Biomedical Engineering Reference
In-Depth Information
and established in vitro aerodynamic performance. The question of the capability of
the CI method to meet this requirement is therefore of crucial importance, regard-
less of the foregoing discussion concerning the capability of EDA with or without
AIM to detect changes in aerosol APSD that can be related in a meaningful way to
in vitro performance change of the OIP.
The sensitivity of the CI method to detect real shifts in aerosol APSD to finer or
coarser sizes is linked to the size-resolving capability of the system as a whole.
Hence, at the simplest level, taking a hypothetical CI whose
d
50
values for the first
and last stage are fixed at 10.0 and 0.5
m aerodynamic diameter representing upper
and lower bounds for a typical OIP APSD, resolving capability can be increased as
more intermediate stages are inserted between the extremes, because the resulting
APSD is defined in terms of the mathematical relationship between API mass and
d
50
at more locations. This reasoning partly drove the decision to design the NGI to
have a minimum of five stages within the critical size range from 0.5 to 5.0
μ
m
aerodynamic diameter throughout its operating flow rate range from 30 to 100 L/
min for MDI and DPI APSD measurements [
83
]. On this basis, rather than simplify
the CI to an AIM-based system, in which two or at most three size-measuring stages
are present within this size range, the number of stages should be further increased
as a means to improve resolution, much as is done with multichannel time-of-flight-
based aerodynamic particle size analyzers [
84
]. However, there are at least two fun-
damental limitations:
μ
was defined in terms of
GSD
stage
, by analogy with the geometric standard devia-
tion for a unimodal and lognormal APSD. At one extreme, a stage with ideal
size-selectivity would have a
GSD
stage
of 1.0 and would be capable of resolving
infinitesimally small shifts in APSD. The most well-designed CI stages from the
aspect of their fluid-handling properties have
GSD
stage
values no smaller than 1.2.
In practice, this limit restricts the number of stages in the critical range to 5, hav-
ing optimally spaced individual stage
d
50
values at equidistant loci on a logarith-
mically scaled aerodynamic diameter axis. Attempts to increase this number of
stages will result in loss of size-resolving power brought about by overlapping
collection efficiency curves for adjacent stages.
2. The measurement precision based on the true stage
d
50
associated with each indi-
vidual stage is also finite, being controlled largely through the degree of control
over the nozzle diameter of a stage comprising an individual jet or the effective
diameter (
D
eff
) for multi-jet stages [
85
].
D
eff
is related in explicit form to the area
mean (
D*
) and area median (
D
median
) diameters of a multi-nozzle stage in accor-
dance with the expression [
84
]:
23
/
13
/
DDD
eff
=
(*)( )
(3.16)
median
Roberts [
85
] reported stage acceptance values with the NGI for
D
eff
varying
between 2.185 ± 0.02 mm for stage 3, 1.207 ± 0.01 mm for stage 4, 0.608 ± 0.01 mm
for stage 5, 0.323 ± 0.01 mm for stage 6, and 0.206 ± 0.01 mm for stage 7. There were
Search WWH ::
Custom Search