Thin Layer Of Epoxy Research Articles

Effect of graphene coating on modified and pristine carbon fibers on the tribological response of carbon fiber epoxy composites

Addition of graphene nanoplatelets (GNP) as nanofillers in carbon fiber reinforced polymer composites improves a range of mechanical properties e.g. flexural, tensile and fatigue. Multiple studies have reported that this occurs through the formation of a carbon-enriched interphase with intermediate properties between the epoxy, graphene and carbon fiber. However, even though GNPs act as a solid lubricant and have exceptional strength and thermal conductivity, the wear performance and failure mechanism of GNP-incorporated composites are still not understood. The current study aims to investigate the effect of GNP-coated-pristine as well as GNP-coated-oxidized carbon fibers on the wear performance of carbon fiber reinforced epoxy composites manufactured through vacuum-assisted resin transfer molding technique (VARTM). It has been found that specific wear rate and coefficient of friction were lower for GNPs added composites compared to pristine laminates and it was the least in the case of composites in which an optimum amount of GNPs was coated on oxidized carbon fibers. Moreover, smaller temperature within the specimens has been found during the wear test in the case of GNP-added laminates owing to the higher thermal conductivity of the GNPs. Worn surface morphology has shown formation of thin epoxy layers at lower number of cycles for all the laminates. Fiber breakage, debonding resulting in clean fibers and a large amount of wear debris has been seen on worn surfaces of pristine laminates. However, GNP-added laminates have shown less wear debris and enhanced fiber-epoxy bonding. The wear response has been correlated with the size of the interphase formed between the fiber and the epoxy.

Assessment of Conductive Sites on Carbon Fiber Reinforced Polymer Plastics' Surfaces Using Electrochemical Methods

The components of automobile bodies have been manufactured primarily from steel since cars were first made. With the goal of energy savings, light-weighting is being adopted by substituting steel with a variety of light metals and composite materials. A DOE sponsored project in collaboration with PPG and Ford studied the use of automotive closure panels comprising an inner carbon fiber reinforced polymer (CFRP) sheet and an outer aluminum alloy (AA) sheet. The ends of closure panels are sealed with hem flange joints. A major concern is the formation of a galvanic cell within the joint wherein CFRP is the cathode that will accelerate the corrosion of the AA sheet. Various combinations of AA6xxx alloys with CFRP having two different orientations of carbon fiber, i.e., random and twill, are studied in this work. CFRP can only exert a galvanic driving force on a coupled AA panel by electrical connection at cut edges where sheared carbon fibers are exposed, or if carbon fibers are exposed on the CFRP panel surface by lack of coverage of the insulating epoxy matrix. It is believed that processing and curing variables lead to varied epoxy cover on the surfaces of CFRP. Electrochemical measurements such as the measurement of oxygen reduction limiting current density during cathodic polarization show that as-fabricated surfaces of CFRP panels exhibit some extent of electrochemical activity, indicating some lack of coverage of C fibers. It is of interest to characterize the epoxy cover on the CFRP surfaces in terms of the nature of the active sites (e.g., whether C fibers are totally uncovered or covered by a very thin layer of epoxy) and the density and distribution of the sites. A technique was developed to identify the exact locations of active sites on both random and twill CFRP by electrodeposition of small amounts of Cu. If the amount of deposited copper is small, the deposits indicate the location and number of the active sites on the CFRP surface. These deposits were also analyzed by serial sectioning using Focused Ion Beam/Scanning Electron Microscope. High resolution images of the deposit/epoxy interfaces provide a description of the nature of the defective/conductive regions on CFRP surfaces. Rotating disk electrode experiments were also performed to determine the oxygen reduction reaction (ORR) limiting current densities of CFRP at different electrode rotation rates i.e., different diffusion boundary layer thicknesses. These aspects are related to the overall mass transport in the electrolyte, which provides a deeper understanding of the kinetics on CFRP and the nature of active sites behaving as microelectrodes. The observations from the copper electrodeposition study and rotating disk electrode experiments will be correlated and the role of CFRP as a cathode contributing to the corrosion of AA in the galvanic couple will be presented.This work was supported by the U.S. Department of Energy through award DE-EE0007760, in collaboration with PPG Industries and Ford Motor Company.

