Fig. 1 A typical stainless-steel shell & tube heat exchanger (S&THE) (Photo courtesy of XLG Heat Transfer, Spain).

Shell & tube heat exchangers (S&THE) have been used for many years in a wide range of industries, including aerospace, oil refineries, chemical processing, beverage industries, and pharmaceutical industries, to name a few.

Brazing has become more and more the desired joining-method to “seal” these heat exchangers from any internal or external leaks. In this article I will look at some of the internal characteristics of S&THE’s, options for sealing them to make them leak-tight, and then at the end of the article, I will discuss some important requirements for proper joint design so that brazing can be used to cost-effectively make many tube-to-header leak-tight joints in S&THE’s at the same time!

A typical shell & tube heat exchanger is pictured in Fig. 1, and a cross-sectional view of one is shown in Fig. 2, where it can be seen that an outer shell surrounds a bundle of tubes within that shell.

Fig. 2 A cross-section of a typical shell & tube heat exchanger (S&THE) in which the tubes are u-shaped, so that the internal fluid enters and exits from the same side of the heat-exchanger (drawing as shown in Wikipedia).

Fig. 2 A cross-section of a typical shell & tube heat exchanger (S&THE) in which the tubes are u-shaped, so that the internal fluid enters and exits from the same side of the heat-exchanger (drawing as shown in Wikipedia).

As can be seen in Fig. 2, one fluid or gas-stream flows through the tubes, while another fluid or gas-stream is directed into the shell itself and flows over and around those tubes. The colder stream of gas or liquid will extract heat from the warmer gas or fluid. As an example, if you had a hot liquid (or gas) flowing through the tubes, and a colder stream of gas or liquid flowing through the shell surrounding the tubes, then the gas/liquid flowing through the shell will extract heat from the hotter gas/liquid flowing through the tubes, causing the exiting gas/liquid from the tubes to be colder than when it entered the heat exchanger. The extent of cooling will, of course, depend on the flow-rates of the gas or liquid in the shell relative to that in the tubes. Thus, the slower the flow rate of a hot liquid (or gas) through the tubes, the longer it will take for that liquid (or gas) to make its way through the heat exchanger, and thus, more heat can be extracted before it exits the heat-exchanger.

NOTE: Please note that shell & tube heat exchangers (S&THE) can be gas/gas, gas/liquid, or liquid/liquid.

Sealing the Tube-to-Header Joints

A critically important part of the success of any S&THE is that the two fluid streams (one through the shell housing and the other through the tubes inside the shell) must be kept completely separate from each other at all times. Thus, as shown in Fig. 2, there must be a complete, tight seal between the tubes and the tube-sheet (header plate) to prevent any cross-leakage of fluids, i.e., there cannot be the slightest bit of intermixing of either flow-stream within the plenums on the right side of Fig. 2. So, the tube-through-header joints must be totally sealed.

Sealing methods for those joints can vary considerably, depending on the chemistry of either (or both) gases/fluids, the service temperatures of the gases/fluids involved, and the pressure at which the S&THE is operating. Based on those factors, leak-tight seals between the tube and header might be achieved by the use of rubber seals, or perhaps by adhesive bonding, soldering, welding, or by brazing.

Rubber-Seals

If the operating temps are below about 275F (135C), operating pressures are less than about 150 psi (1000 kPa) and the chemistries of the gases/fluids are not too aggressive, then rubber gaskets can be used to seal those tube-through-header joints, as shown in Fig. 3.

Fig. 3 Black rubber gaskets used to seal the joints through the header-plate. Such sealing is okay if the temperatures and pressures are low and the fluids/gases are not overly corrosive (drawing courtesy of L&M Radiator, Inc. Hibbing, MN).

Fig. 3 Black rubber gaskets used to seal the joints through the header-plate. Such sealing is okay if the temperatures and pressures are low and the fluids/gases are not overly corrosive (drawing courtesy of L&M Radiator, Inc. Hibbing, MN).

