Tack welding: why it determines weld quality
Tack welds fix the geometry of the part before welding. Their position, size and order decide the final distortion. Explanations and practical rules.
A lot is said about the weld bead, rarely about the small welds laid before it. Yet tack welding decides much of the final quality: the part’s geometry, squareness and distortion are often settled before the first weld run.
What is tack welding?
Tack welding means laying short welds to join the components of an assembly temporarily. It plays three roles:
- Fixing the relative position of the parts, once they have been placed and checked.
- Holding the gap between the parts during welding, despite expansion.
- Making the assembly rigid so it can be handled, turned over or released from its jig.
A tack weld is generally a few millimetres to a few centimetres long. It must be strong enough to resist shrinkage forces during welding, without becoming a defect in the final weld.
Why it governs distortion
When the arc heats the metal, it expands; as it cools, the weld shrinks and pulls on the parts. This is shrinkage, the cause of distortion: an angle closing up, a plate buckling, a frame going out of square.
Tack welds resist this shrinkage. If they are badly distributed, too weak or laid in the wrong order, the part moves:
- too few tacks: the parts open up or close in during welding, and the gap varies along the joint;
- tacks on one side only: shrinkage pulls the part to that side and the angle closes;
- wrong order: each tack pulls on the part, and a poorly chosen order accumulates errors instead of cancelling them out.
Practical tack welding rules
Distribute symmetrically
Place tacks on either side of the part’s axis, and on both sides of a joint when possible. The shrinkage forces then balance out.
Start in the middle, or at the ends, depending on the part
On a long joint, a common rule is to tack the middle first, then work towards the ends, alternating: the gap is controlled along the whole length. On a frame, tack the corners, checking squareness at each step.
Space them evenly
The spacing depends on the thickness and stiffness of the parts: the thinner the plate, the closer the tacks must be to prevent it buckling between them.
Treat the tack like a weld
A tack weld will be absorbed into the final weld, so it must be of the same quality, with no porosity or lack of fusion. Welding quality standards in fact require tacking to be carried out with a suitable procedure, by a qualified welder or operator. A tack that is too large or poorly fused creates excess thickness or a fusion defect in the weld.
Check before welding
Once the part is tacked, check the dimensions and squareness: it is the last moment when a correction is easy.
Manual tacking, robotic tacking
In most workshops equipped with a robot, tacking remains manual: the welder places and tacks the components, then hands the part to the robot for the welds. That is one of the reasons why a welding robot struggles to pay off on small batches.
Automating tacking means solving two problems: holding the components in position without a jig (see welding without jigs), and choosing where and in what order to place the tacks so the part stays held and distorts as little as possible. It is one of the key features of our robotic cell, which tacks the assembly itself before welding it.
What the designer can do
The design office also influences tack quality:
- tabs and slots that position the parts make accurate tacking easier (designing for laser cutting);
- welds accessible from both sides allow symmetrical tacking;
- throats sized to what is needed limit the shrinkage the tacks have to hold back.
Key takeaways
Tack welding is not an unimportant preparatory step: it freezes the geometry of the part and sets up how the metal behaves during welding. Symmetry, order, spacing and tack quality make the difference between a part that comes out square and a part that needs straightening.