Visualization of Water Distribution in Fuel Cell Microporous and Catalyst Layers with 3D Nanoscale X-Ray Imaging

As an efficient green energy technology, polymer electrolyte membrane fuel cells (PEMFCs) are rapidly growing in a wide range of applications. Development of high power-density PEMFCs is under investigation to further reduce the overall cost and footprint. However, in such operational conditions, excessive water production could restrict the transport of reactants due to flooding, and subsequently degrade the performance of the PEMFC. This makes the water management a major challenge and necessitates a good understanding of water distribution and transport within the porous layers of the PEMFC [1, 2].3D X-ray computed tomography (XCT) is a strong tool for visualizing the water distribution within porous media. Micro-scale XCT has shown great promise in understanding and visualizing the water behavior within the gas diffusion layer (GDL) substrate region possessing ~10-100 µm pores [3–6]. This resolution, however, is not sufficient to capture water distribution in microporous and catalyst layers (MPL and CL), which are predominantly composed of nano-scale pores. This work aims to explore a method of water capture and visualization within the MPL and CL of a PEMFC using a lab-based nano-resolution XCT (NXCT) with a spatial resolution as high as 50 nm. Benefitting from the non-invasive NXCT imaging technique, the same spot of a dry/wet material has been studied in Zernike phase contrast mode using a custom-built enclosed fixture with high X-ray throughput to maintain a constant humidity (100% RH) over the tomographic acquisition period. Water domains have been identified after alignment of wet and dry image data sets followed by subtraction of dry from wet data set. Figure 1a shows a 2D projection of the MPL on AvCarb GDL after vacuum-assisted water infiltration. A gold microsphere is fixed on the sample within a thin layer of epoxy to allow for accurate identification of the region of interest. Figure 1b shows a selected segmented 2D projection of wet MPL illustrating how liquid water clusters are distributed within the pores and the 3D volume rendering of structural features is displayed in Figure 1c. Using the above-described method, 18% and ~7% volume fraction of liquid water has been identified within the wetted MPL and CL, respectively. The methodology discussed here is a step toward deep understanding of water transport phenomena in CL and MPL materials dictating the strategies for materials design to prevent flooding under high current densities. Moreover, the set-up and methodology can be further optimized to visualize water domains under controlled temperature and relative humidity, to simulate the real operating conditions. Figure 1. XCT imaging of a wet MPL: (a) A radiograph of MPL on AvCarb GDL after vacuum-assisted water infiltration; (b) a selected 2D projection of MPL showing the segmented solid (green), liquid water (blue), and unoccupied pores (black) domains; and (c) 3D volume rendering of a subdomain in the MPL showing the distribution of liquid water within the pores of the structure (voxel size 32 nm in b and c). Acknowledgement Funding for this research was provided by the Natural Sciences and Engineering Research Council of Canada, Ballard Power Systems, Canada Foundation for Innovation, British Columbia Knowledge Development Fund, and Canada Research Chairs.

Fast response transient behaviour of a coaxial thermal probe and recovery of surface heat flux for shock tube flows

The magnitude of surface heat flux and its accurate estimation are very crucial owing to their applicability in impulsive thermal loads for high-speed flights. The correctness of such heat flux is associated with the implications from highly responsive thermal sensors in ground-based aerodynamic studies. In the present study, the performance effectiveness of an indigenous, in-house-made coaxial thermal probe (CTP) is validated against a standard silver thin film gauge (TFG) counterpart. The thermal probe has been fabricated in-house from two metal alloys; chromel (3.25 mm diameter and 10 mm length) and constantan (0.91 mm diameter and 15 mm length), by disposing one in the annulus of another coaxially, subsequently clubbed by a thin layer of epoxy in between them. Similarly, the thin film gauge is fabricated by putting silver paste (~1 µm thickness) on the pyrex substrate (3.25 mm diameter, 10 mm length). The fabrication process is followed by different calibration activities to determine the characteristics constants for sensors (sensitivity, thermal product value for CTP, and temperature coefficient of resistance for TFG). Both the probes are tested in a highly transient environment in a shock tube that generates high density, high pressure, and high enthalpy flows. The response time of the CTP and TFG are found to be 265 µs and 240 µs, respectively. The heat flux values estimated through temperature responses are having a nice match for CTP as well as TFG with a deviation in peak heat flux value within a range of ±2%. A similar trend and magnitude are also re-assured through numerical simulation within an uncertainty band of ±1.5%. The exhaustive study carried out in a shock tube has assured the performance and reliability of a CTP. Hence, it is recommended as an effective heat flux sensor for routine measurements in impulsive heat loadings.

