A novel fiber-fretting test for tribological characterization of the fiber/matrix interface

https://doi.org/10.1016/j.compositesb.2020.108535Get rights and content

Highlights

  • Novel technique for friction and wear characterization of fiber/matrix interfaces.

  • User defined stress, cycle length, and frequency explore friction dependencies.

  • Access to the tribo-surface enables Raman and other surface characterizations.

  • Adhesive and abrasive wear regimes are observed depending on interphase structure.

Abstract

Ceramic matrix composites exhibit excellent high temperature properties and are candidate materials for structural applications from jet engines to nuclear fuel cladding. The strength and deformation behavior of these composites are intrinsically tied to the properties of the fiber/matrix interface. This research introduces a novel technique to evaluate the fundamental friction and wear characteristics of these interfaces. A case study on SiCf/PyC/SiCm composites applies in situ fiber fretting to investigate the kinetic friction coefficient and cyclic wear characteristics of the interphase. The experimental methodology is presented in explicit detail to encourage other researchers to explore the capabilities of this technique. Testing was carried out as a function of fiber roughness, PyC thickness (≈10, 500, 1200 nm), cycle count (10, 100, 1000), and frequency (1, 5, and 10 Hz). The friction coefficient and wear behavior changed with cycle length due to a transition from adhesive to abrasive mechanisms. SEM fractography provided insights to tribological evolution from adhesive sliding to three-body abrasive wear. The structural disorder at the tribo-surface relating to graphitic content and dangling bond density was investigated using Raman spectroscopy. Properties are contextualized with comparison to uniaxial tensile behavior of the parent composite.

Introduction

Ceramic composites exhibit toughness via microcrack formation and subsequent deflection at the fiber/matrix interface. Under vibrational loading, these micro-cracks open and close causing cyclic wear of the debonded interface [[1], [2], [3]]. The frictional resistance and degradation of this interface govern the local load distribution, where over-loading and frictional pinching of the fiber can initiate composite failure [[4], [5], [6], [7]]. This is especially true in oxidizing environments [8,9]. Fiber push-out and composite hysteresis testing have been industry standards for evaluating the fiber/matrix interface properties responsible for global behavior [[10], [11], [12], [13], [14], [15]]. Fiber push-out tests, as well as micro-pillar compression, provide direct property extraction but are limited in their ability to characterize degradation over time. Hysteresis testing of mini-composites provides information on the macroscopic behavior but requires complex deconvolution of the stress-strain data to estimate the interfacial degradation [5,[16], [17], [18], [19]]. These estimates are then re-integrated into non-dimensional parameters and applied to the predictive models [4,6]. The fidelity of these models could be vastly improved with direct extraction of the tribological properties and responsible mechanisms. Given these characteristics, other parameters and assumptions in the model can be challenged with confidence in the underlying constituent behavior.

In recent years, accelerated evolution of small-scale mechanical testing has provided a variety of systems and expertise that enable unique and approachable experiments [[20], [21], [22]]. With scientific motivation and state of the art equipment, a new technique was developed that could supplement previous methods to provide more applicable friction data for constitutive models. The method, termed “fiber-fretting”, applies a typical scratch stage and modified indenter tip with integrated transducer to grab, debond, and cycle an isolated fiber segment. Control over the applied compressive load alleviates some uncertainties regarding residual stress and shear lag normally associated with fully constrained tests [18,23,24]. Furthermore, the process relies on a relatively simple combination of mechanical polishing and focused ion beam (FIB) milling techniques. This research is intended to be a simple method for probing constituent level tribo-properties of the as-fabricated composite, across a variety of environments.

Here, this methodology is explored with a case study on silicon carbide composites (SiCf/SiCm) with pyrolytic carbon (PyC) bond layer at the fiber/matrix interface. These composites express excellent high temperature strength, toughness, chemical stability, and irradiation tolerance making them preferred candidates for application in advanced nuclear and aerospace applications.

Tribological characterization aims to understand the interplay between the observed experimental properties and the responsible mechanisms. According to the law proposed by Bowden and Tabor [25], the friction force PX evolves by combination of adhesion force FA and ploughing force FP, that is,PX=FA+FPwhere FA depends on active bonding sites and shear strength of the contacting surfaces, and FP on surface roughness and hardness. Additional factors including wear debris, velocity, lubrication, and environment drive the associated wear mechanisms and resulting friction coefficient [25,26].

