Handbook Corrosion (1992) WW Part 8 ppsx

Very little work has been performed on SCC in alcohols other than methanol.The studies that have been reported on alloys other than commercially pure titanium show that certain - alloys,

Nitrogen Tetroxide. As titanium use in aerospace increased, it was found that titanium alloys were highly resistant to corrosion attack in nitrogen tetroxide (N2O4), an oxidizer used with hydrazine rocket fuels Unfortunately, SCC was rather dramatically revealed in an explosion during proof testing of a Ti-6Al-4V storage vessel The vessel that exploded had been exposed to N2O4 at 40 °C (100 °F) at a stress level of 620 MPa (90 ksi) Testing revealed that titanium alloys would crack in NO-free N2O4 (that is, N2O4 with excess dissolved oxygen, or red N2O4) but would not crack in NO-containing

N2O4 (oxygen free, or green N2O4) (Ref 127)

Methanol. Prior to the discovery of SCC in N2O4, methanol was found to cause stress cracking of titanium alloys Methanol and NaBr were shown to be extremely corrosive to titanium, and in some cases, they promoted intergranular SCC of smooth specimens (Ref 157) Soon after this discovery, it was shown that methanol/HCl and methanol/H2SO4

mixtures also caused SCC of commercially pure titanium once again on smooth specimens (Ref 158) Stress-corrosion cracking in methanol was dramatically rediscovered when a Ti-6Al-4V pressure vessel exploded during proof testing with methanol This led to a flurry of research to discover the nature of this cracking phenomenon and the metallurgical factors that promoted cracking

Two types of SCC are observed in methanol solutions and are characterized by the failure mode exhibited In the first type, intergranular fracture is evident This type of fracture is common for commercially pure titanium and titanium alloys, such as Ti-13V-11Cr-3Al, that are exposed to methanol containing a halide ion, such as Cl- or Br- Susceptibility to SCC measured as time to failure of smooth samples indicates that:

• Increasing the halide content decreases time to failure

• Water additions to a critical level decrease time to failure

• Higher halide concentrations increase the critical level of water for maximum susceptibility to SCC

• Water levels beyond the critical level reduce and can inhibit cracking susceptibility

Details of this failure mode are presented in Ref 113, 114, 115, 116 and 159, 160, 161

The effects of both cathodic and anodic polarization have also been investigated Anodic polarization increases the susceptibility of titanium to SCC in methanol/halide mixtures On the other hand, cathodic polarization dramatically reduces SCC susceptibility, as shown in Fig 34 Potentials more negative than -250 mV versus Ag/AgCl prevent cracking

in methanol (Ref 161)

Fig 34 Time to failure versus applied current and water content for cold-rolled and annealed Ti-6Al-4V stressed

to 75% of yield strength in a methanol/HCl mixture For 0.08% H 2 O, 0.15% H 2 O, and 0.20% H 2 O, there was no failure in time shown

Metal ion additions have also been examined and found to affect methanol SCC In general, additions that have altered the cathodic reaction, such as palladium, chromium, iron, and gold, have accelerated cracking This is shown in Fig 35 for palladium

Fig 35 Effect of palladium on time to failure and electrode potential for commercially pure titanium in

methanolic solutions

The other fracture mode, typical of highly alloyed titanium, is characterized by transgranular -phase cleavage Alloys such as Ti-8Al-1Mo-1V typify this failure mode In tests using precracked specimens, cracking changes from intergranular in stage I cracking (SCC initiation, Fig 8) to transgranular in stage II (SCC propagation) Most of the and

- alloys susceptible to SCC in neutral aqueous solutions (discussed later) exhibit this mode of failure In contrast, titanium alloys, such as Ti-11.5Mo-6Zr-4.5Sn, exhibit intergranular fracture in stage II Stage II is the region in which crack velocity is essentially independent of stress intensity (SCC propagation)

