STATISTICAL ANALYSIS OF DYNAMIC FRACTURE TOUGHNESS DATA OF TWO DCI LARGE SCALE SPECIMEN TEST SERIES

Advanced codes and guidelines for the assessment of component safety include fracture mechanics concepts in many cases. Regarding this, extreme loading situations such as accidental conditions are of special interest within the design and proof of safety process. Dynamic (high-rate) loading causes fast changes of the stress and strain states in the material. Therefore, the determination of appropriate corresponding fracture mechanics material properties is an experimental challenge but at the same time it is an essential prerequisite for the practical applicability of the codes. The material in focus here is ductile cast iron (DCI). One major field of application of DCI is the production of casks for radioactive materials. At first the present paper deals with a BAM test series of dynamic fracture toughness tests in 2006. A series of 6 large scale SE(B)140 specimens of DCI was tested at -40 °C and loading rates of dK/dt ≈ 5·104 MPa√m/s. If the validity criteria of the static test standard ASTM E 399 are formally applied here, all data can be regarded as valid dynamic fracture toughness values KId. The statistical analysis of the data was based on a two parameter Weibull distribution function. The data was completed by 4 results of a identical BAM test series in 1998/1999. Weibull analyses of three samples covering different pearlite contents (≤ 4 %, ≤ 9 %, ≤ 20 %) were performed and characteristics of the distribution functions and two-sided confidence intervals were calculated. All investigated samples could be properly described by the applied Weibull distribution function. The calculated characteristics show that KId of DCI decreases with increasing pearlite content. These results are valuable since a statistical analysis of the dynamic fracture toughness of DCI depending on microstructure has been performed for the first time. This is especially important with respect to the production process of heavy-walled DCI castings where small amounts of pearlite are normally tolerated. Nevertheless, the presented results are still based on relative small sample sizes and therefore of preliminary nature. The statistics should be improved 1 by further analyses taking more specimens and especially more ones with higher amounts of pearlite into account.