Determination of Thermal Stress and Elongation on Different Ceramic Coated Ti-6Al-4V Alloy at Elevat

Abstract

Keywords: Finite element analysis; Ti-6Al-4V; thermal stress; thermal expansion; ceramic coating.

1. Introduction

The light metallic alloys working at elevated temperatures and in aggressive environments having potential for corrosion and tribo-corrosion are utilized often in high-tech industrial applications. Only fourteen elements can be defined as light metals because of their density values lower than 4500 kg/m3. Among these, titanium, aluminum, and magnesium are utilized widely as significant design materials [1]. In today’s industrial world, the term of lightness has become extremely crucial due to environmental concerns. Concordantly, the issue of fuel efficiency has been a key point for the automotive and aviation sectors. In this point, it is obvious that structural parts used in the component design must not only have low-density levels but they must also *

Corresponding Author: e-mail: caginbolat@itu.edu.tr, tel: (212) 293 13 00 2013

exhibit sufficient mechanical properties under real service conditions. Compared to other light metals, Ti alloys have a high specific strength, creep resistance, high melting point, corrosion resistance, superior biocompatibility. Hence, from the aerospace industry (landing gears, engine and wing parts) to the health sector (medical implants), many different application areas have been attracted by these versatile materials. [2-4]. Also, titanium displays allotropic transformation at 1115 K (this case is also called as polymorphism) and if the temperature of the metal is below this value, a close-packed hexagonal crystal structure is present, but, above this value, bodycentered cubic crystal is seen [1]. Depending on their phase conditions, Ti alloys usually are classified in three different categories namely α, β and α-β alloys, and mechanical/physical behaviors of the selected alloy are affected hugely by the phase structure [5-7]. Apart from their diversified usage in several real applications, Ti alloys are also chosen as base materials for different coating applications in order to obtain long service life against high temperature, wear and corrosive media [8-11]. With the intention of examination of physical, chemical, and mechanical properties of the coated Ti specimens, many coating techniques like nickel plating, ion nitriding, hard ceramic cladding, boronizing, physical vapor deposition (PVD), chemical vapor deposition (CVD) and high velocity-oxygen fuel (HVOF) have been tried to apply by lots of researchers in the technical literature [9, 12-16]. Moreover, in recent times, investigations about fatigue characteristics, high-temperature oxidation, wear and tensile performance of coated Ti alloys have begun to rise in a noteworthy manner [17-19]. If the coating applications of titanium are probed on a sectoral basis, many different examples can be given according to usage purposes and performance requirements. In the automotive industry, Ti-based exhaust pipes, silencers, exhaust valves, and suspension springs are used frequently and these parts are coated with some hard ceramics to block wear/corrosion threat or with the same alloy by means of cold spray method to perform repairing [20]. In analogy to automotive, titanium fuselage parts, propellers, wing elements and engine parts are preferred in aerospace implementations, however, all these elements should be coated against extreme conditions like heat and ice carrying high damage potential [21]. As for biomedical coating efforts, investigations focusing on dental implants, joint prostheses and bone fracture fixation plates have been increasing so as to improve surface performances of titanium based materials [22-23]. Lately, some new innovative coating investigations have also tried on Ti-6Al-4V metal by different research teams. For instance, Yan et al [24] focused on laser cladding Ti-Ni/TiN/TiW+TiS/WS2 composite coating of Ti-6Al-4V to improve wear performance. Bui et al [25] studied antibacterial coating of Ti-6Al-4V using silver nano-powder mixed electrical discharge machining and they showed that silver content decreased with the increasing depth of coat. In another effort, Prakash et al [26] reported that HA-TiO2 composite coatings could be tailored on Ti-6Al-4V base material by means of plasma spray deposition for orthopedic applications. Ding et al [27] Ta 2O5 based coating created on biomedical Ti-6Al-4V alloy through magnetron sputtering and emphasized that formed coating layers were smooth, dense and homogenous. Liu et al [28] examined the effect of heat treatment on microstructure and mechanical properties of TiC/TiB composite coatings on the Ti-6Al-4V. The researchers indicated that microhardness and fracture toughness enhanced with post heat treatment. Almeida et al [29] studied the effect of nitrogen in the properties of diamondlike carbon coating on Ti-6Al-4V substrate and found that incorporation of nitrogen atoms affected the tribology of the coating films in a positive manner. Although a large number of scientific papers can be found comfortably about coating procedures for Ti alloys, and tribological or corrosion performance at dissimilar temperatures of coated Ti alloys, there is a lack of studies conducted on the thermal stress analysis. This circumstance is considerably significant at high temperature due to completely different physical and mechanical properties of titanium base materials and coating materials such as hard oxides, nitrides, borides, different metals or composites. In particular, for providing the design standards without any setback, correct determination of thermal stress difference between Ti alloys and engineering ceramics is very critical.