In general, halide additions increase crack velocity in stage II It has also been reported that additions of sulfuric acid and acetic accelerate cracking Application of cathodic potential between -1.5 and -1.0 V versus SCE has been shown to prevent crack initiation (stage I) Anodic potentials appear to increase crack velocity As indicated earlier, small water additions prevent methanol SCC initiation This effect is shown in Fig 36 Little work has been performed on the effect

of temperature on methanol SCC; however, the data available indicate that crack velocities increase with temperature

Fig 36 Effect of bromide and chloride additions on SCC of cold-rolled and annealed commercially pure titanium

stressed to 75% of yield strength in methanol/water solutions Arrows indicate no failure in time shown

Titanium alloys known to be susceptible to aqueous SCC (discussed later) are also susceptible to methanol SCC, and the alloys most susceptible to seawater are also most susceptible to methanol In addition, the alloys that are susceptible to SCC in distilled water are not beneficially affected by the inhibiting effect of water (Fig 37)

Fig 37 Effect of water on crack velocity for Ti-8Al-1Mo-1V in methanolic solutions

Other Alcohols. Very little work has been performed on SCC in alcohols other than methanol.The studies that have been reported on alloys other than commercially pure titanium show that certain - alloys, such as Ti-6Al-4V, may be susceptible to cracking in anhydrous ethanol Cracking in ethylene glycol has also been reported for Ti-8Al-1Mo-1V Other work indicates that cracking susceptibility significantly diminishes as the number of carbon atoms in the alcohol increases

Halogenated Hydrocarbons. No testing of commercially pure titanium has been performed in common

hydrocarbons However, widespread use of titanium alloys in the aerospace industry has prompted considerable study of SCC in halogenated hydrocarbons common to aerospace processing Stress-corrosion cracking of certain titanium alloys has been identified in the following hydrocarbons:

In most of these environments, precracked specimens (category 2) are required to identify SCC

Carbon tetrachloride (CCl4) SCC was first noted in Ti-8Al-1Mo-1V (Ref 162, 163, 164, and 165) The threshold stress intensity was approximately the same as that observed for SCC in 3.5% NaCl Crack velocities in CCl4 were approximately ten times faster than velocities in methanol Studies on dynamically loaded smooth specimens (category 3) also showed that Ti-5Al-2.5Sn was susceptible to SCC in CCl4 at stresses approaching the tensile strength of the alloy

The other hydrocarbons identified were found to cause cracking in Ti-8Al-1Mo-1V and Ti-5Al-2.5Sn, alloys known to be susceptible to SCC in distilled water (Ref 165, 166) No other alloys were found to be similarly affected

Freons include any of a number of fluorinated hydrocarbons commonly used as refrigerants Titanium alloys 1Mo-1V and Ti-5Al-2.5Sn have been found to exhibit threshold stress intensities in commercial freons below air threshold stress intensities (Ref 166) The alloy Ti-6Al-4V was also identified as susceptible when exposed in the solution-treated and aged condition

Ti-8Al-Hot Salts. In the late 1950s, cracking of a titanium alloys was discovered during routine creep testing The failure was eventually traced to chlorides from fingerprints on the creep specimen These findings were reproduced in several laboratory studies for a host of titanium alloys Nearly all titanium alloys were found to be susceptible to this cracking phenomenon (termed hot salt cracking) with the exception of the commercially pure grades of titanium With this discovery, a great deal of concern was expressed with regard to the multitude of existing applications similar to this laboratory environment However, after much investigation, it was found that no failure in the field could be attributed to hot salt cracking

Several complete descriptions of host salt cracking can be found in th literature (Ref 167, 168, 169, 170, 171, 172, 173,

174, and 175) Hot salt cracking is primarily influenced by temperature, stress, time, and the alloy itself Cracking is observed in the temperature range from 285 to 425 °C (545 to 800 °F) In general, susceptibility increases with stress and/or temperature and does not occur below 260 °C (500 °F) or above 540 °C (1000 °F)

Cracking is normally characterized by extensive branching and is not necessarily associated with the regions of highest stress intensity; therefore, category 2 specimens are not required to initiate cracking Indeed, it is often difficult to initiate cracks in precracked notches Statically loaded beam specimens (category 1) have been used in most of the laboratory investigations

