Twin-axis tests

The twin-axis strength is very difficult to measure in a twin-axis testing unit which is familiar, it is the strength of the sample and not that of the material that is determined, as the failure always starts from the edge. 

Here it is only necessary to carry out parallel-thread single-axis tests. All other things being equal, the proposals outlined above can be used. Twin-axis tests are not required here. 

Twin-axis tests are not intended for measuring strengths, but to establish the compensation data and to determine the moduli of elasticity. For this reason it is only necessary to carry out the tests in the applicable load range, which thus has to be defined. Here it is only possible to provide general data and each individual case is the responsibility of the engineering consultancy and the building inspectorate. In general a safety factor of five is used, but this should always be discussed and may deviate in individual... 

In addition it is necessary during twin-axis tests to simulate the behaviour in the building structure, in other words under real conditions. Real loads are snow and wind. If the main anisotropic directions are approximately parallel to the warp and weft, then in this load case one of the directions is loaded and the one which is perpendicular to it is relieved of any load. This procedure is then changed round and the previously loaded direction is unloaded and the previously unloaded direction is then loaded. The directions of the warp and weft depend thereby on the design and the web positions. 

The bursting test is capable of answering the question of the twin-axis strength. Therefore the bursting test is required here. The diameter of the base circle should be at least five times the seam width. The results of the bursting tests wiU establish the reduction factor of the parallel single-axis seam test as a result of twin-axis loading. 

Relaxation denotes the decrease in the initially induced stress with constant elongation. Twin-axis relaxation tests are, however, very difficult to carry out. Here it therefore makes sense to carry out single-axis tests and then to attempt to calculate the fabric behaviour. As a first approximation it can be assumed that the stress changes linearly with the relaxation behaviour of the thread. The proportionality constants can be calculated from the E-modulus and the relaxation behaviour of the thread. 

---