Apparently the gyroscopic and caster effects are not the only mechanical forces working on a bicycle to keep it upright as it rolls along. According to a Cornell University based team a bicycle with a high center of gravity at the rear and low center of gravity up front can keep itself upright while rolling along a flat surface even if gyroscopic and caster effects are eliminated or cancelled. The explanation given is that once a suitably designed bicycle with the described characteristics, no caster and minimal gyroscopic effects, is set in motion the front end will always “fall” first and cause the bicycle to steer into the direction of the fall.
This is a striking example of how a commonly accepted engineering theory can turn out to be incomplete. It shouldn’t be a surprise. It happens in science all the time as old theories are found to be incomplete and new ones take their place. This incomplete aspect of accepted theories can delay certain types of innovation. Typically it takes a focused look by motivated engineers and scientists to tease out the most complete explanation and insights into the phenomenon in question. Much of this learning is lost as experienced engineer retire and are not considered for lecturing or teaching positions.
I would also like to note that appropriate depth of theoretical and applied knowledge is very important to create an engineering solution. Just knowing the correct equations to apply is not enough. If that were the case then the CAD designed and FEA modeled automobiles of today would never have recalls for design defects. An engineering team must see, feel and smell the results of their designs first hand in the laboratory to ascertain the limits of their engineering theories and tools. Only then can they correct the identified shortcomings and eventually deliver a well designed product.
For a more complete explanation see below.