Engineering - FSAE MY25 — Sehar's Portfolio

MIT Motorsports (Formula SAE) - MY25 Motor Cooling Sleeve

MIT Motorsports (Formula SAE EV)
Cambridge, MA
September 2024 - Present hello For MIT Motorsports’ MY25 four wheel drive (4WD) car, I am responsible for the design and manufacturing of a custom in-hub electric motor cooling system from scratch. This is the first time the team is designing a 4WD system, so there is no prior knowledge about in-hub AMK motor cooling design. Thus, my design process has involved thorough literature review, learning from the experiences of other FSAE teams, detailed hand-calcs and simulations, and prototyping and testing.

hello

Figure 1: Early motor cooling design concepts

helloEarly in the design process, it was important to consider various modes of heat transfer and different coolants. Four primary cooling strategies were considered: 1) water cooling, 2) air cooling, 3) heat pipes, 4) hybrid, as shown in Figure 1. To properly evaluate these options, I wrote a script in Python to estimate the thermal resistance of each design. Additionally, it was important to consider packaging since the motors would be in-hub and, thus, the space around them would be more restricted. Accounting for the hubs, uprights, and wheels, it was clear that at least 2/3 of the motor would be shielded from reliable air-flow. Moreover, there would be nowhere for the air to escape if forced-convection was used, as one end of the motor would be blocked by the wheel package. Hence, with expected thermal resistance, packaging, manufacturability, and other criteria in mind, it seemed that a water jacket would be the best approach.

hello

Figure 2: Concepts for the channel geometry of a water cooling jacket

hello

Figure 3: Initial CFD analysis of the 4 channel geometries drawn in Figure 2.

helloWater cooling jackets can have circumferential or axial cooling channels. Circumferential channels tend to be efficient for pressure drop since it doesn’t excessively disturb the flow, while axial can be better for heat transfer due to the increase of flow turbulence with the additional 180° bends. Typically, a circumferencial architecture is adopted since it offers a good compromise between heat transfer effectivity and pressure drop.

Considering both architectures, I down-selected to the 4 designs pictured in Figures 2 and 3. Characterizing these designs with hand-calculations along would be too complex, so I employed CFD to better compare the heat removal and pressure drops for each design. Currently, CFD seems to indicate that the helix design and axial snake design may be the most promising. I am continuing to refine this CFD for more accurate results. Bend testing was completed for the conventional helix design and I plan to use this data as a benchmark for the CFD.

hello

Figure 4: Adding spherical dimples internally to the channels can increase flow turbulence and enhance heat transfer

hello

Figure 5: Adding ribs internally to the channels can help enhance heat transfer and ensure better surface temperature uniformity than dimples.

helloWhile I iterate on the channel geometries, I am also exploring ways to continue enhancing thermal exchange without catastrophically increasing pressure drop. One way to do this is by adding spherical dimples to the inisde of the channels. However, this may result in significant hot spots, which would be detrimental to motor performance and safety. So, another concept is to add ribs to the inside of the channels which will similarly promote turbulence, but result in a more uniform temperature distribution. These designs are yet to be analyzed with CFD.

hell

adas

fh hello
hello
hello

f

helhello

<- PREV | NEXT ->