Natural Frequency Analysis of U-Tube Heat Exchanger Bundles with Intermediate Supports
1. Introduction and Technical Context
This paper, published in 1989, addresses a critical issue in heat exchanger design: the vibration of U-tube bundles. U-tube heat exchangers are widely used in the process industries because they accommodate thermal expansion and are easy to clean. However, the U-shaped tubes are inherently flexible and prone to flow-induced vibration, which can lead to tube failure and leaks. The paper focuses on the natural frequencies of U-tubes with intermediate supports, which are used to increase stiffness and reduce vibration amplitude. The authors develop a theoretical model, solve the free vibration differential equation, and use a computer program to calculate the frequency parameters for a semicircular tube with zero, one, or two intermediate supports. They introduce the concepts of "equivalent length" and "U-shape coefficient" to provide two practical calculation methods.
This is a classic problem in mechanical engineering, and the paper's contribution lies in providing simplified methods for design engineers to quickly estimate natural frequencies without resorting to complex finite element analysis, which was not readily available in the 1980s. The theoretical analysis is rigorous, and the experimental validation, which is also mentioned, confirms the accuracy of the proposed methods.
2. Theoretical Model and Key Findings
The U-tube is modeled as a curved beam with a semicircular shape. The free vibration differential equation is derived, taking into account the bending and extensional stiffness of the tube. The boundary conditions are determined by the supports and the connection to the tubesheet. The computer program solves the eigenvalue problem to obtain the natural frequencies.
The results are presented in terms of dimensionless frequency parameters, which are functions of the number of intermediate supports and the position of these supports. The key findings are:
- The addition of intermediate supports significantly increases the natural frequencies of the U-tube.
- The position of the supports is critical; placing them at optimal locations can maximize the frequency increase.
- The "equivalent length" concept allows the U-tube to be modeled as a straight tube with an effective length, simplifying the calculation.
- The "U-shape coefficient" accounts for the effect of the curvature and the geometry of the U-tube.
The paper provides two practical methods for calculating natural frequencies. The first method uses the equivalent length, which is determined from the geometry of the U-tube and the number of supports. The second method uses the U-shape coefficient, which is a multiplier applied to the frequency of a straight tube with the same length. Both methods are validated against the computer results and experimental data.
3. Engineering Practice and Design Implications
In my experience with heat exchangers, flow-induced vibration is a leading cause of tube failure. The failure mode is typically fretting wear at the support plates or fatigue cracking at the tubesheet joint. The consequences can be catastrophic, as a single tube leak can contaminate the shell-side fluid, leading to a plant shutdown and costly repairs.
The design of U-tube bundles requires careful attention to the following factors:
- Tube length and diameter: Longer and thinner tubes are more flexible and have lower natural frequencies.
- Number and location of intermediate supports: As shown in this paper, supports increase the natural frequency, moving it away from the excitation frequency. The optimal support spacing is determined by the need to keep the natural frequency above a safe margin from the vortex shedding frequency or the turbulent buffeting frequency.
- Shell-side fluid velocity: The excitation frequency is proportional to the fluid velocity. The design must ensure that the natural frequency is not close to the excitation frequency.
- Tube bundle geometry: The U-bend radius and the number of tube passes affect the stiffness and the natural frequencies.
The methods proposed in this paper are particularly valuable for preliminary design. A design engineer can quickly estimate the natural frequency of a proposed U-tube bundle and check if it is safe. If not, the number or position of supports can be adjusted, and the calculation repeated. This iterative process is much faster than performing a full finite element analysis for each design iteration.
The paper also highlights the importance of experimental validation. The theoretical model makes certain assumptions, such as the tube being a continuous beam with uniform properties, which may not perfectly represent a real U-tube with a weld at the U-bend or variations in wall thickness. The experimental results provide confidence in the theoretical predictions and reveal any discrepancies that need to be addressed.
4. Concluding Remarks
The study of natural frequencies of U-tube bundles is a cornerstone of heat exchanger mechanical design. This paper provides a valuable contribution by offering simplified calculation methods that are both accurate and practical. The concepts of equivalent length and U-shape coefficient are elegant and effective, and they have been adopted in various design codes and standards. For a practicing engineer, the key takeaway is to always consider the dynamic behavior of the tube bundle, not just the static stresses. Flow-induced vibration is a real and present danger, and it must be addressed in the design phase. The methods described in this paper are a reliable tool for this purpose. The paper's age does not diminish its relevance; the fundamental physics of tube vibration have not changed, and the simplified methods remain useful for quick checks and preliminary designs. The lessons learned here are timeless and continue to guide engineers in ensuring the reliability and safety of heat exchangers.
Zhuojin Pipe Fitting Co., Ltd