Expansion joints evaluation

For a high-temperature system, a separate subheader may be run up to the point where the temperature drops down to the allowable limit of a less expensive material. It may then be connected to the main flare header (either low pressure or high pressure).To properly evaluate this a heat loss calculation is needed. As a rule of thumb a heat loss of 10 BTU/hr/ft may be assumed for a quick estimate for bare pipe. Consideration should also be given to the need for expansion joints. Main flare headers may be as large as 36 to 42 inches in diameter for a large-capacity plant. Expansion joints of such magnitudes may be so expensive as to call for a separate small header for the hot flare system. 

The following are specific considerations to be evaluated by the designer when specifying expansion joint requirements, in addition to the guidelines given in EJMA standards ... 

A thermal expansion joint may be installed on the exchanger shell. As plant size increases, the shell-and-tube heat exchanger becomes more cost effective and will be the exchanger of choice. At intermediate sizes, both types must be evaluated for process and economic reasons. 

The manufacturer s allowable displacement limits for expansion joints for the existing expansion joints (see Hartman, et al. Evaluation of Expansion Joints in the Reactor Process Water Piping, DuPont, July 1988) and the new Model X expansion joints (see Specification for Procurement of Metallic Bellows-Type Expansion Joints in the Reactor Process Water, Westinghouse, July 1989) shall be used. Use NC-3649 as the acceptance criteria for stress computation if the manufacturer s allowable displacement limits cannot be met and code stress computation is used to show acceptability. 

Data for thermal movement of various bitumens and felts and for composite membranes have been given (1). These describe the development of a thermal shock factor based on strength factors and the linear thermal expansion coefficient. Tensile and flexural fatigue tests on roofing membranes were taken at 21 and 18°C, and performance criteria were recommended. A study of four types of fluid-applied roofing membranes under cyclic conditions showed that they could not withstand movements of <1.0 mm over joints. The limitations of present test methods for new roofing materials, such as prefabricated polymeric and elastomeric sheets and liquid-applied membranes, have also been described (1). For evaluation, both laboratory and field work are needed. 

More sophisticated approaches to the evaluation of wfr) exist. Some of them are the technique based on the Yoon-Flory expansion, variations of the Gaussian approximation, and an implementation of the Koyama distribution. Freely jointed chains and rotational isomeric chains have also been used as models of chains. Research is actively pursued on the subject of formulating more realistic closure relationships. In considering the choice of an intramolecular structure factor, a paper by McCoy et al. on model polyethylene melts will be useful. 

Stress relaxation under constant strain has been reviewed. It relates to applications which are kinematically evaluated and for which applied displacements are constant in time. Examples are bolted joints or plastics subjected to a mismatch-fit into a much stiffer structure or having a different coefficient of thermal expansion. 

Recent theoretical studies have become much more complex. New computer-assisted techniques permit the use of finite-element matrix-theory type approaches. The effects of important variables are being determined by parametric studies. More complex joints are also being studied. New adherend materials, including advanced filamentary composites, are also being evaluated. The elastic, low-deflection, constant temperature behavior of scarf and stepped-lap joints has been replaced by elastic-plastic, large-deflection behavior, combined with thermal expansion differences, or curing shrinkage-induced residual stresses. 

For a press-fit joint, the effect of thermal cycling, stress relaxation, and environmental conditioning must be carefully evaluated. Testing of the factory assembled parts under expected temperature cycles, or under any condition that can cause changes to the dimensions or modulus of the parts, is obviously indicated. Differences in coefficient of thermal expansion can result in reduced interference due either to one material shrinking or expanding away from the other, or it can cause thermal stresses as the temperature changes. 

The model Is based on elastic-plastic fracture mechanics principles, and Incorporates effects associated with thermal expansion mismatch and modulus mismatch of various constituents, as well as non-linear material behavior as a function of load and temperature. Key properties of the constituents, such as those of the interlayer, reaction zone, and base material are provided as a data base these data were measured in this program by using bulk samples, The model then uses the processing history, specimen geometry and loading conditions to evaluate the performance of the joint, The results of finite element analysis of cracked specimens have been consolidated In arriving at the engineering model, JADM,... 

---