Cooling Fan Impeller Design
Status: Complete
During my time as a mechanical engineering intern at Milwaukee Tool, I designed and optimized a cooling fan for a 6 bay charger in development. I first found the aerodynamic impedance of the charger geometry, then created multiple fans designed for that operating point, then focused on and optimized one of them for highest efficiency.
Project Type: Internship
Project Timeframe: 6/5/2023 - 8/12/2023
Categories: Turbomachinery Design, CFD Simulation, DOE Optimization
Skills: CAD, CFD, DOE Optimization
Due to an NDA, CAD and simulation files as well as other information about the product in development cannot be disclosed.
When a battery pack gets taken off of a high-power-draw electric tool, it is often too hot to charge it safely. To make matters worse, charging the battery itself also generates heat. Due to these reasons, Milwaukee Electric Tool Company ("Milwaukee Tool" from here on out) developed a dual-bay charger with fans under each pay that delivers air through vented batteries to cool it during the charging process (this product was released August 2023). To increase charging capacity, Milwaukee Tool is developing a multi-bay charger (>2 bays), but this charger will only have 1 fan to reduce the total cost spent on fans. My internship task was designing a fan to be used on this multi-bay charger and to exceed the performance of the released dual-bay charger.
The first step is to determine how much air is needed for cool each battery. The first step was to build a model of the charger-battery system. To save computational costs, the batteries, for which there are flow rate vs. pressure drop data, was curve fitted to a quadratic relationship with viscous and inertial loss coefficients. Then, the batteries were approximated as porous zones in Ansys fluent, which eliminated the need for all the complicated geometries inside the batteries while maintaining very good pressure drop accuracy. These porous batteries were then attached to the charger to form the whole flow path (above picture), and I obtained a flow rate vs. pressure drop curve for this system by running a few different initial flow rate conditions.
From experimental testing, I found the amount of airflow needed per 1 battery needed to meet the dual-bay charger's benchmark. From taking 6 times this number plus other simulated losses throughout the system, I found the airflow needed for the whole system, and extracted the pressure drop for this corresponding flow on the system flow rate vs. impedance curve. This flow rate and pressure drop information was then used as initial constraints for the fan design.
Aside from flow rate and pressure drop, there were other constraints, such as the hub and shroud diameters, which is a result of the charger inlet geometry. The motor power is also fixed. As such, a brief overview of the design and optimization process is shown on the right.
Aside from the abovementioned constraints, the blade geometry itself is completely open to variation. As such, I selected 4 variables to change: the tip sweep angle, the blade aspect ratio (tip chord length / height), the blade trim ratio (tip chord length / hub chord length), and the hub loading parameter (relative gas inlet angle at hub vs tip). From these variables, I conducted a DOE (design of experiments) optimization, where points in the design space are randomly sampled and simulated to generate a response surface (shown on the right). The output parameter on the right is the downstream total-to-static efficiency, and the surface indicated that a high blade trim ratio and a low blade aspect ratio would result in a high efficiency given my design constraints.
I optimized a total of 7 output parameters, including total-to-total efficiency, total-to-static efficiency, hub DeHaller number, etc. For each optimized design, I ran a full 3D turbomachinery CFD simulation, as shown below. These tests gave me a expected outcome of the performance of the fans relative to each other.
Lastly, all the design models were 3D printed and tested in real life by attaching them to a motor and fixing then onto a model of the charger. The end-of-charge temperature was measured, and the lower this number, the better the performance of the fan will be. One of the best fans is shown on the right, which had a 40.9 celsius end of charge temperature, compared to the 45 celsius of the released dual-bay charger.