The alloys that are most susceptible to hot salt cracking are alloys with more than 3% Al, such as Ti-5Al-2.5Sn Commercially pure titanium is apparently immune (Ref 176) Alpha-beta alloys are less susceptible to cracking, although alloys with high aluminum contents are most susceptible Apparently, the least resistant titanium alloy is Ti-8Al-1Mo-1V Alloys with higher molybdenum content, such as Ti-4Al-3Mo-1V, are most resistant (Ref 169) The combined effect of time, temperature, and stress is shown in the Larsen-Miller diagram in Fig 38 for several alloys From Fig 38, it is clear that alloy type and microstructural condition are important

Fig 38 Larsen-Miller plot for hot salt cracking of several annealed - titanium alloys T is temperature (°R),

and t is exposure time (hours)

Oxygen has been reported as necessary for hot salt cracking At least one study has shown that cracking will not occur in Ti-5Al-2.5Sn when the environmental pressure is reduced below 10 m (Ref 169) Although the role of water (moisture) has not been clearly established, it appears that water is also a necessary environmental component in the cracking process (Ref 170, 171)

Chloride, bromide, and iodide salts have all been shown to produce similar cracking Fluoride and hydroxide salts have not The cation associated with the salt has also been reported to affect cracking susceptibility The severity of attach has been shown to increase as follows (Ref 170, 171):

MgCl2 > SrCl2 > CsCl > CaCl2 > KCl >

BaCl2 > NaCl > LiCl

Table 28 lists titanium alloys in order of their susceptibility to hot salt cracking This list is taken from Ref 177 and has not met with unanimous agreement

Table 28 Relative resistance of titanium alloys to hot salt cracking

From a practical standpoint, hot salt cracking appears to be a phenomenon that is restricted to the laboratory As indicated earlier, no in-service failure has been attributed to hot salt cracking The likely reason for this is the critical relationship among environment, stress level, and alloy type Unless all of the conditions are met simultaneously and for extended time, cracking will not occur

Molten Salts. It would appear that Ti-8Al-1Mo-1V is the only titanium alloy tested for SCC in molten salt environments Cracking has been observed in pure chloride and bromide eutectic melts at temperatures between 300 and

500 °C (570 and 930 °F) In general, increasing temperature increases crack velocity Cathodic protection has been observed to inhibit or stop cracking

Nitrate salts below 125 °C (255 °F) do not induce cracking even when Cl-, Br-, or I- anions are present At higher temperatures in pure molten nitrates, cracking can occur only when halides are present (Ref 164)

Liquid/Solid-Metal Embrittlement. Several metals, both in liquid and solid form, have been found to induce cracking in contact with titanium alloys (Ref 127, 164, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, and 190) The first reported incidence stemmed from a cracked compressor disk in contact with cadmium-plated steel bolts (Ref 179) Initial speculation hinted that the exposure temperature may have been above the melting point of cadmium, leading

to liquid-metal embrittlement However, later work found that cracking would occur well below the melting point of cadmium (Ref 186), such as at room temperature for Ti-6Al-4V

Those metals known to cause cracking of titanium alloys include cadmium, mercury, zinc, and certain silver brazing alloys The titanium alloys that are known to be susceptible to cracking in cadmium include commercially pure titanium (ASTM grade 3) with more than 0.2% oxygen, Ti-4Al-4Mn, Ti-8Mn, Ti-31V-11Cr-3Al, Ti-6Al-4V, and Ti-8Al-1Mo-1V

It is likely that most other titanium alloys are susceptible, but have not been tested

Alloys tested and found to crack in mercury include commercially pure titanium (ASTM grade 4, 0.3% oxygen), 8Mn, Ti-13V-11Cr-3Al, Ti-6Al-4V, and Ti-8Al-1Mo-1V As with cadmium, other alloys are probably susceptible, but have not been tested

Ti-Zinc, in both solid and liquid form, has been reported to cause cracking of titanium alloys However, there is conflicting evidence in the literature as to whether this is actually the case

