09-04-13 Topic — Impact of baffle design on vibration in shell-and-tube heat exchangers (Prize available for participation)

In traditional bow-shaped baffle heat exchangers, the function of the bow-shaped baffles is to cause the flow on the shell side to become a curved zigzag pattern, allowing the fluid to strike the tube bundle perpendicularly and thereby enhancing heat transfer. However, this arrangement of baffle plates also brings about many problems. 1) The arc-shaped baffle causes the fluid to impact the shell wall perpendicularly, resulting in a large pressure drop along the flow path ； 2) Flow stagnation zones are formed at the junctions of the baffle plates with the shell walls, which reduces the heat exchange efficiency. 3) Due to manufacturing tolerances and installation requirements, there are gaps between the tube bundle and the shell, resulting in significant bypass flow on the shell side, namely tangential flow near the shell walls. There is leakage between the baffle and the shell wall, as well as between the heat exchange tubes and the baffles. These bypass flows and leaks reduce the effective mass flow rate of the transverse tube bank, thereby decreasing the heat transfer efficiency on the shell side ； 4) A high mass flow rate flowing across the heat exchange tube bundle can induce vibrations in the tubes, thereby reducing the lifespan of the heat exchanger. Therefore, with this traditional bow-shaped baffle heat exchanger, to achieve a higher heat transfer coefficient, a considerable pressure drop along the flow path is generated, which means an increase in pump power consumption as a cost. To save energy and materials, there is an urgent need to change the traditional baffle structure. Modern industrial equipment has increasing demands on heat load, which requires the installation of heat exchangers with larger surface areas. However, increasing the number of heat exchangers in the existing process flow will inevitably lead to an increase in the total pressure drop along the flow path, which means higher pump power consumption is required. In some cases, this approach is not feasible. For example. Improvements are being made to the existing equipment; due to space constraints, it is not possible to install more heat exchangers ； Alternatively, the pump’s capacity was determined at initial installation without leaving enough headroom. With the development of industry, energy conservation has become increasingly important, and the requirements for heat exchangers have also become more stringent. It is necessary to achieve the desired heat transfer capacity while consuming less pump power. For bow-shaped baffle heat exchangers, achieving the desired heat transfer amount necessarily results in a considerable pressure drop. Therefore, it is necessary to reconsider the arrangement of the baffles. In summary, in bow-shaped baffle heat exchangers, the bow-shaped baffles cause the fluid to flow laterally across the tube bundle, thereby improving the heat transfer capacity of the fluid on the shell side. However, the sudden change in direction of the fluid near the shell wall leads to a rapid increase in energy loss, resulting in an increased pressure drop along the shell side. Furthermore, the presence of bypass flow between the baffles and the shell, as well as leakage and dead zones between the heat transfer tubes and the baffles (as shown in Figure 11), makes the drawbacks of the flow characteristics on the shell side quite evident. The spiral baffle heat exchanger was developed to address the shortcomings of the shell-side flow in bow-shaped baffle heat exchangers. In a spiral baffle heat exchanger (as shown in Figure 12), the flow direction of the fluid on the shell side changes continuously. There are no sudden direction changes in the flow, which greatly reduces flow resistance. The fluid has a spiral inclination relative to the heat exchange tubes; it flows around the tube bundle in a spiral pattern. This results not only in low flow resistance but also in the absence of dead zones, and the advantages of such flow characteristics are evident. The fundamental difference between the spiral baffle and the arc-shaped baffle lies in the variation of the baffle’s structural form within the housing. The arc-shaped baffle is arranged vertically within the shell in front of the heat exchange tubes, creating several parallel return channels on the shell side; the sudden change in flow direction of the fluid inevitably leads to significant pressure losses. This is the main reason for the high energy consumption of such heat exchangers. At the same time, in the transition zones where the direction changes in the two baffled channels, the fluid moves diagonally along the shortest path, thus forming a triangular area in which the medium remains relatively stationary. As shown, this stationary fluid reduces the heat transfer capacity of the tube surface, which is equivalent to a reduction in the heat transfer area. This also explains one of the main reasons why bow-shaped baffle heat exchangers cannot significantly improve heat transfer efficiency. The spiral baffle heat exchanger features a structure in which several fan-shaped plates with a quarter of the shell-side cross-section are assembled to form spiral baffles, enabling the fluid to move in a spiral pattern from the inlet to the outlet of the shell. This approach avoids the severe pressure losses associated with large-angle turns, resulting in low pressure drops. Therefore, it is possible to adjust the flow cross-section by using different angles. The high heat transfer coefficient of the medium stems from a high Re number, while a high flow velocity is an important factor in increasing the Rc number. Traditional arc-shaped baffle heat exchangers suffer from significant velocity reductions due to their large dead zones. High pressure drops are a problem; the introduction of a spiral baffle structure has changed the internal layout of the shell. By maintaining low pressure drops, it is possible to increase the flow velocity of the fluid, thereby increasing the film heat transfer coefficient through an increase in the Re number. The main reason for this is that this spiral structure causes the medium to form vortices. A large velocity gradient is generated in the radial direction from the center, causing turbulence in the fluid on the tube surface and thereby thinning the boundary layer. It helps to increase the heat transfer coefficient. Furthermore, the continuous spiral supports reduce the span between the tubes, allowing the natural frequencies of the tubes to avoid the excitation frequencies of the fluid and thus preventing damage caused by resonance. This thus extends the lifespan of the equipment and reduces maintenance costs. Due to the effective scouring by the fluid, the deposition of dirt is also reduced, allowing the heat exchanger to operate efficiently over a long period of time and thus achieving energy savings. This post was last edited by zx9527007 on 2009-4-13 00:04.]

