Position within the page tree

Projects of Reliability Engineering

Institute of Machine Components (IMA)

Overview of current projects, recently completed projects and the project archive

Current projects

Reliability assurance of safety-critical systems at CERN
Development of a guideline for the reliability assurance of safety-critical small series systems

Cost-efficient reliability of PV power plants with battery systems
Optimizing the availability of PV power plants with battery storage by forecasting the remaining useful life and condition monitoring using integrated PHM solutions and predictive maintenance.
Safety of the intended functionality
Functional safety and safety of the intended functionality for powernets
Reliability of passive components
Reliability and lifetime model of passive components
Reliability Assurance of Variants
Effort-reduced reliability assurance in product variant development considering prior knowledge
Data-driven condition monitoring of electronic systems at CERN
Condition monitoring of electronic systems by integrating expert knowledge and hardware tests into machine learning algorithms
Reliability of Adaptive Load-Bearing Structures
Collaborative Research Centre 1244: adaptive skins and structures for the built environment of tomorrow
Lifetime Tests of HV Batteries with Prior Knowledge
Efficient planning of lifetime tests for reliability assurance of hv batteries considering prior knowledge
Lifetime Model for a Timing Belt Drive
Lifetime model for a timing belt drive in steering gears based on design of experiments (L-DoE)
Reliability Process for Photovoltaic & Battery Inverters
Development and evaluation of an efficient and reliable methodology for reliability prediction and lifetime prediction of photovoltaic & battery inverters and their critical components
Lifetime Model Winding Insulation
Development of a holistic lifetime model for the winding insulation of electrical machines on the basis of the acting damage mechanisms and their different load levels

Recently completed projects

Zuverlässigkeitsbetrachtung von Wechselrichtern
Erarbeitung und Evaluation einer effizienten und verlässlichen Methodik für die Zuverlässigkeits- und Lebensdauer-Vorhersage von Photovoltaik & Batteriewechselrichtern und ihren kritischen Bauteilen
Entwicklung modulintegrierter Elektronik für Photovoltaik-Anlagen
Zuverlässigkeitsbewertung der modulintegrierten und systemoptimierten Elektroniklösungen von Photovoltaik-Anlagen
Feuchtetransfer
Virtuelle Sensoren zur Bestimmung der Feuchteverhältnisse in Solarwechselrichtern
Zuverlässigkeitsnachweis von Systemen mit Vorkenntnis
Zuverlässigkeitsnachweis von Systemen unter Berücksichtigung von unscharfer Vorkenntnis

Project archive

Modular and Scalable Safety Concept

Philipp Kilian, in Cooperation with Robert Bosch GmbH

Innovative and holistic safety concepts for the power supply system are required due to the megatrends electrification, automation and connectivity. To cope with these trends and to be able to react dynamic on changing requirements, the corresponding safety concepts shall be structured in a modular and scalable manner. Therefore, we do research on innovative approaches to automatically derive efficient safety concepts including their validation and verification based on artificial intelligence and/or optimization algorithms. Because the evaluation of different hardware-architecture is very time consuming and error prone, we want to investigate an automated approach based on fault injection to evaluate hardware building blocks in a modular way. The fault injection shall be based on generic fault models to enable automation and increase objectivity.

Future-Load-Model for Predicting Representative Load Collectives

Tamer Tevetoglu, IMA

As part of a predictive diagnostic model, the Future Load Model is being used to predict future loads (time series data) that significantly influence the remaining lifetime of the lead-acid battery. For this purpose, past sensor data of the battery are first stored and grouped into representative load cases using unsupervised machine learning methods. Then, a feedback neural network (RNN) based on a so-called long short-term memory (LSTM) is trained and being used to predict future signal data. The now predicted data is assigned to the representative load cases by a classifier, e.g. Random Forest. By using the scattering of the data points within a load case, the confidence intervals of the predicted signal data are derived. Now the remaining lifetime can be calculated based on the future loads.

Close-to-reality system analysis with a confidence interval

Frank Müller, DFG

With stochastic modelling techniques, it is possible to model complex, technical systems and to simulate their reliability and availability. At the moment, the statistical quality of the initial data cannot be considered within the analysis. Objective of the research project is the continuously consideration of the statistical quality of the input data within the analysis method, in order to combine the expressiveness of confidence intervals with the performance of the modelling techniques.

Development of algorithms for soft sensors to determine damaging load spectra for components of large diesel engines

Martin Diesch, MTU Friedrichshafen GmbH

Knowing the actual operational load is of utmost importance for the development of a reliable product. The goal of this project is to determine load spectra, which describe the load on components of large diesel engines. Soft sensors will be implemented to determine the correlation between physical measured variables and damaging parameters.

