Back to Basics: Inspecting Welds in Galvanized Steel 2005-10-023
 ITRENDS - BACK TO BASICS
Inspecting Welds in Galvanized Steel

Inspectors need tounderstand how zinc
galvanization affects welds and the unique
magnetic particle testing indications that
can be produced during inspections

Fig. 1 — A part freshly hot-dipped in zinc.

BY
GORDON E. SMITH AND RICHARD M. BELDYK
 
The following is the final article in the “Back to Basics” series. Here, Gordon Smith and Richard Beldyk tell you what you need to know about magnetic particle inspection of galvanized materials. Smith is an AWS CWI and an ACCP NDT Level II and III – UT, and a Level III in MT, PT, and RT. Beldyk is a Registered Engineer in ten states, an AWS Certified Welding Engineer, an Association of Facilities Engineer, a Certified Plant Engineer, and a NASP Certified Accident Investigation Technician.

Zinc galvanization, per square foot per year, is the most cost-effective means of corrosion protection in the world today, yet the process of galvanizing steel has been a little understood technology for more than a hundred years. Most persons, including welding inspectors, think of galvanization as a covering layer that sometimes rusts and sometimes peels off. Both of these conditions are thought to be the fault of the galvanizer. That is probably the most incorrect thought about the metal protection industry today.

What is Galvanization?
Zinc galvanization is a process where the electromotive metal is applied by dipping a steel item (Fig. 1) into molten zinc, which then forms a durable bond to the iron at the atomic level. The formation of this zinc-iron bond leaves a coating that has penetrated the iron in the steel and has iron trapped in it to varying degrees, with a protective top layer of almost pure zinc. The exterior surface of a galvanized item is just one part of many layers (Fig. 2) of different zinc-iron alloys that make up a galvanized, corrosion-resistant metal protection system.
   
Fig. 2 — The various zinc layers in a galvanized part.

The great tenacity of zinc galvanization makes it detrimental to welding, often resulting in poor weld strength and producing unique magnetic particle testing (MT) and visual indications that the inspector needs to understand. Additionally, the welding process can change the galvanization characteristics. These changes range from destruction of the protective effect of galvanization to the increased buildup of zinc until it can exceed 500% of that specified for corrosion protection of the item. Galvanization of structural bolts has recently come under scrutiny with the reminder that fully tensioned ASTM structural fasteners should always be qualified after galvanization.

Appearance of Galvanized Steel
When properly designed, applied, and installed, a zinc galvanized surface should protect the underlying steel for many years. Initially, it will have a shiny appearance, which fades to a soft gray color in ten or more years.
   
Fig. 3 — White rust spots on a newly installed dual galvanized panel.

Sometimes, however, it can have a different appearance, such as what looks like white rust — Fig. 3. These white rust spots are a mixture of zinc oxides, hydroxides, and carbonates formed from the available reactive materials: oxygen, water, carbon dioxide, etc. This layer actually forms a barrier to further deterioration of the protective zinc galvanization. It has a gem form “Smithsonite” that was named after James Smithson (1754–-1829), a British chemist and mineralogist, and founder of the Smithsonian Institute in Washington, D.C. — Fig. 4.
   
Fig. 4 — Smithsonite geode.

Other times, new welded items may rust along every weld soon after installation — Fig. 5. This happens because most arc welding filler metals have a lot of silicon in them. Silicon can  transport and trap iron in it as the weld cools to room temperature. However, we now know that steel is only semiquenched at room temperature and for some alloys quenching to –300° or even –450°F will not freeze the metal atoms trapped in these alloys. This is true of the zinc-iron alloy that makes up galvanization. There is a lot of iron relatively free to move about. Under one condition, annealing, the iron will return to the substrate iron and leave relatively pure zinc, which quickly disbonds and separates from the iron substrate. That is one reason to remove the zinc to a specified distance away from the weld zone prior to welding. If the zinc is heated close to the 800°F range, it may remelt locally and, upon cool-down, the iron may return to the steel underneath leaving a poorly bonded layer of zinc.

Magnetic Particle Inspection of Galvanized Steel
Welds completed prior to hot dipping may require special welding materials with reduced silicon levels. If these special welding materials are not used, then excessive zinc may be deposited upon the weld surfaces. When a galvanized layer of 0.004-in. thickness is achieved on the item surfaces, there often may be 0.012 or even 0.025 in. of zinc deposited upon the weld. This can lead to magnetic particle indications that may be distantly related to other more serious problems — Fig. 6.
   
Fig. 5 — A — Rusted new weld under zinc; B — close-up of weld joint corrosion.