In this study, our main aim is to figure out thermal properties and thermal stress distributions of hard ceramic coated Ti-6Al-4V alloy by means of the finite element analysis (FEA) at high temperatures. Thanks to detailed FEA, the difference between the base material and coating materials can be interpreted and analyzed in detail. In addition, this paper introduces a comprehensive comparison between different hard coating materials, coating thickness values and coating models to find out their probable effects on the thermal stress features.

3. FEA Results

Firstly, in the FEA, thermal stress distribution on the ceramic coated models was observed. As a consequence of the comparative evaluation, it is very clear from the Figure 3 that TiB 2 coated Ti-6Al-4V samples are exposed to higher stress levels than the others regardless of coating layer thickness, ambient temperature and coating model. This circumstance can be attributed to the high modulus of elasticity values of TiB2 between 373 K and 873 K. Concordantly, the highest Von Mises stress (σvm) value of 3729.5 MPa belongs to full-surface TiB2 coated sample with coating thickness of 800 µm at 873 K even though the lowest σ vm value of 230.7 MPa is measured at 373 K for single-surface AlN coated sample having 800 µm coating thickness. Typically, for all types of coating materials, single-surface coated samples display a downward tendency in the thermal stress values owing to decreasing rigidity and escalating thermal expansion coefficients of ceramics depending upon increasing coating thickness. Compared to the single-surface coating model, an upward trend is observed for samples having full-surface coating as a result of rising volume fraction of the hard-ceramic coated zones in the sample. Besides, if effects of the ambient temperature are examined, it can be claimed that since the base material of Ti-6Al-4V has better expansion ability, which fluctuates between 9 and 9.7×10-6 1/K, than used hard coating ceramics. Therefore, calculated σvm values increase for all

models in the same thickness because of the poor elongation ability of titanium (limited by the coating material).

Figure 3. Comparison graphs for σvm values at various temperatures. Figure 4 given below shows thermal deformation behavior of simulated models by taking maximum and minimum stress distributions into account. From the simulations, even though single-surface coated samples deform more visibly in the direction of expansion of Ti-6Al-4V (Figure 4a-f), full-surface coated samples remain less deformed due to ceramic coating layers covering the Ti alloy all around (Figure 4g-l). At 373 K, maximum stress concentration accumulates in top right corner on the TiB2 layer with thickness of 400 µm (Figure 4a), however AlN coated sample having 800 µm coating thickness bears the lowest stress concentration in the same location (Figure 4b). As the temperature rises to 573 K and 873 K respectively, observed σvm trend continues obviously, and the highest and the lowest σvm concentration are observed for TiB2 and AlN coated samples (Figure 4c-f). On the other hand, when the full-surface coated samples are analyzed, it is true to say that stress concentration is localized at bearing point for all temperatures. From Figure 4g, Figure 4i and Figure 4k, at all ambient temperatures, maximum concentration of σvm values are detected for 800 µm TiB2 coated samples, at the top left corner of the bearing point. The lowest σvm concentration on the coating layer is calculated for AlN with coating thickness of 400 µm (Figure 4h, Figure 4j and Figure 4l). 2018