Orbital Overlapping through Induction Bonding Overcomes the Intrinsic Delamination of 3D-Printed Cementitious Binders.

3D printing of cementitious materials holds a great promise for construction due to its rapid, consistent, modular, and geometry-controlled ability. However, its major drawback is low cohesion in the interlayer region. Herein, we report a combined experimental and computational approach to understand and control fabrication of 3D-printed cementitious materials with significantly enhanced interlayer strength using multimaterial 3D printing, in which the composition, function, and structure of the materials are programmed. Our results show that the intrinsic low interlayer cohesion is caused by excess moisture and time lag that block the majority of valuable interactions in the interlayer zone between the adjacent cement matrices. As a remedy, a thin epoxy layer is introduced as an intermediator between the adjacent extruded layers, both to improve the interlayer cohesion and to extend the possible time delay between printed adjacent layers. Our ab initio calculations demonstrate that an orbital overlap between the calcium ions, as the main electrophilic part of the cement structure, and the hydroxyl groups, as the nucleophilic part of the epoxy, create strong interfacial absorption sites. These electronic absorptions lead to several iono-covalent bonds between the cement matrix and epoxy, leading to significant improvements in tensile, shear, and compressive strengths as well as ductility of the 3D-printed composites. This is verified by our experimental data, which showed an average of 84% improvement in interlayer bonding. The upward augmentation of interlayer bonding helps 3D printing cementitious material to overcome their intrinsic limitation of weak interlayer cohesion, thereby mitigating/eliminating the key bottleneck of additive manufacturing in constructing materials.

Thin epoxy layers on native aluminium oxide: Specific ageing processes in mild conditions

Post-cured epoxy layers (EP, approx. 100 nm and 1 µm) of three amine hydrogen – oxirane ratios are prepared in dried air on natively oxidised aluminium substrate. A less crosslinked interphase forms as the consequence of adhesive interactions. Even though the layers exhibit very good durability in moist-warm air for up to 100 d, two stages of chemical ageing are revealed. During a short initial stage, some water absorbs firmly in the EP networks. Residual oxirane rings and secondary amine groups are shown to hydrolyse, while network chains remain intact. Marking the 2nd stage, dedicated autoxidation reactions are proven to simultaneously attack tertiary amine crosslinks and α-CH2 groups at phenylene ether units. These reactions destroy network chains, and the release of network fragments was established. Moreover, network destruction proceeds more effectively in 100 nm layers as the surface erodes slightly, even under the given environmentally mild conditions. That observation proves a specifically reduced durability for the interphase at the contact with native aluminium oxide.

Enhancement of piezoelectric vibration energy harvesting with auxetic boosters

This paper proposes a new concept to enhance the efficiency of the vibration energy harvesting via an intermediate booster. The boosters have auxetic structures and exert extra stretching strain in two perpendicular directions. The concept is tested on a conventional cantilever beam under the base excitation. The problem consists of a cantilever beam subjected to a body load at low frequencies. An auxetic substrate is bonded to the beam with a thin epoxy layer, and the piezoelectric (PZT) element is attached on top of it. Two different auxetic structures are investigated in this study. It is shown that employing these kinds of boosters can remarkably enhance the performance of the energy harvesting system. The harvesting efficiency is numerically evaluated in different load amplitudes and frequencies. A parametric study is then carried out, and effects of different geometrical design parameters of the auxetic boosters on the performance of the energy harvesting system are investigated. Comparing with the case in which the PZT is straightly attached to the cantilever, it is shown that adding such intermediate boosters at low-frequency range can increase the extracted power by factors of 3.9 and 7.0 for the two proposed geometries.