Silver and silver brazing alloys have been shown to cause cracking in titanium alloys that are particularly sensitive to SCC These alloys include Ti-8Al-1Mo-1V, Ti-5Al-2.5Sn, and Ti-7Al-4Mo (Ref 187, 188) As with cadmium, both solid and liquid forms of silver may produce cracking Susceptibility for Ti-6Al-4V is considered to be above 345 °C (650 °F)

Aqueous Environments. Under certain metallurgical conditions, several titanium alloys have been shown to be susceptible to SCC in distilled water These include Ti-8Al-1Mo-1V, Ti-5Al-2.5Sn, and Ti-11.5Mo-6Zr-4.5Sn Microstructural variation for each alloy affects the degree of susceptibility For example, mill-annealed Ti-8Al-1Mo-1V is less susceptible than step-cooled Ti-8Al-1Mo-1V Testing has been performed with category 2 type specimens where crack velocity and threshold stress intensity are determined In these alloys, the degree of susceptibility is highly dependent on heat treatment

Test results in neutral-pH environments with category 2 type specimens indicate that titanium alloys exhibit a threshold

stress intensity, KISCC, below which cracks will not propagate The individual effects of ionic species, concentration, potential, pH, and so on, have been extensively studied and are discussed below

Ionic Species. The anions Cl-, Br-, and I- are the only species shown to promote and/or induce SCC in titanium alloys The few alloys susceptible to cracking in distilled water become more susceptible, while alloys that are not susceptible in distilled water may become susceptible when these species are present Anions such as , , OH-, F-,

Cr2 , and may reduce sensitivity

In general, cations do not alter SCC sensitivity However, oxidizing cations, such as Fe3+ or Cu2+, can increase KISCC in more susceptible alloys This effect is analogous to anodic polarization, which is discussed in the section "Potential and pH" in this article

Concentration. As shown in Fig 39, increasing the concentration of anions that promote SCC generally increases crack

velocity (Ref 164, 191, and 192) and decreases KISCC Additions of and to distilled water can completely inhibit SCC in alloy/heat-treatment combinations that are moderately susceptible, such as mill-annealed Ti-8Al-1Mo-1V More susceptible combinations, such as step-cooled Ti-8Al-1Mo-1V, are not similarly affected

Fig 39 Effect of halide ion concentration on KISCC and crack velocity in aqueous solutions A, highly susceptible;

B, moderately susceptible; C, slightly susceptible

Potential and pH. It is difficult to discuss the effect of potential and pH on SCC independently In neutral brines

containing halides, KISCC varies significantly with potential Both cathodic and anodic polarization increase KISCC Within

a narrow potential region, KISCC reaches a minimum Both the start and breadth of this range are dependent on the alloy

and its metallurgical condition In general, the range of maximum susceptibility (minimum KISCC) occurs at approximately -500 mV versus SCE Examples of this behavior are shown in Fig 40

Fig 40 Stress-corrosion cracking of titanium alloys as a function of potential in 0.6 M KCl at 24 °C (75 °F)

Source: Ref 164

Stress-corrosion crack velocity is also affected by potential For alloys such as Ti-13V-11Cr-3Al, crack velocity increases linearly with anodic polarization (Ref 192) The slope of the crack velocity curve depends on the alloy type and metallurgical condition, and application of sufficient cathodic potential can halt a stress-corrosion crack

In acidic solutions, potential effects are somewhat different (Ref 193) First, cathodic polarization will not stop a

propagating crack Second, lowering pH usually decreases KISCC and increases crack velocity at a constant potential, as shown in Fig 41 In alkaline solutions, SCC is similar to that found in neutral solutions (Ref 194, 195)

Fig 41 Stage II crack velocity as a function of pH and potential in aqueous solutions

Temperature. Limited published data were available on the effect of temperature on SCC in aqueous environments

(Ref 164, 191) Within a narrow temperature range (0 to 93 °C, or 32 to 200 °F), KISCC in NaCl solution was found to be independent of temperature for Ti-8Al-1Mo-1V However, crack velocity was found to be strongly dependent on temperature