Adhesives or Soldering

Similarly, for situations with reasonably low operating temps, pressures, and non-corrosive fluids, some of today’s epoxy sealing materials might be used, or even low-temp soldering with silver-tin solders. I usually encourage people to evaluate the use of mechanical joining techniques (such as the rubber gaskets), or adhesive bonding, so as to keep the labor costs down and assembly fairly easy, if at all possible. Costs are always a business factor that should be evaluated before jumping into any kind of manufacturing or joining process.

Welding

A number of companies have incorporated either hand-welding or automated welding equipment (e.g., orbital welders) to weld (MIG, TIG, laser, etc.) the joint-circumference around each individual tube exiting a tube-sheet (header plate). An example of this is shown in Fig. 4.

Fig. 4 Example of automated welding of tube-through-header in order to seal each tubular joint. Welding is accomplished one tube at a time.

Fig. 4 Example of automated welding of tube-through-header in order to seal each tubular joint. Welding is accomplished one tube at a time.

Each tubular joint is made, one at a time, until all tubes have been sealed. As shown in Fig. 4, the seal is external and depending on how thick the tube-sheet or header plate is, much of the open space in the header between the tube/pipe and header (tube-sheet) remains open, allowing gas/fluid to seep into that area, which could cause internal corrosion over time. This is illustrated in Fig. 5, and is not being shown to discourage such joining, but merely to create awareness of an issue that must be considered and dealt with.

Fig. 5 With external welds, the actual seal is on the outside of the tube-through-header assembly, thus leaving open the gaps between the tubes and the header-plate (tube-sheet). This could theoretically allow corrosive material into those joint-gaps.

Fig. 5 With external welds, the actual seal is on the outside of the tube-through-header assembly, thus leaving open the gaps between the tubes and the header-plate (tube-sheet). This could theoretically allow corrosive material into those joint-gaps.

Brazed Joints

If the operating temperatures of your S&THE are high (above 700F/350C), and/or the unit is operating at high pressures, or if the fluids involved have any corrosive characteristics, then brazing is probably the best joining technique to use in order to ensure a strong and leak-tight joint between S&THE tubes and header plates (tube sheets), especially since the molten brazing filler metal (BFM) should fill the entire gap between the tubes and the header-plate (tube-sheet), preventing any entrapment of anything else between the tubes and the header assembly, such as that illustrated in Fig. 6.

Fig. 6 Brazing filler metal (BFM) will melt and then flow as a liquid down and around each of the tubular joints in the header-plate (tube-sheet), to completely fill those joints from one end to the other.

Fig. 6 Brazing filler metal (BFM) will melt and then flow as a liquid down and around each of the tubular joints in the header-plate (tube-sheet), to completely fill those joints from one end to the other.

One significant advantage of brazing is that all these brazed joints are made at the same time within the brazing furnace, along with perhaps many other S&THE’s being brazed in the same furnace load. This can amount to huge time savings vs. individually welding each tube to header joint.

But how, you might ask, is that BFM applied to the surface of the header-plate (tube-sheet) so that the BFM surrounds each tube completely, to ensure that the BFM will melt and flow into each joint and guarantee a complete seal of each tube-to header joint? Well, one way to do this would be to apply the BFM as a slurry onto the outside surface of the header AFTER the tubes have been pushed through each of the holes in the header. When this technique is used, I recommend that the tubes be made longer than needed, so that they will extend beyond the front surface of the header, as shown in Fig. 6. Once assembled, the slurry can be poured around each of the tubes to provide sufficient BFM to fill each of the joints. After brazing has been completed, the tubes can be cut by a variety of techniques so that they are made flush with the outer face of the header, if required.

Another highly recommended procedure (preferred) would be to put the BFM onto the inside surface of the header plate and allow it to be drawn to the outside surface of the header by capillary action during brazing. Three (3) ways that are often used for this are; silk-screen printing, BFM-powder spraying, or by applying a layer of brazing tape.

