How to Prevent Common FSW Defects in Aluminum Welds

Most common FSW defects in aluminum welds follow a pattern. A process parameter drifts, tool engagement changes, or heat input drops below the point where material plasticizes and consolidates. The result is a tunnel, a void, flash, or lack of fill that shows up only after cross-sectioning or NDT. My position is simple: defect prevention belongs in machine capability and tool selection, not in inspection alone. When force control, tool geometry, and heat input are specified correctly before welding begins, most of these defects never form. This article explains the root causes and prevention measures I use when developing production FSW processes.

What Causes Tunnel and Void Defects in FSW?

A tunnel defect is a continuous or intermittent cavity that runs along the advancing side of the weld, usually just below the surface. It forms when material behind the tool cannot consolidate before cooling. Insufficient heat input is the usual driver. The spindle speed is too low for the travel speed, or the tool has too little plunge force to generate the required forging pressure. On 6xxx aluminum, I often see this when a process is speeded up without increasing spindle speed or changing the tool shoulder.

A void is more localized. It differs from a tunnel in shape and origin. Voids often come from folded oxide layers, insufficient shoulder contact, or a loss of forging pressure during a thickness change. The material may look consolidated on the surface while internal porosity remains. In cross-sections, a void appears as a rounded or irregular cavity near the weld interface rather than a long channel. The fix is not always more heat; it is often better shoulder engagement and cleaner faying surfaces.

How Do You Tell a Tunnel From a Void?

A tunnel is long and directional. It follows the advancing side and points to travel speed or heat input problems. A void is more compact and rounded, often associated with poor interface consolidation or contamination. If a defect changes length when travel speed changes, suspect a tunnel. If it appears only at certain locations along the joint, check part fit-up, oxide removal, and shoulder pressure first. Destructive sectioning at multiple locations is the fastest way to distinguish them before NDT confirmation.

Why Does Flash Form and When Does It Threaten Weld Quality?

Flash is the material pushed out beyond the shoulder edge during FSW. A small amount of flash is common with conventional tools, but heavy flash means excess material is being expelled. The root cause is almost always excessive plunge depth or excessive shoulder engagement. When the shoulder pushes too hard into the surface, the softened material has no space to stay in the joint, so it escapes sideways. Overheating makes the metal too soft and increases the effect.

Flash becomes a defect when it changes the joint geometry. It reduces the effective sheet thickness at the weld edges, creates stress concentrations, and can fail visual inspection before any mechanical test. On thin-wall parts such as battery tray side rails or cooling plates, even 0.2 mm of excessive flash matters because the wall thickness is already tight. A shoulder designed for thicker plate may produce flash on thin material.

How Much Flash Is Acceptable?

Acceptable flash is usually defined by the drawing or customer specification. For most structural aluminum parts, flash up to the thickness of the original sheet is considered too much. Anything that alters the edge radius or reduces the load-bearing section should be removed. I prefer to separate flash produced by excessive plunge from flash produced by thermal over-softening. The first is fixed mechanically, the second by heat input and travel speed.

Stationary shoulder tools solve most flash problems at the source because the shoulder does not rotate. It slides over the surface while only the pin rotates, which lowers heat input and keeps material contained. AEE offers a stationary shoulder FSW tool for aluminum from 1.0 mm upward, and the surface is typically clean enough to skip post-weld machining on cosmetic surfaces.

If your program involves thin-wall aluminum enclosures or cosmetic surfaces where flash removal adds rework, it is worth confirming shoulder design and plunge parameters before freezing the tool order. Send your section thickness and weld length to [email protected] and we will check whether a stationary shoulder or conventional tool fits.

How Do Tools and Parameters Prevent Lack of Fill?

Lack of fill is a surface or near-surface defect where the weld zone does not contain enough plasticized material to complete the joint. It appears as a depression, incomplete edge, or missing material at the end of the weld. The most common drivers are low heat input, insufficient shoulder contact, too short a pin, or travel speed beyond what the tool can feed material.

For long linear welds, the answer is usually parameter work. Raise spindle speed or lower travel speed, increase plunge depth within the tool’s shoulder capacity, and verify the pin length matches the material stack. For closed-contour welds, lack of fill is often the exit hole. No parameter change can eliminate that. The tool has to withdraw while the material is still plastic and the shoulder remains engaged. Retractable FSW tools are designed for this. The pin slowly withdraws into the shoulder at the weld end, and the surrounding material fills the cavity.

For closed-contour parts such as pressure vessels and battery housings, exit-hole lack of fill can pass through an entire batch if the tooling is wrong. <AEE provides customer with optional features FSW tool> covers how retractable pin and stationary shoulder options are selected for customer applications, and why the tool choice is made before process qualification rather than after first article inspection.

What Is the Exit Hole, and Is It a Lack of Fill?

