Simplified calculation of thermal stresses — On the reduction of effort in the stress analysis of reactor components

Abstract

The fatigue behaviour of reactor components is predominantly determined from the in-service thermal stresses. The calculation of such stresses for a number of temperature transients in the adjacent fluid may be expensive, particularly with complicated structures. Under certain conditions this expense can be reduced considerably with the aid of a rule, which permits interpolation of thermal stresses from known reference values instead of calculation. This paper presents the derivation and method of application of this interpolation rule.The derivation procedure is based on well-known proportionalities between thermal stress range Δσ in the structure and temperature change ΔT and rate of change Ṫ of the fluid in the extreme cases of an ideal thermal shock and quasi-steady-state conditions, respectively. For the real transients in between the relationship Δσ ∝ (ΔT)xṪ1 − xαy is proposed, where x is the shock-degree and lies between 0 and 1, and, additionally, y designates the influence of the heat transfer coefficient α. This formula yields the interpolation rule.The rule permits interpolation of stress ranges for additional thermal transients from at least 3 reference stresses via x and y. The procedure is applicable to any metallic structure, reduces fatigue analysis effort considerably and yields excellent results.The paper is split up into 2 parts. In the following the derivation of the rule is presented. The second part describes its application and will be published shortly.