In-process measurement of Barkhausen noise and resulting productivity increase potential in grinding of case hardened steel

The micromagnetic Barkhausen Noise (BN) measurements bear the capability of providing non-destructively information about the condition of the surface integrity of ground workpieces. No unambiguous statements about the surface and subsurface area state can be made by evaluating single measured micromagnetic values if a superposition of different effects in the microstructure as a consequence of high thermomechanical loads occur. Especially for highly stressed components in automotive industry or in wind energy systems, case hardenend steels are deployed. These steels are often finished at the end of their process chain by a grinding process. At this point, the workpieces to be ground have already achieved enormous created added value. For productivity reasons, grinding should be carried out close to the process limit. To avoid undesirable phase transformations in the microstructure, process models such as the process model from Malkin can be used. If a grinding process consists of several process steps it must be ensured that the final process step leads to the intended surface and subsurface properties. In order to make the process design as productive as possible, an in-process measurement signal which correlates to the current state of the surface and subsurface area can help to achieve this result. Ideally this signal is applicable to generate an adaptive grinding process.
Such a signal is investigated in this paper by application of BN as an in-process measurement technology. The paper shows the reliability of BN measurements as well as that there are different maximum BN amplitudes for different material removal rates. It is expected that these BN values correlate particularly to the residual stresses in the surface and subsurface area since Malkin's grinding burning limit was used to exclude (strong) phase transformations. Considering a micromagnetic parameter measured in-process a reduction of process time of 37% and thus a significant process improvement was achieved.