica in mines, foundries, potteries and so on is an important substance controlled by legislation. In a workplace with dust suspensions, not only the measurement of the dust concentration but also the determination of the free silica

*To whom correspondence should be addressed.

fraction in the dust are required, because the health risk (mainly silicosis) depends on the mass of free silica (quartz, tridymite and cristobalite) in the dusts. In the Working Environment Measurement Standard2), the X-ray diffraction method and the phosphoric acid method are stipulated as the determination methods for the concentration of free silica in dusts. The absorption collection X-ray diffraction method using a metal substrate standard is widely used for determining free silica in Japan. Although the NIOSH method 75003) employs a silver membrane filter and is very accurate, the silver membrane filters are very expensive and occasionally not available in Japan. Kohyama developed an X-ray diffraction method using a metal substrate standard (zinc or aluminum plate) instead of the silver

250
membrane filter and this technique has been accepted as the Japanese standard4). In Japan, the phosphoric acid method is also widely used for the determination of free silica because of its low-cost. The principle of the phosphoric acid method uses the difference in the solubility of quartz and that of other minerals in hot phosphoric acid. The method for free silica determination was developed by Talvitie5) and has been studied by many researchers in Japan610). The reliability of both methods depends significantly on the particle sizes in the sample. A standard sample is used for calibration in the X-ray diffraction method. In the phosphoric acid method, it is used for correcting the solubility of quartz because some of the quartz is also digested by hot phosphoric acid. The difference in particle size composition between a standard quartz samples and field samples may cause determination error. For example, the intensity of Xray diffraction depends on the particle size of quartz11). Therefore the standard quartz sample should have not only pureness but also a size distribution representative of field samples. In this study, we focused on the size effect of quartz particles in the X-ray diffraction method and the phosphoric acid method. The X-ray diffraction intensity and the solubility in hot phosphoric acid were measured using size classified quartz samples. The size distributions of quartz particles in field samples were also measured and compared with that of the standard quartz sample.

J YABUTA et al.

Fig. 1. Schematic view of liquid sedimentation device used for size classification.

Methods
Preparation of size-classified quartz samples Quartz powder produced in India (product name: Crystallite VX-S) was sieved through a 200-mesh standard sieve. Then the under 200-mesh powder was treated successively by acid and alkali for a few hours each to leach away impurities in the powder. After the powder was washed with distilled water, it was placed in a liquid sedimentation device (Fig. 1). About 6 g of sample were placed in it and distilled water was added to suspend them. The settling time of particles was calculated from Stokes law as shown below, t ={18 108 / (1 2) G d2} L (1).

t : Settling time (s), : Viscosity coefficient of the medium (g/cms), 1: Density of the powder (2.60 g/cm3), 2: Density of the medium (g/cm3), G : Gravitational acceleration (cm/s2),

d : Particle diameter (m), L : Settling distance of particles in time (10 cm in our device). After the settling time of 1 m particles had passed, the water suspension above the cock was collected by opening the cock screw. Then distilled water was added to the device and was shaken for re-suspension. The same procedures were repeated until the water above the cock became almost clear. All of the collected water were filtrated by a membrane filter of 0.8 m pore size and then the filter was dried to recover 1 m and smaller particles. A similar procedure was conducted using the rest of quartz samples settled on the bottom of the device and the next separated fraction was named as 13 m quartz sample. Successive operations to prepare 35 m, 57 m and 10 m and larger samples were conducted to obtain 6 groups of size classified quartz samples (Fig. 2). In this experiment, the settling time at 20C was 32 hours 2 minutes 37 seconds for 1 m, 3 hours 33 minutes 37 seconds for 3 m, 1 hour 16 minutes 54 seconds for 5 m, 39 minutes 14 seconds for 7 m, and 19 minutes 14 seconds for 10 m for our device. The particle size distributions of these 6 samples were further measured using a centrifugal settling type particle size analyzer (Shimadzu SA-CP3L, Kyoto). The results were expressed in terms of frequency of sample weight in each

