PSI - Issue 43

26 Stanislav Žák et al. / Procedia Structural Integrity 43 (2023) 23 – 28 Stanislav Žák and Alice Lassnig / Structural Integrity Pr ocedia 00 (2022) 000 – 000 The Mo film on PI was synthesized by direct current magnetron sputtering using an industrial scale magnetron sputter system (FHR.Line.600-V) equipped with a rotary Mo target ( ∅ 125 × 600 mm, 99.97% purity, Plansee). Prior to deposition, the 50 µm thick PI foils (Upilex-S from UBE) was ultrasonically cleaned in an ethanol bath for 10 min. Next, the substrate was plasma etched in the deposition chamber for 48 s at a direct current power of 0.8 kW and 180 sccm Ar flow rate (0.34 Pa). The base pressure in the deposition chamber was 2×10 -6 Pa. The film was deposited with an Ar flow rate of 300 sccm, corresponding to a pressure of 0.52 Pa. A direct current power of 2 kW for 154 s yielded a 300 nm thick Mo film. The elastic material properties of E 1, Mo = 320 GPa, ν 1, Mo = 0.29 and E 2, PI = 8.5 GPa, ν 2, PI = 0.4 for the film and substrate, respectively, yielded the Dundurs parameters values of α = 0.94 and β = 0.15, resulting in ω = 69.77°. The Mo-Cu-glass material system data were used directly from the raw data from previously published research by (Lassnig et al., 2021). For the detailed information on the films manufacturing process the reader is referred to the original research paper. The equivalent Young’s modulus and Poisson’s ratio of the Mo -Cu bilayer with a 100 nm thin Cu film and a 500 nm Mo overlayer on top lead to thickness- weighed Young’s modulus E 1, Mo/Cu = 285.3 GPa and Poisson’s ratio ν 1, Mo/Cu = 0.31. In combination with the glass substrate ( E 2, glass = 64 GPa and ν 2, glass = 0.2), the resulting elastic mismatch was characterized by the parameters α = 0.65 and β = 0.26, yielding the phase factor value of ω = 49.95° . 3. Results and discussion To quantify the differences in fracture criteria caused by different assumptions of mode-mixity angle Ψ , the experimental buckle measurements for two sets of Cu-Mo-glass material systems and one set of Mo-PI material system were evaluated two times (once for the ω = 52.1° and second tim e with the modified ω according to eq. (4)). All of the data produced different Ψ - Γ plots, where the changes in fitting the mixed-mode fracture criteria (according to Table 1) for different ω -values could be quantified. 4

Fig. 2. Example of processed data – blue markers and lines represent the original approach with ω = 52.1°, the red ones are for the respective ω -values according to eq. (4): (a) Mo-Cu-glass (0.5 Pa pressure during the deposition) material system with ω = 49.95°; (b) Mo -PI material system with ω = 69.77° Note that a large shift in mode-mixity angles in Fig. 2 is caused even by a slight change in ω -value. In case of Mo thin film on PI substrate a correct value of ω even changes the initially pure mode II loading into significant I+II mixed-mode loading. Moreover, the mode-mixity angle for each data point visibly influences the shape of each of the fitted semi-empirical functions. By closer look on the resulting G I, c , λ and Ψ 0 parameters in Table 2, large differences between the different fracture criteria are evidenced. This may be caused by allowing the parameter λ be evaluated during the fit, however, as it represents the actual involvement of shearing during the fracture process, it should be independent of the method. Moreover, the λ parameter is sensitive to the data close to Ψ = 0°, which are not available for presented material systems. This may cause also extreme differences when applying the proper ω -value. Therefore,