Annals of Occupational Hygiene Advance Access originally published online on April 24, 2007
Annals of Occupational Hygiene 2007 51(4):371-378; doi:10.1093/annhyg/mem011
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Redesign of a Static Horizontal Elutriator to Perform According to the ISO 7708 Respirable Convention
Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, Iseigaoka 1-1, Kitakyushu 807-8555, Japan
* Author to whom correspondence should be addressed. E-mail: tmyojo{at}med.uoeh-u.ac.jp
| ABSTRACT |
|---|
|
|
|---|
A static horizontal elutriator (multi-channel elutriator C-30, Sibata Scientific Instruments Ltd, Tokyo) has been widely used as a dust size classifier for a low-volume air sampler in Japan. The sampler uses the historical criterion defined by the British Medical Research Council (BMRC). However, a new sampling convention based on the ISO 7708 respirable dust convention was recently introduced into the Japanese standard for work environment measurement. It is necessary to modify the multi-channel static horizontal elutriator to satisfy the ISO 7708 respirable convention. We propose a modification of the horizontal elutriator, involving the shortening of 11 of the 36 plates to meet the ISO 7708 respirable convention. The relationship between aerosol particle size and penetration for the elutriator was measured in calm air. The measured penetrations were compared with the calculated performance of the sampler and with the sampling convention for the ISO respirable dust. The calculated bias of sampled masses with respect to the ISO respirable mass was almost zero for the workplace aerosols.
Keywords: aerosol respirable dust sampling convention
| INTRODUCTION |
|---|
|
|
|---|
In determining health hazards from inhaled aerosol particles, the respirable convention defined in the international standard ISO 7708 (1995) is commonly accepted by industrial hygienists. Owing to the size selective nature of the particle removal mechanisms in the nasal passages and lung airways, the selection curve embodied in the respirable convention is a function of particle size. The BMRC (British Medical Research Council) definition of the respirable aerosol fraction was the first recognized internationally (Orenstein, 1960). Other national and international conventions have also been proposed since then. However, new definitions of three health-related aerosol fractions (inhalable, thoracic and respirable) were proposed by Soderholm (1989) and adopted by CEN (1993); ISO (1995); ACGIH (1995). In Japan, the criterion defined by the BMRC was adopted more than three decades ago. The new sampling convention (2004) based on ISO 7708 respirable convention (1995) has recently replaced the BMRC sampling convention.
The mass of respirable particles in an aerosol can be determined by any gravimetric sampler attached to a size classification device with penetration characteristics similar to the respirable convention. The size classification devices that have been developed and used to approximate the respirable convention are the horizontal elutriator, the cyclone and the impactor. In order to confirm the performances of these currently used samplers to the ISO respirable fraction and other fractions, the following works were done.
A horizontal elutriator were originally developed in the UK (Wright 1954; Walton 1954; Hamilton and Walton 1961; Dunmore et al., 1964) and used in mines and other workplaces. Cyclones collect particles by the principle of centrifugal force and are more convenient than elutriators for field use since they are smaller and do not need to be operated in a horizontal position as do elutriators. In particular, miniature cyclones worn by workers are suitable for personal sampling (Maynard and Kenny, 1995; Kenny and Gussman, 1997; Harper, 1997; Gautam and Sreenath, 1997). However, the penetration characteristics of the cyclone must be determined experimentally. Impactors and/or virtual impactors are also used to simulate respirable and other sampling conventions (Marple, 1978; Koch et al., 1999). Porous foam size selectors for respirable dust samplers also have been studied (Brown, 1979; Chen et al., 1998; Mulmann et al., 2002) as a new technique.
Personal samplers are the main gravimetric sampling devices used at present in workplace monitoring. However, both low- and high-volume static samplers have a significant role in the calibration of both these personal samplers and direct-reading aerosol measuring instruments (Courbon et al., 1988).
In Japan, area sampling using direct-reading aerosol instruments is conducted at each assigned dusty workplace twice a year. A static horizontal elutriator with penetration characteristics following the BMRC convention has been widely used as the size-classification device of low-volume sampler to monitor respirable dust concentration. It has been used for parallel sampling with direct-reading aerosol measuring instruments at workplace. Although the ISO 7708 respirable convention was introduced as a measurement standard (Ministry of Health, Labour and Welfare, 2004), there is no suitable device meeting the required penetration characteristics for a low-volume sampler.
