|
romac |
|
University of Virginia Rotating Machinery and Controls Laboratory P.O. Box 400746 122 Engineer’s Way Charlottesville, VA 22904 (434)924-3292 (434)982-2246(fax) |
|
Rotating Machinery and Controls Laboratory |
|
BALOPT
Students: Guoxin Li, Coston Untariou, Bin Huang Advisors: Paul Allaire, Zongli Lin, Daili Fujimura, Pradip Sheth Funding: ROMAC
Industrial members are interested in new, innovative balancing procedures. We have been working on new methods that have resulting in new balancing codes developed by Bin Huang and Guoxin Li. There are four specific balancing methods being developed. They are influence coefficient method optimization which employs 1) least squares balancing, 2) min-max balancing, 3) optimum stochastic balancing which takes into account a known distribution of uncertainty in the measured influence coefficient data, and 4) a balancing method to take into account uncertainty in measured influence coefficient data when the distribution of uncertainty is not known.
The optimum balancing code BALOPT, Release 1.1, has been released to member companies in November 2005. BALOPT, release version 1.2, is expected to be available to members in March 2006. The highlights of the new BALOPT are the two new balancing methods that take the system uncertainties into consideration and the advanced weight splitting function. The weight splitting function enables users to split the calculated weight solution at multiple holes with the total number of weight splitting holes pre-specified. This advanced weight splitting function is already available in a BALOPT beta version. Members are welcomed to request it from the developers.
Testing of BALOPT continues in the UVA ROMAC Labs on a three mass rotor. Additional balancing analysis on industrial rotors is also being carried out with Mitsubushi Heavy Industries with Mr. Daiki Fujimura, who was a Visiting Researcher in our lab for 2005. We are applying BALOPT to a large MHI turbine generator set with excellent results, better than conventional influence coefficient methods, using the different BALOPT balancing methods described in this article.
Finally, a different optimum minmax balancing method of splitting the balancing objectives into critical locations and other locations with less severe constrains has been carried out. Dr. Costin Untariou of UVA is conducting research into this method. This has been applied to a GE turbine generator set with excellent results compare to conventional least squares methods.
|