Long, parallel magnetic particle indications may often appear on “stock” welded or “weld-clad” areas that have been ground smooth. These are thought to be due to concurrent field leakages between individual weld bead grain orientations. They typically have a “fuzzy” appearance on top of the galvanization. Different dissolution rates for different iron concentrations in the weld beads cause them. With galvanization of less than 0.006 in. thick and wide, thick welds, after grinding the galvanizing patterns may appear to look like dendrites or have numerous small branches at 90-deg intervals to the direction of the weld beads.
   
Fig. 6 — MT indication across a bridge flange.

For other welds, especially where different thicknesses are joined with wide welds, a fuzzy indication may appear oriented along the weld midline. This is most prevalent with AC dry yoke MT techniques. It has sometimes been found that the weld had later developed a longitudinal weld crack at this inspection location. On occasion, when using dry AC MT, a magnetic particle indication would increase in strength of appearance or definition and then fade away as the zinc was carefully removed. In this case, excess silicon was thought to aid the dissolution of iron from the weld metal during galvanization, resulting in a concentration of iron trapped in the upper surface of the galvanization.
   
These last conditions should be noted with care as sometimes they may be linked to liquid metal embrittlement (LME) resulting in very large and sometimes sudden fractures. There is some thought that these fractures may be due to a thermal or time-related change in grain structure, i.e., austenite to ferrite, martensite, etc., wherein a volume change in the weld takes place and increases stress levels dramatically to the point of fracture.
   
The inspector should always obtain a copy
of the galvanization bath chemical analysis and look at the tin, bismuth, and lead levels.

The presence of low-melting-point elements such as tin and bismuth in the galvanizing melt have a role in causing LME. However, the relative significance of melt composition in the galvanizing bath on the potential for steel cracking is not fully understood.
   
Fig. 7 — Newly welded deck metal.

The inspector should always obtain a copy of the galvanization bath chemical analysis and look at the tin, bismuth, and lead levels. Scanning electron microscopy and X-ray dispersive spectrometry analysis of the crack surfaces have indicated high concentrations of these elements in the cracks. If the % tin (wt) +  % lead (wt) is greater than 1.3% (by weight) or the % bismuth (wt) is greater than 0.1% (by weight), a closer examination for cracking is warranted, preferably with dry AC yoke MT. Wide, fuzzy MT indications running down the length of welds, sometimes for many feet, are thought to be due to the presence of dissolved iron suspended in the thicker galvanization buildup on welds — Fig. 6.

Other Causes of Rust
Nonwelded, newly galvanized items include all sorts of cold- and hot-formed products, HVAC supports, steel decking, etc., where the distribution of iron may change in the first few months due to temperature or time or electrical influences resulting in iron migrating to the surface of these items. This  migration generates a rusty orange appearance as the iron reacts with surface zinc oxides, hydroxides, etc., producing an unsightly appearance with the ability to produce rust stains on items below. This is known to happen to large areas of galvanized decking after welding to steel joists. On occasion, it has been found to stop advancing in size after welding operations were completed — Figs. 5, 7.

Fig. 8 — Schematic of iron distribution changes.

Storage practices are thought to influence the tendency toward top surface corrosion of galvanized steel deck metal. Referring back to the cross section of the zinc galvanized layer (Fig. 2) and remembering that the surface is almost pure zinc right after galvanization, we see that if two freshly galvanized surfaces are brought together, the iron trapped in each galvanized layer has a tendency to move to the centerline of the two layers when in contact or connected by an electrolyte such as rainwater — Fig. 8. Upon separating the two pieces of deck metal, the iron at the centerline is exposed to a more corrosive environment such as is found at most construction sites. This surface iron is quickly chemically trapped by reacting with the zinc oxides, etc., on the surface of the galvanization, where it quickly takes on rusty colors.
   
The inspector should be wary that magnetic coating thickness gauges may produce erroneous and typically thinner readings than what is in the actual galvanized layer. A through-cutting destructive coating gauge is a good backup if you have questions on the accuracy of the coating measurement. The inspector should always calibrate any measuring equipment used on a regular basis, and be knowledgeable of the variables surrounding these measurements. For many of these conditions, rust-bonding chemically reactive paints are a good solution for this problem. These paints can be expected to last 5–10 years. The galvanization industry has been recently promoting paint after galvanization as a solution for this problem. In the authors’ opinion, after more than ten years of observation, zinc-rich spray-on galvanization is a good solution for regalvanizing weld repaired areas.


GORDON E. SMITH (gsmith@hcnutting.com) is Senior Consultant, H. C. Nutting Co., Columbus, Ohio.

RICHARD M. BELDYK, PhD (welding_engr_rich@charter.net), is an Engineer with Ohio Bridge Corp., U.S. Bridge Division, Cambridge, Ohio.