Seam defects in arc brazing: pores, spatter, lack of bonding and braze cracking

By Robert Lahnsteiner, International Welding Engineer (IWE) · Updated: November 2026

Which defects occur in arc brazing?

Above all pores, spatter, zinc burn-off beside the seam and shape defects such as a seam that is too high. Also typical of brazing is braze cracking (liquid metal embrittlement): filler metal penetrates the steel along the grain boundaries.

DVS 0938-3 describes the imperfections of arc-brazed seams, based on the catalogues for brazing (ISO 18279) and gas-shielded welding (ISO 5817), and assigns measures to each defect. Defects caused by flux or slag do not occur, because brazing is done without flux and usually in a single pass.

Defect	common cause	first remedy
Pores	zinc vapour, contamination, disturbed gas shielding	less heat, escape space, check gas supply
Spatter	zinc vapour, unsuitable arc type, thick zinc layer	pulsed arc instead of short arc, start programme, clean gas nozzle
Zinc burn-off, white flakes	too much heat input per unit length	reduce energy, conduct heat away via clamping jaws
Lack of bonding and incomplete filling	torch position, gap, arc blow	correct torch, check gap, move earth connection
Excess seam height	too much filler metal, voltage too low	increase voltage, adjust amount of filler metal
Erosion of the base material, burn-through	too much heat	reduce energy, correct torch position, inert gas
End crater cracks and hot cracks	shrinkage, clamping, long seams	crater-fill programme, free shrinkage, shorter seams
Filler metal penetration, braze crack	stress in the sheet plus heat	reduce energy, move seam out of formed areas
Tungsten inclusion	scratch start, electrode in the pool	HF ignition, keep distance

Summarised from DVS 0938-3:2012, Table 1. A defect can have several causes, and several measures often help.

Pores: causes and remedies

The most common cause is zinc vapour. It forms under the arc and must escape before the filler metal solidifies. The more heat and the tighter the space, the more pores.

Heat: short arc, slightly pushing torch, no more energy than necessary. Give the pool time to degas. Settings: MIG brazing.
Design: Provide escape space for the zinc vapour. A gap helps, but above 0.5 mm the number of pores rises markedly again; see Joint types and seam geometry.
Cleanliness: Remove oil, wax and adhesive from the sheet. The wire can also cause pores: follow the manufacturer’s storage instructions and avoid condensation.
Gas: Check hose, connections, cooling water, draughts and gas flow; see Shielding gas.

How bad are pores? In tests on the fillet seam on a lap joint, porosity played a minor role for strength as long as the seam was continuous. Only pores in the bonding area between filler metal and sheet were critical (Kusch and Podlesak, DVS-Berichte 275, 2011). Open surface pores are nevertheless not permitted in visible areas and areas at risk of corrosion.

Spatter and zinc burn-off

Both come from too much energy or an unsteady arc on the zinc layer. White zinc oxide flakes beside the seam are a clear sign that the power is too high.

Arc: pulsed arc instead of short arc where possible. TIG and plasma brazing are considered spatter-free. A start programme helps against spatter on ignition.
Zinc layer: The type of galvanising affects spatter and how well it adheres. DVS 0938-3 states, for hot-dip galvanised sheet, preferably no more than 15 µm.
Filler metal: no filler metals with readily evaporating constituents such as zinc.
Zinc burn-off: Reduce the heat input per unit length and conduct heat away from the vulnerable area via clamping jaws or backing plates.

Brazing spatter adheres flat to the galvanised sheet, unlike the spherical spatter from welding. It hardly comes off with hammer and chisel; on visible surfaces it has to be ground. It does not, however, cause rust spots. How the zinc-free margin beside the seam corrodes is explained in Base materials and coatings.

Lack of bonding, incomplete filling and seams that are too high

Lack of bonding usually comes from torch angle, gap or arc blow. A seam that is too high with a steep transition, on the other hand, is often a matter of voltage and amount of filler metal.

Lack of bonding: Correct torch angle and position, check the gap. If arc blow deflects the arc, move the earth connection to a more favourable point. Avoid filler metal running ahead, for example in positional brazing.
Incomplete filling: By the nature of the process, arc brazing does not fill a gap deeply. If the gap is too narrow, the vapour pressure can deform the seam; if it is too wide, the filler metal is not enough.
Excess seam height: Increase the brazing voltage, adjust the amount of filler metal and the clamping. Active gas components lower the surface tension of the pool and let the filler metal flow out better.

