CFD For A Complete Dust Collector

Having performed Computational Fluid Dynamics (CFD) simulations for each component of Matthias Wandel's "Small Dust Collector," the time has come to assemble and test the entire system - virtually, of course. Recall that the main components of the dust collector are a cyclone, a filter assembly, and a blower.

Dust Collector CFD SimulationDust Collector CFD Simulation

3D Model Construction

The fully assembled model required the repositioning of the three main components and then virtual plumbing to connect them as depicted in Matthias' SketchUp assembly model.

Small Dust Collector SketchUp ModelSmall Dust Collector SketchUp Model

Small Dust Collector SketchUp Model SectionSmall Dust Collector SketchUp Model Section

First the pipes for the complete CFD model were trimmed to the required lengths. Then the final multi-volume flow-domain was created using a combination of edge lofting to create joint faces and volume stitching.

Small Dust Collector CFD ModelSmall Dust Collector CFD Model

Performance Test

Unlike the previous CFD simulations where we had to consider a range of volume flow rates to determine the pressure drops at each condition, this test has only one operating parameter and it is fixed according to the rotation speed of the motor at 3500 RPM. The single inlet and outlet pair are assumed to be open to atmosphere, equivalent to a gauge pressure of zero.

Residuals MonitorResiduals Monitor


Recall that from our individual cyclone and filter assembly CFD simulations we found that their combined demand curves (pressure drop vs volume flow rate) intersected the blower's fan curve at a volume flow rate of 0.063 m3/s. The results for the fully assembled system indicate a volume flow rate of 0.08 m3/s.

Volume Flow Rate MonitorVolume Flow Rate Monitor

Notice that there is some discrepancy in flow rate between the two CFD results for the discrete components and the fully assembled dust collector. Given that the cyclone contributes to the majority of the pressure drop through the dust collector, it is the obvious place to look for differences. It appears that the outlet location for the lone cyclone CFD analysis intersected a vortex core, whereas the cyclone exhaust piping in the assembly CFD analysis allows the vortex to dissipate and the pressure to recover. This results in a lower overall pressure drop prior to entering the filters and therefore a higher overall flow rate for the assembly.

Flow Visualization


Pressure Iso-SurfacesPressure Iso-SurfacesDust Collector CFD Simulation

Clipped Pressure Iso-SurfacesClipped Pressure Iso-SurfacesDust Collector CFD Simulation

Velocity Iso-SurfacesVelocity Iso-SurfacesDust Collector CFD Simulation

Clipped Velocity Iso-SurfacesClipped Velocity Iso-SurfacesDust Collector CFD Simulation


Velocity Vectors For BlowerVelocity Vectors For BlowerDust Collector CFD Simulation

Streamlines For CycloneStreamlines For CycloneDust Collector CFD Simulation

Pressure Iso-Surfaces For FiltersPressure Iso-Surfaces For FiltersDust Collector CFD Simulation

Velocity Iso-Surfaces For FiltersVelocity Iso-Surfaces For FiltersDust Collector CFD Simulation

Streamlines For FiltersStreamlines For FiltersDust Collector CFD Simulation


The assembled geometry was created in Caedium Professional. The CFD simulation was performed using the incompressible, steady-state RANS solver with porous media, a Moving Reference Frame (MRF), and the k-omega SST turbulence model.


Good Analysis

Good work, In this model you have cyclone->filter-> impeller this makes theoretical sense as the air going through the impeller is clean, but as other makers often may put impeller->cyclone->filter or cyclone->impeller->filter. I wonder if there is any order that is better optimized for flow.

Improvement Idee

I'm currently working on an idea for my dust collector in which i put the impeller on top of the cyclone.

The cyclone creates spinning air an the impeller has to do the same. so why not use the sinning air of the cyclone to help the impeller.
In this idea i wanted a minimum of sharp corners that restrict the airflow

the Cyclone I have in mind is a bit different from the normal ones. On the bottom is a cone with a 4 cm gap at the sides the wall with the inlet is tilted 20 degrees inward (bottom wide top narrow) and after that it has a 45 degree angle toward the intake of the impeller.
The impeller has two outlets but not at 180 degrees from each other but 187.5 and 172.5 degree this combined with a 12 blade impeller, makes that the outlets have there pressure alternating instead of simultaneously. This will cancel out a part of the noise.
after this comes 2 halve cartridge filters one for each outlet

if there are questions please ask. I got some drawings on paper.

The impeller is protected by

The impeller is protected by being after the filter. A nail could bounce around in the cyclone and trigger a bad wreck of the impeller. Maybe ad an energy recovering diffusor at the cyclone and then stacking the two filters to create a long straight run of the air before the blower inlet? Fans are sensitive to turbulence and flow misalignment at their inlets.