Figure 4. Von-Mises (σvm) stress distribution on coating models; max. at 373 K for SSCM (TiB2; 400 µm)(a), min. at 373 K for SSCM (AlN; 800 µm)(b), max. at 573 K for SSCM (TiB2; 400 µm)(c), min. at 573 K for SSCM (AlN; 800 µm)(d), max. at 873 K for SSCM (TiB2 ; 400 µm)(e), min. at 873 K for SSCM (AlN; 800 µm)(f), max. at 373 K for FSCM (TiB2; 800 µm)(g), min. at 373 K for FSCM (AlN; 400 µm)(h), max. at 573 K for FSCM (TiB2; 800 µm)(i), min. at 573 K for FSCM (AlN; 400 µm)(j), max. at 873 K for FSCM (TiB2; 800 µm)(k), min. at 873 K for FSCM (AlN; 400 µm)(l) During the FEA study, shear stress (τ xy) values on the coated samples were recorded and analyzed in detail. Figure 5 depicts a comparison for two types of coating model depending on temperature and coating layer thickness. First of all, contrary to the single layer model, higher τxy levels are seen on full-surface coated models due to increasing volume fraction of hard/rigid ceramic layer material in the sample. For both coating models, as the ambient temperature goes up, calculated τxy also rise on account of climbing thermal expansion coefficients of base material and coating ceramics. Similarly, a parallel trend is also recorded for alternating coating layer thicknesses owing to the same reason. When the peak and bottom points are probed, it is correct to express that while the lowest value of 20.3 MPa belongs to single layer 400 µm TiB 2 coated sample in 373 K, the highest value of 865.5 MPa is read for full-surface 800 µm AlN coated sample (873 K). This outcome indicates that mechanical and thermal properties of coating materials at a specific temperature can affect the deformation mechanism sharply. Also, TiB2 coated sample is subjected to more shear deformation than Al2O3 at 573 K and 873 K for both models and at 373 K for FSCM. However, it is observed that TiB2 loses its superiority over Al2O3 at 373 K for SSCM. Regardless of coating model, along with rising ambient temperature and coating thickness, AlN coated samples experience more shear stresses than the others. This 2019

circumstance can be explained with its relatively low thermal coefficient of expansion value in comparison with the Al2O3 and TiB2.

Figure 5. Comparison graphs of ceramic coated models for shear stress (τxy) values In order to determine τxy localizations on the modelled samples, simulation images obtained from FEA are given in Figure 6. Due to high thermal expansion coefficient of Ti-6Al-4V base material, all single-surface coated model samples deform upward direction as estimated before (Figure 6a-f). In terms of shear stress localization, top right sections of the single-surface coated samples are under a noteworthy risk. If the peak τ xy values of 41.1, 109, 212.9 MPa (all measured for AlN coating) are examined, from 373 K (Figure 6a) to 873 K (Figure 6e), an increasing pattern can be seen for both average τxy and magnitude of the affected zone. In a similar vein, if the lowest τxy values (all measured for Al2O3 coating) are taken into consideration, while 1.3 MPa is found at 373 K (Figure 6b), 2.2 and 3.7 MPa are calculated at 573 K (Fig 6d) and 873 K (Figure 6f) respectively. As for full-surface coated model samples shown between Figure 6g and Figure 6l, the bearing point plays an important role for probable increment in the τxy concentration since thermal expansion of Ti based alloy is confined by hard ceramic layers. Due to large difference between thermal expansion coefficients of AlN and Ti-6Al-4V alloy, the highest τxy concentration is detected at 873 K for AlN coated sample with thickness of 800 µm (Figure 6k) although the smallest effect is seen at 373 K for Al2O3 coated model with thickness of 400 µm 2020

(Figure 6h). Apart from the stress levels calculated on the models, resultant displacement values are also significant for certain applications by reason of control necessity of dimensional accuracy of the coated parts under different service conditions. In this context, Figure 7 illustrates resultant displacement values of all samples having dissimilar coating models and deformation temperature to compare coating characteristics.