Screen-Printing

Screen-printing is a method whereby a closely controlled amount of paste can be applied to the metal surface through a fine-mesh screen. Both the size of the powder particles (powder mesh-size) and the size of the openings in the screen (the screen mesh-size) must be carefully selected and controlled to optimize any screen-printing application in brazing. Additionally, the paste-binder system must be formulated to properly allow the paste to go through the screen, and to dry and remain adherent to the header-plate during subsequent hole-drilling and other assembly operations so that the dried-paste does not flake off prior to actual brazing operations. Fig. 7 shows an example of a silk-screened nickel-based BFM applied to a header and then drilled to allow tubes to be pushed through.

Fig. 7 Header-plate coated with BFM-paste via a screen-printing technique. (Photo courtesy of Hoganas AB, Sweden).

Fig. 7 Header-plate coated with BFM-paste via a screen-printing technique. (Photo courtesy of Hoganas AB, Sweden).

BFM-Spray

In a similar manner, spray guns can be used to either spray braze-powder along with a sprayable binder system so that a uniform, thin coating of BFM powder will be deposited onto the header plate, such as that shown in Fig. 8

Fig. 8 Header-plate coated with BFM powder from a spray-gun. (Photo courtesy of Hoganas AB, Sweden)

Fig. 8 Header-plate coated with BFM powder from a spray-gun. (Photo courtesy of Hoganas AB, Sweden)

BFM-Tape

BFM-tape is a flexible sheet of BFM powder, bonded together with a binder that will disappear (volatilize) during brazing, and is illustrated in Fig. 8.

Fig. 9 BFM in tape form, with an adhesive backing for ease of applications, and strong surface-adherence.

Fig. 9 BFM in tape form, with an adhesive backing for ease of applications, and strong surface-adherence.

BFM-tape is available in rolls (many options re thickness and width), and often have an adhesive backing to the tape. When the protective paper is peeled away from the adhesive layer, the tape can be carefully placed on any metal surface, and then handled just as the screen-printed or sprayed coatings are handled.

How much BFM should I use? The amount of BFM to use for any given brazing application has been discussed at length in some of my other articles on this website.

Once the BFM has been applied by silk-screening, spray coating, or by BFM-tape, a number of tube-holes would be drilled through the BFM layer and then through the header plate, leaving a nice layer of BFM in place, tightly bonded to the inside surface of the header (as shown in Figs. 7 & 8). The tubes would then be carefully inserted through the holes in the BFM-coated header. By contrast with the slurry technique, this silk-screening/spraying/tape method puts the BFM on the inside surface of the header, rather than on the outside, and, when properly brazed, the BFM will melt, be pulled by capillary action into and through the tube-joints, and reveal itself as a tiny fillet (meniscus) on the outside of each tubular joint, clearly verifying that the brazing was done properly. This method also allows easy adherence to my rule of brazing:

RULE OF BRAZING — You should always apply BFM from one side of the joint and inspect on the other!

Thus, by applying the BFM on the inside surface of the header, it makes external inspection of each joint very easy.

Brazing Joint Design

Tubular joints for brazing must be close-fitting. Thus, the shape and outside-diameter (OD) of tubes/pipes used in S&THE’s must closely match the inside diameter (ID) of the holes through the header-plate/tube-sheet so that a leak-tight joint can be made. Additionally, you’ll need to decide if the joint should contain a shoulder in it, so that the tube/pipe bottoms-out on the shoulder (as shown in Fig. 10), or if the tubes will go all the way through the header (tube-sheet). From a joint-design perspective, the shouldered design would be preferred in order to ensure a strong, leak-tight brazed joint for each tube, but inspection of the BFM flow through the joint will be a bit more difficult to confirm.

Commercial tolerances? I do NOT recommend the use of commercial tolerances for hole-drilling or for sizing tubes/pipes since common commercial size tolerances often lead to larger-than-desired gap clearances for brazing, and thus can often result in poorly filled (or non-filled) joints, thus resulting in a lot of headaches in trying to repair such defects.