The penetration characteristics of the horizontal elutriator used at present, following the BMRC sampling convention, are for particles larger than the 50% cut-size much sharper than the ISO 7708 sampling convention for respirable dust (Myojo, 2005). It would be useful for hygienists monitoring workplace conditions to have a multi-channel static horizontal elutriator performing according to the ISO 7708 respirable convention. In this paper, we propose and evaluate a simple device to meet the ISO 7708 respirable convention and evaluate its performance.
| A HORIZONTAL ELUTRIATOR |
|---|
|
|
|---|
In Japan, a static horizontal elutriator with penetration characteristics of BMRC has been widely used as a standard size-selective low-volume sampler to monitor dust concentration at workplaces. A typical multi-channel horizontal elutriator (Sibata C-30 type, see Figure 1) consists of a rectangular box made of brass in which are stacked 36 stainless steel plates 53 mm wide, 85 mm long, 0.3 mm thick and 1.22 mm apart, forming 37 ducts. The flow rate required for the device to satisfy the BMRC criterion, with zero penetration at 7.07 µm is 15 l/min. The plates are inserted into grooves of the rectangular box (see Figure 1). If the rectangular front cover is removed all of the plates can be taken out for cleaning.
|
The penetration of aerosols of aerodynamic spherical diameter dp through a horizontal elutriator can be calculated from physical theory (Walton, 1954) to give in MKS units:
![]() | (1) |
It can be shown that the critical diameter, dp0 of the horizontal elutriator is calculated from:
|
| (2) |
is air viscosity,
is density of particles and g is acceleration due to gravity. If the values corresponding to the Sibata C-30 elutriator (L = 85 mm, W = 53 mm, N = 37, Q = 15 l/min) are substituted into Equation (2), its critical diameter is calculated to 7.07 µm. | REDESIGN OF A MULTI-CHANNEL HORIZONTAL ELUTRIATOR TO SATISFY THE ISO RESPIRABLE CONVENTION |
|---|
|
|
|---|
Particle penetration characteristics of impactors are, like those of simple elutriators, much sharper than the respirable convention. The problem of the sharp penetration characteristics of impactors has been remedied by using several impactors with different nozzle sizes in a parallel flow arrangement (Marple, 1978).
We redesigned the multi-channel horizontal elutriator in a similar way to the multi-impactor. If some of the plates in the elutriator are shortened, the flow rate of through the channel changes in inverse proportion to the length, L if the flow is laminar (see Appendix). Each channel has its individual penetration characteristic, and the combined penetration through all the channels is different from the original BMRC curve. Figure 2 shows the general concept of the redesign. The design is based on the following simple assumptions: (1) All elutriator plates start from the elutriator inlet; (2) Longer plates are mounted below shorter plates; (3) Pressure drop from the inlet to outlet of all channels is the same; (4) The flow in the channel is laminar; (5) Aspiration efficiency of all channels is unity for particles penetrating the elutriator; (6) The effects of ducts formed after shorter plates, such as additional pressure drop or additional separation of particles on the longer plates are ignored.
|
The detailed calculation to fit the ISO respirable convention is given below.
Flow rate through n-th channel, qn;
|
| (3) |
|
| (4) |
The critical settling velocity of micrometer-size particle, vn;
![]() | (5) |
|
| (6) |
At first step of calculation, the critical diameter, dp0 of basic plate length (L = 85 mm) sets as 5 µm in the elutriator and dp0 of minimum length (half length of basic plate, L = 42.5 mm) also sets at 10 µm. In the case studied, the flow rate of each channel is 0.202 l/min for 5 µm, which means that the original horizontal elutriator operates at 7.5 l/min. If we set the value of the critical diameter for n-th channel ranging from 5 to 10 µm, we obtain n-th plate length, Ln, by Equation (6) and penetration of n-th channel, p(dp, dp0) can be calculated by Equation (1). Total penetration of the elutriator is indicated as a summation of all channel penetrations as below.