Distributed fiber-optic pressure sensor based on Bourdon tubes metered by optical frequency-domain reflectometry

We report a distributed fiber-optic pressure sensor based on Bourdon tubes using Rayleigh backscattering metered by optical frequency-domain reflectometry (OFDR). In the proposed sensor, a piece of single-mode fiber (SMF) is attached to the concave surfaces of Bourdon tubes using a thin layer of epoxy. The strain profiles along the concave surface of the Bourdon tube vary with applied pressure, and the strain variations are transferred to the attached SMF through the epoxy layer, resulting in spectral shifts in the local Rayleigh backscattering signals. By monitoring the local spectral shifts of the OFDR system, the pressure applied to the Bourdon tube can be determined. By cascading multiple Bourdon tubes and correspondingly attaching SMF sections (i.e., a series of SMF-modified Bourdon tubes), distributed pressure measurements can be realized. Three Bourdon tubes are employed to demonstrate the proposed spatially distributed sensing scheme. The experimental results showed that linear relationships between spectral shift and pressure were obtained in all three SMF-Bourdon tubes (i.e., at three spatial locations). It is expected that the proposed sensing device, the SMF-Bourdon tube, can be used in applications where distributed/multipoint pressure measurements are needed.

Enhancement of Magnetoelectric Conversion Achieved by Optimization of Interfacial Adhesion Layer in Laminate Composites.

We report the effect of epoxy adhesion layers with different mechanical or physical property on a magnetoelectric (ME) composite laminate composed of FeBSi alloy (Metglas)/single-crystal Pb(Mg1/3Nb2/3)O3-Pb(Zr,Ti)O3/Metglas to achieve an improved ME conversion performance. Through theoretical simulation, it was revealed that the Young's modulus and the thickness of interfacial adhesives were major parameters that influence the conversion efficiency in ME composites. In the experimental evaluation, we utilized three epoxy materials with a distinct Young's modulus and adjusted the average thickness of the adhesion layers to optimize the ME conversion. The experimental results show that a thin epoxy layer with a high Young's modulus provided the best performance in the inorganic-based ME conversion process. By tailoring the interfacial adhesion property, the ME laminate generated a high conversion coefficient of 328.8 V/(cm Oe), with a mechanical quality factor of 132.0 at the resonance mode. Moreover, we demonstrated a highly sensitive alternating current magnetic field sensor that had a detection resolution below 10 pT. The optimization of the epoxy layers in the ME laminate composite provided significant enhancement of the ME response in a simple manner.

Overall Thermal Conductance and Thermal Contact Resistance in No-Insulation REBCO Magnet

In no-insulation (NI) REBCO coils, thermal contact resistance (TCR) between winding turns has important influence on thermal stability of REBCO magnets under varying magnetic field or even in case of quench. Reduction of TCR between winding turns is one of the crucial factors to develop thermally stable magnets cooled by conduction. To avoid a delamination problem, epoxy is not usually used between winding turns and the significant TCR exists in NI magnets. Then, thin epoxy layers between coil surfaces and metallic conduction plates are used to enhance thermal conduction in conduction cooled magnets. To design thermally stable NI REBCO magnets cooled by conduction, TCR, induced by dry contacts of winding turns, should be properly evaluated. This paper describes measurements of overall thermal conductance using a model coil without conduction plates and epoxy layers from 10 to 40 K in a conduction test apparatus. TCR is evaluated using the measured overall thermal conductance and previously reported transverse thermal conductivity of a REBCO wire. In another model coil with conduction plates and epoxy layers, enhanced overall thermal conductance is measured and compared with finite element analysis results that include the measured TCR.

Evidence of a broadband gap in a phononic crystal strip

We experimentally demonstrate a very large ultrasonic band gap in a one-dimensional phononic crystal. The structure consists of periodic tungsten pillars fixed to a tailored silicon strip with a layer of epoxy. Combining local resonances and Bragg scattering, the gap ranges from 450kHz to 1250kHz, which corresponds to a gap-to-midgap ratio of 94%, and the attenuation exceeds 35dB with only three periods. Numerical calculations with the Finite Element Method are performed to support the analysis and provide a better understanding of the behavior of the structure. In particular, the role of the thin layer of epoxy is studied and is shown to have a strong influence on the dispersion. This phononic structure with a very large band gap can be considered as a new tool to design acoustic devices with high performances.