Adhesive wear is governed by local welding at the contacting surfaces and subsequent debond and transfer of that material [[25], [26], [27]]. Resistance increases when surface bonding is favorable, e.g., between similar materials, when dangling bonds are present, and in vacuum. Resistance decreases when the active bonding sites are passivated, for example when surfaces oxidize. Abrasive friction is common for hard ceramic materials and typically governed by asperity fracture and plowing that causes third body wear debris [27]. These particles exacerbate plowing and thereby increasing relative contribution of FP. Both friction coefficient and wear rates typically increase in abrasive conditions compared to adhesive.

In this study, the friction system consists of interfacing SiC surfaces with the PyC solid lubricant bond layer. The fracture location and PyC thickness (hPyC) relative to the max fiber surface roughness (Rmax) determines the degree of SiC on SiC asperity interaction. For hPyC/Rmax < 1, the sliding resistance is likely governed by SiC asperity shearing and plowing. For hPyC/Rmax ≈ 1 mixed shearing contribution from SiC and PyC is expected. For hPyC/Rmax » 1, failure can be adhesive at the SiC/PyC interface or cohesive within the PyC. The former suggests mixed shearing from PyC and SiC, while the latter is exclusively shearing of the PyC.

Tribological behavior of SiC and graphite are well studied across many material allotropes, couples, and environments [[28], [29], [30], [31], [32], [33], [34]]. As a brittle ceramic, SiC usually degrades following abrasive mechanisms like asperity chipping and surface cracking. The wear rate is dependent on grain size, texture, and additives, with typical kinetic friction coefficient (μk) values between 0.2 and 0.8 depending on environment. Graphitic materials, like PyC, provide desired lubricity due to weakly bound basal planes of the anisotropic sp2 ring structure. However, the relatively soft material is subject to structural damage from both mechanical and thermal loading. Degradation of the sp2 ring structure can introduce active carbon sites that compromise the lubricious quality [35]. Orientation of the basal planes also plays an important role relating to active site density at the tribo-surface [36]. Graphite-like materials are particularly sensitive to environment where both physio-and chemisorption of gasses and moisture can reduce the van der Waals bonding between the basal planes [31,35] Increasing temperature, from the working environment or flash heat generated by asperity contact, can promote gas and moisture absorption that helps maintain low friction by passivating these active sites [32]. However, as temperatures approach 500 °C, oxide and hydrocarbon volatilization become a major concern. With all of this considered, friction values for graphitic carbon have shown to range from 0.05 to 0.5 depending on environment [31].

Raman spectroscopy is commonly applied to understand the relationship between carbon structure and friction values [34,[37], [38], [39], [40]]. Raman is an excellent tool for carbon characterization with robust analyses developed to quantify myriad characteristic including crystallinity, sp2 fraction, structure clustering, and bond degradation. There are two prominent peaks in the Raman spectra of graphitic carbon, the D band appearing at ≈ 1355 cm−1, the G band at ≈ 1581 cm−1. The D band appears due to the A1g breathing mode and requires a defect for its activation. The G band originates from the E2g stretching mode and occurs at all the available sp2 sites, ring or chain [39]. For nano-crystalline graphite-like CVI PyC, the intensity of the D band relative to the G band (ID/IG) reveals information regarding the graphite cluster size and thereby degradation of the clusters. The G peak position provides insight to the amount of structural degradation as sp2 phonon modes soften with reduced ring structures. Ferrari et al. proposed a phenomenological trajectory that maps the expected G-peak shift and ID/IG from graphite to diamond in three stages [39]. Moreover, the G peak width can indicate possible surface structural disorder, such as variations in bond lengths and angles in sp2 clusters [41]. Given the spectra evolution for different stages of the test, conclusions can be drawn regarding the damage state of the PyC.

This experiment was developed in SEM vacuum at ambient temperatures for controlled investigation of the material specific properties and mechanisms. However, testing in ambient pressure or in artificially hydrated atmospheres is similarly possible.

Section snippets

Experiment design

A single fiber, tow, or woven composite can be polished along the fiber's longitudinal axis to expose the underlying fiber. It should remain embedded in the matrix and have at least one-half of its diameter polished away to minimize residual stresses and avoid complex stress states during testing. FIB milling is subsequently used to polish additional material off the top and to trench two locations that results in an isolated fiber segment with a semi-cylinder cross section. Fig. 1 shows a

Fretting test

Early testing with the conical tip on composite S-thick found that the conical tip splits the fiber segment as loading approached ≈40 mN. This limited the achievable stress and motivated exploration of cycle and frequency dependence at a fixed pressure. A successful conical tip test is shown in the vertical sequence on the left of Fig. 6. This tip-segment coupling allowed for tilting during translation, highlighted by the red lines. The test video (see supplementary video, V_S-thick-1) showed