1. Only when the frequency of vibrations induced by the fluid matches or is quite close to the natural frequency of the heat exchange tube, will the amplitude of vibration in that tube increase sharply, leading to its damage. 2. No single type of baffle can be considered as inherently causing vibration; the occurrence of vibration is related to many factors such as the spacing between baffles, the spacing without supports, the natural frequency of the heat exchange tubes, the precision of baffle manufacturing, the excessive thinness of the baffles, as well as the feed rate, the flow rates in the shell side and tube side, and the installation of anti-erosion plates. 3. The pressure drop of disc-type baffle plates is relatively low; given that the heat exchange area of the equipment and other material parameters remain the same, theoretically, the likelihood of vibration should be low. 4. If pipes are arranged at the arc-shaped notch of the arc-shaped baffle, it is easy to cause vibrations

Baffles in the form of double arches, spirals, or baffle rods can effectively prevent vibrations.

I also read in the heat exchanger design manual that the pressure drop associated with disc-type baffle plates is lower than that of arc-shaped baffle plates. Do you guys use the same flow velocity for these two types when designing? As mentioned on the second floor, any type of baffle can cause vibration; if I were to make a comparison, the situation should be the same. Could you explain the conditions in detail?

The flow of the medium across the disc-shaped baffle is mostly a parallel flow directed toward the tube bundle; therefore, the flow resistance is lower than that with single arch-shaped baffles, and the vibrations caused by the transverse flow perpendicular to the tubes are reduced. This is what the book says: that is, when vibration is caused by transverse flow, the disc-ring shape has an advantage over the single-arch shape!

I haven’t used anything else; for arc-shaped baffles, those with double arc-shaped baffles result in the lowest pressure drop, while those with single arc-shaped baffles and NTIW (no tube in window) exhibit a higher pressure drop. However, according to the HTRI selection calculations, it is not necessarily the case that double-arch baffle plates result in a lower probability of vibration. I don’t know why; could some expert explain it? The NTIW format feels relatively solid. There are many factors that can cause vibrations, and they are interrelated. The type of baffle, its spacing, the way in which pipes are arranged, the distance between pipes, the presence of supports, and even the size of the pipe ends (which can cause acoustic vibrations) all play a role.