Fig. 10 Sizing the holes in the header/tube-sheet so that the tubes/pipe bottom-out on shoulders inside the header is preferred from a strength and from a leak-tightness (hermeticity) perspective.

Fig. 10 Sizing the holes in the header/tube-sheet so that the tubes/pipe bottom-out on shoulders inside the header is preferred from a strength and from a leak-tightness (hermeticity) perspective.

Diametrical clearances. I strongly recommend that diametrical clearances for brazed joints (especially if nickel-brazing is to be used) be no larger than 0.003” (0.75 mm). Yes, I said diametrical, not radial. Tubes/pipes are often not round and will tend to move to one side of the joint (tubes do NOT center themselves), resulting in metal-to-metal contact on one side of the joint, and a larger, open gap on the other side of that joint clearance. When nickel-brazing, the diametrical clearance should be no greater than 0.003” (0.075 mm), or else hard, non-ductile precipitates WILL form in the center of the widest part of the joint (discussed in detail in some of my other online articles), possibly resulting in joint-cracking (and leakers) under service conditions. TAKE THE TIME TO MAKE SURE TUBES/PIPES ARE ROUND AND SMOOTHLY FIT INTO CLOSE-FITTING ROUND HOLES IN THE HEADERS.

Tube expanders. These are commonly used, in a variety of forms and sizes, to help ensure closely matched fits of tubes/pipes into fittings and holes. Such tooling may be used, depending on the difficulty in closely matching tube/pipe roundness and size relative to hole-diameter. The important thing is that the capillary gap between the tube and hole must be very close and uniform for good brazing. Find the method that works for you to achieve this.

Furnace, Torch, or Induction Brazing?

Much of the choice between whether to use furnace, torch, or induction brazing will depend on the size of the S&THE. Where the S&THE’s are about 48” long, or shorter, then the use of brazing-furnaces can be very effectively implemented, especially since several S&THE’s can often be brazed at the same time in the same furnace load.

When S&THE’s are significantly longer, such as 10-feet (3 meters) or more, then the cost for a using a furnace large enough to hold the entire S&THE can be very high. In such a situation, you may want to consider using either torch or induction brazing to heat up only the header-plate (tube-sheet) where the actual braze needs to be made. This can save a huge amount of time and money for the needed brazing.

Torch brazing has two potential drawbacks to it: worker skill, and the need to use a flux when brazing out in the open air. As a torch-brazing trainer for many years, I’ve seen a large amount of skill-variation between torch-brazers, and only recommend that process when the operation can be closely monitored by skilled persons in the art and science of torch brazing. Plus, the flux that is used in the process WILL get trapped to some extent in the joints, and such trapped flux voids need to be kept to a minimum. Much of that depends on the torch brazer’s skill.

Rather than torch-brazing, I would suggest that the reader might want to consider induction heating for bringing the header/tube-sheet up to brazing temp. The use of a “pancake coil” design for the induction coil might be worth investigating. Such an induction heating method is shown in Fig. 11, and is a coil design used in a lot of brazing applications.

Fig. 11 A pancake-coil inductively heats a steel face quite rapidly, and when proper induction-heating variables are employed, can effectively heat fairly deep-down into structures for brazing.

Fig. 11 A pancake-coil inductively heats a steel face quite rapidly, and when proper induction-heating variables are employed, can effectively heat fairly deep-down into structures for brazing.

Please be aware, too, that an option in induction heating is to do it inside an inert-atmosphere box, which could be specially designed (with the help of the induction equipment manufacturer) to fit around the base of the S&THE so as to exclude any air, and thus to keep any of the surfaces being brazed from oxidizing while being heated up to brazing temp. This kind of atmosphere container and induction coil design must be developed carefully in conjunction with one of the many excellent induction-heating equipment manufacturers out there.

Problem solve and improve with our technical experts.