by A. A. Brouwer, The Dow Chemical Co.
D CORROSION A
W O R K B O O K
F E A T U R E
E
C
Corrosion Experiences with Dissimilar Metals It is hard to predict the performance of metal or alloy combinations, but a tabulation based on the galvanic series of metals helps GALVANIC corrosion has been the subject of numerous articles in the past decade. T h e fundamental aspects have been ably described, along with general pitfalls and preventive techniques. With this wealth of information, bimetallic corrosion should be a thing of the past. Unfortunately, failures due to this phenomenon continue to occur. Improper material selection or design usually accounts for a majority of these cases. However, the performance of certain metal or alloy combinations is often difficult to predict, particularly those involving magnesium, aluminum, or stainless steel. Extensive research is currently in progress to understand better the galvanic corrosion behavior of these materials.
When dissimilar metals in electrical contact are exposed to a corrosive or conductive medium, galvanic current flows from one metal to the other. Corrosion of the more active or anodic member of the couple usually increases by an amount directly related to the galvanic current flow by Faraday's law. This portion of the total anode corrosion is called galvanic corrosion. Attack of the more noble or cathodic element may be unaffected, partially reduced, or completely eliminated, depending on specific circumstances. In certain cases, particularly where a passive metal is the cathodic member of a couple, galvanic current flow may cause cathodic corrosion due to the breakdown of inherent passivity. The standard e.m.f. series of the elements cannot be relied upon to predict potential differences between metals in a corrosive environment. It should not be used to pre-
dict the corrosion behavior of dissimilar metal combinations. In the e.m.f. series the metals are arranged according to the potentials they exhibit when immersed in a lAf solution of their respective ions. Conditions are seldom, if ever, standard in irreversible corrosion processes. A more useful tabulation is found in the galvanic series of metals, based upon actual corrosion testing exGalvanic Series of Metals and Alloys" Corroded end (anodic, or least noble) Magnesium Magnesium alloys Zinc Aluminum 2S Cadmium Aluminum 17ST Steel or iron Cast iron Chromium-iron (active) Ni-Resist 18-8 stainless (active) 18-8-3 stainless (active) Lead-tin solders Lead Tin Nickel (active) Inconel (active) Brasses Copper Bronzes Copper-nickel alloys Monel Silver solder Nickel (passive) Inconel (passive) Chromium-iron (passive) 18-8 stainless (passive) 18-8-3 stainless (passive) Silver Graphite Gold Platinum Protected end (cathodic, or most noble) " Based on International Nickel Co. data.
I/EC
perience with numerous corrosive media, both in the laboratory and under service conditions. Although this series indicates only the tendencies of the metals to promote galvanic corrosion, the general order in which the materials are grouped is valid for most aqueous systems. The corrosion characteristics of the components of a bimetallic system may vary markedly from those that normally would be predicted from the galvanic series. Environmental factors (anions, cations, pH, aeration, and velocity) influence the static potential of most metals. Passivity induced by oxide formation, corrosion product films, adsorption, and surface reactions has a similar effect. Anodic and/or cathodic polarization phenomena may ultimately decide the utility of a particular dissimilar metal combination. Corrosion experiments with questionable bimetal systems are best conducted in the anticipated service environment. When bimetallic combinations must be used, several methods are available to limit the effects of galvanic corrosion. T h e materials selected should be in the same or at least an adjacent group in the galvanic series. Insulating materials— fittings, inserts, or nipples—may be used to isolate the components of the couple. A protective coating may serve as a barrier between metal and electrolyte. The coating should be applied to both anode and cathode elements, or, if this is not feasible, to only the cathodic metal. T h e anode-to-cathode area ratio should be large, thus minimizing the anodic current density. Corrosion inhibitors or cathodic protection may often effectively mitigate galvanic corroORKBOOK
FEATURES
73 A
I
/
EC
CORROSION
•
A Workbook Feature
Case Histories of Galvanic Corrosion Case I. Stainless steel heating coils were installed in an ammonium nitrate storage tank fabricated from Type 5052 aluminum. The heater tubes were parallel to and within 12 inches of the bottom of the tank. Average solution temperature was maintained at 95° C. Selective attack of aluminum occurred adjacent to weldments and in areas subjected to mechanical stresses due to bending or buckling along the tank bottom. Complete perforation was observed within 4 weeks. Although stress corrosion cracking was indicated, it was suspected that the failures were accelerated, and possibly initiated, by galvanic attack. After repairs, the stainless steel coils were completely insulated from the aluminum tank. When the vessel was examined several months later, no further attack could be detected. Case II. A drainage system for a brine filter was constructed from 2-inch-diameter perforated steel tubes and cast iron fittings (cross joints and caps). This steel-cast iron network was embedded in granular anthracite coal which completely filled the filter tank. The system failed in 2 years. The cast iron fittings graphitized severely in the carbonaceous brine environment. Localized galvanic attack of the steel tubing immediately adjacent to the graphitic cast iron members caused circumferential grooving and subsequent failure. A laboratory corrosion cell utilizing equal-area specimens of graphitized cast iron and steel tubing from the filter tank produced a steady-state galvanic current flow in the coal -brine mixture equivalent to a steel dissolution rate of 0.083 inch per year. Cast iron fittings were replaced with steel. No subsequent failures have been encountered in 4 years. Case III. Aluminum clamps were used to support copper bus bars at a transformer station. The clamps were secured with steel nuts and studs and the entire assemblies painted with an alkyd resin system. Atmospheric contamination from an adjacent power house utilizing coal for fuel caused rapid failure of the coating and, subsequently, severe galvanic corrosion of the aluminum. The clamps eventually cracked along the stud holes because of stress concentrations built up by the voluminous aluminum corrosion products. This condition was corrected by using copper clamps, studs, and nuts. Case IV. Brass nozzles were used on small-diameter steel piping in the jet aerating system at a waste water aeration pond. This effluent water contains residual organic materials and hydrogen sulfide, and is high in chloride, sulfate, and phosphate. Galvanic attack concentrated at the steel—brass interface and was particularly severe on the exposed pipe threads. A gradual circumferential decrease in pipe wall thickness was observed and within 2 years all units required new component parts. The brass nozzles were subsequently insulated from the steel piping with high-impact PVC nipples. Protective coatings are being used to mitigate corrosion of the steel network in this aggressive environment, particularly crevice attack at steel-to-steel threaded joints.