Industrial Health 2003, 41, 249259

FREE SILICA IN DUST PARTICLES

251

Fig. 2. Procedure of size classification.

particle size range. The measuring ranges were 0.1 to 40 m for the 1 m and smaller sample, 0.6 to 40 m for the 13 m sample, and 1 to 40 m for the other samples. X-ray diffraction method To know the X-ray diffraction intensity of the 6 groups of quartz sample, known amounts of each group sample were suspended in a volumetric flask with water and part of the suspension was filtered on a Teflon-bound glass fiber filter (Palflex T60A20; 25 mm diameter). Three different amounts of sample, corresponding to 0.5, 1.0 and 2.0 mg/cm2 of quartz on the filter, were prepared for the X-ray diffraction intensity measurements. The (101) quartz peak was used to obtain peak intensity data. The filter sample was mounted on a rotational sample attachment of X-ray diffraction apparatus and scanned from 29 to 25 degrees (Cu and Kd radiation)

and the peak height of the (101) peak at 26.6 degrees was measured. The measured intensities were treated both with the absorption correction and without the correction. The absorption effect due to the differences of the sample mass (or matrix) was corrected by the metal substrate correction method using a Zn plate 4). The correction factor was calculated from an equation similar to that of NIOSH method 7500, except for the Ag angle, and applied to the measured intensity data. After the correction, the quartz intensity was recalculated to the intensity per unit mass. The intensities without the absorption correction (direct method) were also treated in the same manner. Phosphoric acid method Single quartz sample of one size range: In order to

252

J YABUTA et al.
Table 1. Mixing ratios of quartz samples of respective particle size ranges in respective mixed samples No. Sample name 1 m and smaller 1 2 3 A B C 30 20 10 13 m 30 20 10 Mass percent (%) 35 m 20 20 20 57 m 10 20 30 710 m 10 20 30

In the table, 10 corresponds to a mass of about 20 mg

Table 2. Workplace names and work descriptions in a foundry from which samples were obtained No. 1 2 3 4 5 6 Workplace name Weighing Pouring New sand Waste sand Raw material of casting sand Sand control Work description

Process of weighing raw materials used in cupola molten Pouring molten iron into a sand mold using a ladle Process of drying of sand (two kinds of silica sand) used for casting mold Recovery of sand for reuse Process of grinding, grading and drying sand mixtures described in 3 and 4 Process of recovery of sand after shake-out, removal of scrap iron and addition of additives 7 Molding Process of sand mold molding using a feeder 8 Core making Sand near a core making machine 9 Shake-out (oscillating conveyor) Process of dropping the sand mold after pouring on an oscillating conveyor for separation of sand and casting 10 Dismantling Process of mechanical dismantling; separate room 11 Finishing (shot blast) Sand mold shake-out using a shake-out machine, and grinding with shot blast 12 Finishing (shot blast) Grinding with shot blast 13 Finishing (housing) Housing work 14 Finishing (hanger blast) Mostly the dust inside a casting 15 Finishing (cylinder block) Grinding using cylinder block

determine the residual ratio of the size classified quartz sample, phosphoric acid digestion of the sample was conducted in a conical beaker with a short-stemmed funnel and an electric heater. The procedure2) was modified the Talvits method5) and optimized for the analysis of small amounts of respirable size particles. The details are as follows. Two hundred milligrams of size-classified quartz samples were precisely weighed and put into the beaker, then 15 ml of phosphoric acid was added. After the beaker was covered with a bent-stem funnel, it was placed on an electric heater applied 89 V and 2.7 A. After 6 minutes, swirling for 5 seconds by hand was started, and this short swirling was repeated every minute during heating. After 11 minutes heating, the beaker was removed from the heater and left to stand until the solution reached room temperature. After cooling, hot water was poured into the beaker which was swirled vigorously. Then diluted fluoboric acid solution was added. The whole content of the beaker was filtered using a membrane filter of 0.8 m pore size. The filter with

