Malcolm Scott

I grew up in the Scottish Borders, but have been in Sheffield for the last nine years where I completed my Meng and PhD. I now work as an RA as part of the engineering Nonlinearity programme grant.  When not embroiled in structural dynamics I mostly like to spend my time out in the Peak District.

Research interests:

My research interests include Bayesian sensitivity analysis and system identification techniques, Validation of nonlinear models, Bayesian V&V techniques and Nuclear power.

Current research:

I am currently looking at the Validation of nonlinear models. Nonlinearity makes model validation far more difficult. One reason for this is that, in nonlinear systems, small quantitative changes in inputs or parameters can cause drastic qualitative changes in the dynamic response, known as bifurcations. I am currently investigating a three story building structure with a harsh nonlinearity between the top two floors. Upon changes in forcing frequency the structure displays both saddle node and period doubling bifurcations. A model of this system has been developed and the model behaviour is now being compared in detail to the structure in order to assess which features are suitable for model validation.

I am also looking at Bayesian V&V techniques. The general idea behind these is to be able to make model predictions which take into account all possible sources of model uncertainty. While there are techniques for this which have been developed in the statistics community there are still challenges to applying them in a real world structural dynamics context.    

Below is a sketch of the three story building test rig, with a harsh nonlinearity between the top two floors. The plot shows the response of the strucutre to a periodic forcings at the same amplitude but very slightly different frequencies – 52.920Hz and 52.921Hz. There is a drastic difference between the two responses due to a saddle node bifurcation.

Publications:

Journal papers

  • Scott M & Worden K (2015) A bee swarm algorithm for optimising sensor distributions for impact detection on a composite panel. Strain, 51(2), 147-155.
  • Scott M, Green PL, O'driscoll, D, Bateman M, Worden K & Sims ND (2014) Sensitivity analysis of an Advanced Gas-cooled Reactor control rod model. Journal of Nuclear Engineering and Design (under review)

Conference papers

  • Scott M & Worden K (2015) A bee swarm algorithm for optimising sensor distributions for impact detection on a composite panel. Proceedings of ICEDyn 2013, International Conference on Structural Dynamics - Sesimbra, Portugal, 17 - 19 June, 2012  
  • Scott M, Green PL, Grant-Wilson C, Bateman M, Worden K & Sims ND (2014) Sensitivity analysis of an Advanced Gas-cooled Reactor control rod model. Proceedings of ISMA 2014, Conference on Noise and Vibration Engineering - Leuven, Belgium.
  • Scott M, Sims ND, Worden K & O'driscoll, D Control rod mechanism condition monitoring. . Proceedings of SMIRT 23, Manchester, 2015
  • Scott M, Tiboaca D,  Barthorpe R.J, Wagg DJ, Worden K. On the Validation of Nonlinear MDOF System Models. Proceedings of ISMA 2016, Conference on Noise and Vibration Engineering - Leuven, Belgium.