Railway safety stays on track

A computer modelling program that may help prevent rail accidents by more accurately predicting vehicle behaviour is being developed by Manchester Metropolitan University and Phoenix Inspection Systems.


A computer modelling program that may help prevent rail accidents by more accurately predicting vehicle behaviour is being developed by Manchester Metropolitan University (MMU) and Phoenix Inspection Systems.

MMU’s rail technology unit and Phoenix, a manufacturer of ultrasonic testing equipment, have drawn on work by universities and rail companies throughout the world to create an optimised computer modelling approach that aims to be faster and more precise than existing systems.

Computer simulation programmes are widely used to evaluate problems such as rolling contact fatigue on railway vehicles.

The simulations predict the stresses that occur in the area the wheels come into contact with the rail.

The area, typically the size of a thumbnail, has to contend with a number of forces from the train at any one time.

Prof Simon Iwnicki, director of studies at MMU’s rail technology unit, told The Engineer that contact conditions are difficult to predict because of the high forces and the short time in which they pass a particular point on the rail, and because the material behaviour is highly non-linear at these forces. ‘All computer packages simplify matters to some extent to allow the analysis to be completed within an acceptable time but the effect of these simplifications is not well understood,’ he said.

‘As a result, different packages can produce very different results. Our research aims to determine the limitations of existing models and produce an optimised tool that will provide more accurate predictions across different test situations.’

The Hertz Analytical theory, developed in the 1890s by German physicist Heinrich Hertz, remains the basis of most prediction techniques. It took observations from classical theory of elasticity and continuum mechanics to summarise the way in which two objects in contact would behave under pressure.

Although new methods have been developed to improve the theory, the results can vary significantly depending on different variants.

MMU’s computational model intends to combine the best aspects from each of these methods to select approximations that will have a minimal effect on the final results.

Working alongside MMU, Phoenix has developed a scanner for use on a scale version of a wheel and rail that emulates the stresses that would occur on a real railway.

The results from the simulation will be analysed against the data created by MMU’s predictions in the university’s laboratories later this year.

Dr Chris Gregory, head of the transducer development team at Phoenix, said: ‘By analysing the quality of the ultrasound reflections from the trial rig while it is in operation, we can detect the pressure distribution, which indicates where strains are occurring.

‘These findings will be used to corroborate the accuracy of the computational model.’

The collaboration between MMU and Phoenix has gained support from the EPSRC through Rail Research UK.

The work has already generated interest from the industry and the university hopes this will pave the way for more research into making the UK railway system safer and more efficient.

Ellie Zolfagharifard