Producing residual stress reference pieces for Barkhausen noise inspection
Lehtinen, Olli (2023)
Lehtinen, Olli
2023
Master's Programme in Materials Science and Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. Only for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2023-10-30
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202310259048
https://urn.fi/URN:NBN:fi:tuni-202310259048
Tiivistelmä
This master’s thesis examines the effect of residual stress on Barkhausen noise (BN) signal and how a set of reference pieces can be produced. It is desirable to be able to inspect components in a defined interval and discover if there has been changes in the residual stress state. This can help in preventing detrimental effects on critical components. Since BN measurement RMS signal itself does not provide absolute residual stress values it is desirable to compare the signal from an inspected component with a reference piece that has been made of same material and gone through the same processing. Therefore, this reference piece has the same material properties e.g., hardness, yield strength and magnetic characteristics.
Totally 24 samples were made of 300M steel and processed by heat treating, grinding, and shot peening. With the last, a compressive residual stress was induced to the surface of the samples. This was done in order to make the samples same as components in aviation industry. The compression is induced to hinder crack initiation and growth and prolong the service life. The samples were bent using four-point-bending in order to induce residual stresses in a controlled way and decrease the shot peened compressive residual stresses from the surface in a way to simulate relaxation of these stresses during component use. This way the reference pieces can be compared to a component in use, and it can be determined if it is no longer suitable for its purpose. Since BN inspection using Rollscan 350 does not provide absolute values on residual stress, X-ray diffraction (XRD) is used to correlate BN signals to real stress values.
In total 24 reference pieces were produced with a varying range of surface residual stresses. The residual stress results correlated well with literature. RMS signal was sensitive to the change in residual stress when compression decreased and especially when it changed into tensile. In high compression the RMS signal was very stable. This means that using BN inspection is possible to find areas in material surface where the residual stresses have significantly changed but further examination using XRD is needed for absolute values of residual stresses. It was concluded that by four-point-bending, it is possible to induce desirable residual stress state into the sample surface. Two types of BN sensors were used, and a clear difference was found in the signal level. It would suggest that their results cannot be directly compared to one another and measured values from reference pieces must be performed with the same sensor as the actual component inspection is done.
Totally 24 samples were made of 300M steel and processed by heat treating, grinding, and shot peening. With the last, a compressive residual stress was induced to the surface of the samples. This was done in order to make the samples same as components in aviation industry. The compression is induced to hinder crack initiation and growth and prolong the service life. The samples were bent using four-point-bending in order to induce residual stresses in a controlled way and decrease the shot peened compressive residual stresses from the surface in a way to simulate relaxation of these stresses during component use. This way the reference pieces can be compared to a component in use, and it can be determined if it is no longer suitable for its purpose. Since BN inspection using Rollscan 350 does not provide absolute values on residual stress, X-ray diffraction (XRD) is used to correlate BN signals to real stress values.
In total 24 reference pieces were produced with a varying range of surface residual stresses. The residual stress results correlated well with literature. RMS signal was sensitive to the change in residual stress when compression decreased and especially when it changed into tensile. In high compression the RMS signal was very stable. This means that using BN inspection is possible to find areas in material surface where the residual stresses have significantly changed but further examination using XRD is needed for absolute values of residual stresses. It was concluded that by four-point-bending, it is possible to induce desirable residual stress state into the sample surface. Two types of BN sensors were used, and a clear difference was found in the signal level. It would suggest that their results cannot be directly compared to one another and measured values from reference pieces must be performed with the same sensor as the actual component inspection is done.