Feasibility of degraded adhesion tests in a locomotive roller rig

Abstract

In railway applications, the testing of on-board components is necessary to optimize the efficiency of the systems and to allow high safety levels. In order to reduce the time and the cost of the testing phase, the use of dedicated test rigs is being increased. The current paper summarizes some studies for the realization of a full-scale locomotive roller rig. The main mechanical and control problems that arise in the design of this type of test rig have been highlighted, and in particular, the feasibility of tests with degraded adhesion conditions between the wheel and the rail is simulated.

A numerical model of a HIL scaled roller rig for simulation of wheel–rail degraded adhesion condition

“…The numerical control performance was evaluated in terms of angular speed error (between the simulated train model and the roller velocity) and torque estimation error (between the real torque and the estimated torque). The results show that the architecture proposed in [4] was robust in terms of angular speed error for both the scenarios while, as regards the dynamical imbalance disturbance, it was necessary to modify the estimation procedure. The proposed estimation strategy contains a multi-state nonlinear filter that consistently reduces the particular disturbance produced by the imbalance.…”

“…HIL testing may be performed considering different levels of integration of the tested system with the surrounding vehicle as in the example of Firenze Romito MI-6 test rig [3]. More recently, in the research Centre of Firenze Osmannoro, Trenitalia has built an innovative full-scale roller rig in which it is also possible to simulate degraded adhesion conditions in railway vehicle according to a control scheme previously described in [4]. The control scheme proposed requires the on line estimation of the vehicle motor torque (to reduce the time of the setting up phase) by means of an estimator strategy that uses the roller angular velocity and the tangential component of the contact forces (measured on the support).…”

“…The control scheme proposed requires the on line estimation of the vehicle motor torque (to reduce the time of the setting up phase) by means of an estimator strategy that uses the roller angular velocity and the tangential component of the contact forces (measured on the support). Within this research activity, a scaled roller rig was designed and realised, with the aim of investigating the feasibility and the calibration of different control layouts that may be used also for the full-scale version [4,5]. In [5], the control laws for the definition of the roller torques and speeds were developed by means of considerations based on a simplified model that considered the system dynamics only in the longitudinal plane.…”

“…The main contribution of the paper is the study of the complete three-dimensional dynamics of the hardware part of the test rig, in particular, the contact between wheels and rollers [6] was modelled by adapting a previously existing algorithm, described in detail in [7,8], initially defined for the wheel/rail contact model: the algorithm was then modified to simulate the contact between two revolute surfaces and to work with variable step solvers (that improve the numerical efficiency and robustness). The software part of the roller rig and the control laws of the roller motors were derived substantially from those described in [4], even if some improvements were necessary: first of all, since the roller rig described in that paper tests a whole vehicle, while the scaled version tests only one bogie, the controller has to include a virtual vehicle, in which one of the bogies is real and the other is simulated. Furthermore, the roller motor control laws had to be modified taking into account the scaling factor.…”

“…Furthermore, the roller motor control laws had to be modified taking into account the scaling factor. Since the same software environment (Matlab-Simulink) was used for the development of the real-time software of the scaled test rig (controller and virtual vehicle [4]), and the simulator of the hardware part of the test rig, the proposed study may be seen as a software in the loop (SIL) calibration of the control algorithms that will be directly compiled and implemented in the scaled rig.…”