Steer Torque Measurement Designs
Some ideas on measuring steer torque on a bicycle.
We are planning on measuring the steer torque the rider applies to control a bicycle. This will be used for human control model identification and for use in the necessary feedback loops required control the riderless bicycle. Measuring the steer torque is not trivial. This is because various models predict torques ranging in the 0-2 Nm (0-1.5 ft lbs) range with signal variations and reversals requiring +/- 0.01 Nm (0.01 ft lbs) in measurement accuracy. The range and accuracy are easily measured with modern torque sensors, but the fact that large moments can be applied to the fork and handlebars by the ground and/or rider introduces the problem of crosstalk. The forces and moments applied to the fork will corrupt the relatively small torque measurements as they can be hundreds of times larger in magnitude. With this in mind, we are trying to come up with a way to isolate the torque measurement to eliminate or minimize the crosstalk and get good, noiseless, accurate readings. The following are some basic designs we are working with:
1. Åström DesignThis is a sketch of what was designed for the UCSB instrumented bicycle and presented in a 2005 paper by Karl Åström et al. It uses an off-the-shelf axial load cell mounted between a floating handlebar and a bar extending from the steer tube. This seems to be a good design, but it would be nice to eliminate the handlebar bearings and the rod ends.
My professor, Drew Landman, from Virginia who I worked with designed force balances for wind tunnel testing at the LFST suggested a redesign that eliminates the bearings and replaces them with flexures.
2. Weir Design
David Weir designed a motorcycle steer torque measurement system in his 1979 technical report that also floats the handlebars on bearings but uses an off-the-shelf torque sensor instead. The sketch shows the basic concept. The handlebars are floating on bearings and the torque sensor connects the handlebars to the steer tube. He claimed that the design lacked low range resolution. Motorcycles can experience torques that are as high as 50 Nm according to some models.
3. Internal Stem Design
This is a design that we came up with when preparing our abstract on the topic. It is fundamentally the same as the Åström design but includes flexure elements instead of rod ends and is a bit smaller in scale.
4. Double Steer Design
This design separates the handlebar and stem's rotation axis from the steer tube and fork's rotation axis much the way many long wheel base recumbents or bakfiets are designed. The load cell is then place on the connecting rod. This design is is prone to slop in the steer mechanism.
5. Bearing-less design
Luke came up with this design and was able to eliminate the need for bearings. Two arms are clamped to the steer tube and a load cell is placed between the arms. The difference in this is that not all of the torque is transferred through the load cell, but maybe enough is that we can measure it.