export@ezsteelpipe.com
+86 731 8870 6116
In a shell-and-tube heat exchanger, the tube-to-tubesheet joint does the real work. It has to hold pressure on one side, transfer heat across the wall, and survive thousands of thermal cycles without leaking. The way that joint is made is by expanding and rolling the tube into the tubesheet hole — a cold-working operation that sounds simple but actually controls whether the exchanger runs for thirty years or fails in three. This guide walks through how the process works, what tools are used, and what separates a sound joint from a leaking one.
Tube expansion is a cold-working process. The tube is pushed into the tubesheet hole with a small clearance, and a tool is then used to push the tube wall outward until it plastically deforms and presses hard against the hole wall. The plastic flow creates a friction-based mechanical joint that is also leak-tight.
"Rolling" is the most common way to drive that plastic deformation. A tube expander — a tapered mandrel inside a cage of three to five hardened steel rolls — is pushed into the tube end. As the mandrel advances, it forces the rolls outward against the tube's inner wall, which in turn presses the tube's outer wall into the tubesheet hole. The result is a permanent, gas-tight interference fit.
The two standards every rolled joint has to meet are leak tightness and pull-out strength. Leak tightness is achieved when the residual contact stress between the tube and the tubesheet exceeds the internal pressure of the service. Pull-out strength is the axial load needed to drag the tube back out of the hole, and it depends on how much the tube was expanded, the surface finish in the hole, and the tube material's work-hardening behaviour.
A tube expander has three working parts: a tapered mandrel, a roll cage, and three, four, or five rolls. The mandrel is driven by a torque-controlled motor (electric, pneumatic, or hydraulic). As it rotates and feeds forward, the taper pushes the rolls outward. The rolls ride against the tube's inner wall, applying radial pressure that exceeds the tube's yield strength so the wall flows.
There are two main geometries, and the choice matters:
In practice, the choice comes down to joint design. For most heat efficiency tubes in shell-and-tube exchangers, parallel rolling is the default because it gives predictable wall thinning and clean leak-test results.
A proper rolling job follows a controlled sequence. Skipping steps or rushing the cycle is the fastest way to produce a joint that either leaks at the pressure test or cracks the tubesheet ligament during service.
A rolled joint that follows this sequence typically achieves 4–10% wall thinning in the expanded zone, well within TEMA and ASME guidelines. Hydraulic expansion produces a more uniform 2–5% wall reduction and is preferred for thick tubesheets, thin-wall tubes, and sour-service applications under NACE MR0175.
The two methods produce similar joints but use very different tools.
Mechanical roller expansion uses the tapered-mandrel expander described above. It is fast, economical, and the standard choice for most shell-and-tube units. It works well for tubesheet thicknesses up to about 50 mm in routine service. The trade-off is that wall thinning varies along the joint, and residual stresses are higher than in a hydraulically expanded joint.
Hydraulic expansion uses a sealed elastomeric or metal-to-metal mandrel inflated with high-pressure water, typically 5,000 to 100,000 psi (35 to 700 MPa). The pressure pushes the tube wall outward uniformly along the entire tubesheet thickness. The result is lower residual stress, more uniform contact pressure, and less work hardening — useful for stainless steel, titanium, and nickel alloy heat exchanger tube materials in cyclic service.
For most projects, the choice is governed by the design code, the tube material, the tubesheet thickness, and the service fluid. Many exchangers use a combination: a mechanical roll for the bulk of the joint, then a seal weld at the tubesheet face to handle thermal cycling at the tube mouth.
Tube-to-tubesheet joints are only as good as the materials on either side of the hole. For a heat exchanger running in refining, petrochemical, power, or marine service, the most common combinations are:
A frequent mistake is mixing tube and tubesheet materials with very different thermal expansion coefficients. Under thermal cycling, the tube can work loose from the hole. The fix is to choose material pairs with similar expansion rates, or to use a seal weld at the tube face to take the thermal load.
Most rolled-joint problems trace back to one of a handful of process errors. Knowing them up front is the cheapest way to avoid rework in the field.
The rolled joint is not finished until it has been measured and tested. A typical quality control sequence includes:
Not all tubes roll the same way. A few practical rules of thumb based on the materials used in real heat exchangers:
These differences are exactly why the same rolling parameters cannot be used across the whole exchanger. A good fabricator sets the rolling schedule per material and verifies the first few joints before starting production rolling.
The rolled joint is not just a workshop detail — it is covered by major design and manufacturing standards. The most commonly referenced documents are:
A rolled joint that meets these standards is accepted worldwide for refinery, petrochemical, power, and marine service.
A few habits separate a reliable rolled joint from one that fails in service:
Tube expanding and rolling is the cold-working process that turns a loose-fitting tube into a leak-tight, high-strength joint with the tubesheet. It works by forcing a tapered mandrel through a cage of hardened rolls inside the tube, plastically deforming the tube wall until it presses hard against the tubesheet hole. Done right, the joint is gas-tight, mechanically strong, and durable enough to survive decades of thermal cycling. Done wrong, it leaks at the pressure test or fails in service.
The keys to a good rolled joint are simple: choose the right expander for the tube size, material, and tubesheet thickness; control expansion to the value allowed by the design code; stop on target rather than on feel; measure the joint and leak-test it before sign-off; and document everything. Whether the joint is rolled mechanically or expanded hydraulically, those rules do not change. They are the difference between a heat exchanger that runs for thirty years and one that fails in three.
Related Products