Bolted Joint Analysis: Fatigue Considerations

Bolted connections are the workhorses of mechanical design, but in dynamic applications—from vibrating machinery to aircraft structures—they face a silent, insidious enemy: fatigue. A bolt that can easily sustain a massive static load may fracture unexpectedly under a much smaller, repeated force. The core principles for analyzing and preventing fatigue failure in bolted joints involve understanding how cyclic loading causes cracks to initiate and propagate, typically in the highly stressed bolt threads or at the fillet under the bolt head. Understanding how to manage the stress state within the bolt is the key to designing joints that last for millions of cycles.

The Critical Role of Preload in Fatigue Resistance

The single most important factor in combating bolt fatigue is not the bolt's ultimate strength, but the preload introduced during tightening. Preload is the tensile force intentionally created in the bolt when the nut is torqued, which in turn creates a clamping force that squeezes the joined plates together. In a perfectly tightened joint, the preload should be a significant percentage of the bolt's proof strength.

Why does this help fatigue? Imagine an external cyclic force trying to separate the joint. In a joint with little or no preload, that external force is borne almost entirely by the bolt itself, causing its stress to cycle from a low to a high value. However, in a properly preloaded joint, the clamping force is so high that the applied cyclic load first must overcome the friction between the plates and then simply reduce the clamping pressure. The bolt only sees a small fraction of the external load. This dramatically reduces the alternating stress amplitude (${\sigma }_a$) on the bolt, which is the primary driver of fatigue damage. The preload establishes a high mean stress (${\sigma }_m$) from which the stress only oscillates slightly.

Analyzing Stresses: Mean and Alternating Components

To perform a quantitative fatigue analysis, you must separate the total stress in the bolt into its constant (mean) and variable (alternating) parts. The mean stress (${\sigma }_m$) comes predominantly from the preload. You calculate it using the bolt's tensile stress area ($A_t$) and the preload force ($F_i$): ${\sigma }_m=F_i/A_t$.

The alternating stress (${\sigma }_a$) is more complex. When an external cyclic force ($F_{ext}$) is applied, only a portion of it acts to further stretch the bolt; the rest works to relieve compression on the clamped plates. This portion is defined by the bolt fraction or load factor ($\Phi$). For a simple joint with identical plates, $\Phi$ is determined by the relative stiffness (spring rates) of the bolt and the clamped material. A stiffer bolt or more compliant clamped members leads to a higher $\Phi$, meaning the bolt carries more of the external load—a worse scenario for fatigue.

The alternating stress is then calculated from the portion of the external load range ($\Delta F_{ext}$) that the bolt feels: ${\sigma }_a=(\Phi \cdot \Delta F_{ext})/(2A_t)$. The factor of 2 appears because $\Delta F_{ext}$ is the peak-to-peak load range, and the stress amplitude is half of that range.

The Modified Goodman Diagram: A Tool for Prediction

Engineers use fatigue failure criteria to determine if a given combination of mean and alternating stress is safe for a required number of cycles. The modified Goodman diagram is a common and conservative tool for this assessment. It is a graph with mean stress (${\sigma }_m$) on the horizontal axis and alternating stress (${\sigma }_a$) on the vertical axis.

The diagram plots several key lines:

• The vertical line at the bolt's ultimate tensile strength ($S_{ut}$).
• The horizontal line at the bolt's fully reversed endurance limit ($S_e$)—the alternating stress it can withstand for infinite life when the mean stress is zero.
• The modified Goodman line, which connects $S_e$ on the vertical axis to $S_{ut}$ on the horizontal axis.

To use the diagram, you plot a point with coordinates (${\sigma }_m$, ${\sigma }_a$) representing the bolt's operating stress state. If this point lies below the modified Goodman line, infinite life is predicted. The factor of safety can be found graphically or by using the Goodman equation:

$$\frac{{\sigma }_a}{S_e}+\frac{{\sigma }_m}{S_{ut}}=\frac{1}{n}$$

where $n$ is the factor of safety. A value of $n>1$ indicates a safe design for infinite life. This model clearly shows the dual benefit of high preload: it raises ${\sigma }_m$ but drastically reduces ${\sigma }_a$, often moving the operating point into a much safer region of the diagram.

Improving Fatigue Life: Design and Manufacturing Choices

Analysis guides design, but specific choices directly enhance fatigue performance. First, ensure preload is consistently high and accurate. Use torque-turn methods or direct tension indicators instead of simple torque, which is highly affected by friction. Second, consider joint stiffness. Designing stiffer clamped members (e.g., using thicker plates or higher-modulus materials) lowers the bolt fraction ($\Phi$), diverting more of the cyclic load away from the bolt.

Perhaps the most effective manufacturing improvement is the use of rolled threads. Unlike threads that are cut (machined), which cut through the material's grain structure, rolled threads are formed by cold-working the material. This process creates a favorable grain flow that follows the thread contour, introduces compressive residual stresses at the critical root of the thread, and results in a smoother surface finish. All three factors—better microstructure, compressive stress, and reduced surface flaws—combine to significantly raise the bolt's endurance limit ($S_e$), directly improving its performance on the Goodman diagram.

Common Pitfalls

1. Ignoring Preload Variability: Assuming the specified torque yields the exact desired preload is a major error. Friction can cause preload to vary by ±25% or more. Correction: Account for this variability in your safety factor calculations or use more precise tensioning methods for critical joints.

1. Overlooking the Bolt Fraction: Treating the bolt as if it carries the full external load ($\Phi =1$) is excessively conservative for some designs but dangerously non-conservative for others (e.g., with soft gaskets). Correction: Always perform a stiffness analysis to estimate the true load factor $\Phi$ for your specific joint geometry and materials.

1. Misapplying the Endurance Limit: Using the endurance limit for a polished lab specimen ($S_e^{\prime }$) without applying the necessary modifying factors for your real bolt. Correction: Calculate the corrected endurance limit: $S_e=k_a{k}_b{k}_c{k}_d{k}_e{S}_e^{\prime }$, where factors account for surface finish ($k_a$), size ($k_b$), load type ($k_c$), temperature ($k_d$), and other effects ($k_e$).

1. Under-Tightening to "Reduce Stress": Intuitively, one might think less preload means less stress. For fatigue, this is backwards. Correction: Maximize the permissible preload to minimize the alternating stress component, staying within the bolt's proof strength to avoid yield during installation.

Summary

• Preload is paramount. A high, consistent clamping force dramatically reduces the alternating stress amplitude the bolt experiences under cyclic external loads, shifting the stress state to a more favorable region for fatigue life.
• Fatigue analysis separates stress into a mean component (from preload) and an alternating component (from the bolt's share of the external load), which are evaluated together using tools like the modified Goodman diagram.
• The bolt fraction or load factor ($\Phi$) determines what portion of an external load is carried by the bolt itself; designing for lower $\Phi$ (stiffer clamped members) improves fatigue performance.
• The modified Goodman criterion provides a conservative method for infinite-life design, requiring the sum of the ratios of alternating stress to endurance limit and mean stress to ultimate strength to be less than 1 (divided by the safety factor).
• Manufacturing matters. Rolled threads, through grain flow and compressive residual stresses, offer a significantly higher fatigue resistance compared to cut threads and are preferred for demanding dynamic applications.