Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence
Recent research has shown that fatigue life in non strain-crystallising rubber increases if the material is subjected to pre-loading. The central research question posed in this project is ‘whether preloading and surface treatments improve the surface finish of rubber components, reduce stress concentration and can this reduction partially account for greater fatigue resistance’. The specific objectives of the research are:- i) To assess surface finish measurements and to investigate the changes in stress concentration in preloaded rubber samples using White Light Interferometry. ii) To consider other non-contact methods of observing changes in stress concentration in preloaded samples through use of: i) Electronic Speckly Pattern Interferometry (ESPI) methods, ii) Ultrasonic methods iii) To examine whether coatings or other surface treatments can play a part in diminishing the surface flaws produced in the various manufacturing processes. The coatings/treatments investigated were: i) Diamond Like Carbon Coating (DLC), ii) Sol-gel Coating Technology, iii) Surface Peening treatments. iv) To use MSC/MARC Non-linear Finite Element software to model: i) fatigue behaviour of DLC coated EPDM rubber, ii)EPDM rubber elastomeric crack propagation, iii) flaw behaviour in EPDM rubber. v) To advise on design standards for pre loaded rubber components and to establish criteria for selecting elastomers that will minimise the likelihood of fatigue failures in the light of the aforementioned research and of this study. The Deutsches Institute fur Kautschuktechnologie (DIK) supplied the test material for this research and provided facilities for the rubber vulcanisation process described in this work. DIK also allowed use of their test equipment. The Research involved the design of a clamp for tensile testing of elastomers, a compression mould and an ESPI in-plane test rig. An initial investigation determined that stress concentration diminishes with prestressing rubber components. White Light Interformetry was found to be the most suitable method of non-contact analysis. The only disadvantage proved to be the small field of view. The ESPI Method did not show results as positive as those achieved using WLI, however further investigation is required before this method is discounted. The Ultrasonic non-contact method was not evaluated but a schematic system is recommended for further research. Analyses for surface roughness and stress concentration at the surface flaws of EPDM rubber using the WLI confirmed that prestressing decreases surface roughness values. The measurement of stress concentration at surface flaws using the WLI method was not successful, because the test rig used was inappropriate to perform this test. However, it was shown that WLI could detect surface defects. Another method was evaluated which uses the Olympus BX60M System Microscope in combination with the Omnimet Archive Digital Imaging System. This method was used to measure stress concentration at the tip of an initiated edge crack. The Griffith theory for measuring stress concentration at the crack tip was employed. The results showed that pre-stressing of rubber samples results in reduction in stress concentrations at the crack tip. The first surface treatment applied was Diamond Like Carbon (DLC) coating. This coating was used because carbon particles exhibit an appropriate size to fill the surface defects present in rubber. The coating process was performed in the laboratories of Enterprise Ireland (EI). The coating procedure involved heating the samples to a temperature of 120°C and this caused a change in the physical properties of the rubber samples. The fatigue test results show that DLC treatment does not improve the fatigue life of elastomers. However, data from the same tests also show that the complex modulus E* fell to approximately 76% for the first cycle. This supports the hypothesis in the earlier research that the complex modulus can be used for safely predicting the service life of rubber components. Following these tests, further coating was attempted, this time employing a rubber like material using Sol-gel technology. Tetraoxidsilicate (TEOS) liquid solution was chosen as the coating material because it has a low cyrstallisation temperature. The results were disappointing because the coating id not react with the rubber material and so did not successfully crystallise on the surface of the rubber specimen. Shot Peening was also used to treat the EPDM rubber surface of same samples pior to tensile testing. The tests showed that no change in tensile properties of EPDM rubber occurred in comparison with untreated EPDM rubber. However, microscopic analysis showed improvement in the surface roughness (Ra). For carrying out a Finite element Analysis (FEA), material data was obtained from uniaxial tests. The first test was an axisymmetric analysis of fatigue behaviour of the DLC coated EPDM rubber. For these analyses data was used from the first recorded cycle and the last recorded cycle of the fatigue tests. Material models were determined for implementation in the analysis using MSC/MARC software. The axisymmetric analysis of the fatigue specimen simulated the model being subjected to the same displacements that produced values of maximum stress in the tests. FEA reasonably accurately predicted the physical test results. The axisymmetric FEA of the crack models showed a large concentration of stress at the tip of the crack. The FE plain strain analysis of the EPDM rubber sample, with a flaw of 60um in diameter modelled, gave the best correlation with physical tests. The Ogden material model represents the most plausible model and is the most popular function using stretch ratios. For the Ogden model, a two-term function was used and constants were calculated from uniaxial tests. In future an emphasis must be placed on creating appropriate material models. Conclusions can be summarised as follows. While Light Interferometer is the best metrology method considered for measuring surface roughness and profiling the geometric of surface defects in rubber materials Tensile loading reduces stress concentration Surface finish improvements may contribute to improve fatigue life. An exponential decay formula is proposed for determining improvement in surface finish with tensile load. The coatings or surface treatments did not increase fatigue life. The dynamic stored energy theory of fatigue in non-strain crystallising rubbers is supported. That ‘loss in complex modulus’ can be used as a predictor of fatigue life for non-strain crystallising rubbers are confirmed.
Tabakovic, A. (2002). The Influence of Surface Treatments and Surface Finish on the Fatigue Properties of Elastomers. Masters dissertation. Dublin Institute of Technology. doi:10.21427/D7R62V