The design of a Milner CVT using simulation based design of experiments

Research output: Contribution to conferencePaperpeer-review

4 Citations (SciVal)


The Milner continuously variable transmission (MCVT) is a traction drive transmission based on rolling contacts analogous to those found in angular contact roller bearings. The MCVT is applicable to very high power density applications or those with demanding packaging constraints, due primarily to the existence of multiple traction contacts on each of the rolling elements. The transmission has been modeled analytically and then Design of Experiments (DoE) techniques have been applied to create response surface model sets describing performance criteria over a wide range of design input factors. The response surface models have then been interrogated to find optimal solutions. The results show that it is possible to create a transmission with a significantly higher ratio range (10:1), compared with 5:1 that is typical for existing designs. The results also indicated that the traction patch spin factor could also be reduced over the existing designs and it is hypothesized that this should lead to an increase in efficiency. Finally, research into the stresses set up by the rolling contact indicated that stress is highest at the greatest reductions in ratio and that maximum stress appears to be proportional to the ratio range of the transmission. Hence a design optimization tool has been developed allowing engineers to rapidly optimize geometries of the MCVT to achieve desired trade-offs between ratio range, ratio position, transmission life/durability and transmission efficiency.
Original languageEnglish
Publication statusPublished - 20 Apr 2009
EventSAE 2009 World Congress - Detroit, Michigan, USA United States
Duration: 20 Apr 200923 Apr 2009


ConferenceSAE 2009 World Congress
Country/TerritoryUSA United States
CityDetroit, Michigan

Bibliographical note

SAE paper: 2009-01-1539


Dive into the research topics of 'The design of a Milner CVT using simulation based design of experiments'. Together they form a unique fingerprint.

Cite this