Non-linear modelling of vortex induced lock-in effects on circular chimneys

• Published in: Journal of wind engineering and industrial aerodynamics 2020-07, Vol.202, p.104201, Article 104201
• Main Authors: Arunachalam, S., Lakshmanan, N.
• Format: Article
• Language: English
• Subjects: Circular chimneys-modelling-vortex shedding-full-scale data
• ISSN: 0167-6105; 1872-8197
• DOI: 10.1016/j.jweia.2020.104201

The paper presents results of an extended study of the model developed earlier by the authors for across-wind response of tall circular chimneys to vortex shedding. A non-dimensional parameter,“fact” is used to suitably decrease the spectral bandwidth of the generalized across-wind load spectrum, implicitly accounting for an increase in load correlation. This paper discusses similarity between the parameter “fact” of authors’ model and the ratio of the aerodynamic damping coefficients, (ka/ka,0) in Vickery’s model, with respect to enhancement of correlation of force/response due to vortex shedding under lock-in condition. However, in Vickery’s model, as recommended in CICIND code, while the variation of (ka/ka,0) with increasing normalized amplitude follows a negative curvature, the corresponding variation of “fact” exhibits a positive curvature. Recent wind tunnel experimental results by Lupi et al., showed a positive curvature for the variation of (ka/ka,0) with increasing normalized amplitude. This lends support for authors’ observation. With a suitable scaling factor, an excellent agreement is found between the authors’ results and Lupi’s experimental results. The suggested method was applied to 27 full scale steel chimney data, referred in Lupi’s paper and the response predictions show a good comparison within ±20% limits with full scale values. The usefulness of the mass damping parameter, (2me(ρD2)η) for predicting the across-wind response is also discussed. •Parameters “fact” by authors and (ka/ka,0) by Vickery are shown to be similar.•However, “fact” shows a positive curvature with (σ/D), distinctly from results by Vickery.•Authors’ observation is supported by experimental results by Lupi et al., and Cheung and Melbourne.•Application of authors’ method to 27 full-scale chimney data.•Use of new mass damping parameter, ((2me/ρD2)∗√η) is discussed.