Detection of Structural Damage of Technical Objects with the Use of Multidimensional Analysis

The paper presents a damage identification method of technical objects with the use of multidimensional analyses. The studied objects were rolling bearings tested on a laboratory test stand as well as vehicle wheel rims. The rims were examined by an Automated Tester. A group of variables describing the condition of the bearings included basic parameters of acceleration signal such as: RMS, peak value, functions describing the envelope of vibration spectrum and additionally the experimental coefficient of friction for the oscillatory motion. Diagnostic measurements data can be input into the proposed mathematical models and compared with a database of cases of the object technical states. Results of the multidimensional analysis of diagnostic parameters are presented in the form of diagrams (cluster trees). On this basis, the technical condition of the tested object can be easily interpreted. The proposed models supported the classification of wheel rims into groups characterized by similar frequencies and amplitude based on shock tests.