BAJA SAE Transfer Case Design

This is an ongoing project, that I plan to update this page with regarding future developments. In my last semester at Northeastern, I was tasked with designing the transfer case for our Baja SAE car.

The purpose of this transfer case is to enable 4WD in our off road cars, by redirecting torque from the rear gearbox at a 90 degree angle to transmit torque to the front axle via a driveshaft. This is accomplished by 2 bevel gears: a smaller pinion gear at the gearbox output, and a larger drive gear meshing to it to create an additional ratio reduction.

The previous design from the 2024 racing season was only responsible for this torque redirection, not for toggling 4WD on and off. Engagement of power to the front axle was controlled by a custom slip clutch with a linear actuator controlled by a wheel mounted button.

Early iteration of 2025 Transfer Case

For the first part of the 2025 design cycle, the tentative plan for 4WD actuation was very similar, so my primary goals were to improve on the weight and footprint of that design. Later in the design stage, the drivetrain subteam decided to pursue a dog clutch inside the transfer case instead of a seperate active slip clutch. This has lead to extensive iteration of the transfer case in both of these configurations. As machining deadlines approach, the team is currently working to finalize the specifics of this actuation mechanism.

The following images are from design review decks detailing progress on my early non-actuated variant.

One of the first significant assembly changes I made to the previous years design was replacing the mounting bolts with studs threaded into the rear gearbox. This was in response to the larger stack up of components between the gearbox and transfer case which required increased clamping force and rigidity. It also solved clearance issues created from mounting holes that allowed enough space to tighten a nut onto a stud, but did not have enough passthrough to accommodate a large enough bolt being installed.

Subsequent improvements to this version were focused on further reducing weight, improving machinability, and conducting internal FEA based on the maximum torque produced by the drivetrain.

After this point in development, after deliberation with other subteams, we decided to change directions with the transfer case functionality, and include an internal, actuated dog gear to toggle 4WD instead of an external slip clutch. Another member became responsible for the design of the dog clutch and gear teeth themselves, while I continued to develop the case, fitment, and actuator design.

Output bevel gear with integrated dog clutch and engagement spring

First iteration of toggled transfer case

The first embodiment of this design involved dog teeth machined directly into the front side of the output bevel gear. A tuned spring would engage the teeth into 4WD, and to disengage, tension would be applied to two control cables extending into the case, retracting the

toothed cage.

I conducted FEA on the case based on the new loading conditions, which would involve enough cable tension braced against the back of the case to compress the internal spring. These results were promising, and even with these modifications, the total aluminum components were still significantly lighter than last years transfer case.

The basic principles here were generally received well during their design review, but there were some key concerns which influenced further changes. Firstly, there was the problem of creating a reliable seal for the entry point of the cables to prevent oil leakage and dust ingress. Typical grommets or rod seals may have sufficed, but the inconsistent surface finish of a cable combined with it's lack of rigidity would have likely performed worse than polished rods.

Having cables operating inside the case would have also been a source of risk, because even though the sliding cage is suspended on its own bearings and does not see any torsional loading, even a temporary rotational force could misalign the cables. In extreme cases, this could cause interference between the cables and gear mesh which would spell a

catastrophic failure.

The following revision addressed these concerns:

The two cables were replaced with 3 ground and polished .25" rods, entering the case through rod seals. They were connected with 8-32 screws on their axis to the trident shaped yoke which could be easily waterjet out of sheet steel. In this configuration, one cable could pull the center of the yoke, which also addressed concerns about symmetrical force application to the dog clutch.

Solving all of these problems introduced a new one, which once again catalyzed a significant redesign. The black circular shape shown above is the rotating spline for the rear CV axle, which was already in very close quarters to the transfer case (pink). Whereas the two independent cables could route around this CV axle, the yoke and pushrod design had no additional clearance to fit a central cable.

In response, the team member designing the dog clutch and I collaborated to move the teeth to the opposite side of the bevel gear, allowing for actuation through the front side of the transfer case.

In this configuration, the tri-spoke yoke was replaced by a ring, with 2 pushrods entering the case, and 2 opposing cable connections. When tension was applied to the cable, this would plunge the ring and pushrod assembly inward, engaging the teeth.

A new limitation for this design was the lack of space for an internal spring. In the first cable actuated design, the internal spring force is what engaged 4WD, and the cable tension separated the dogs, disengaging it. For this embodiment, the cable tension would engage the clutch, while the springs would separate it. Based on friction calculations, the force output required of these springs was not significantly different for either direction of operation, however, none strong enough could fit into this new version.

To address this, I replaced the single internal springs with an array of external springs. This solved the internal clearance issues, and also doubled as an automatic alignment feature. By using 4 separate springs that symmetrically opposed the pushrods, if any one point of the ring was having force applied unevenly, it would receive proportionately greater resistance from the springs, ensuring a parallel, axial plunge of the pushrod assembly.

As a revision to this design, I wanted to create a version with a simpler cable interface. The one shown above fastens the cable housing to the floating ring, while the cable end is threaded directly into the case. This would require an unconventional lever setup on the force application side, which would limit the use of OTS parts. More conventional cable operation is shown below.

By adding concentric ears to the case lid and floating ring, the cables are free to operate in a standard configuration with the housing fixed, and the cable end receding into it. Even though the cables would have to run rearward in this version, they have sufficient clearance to route safely around the moving components of the rear drivetrain.

Moving forward, I am working to validate this design before machining, and finalize the driver interface for actuation. For a manual lever option, there needs to be a mechanism to lock tension in the engaged position, and then easily disengage it. A gated shifter style track would be one option for this, while another is to use an OTS parking brake lever assembly, which already has a ratcheting mechanism for pulling a cable, and a simple push button release.

An electronic linear actuator may also be used to similar effect, and could be positioned anywhere on the frame without the same ergonomic considerations of a lever. The team is continuing to explore and validate these options as the machining and build cycle approaches. Regardless of the final outcome, one certainty is that this project will involve some very ambitious and exciting additions to our 4WD system.