the convention more exactly will exhibit smaller biases throughout the entire size distribution range. Bias maps are available for several respirable samplers [38, 39]. It should be mentioned that in some cases, e.g., coal mine dust sampling, a single sampler is specified by regulation. This sampler specification eliminates the question of bias for that type of measurement. Investigation of the effect of changing the physical dimensions of a commercial cyclone resulted in modifications that improved the match to the respirable sampling convention [52]. Chen and coworkers developed a virtual cyclone that appeared to give excellent compliance with the respirable curve [53]. It is possible to make samplers that have predicted biases less than 10% over the entire range of likely workplace size distributions. While the behavior of certain samplers, such as impactors, can be predicted theoretically, it is still important to measure penetration curves experimentally to ensure correct application of the theory. Bias maps based on these data then allow estimation of accuracy for a specific workplace application. As improved samplers are tested and become commercially available, more accurate thoracic and respirable aerosol measurement on a routine basis will be possible. As mentioned above, several inhalable samplers were investigated in a wind tunnel to evaluate their sampling efficiency compared to the inhalable convention [9]. Based on these data, sampler performance (maximum bias confidence limit) was ranked [Bartley 1998] and the IOM, GSP, and CIP-10 samplers were rated the best. As interest in the new particle size-selective conventions by standards setting bodies has grown, efforts have been made to define protocols to guide the testing and validation process for available samplers. One approach was developed by the CEN [54, 55]. In the CEN model, for any given sampler to be tested, the first step is a critical review of the sampling process for the instrument in question. This is intended to identify factors that may influence the performance of the sampler, including particle size, windspeed, aerosol composition, filter material, etc. This is essential in the process of sampler evaluation, determining under what conditions the sampler will need to be tested. Three options are then presented for the testing of samplers: (a) the laboratory testing of samplers to compare performance with the sampling conventions, (b) the laboratory comparison of instruments, and (c) the field of comparison of instruments. Research projects have been conducted in recent years to define testing protocols (option a), funded both by the European Community and by NIOSH, to consolidate the scientific basis for such protocols and to identify improved and more cost-effective methods. 4. SAMPLER ASSEMBLY Some samplers are designed such that improper assembly can result in internal leakage, i.e., aerosol particles bypassing the filter. This bypass leakage has been noted in the 37-mm closed-face cassette [56,57]. Although at least one study found no problem with hand assembly of these cassettes [58], NIOSH and others [59] have occasionally observed, after sampling black or colored dusts, streaks of dust on the filter’s compression seal region or an incomplete compression mark, indicating aerosol leakage bypassing the filter. An airtight seal in these cassettes is achieved by compression of two plastic parts that must be parallel and joined with the proper force. If this seal is not compressed with sufficient force, the vacuum behind the filter may pull the filter from the seal, especially at high flow rates. Too high a compression force results in cracking the cassette or cutting the filter, also producing leakage.
3/15/03
189
NIOSH Manual of Analytical Methods
