Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

How can friction stir welds be qualified?


Frequently Asked Questions

The rapid and successful industrialisation of FSW across a range of fields has generated significant interest in standardisation of the process, particularly for safety critical applications.

The current status of standardisation activities in the field of FSW is as follows:

1. AWS Standard
AWS - D17.3/D17.3M:2016 “Specification for friction stir welding of aluminum alloys for aerospace applications”. As the title suggests this document is targeted towards aerospace applications, however the guidelines presented have general relevance for all FSW users.

2. ISO Standard

ISO 25239:2020 “Friction stir welding — Aluminium” is a five part standard.

Part 1: Vocabulary
Part 2: Design of weld joints
Part 3: Qualification of welding operators
Part 4: Specification and qualification of welding procedures
Part 5: Quality and inspection requirements

There are no ISO standards governing FSW of other materials other than aluminium alloys.

3. Registration of FSW

Prior to the publication of FSW standards, various certification agencies (including Lloyds Register, DNV, Germanischer Lloyd) have issued insurance approvals for FSW users for the production of specific components. These fabricators have been certified against standard welding code requirements by these bodies, on a case by case basis. This approach typically involves the development and verification of welding procedures, followed by qualification testing, as described below:

Visual appearance

No obvious flaws on surface, root, or in exit hole. Good quality surface finish, minimal flash.

Bend tests

Root and face bends should be made using guided rollers, as hardness may vary considerably across the weld. Three-point bend tests can concentrate the strain into weaker areas, giving results which can be difficult to interpret. Bend radius should be no more severe than used on parent material of the same specification, temper and thickness.

For some alloys with very low ductility (e.g. some Mg die casting alloys) the value of bend tests is questionable. On ductile materials, 180° bends should be obtainable. The aim of the test is to expose flaws, not to demonstrate material ductility. Flaws in the root of the weld will normally show up well on a root in tension bend test. On thicker materials, side bend tests should be used.

Tensile tests

Cross-weld tensiles on full thickness samples should be used to demonstrate weld strength. The test report should note, as a minimum, tensile strength, elongation and failure location. Reduction in area is also very useful, and gives a more accurate assessment of ductility. Low elongation values may be obtained where a locally softened region exists, and is not necessarily deleterious. For alloys which can naturally age after welding (e.g. 2xxx and 7xxx aluminium alloys), the time of welding and testing should be noted.

Macro sections

These should be used to check for buried flaws, and to gain an overall assessment of weld quality. There should be no obvious concentration of oxides along the original joint line, and no evidence of flaws, particularly in the root, or close to the surface of the weld where the material will have been extensively sheared. Small volumetric flaws can occur at any depth.

Hardness tests

These provide a useful insight into variations in properties across a weld. Care should be taken to avoid mounting samples for hardness testing in bakelite, as the curing cycle (typically 10 minutes at 180°C) will affect the hardness of some age-hardening materials. For alloys which can naturally age after welding (e.g. 2xxx and 7xxx aluminium alloys), the time of welding and testing should be noted.

Other tests

Where specific properties are required, e.g. corrosion performance, fatigue performance, formability, etc, additional tests are required. However, these should not be attempted on welds which do not satisfy the criteria established for the earlier tests.

Quality Control

Although not primarily a tool for procedure qualification, methods of NDE and/or in-service monitoring of parameters need validation as part of the qualification process.


The first approved standard for FSW was published by AWS in 2010, and this was shortly followed by ISO. Both these standards have now been revised and are widely used in specifying the required weld quality. However where applications have additional requirements these are developed by the industrial users and approved by the relevant certification bodies where appropriate.

TWI has been actively involved in the development of welding procedure specifications for FSW, and has assisted many companies in the adoption of the technology. In addition TWI has worked with industrial users and insurance agencies to review and audit quality assurance procedures for a range of FSW applications. For further information on welding procedure specifications and quality assurance in FSW please contact us.

For further information also please see Joining Technologies.  

For more information please email: