Impactors and Aerodynamic Particle Separation
In the realm of aerosol science, measuring the size distribution of airborne particles is crucial for environmental monitoring (e.g., PM2.5 and PM10 measurements) and occupational health. In the pharmaceutical industry, similar measurements are required to test exactly where inhaled drugs (from inhalers) will deposit in the human respiratory tract. The primary instruments used for these precise size classifications are inertial impactors.
The most common variant, the cascade impactor, consists of multiple stages, each designed to capture progressively smaller particles. The operational principle of an impactor stage is relatively simple: a jet of particle-laden air is accelerated through a nozzle and directed perpendicularly (at a 90-degree angle) toward a flat collection plate. Just above the plate, the air sharply diverges and flows radially outward. Small particles with low inertia follow the turning air streamlines and flow to the next stage. Large particles, unable to make the sharp turn due to their high inertia, crash into the plate and are collected (impaction). The sole parameter that governs whether a particle impacts the plate or escapes is the Stokes number ($Stk$). You can simulate parameters for your laboratory designs using our Stokes Number Calculator.
The Critical Stokes Number ($Stk_{50}$) and the $d_{50}$ Cut-Point
Ideally, an impactor stage would act as a perfect sieve, capturing 100% of particles above a certain size and 0% below it. However, fluid dynamics dictates that the transition is not a vertical step, but an S-shaped efficiency curve. Therefore, the characteristic capture size of an impactor stage is defined by its $d_{50}$ (cut-point diameter)—the particle size that is collected with exactly 50% efficiency.
In impactor design theory (such as the foundational work by Marple and Willeke), the Stokes number corresponding to this 50% collection efficiency for a specific nozzle geometry is termed the Critical Stokes Number ($Stk_{50}$).
For impactors featuring rectangular slit nozzles, the theoretical $Stk_{50}$ is approximately $0.59$. For round-jet (circular) nozzles, the generally accepted value is $Stk_{50} \approx 0.24$. (Note: In impactor theory, the characteristic length $L$ in the Stokes equation is traditionally defined as the nozzle radius, $W/2$).
The modified Stokes equation utilized specifically for impactor design is:
$$ Stk = \frac{\rho_p \cdot d_p^2 \cdot U_j \cdot C_c}{9 \cdot \mu \cdot W} $$
In this formula:
- $U_j$: The average velocity of the air jet exiting the nozzle.
- $W$: The diameter of the round nozzle or the width of the rectangular slit. (Useful note: The '18' in the denominator of the standard Stokes equation becomes a '9' here because the characteristic length is $W/2$).
Why the Cunningham Slip Correction ($C_c$) is Non-Negotiable
The standard Stokes drag law is built on the assumption that the fluid is a "continuum." In a continuum, air molecules are so densely packed relative to the particle that the fluid acts as a solid, continuous medium. The average distance gas molecules travel before colliding with one another is called the "mean free path" ($\lambda$), which is roughly $0.066 \ \mu m$ for air at standard conditions.
Cascade impactors are frequently designed to isolate very fine particles (e.g., $1 \ \mu m$ and below). As the particle diameter ($d_p$) approaches the mean free path of the air, the particle essentially begins to "slip" between the gaps of the air molecules. The actual aerodynamic drag acting on the particle becomes significantly less than what classical continuum theory predicts.
To correct this physical discrepancy, the Cunningham Slip Correction Factor ($C_c$) is applied to the numerator of the Stokes equation. The value of $C_c$ increases as the particle size decreases. For a $d_p = 10 \ \mu m$ particle, $C_c \approx 1.016$ (which can often be ignored). However, for a $d_p = 0.1 \ \mu m$ particle, $C_c \approx 2.89$. If you omit the Cunningham correction during design, you will underestimate the inertia of a $0.1 \ \mu m$ particle by nearly a factor of three, resulting in a completely flawed impactor stage.
Realistic Example and Practical Scenario
Scenario: You are designing a custom round-jet impactor stage for a laboratory air quality monitor. You need this stage to have a $d_{50}$ cut-point of precisely $2.5 \ \mu m$ (to act as a PM2.5 separator). Your goal is to determine the required air jet velocity ($U_j$) exiting the nozzle to achieve this exact cut-point.
Data:
- Critical Stokes Number ($Stk_{50}$): $0.24$ for a round-jet nozzle.
- Target Cut-Point ($d_p$ = $d_{50}$): $2.5 \ \mu m \ (2.5 \times 10^{-6} \ m)$
- Nozzle Diameter ($W$): $2 \ mm \ (0.002 \ m)$
- Dynamic Viscosity of Air ($\mu$): $1.81 \times 10^{-5} \ Pa\cdot s$
- Particle Density ($\rho_p$): $1000 \ kg/m^3$ (Standard calibration density)
- Cunningham Correction ($C_c$): Approximately $1.06$ for a $2.5 \ \mu m$ particle.
Calculation Steps:
Rearrange the impactor Stokes formula to solve for velocity ($U_j$):
$U_j = \frac{Stk_{50} \cdot 9 \cdot \mu \cdot W}{\rho_p \cdot d_p^2 \cdot C_c}$
- Calculate the Numerator:
$0.24 \cdot 9 \cdot (1.81 \times 10^{-5}) \cdot 0.002 = 7.8192 \times 10^{-8}$ - Calculate the Denominator:
$1000 \cdot (2.5 \times 10^{-6})^2 \cdot 1.06 = 6.625 \times 10^{-9}$ - Jet Velocity ($U_j$):
$U_j = \frac{7.8192 \times 10^{-8}}{6.625 \times 10^{-9}} \approx 11.8 \ m/s$
Interpretation: To successfully separate $2.5 \ \mu m$ particles at a 50% efficiency rate, you must drive the air through your $2 \ mm$ nozzle at exactly $11.8 \ m/s$. If the flow rate drops and the velocity falls below this, the impactor will fail to capture the $2.5 \ \mu m$ particles, skewing your PM2.5 measurements. You can utilize a reverse-engineering approach in our Stokes Number Calculator to quickly evaluate the required velocities for different nozzle diameters.
Important Warnings and Boundary Conditions
When utilizing the Stokes number for impactor design, researchers frequently encounter practical limitations that cause deviations from theoretical calculations:
- Jet Reynolds Number ($Re$) Limits: The theoretical constant $Stk_{50} \approx 0.24$ is only valid when the airflow inside the nozzle maintains a specific Reynolds number range (typically $500 < Re < 3000$). If the nozzle is too narrow and the velocity is extremely high, the jet becomes highly turbulent upon exit. This turbulence disrupts the smooth radial streamlines above the collection plate, completely destroying the sharpness of the efficiency cut-curve.
- Particle Bounce-Off: Just like in filtration, hard solid particles (like silica dust) possess high kinetic energy upon impact. Instead of adhering to the collection plate, they may strike it, bounce off, and re-entrain into the airflow, passing to the next stage. To prevent this, impactor plates are routinely coated with a sticky substrate, such as silicone oil or specialized greases.
- Particle Loading Effect: In highly polluted environments, massive amounts of dust can build up on the collection plate directly under the jet, forming small "mountains." This accumulated mass alters the physical geometry (specifically the nozzle-to-plate distance, $S/W$), which locally accelerates the air and changes the effective Stokes number during the run, invalidating the calibration.