Experimentation with category 1 and category 2 specimens has shown that the alloy Ti-3Al-8V-6Cr-4Mo-4Zr does not crack in NaCl brines up to 205 °C (400 °F) (Ref 196) Tests above that temperature in NaCl brines containing H2S may produce cracking with category 3 specimens

Hydrogen Sulfide. Tests in acidic NaCl brine environments containing hydrogen sulfide employing category 1 specimens indicate that titanium alloys are immune to SCC at stress levels up to their yield strength (Ref 138) It is, however, unclear whether the category 1 type specimens are capable of determining SCC susceptibility in these environments As noted in the previous section, elevated-temperature tests in H2S brines indicate that Ti-3Al-8V-6Cr-4Mo-4Zr is susceptible to SCC Unfortunately, no similar tests in environments with H2S have been reported Therefore,

H2S may not be the species responsible for SCC in this alloy

Metallurgical Effects. In addition to environmental effects, the metallurgical condition of a particular titanium alloy will influence its susceptibility to SCC Both chemistry and heat treatment (microstructure) play a part in these effects

In alloys, the two most important alloying elements are aluminum and oxygen Binary titanium-aluminum alloys have been extensively studied In this investigation, it has been established that 5% Al is necessary for SCC to occur in

aqueous environments As the aluminum level is increased, KISCC decreases and crack velocity increases (Ref 193, 197) The most susceptible heat treatments in aluminum-containing alloys are those that produce the 2 phase

Binary titanium-oxygen alloys also exhibit a critical level of oxygen below which SCC does not occur, because of the transition from wavy to planar slip (Ref 198) This level, generally taken as less than 0.20 to 0.25%, represents the break between ASTM grade 2 and ASTM grade 3 Titanium-aluminum-oxygen alloys generally suffer from the cumulative effects of both aluminum and oxygen on SCC (Ref 198) Tin additions to titanium-aluminum alloys generally cause a loss

in SCC resistance This is clearly demonstrated by Ti-5Al-2.5Sn, one of the more susceptible titanium alloys

The - titanium alloys are considerably more difficult to generalize with regard to SCC behavior This stems from the wide variety of microstructures that can be produced and the number of alloying elements involved

Stress-corrosion cracking of Ti-8Al-1Mo-1V has been extensively studied because of its sensitivity to microstructure The martensitic structures produced by quenching from high-temperature solution treatment are immune to SCC (Ref 190, 191) Lower-temperature solution treatment produces an equiaxed - structure that is susceptible to SCC The degree

of susceptibility is determined by the grain size, volume fraction, and mean free path of the susceptible phase

Tempered martensitic structures, produced by annealing a martensitic microstructure, are also susceptible to SCC (Ref 190) Basket-weave or Widmanstätten microstructures produced by working and/or heat treatment above the transus generally exhibit better toughness both in and out of an aqueous environment

In general, titanium alloys with higher aluminum, oxygen, and tin contents are the most susceptible to SCC; the effect of aluminum is shown in Fig 42 Molybdenum is usually beneficial in increasing SCC resistance Microstructural effects in these alloys are similar to those discussed for Ti-8Al-1Mo-1V

Fig 42 Effect of alloy composition on SCC resistance of mill-annealed titanium alloys in aqueous 3.5% NaCl

solution at 24 °C (75 °F)

With the exception of Ti-13V-11Cr-3Al, all of the commercial titanium alloys are immune to SCC in the -phase condition (Ref 199, 200, 201, and 202) However, aging decomposes the phase and produces a variety of phases The phase, produced by low-temperature aging of many alloys, does not induce SCC susceptibility Aging at higher temperatures produces the phase, which nearly always leads to SCC The degree to which a given alloy in the + condition is susceptible appears to be related to the alloy chemistry and to the quantity and morphology of the phase

Aging at higher temperatures produces a coarser, which is less susceptible to SCC than the finer Alloys containing molybdenum are less susceptible to SCC, especially those without tin