Yes, you’re right – what I meant is to make comparisons under identical design conditions; without this constraint, the discussion makes no sense. Currently, there’s a discrepancy between what’s stated in the design documents and the actual reality, and we should focus our discussion on this issue as the key point

1. Only when the frequency of vibrations induced by the fluid matches or is quite close to the natural frequency of the heat exchange tube, will the amplitude of vibration in that tube increase sharply, leading to its damage. 2. No single type of baffle can be considered as inherently causing vibration; the occurrence of vibration is related to many factors such as the spacing between baffles, the spacing without supports, the natural frequency of the heat exchange tubes, the precision of baffle manufacturing, the excessive thinness of the baffles, as well as the feed rate, the flow rates in the shell side and tube side, and the installation of anti-erosion plates. 3. The pressure drop of disc-type baffle plates is relatively low; given that the heat exchange area of the equipment and other material parameters remain the same, theoretically, the likelihood of vibration should be low. 4. If pipes are arranged at the arc-shaped notch of the arc-shaped baffle, it is easy to cause vibrations

It’s not possible to generalize which type of baffle will cause vibration; the occurrence of vibration is related to many factors such as the spacing between baffles, the spacing without supports, the natural frequency of the heat exchange tubes, the precision of baffle manufacturing, the excessive thinness of the baffles, as well as the feed rate, the flow rates in the shell side and tube side, and the arrangement of anti-erosion plates

Baffles cause the fluid to flow across the tube bundle; while this enhances heat transfer, it also induces vibrations in the fluid. For baffled tubular heat exchangers, measures to reduce the amplitude of these induced vibrations include: 1) reducing the flow velocity of the fluid across the tube bundle； 2) Increase the natural frequency of the heat transfer element, such as by increasing the wall thickness of the tube or reducing its span. Obviously, these two measures are contradictory, because to reduce the span of the tubes, it is necessary to increase the number of baffle plates, and an increase in the number of baffle plates in turn increases the flow velocity within the transverse tube bundle. The vibration generated by a baffle rod heat exchanger should be relatively low; the tubes in the notched area can be removed when installing the arc-shaped baffles

I learned a lot from the answers given by the people above. Can we consider this issue from another perspective? For example, whether resonance occurred under their operating conditions.

I don’t understand. By reading your discussions, I’ve learned*.

I’m not very familiar with the internal structure of heat exchangers, but the heat exchangers used in our factory cost tens of millions of yuan each and can withstand temperatures of up to 850°C – which is already considered very advanced. There are also smaller water heaters that use steam. I would like to ask how to determine the direction of flow in the tube side and shell side, whether it should be in the same direction or opposite directions. Additionally, could you please send me some diagrams of their internal structure? To be honest, I don’t even know what baffle plates are or where they are located. Thank you

It seems that the shock-absorbing effect of arc-shaped baffle plates is greater than that of circular baffle plates

The main reason for the vibration, I believe, is not whether the heat exchanger is of disc or bow type; the key issue, I think, lies in the fact that the operating conditions and process requirements were not properly taken into account during the design phase. When I visited a chemical plant before, I noticed that the outlet pipe of the flash heat exchanger there was vibrating severely. It was intermittent vibration, occurring every ten or so minutes, which seemed strange. Upon closer inspection, it was found that a set of control valves were installed on the outlet pipe of the heat transfer medium in that heat exchanger, intended to regulate the pressure in the flasher ahead of the heat exchanger. The design was unreasonable, as this caused the pipe to carry either condensed water or gas at different times; the flow rates of these two media differed greatly, which is why vibration occurred Later, when our unit was installed, we made modifications in that area, and the vibration completely disappeared.

There are many reasons for vibration. Just looking at baffle plates, I believe that the vibration caused by single-arch baffles is greater than that caused by disc-type baffles. Additionally, vibration is also related to the spacing between the baffle plates; when the spacing is smaller, the pressure drop increases and so does the vibration.