A survival analysis on endurance tests of chainsaws and power cutter components at Husqvarna Group AB

Sammanfattning: Due to the steadily growing world chainsaw market and the market’s relative competitiveness, high performing chainsaws are important for Husqvarna Group AB to remain competitive. In the development phase a large number of endurance tests are run to evaluate the product’s performance. Given the data gathered from these tests, the main goal of this thesis was to model the time until failure and evaluate the relationship between the measured predictors and the failure times. To model the data, traditional survival theory was used, which included fitting a Cox proportional hazard model and an extended Cox model for time-varying variables. The results of the analysis showed that the temperature of the surrounding equipment has a significant effect on the failure times of chainsaws for both test methods. For the first test method the volatility of the revolutions per second(RPS) in the engine also proved to have a significant effect. Component cracks in power cutters seemed to increase with dry weather and low temperatures, confirming what the engineers at Husqvarna Group AB believed to be the case before this thesis was carried out. Other future analysis methods were discussed together with some identified data collection challenges. When the technical difficulties for collecting the proposed new data sets have been resolved, the models suggested in this thesis may be improved by linking a number of new predictors to the failure times.