Leveraging comprehensive baseline datasets to quantify property variability in nuclear-grade graphites

• Published in: Nuclear engineering and design 2016-10, Vol.307 (C), p.77-85
• Main Authors: Carroll, Mark C., Windes, William E., Rohrbaugh, David T., Strizak, Joseph P., Burchell, Timothy D.
• Format: Article
• Language: English
• Subjects: Billets; Compressive strength; Economics; Graphite; Irradiation; Nuclear reactor components; Process parameters; Raw materials
• ISSN: 0029-5493; 1872-759X
• DOI: 10.1016/j.nucengdes.2016.06.028

[Display omitted] •An effort is underway to fully quantify the properties of nuclear-grade graphites.•Physical and mechanical properties of graphite are best characterized by distributions.•The Weibull distribution is most representative of graphite based on goodness-of-fit.•Fine-grained isomolded grades exhibit higher Weibull modulus values, indicative of more homogeneous properties. The full characterization of the physical and mechanical properties of candidate nuclear-grade graphites is highly dependent upon an understanding of the distribution of values that are inherent to graphite. Not only do the material properties of graphites vary considerably between grades owing to the raw materials sources, filler particle type and size, methods of compaction, and production process parameters, but variability is observed between billets of the same grade from a single batch and even across spatial positions within a single billet. Properly enveloping the expected properties of interest requires both a substantial amount of data to statistically capture this variability and a representative distribution capable of accurately describing the range of values. A two-parameter Weibull distribution is confirmed to be representative of the distribution of physical (density, modulus) and mechanical (compressive, flexure, and tensile strength) values in five different nuclear-grades of graphite. The fine-grained isomolded grades tend toward higher Weibull modulus and characteristic strength values, while the extruded grade being examined exhibits relatively large distributions in property values. With the number of candidate graphite specimens that can undergo full irradiation exposure and subsequent testing having limited feasibility with regard to economics and timely evaluations, a proper capture of the raw material variability in an unirradiated state can provide crucial supplementary resolution to the limited amount of available data on irradiated candidate grades.