Computational Modeling of Twin Screw Pumps for Thermal Management
Applications 15-15-01-0004
This also appears in
SAE International Journal of Passenger Vehicle Systems-V131-15EJ
Electrification has become less of a catchphrase and increasingly commonplace
when discussing today’s locomotives. Engineers developing thermal management
strategies (both component suppliers and system-level analysts) must be armed
with effective tools to design and analyze essential components such as coolant
pumps and study their behavior in an actual system. This study focuses on the
analysis of twin screw pumps for cooling battery packs in hybrid and battery
electric vehicles via three different approaches—experimental measurements, a
one-dimensional (1D) thermodynamic chamber model, and a three-dimensional (3D)
computational fluid dynamics (CFD) model. Experimental measurements are
conducted to quantify the coolant’s volume flow rate and estimate hydraulic
power consumption over a range of operating speeds and pump discharge pressures.
While these measurements provide some insight into the overall internal leakages
and pumping efficiencies, more comprehensive tests at a higher cost are required
to fully understand the detailed thermodynamic processes occurring within the
pump. Two computational modeling approaches are presented and extensively
validated against these measurements. The 1D chamber model demonstrates a good
agreement of all measured quantities at a very low computational cost. It also
provides useful information regarding the relative importance of the various
leakage paths along with the working processes and pressure pulsations. This
makes it an effective tool to quickly analyze operating conditions where test
data may not be available and iterate toward improved designs via parametric
analysis. A 3D CFD yields very good agreement compared to the measured results
and provides a more complete picture with greater spatial accuracy that is
sacrificed in the 1D approach. However, this is available at a significantly
higher computational cost. A combination of both methodologies can guide
engineers in designing screw pumps for optimal performance.