Session: MF-20-01 Material Quality and Failure Analysis - 1

Paper Number: 106094

106094 - Effect of Normalizing Temperature on Impact Toughness of Asme Sa - 350 Lf2 Cl1 Forgings 

The present work is a continuation of paper PVP2022-84549.

Four equal pieces of 5.7 kilograms from a ring forged by ring rolling were used for this study. Normalizing was done in a laboratory electrical furnace at 870ºC, 885ºC, 900ºC and 915ºC. In all cases holding time was 3 hours, which was enough to ensure a homogeneous temperature distribution along the thickness of the pieces. Cooling from normalizing to room temperature proceeded under forced air with the help of a blower. Temperature in the center of each piece was registered by thermocouples every two seconds. Cooling rates were calculated through Newton equation considering the temperatures registered between 700 and 300 ºC. Values obtained were similar, therefore differences between microstructural features of the four pieces could be connected to the normalizing temperature solely.

Once the normalizing was completed each piece was metallographically assessed. Furthermore, open-source image analysis software ImageJ was used on SEM micrographs to evaluate the size distribution of pearlite nodules.

Finally, twelve Charpy (simple-beam) impact test full size specimens following ASTM SA – 370 were taken from each piece. Impact test was performed at – 46 ºC.

The four different normalizing temperatures yielded impact test values above the minimum threshold of SA – 350 for LF2 Cl1 classification. However, the specimens normalized at 870 ºC showed the best performance, being significantly higher (122 Joules average – 53 Joules minimum) than specimens normalized at 915 ºC (86 Joules average – 33 Joules minimum). Normalizing temperatures of 885 ºC and 900 ºC behaved similarly (104 J and 113 J average, respectively).

From a metallographic stand point, it is remarkable that the ferrite – pearlite balance was not affected by the normalizing temperature. All samples showed a content of 76% ferrite and 24% pearlite. It is only the sample for normalizing temperature of 900ºC which deviated somewhat from these figures (73% ferrite and 27% pearlite). The effect of normalizing temperature on grain size is subtle. The 870 ºC samples showed a slightly refined grain size (ASTM No. 9 – 10) when compared with the rest of temperatures where the grain size obtained was ASTM No. 9. However, the effect of the normalizing, though mild, is more evident on the prior austenite grain size, which becomes coarser as the normalizing temperature increases (915 ºC yielded an ASTM No. 8, while 870ºC led to ASTM No. 9).

The morphological analysis of SEM micrographs by means of ImageJ open-software revealed that even though the size of the coarsest measured pearlite nodule increases with the normalizing temperature, such trend is missing when the average sizes of the pearlite nodules are considered. Consequently, average nodules size seems unrelated to the normalizing temperature. As a matter of fact, approximately 90% of the nodules measured are relatively small in size (< 200 µm²) for all the samples. It is in these small-size nodules where a clear difference is appreciable among the samples. At 915 ºC nodules are made of one colony and therefore, they show a single orientation of the cementite and ferrite lamellae. Nevertheless, at lower normalizing temperatures the nodules are made of more than one pearlite colony and consequently, the misorientation within the nodule increases allowing a more effective resistance to brittle crack propagation since colony boundaries function as an additional barrier.

This dissimilarity in the pearlite nodules has its origin in the differences found in the prior austenitic grain size of samples. Finer sizes imply a larger austenite grain boundary area, and hence greater nucleation sites for proeutectoid ferrite in the cooling process. Since pearlite colonies are formed from cementite thin films which previously nucleated at the proeutectoid ferrite – austenite (α/γ) interface, a larger number of pearlite colonies would be formed from finer austenitic grains. These colonies would eventually impinge on each other forming a pearlite nodule. As mentioned above, these nodules are more effective to arrest the propagation of a crack.

Presenting Author: Ricardo Hernández Soto TECNICAS REUNIDAS

Presenting Author Biography: Ricardo Hernandez Soto is the Head of Inspection at TECNICAS REUNIDAS. He holds a M.S in Chemistry at the University of Seville, Spain. He is certified as International Welding Engineer by International Institute of Welding (IIW) and as Welding Educator by AWS. He has 19 years of experience. He has previously worked as Site Quality Control Manager in different international assignments with TECNICAS REUNIDAS and Wood. He is currently pursuing a Ph Degree in Chemical Engineering and Material Science at Complutense University of Madrid.

Authors:

Ricardo Hernández Soto TECNICAS REUNIDAS
José María Gómez De Salazar Complutense University of Madrid