PSI - Issue 18

Christoph Bleicher et al. / Procedia Structural Integrity 18 (2019) 46–62 Author name / Structural Integrity Procedia 00 (2019) 000–000

49

4

2.3. Non-destructive tests on the tests blocks After being cast, the blocks with Dross were cut, so that only the upper 200 mm were used for the further tests. These flatter blocks were investigated non-destructively with ultrasonic testing (GINA (2003)) and magnetic particle inspection (GINA (1997)), to detect the positions of Dross and its expansion of the blocks’ surface. The ultrasonic tests were carried out in the foundries with handheld devices and subsequently with mechanised devices, providing a better resolution as well as a documented position of the test device and the Dross relative to a fixed point on the block. This offers the possibility to subsequently transfer the gathered data into numerical simulation, for instance to recalculate a cast component with the Dross region in the numerical model much more precisely. The mechanised ultrasonic tests were performed on the side of the test blocks with Dross and from the side opposite to the Dross region. The results of the ultrasonic tests are shown, in Fig. 3, for the measurements from the side with Dross and, in Fig 4, from the opposite side. The tests showed that Dross can be detected from the Dross side and from the opposite side from the Dross. Moreover, the ultrasonic test from the Dross side showed an even better resolution of the Dross. During the ultrasonic tests, depths of Dross of more than 60 mm could be determined precisely. During the ultrasonic tests, it emerged that all cast blocks have a certain amount of convexity on the surface, of 16 mm and more. This leads to challenges during the ultrasonic tests and also during the subsequent magnetic particle inspection. The magnetic particle inspection was performed according to GINA (1997) and done both in the black and-white and fluorescent methods. Results for both methods are shown in Fig. 5 and Fig. 6. While performing the magnetic particle inspection, it became obvious that the blocks had not only a convex surface, but also intolerable signals resulting from surface roughness of several millimetres in depth. For both reasons, all cast blocks were milled on the surface with Dross to create a more or less plane surface enabling a proper magnetic particle inspection on each block. Based on the magnetic particle inspection, fatigue specimens, Fig. 7, were positioned on each of the cast block’s surfaces. After the removal of the fatigue specimens, they were subsequently characterised non-destructively by means of computer-tomography and magnetic particle inspection to identify further information regarding the Dross formations in the specimens leading later to the fatigue of each specimen. Example of results of the computer tomography on a fatigue specimen without Dross and with Dross, as well as a magnetic particle result, are given in Fig. 8. The sulphur test, according to Baumann described in Kauczor (1957), is used to show where sulphur is present in the specimens, being accompanied by Dross. The sulphur test is conducted with the help of sulphuric acid and photographic paper, which contains bromine. For the test, the photographic paper is saturated with 5 % sulphuric acid and then pressed on the specimen for approximately 5 minutes. During this time, the reactions in Equations (1), (2) and (3) take place, leading to a brown to black coloured copy of the sulphuric regions in the specimen on the photographic paper Fig. 8. MnS � H � SO � → MnSO � � H � S (1) FeS � H � SO � → FeSO � � H � S (2) H � S � ����� → �� � S � �H�� (3) All of the applied non-destructive test methods are able to detect Dross, either in the volume (ultrasonic and X-ray computer tomography) or at the surface (sulphur tests and magnetic particle inspection). While the sulphur test is only able to detect Dross of a laminar form, magnetic particle inspection shows much more details for Dross forms, such as strings, cycles, pits and chunks. The determined NDT results help to identify the characteristics of Dross and provide a first step towards a method of transferring results from NDT to lifetime assessment. This could result, for instance, in a catalogue of pictures that foundrymen can use to compare the detected Dross on the component’s surface with results determined during the fatigue tests, for an initial lifetime estimation.