Galvanic Corrosion in Specific Media

In their normal passive condition, titanium alloys are most often the cathode when galvanically coupled to most common engineering alloys in service As a result, galvanic corrosion of titanium is very rare and occurs only under very unusual conditions This rare situation could occur in a medium such as reducing acid, in which titanium is actively corroding In this case, coupling to a more active metal (Ref 203) or a more noble metal could accelerate titanium alloy corrosion only

if full passivation is not achieved Coupling titanium alloys to a more noble material is most often very beneficial, resulting in establishing passivity when titanium is marginally active (corroding) or further maintaining passivity of titanium at more noble potentials This form of anodic protection explains the excellent galvanic compatibility of titanium with noble metals (precious metals) and graphite composites in most environments

Titanium alloys exhibit relatively noble corrosion potentials in the many environments in which full passivity is achieved This is evident from Table 2 Similar data in natural seawater from other sources indicate that corrosion potentials for titanium and its alloys fall in the range of +0.1 to -0.3 V versus SCE (Ref 41, 44, 204, 205, and 206) Deaeration, increasing temperature (Ref 206), and sunlight (Ref 207) were all shown to cause a slight shift in titanium corrosion potential in the active direction

In all cases, these referenced sources reveal that titanium exhibits corrosion potentials that are very similar to those of other more resistant alloys in the passive conditions, including the stainless steels and nickel-base alloys The minor potential differences between these resistance, alloys result in very small, benign galvanic interactions as long as passive alloy conditions exist Thus, galvanic compatibility can be expected in most environments when a titanium alloy is coupled to another resistant alloy, assuming both exist in a fully passive condition The data in Table 29 for stainless steel and nickel-base alloys coupled to titanium in marine environments support this point

Table 29 Corrosion rates of various metals galvanically coupled to titanium in marine exposures

Corrosion rate after indicated exposure, mm/yr (mils/yr)

193 days at half tide 56 months in sea air Coupled material

Uncoupled Coupled (a) Coupled (b) Uncoupled Coupled (b)

Alclad 2024-T3 0.015 (0.6) 0.03 (1.2) 0.043 (1.7) 0.001 (0.04) 0.007 (0.28)

Copper 0.013 (0.5) 0.023 (0.9) 0.025 (1) 0.002 (0.08) 0.006 (0.24)

Low-carbon steel 0.15 (6) 0.31 (12.2) 0.43 (17) 0.156 (6.1)

Monel alloy 400 0.025 (1) 0.003 (0.12) 0.003 (0.12) nil 0.001 (0.04)

AISI type 302 stainless steel 0.002 (0.08) nil 0.003 (0.12) nil nil

AISI type 316 stainless steel nil nil nil nil nil

(a) Area ratio of titanium to other metal: 1 to 7

(b) Area ratio of titanium to other metal: 7 to 1

However, when a titanium alloy is coupled to a metal that is active (corroding or pitting) in an environment, accelerated anodic attack of the active metal may ensue Depending on the environment, active metals may include carbon steel, aluminum, zinc, copper alloys, or stainless steels that are active or pitting

The effect of enhanced galvanic corrosion of various copper alloys coupled to commercially pure titanium in ambient temperature seawater is illustrated in Fig 43 and 44 These sources and others reveal a wide variation in the extent of galvanic corrosion of copper alloys and steel (Ref 43, 69, 208, 209, and 210) The degree of galvanic attack depends on many (often interacting) factors, including cathode-to-anode surface area ratio, concentration of dissolved cathodic depolarizers such as oxygen and atomic hydrogen, temperature (Ref 44, 208), medium flow velocity and flow characteristics (that is, turbulence, angle of impingement), and medium chemistry The presence of sulfide, increasing cathode-to-anode area, oxygen content, and flow velocity all aggravate the galvanic attack of copper alloys in seawater

As shown in Fig 45, deaeration of NaCl brines drastically reduces galvanic attack (compare to Fig 43) Discussions of galvanic interactions between titanium and a wide array of engineering alloys in flowing seawater are presented in Ref 74

Fig 43 Corrosion of various copper alloys that were galvanically coupled to titanium in aerated seawater at 25

°C (77 °F) Compare with Fig 45 Source: Ref 41, 44

Fig 44 Corrosion of dissimilar metals coupled to titanium in flowing ambient-temperature seawater Source:

Ref 56

Fig 45 Corrosion of copper alloys that were galvanically coupled to titanium in boiling, deaerated 6% NaCl at

100 °C (212 °F) Compare with Fig 43, which shows corrosion rates in aerated seawater Source: Ref 44

As a result of the relatively high polarization resistance (2.6 × 106 · cm2) and hydrogen overvoltage of titanium in ambient seawater, the cathodic behavior of titanium is similar to that of the 18-8 stainless steels (Ref 211) Therefore, the galvanic effects of titanium on active metals are quite similar to those for 18-8 stainless, as observed in salt spray tests (Ref 212) Although these cathode characteristics tend to mitigate galvanic current, they also result in increased cathodic current throwing power in conductive electrolytes; therefore, the effective surface area of titanium involved in a galvanic couple may be quite large For example, studies show that the effective length of titanium (and stainless steel) condenser tubing involved in the galvanic attack of copper alloy tubesheets in seawater is in excess of 6 m (20 ft) (Ref 208, 211, and 213) This differs significantly from the two-tube diameter effective length rule used for copper alloys in seawater Thus, significant cathodic polarization of titanium may occur in nonoxidizing media, such as seawater, when coupled to a more active metal This often leaves the galvanic couple in cathodic control

Effective design strategies for limiting or avoiding galvanic attack of active metals include:

• Coupling to more compatible (passive) alloys (including all-titanium design)

• Use of dielectric (insulating) joints between dissimilar metals

• Cathodic protection of the active metal by either impressed current or sacrificial anode means (Ref 44,

208, and 210)

Coating of titanium cathode surfaces may also mitigate galvanic current, assuming that the galvanic couple is in cathodic control

Erosion-Corrosion in Specific Media

Unalloyed titanium and grade 5 titanium have both been shown to withstand silt-free flowing seawater to velocities as high as 30 m/s (100 ft/s) (Ref 41, 57, 214, 215, 216, 217, and 218) In fact, high-speed water wheel tests in seawater indicate erosion rates for grade 5 titanium of approximately 0.013 mm/yr (5 mils/yr) at 46 m/s (150 ft/s) (Ref 216) Jet impingement tests also involving seawater velocities of 46 m/s (150 ft/s) reveal rates of 0.03 to 0.06 mm/yr (1.2 to 24 mils/yr) for commercially pure titanium, with values of approximately 0.03 mm/yr (1.2 mils/yr) for welded and unwelded grade 5 titanium samples alike (Ref 218) Extremely low erosion rates are also reported for grade 2 titanium at various

seawater locations, as shown in Table 30 The superior erosion-corrosion resistance of titanium has also been reported in other media (Ref 103)

Table 30 Erosion-corrosion of grade 2 titanium in seawater at various locations

Brixham Sea 9.8 32 Model condenser 12 0.003 0.12

Kure Beach, NC 1 3.3 Ducting 54 7.5 × 10-7 0.00003

Kure Beach, NC 8.5 28 Rotating disk 2 1.3 × 10-4 0.005

Kure Beach, NC 9 29.5 Micarta wheel 2 2.8 × 10-4 0.01

Kure Beach, NC 7.2 23.6(a) Jet impingement 1 5 × 10-4 0.02

Wrightsville Beach, NC 1.3 4.3 6 1 × 10-4 0.004

Wrightsville Beach, NC 9 29.5 Micarta wheel 2 1.8 × 10-4 0.007

Mediterranean Sea 7.2 23.6(a) Jet impingement 0.5 0.5 mg/day

Dead Sea 7.2 23.6(a) Jet impingement 0.5 0.2 mg/day

(a) Included air

Studies involving sand and emery particle-laden seawater indicate satisfactory erosion-corrosion resistance to flow rates

of approximately 6 m/s (20 ft/s) (Ref 57) Data generated from rotating disk tests are presented in Table 31 The immunity

of titanium to erosion-corrosion in silt-laden seawater flowing at approximately 2 m/s (6.5 ft/s) has been demonstrated in more than 20 years of power plant surface condenser tube service (Ref 69, 72) The outstanding resistance of titanium alloys to cavitation damage has also been documented (Ref 217, 219, and 220), and it has been confirmed that the harder higher-strength titanium alloys are more resistant to cavitation (Ref 221)