residue was put into a pre-weighed porcelain crucible and incinerated at 800C using an electric furnace. After incineration and cooling, the crucible was weighed. This was repeated until a constant weight was obtained. The residue of the phosphoric acid dissolution, i.e. quartz weight in sample, was obtained from the weight difference of the crucible before and after incineration. We can calculate the ratio of phosphoric acid dissolution (=residue of the phosphoric acid dissolution/ sample weight at first, %) Preparation of a mixed quartz sample with different particle size ranges: To simulate an actual sample, we prepared three samples having different particle size distributions by mixing the size-classified samples previously obtained as described above. Known amounts of the size classified quartz samples were mixed to be totally 200 mg. The mixing ratios of the three samples are shown in Table 1. For samples A to C, the proportion of fine particles in the sample decreases in the same order. These samples were processed by the phosphoric acid method as described above.

Industrial Health 2003, 41, 249259

FREE SILICA IN DUST PARTICLES

253

Fig. 3. Size distributions of size-classified quartz samples.

Measurement of size distribution of quartz particles in field samples Settled dust samples at fifteen workplaces were collected to obtain information on the size distribution of quartz particles in field samples. Table 2 describes the workplace names and the type of work which took place in them. Analytical samples of particle size smaller than 10 m were prepared using the liquid settling method similar to the quartz sample preparation. These size adjusted field samples were digested by hot phosphoric acid to remove materials other than quartz, and the residue was regarded as quartz. The obtained residues were analyzed for their size distribution using a particle size analyzer. Due to the dissolution by hot phosphoric acid, the residual quartz particle size may be

smaller than that before digestion. However, since it is considered that the dissolution proceeds constantly from the surface of each particle, the size distributions are proportional to each other both before and after the phosphoric acid digestion.

Results and Discussion

Particle size distribution of size classified quartz samples The size distributions of size-classified quartz samples are shown in Fig. 3. The modal diameters of the respective samples were 0.60.8 m (14.8%) for the 1 m and smaller sample, 1.52 m (38.3%) for the 13 m sample, 23 m (38.0%) for the 35 m sample, 56 m (39.8%) for the 5

254

J YABUTA et al.

Fig. 4. Comparison of the normalized X-ray intensity ratios in the direct method at different sample concentrations.

Fig. 5. Comparison of the normalized X-ray intensity ratios in the X-ray diffraction method using a metal substrate standard at different sample concentrations.

7 m sample and 810 m (31.2% of total) for the 710 m sample. The fraction within the specified particle size range of each group was 50.5%, 75.0%, 50.3%, 47.3%, and 41.8% for 1 m and smaller, 13 m, 35 m, 57 m and 710 m samples, respectively. Thus, almost half of the particles were within the nominal size range. Although there are few studies with which to compare the size distribution of quartz

samples prepared by the sedimentation method with other size analysis methods, these quartz samples were classified fairly well into the nominal size range except for the 1 m and smaller sample. This exception may be due to the fact that it is difficult to obtain particles smaller than 1 m using the ordinary sedimentation method.

Industrial Health 2003, 41, 249259

FREE SILICA IN DUST PARTICLES

Table 3. Relative X-ray intensities of the quartz samples for respective particle size range Direct determination method Sample 1 m and smaller 13 m 35 m 57 m 710 m 10 m and larger 0.5 (mg/cm2) 610 625 781 793 819 905 1.0 (mg/cm2) 552 540 692 731 734 824 2.0 (mg/cm2) 458 460 550 580 592 642 (cps/mg)

255

X-ray diffraction method using metal substrate standards 0.5 (mg/cm2) 641 678 839 882 894 941 1.0 (mg/cm2) 612 626 803 819 831 935 2.0 (mg/cm2) 611 618 780 805 840 910

Fig. 6. Ratio of the relative X-ray intensities of quartz samples of respective particle size ranges.

