This paper compares material issues for cermet and graphite fuel elements. In particular, two issues in NTP fuel element performance are considered here: ductile to brittle transition in relation to crack propagation, and orificing individual coolant channels in fuel elements. Their relevance to fuel element performance is supported by considering material properties, experimental data, and results from multidisciplinary fluid/thermal/structural simulations. Ductile to brittle transition results in a fuel element region prone to brittle fracture under stress, while outside this region, stresses lead to deformation and resilience under stress. Poor coolant distribution between fuel element channels can increase stresses in certain channels. NERVA fuel element experimental results are consistent with this interpretation. An understanding of these mechanisms will help interpret fuel element testing results. Nomenclature a = flaw length, m CTE = coefficient of linear thermal expansion, m/m-K E = modulus of elasticity (Young's modulus), MPa g = acceleration due to gravity, m/s 2 G, G P = total energy dissipated, energy dissipated due to plastic deformation, J/m 2 I sp = specific impulse, s k = coefficient of thermal conductivity, W/m-K %ΔL/L o = percentage change in length due to linear thermal expansion, % MW = molecular weight, kg/mol MW th = megawatts of thermal energy, MW R u = universal gas constant, J/mol-K T = temperature, K y + = normal boundary layer grid spacing at wall, normalized α = coefficient of linear thermal expansion (CTE), m/m-K γ = ratio of specific heats γ = surface energy, J/m 2 σ f = stress at fracture, kPa A -25% B = metal matrix/ceramic particle mixture of metal A and ceramic B, % volume A /25% B = alloy of metals A and B, % volume % = compositions are percentage by volume, unless otherwise marked as % weight