Premium
An error analysis for visioplasticity data processing
Author(s) -
Pehle Hans Joachim,
Kopp Reiner
Publication year - 1986
Publication title -
steel research
Language(s) - English
Resource type - Journals
eISSN - 1869-344X
pISSN - 0177-4832
DOI - 10.1002/srin.198600772
Subject(s) - smoothing , grid , observational error , algorithm , data processing , function (biology) , process (computing) , error function , experimental data , computer science , mathematics , statistics , geometry , evolutionary biology , biology , operating system
The conducted investigations are concerned with the effect of data errors on the calculations of stress and strain components according to the visioplasticity method. The error analysis was based on an evaluating process suitable both experimentally and theoretically for exact calculations. Longitudinal sectional areas of rod drawing stock were marked photochemically with measuring grids with a line spacing of 0.1–1 mm. Specimens thus prepared were partially drawn and the deformation patterns were registered by means of a measuring microscope. The measured data were evaluated with the aid of a smoothing and calculating procedure based on cubic spline approximations. This procedure was checked with a flow model and proved to be suitable for exact visioplasticity calculations. Error investigations were carried out in order to examine the influence of a measured data scatter on the visioplasticity calculation results and an error size was defined and used to establish an error curve according to three different methods. By means of this function it is possible to quantify – as has been indicated by means of an example – the calculation error for a given quantity of data or, depending on the accuracy and aim of the calculations, the demands on the measuring grid and the data quality can be derived. By the example of the calculation of frictional stresses it has been demonstrated that the spacing of the measuring grid lines must be adjusted to suit the inhomogeneity of the plastic flow. If need be, a special more sophisticated method of evaluation is necessary – as described in 5 ) – to determine accurately such inhomogenous strain patterns.