DVS 0938-3 points out that a certain excess seam height and a non-tangential transition are typical of modern energy-controlled brazing processes. Whether this is acceptable must be clarified case by case; for components under cyclic loading, if necessary with fatigue tests.

End crater cracks and hot cracks

They form during cooling when filler metal and sheet cannot shrink freely. Long seams, rigidly clamped components and the end of the seam are at risk.

End of seam: Use a crater-fill programme; otherwise reduce the energy by hand at the end and add a little filler metal.
Seam length: Avoid seams that are too long; DVS 0938-3 gives 70 mm for deep-drawing steels as an example. Shorten seams and optimise the seam sequence.
Clamping: Clamping parallel to the seam can trigger hot cracks. Allow free shrinkage and conduct heat away, for example with copper bars.
Design: sufficient distance from the sheet edge, larger overlap, no brazing in heavily cold-formed areas.

What is braze cracking?

Liquid copper filler metal penetrates the steel along the grain boundaries. This happens mainly where the sheet is under tensile stress, and the more heat is introduced, the more it happens. The small penetrations are often only visible under the microscope; larger ones turn into cracks.

DVS 0938-3 lists filler metal penetration as foreign metal or liquid metal penetration, which only becomes visible from 50× magnification. SLV München investigated it systematically in an AiF project and classified it by depth: small up to 70 µm, medium 70 to 300 µm, large over 300 µm (Aicher and Cramer, DVS-Berichte 244, 2007).

What influences braze cracking

Filler metal: markedly more grain boundary damage with the tin bronze CuSn10 than with CuAl8, and more with CuSi2Mn1 than with CuSi3.
Base material: up to 35 % more damage on high-strength dual-phase and complex-phase steels. Among the soft grades, DX54D was more susceptible than H340LAD.
Clamping: Rigid clamping increases residual stress and roughly doubled the number of filler metal penetrations on high-strength sheet.
Cold forming: DX54D with 20 % pre-strain showed penetrations and small cracks, with 10 % it did not. On flange joints and formed joints, the damage lay in the zone of bending tensile stress.
Process: Short arc, DC pulse and pulse with an AC component were comparable with each other. TIG brazing caused less grain boundary damage.
Galvanising: Whether hot-dip or electrolytically galvanised made hardly any difference.
Annealing: Stress-relief annealing at 700 °C for 30 minutes counteracted the damage.

The remedy according to DVS 0938-3 is short: reduce the energy input and move the seam out of heavily cold-formed areas. With optimised working ranges, both MIG and TIG brazing can produce high-quality seams. On choosing the filler metal: Filler metals and standard designations.

Never weld over a brazed seam, not even after grinding it off. According to DVS 0938-2, there is otherwise a risk that the base material becomes brittle and cracks.

How are brazed seams assessed?

There is no dedicated standardised assessment system for arc-brazed seams. DVS 0938-1 recommends deciding case by case or following ISO 18279 and ISO 5817.

The reason is the dual nature of the seam: in design terms it is built like a weld, metallurgically it is a brazed joint. In the Chemnitz tests, imperfections had little influence on strength as long as the seam was continuous. What mattered was the amount of filler metal and the size of the bonding area.

Test methods and limits: Testing, rework, health and safety.

Frequently asked questions

Why does my brazed seam have so many pores?

Usually because of zinc vapour: too much heat, dragging torch, arc too long or a gap that is too narrow with no escape space. Then check the cleanliness of sheet and wire and the gas supply.

What does braze cracking mean?

The penetration of liquid filler metal into the steel along the grain boundaries, mainly in areas under tensile stress. This can lead to cracks in the base material beside the seam.

Which filler metal is least prone to braze cracks?

In the SLV München tests, CuAl8 showed markedly fewer filler metal penetrations than the tin bronze CuSn10, and CuSi3 fewer than CuSi2Mn1.

How can you tell there is too much heat?

By white zinc oxide flakes beside the seam, wide zinc burn-off, a melted upper sheet and pores.

Can a brazed seam with pores be re-brazed?

Yes, with filler metal of the same type. Welding over it, on the other hand, is never permitted. More on rework in Testing, rework, health and safety.