X-ray diffraction method Relative X-ray intensity and sample mass on the filter: The sample collected on the filter is regarded as a thin powder layer. In such a case, X-ray intensities are affected also by the degree of absorption of the sample itself. Thus, change of the X-ray intensity with the difference of sample amounts on the filter was examined first. Figures 4 and 5 show the ratio of relative X-ray intensities of 1.0 and 2.0 mg/cm2 samples normalized to that of a 0.5 mg/cm2 sample. Table 3 also summarizes these results. Figure 4 shows that, in the direct method, the mean value of the ratio decreased from 0.90 to 0.73 for 1.0 mg/cm 2 to 2.0 mg/cm 2 samples, respectively. However, within the 6 groups, there was no significant difference in this trend. Thus, difference of the particle size did not affect the sample absorption in the present study. Figure 5 shows the result obtained after the absorption

correction was applied. The mean values are 0.95 and 0.94 for 1.0 and 2.0 mg/cm2 samples, respectively. This shows that the absorption correction was fairly accurate within this sample mass range. It the confirms that within these sample masses, the attenuation of X-ray intensity caused by absorption of the sample can be corrected by the X-ray diffraction method using a metal substrate standard. X-ray intensity and particle size range: Figure 6 shows relative X-ray intensity ratios of the 6 size-classified quartz samples normalized to those of the 1 m and smaller samples. There is no remarkable difference between the ratios obtained by the direct method and by the absorption correction method. A general trend can be sent in which the relative X-ray intensity ratios increase with the increase of sample particle size. Although the reasons for this result are not clear, Bhaskar et al.11) have shown a similar result. They mentioned that particle shape and surface characteristics may play an

256

J YABUTA et al.

Fig. 7. Phosphoric acid dissolution residual rates of quartz samples of respective particle size ranges.

important role. It seems remarkable that the difference of the ratios between the particle sizes in groups over 3 m and under 3 m is fairly large. Shimazu12) reported that a quartz sample of 13 m particle size range showed maximum X-ray intensity compared with the other size ranges and that the intensity difference was less than 6% in 110 m size range particles. The present result showed that the intensity difference within 110 m size group was over 30%. One explanation for this discrepancy may be the counting method which we used. In the paper of Shimazu12), particle size distributions were expressed in terms of particle number, whereas we expressed them in terms of particle weight. The difference between our results and the results of Shimazu may be due to mass base size distribution and count base size distribution. Phosphoric acid method Phosphoric acid dissolution residual ratios of quartz samples for respective particle size ranges: Figure 7 shows the phosphoric acid dissolution residual ratios of quartz samples for respective particle size ranges. It can be seen that samples smaller in particle size are smaller in the dissolution residual ratio. This suggests that the samples of smaller particle size are dissolved more in hot phosphoric acid. Although the residual ratio decreased almost uniformly

Table 4. Phosphoric acid dissolution residual rates of mixed samples Sample Phosphoric acid dissolution residual rate (%) Measured value Estimated value 83.1 89.2 92.4 84.7 87.4 90.2

A B C

from large samples to 13 m sample, it suddenly decreased with 1 m and smaller samples. Table 4 shows phosphoric acid dissolution residual ratios of mixed samples. Also, it shows the estimated residual ratios obtained from these of single quartz samples of respective particle size ranges (Fig. 7). The fine particle content in the mixed samples decreased in the order of A, B and C, and the dissolution residual rate also became smaller in the same order. In the comparison between the measured dissolution residual ratios and the estimated ones, the difference was within a range of 2.5% with every sample. It shows that from the dissolution residual ratios of single quartz samples of respective particle size ranges, the dissolution residual ratios of mixed quartz samples could be estimated. This suggests that the dissolution residual ratios in hot phosphoric acid are mostly determined by particle size. Furthermore, it can be considered that finer particles are dissolved almost at the same ratios irrespective of the

