Experimental investigation of friction in compliant contact: The effect of configuration, viscoelasticity and operating conditions

Highlights

• Compliant contacts were investigated using MTM in a ball-on-disc configuration.

• Kinematic conditions were studied using different values of the slide-to-roll ratio.

• Ball-on-disc configurations were studied in soft-hard, hard-soft and soft-soft configurations.

• The slide-to-roll ratio was found to affect rolling friction in compliant contacts.

• Configurations affect the viscoelastic response and thus the whole Stribeck curve.

Abstract

This work investigates the effects of kinematic conditions, configuration, viscoelasticity, and lubricant viscosity on friction in lubricated compliant contacts. Experimental data were also used to develop a numerical simulation capable of predicting fluid friction in compliant contacts. Mini Traction Machine (MTM) in the ball-on-disc configuration was used to successfully gain insight into the behaviour of compliant contacts, allowing the investigation of the mentioned effects. The findings have confirmed that viscoelastic effects are present in all configurations, being soft-on-hard (S/H), hard-on-soft (H/S) and soft-on-soft (S/S), where they seem to be more profound in the configurations using compliant discs. The experimental data also suggest that the slide-to-roll ratio affects rolling friction in all configurations which is contrary to current literature.

Introduction

Mechanical systems are an integral part of today’s ever-improving world, where one of the most closely monitored parameters is efficiency. These systems mostly transfer energy through contacts in which it is estimated to be lost up to 23% of the world’s total energy consumption [1]. This makes tribology an important discipline with the potential of significant savings in lost energy. The development of these devices entails the use of many new materials including polymers, which are extensively used for their simple design providing a range of different mechanical and tribological properties while being cost-efficient. These materials are beginning to appear in several applications such as bearings [2], tyres [3], wipers, gears [4], and sealings. Nevertheless, these materials are also extensively used in biomedical implants [5], contact lenses [6], joint replacements [7], and smart devices [8] as biocompatibility is one of their many properties.

This creates several tribological interfaces frequently entitled as compliant, soft, or isoviscous-elastohydrodynamic (i-EHL) contacts. A decisive factor of these contacts is that viscoelastic effects may occur. These effects can be negligible or could have the potential to increase or modify friction. A better understanding of the compliant contact's behaviour could lead to fine-tuning their behaviour and thus improving the comfort of wearing contact lenses, decreasing wear in artificial joints or improving traction of tyres etc. While numerical approaches have recently been applied to tackle the problem of lubricated soft contacts, compliant contacts have been predominantly investigated in experimental studies. Therefore, in particular, the literature on the latter will subsequently be reviewed.

Bongaerts et al. [9] examined the influence of roughness and the effects of hydrophobicity on lubrication using a Mini Traction Machine (MTM) in the S/S configuration using polydimethylsiloxane (PDMS). The influence of surface roughness and hydrophobicity was found to be negligible in the EHL regime. The increase of surface roughness, however, shift the transition of boundary and mixed lubrication regimes. Hydrophobicity also affected boundary and mixed regimes to a large extent.

The next research group investigating friction in compliant contacts was led by de Vicente. In one of his works [10], the authors combined both experimental and numerical studies in a rolling-sliding soft-EHL contact at a slide-to-roll ratio (SRR) of 50%. The numerical study of the i-EHL circular contact was used to derive a predictive equation of Couette and Poiseuille friction components. The Couette friction component arises from sliding in the contact and the Poiseuille friction component to the rolling friction. The experimental data were averaged using data from positive and negative SRRs. However, this methodology made it unable to directly compare the Poiseuille friction component with the predictions, thus only the Couette friction component was compared with theory, showing good agreement. Yet, it was still demonstrated, that the Poiseuille friction component cannot be neglected even in pure sliding contact, as this component is comparable to the Couette components, making the portion of rolling friction a significant part. In a follow-up study, de Vicente et al. [11] separated the mentioned friction components, being rolling (Poiseuille) and sliding (Couette) friction, using a novel experimental technique. Most of the tests had SRR set to 50% with some measurements carried out at lower SRRs, over a wide range of entrainment speeds from 4 mm/s to 1200 mm/s. Thanks to the authors’ novel approach it was possible to separate rolling and sliding friction. Rolling friction was found to mostly be a product of Poiseuille flow and elastic hysteresis. Also, an unexplained rolling friction component was observed for high reduced velocities. The author also suggested that friction was mostly independent of sliding speed due to surface adhesion, elastic hysteresis, and Poiseuille flow. Nevertheless, Couette flow was found to be proportional to the SRR. Both of the studies from de Vicente et al. [10], [11] used a H/S configuration.

