Design and multi-physics optimization of an energy harvesting system integrated in a pneumatic suspension

Highlights

• An energy harvester integrated in an air spring was designed, prototyped and tested.

• A multi-physical model of the energy harvesting device was defined and validated.

• A multi-physics procedure for optimal design definition was determined.

Abstract

This paper describes a research activity concerning the design and the development of an energy harvesting system integrated in a pneumatic spring for railway application to recover otherwise wasted energy sources from the train suspension vibration. The final scope of this research is to harvest energy and create a self-powered smart component capable of supplying useful information for the monitoring and the diagnostics of a vehicle or its subsystems.

Starting from a common air spring for a metropolitan train application, the boundary volume of the new device was defined by means of reverse engineering techniques. Exploiting commercial component, two alternative transduction mechanisms were evaluated to select the best one in terms of flexibility and functionality. The defined concept design was thus modelled and optimized by means of a multi-physical approach. The reiterative optimization process, based on the use of a specific cost function, led to define the optimized device layout. The effectiveness of the proposed device, in terms of power generation, was evaluated realizing a physical prototype and a test rig. The results of the experimental tests validated the design process.

Introduction

In the past decades, the research effort in energy harvesting (EH) from both academic and private sector has increased exponentially due to a plethora of possible applications. This increased interest is due to the convergence of several factors, with the most prominent being the enormous progress made in the field of semiconductors. It substantially reduced energy consumption and size of electronic devices leading to the widespread use of Micro Electro-Mechanical Systems (MEMS). In turn, the development and widespread adoption of these low-powered systems led to the birth of the age of portable electronics, miniature sensors and actuators that have a wide range of applications. In the industrial field, more specific use of these technologies concerns Wireless Sensor Nodes (WSNs), Structural Health Monitoring (SHM), Non-Destructive Evaluation (NDE) and more recently the Internet of Things (IoT). These applications are based on the use of wireless self-powered sensors with the aim of measuring and transmitting extensive data for monitoring of objects or activities like industrial processes, structures, machines, vehicles, environment, healthcare and traffic.

The railway industry is also engaged in the use of this kind of technology. Nowadays, the railway industry is moving forward at a fast pace and railway vehicles have changed from being an essentially mechanical system to one that increasingly includes electronics, computer processing and sensors [1]. Soaring passengers number worldwide and a shift to stricter regulations the railway industry focus has moved to improve capacity, punctuality, reliability and safety of the whole system. These challenges encourage the industry to invest in new technologies.

Safety is one of the most important attributes of railways compared with other means of transport. Modern technology allows to preserve, if not enhance, this characteristic i.e. reducing the cost of maintenance. An important improvement in this field can be reached by changing from a calendar-based maintenance to a condition-based strategy or a combination of the two.

Modern railway vehicles provide sophisticated monitoring systems for data acquisition and processing both for rolling stock and infrastructure status; they must be fitted with high-capacity communication buses and multiple sensors in order to allow advanced processing for data collection and management. Roberts and Goodall [2], building on the concept that the railway is a complex, distributed, closely-coupled system, describe four different types of monitoring system: infrastructure-based infrastructure monitoring, rolling stock-based infrastructure monitoring, rolling stock-based rolling stock monitoring, and infrastructure-based rolling stock monitoring.

Nigigi et al. [3] provide a complete review of the modern techniques used for condition monitoring of railway vehicle dynamics. Reviewed solutions include both infrastructure-based and rolling stock-based techniques, referring to both model-based techniques (i.e. Kalman filter and sequential Monte Carlo method) and signal-based techniques (band-pass filter, spectral analysis, wavelet analysis and Fast Fourier Transform). Authors remark the importance of sensors as a starting point for practical applications of train dynamics condition monitoring. Signals can be carried out either through the employment of track-based sensors or vehicle-based sensors.

Every signal carried from the instruments can be used, on its own or with other signals, to obtain information on vehicle or track. For example, axle box accelerometers can be used to measure shorter wavelength vertical irregularity.

From a practical point of view, whatever is the final goal of the sensors, on-board sensors and sensor networks imply the presence of wire cablings: firstly, to supply the electric power generated by the engine or stored in high-capacitance batteries; secondly to carry the signals from the field to the control room.

The proliferation of measurement points, especially on some outdoor components like the bogies, has led to an undesirable cabling architecture that often creates several operative problems in installation, reliability and maintenance. Furthermore, measurement instrumentations should be placed in a protected location in order to prevent false warning detection or disconnection.

Self-powered sensors could be a suitable solution to overcome this issue. In this context, EH generators coupled with wireless technology can be a viable option for railway applications. Most EH applications in railway industry refer to the infrastructure-side of systems. Indeed, several researchers have concentrated their studies on the EH from train-induced track vibrations [4], [5], [6]. Looking at train-side, a smaller number of EH devices can be found. Socie and Barkan in the 2008 [7] proposed a smart sensor device with the scope of monitoring a vehicle braking system. De Pasquale et al. [8] proposed a piezoelectric EH generator to power autonomous sensors for sensors network applications. Electromagnetic strategies to harvest energy in freight trains, where electricity is not available in all carriages, are described in [9] and, more recently, by Bradai et al. in [10]. All these applications and prototypes aim to fabricate a compact integrated platform, a sort of sand box, to be installed on the bogie or on the car body. These boxes are in addition to other components and systems and must find an appropriate place to be installed. In the context of a sensors network, the autonomous node must be properly optimized in dimensions, weights and packaging in order to have a negligible impact and no interferences with the main systems.

This paper focuses on a multi physics methodology to carry out an optimized energy harvesting system (EHS) integrated in a railway pneumatic spring whose main aim is to recovery otherwise wasted energy generated by the suspension vibration. Its main advantage consists in harvesting energy by means of exploitation of existing components. The innovative vision is the integration of a generator, electronic and sensing system in the existing and protected space available in the pneumatic spring to form a smart-spring. Two main objectives are pursued and described in the paper:

• the design and the development of an EH device that exploits commercial transducers to convert the mechanical energy into an electrical one;

• The compliance with the constraints due to the embed of the system in an existing volume.

The respect for geometric constraints due to the available space represents a strict design challenge in comparison with ad-hoc EH solution.

A commercial pneumatic spring with the same size of bellows used for the metropolitan trains’ suspension was used in this study. A reverse engineering process was adopted to determine the geometric boundaries and the available volume. After selecting a suitable transduction mechanism, a virtual prototype and a multi-physical model were realized. Exploiting this model and once the design variables were determined, an optimized design was carried out. Then the reference air spring was modified to allow incorporation of the designed harvester. The successive assembling phase allowed to realize the full-scale physical prototype that was later tested in order to verify the energy harvesting system. The experimental results highlighted the functionality of the EHS and validated the design procedure.

The paper is organized as follows: Section 2 describes the EHS embedded in an air spring, Section 3 describes the device architecture definition by means of CAD modelling, the main optimization process is discussed in Section 4, while the experimental activity and results are illustrated in Section 5.