Overview:
I designed and analyzed the 4WD engagement/disengagement dog clutch for our car. In addition to the design of the actual clutch components, I also developed a new packaging architecture that allowed frame weight to be locally reduced by ~10% and massively increased manufacturability. This is the first dog clutch designed in our team’s history that has functioned continuously with no issues through all of testing and competition.
In previous years, reliability with four-wheel drive engagement and disengagement has been a massive issue that has lost the team significant points in endurance, the largest points scoring event. Reliability and efficient mass usage were the priorities for the redesign of the system, which is reflected in the requirements.
The input dog gear will slide inside of female splines on the gearbox output shaft. This reduces the total axial length of the assembly significantly and allows it to package nicely inside of existing material
Teeth on both gears would shift from a ramp design to a flat design. This increases ease of machining and reduces the stress concentration at the tooth root.
The stationary dog gear and the sliding dog gear are made into identical parts, but used differently.
Use the bearing that constrains the stationary dog gear inside of the housing as the connection between the throwout fork and the sliding dog gear. This prevents frictional wear from the throwout fork, which was a major issue with previous designs.
Material - AISI 4340 per AMS 6414
Actuation Method - Push-pull cable
System level redesign of the rear inboard assembly.
I determined spline length using equations from Dudley’s “When Splines Need Stress control”, calculated contact and bending stress using equations from Shigley’s Mechanical Engineering Design textbook, and performed bulk structural analysis in Ansys to prove out the housing redesign concept.
Note that the input and output dog gears are identical parts. On the input sliding dog gear, the bearing is used as the interface point between the rotating gear and the throwout fork. This prevents friction wear, which was a problem in previous years. The same bearing on the stationary gear is used to constrain it radially and take any of the dynamic loads that occur during engagements and disengagements. There is also a linear bearing on the fork which guides the engagement path.
This simulation encompasses the entire rear drivetrain and frame. I wanted to investigate the effects of the bevel gear reaction forces + full drivetrain torque on the bulk structure of the rear inboard and how it would affect the interface of our rear transfer case with the propshaft system. I have the gearbox, rear frame structure, bevel gears, and a stiff body to ground spring to simulate the torsional stiffness of a locked propshaft/front driveline. What I found is that the stresses induced in the housings are minimal and the maximum deflection of the assembly was approximately 0.02 inches, which is well within the misalignment range of the crowned spline on our propshaft. The diameter and thickness of my structural housing was largely a function of the clutch component’s outer diameter and not a bending stiffness requirement. Despite being loaded by the bevel gear reaction forces, the housing was still able to reduce in weight from 0.53 lbs to 0.29 lbs. This is a much more effective mass usage than the previous housing which existed strictly to fulfill safety requirements.
My tooth simulation was set up to simulate the absolute worst case engagement scenario. In this situation, max torque is being applied to the top half of only 2 of the teeth. This was derived from the maximum possible misalignment of the two gear’s axis and applying the moment along that misaligned axis. Even with the worst case situation, the minimum factor of safety is still above 1.5, which is the requirement for all drivetrain components. After further analysis, I was able to reduce the total weight of the component and assembly by over 20%, even with the 30% increase in torque from last year’s car. This is not even factoring in the enormous frame weight loss from the packaging redesign.
Removing the reinforced members highlighted in purple significantly reduced weight in the rear subframe area which allowed for an elongation of the wheelbase without dramatically increasing overall weight
The assembly integrated into the rear inboard drivetrain of the car. Another benefit of this design was that the push pull cable and interior of the housing are accessible from the top and are able to be completely removed and serviced without taking out any other drivetrain components, which was a major issue with last year’s design.
For the first time in our team’s history, our four wheel drive system worked for the entire duration of the period it was installed on the car. This allowed us to fulfill our vehicle level goals of being both consistent and nimble, and in doing so we were able to secure a 1st place overall finish at SAE New York.