![]() | (7) |
MS Excel (Microsoft office 2003, Microsoft Inc) was used to calculate the combination of plate lengths, shown in Table 1. The first column is the channel number, n. The second to fourth columns are critical diameters for the channel, dp0n, unit flow rate through the channel, q (1/37 = 0.027 for the unmodified channels) and plate length of the channel, Ln based on Equations (3–6). The next 10 columns are the calculated penetrations, P(dp, dp0n)qn for each channel from 1 to 37 and each diameter from 1 to 10 µm according to Equation (1). dp0n for each channel has been manually adjusted to obtain a small deviation between the calculated total penetration according to Equation (7) and the ISO sampling convention for respirable dust, shown in last two rows of Table 1. The calculated results are close to the ISO respirable convention, and the maximum deviation is 4% at 5 µm. As the factor of flow rate is 1.19 in the table, the total flow rate through the redesigned elutriator should be set to 9.0 l/min (= 7.5 l/min x 1.19). It should be noted that the one shortest plate makes the two shortest channels.
|
The original plate length of C-30 elutriator was 85.0 ± 0.1 mm. Eleven plates out of the 36 plates in the elutriator were removed from the box and sharply cut to within ±0.3 mm tolerance of the values in the Table 1 by a shearing machine. The plate length error of 0.5 mm corresponds to error of 0.12 µm for the shortest plate (42.5 mm and 10 µm).
The combined penetration curve shown in Figure 2 is a function of the number and the length of the plates and total flow rate through the elutriator. The redesigned elutriator based on Table 1 is expected to show cut-off characteristics, which approximate the ISO respirable curve.
| EXPERIMENTAL EVALUATION OF PENETRATION CHARACTERISTICS |
|---|
|
|
|---|
The penetration of aerosol through the elutriators was measured in a vertical plastic cylinder (24 cm in diameter and 100 cm in length) in almost calm air conditions; the downstream wind speed was 1 cm/s. Suspensions of seven sizes of polystyrene latex particles (PSL) were aerosolized as test aerosol by a glass nebulizer sequentially. The size of monodisperse PSL particles are listed in Table 2. The generated aerosol particles were neutralized by mixing with air passing through an ionizer (ZStat 6110A, Ion Systems, Berkeley, USA) before being injected into the cylinder. The aerosol flow rate was 15 l/min and the flow rate of mixing air was 10 l/min. The general configuration of the experimental apparatus is shown in Figure 3.
|
|
The penetration of aerosol through the elutriator was measured with a time-of-flight aerosol spectrometer APS (Aerodynamic Particle Sizer 3300, TSI Inc., St Paul, MN, USA). Particle number concentrations were measured as functions of the diameter given by the APS at the outlet of the test sampler and at the outlet of the reference probe, respectively. The sizes of PSL particles were their nominal diameters and not APS diameter because the APS was calibrated with standard PSL particles. As the sampling flow rate of the APS is 5 l/min, an additional mass-flow type pump was used to increase the flow rate in this experiment as the same manner of Görner et al. (2001).
The elutriator and reference probe, or two reference probes were set parallel at the same height, i.e. 60 cm from top of the cylinder, as shown in Figure 1. At first, two reference aerosol probes, each of which had a bell mouth inlet (74 mm open end and 30 degree cone) and provision for attachment of an elutriator, were set to the same flow rate as that of the sampler. The aspiration efficiency ratio of both probes caused by their positions in the cylinder was measured for each size of PSL particle. Then, a reference probe was switched to the test elutriator for penetration efficiency measurements.
A 1 min measurement by the APS was repeated five times using the reference probe or the elutriator outlet, the line being switched by two butterfly valves as shown in Figure 1. The first minute's data were discarded because of the unsteady state of the aerosol. The remaining 4 min data were summed to give the number of incoming PSL particles. The 4 min data are a group of reference or elutriator data. Four sets of measurements, starting from reference probe, elutriator, reference probe and then finally reference probe, were made, resulting in nine groups of data. Consequently, we obtained four values of penetration efficiency for each size of PSL particles.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
Sampler calibration
The aspiration efficiency ratios between the two reference probes at the two positions are shown in Table 2. The ratio of results for the probes deviated from unity for larger particles. As larger PSL particles showed a larger difference, it seems that they were not uniformly mixed in the cylinder at low-flow conditions, but the ratios were constant and reproducible. The aspiration efficiency ratios were used as correction factors for subsequent penetration measurements and the measured aspiration efficiency for each size PSL particle was multiplied by the measured penetration.