Quantum dot-based sensor layer in lightweight structures

Quantum dots can be used to detect, store and make visually apparent mechanical loading conditions. The fluorescent properties of the nanocrystals are selectively influenced by the injection of electric charges. By applying an external electric voltage, it is possible to suppress photoluminescence completely. If the quantum dots, as part of a functional layer system, are integrated in smart components, the integrated material system allows for energy-autonomous condition monitoring. We present for the first time a quantum dot-based system in a glass fiber-reinforced epoxy composite with a layer structure which is suitable for impact visualization. The quantum dots dispersed in poly (9-vinylcarbazoles) were applied on a PEDOT:PSS layer on an ITO-coated PET substrate. Silver electrodes were sputtered as a structured layer. For integration of the layer stack, which measured 25×25×0.1mm, in an epoxy composite, two process variants and sample geometries were used: a 2D curved component for hand lay-up and a plate for resin transfer molding. The epoxy composite components had a material thickness of 1.5mm and included eight layers of fiberglass cloth. The quantum dot-layer stacks were positioned either between the first two layers of glass fiber or directly at the component surface which had only a thin epoxy layer. By applying an external voltage, we suppressed the photoluminescence of the integrated quantum dots; the suitability of the coating system for integrated material sensors was evident.

An investigation of the tensile deformation and failure of an epoxy/Cu interface using coarse-grained molecular dynamics simulations

In this study, a coarse-grained model is developed to describe the interatomic interactions between a cross-linked epoxy and a copper substrate. Based on this model, the tensile deformation and failure of an epoxy/Cu bimaterial is studied. Attention is given to the microstructural evolution near the epoxy/Cu interface, and its effects on the overall stress–strain behavior of the bimaterial. It is found that, under uniaxial strain, plastic deformation in the epoxy/Cu bimaterial is localized to a thin layer of epoxy next to the epoxy/Cu interface. This discovery enables the definition of an interfacial zone. Consequently, a cohesive zone model at the continuum level could potentially be constructed where the separation of the interface is given by the stretching of the interfacial zone computed from the coarse grain simulations.

Reinforcement role of carbon nanotubes, Part I: High‐performance epoxy resin

The reinforcement role of single‐walled carbon nanotubes (SWCNT) in a high‐performance epoxy resin is studied in this article, with the emphasis on the deformation and load transfer mechanism. Tensile tests show that introducing carbon nanotubes (CNTs) into the epoxy resin changes the tensile strength of the material, and that the increase of Young's modulus is in good agreement with the increasing of glass transition temperature (Tg) measured by differential scanning calorimetry (DSC). Residual strains in CNTs induced during curing process have been investigated both experimentally and theoretically. Raman spectroscopy study shows that load transfers more efficiently in tension than in compression. The Raman shifts of deformed CNTs under repeated loading indirectly reveals that the load can be further transferred via friction after debonding fully occurs. Thin epoxy layers were observed on the pull out ends of CNTs, which implies the existence of the immobilized epoxy layer between CNTs and bulk epoxy. POLYM. COMPOS., 36:2212–2219, 2015. © 2014 Society of Plastics Engineers

Heat dissipation performance of metal-core printed circuit board prepared by anodic oxidation and electroless deposition

An aluminum anodic oxide layer was applied as the insulation layer of metal-core printed circuit board (MCPCB). Copper electroless- and electro-deposition were conducted on the insulation layer to form a copper circuit layer. In the present work, thin epoxy layer was coated on the aluminum anodic oxide to secure the electrical insulation between aluminum alloy substrate and copper layer. The MCPCB with the aluminum anodic oxide layer showed a good break down voltage and dielectric strength. The thermal resistance with respect to the thickness of oxide layer was measured using a thermal transient method. The thermal resistance of the MCPCB was increased by increasing the thickness of aluminum anodic oxide layer, and a lower thickness of aluminum anodic oxide was effective to reduce the LED junction temperature.

TEM Study of Local Conduction Mechanisms in Model Specimens of Ag-Based Conductive Adhesive

The conduction mechanisms of silver (Ag)-based conductive adhesives, which consist of metallic Ag particles and an epoxy resin, are discussed on the basis of transmission electron microscopy (TEM) studies and local conductivity measurements of model specimens. A small electrical current was observed in the model specimens, in which a thin epoxy layer was sandwiched between Ag electrodes. TEM observations demonstrated that the irreversible changes in the current versus voltage characteristics occurred along with microstructural changes in the Ag electrode, which was subjected to a large electrical current. Calculated equipotential lines indicated a considerable change in the electric field distribution near a small horn that was formed in the surface of the Ag electrode. The results provide useful information for understanding local conduction in the cured adhesive.