sion. Laboratory corrosion tests yield valuable information as to galvanic polarization characteristics, carbonate scale deposition, and passivation phenomena. Singly or in combination, these techniques have proved effective in controlling galvanic corrosion. Dissimilar metal corrosion can be 74 A
Case V. Severe corrosion of Type 312 weld metal occurred in a heat exchanger, where the welds joined Type 304 stainless steel tubes to a mild steel jacket. The unit was acidized frequently with inhibited 10% hydrochloric acid at a temperature of 70° C. Laboratory corrosion tests on this combination of metals in the inhibited acid media revealed that attack occurs predominantly in the weld metal (>10 inches per year). The corrosion rate of isolated 312 weld metal exceeded that of uncoupled mild steel and Type 304 stainless steel 12- to 15-fold. The deterioration of the weld metal in the heat exchanger was due to galvanic action between the ferritic and austenitic phases of the alloy during acidizing. Static potential and galvanic polarization studies with typical austenitic and ferritic stainless steels substantiated this conclusion. Inco A and Type 310 welding rod were found to be suitable for this application. A chloride-free solvent was recommended for acidizing stainless steel equipment. Casé VI. Stainless steel valves (selected because of immediate availability) were installed on an aluminum pipeline handling slightly acidic hot water, pH 6. Within a few weeks intense pitting of the aluminum pipe was observed adjacent to each valve. Aluminum valves were subsequently used and the attack was soon neutralized by oxide formation. Case VII. Extensive laboratory investigation has shown that copper-steel and aluminum-steel bimetallic systems are satisfactory for handling service cooling water. This particular water has an average total dissolved solids content of 550 p.p.m., it possesses moderate scaling characteristics, and pH varies from 7.8 to 8.2. Experiments with copper-steel and aluminum-steel couples in flowing water at 50° C. with anode-to-cathode area ratio varying from 10:1 to 1:10 revealed insignificant galvanic corrosion. Rapid polarization, predominantly of the cathodic element, restricts the flow of galvanic current to extremely low values. The polarization phenomenon is supplemented in its action by the deposition of calcareous material on the cathodic surface. In prolonged exposure tests these combined factors have effectively lessened galvanic action. Several heat exchangers with copper tubes and steel tube sheet and shell have been handling service water for over 5 years at temperatures in the various applications ranging from 40° to 70° C. When examined annually for galvanic corrosion, particularly at the tube -tube sheet junction, no accelerated attack of the steel members has been observed. Case VIII. Cast iron fittings (tees, els, reducers) were, used on large diameter steel lines handling calcium chloride brine at a temperature of 90° to 100° C. The brine velocity varied from 5 to 7 feet per second. Failures occurred in the steel piping adjacent to these fittings in less than 2 years. The attack was highly localized. Examination of the cast iron members revealed an adherent graphitic corrosion product deposit nearly '/i inch thick. In this environment the graphitized cast iron was found to be 400 mv. cathodic to steel. Both elements of the couple exhibited a marked resistance toward galvanic polarization under high velocity conditions. These factors produced a localized corrosion rate for steel piping in excess of 0.1 inch per year.
an aggravating factor in maintaining equipment, although in some corrosive environments apparently aggressive couples may not cause excessive corrosion. Unless experience can dictate the extent of corrosion due to galvanic coupling, a good procedure is to investigate the effect of the corrosive on the couple to be used.
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Our authors like to hear from readers. If you have questions or comments, or both, send them via The Editor, l/EC, 1155 16th Street N.W., Washington 6, D C . Letters will be forwarded and answered promptly.