Figure 4a shows the particle cut-off characteristics of the original C-30 elutriator for three flow conditions. The three lines are the results of theoretical calculation. The points are experimental data for flow rates of 7.5 l/min obtained in this work and for flow rates of 9.6 and 15 l/min obtained previously (Myojo, 2005). The flow rate of 15 l/min is the original setting of this elutriator following the former BMRC sampling criterion. The flow rate of 9.6 l/min is a revised rate to meet the ISO respirable convention (adjusted BMRC). The flow rate of 7.5 l/min is based on calculations for the ISO-redesigned elutriator. The performances of the original elutriator showed good agreement with the theory based on Equation (1) at any flow rate. Figure 4b shows the particle cut-off performance of the redesigned C-30 elutriator tested at 9.0 l/min. The flow rate was determined by the calculation shown in Table 1. The performance of the redesigned elutriator also showed good agreement with calculated results.
|
The performance of the ionizer used as an aerosol neutralizer was not measured quantitatively and we cannot be sure whether the test aerosols reached a Boltzmann charge distribution or not. However, the measured penetration of 4 µm PSL particles for the redesigned elutriator was 50.1% with the ionizer on and 48.3% with it off. A higher penetration for neutralized aerosol would be expected.
Bias map of these samplers
Knowledge of the sampling efficiency for each particle diameter allows calculation of the sampler's behavior with various polydisperse aerosols. The normalized mean aerosol concentrations, C that would be sampled from a range of unimodal log-normally distributed aerosols were estimated, both by using the polygonal approximation and by numerical integration of the fitted curve. The bias,
in sampled aerosol concentration, relative to a sampler perfectly following the ISO respirable sampling convention was also estimated (Soderholm, 1989; Liden and Kenny, 1991, 1992; Kenny and Bartley, 1995; Kenny and Gussman, 1997; Görner et al., 2001).
|
| (8) |
Figure 5 shows three bias maps calculated for ideal penetration curves of three types of elutriator. In this work, thirty-five combinations of mass median aerodynamic diameter (MMD) and geometric standard deviation (
) were chosen for simulated size distributions of work environment aerosols and plotted in the figures as (X). The 3D contours were depicted by drawing software named Origin ver. 6.1 (MicroCal).
|
Figure 5a is the bias map between the BMRC (original elutriator at 15 l/min) and the ISO respirable convention. The bias maps as the same sampler combination as this map has been presented in several papers (Soderholm, 1989; Kenny and Bartley, 1995; Görner et al., 2001). The BMRC sampler gave higher values than the ISO respirable convention for respirable dust for most workplace sampling situations.
Figure 5b is the bias map between the adjusted BMRC (original elutriator at 9.6 l/min) and the ISO respirable convention. It clearly underestimates the ISO respirable convention for all situations. In particular, it seriously underestimates it for particle larger than 5 µm in dp50. Figure 5c is the bias map between the redesigned elutriator proposed in this work and the ISO respirable convention. It is difficult to show the results clearly because almost all values are within ±5%. The redesigned elutriator simulated the ISO respirable convention very well compared with fifteen respirable aerosol samplers studied by Görner et al. (2001).
An application of the elutriator
In order to check the efficacy of this redesigned elutriator, aerosols containing larger size particles than 4 µm are preferable. We used a kind of toner for copy machines. The black toner was aerosolized by a fluidized bed aerosol generator (Tanaka and Akiyama, 1984). Generated toner aerosols were mixed with clean air in pre-chamber and then introduced into the top of an inhalation chamber (80 cm square floor, height 60 cm and volume 0.38 m3, aerosol flow rate 150 l/min). Then the toner aerosol was introduced from the ceiling of the chamber to floor at 0.4 cm/s velocity with 6 mg/m3 concentration. In the inhalation chamber, we used two elutriators and one Andersen sampler (eight stages, KANOMAX JAPAN, Inc. Osaka, Japan) to determine the ratios of BMRC/ISO-redesigned and adjusted BMRC/ISO-redesigned, respectively and the size distribution of the test aerosols.