Furthermore, Myant et al. [12] studied the influence of load and elastic properties on the rolling and sliding friction using MTM with a steel ball that was loaded against a PDMS disc (H/S configuration). Throughout the measurements, loads and soft discs were both varied, using three different discs each with a different order of elastic modulus. The influence of load on the isoviscous-elastic sliding friction coefficient was in quite close agreement with the numerical models for all three polymers. However, the author states that the friction coefficient increases with decreasing elastic modulus within the i-EHL regime. This in turn does not support the prediction made by de Vicente et al. [11], according to which the influence of the elastic modulus should be negligible.

Selway et al. [13] studied the influence of fluid viscosity and wetting on different scales of viscoelastic lubrication. The authors found that viscosity and wetting substantially affected the resulting tribological profiles. For smooth contacts, viscosity was expected to influence dewetting and squeeze-out at an interfacial scale, however, this effect was not as profound for rough contacts. In turn, rough contacts rather produced higher interfacial friction stimulating the viscoelastic hysteresis.

Putignano and Dini [14] found that viscoelasticity affects friction, where it adds more dissipation and reshapes the Stribeck curve. Also, the authors defined a new lubrication regime called visco-elastohydrodynamic lubrication (VEHL), which can be determined using a simple parameter Γ = h_hydro / δ_cr, where h_hydro corresponds to minimum film thickness in hydrostatic conditions and δ_cr to the solid penetration at critical speed when the viscoelastic response occurs. If Γ < 1, the contact operates in the proposed VEHL regime. Furthermore, Sadowski et al. [15] studied friction in a pure-sliding soft contact using three configurations: S/H, H/S and S/S, focusing on the effects of surface roughness and configuration. Configuration showed negligible effects, especially in the full-film regime after being corrected for the viscoelastic effects. Surface roughness affected friction at λ > 10 and film breakdown leading to asperity contacts beginning at λ = 3. Putignano [16] later proposed a numerical methodology capable of calculating both film thickness and friction coefficient in soft contacts including the viscoelastic effects. The proposed methodology showed good agreement with the experimental data from Sadowski et al. [15] in the full-film regime. Kim et al. [17] found that the loss modulus or loss tangent determines the friction coefficient of soft contacts depending on the lubrication regime. From this observation, a regression equation was further developed for dry and lubricated contacts in the boundary lubrication regime. Moyle et al. [18] showed a possible way for modifying the EHL friction in compliant contacts using patterns of compliant and stiff regions for the compliant specimen and proposed a new form of elastic hysteresis provided by the lubricating fluid in the contact.

To summarise, compliant contacts have been studied using experimental and numerical methods, where certain configurations were studied but were not directly compared. Merely, Sadowski et al. [15] investigated and compared all three configurations (S/H, H/S and S/S) limited to pure sliding conditions. The investigation of the effect of SRR on friction in compliant contacts has been studied by de Vicente et al. [10], [11], however, only for SRRs up to 50%. Concerning the limitations of the previous studies, this contribution aimed at the investigation of the effects due to different SRRs in compliant rolling-sliding contacts in different configurations and their underlying hysteresis response. Attention is paid to revealing the fundamental mechanisms and phenomena associated with the viscoelastic response of compliant materials.