Acceleration of concrete tests with very high numbers of load cycles

Table of contents

1. 1. 1. Project data
      2. Report in the year book 2016

Project data

Titel | Title Möglichkeiten zur Beschleunigung von Untersuchungen zur Betonermüdung unter sehr hohen Lastwechselzahlen | Possibilities for accelerating investigations of concrete fatigue under very high numbers of load cycles Förderer | Funding Deutscher Ausschuss für Stahlbeton (DAfStb) Zeitraum | Period 01/2012 – 12/2016 Leiter | Project manager Prof. Dr.-Ing. Dr.-Ing. E.h. Manfred Curbach Bearbeiter | Contributors Dr.-Ing. Torsten Hampel, Dipl.-Ing. (FH) Sabine Wellner, Dipl.-Ing. Robert Schneider

The behaviour of concrete at over 2 million load cycles, as arise, e.g., in wind power plants is not yet sufficiently explored. One of the reasons for this is the long duration of a single experiment. Within this research project possibilities for reduction of testing are examined.

Report in the year book 2016

Report in the year book 2016Report in the year book 2016

Concrete under high load cycles

For the design of concrete structures under cyclic loading, it is common to use characteristic material properties determined after 2 · 10⁶ load cycles. In the case of wind energy plants, significantly higher load exchange numbers N occur in a relatively short time. For this reason, we must ask ourselves whether known design values can be used, as no permanent strength limit could be detected for concrete until today. According to current knowledge, specific tests for very high load cycles of N ≥ 10⁷ are necessary. In a feasibility study, concrete cubes under cyclic loading were tested in modified experimental set-ups in order to examine experimentally whether the extremely long duration for such experiments can be reduced.

For the tests, normal strength concrete was used. In the beginning, cubic specimens with 10 cm edge lengths were fabricated. In the first work package, force controlled compression tests with varying frequencies of 5, 20 and 30 Hz and approximately one million load cycles were carried out to check if the frequency influences the test results. Such an influence was observed. This result is in line with recent findings from other researchers. An arbitrary increase of the test frequency in order to reduce the duration of an experiment must therefore be excluded.

Furthermore it was investigated, if a reduction in the total test time is possible by a simultaneous testing of several samples. For this reason the test specimens were loaded in parallel and in series with 5 Hz. Load cycle numbers up to 3 × 10⁷ were reached in tests of three parallel arranged cubes without a premature failure of one or more test specimens. In contrast, the experimental set-up with several test cubes in series was proven to be unsuitable. When one cube failed, stress peaks were introduced into the remaining samples due to an insufficiently sensitive machine control.

Accompanying, it was checked at different frequencies up to 30 Hz, whether the samples heat up as a result of the cyclic loading. This was not the case for the tested frequency range for the normal strength concrete samples.