Designing for stability is especially important because many variables that affect joint behavior cannot be directly controlled. Some of the major causes of uncertainty include:
Assembly: Clamping force is established by assembly operations, and it is subject to a great deal of variation. About 75 factors affect the tension created by applying torque to a single bolt, and more than 250 variables come into play in a group of bolts. Many of these variables can be identified, but they rarely can be characterized for a production lot.
Service loads: Through testing and experience, some industries can quantify the loads on their joints with a fair degree of certainty. However, while manufacturers can predict the loads their products should be exposed to, they cannot quantify conditions the product might experience through customer ignorance or abuse. Even when products are used properly, service conditions may vary.
Load distribution: Another area of uncertainty is the way parts of a bolted joint share external loads. Here, the stiffness ratio between bolts and joint members is key. Relatively flexible bolts will experience less change in stress for a given change in tensile load on the system, which can enhance bolt fatigue life.
Bolt stiffness may be uncertain because it depends on effective bolt length, the length of the bolt that acts as a spring. There are several ways to estimate effective length. One rule of thumb assumes that one-half the nut or hole threads are loaded. However, the accuracy of this rule varies with joint geometries and conditions.
The greater the length-to-diameter ratio of the bolt, the less effect thread engagement has on effective length. However, on short bolts, thread engagement can significantly affect effective length and stiffness estimates.
Thermal effects: The response of bolts and joints, especially gasketed ones, to temperature change is also uncertain. It is known that high temperature affects the strength and modulus of bolting materials. However, these properties may vary from lot to lot and manufacturer to manufacturer, so quantitative evaluations of temperature effects are difficult.
Data about expansion coefficients of bolting materials and stiffness ratios, needed to estimate the change in clamping force created by expansion or contraction, are also limited. Even less information is available about other material properties. Temperature influences failure modes such as stress-corrosion cracking, hydrogen embrittlement, fatigue, and corrosion, making exact predictions rarely possible.
Also, gasketed joints often start leaking after a thermal change or series of thermal cycles, but the reasons are unknown. Factors may include creep of gaskets and metal parts, degradation and change in gasket-material properties, and loss of clamping force on the gasket caused by cycles of thermal expansion and contraction.