Figure 6. Shear stress (τxy) distribution on coating models; max. at 373 K for SSCM (AlN; 800 µm)(a), min. at 373 K for SSCM (TiB2; 400 µm)(b), max. at 573 K for SSCM (AlN; 800 µm)(c), min. at 573 K for SSCM (Al2O3; 400 µm)(d), max. at 873 K for SSCM (AlN; 800 µm)(e), min. at 873 K for SSCM (Al2O3; 400 µm)(f), max. at 373 K for FSCM (AlN; 800 µm)(g), min. at 373 K for FSCM (Al2O3; 400 µm)(h), max. at 573 K for FSCM (AlN; 800 µm)(i), min. at 573 K for FSCM (Al2O3; 400 µm)(j), max. at 873 K for FSCM (AlN; 800 µm)(k), min. at 873 K for FSCM (Al2O3; 400 µm)(l) According to FEA results, single-surface coated samples have better displacement ability than their full-surface coated versions at the same temperature values because of the fact that Ti-6Al4V base material can elongate easily without bottom and right-side confinements. From Figure 7, single layer coated models have the largest elongation capacity than the full-surface models and the highest value is read as 2.69 mm at 873 K for sample having single 800 µm AlN layer. In addition, it can be alleged that as the temperature escalates, elongation capacities of the all ceramic coated models improve due to diminishing rigidity (modulus of elasticity) and rising thermal expansion coefficients of both ceramics and the base alloy. When the full-surface models are analyzed in terms of resultant displacement, compared to AlN and TiB 2, thanks to their moderate elastic modulus and high thermal expansion capacity, Al 2O3 coated samples exhibit 2021

superior behavior than the others. What’s more, since the base metal is covered from all directions by ceramic coating materials, there is a negative relation between coating thickness and displacement ability properties in full-surface coating model samples. This case is observed separately at all temperatures. This situation can be explained with displacement values along y axis. In FSCM, Ti-6Al-4V base material elongates through x axis much more than y axis. Therefore, the effect of elasticity modulus and thermal expansion coefficient values are highly important factors. Moreover, as expected, AlN coated sample exhibits lowest resultant displacement because of its low thermal expansion coefficient. On the other hand, the amount of elongation in the y-axis significantly affects the resultant displacement of the AlN coated sample in the SSCM. In this context, with the help of finite element analysis, a prediction can be made about the thermo-mechanical behavior of the coated samples under real operating conditions.

Figure 7. Comparison graphs of ceramic coated models for resultant displacement (U sum) When the simulation details demonstrated in Figure 8 are glanced at carefully, it can be noticed easefully that single-surface coating model samples performs better than the full-surface coating model samples from the point of thermal elongation ability. For single-surface design, at 373 K, maximum and minimum resultant displacement values are 0.27 mm and 0.41 mm for Al2O3 and AlN coatings respectively (Figure 8a and Figure 8b). At 573 K (Figure 8c-d) and 873 2022

K (Figure 8e-f), together with the climbing temperature, deformation occurred on top right side of the samples become more apparent. At 573 K, the highest elongation can be found as 1.22 mm (Figure 8c) while this value is 2.69 mm at 873 K (Figure 8e). Nevertheless, if the full surface coating model samples are evaluated, it can be realized that the situation is quite different (Figure 8g-l). All full-surface design samples substantially elongate in the x axis because of confinements stemming from bearing point and frame ceramics. Furthermore, this one-way elongation style is also caused by symmetric structure of coating materials. As for numbers obtained from the FEA, the highest value is determined as 1.48 mm for 400 µm Al2O3 coated sample at 873 K (Figure 8k). The lowest displacement is observed as 0.211 mm at 373 K for 800 µm AlN coated sample (Figure 8h). At the intermediate temperature of 573 K, maximum calculated displacement is 0.716 mm for 400 µm Al2O3 coating even though minimum value of 0.66 mm belongs to 800 µm AlN coating (Figure 8i-j).