Figure 6 shows an example of the deposited toner particles on filters (T60A20, Pall Co. NY, USA) behind both elutriators. The sampling time was 295 min and flow rate was 9.6 l/min for original elutriator (adjusted BMRC) and 9.0 l/min for ISO-redesigned elutriator. The filter set in the original elutriator shows typical stripe pattern of deposited toner particles as shown in Figure 6a. In each slit of the elutriator, the lower part shows a dense colouration by dust particles, and the upper part shows a thin colouration due to the sedimentation loss of particles in the elutriator channel. Inside the elutriator channels the flow pattern was laminar, but changed from rectangular to circular cross section upon entering the filter. There are 37 stripes, equal to the number of slits. However, a part of aerosol particles through the bottom slit may deposit on elutriator inner wall. On the other hand, the filter in the ISO-redesigned elutriator, Figure 6b, is well-mixed at the upper half and a stripe at the lower half. The mixing may be associated by different velocities at outlets of redesigned slits.
|
The sampled masses of the toner aerosol on the filters in Figure 6a and b were 3.3 and 6.1 mg, and the concentrations were 1.2 and 2.3 mg/m3. Other sampling was conducted and the sampling time was 180 min and flow rate was 15.0 l/min for original elutriator (BMRC) and 9.0 l/min for ISO-redesigned elutriator. Sampled mass of the toner aerosol on the filter of original elutriator was 4.3 mg and the concentration was 1.6 mg/m3. Sampled mass on the filter of ISO-redesigned elutriator was 2.5 mg and concentration was 1.6 mg/m3. The concentration ratio of BMRC/ISO-redesigned was 1.0 and adjusted BMRC/ISO-redesigned was 0.51. According to the Andersen sampler, the mass mean aerodynamic diameter MMAD of the aerosol was 5.5 µm and geometric standard deviation (
) of larger particles was 1.6 graphically. The shape of toner particles was fairly spherical according to SEM observation. However, the toner aerosol had bimodal size distribution, i.e. micron size toner particles and submicron size fragment of toner. Compared with calculated results (BMRC/ISO and adjusted BMRC/ISO) shown in Figure 5a and b, the measured ratio and size distribution are slightly different. If we assumed that dp50 of the toner aerosol is 7.7 µm, other values are consistent with Figure 6. Other aerosols should be used to test the performance of the redesigned elutriator as a quasi-ideal respirable sampler. | CONCLUSION |
|---|
|
|
|---|
Horizontal elutriators have been used historically with static samplers to meet the criterion defined by the British Medical Research Council (BMRC). A multi-channel sampler (C-30, Sibata Scientific Instruments Ltd, Tokyo) has the potential ability to meet not only ISO 7708 respirable convention but other conventions by changing the slot plates combination. We proposed a modification method for the horizontal elutriator, involving the shortening of 11 of the 36 plates to meet the ISO 7708 respirable convention. The relationship between aerosol particle size and penetration for the elutriator was measured in calm air. The calculated bias of sampled masses with respect to the ISO respirable mass was almost zero for the workplace aerosols. An application of this redesigned elutriator to toner particle aerosol showed reasonable results, but it needs a more precise size distribution technique for micrometre size aerosol to evaluate the performance.
| APPENDIX |
|---|
|
|
|---|
Air flow rate of each channel
Flow rate through n-th channel, qn was assumed in inverse proportion to channel length, Ln, as shown in Equation (3).
|
|
The pressure drop,
Pn can be approximated by
|
| (A-1) |
|
| (A-2) |
![]() | (A-3) |
From these equation, we can obtain the equation
|
| (A-4) |
In this study, we assumed dn = 2h and constant for all channels, because the height between plates h = 1.22 mm, W = 53 mm and h/W <<1 in (A-3).
|
| (A-5) |
If simply assumed that pressure drop of each channel is constant and the same as pressure deference between inlet and outlet of the elutriator,
P
|
| (A-6) |
Flow rate through a channel n, qn was in proportion to the cube of opening height h, and in inverse proportion to channel length, Ln. Hamilton and Walton (1961) have already pointed the accuracy in construction of horizontal elutriator was dependent on the cube of the channel height. Precise and fine machining techniques for the elutriator grooves to insert thin plates were needed to control penetration efficiency.
| ACKNOWLEDGEMENTS |
|---|
|
|
|---|
The authors greatly appreciate Dr Richard C. Brown (UK) for his suggestions and assistance with the editing of this paper.