Industrial Health 2003, 41, 249259

FREE SILICA IN DUST PARTICLES

257
indicates that samples containing large amounts of fine particles are dissolved further than is estimated by the equation (2). That is, if the proportion of fine particles in a standard sample is smaller than analytical samples, the value obtained by the phosphoric acid method will be underestimated. Therefore, similar to the X-ray diffraction method, it is desirable to use a standard quartz sample having a size distribution approximate to that of the test sample. Particle size distributions of quartz in field samples It is important to know the practical particle size distribution of the quartz in the field sample, because the size distribution significantly affects both the X-ray diffraction method and the phosphoric acid method. Table 6 shows the particle size distributions of phosphoric acid dissolution residues (virtually quartz) in samples obtained at 15 workplaces in a foundry. In the particle size distributions shown in Table 6, the quantity of 10 m and smaller particles was regarded as 100%. The contents of 5 m and smaller particles in the field samples were in a range of 39.2 to 59.4%, and the mean value was 52.0%; 3 m and smaller particles, 13.7 to 31.8%, and 26.3%; 2 m and smaller particles, 8.8 to 20.9, and 26.3%; and 1 m and smaller particles, 3.1 to 9.4%, and 6.2%. Although the ratios were different from workplace to workplace, the approximate size distribution of quartz particles in this foundry was of 5% of 1 m and smaller particles, 20% of 13 m particles, 25% of 35 m particles, and 50% of 510 m particles. As relatively large amounts of 3 m and smaller particles were contained in the field samples and these samples were obtained after dissolution in phosphoric acid, the actual content of 3 m and smaller particles was considered to be larger, about 30%. The results of measurements at other workplaces should be accumulated, but judging from present results, it is desirable that the standard quartz should contain about 30% of 3 m and smaller particles. The particle size

Fig. 8. Relation between dissolving ratio and surface area of particles.

mixed condition. Relation between dissolving rate and surface area: Figure 8 shows the relationship between the dissolving ratio (=100 dissolution residual ratios) and the calculated surface area of test samples. Total surface area of 200 mg of particles was calculated from the size distribution data shown in Figure 3. The particles shape was assumed to be spherical and the mid-value of each size range was adopted to be the representative diameter. The value of the 13 m quartz sample slightly deviated from the regression curve, but the correlation between the dissolving ratio and the surface area was good with at a correlation coefficient of 0.95. It was significant at a significance level of 5%. y = 202.2 x 213.3. (2)

where y is the surface area and x is the dissolving ratio. Table 5 shows the results for mixed samples. For the sample A, which contained a large proportion of fine particles, the surface area obtained from equation (2) is larger than the calculated value expected from its size distribution. This

Table 5. Relation between dissolving ratios and surface areas of mixed samples Item Surface area (cm2) Samples Theoretical value From the formula of Fig.8 / A (3:3:2:1:1) 2623.1 3204.2 1.22 16.9 14.0 0.83 B (1:1:1:1:1) 2114.3 1970.7 0.93 10.8 11.5 1.06 C (1:1:2:3:3) 1351.1 1326.6 0.98 7.6 7.8 1.03

Dissolving Measured Value ratio (%) From the formula of Fig.8 /

258
Table 6. Particle size distributions of phosphoric acid dissolution residue samples (mostly quartz) in 15 workplace of a foundry Workplace No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean SD CV (%) Maximum Minimum 5 m and smaller 52.1 59.4 59.4 55.4 39.2 45.3 55.8 52.0 56.0 59.1 45.4 45.1 49.2 57.4 48.9 52.0 6.26 12.0 59.4 39.2 3 m and smaller 26.5 31.8 31.8 26.1 13.7 19.7 30.4 25.7 27.7 29.7 20.4 23.0 28.3 30.8 28.8 26.3 5.14 19.5 31.8 13.7 2 m and smaller 17.0 20.9 19.8 16.5 8.8 12.7 19.7 16.7 17.6 18.9 13.0 15.1 19.0 20.2 19.5 17.0 3.39 19.9 20.9 8.8 1 m and smaller 5.8 8.1 5.0 4.6 3.1 4.2 7.1 6.9 5.2 6.8 4.4 5.8 8.5 8.0 9.4 6.2 1.82 2.93 9.4 3.1

J YABUTA et al.

distribution13) of the standard quartz commercially available from the Japan Association for Working Environment Measurement (JAWE) contains about 30% of 3 m and smaller particles and about 4% of 1 m and smaller particles. From our results, the quartz sample provided by JAWE is a suitable standard for analysis of field samples.