Methods and technologies for fail-safe adaptive bearing structures and their elements and structures

Andreas Ostertag, DFG Sonderforschungsbereich (SFB 1244)

In the SFB 1244 the aspired adaptive structures need to adapt to the current load situation like wind and weather and the user. Beside of guaranteeing a high reliability the behavior of the system when it comes to a failure is very important. The objective is to have a Fail-Safe-concept managing an adaptive fault response which keeps the operating condition sustainable as far and as long as possible. Furthermore supporting Fail-Safe-elements will be developed and applied.

Reliability prediction based on findings of remanufactured car transmissions

Andreas Kroner, Daimler AG

One goal of remanufacturing is to maximize the re-use rate of parts of the product to save resources and capital, but without reducing the quality and reliability of the remanufactured products. To achieve this, it is important, that the selected parts for remanufacturing ensure a second lifetime of the whole product.

Test and reliability-oriented product design

Thomas Herzig, IMA

The testing effort required to prove the reliability target can already be taken into account in the early development process. The aim is to achieve an optimum between product development costs (material, production, ...) and testing costs (test type, number of test items, ...) by adjusting the overdimensioning. (material, production, ...) and testing costs (type of test, number of test items, number of test rigs required, choice of load levels in accelerated testing, ...). The choice of the appropriate testing strategy is thereby objectively evaluated with the help of test costs, total time and probability of success and is influenced by the oversizing of individual failure mechanisms considered in the development process.

Prognostics and Health Management (PHM): Theoretical approaches and implementation of PHM relevant prior knowledge in machine learning algorithms

Mark Henß, IMA

Prognostics and Health Management is a technical discipline dealing with the evaluation and control of the remaining useful life of systems. One goal of this work is to bundle the spectrum of methods and models from reliability engineering, statistics, machine learning, measuring technology, etc. in a common theoretical approach.
The use of machine learning enables automatic recognition of relevant patterns and trends in data that are difficult to access for humans. Missing training data, labels and a low data quality are obstacles for implementation. This project investigates how prior knowledge can be implemented in the algorithms to eliminate these deficits.

Expert system for application-based lifetime prognosis of timing-belt drives

Alexander Kremer, Walther Flender GmbH

Despite a long history of available belt drives there is whether standardized designs or engineering method. In addition, there is a lack of quantitative information on the reliability and wear of belt drives. Therefore, a methodology for application-based lifetime modelling and lifetime prediction of belt drives should be developed. This lifetime model has to consider relevant design and use parameters.

Working Group Standard Load Spectra for Commercial Vehicle Brakes

Kevin Lucan und Mark Henß, Bergische Achsen, DAF, Daimler, Haldex, IVECO, Knorr-Bremse, MAN, MERITOR, SAF Holland, WABCO

Knowledge of the strength behavior and operational loads is necessary for the development of durable and reliable products. For this reason a working group in combination with ten participants of the commercial vehicle industry was established to derivate standard load spectra for commercial vehicle brake systems. Goal of the working group was the development of an industry-wide standard for designing and dimensioning of commercial vehicle brakes.

Reliability analysis of mechatronic systems

Frank Müller und Jan Gröber, Festo AG & Co. KG

Reliability analysis of mechatronic systems considering mechanic and electronic components as well as aspects of software. The reliability analysis is shown on the example system of a novel valve terminal. Design and implementation of reliability tests. Systematic work out of the basics of reliability of mechatronic systems as well as working out of the possibilities to determine the systems’ reliability in early development phases.

Development of a fail-safe mechatronic drive system for deep – sea application

Julian Popp, HSA Aalen, ISSA Forschungsprojekt

The demand for raw materials from the deep sea is constantly increasing. In order to extract these raw materials, systems are needed that are environmentally friendly, safe, reliable and also available. The project will develop a system architecture with a focus on mechatronic drive technology that meets the enormous requirements that prevail in the deep sea. Verification and validation on a self-developed test bed environment is also part of the project.

Dependability of electronic particle accelerator systems at CERN

Volker Schramm, CERN

Analysis of surveillance systems for particle accelerators to improve the reliability, availability and the protective function. Development of a methodology on how to design, produce, test and operate dependable electronic systems. In addition, application of the methodology for the system reliability analysis, definition of the operational strategy and execution of functional and reliability tests.

