Search Results

The experimental methods focused on utilizing the newly developed NOVA induction heating and hardening manufacturing process as an adapted method to produce high performance engine valve springs. A detailed testing plan was used to evaluate the expected and theorized possibility for fatigue life enhancement. An industry standard statistical analysis method and tools were employed to objectively substantiate the findings. Fatigue cycle testing using NOVA induction-hardened racing valve springs made of ultra-high tensile material were compared to data for springs with traditional heat treatment and those with standard processing. The results were displayed using Wöhler and modified Haigh fatigue life diagrams. The final analysis suggests that NOVA processed springs have a seemingly slight, yet significant benefit in fatigue life of 5 - 7% over springs processed through a competing method.

The engine valve spring is a very important component in automotive engine systems. The non-metallic inclusions in an engine valve spring will significantly reduce its reliability. In this study, an attempt was made to establish a correlation between fatigue failures and non-metallic inclusions by applying statistical methods. Fatigue tests with BZ and OTEVA-90 materials are performed with two different types of experiments, which are rotating bending fatigue test (Nakamura test) and spring fatigue test. By using RELIASOFT, the data of these two tests are analyzed with the Weibull distribution in order to statistically estimate BZ and OTEVA-90's fatigue lives at 90% low confidence under different stresses. On the other hand, fatigue strength of these materials can be estimated by Murakami and Endo's model with maximum inclusion size predicted from the Gumbel distribution.

The reliability of engine valve springs is a very important issue from the point of view of warranty. This paper presents a combined experimental and statistical analysis for predicting the fatigue limit of high tensile engine valve spring material in the presence of non-metallic inclusions. Experimentally, Fatigue tests will be performed on valve springs of high strength material at different stress amplitudes. A model developed by Murakami and Endo, which is based on the fracture mechanics approach, Extreme value statistics (GUMBEL Distribution) and Weibull Distribution will be utilized for predicting the fatigue limit and the maximum inclusion size from field failures. The two approaches, experimental and theoretical, will assist in developing the S-N curve for high tensile valve spring material in the presence of non-metallic inclusions.