Conclusion
Size classified quartz samples prepared in this study were used to measure their relative X-ray intensities by the Xray diffraction method and their dissolution residual ratios by the phosphoric acid method. We obtained the following findings. 1) The relative X-ray intensity depends on the particle size, larger or smaller than 3 m. Larger particles showed a large relative X-ray intensity. The difference of relative X-ray intensities could be observed among respective ranges of 3 m and smaller particles, 310 m particles and 10 m and larger particles. 2) The hot phosphoric acid dissolution residual ratio depends on the particle size, larger or smaller than 1 m. A sample of a smaller particle size range shows a smaller residual ratio. The 1 m and smaller sample

showed about 70% of the residual ratio of 110 m particles. The phosphoric acid dissolution residual ratio of each mixed sample was approximately the value estimated from the single quartz samples of respective particle size ranges. 3) The hot phosphoric acid dissolution ratio of a sample showed a fairly good correlation with the surface area of the sample. 4) The quartz particles in field samples from one foundry contained about 30% of 3 m and smaller particles that greatly affect the analytical value. To reduce error due to differences in particle size, it is desirable that the quartz particles used as a standard also contain a similar percentage of 3 m and smaller particles.

References
1) The Work Environment Improvement Office, Department of Safety and Health, The Ministry of Labour (1998) Industrial Health Control and Practice of Design Sampling, The Japan Association for Working Environment Measurement (in Japanese). 2) The Work Environment Improvement Office, Department of Safety and Health, The Ministry of

Industrial Health 2003, 41, 249259

FREE SILICA IN DUST PARTICLES Labour (1998) Working Environment Measurement Guidebook 1 (Mineral Dust), The Japan Association for Working Environment Measurement (in Japanese). NIOSH (1979) Criteria for a recommended standard Vol. 5, 3091, National Institute of Occupational Safety and Health, Cincinnati, OH 45226. Kohyama N (1985) A new X-ray diffraction method for the quantitative analysis of free silica in the airborne dust in working environment. Ind Health 23, 22134. Talvitie NA (1964) Determination of free silica: Gravimetric and spectrophotometric procedures applicable to air-borne and settled dust. Am Ind Hyg Assoc J 25, 16978. Shinohara Y (1993) Application of the microwave acid digestion technique to the pyrophosphoric acid method for quantitative analysis of free silica in dust. Ind Health 31, 91100. Terayama R, Konishi Y, Yabuta J (2001) Effects of Particle Sizes of Particulates on Free Silica Analysis, 3rd report. Working Environment Special Issue No.41, 145 (in Japanese). Yabuta J, Seki Y (2000) Examination on X-ray

259
Diffractometry and Phosphoric Acid Method for Analysis of Free Silcic Acid in Particulates. Working Environment 21, 537 (in Japanese). Yabuta J, Konishi Y, Takata T (1990) Effects of Particle Sizes on the Phosphoric Acid Method. Working Environment 11, 6571 (in Japanese). Yabuta J, Seki Y (2001) Estimation of Particle Size Distributions of Quartz in Field Samples. Working Environment Special Issue No. 41, 167 (in Japanese). Bhaskar R, Li J, Xu L (1994) A comparative study of particle size dependency of IR and XRD methods for quartz analysis. Am Ind Hyg Assoc J 55, 6059. Shimazu M (1962) X-ray diffraction intensity of the NaOH-leached quartz powder specimens with various particle sizes and the structural aspect of crushed very fine particles. I. The relationship between the X-ray diffraction intensity and the particle size of quartz. Bull Nat Ind Health 7, 1424. The Japan Association for Working Environment Measurement (1993) Description of JAWE Standard Samples. 31, JAWE, Tokyo (in Japanese).

3)

9)

4)

10)

5)

11)

12)

6)

7)

13)

8)