On the mechanical behavior of rubber springs for high speed rail vehicles. (October 2018)

• Record Type: Journal Article
• Authors: Pintado, P.; Ramiro, C.; Berg, M.; Morales, A.L.; Nieto, A.J.; Chicharro, J.M.; Miguel de Priego, J.C.; García, E.
• Is Part Of: Journal of vibration and control. Volume 24:Number 20(2018)
• Page Start: 4676
• Page End: 4688
• Publication Date: 2018-10
• Subjects: Rubber springs -- force relaxation -- creep -- smooth friction -- viscosity -- characterization -- simulation; Vibration -- Periodicals; Damping (Mechanics) -- Periodicals; 620.3
• DOI: 10.1177/1077546317732206
• Languages: English
• ISSNs: 1077-5463

Abstract

There are many engineering design problems that call for rubber components as the best solution. Vulcanized rubber has found its way into all sorts of devices, from the universal automobile pneumatic tire to the ubiquitous compliant bushing. Some high-speed rail vehicle suspensions make use of rubber, not only in the air spring itself, but also in the auxiliary spring. The mechanical characteristics of this component influence vehicle dynamics and, therefore, accurate spring models with which to conduct dynamic analysis would make for powerful design tools. Nevertheless, the mechanical behavior of rubber defies simple modeling on account of stress relaxation, creep, set, viscosity, internal friction, and nonlinear stress–strain relations. Despite the advances in the micromechanical understanding of these phenomena, as well as in the macroscopic modeling of rubber spring behavior, there is ample room for refinement, and this is precisely the goal of this paper. The mechanical behavior of a particular rubber spring for high speed rail vehicles has been characterized. The results reveal the necessary components of the model, and suggest the appropriate procedure for parameter extraction. Our model proposal consists of three elements in parallel: a nonlinear elastic spring; a "soft friction" element; and a Maxwell viscous component. The characterization procedure takes into account both stress relaxation and nonlinear elasticity. The proposed model accurately reproduces experimental results and may then be used with confidence in any type of numerical simulation. Nevertheless, for this statement to be true, the problem of "numerical softening" potentially induced by soft friction models should be resolved. The paper will show that a trailing moving average filter, seamlessly tied to the model, wipes out the softening effect.