Suggested Citation:
Moore, Jason and Mont Hubbard (2019)

Expanded Optimization for Discovering Optimal Lateral Handling Bicycles

Previously, we introduced a method of optimizing four primary geometric parameters of a bicycle's design to maximize its lateral handling qualities. Here we expand that method to optimize over all of the geometric and inertial parameters in the linear Whipple-Carvallo bicycle model. To ensure physically realizable bicycle designs we include 7 equality constraints, 21 inequality constraints, and lower and upper bounds on each free optimization parameter. This improves over the prior work by expanding the search space with many more parameters and the guarantee of physical realizability. We present four bicycle designs discovered by the optimization procedure that have optimal later handling qualities. The bicycles are similar in design to familiar bicycle designs but are not generally self-stable and exhibit unusual characteristics such as large positive and negative trail, large size relative to the rider, and minimal steering inertia. The method is a useful tool for generating atypical bicycle designs that exhibit desired dynamical qualities and could be broadly applied to other vehicle designs.