Experimental and modelling methods for high-accuracy performance characterisation of chain drives

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

In recent years there has been a growing interest in measuring and predicting the performance of chain drives in applications such as elite cycling and industrial drives since improvement in the life-time performance of transmission promotes efficient use of finite energy sources. This thesis investigates and develops testing and modelling methods for achieving high-accuracy estimation of transmission power efficiency of chain drives.

At present many test rigs used in industry are not able to replicate real-world loading and have limited accuracy in torque measurements. To address these limitations a dynamometer test rig was developed at the University of Bristol which has the capability of testing in representative boundary and loading conditions. This thesis contributes to the development of this dynamometer rig with new testing protocols and applying advances in data processing.

A significant area of uncertainty in measurements on the test apparatus was the power losses in the support bearings on the shafts. To isolate the performance of the transmission system, the friction in the bearings must be accurately compensated. Analytical methods from the literature were evaluated and found to poorly predict frictional moment in light lubrication regimes at low operating range relative to the bearing rated loads and speeds. A purpose-built bearing test rig was developed to test identical bearings in similar conditions, and empirical results used to compensate bearing friction in the dynamometer test rig. The remaining uncertainty of the power loss measurand was determined, encompassing the combined uncertainty of measured variables across two test rigs.

High frequency chain dynamics from polygonal action and chain span resonance were measured in torque data on the test apparatus. Tested transmissions were shown to have unique vibration signatures across the test envelope. Chains with smaller pitch were seen to have reduced measured dynamics. On a bicycle, these measured disruptions are sensed as vibrations in the crank by a rider and smoother transmissions are more desirable for maximal effort exertion.

In power loss measurements, a transmission with a cyclic driving torque profile incurred more losses when compared with the same transmission at constant torque. Results showed the greater the variation about a mean torque, the greater the transmission losses. This is important because a cyclist delivers cyclic torque during cycling and until now test rigs have applied a constant torque based on the assumption that this gives the same result. Results also demonstrate that a cyclist with smoother pedal action will incur reduced transmission losses.

The primary application of this research has been to inform decision making in the development of a novel small-pitch chain drive by Renold Plc for British Cycling at the 2020 Tokyo Olympics.
Date of Award3 Oct 2023
Original languageEnglish
Awarding Institution
  • University of Bristol
SponsorsEPSRC RCUK, Renold PLC & British Cycling Federation
SupervisorStuart Burgess (Supervisor), Jason M Yon (Supervisor) & Robert Wragge-Morley (Supervisor)

Keywords

  • Losses
  • Efficiency
  • Performance
  • Chain
  • Dynamometer
  • Friction
  • Cycling

Cite this

'