The exit hole is the round void left when a conventional FSW tool simply retracts at the end of a closed weld. It is a form of lack of fill, but it has a different solution. The weld itself may be sound; the hole is an artifact of tool withdrawal. For parts with no run-off tab, the exit hole must be eliminated by a retractable pin or filled with a plug, which adds cost and inspection. I treat the exit hole as a tooling decision, not a parameter defect.

Which FSW Machine Features Reduce Defects in Production?

Parameters and tooling solve defects in development, but production machines have to hold those solutions. The two machine capabilities that matter most for defect prevention are force control in the welding axis and continuous process monitoring. If the machine only controls position, a thickness variation of 0.1 mm can change plunge depth enough to change flash or lack of fill. Force control holds axial load constant and lets the tool move with the part.

Torque and temperature data are the first indicators that a process is drifting toward a tunnel defect. On our production lines, spindle torque and Z-axis force are recorded with the weld number, so a shift in travel speed or tool wear shows up before the NDT station finds a discontinuity. The C-type FSW machine platform from AEE carries 12 kW spindle power, 112 Nm torque, and Z-axis force up to 30 kN on its 8 mm system, with positioning accuracy and repeatability of 0.03 mm. Those figures set the control envelope for a consistent weld.

Feature selection should follow the defect risk. If the weld path is closed, specify a machine with spindle retraction control. If the section is thin and cosmetic, specify a stationary shoulder tool. If the material is thick or high-strength, specify a spindle with enough torque at the planned rpm. The machine choice can remove entire defect classes before the first weld is made.

For aluminum-to-copper or aluminum-to-steel joints, heat balance and material flow are more sensitive than in single-alloy FSW. <Customized Dissimilar Material Welding Manufacturers in China> covers how tool geometry and machine parameters are adapted for dissimilar material combinations, including the trial work needed to keep defect rates down in production.

Why Does Force Control Matter More Than Speed?

Speed determines production rate, but force determines whether the material stays consolidated. A travel speed increase with no force reserve can produce a tunnel that looks good on the surface. I would rather hold force control and process monitoring in place and then push speed in small increments with cross-section verification. Once a parameter is qualified, force control protects it against thickness and fixturing variation.

What Should You Do Before Releasing an FSW Weld to Production?

Releasing a weld with tunnel, void, flash, or lack of fill is expensive. It shows up at the worst time: after tooling is installed, after first article approval, and sometimes after the customer has run NDT on production parts. The common reaction is to add more inspection. Inspection catches defects; it does not remove the cause.

The better path is a short pre-production loop. Weld a small sample set, section it, then confirm spindle speed, travel speed, plunge depth, force control, and tool geometry. If the machine has process monitoring, review the force and torque traces before accepting the parameter card. When a defect appears, change one variable at a time and re-section. This resolves cause before the part reaches full rate.

If your team is qualifying aluminum FSW for battery trays, cooling plates, rail profiles, or other structural parts, I recommend specifying machine and tooling options that match the defect risk instead of fixing the process later. Share your part drawings, cross-section thickness, and planned weld length with AEE at [email protected] or call +86 18325808715. We will confirm which spindle, tool type, and control configuration fit your application before you commit.

What Else Do Buyers Ask About FSW Weld Defects?

Can NDT find tunnel defects before mechanical testing?

Yes. Phased-array ultrasonic testing and X-ray can detect tunnel defects below the surface, but they are only as effective as the calibration and scan plan. A tunnel running along the advancing side can be missed by a single normal-beam UT pass if the scan direction is wrong. I prefer to combine linear scanning with a few destructive sections during qualification. Once the NDT signature is correlated to a known defect, it becomes a reliable production screen.

Is tunnel defect caused directly by tool wear?

A common assumption is that tool wear directly produces voids. It does not create them in isolation; it changes the effective pin diameter and surface heat generation. A worn pin can reduce material flow and shift the process toward a tunnel defect. The real signal is a slow torque drop or a change in weld surface texture. Checking pin wear at set intervals is more useful than waiting for a void to appear.

Is flash always a sign of too much heat?

It depends on the shoulder and plunge depth. Flash can appear with normal heat input if the shoulder is plunged too deeply or the tool is tilted excessively. It can also appear in hot, slow welds where material is over-softened. I separate the two by checking the surface: severe edge expulsion with a flat surface suggests mechanical over-engagement, while heavy flash with a very smooth and glossy surface suggests excessive heat.

Which FSW tools work best for closed-profile parts?

In closed-profile programs we have run, retractable pin tools and bobbin tools cover most risk. A retractable pin eliminates the exit hole on closed paths. A bobbin tool welds hollow sections without a back support and removes root lack-of-penetration. The right choice depends on the cross-section access and load path. Share your closed-profile geometry and wall thickness at [email protected] and we will confirm which machine and tool combination fits your application.

If you’re interested, check out these related articles:

Refill Friction Stir Spot Welding (RFSSW)
AEE attend 13th International Symposium on FSW in Kyoto, Japan on 21 – 23 May 2024