Nanotube attachment for prevention of interfacial delamination

A new approach to suppressing interfacial delamination in composites has been investigated. It involves growing strongly attached nanotubes on the surface of the core phase prior to matrix infiltration. Unusually durable interfaces between epoxy and graphite have been demonstrated using this technique. Two types of graphitic core materials have been studied: complex cellular foams having open-interconnected porosity and highly oriented pyrolitic graphite (HOPG) providing a model flat interface. When untreated foam is infiltrated with epoxy, the resulting composite is brittle, and shatters before 10% compression. However, when carbon nanotubes (CNTs) are grown on the foam prior to epoxy infiltration, the specimen becomes pliable, and visibly flattens out rather than fracturing. Model studies on a flat graphite–epoxy interface were performed by joining two HOPG specimens with a thin layer of epoxy, and testing the flexural response of the ‘seam’ using the three-point bend test. The untreated HOPG sandwich fails easily, whereas nanotube-attached HOPG sandwich shows an over three times increase in flexural load-carrying capacity, close to that of seamless monolithic graphite having identical dimensions. Microscopic evaluations of fractured interfaces indicate that, in all geometries, CNT grafting prevents delamination at the graphite–epoxy interface, and forces any crack(s) to propagate through the graphitic phase. This added inter-laminar strength and toughness can be related to the hierarchical morphology of the interface created by CNT attachment, and unprecedented composite structures can be envisioned.

Electrophoretic Deposition of YAG Phosphor on SMD LED

The uniform YAG phosphor layers were deposited onto the ITO glass substrates by electrophoretic deposition (EPD) technique. The more stable deposition condition was found when the electrolytic suspension was stirring. White SMD LEDs were easily achieved by using the directly EPD method, i.e. it is easy to achieve a white light point. For two white EPD LEDs, with and without thin epoxy layer, the chromaticity coordinates were x = 0.311, y = 0.29 and x = 0.311, y = 0.30, respectively. Although there has a uniform light shape for the EPD white LED with a thin epoxy layer, however, the lower luminescence efficiency was found due to the reduction of blue light extraction.

108 2個の異方性半無限体の接合層に存在する2個のき裂周囲の応力拡大係数について(材料力学(2))

Woven composite materials are made of epoxy layers reinforced by textile, and thin epoxy layers that bond the reinforced layers. In the present paper, the composites are modeled by an isotropic epoxy layer and two reinforced elastic half-planes that have anisotropic elasticity. A crack is situated in the layer parallel to the interfaces, and stress intensity factors are solved under normal tensile stress applied perpendicular to the crack. Fourier transformations are utilized to reduce the boundary conditions with respect to the crack to two pairs of dual integral equations. To solve the equations, the differences in crack surface displacements are expanded to a series of functions that are zero outside the crack. The unknown coefficients in the series are solved by the Schmidt method so as to satisfy the conditions inside the crack. Numerical calculations are carried out for some crack configurations.

Effect of Surface Morphology on Crack Growth at a Sol-Gel Reinforced Epoxy/Aluminum Interface

The Boeing sol-gel conversion coating (Boegel-EPII), derived from an acid-catalyzed aqueous solution of organofunctional silane and zirconium alkoxide precursors, is being used as an adhesion promoter for adhesive bonding and painting applications in the aerospace industry. A unique advantage of the sol-gel process is that strong and durable bonds are produced without the hazardous chemical usage and rinse-water requirements of conventional anodizing or etching processes. In this study, a fracture mechanics method was used to investigate the adhesion properties of sol-gel-reinforced epoxy/aluminum joints. The Hugh Brown asymmetric double cantilever beam (ADCB) wedge test was employed, which allowed the measurements of the critical energy-release rate, subcritical crack-growth kinetics, and threshold energy-release rate on a single sample in a reasonably short period of time. These experiments were carried out with aluminum substrates on which the surface morphology was systematically varied by polishing, sanding, grit-blasting, and chemical etching. X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM) were used to identify the locus of failure. The surface morphology of the substrates was characterized with SEM, optical profilometry, and spreading kinetics. The macrorough structures drive the crack to within a thin epoxy layer close to the polymer/metal interface, which enhances the initial strength of the sol-gel-reinforced interface. The microroughness of the substrate is, however, more effective than the macroroughness in enhancing the durability. Lastly, an attempt has been made to correlate the energy-release rate with the fractal dimension for sol-gel-reinforced joints with macrorough substrates.