Comprehensive evaluation

The experimental results demonstrate the capability of the fiber fretting technique to extract friction data and related mechanisms. The most pronounced trend is friction dependence on wear debris as it relates to adhesive and abrasive wear mechanisms. This is showcased clearly by composite H-thick that expresses both low and high friction regimes correlated with smooth and rough tribo-surfaces. Fig. 16A and B compares the adhesive and abrasive steady-state regimes for all composites except

Conclusion

A novel fiber fretting test was developed to probe friction properties and tribological mechanisms of continuous fiber composites. Details regarding experimental execution and analysis were laid out explicitly with intention for repeatability and technique advancement. The design was relatively simple, applying existing scratch techniques to a FIB milled surface structure. Four different SiCf/PyC/SiCm composite interfaces were evaluated in a SEM vacuum environment. Friction data was accurately

CRediT authorship contribution statement

Joey Kabel: Conceptualization, Investigation, Methodology, Formal analysis, Writing - original draft, Visualization. Thomas E.J. Edwards: Investigation, Methodology, Formal analysis, Supervision, Writing - review & editing. Caroline Hain: Investigation, Formal analysis, Writing - review & editing. Tatiana Kochetkova: Investigation, Formal analysis, Writing - review & editing. Darren Parkison: Investigation, Formal analysis, Writing - review & editing. Johann Michler: Supervision, Resources,

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was made possible through the collaborative efforts between UC Berkeley and Swiss Federal Laboratories (Empa, Thun). The first author and corresponding author would like to thank the DOE NEUP program [DE-NE0008460] for student and research support. The first author also wants to thank the Empa funding sources and the ThinkSwiss Scholarship for enabling and supporting this developmental research. The work would not have been possible without expert guidance from the Alemnis team

References (57)

  • G. Dehm et al.

    “Acta Materialia Overview on micro- and nanomechanical testing : new insights in interface plasticity and fracture at small length scales

    Acta Mater

    (2018)
  • D. Frazer

    Localized mechanical property assessment of SiC/SiC composite materials

    Composites Part A Appl Sci Manuf

    (2015)
  • M. Sebastiani et al.

    Depth-resolved residual stress analysis of thin coatings by a new FIB-DIC method

    Mater Sci Eng

    (Oct. 2011)
  • J. Xiao et al.

    Anisotropic friction behaviour of highly oriented pyrolytic graphite

    Carbon N. Y.

    (Dec. 2013)
  • X. Luo et al.

    Nuclear graphite friction properties and the influence of friction properties on the pebble bed

    in Nuclear Engineering and Design, Oct

    (2010)
  • D.W. Kim et al.

    Effects of sliding velocity and normal load on friction and wear characteristics of multi-layered diamond-like carbon (DLC) coating prepared by reactive sputtering

    Wear

    (2013)
  • T. Xing

    Disorder in ball-milled graphite revealed by Raman spectroscopy

    Carbon N. Y.

    (2013)
  • P.J. Bryant et al.

    A study of mechanisms of graphite friction and wear

    Wear

    (1964)
  • J. Xiao et al.

    Anisotropic friction behaviour of highly oriented pyrolytic graphite

    Carbon N. Y.

    (Dec. 2013)
  • P.K. Chu et al.

    Characterization of amorphous and nanocrystalline carbon films

    Mater Chem Phys

    (2006)
  • Y. Katoh

    Continuous SiC fiber, CVI SiC matrix composites for nuclear applications: properties and irradiation effects

    J Nucl Mater

    (2014)
  • S. Bertrand et al.

    TEM structure of (PyC/SiC) n multilayered interphases in SiC/SiC composites

    J Eur Ceram Soc

    (2000)
  • E. López-Honorato et al.

    Fluidized bed chemical vapor deposition of pyrolytic carbon - I. Effect of deposition conditions on microstructure

    Carbon N. Y.

    (2009)
  • K. Thompson et al.

    In situ site-specific specimen preparation for atom probe tomography

    Ultramicroscopy

    (2007)
  • Y. Zayachuk et al.

    Linking microstructure and local mechanical properties in SiC-SiC fiber composite using micromechanical testing

    Acta Mater

    (2019)
  • A.G. Evans et al.

    The physics and mechanics of fibre-reinforced brittle matrix composites

    J Mater Sci

    (1994)
  • N.P. Bansal et al.
    (2014)
  • W.A. Curtin

    Stochastic damage evolution and failure in fibre-reinforced composites

    Adv Appl Mech

    (1999)
  • Cited by (0)

    View full text