Corrosion, Failure Analysis and Materials Selection Specialists






Corrosion Testing / Failure Analysis




Corrosion Testing

Failure Analysis

Field Investigations



Technical Papers

CTL Profile

Pricing & Policies

Contact CTL

Quality Assurance

Return to Failure Analysis Case Histories

Stress Corrosion Cracking of Unused Copper Tube from Air Conditioner Coil




Air Conditioner Unit


SERVICE TIME: Never Used  


Stress Corrosion Cracking



The received section was approximately 5-1/2-inches long by 2-3/4-inches wide by 4-inches in height (Figure 1).  The section consisted of U-shaped copper coils, presumably fabricated of UNS C12200 DHP copper, that passed through a galvanized steel tube sheet and numerous thin corrugated aluminum radiator fins.  The copper tubing had an outer diameter (OD) of approximately 0.386-inches, a wall thickness of approximately 0.011-inches, and possessed internal rifling that ran parallel to the longitudinal axis of the tubing. 

According to information provided by the manufacture, this coil, from which the supplied section was taken, had been shipped to the Middle East.  The coil had never been used but was found to have leaks in the copper tubes at the tube sheet.  The coil was returned to the US, tested in a water dip tank, and the leak locations marked.  One of the separate copper tube hairpins was similarly marked with an arrow indicating a leak at the same location as the tubes on the larger section.  




The coil section and two hairpins were examined in the as-received condition at the indicated leak sites with an optical microscope at up to 40X magnification.  As previously mentioned, the marked leak sites were on the copper tube as it emerged from the tube sheet, at the start of the U-bend.  Black and white deposits were noted on all the tubes in the general area of the leak sites, but no obvious leak sites were observed. 

Figure 1. As-received copper coil section.  Arrows indicate leak sites found. 


One of the copper U-bends marked as possessing two leak sites was carefully removed from the coil section.  One end of the U-bend was crimped and soldered shut, and a valve was clamped to the other end.  The U-bend was leak tested by pressurizing it to 10 psig with air and holding it under water.  Two very small leak sites, revealed by air bubble streams from the tube surface, were observed approximately 180° apart around the tube outer diameter (OD) surface on one leg of the U-bend, one site being in one of the indicated areas.  No leaks were observed at the other marked leak location. 

The pressurized tube was again examined at up to 40X magnification with an optical microscope.  Residual water on the tube surface continued to bubble during the examination, which aided in pinpointing the leak sites.  A very faint crack in the copper tube was observed at one leak site.  The other leak site, 180° around the tube, appeared to be associated with the black deposit on the copper tubing mentioned earlier, but no obvious hole or crack was observed. 

An approximately 1-inch length containing the two leak sites was cut from the examined tube.  This length was then split longitudinally.  An iridescent blue discoloration was observed on the tube inner diameter (ID) surface at both leak sites, as well as an obvious crack at one leak site (Figure 2). 


Figure 2. Blue discoloration on tube ID surface at leak site.  Yellow arrow indicates crack.

In order to determine if failure modes other than cracking were producing the tube leaks; the leak site without the obvious crack was chosen for a metallographic examination (Figure 3). The chosen tube section was mounted in a cold-curing epoxy and metallographically ground and polished in accordance with standard procedures.  The prepared metallographic specimen presented the leak site in the transverse cross-sectional direction.  Examination of the specimen revealed four branching cracks, indicative of stress corrosion cracking (SCC), originating at the tube OD surface and proceeding either nearly or completely through the tube wall.  Negligible corrosion on the OD surface was noted except for shallow pits (approximately 0.4 mils deep) at a through wall crack.  The specimen was etched, revealing that the cracking was intergranular.  The intergranular attack was severe enough to produce grain dropping.  Etching revealed an apparently normal annealed copper microstructure (Figure 4). 

Figure 3. As-polished metallographic specimen showing through-wall crack.  Note shallow pits at crack origin at bottom of photomicrograph.  (250X original magnification) 

Figure 4. Etched metallographic specimen showing intergranular nature of cracking.  (Potassium Dichromate Etch, 1250X original magnification) 


The black and white deposits on another tube from the coil section were examined using a scanning electron microscope (SEM) fitted  with an energy dispersive spectroscopy (EDS) microprobe.  EDS revealed that the deposits were mostly zinc (approximately 30 weight percent; from the galvanized tube sheet) with magnesium, aluminum, silicon, phosphorus, sulfur, chloride, potassium, and calcium (i.e., “dirt”) each present in quantities from approximately 1 to 3 weight percent.  (Copper made up the balance.)  EDS analysis of matter present in the branching cracks indicated that it was copper oxide. 



Stress corrosion cracking (SCC) refers to cracking of an alloy that occurs under the simultaneous action of corrosion and sustained tensile stress.  Cracking can be either intergranular (as in the present case) or transgranular, depending on the alloy and/or the corrodent.  The classic example of SCC is intergranular SCC of stressed austenitic stainless steels (e.g. Type 304L stainless steel) exposed to hot, aqueous, chloride-containing environments.  Branching of the cracking occurs in the direction of crack propagation.   

High-purity coppers, such as phosphorized copper (C12200), are generally considered almost immune to SCC; however, intergranular SCC of this alloy has been observed.  The usual culprit in SCC of copper and copper alloys is ammonia (or other amines capable of reacting with copper to form complex ions) acting in tandem with stresses in the alloy from forming and/or service conditions.  The source of ammonia can be enormously diverse, from the decomposition of organic matter, to flooring adhesives, to cleaning products. 

In the present case, cracking of the copper tubing initiated at the OD surface and propagated intergranularly through the tube wall in a direction normal to the tube surface.  Stress in the copper tubing was due to the forming of the hairpin bends (where the failures occurred); away from the hairpin bends, the copper tubing was in a stress-free annealed condition. 

EDS analysis did not reveal a definitive “bad actor” chemical species in the cracking failures; but, if ammonia was the culprit, very little or no evidence of its presence would be expected to be found by EDS or other methods.  The presence of the blue discoloration on the tube ID surface at the failure sites was indicative of an ammonia-copper complex forming species. 

Site Index

Site Copyright © 1995 - 2007, All Rights Reserved,

Corrosion Testing Laboratories, Inc.

60 Blue Hen Drive

Newark, Delaware USA 19713

Phone: 1-302-454-8200

Fax: 1-302-454-8204