Figure 8. Resultant displacement (Usum) distribution on coating models; max. at 373 K for SSCM (AlN; 800 µm)(a), min. at 373 K for SSCM (Al2O3; 400 µm)(b), max. at 573 K for SSCM (AlN; 800 µm)(c), min at 573 K for SSCM (Al2O3; 400 µm)(d), max. at 873 K for SSCM (AlN; 800 µm)(e), min. at 873 K for SSCM (Al2O3; 400 µm)(f), max. at 373 K for FSCM (Al2O3; 400 µm)(g), min. at 373 K for FSCM (AlN; 800 µm)(h), max. at 573 K for FSCM (Al 2O3; 400 µm)(i), min. at 573 K for FSCM (AlN; 800 µm)(j), max. at 873 K for FSCM (Al 2O3; 400 µm)(k), min. at 873 K for FSCM (AlN; 800 µm)(l) In the light of theoretical simulation efforts, it can be claimed that different coating materials, coating thickness and ambient temperatures can change the thermal stress properties of ceramic coated Ti-6Al-4V base material. In addition, since some physical and mechanical properties of the base alloy like thermal conductivity and young modulus alter depending on ambient temperature as some investigators reported [39], single-layer ceramic coated samples are forced an upward motion due to the structure of simulation models while full-surface model samples expand to right side. In this paper, two FEA models were set for three different ceramic coating materials to analyze stress distribution. The FEA study is considerably significant because, sometimes, 2023

experimental results are not enough to figure out thermal stress examination after coating operations on metallic materials [40]. It can be also asserted that although Al 2O3, AlN and TiB2 coatings can be formed on Ti-based alloys by means of some different methods such as boronizing, HVOF, and micro arc oxidation, investigations on mechanical properties [13,14,41], especially stress distribution analysis, are difficult tasks for researchers without detailed FEA works. In this point, in this study, following to the FEA study, it was observed that thermal stress and elongation distribution of the coated samples could be predicted. For all ceramic types, the highest stress concentration was detected on the point of support for full-surface model samples whereas the highest shear and Von-Mises stress values were read on the top right side sections of the single-surface samples. According to the FEA results, thermal stress distribution and elongation properties can change depending on dissimilar coating ceramics, so it can be also alleged that alternative coating models, optimum coating thickness and proper part geometry with round corners should be taken into consideration in order to keep corrosion/oxidation resistance and improve the design conditions against the thermal loadings.

4. Conclusions

As a result of this simulation study focusing on thermal stress and elongation features of hard engineering ceramic coated widely used Ti-6Al-4V alloy, the followings can be listed;  According to FEM analysis, TiB2 and AlN coated samples show the highest σvm and τxy values for all temperatures respectively.  AlN coated SSCM samples have superior displacement ability than the others, however, this situation is in favour of Al2O3 for FSCM model design.  Compared to σvm values, the resultant displacement values and τxy values calculated for FSCM are obtained considerably different than SSCM due to increasing volume fraction of hard ceramic coating materials.  σvm and τxy localizations largely depend on type of the coating design since both of them localize in top right sides of the parts for SSCM models while the bearing points on the left side are subjected to higher stresses for FSCM models.  From 373 K to 873 K, FSCM samples possess higher thermal σ vm and τxy stress levels in comparison with SSCM samples on account of limited displacement of base metal leading to internal stresses.  For all types of coating materials, SSCM samples display a downward tendency in terms of the thermal stress values owing to decreasing rigidity and escalating thermal expansion coefficients of ceramics depending upon increasing coating thickness. However, an upward trend is observed for samples having FSCM because of rising volume fraction of the hard-ceramic coated zones in the sample.

Share and Cite