Design of a condition monitoring system for a safe mechatronic drive system

Sebastian Imle, WITTENSTEIN motion control GmbH

The extraction of raw materials in the deep sea requires an environmentally friendly and safe propulsion solution. The system architecture defines the components and interfaces of the system in terms of type and quality. The processing quality of this information via a secure communication architecture determines the diagnostic coverage and functional safety of the system. The design of a condition monitoring system for use in the deep sea is the goal of this project.

Consideration of simulation results in reliability assurance

Alexander Grundler, Martin Dazer; Daimler AG

In addition to physical tests of products, simulation results can also be taken into account to ensure reliability. These have to be understood from a statistical and reliability point of view, so that they can be converted into the necessary reliability statement with a statement probability. Depending on the simulation and the failure mechanism, special hurdles have to be overcome. The simulation results can then be combined with results from physical tests to make the reliability statement of the product more precise.

Methods for consideration of prior knowledge

Alexander Grundler, ZF Friedrichshafen AG

The consideration of prior knowledge does not only allow for a more precise reliability statement, but also a reduction of costs and time in reliability test planning. Existing methods, however, show very conservative behavior and in particular the quantification of prior knowledge stemming from simulation results, as well as a transformation of reliability information between different system levels do require practicable methods.

Analyzing reliability and functional safety of electrical onboard power supply systems and their components

Frederic Heidinger, Patrick Münzing and Andreas Ostertag, Robert Bosch GmbH

Basics for automated driving are recognizing the surroundings, planning and realizing the trajectories for steering the vehicle. The energy supply is required for fulfilling these functions. Because of a high criticality of these functions the onboard power supply system gets unprecedented huge safety requirements. To ensure a flawless function of the automated vehicle it is indispensable to regard reliability and functional safety.

Realistic modelling and analysis of the availability of production systems in Industry 4.0

Fei Long, GSaME, DFG

Modelling and analyzing the availability of production systems in Industry 4.0 based on Petri nets, taking into account all aspects such as flexibility, adaptability, learning ability, numerous dependencies and interactions. The aim is to develop a powerful method for modelling and analyzing the availability of production systems in Industry 4.0.

Investigation of the damage mechanism pitting on gears of a vehicle transmission

Zeljana Beslic, IMA

The lifetime of gears of serial transmissions are investigated on the testing bench in consideration of the failure mechanism pitting. A statistical liftetime proof of the gear with respect to the fatigue damage and the influence on the scatter of the pitting degradation model, under the influence of the load, is presented. A test methodology with regard to resource planning and a simulation approach are verified with the measurement data.

Development of Prognostics and Health Management (PHM) concepts for commercial vehicle components

Sebastian Held, Knorr-Bremse Systeme für Nutzfahrzeuge GmbH

PHM focusses on the assessment of the remaining useful lifetime of a component, as well as on deriving corresponding actions. The usage of PHM in commercial vehicles is becoming more and more promising in the context of digitalization, electrification and automated driving. Within this project PHM applications for commercial vehicle components are being developed in regards to reduced life cycle costs and increased vehicle safety and availability.

Optimal reliability test planning using prior knowledge from stochastic fatigue simulations

Martin Dazer, Knorr-Bremse Systeme für Nutzfahrzeuge GmbH

Optimal reliability test planning can only take place when there is already credible prior knowledge of the product-specific failure behavior. In this project it is obtained from stochastic fatigue simulations. A virtual Weibull distribution is developed via systematic modeling of uncertainties using probability functions. With this virtually determined prior knowledge it is possible to give recommendations for determining an optimal test strategy.

Cost-related reliability development

Target values such as availability (A) and reliability (R) are defined in a cost-optimal way (€) based on function-related customer requirements, the impact of suitable maintenance strategies on the business model is investigated and the business model selection is supported. The symbiosis of stakeholders, values, functions and budget in the reliability context is the focus of the study.

Simulation methodology for reliability analysis of cylinder crankcases

A method is being developed for the design of cylinder blocks which makes it possible to provide information on the reliability of cylinder blocks at an early stage of development. The calculation is based on load collectives from driving operation and experimentally determined material data. By linking and comparing calculation and field analysis, a consistent strategy for reliability analysis is to be developed.

Robust design in commercial vehicle braking systems

Based on elements of the Robust Design methodology (Axiomatic Design and Taguchi), a holistic method for designing robust and reliable components of commercial vehicle braking systems is developed.

Download Project Description in German

Planning of extended service life tests of commercial vehicle brakes

By developing a methodology for accelerated testing of commercial vehicle braking systems, their reliability has to be investigated. In addition to end-of-life tests, reliability demonstration tests and preliminary investigations are also planned. At the same time, a procedure will be shown for demonstrating the reliability of the systems during development for future projects.