Received September 7, 2006; in final form December 27, 2006
| REFERENCES |
|---|
|
|
|---|
ACGIH. (1995) Threshold limit values. American Conference of Governmental Industrial Hygienists.
Brown RC. Porous foam size selectors for respirable dust samplers. J Aerosol Sci (1979) 11:151–9.[Web of Science]
CEN. (1993) Workplace atmospheres: size fraction definitions for measurement of airborne particles in the workplace. EN481, CEN.
Chen CC, Lai CY, Shih TS, et al. Development of respirable aerosol samplers using porous foams. Am Ind Hyg Assoc J (1998) 59:766–73.[Web of Science][Medline]
Courbon P, Wrobel R, Fabri JF. A new individual respirable dust sampler: The CIP 10. Ann Occup Hyg (1988) 32:129–43.
Dunmore HH, Hamilton RJ, Smith DSG. An instrument for the sampling of respirable dust for subsequent gravimetric assessment. J Sci Instrum (1964) 41:669–72.[CrossRef][Web of Science]
Gautam M, Sreenath A. Performance of a respirable multi-inlet cyclone sampler. J Aerosol Sci (1997) 28:1265–81.[CrossRef][Web of Science]
Görner P, Wrobel R, Micka V, et al. Study of fifteen respirable aerosol samplers used in occupational hygiene. Ann Occup Hyg (2001) 45:43–54.
Hamilton RJ, Walton WH. The selective sampling of respirable dust, in Inhaled Particles and Vapours. Davies CN, ed. (1961) Oxford, UK: Pergamon Press. 465–475.
Harper M. Measurement and monitoring of airborne particles—Techniques and regulations, Processing and handling of powders and dusts. In: The minerals, metals and materials society—Battle TP, Henein H, eds. (1997) Orlando, FL, USA: 1997 TMS Annual Meeting.
ISO 7708. Air quality—Particle size fraction definitions for health-related sampling. (1995) International Organisation for Standards.
Kenny LC, Bartley DL. The performance evaluation of aerosol samplers tested with monodisperse aerosols. J Aerosol Sci (1995) 26:109–26.[CrossRef][Web of Science]
Kenny L, Gussman RA. Characterization and modelling of a family of cyclone aerosol preseparators. J Aerosol Sci (1997) 28:677–88.[CrossRef][Web of Science]
Koch W, Dunkhorst W, Lodding H. Design and performance of a new personal aerosol monitor. Aerosol Sci Technol (1999) 31:231–46.[CrossRef][Web of Science]
Lidén G, Kenny LC. Comparison of measured respirable dust sampler penetration curves with sampling conventions. Ann Occup Hyg (1991) 35:485–504.
Lidén G, Kenny LC. The performance of respirable dust samplers: Sampler bias, precision and inaccuracy. Ann Occup Hyg (1992) 36:1–22.
Marple VA. Simulation of respirable penetration characteristics by inertial impaction. J Aerosol Sci (1978) 9:125–34.[CrossRef]
Maynard AD, Kenny L. Performance assessment of three personal cyclone models, using an aerodynamic particle sizer. J Aerosol Sci (1995) 26:671–84.[CrossRef][Web of Science]
Ministry of Health, Labor and Welfare, Japan. (2004) Notification 368.
Mulmann C, Aitken RJ, Kenny LC, et al. Size-selective personal air sampling: a new approach using porous foams. Ann Occup Hyg (2002) 46:386–9.
Myojo T. Assessment of measured respirable dust sampler penetration and the sampling convention for work environment measurement. San Ei Shi (2005) 47:239–245.
Orenstein AJ. Recommendations adopted by the Pneumoconiosis conference. In: Proceedings of Pneumoconiosis conference. (1960) London: Johannesburg, J and A Churchill Ltd. 610–621.
Soderholm SC. Proposed international conventions for particle size-selective sampling. Ann Occup Hyg (1989) 33:301–20.
Tanaka I, Akiyama T. A new dust generator for inhalation toxicity studies. Ann Occup Hyg (1984) 28:157–62.
Walton WH. Theory of size classification of airborne dust clouds by elutriation. In: Brit J Appl Phys (1954) S29–S37.
Wright BM. A size-selecting sampler for airborne dust. Brit. J industr Med (1954) 11:284–8.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||









