If the width factor is greater than one (μ>1), it signifies that the CPMA transfer function is narrower than
the idealized triangular transfer function.
These two factors are determined through multivariate nonlinear functions that are fitted to TCPMA
measurements. The functions take into account either the CPMA mass set point (mp),
CPMA resolution (Rm), and CPMA flow rate (Q), or the CPMA rotational speed (ω) and
voltage (V). If the CPMA flow rate is set to 0.3, 1.5, 4 or 8 LPM, the loss factor is directly
calculated using the fitting function corresponding to the relevant CPMA flow rate. However,
for other flow rates, the loss factor is derived through simple interpolation between the
nearest lower and higher CPMA flow rates.
Under the specified CPMA configuration settings, wherein the width (μ) and height (𝜂) factors of the CPMA transfer function are calculated as below, particles with specified mass exhibit a transmission efficiency of:
The diagram below compares the empirical triangular models with the theoretical triangular, trapezoidal, and diffusion models using the CPMA configuration settings as input. The trapezoidal and diffusion models are alternative transfer functions for CPMA that were derived theoretically by Olfert and Collins (2005) and Sipkens et al. (2020), respectively, based on slightly different assumptions from the triangular transfer function. For instance, the trapezoidal model assumes the axial flow to be parabolic and forces to vary radially, while the diffusion model incorporates particle diffusion into the models. The diffusion and trapezoidal models utilized here correspond to Sipkens et al.'s (2020) study.