1912 2022 eng 30 preprint 0 -- -- -- Varying thermal stresses influence significantly the time to failure of electric components as used, for instance, in automotive devices. For applications as autonomous driving a high reliability has to be guaranteed. In this article we discuss how to combine probability distributions for failure. Discrete and continuous changes of the probability distribution in time are both considered. It turns out that the temporal order of the distributions, corresponding to the succession of stresses in applications, is essential. The latter observation restricts the general applicability of the widely used temperature collectives where only the total time of a temperature stress is considered neglecting the order of the stresses. An application of our results are thermal overstress tests on electric components. We may explain yet not well understood measurements for automotive electric cables. urn:nbn:de:bvb:861-opus4-19126 Ja publish Creative Commons - CC BY-NC-ND - Namensnennung - Nicht kommerziell - Keine Bearbeitungen 4.0 International Sven-Joachim Kimmerle Karl Dvorsky Hans-Dieter Liess Rudolf Avenhaus eng uncontrolled Reliability eng uncontrolled Applied Probability eng uncontrolled Data Science eng uncontrolled Physics of Failure eng uncontrolled Arrhenius Model Mathematik Fakultät für Angewandte Natur- und Geisteswissenschaften Technische Hochschule Rosenheim https://opus4.kobv.de/opus4-rosenheim/files/1912/KimmerleDvorskyLiessAvenhaus_ExtendedVersion.pdf