Table 31 Erosion-corrosion of titanium grade 2 in seawater containing suspended solids

The relative erosion resistance of unalloyed titanium was investigated in two special versions of a rotating drum test involving erosive wear by wet TiO2 filter cake solids and feed slurry (Ref 224) Titanium exhibited minor metal loss, being measurably superior to the steels, stainless steels, and nickel alloys tested The superior erosion-corrosion resistance

of titanium compared to Hastelloy alloy C was noted in a high-velocity gas scrubber venturi device in which a rich solution was used to scrub hot (315 °C, or 600 °F) process gas (Ref 224)

chloride-Erosion-corrosion testing in coal-water slurries representative of those associated with coal-washing plants also affirmed the superior performance of unalloyed titanium as compared to carbon steel, Ni-hard cast iron, and stainless steel alloys (Ref 225) Test results revealed that titanium was inert to attack to slurry velocities of 5 m/s (16 ft/s) but that types 304 and 316, alloy 904L, and type 440C stainless steels exhibited significant wear above 2 m/s (6.5 ft/s) Unalloyed titanium that was thermally oxidized at 700 °C (1290 °F) provided full wear resistance to velocities as high as 8 m/s (26 ft/s) in these rotary tests

Extensive erosion-corrosion testing of Ti-6Al-4V, Ti-5Al-2.5Sn, and Ti-7Al-4Mo alloys has been conducted in velocity wet stream environments for application in low-pressure steam turbine blading in power plants These alloys have demonstrated superior resistance to 403 stainless steel (12 to 13% Cr steel) in operating turbines and in water droplet erosion and water jet impingement tests (Ref 220) Full erosion resistance of Ti-6Al-4V blades to velocities of 440 to 530 m/s (1450 to 1740 ft/s) at 10% steam moisture has been noted in turbines In fact, these studies suggest useful erosion resistance of Ti-6Al-4V in approximately 8% steam moisture to 549 m/s (1800 ft/s), and in 11% steam moisture to 488 m/s (1600 ft/s) (Ref 220) Single-shot water jet impingement testing has shown that annealed Ti-7Al-4Mo alloy is significantly more erosion resistant than 12% Cr steel, type 303 stainless steel, or Stellite alloy 6 at jet velocities of 610 and 915 m/s (2000 and 3000 ft/s)

high-Expanding and Enhancing the Corrosion Resistance of Titanium

The general corrosion resistance of titanium can be improved or expanded by one or a combination of the following strategies:

• Alloying

• Inhibitor additions to the environment

• Precious metal surface treatments

• Thermal oxidation

• Anodic protection

Alloying. Perhaps the most effective and preferred means of extending resistance to general corrosion into reducing environments has been by alloying titanium with certain elements Beneficial alloying elements include precious metals (>0.05 wt% Pd) (Ref 226, 227, 228, and 229), nickel ( 0.5 wt%) (Ref 226, 229, 230, 231, and 232), and/or molybdenum ( 4 wt%) (Ref 13, 229, 233, and 234) These additions facilitate cathodic depolarization by providing sites of low hydrogen overvoltage, which shifts alloy potential in the noble direction where oxide film passivation is possible Relatively small concentrations of certain precious metals (of the order of 0.1 wt%) are sufficient to expand significantly the corrosion resistance of titanium in reducing acid media

These beneficial alloying additions have been incorporated into several commercially available titanium alloys, including the titanium-palladium alloys (grades 7 and 11), Ti-0.3Mo-0.8Ni (grade 12), Ti-3Al-8V-6Cr-4Zr-4Mo, Ti-15Mo-5Zr, and Ti-6Al-2Sn-4Zr-6Mo As shown by the corrosion data in this article, these alloys all offer expanded application into hotter and/or stronger HCl, H2SO4, H3PO4, and other reducing acids as compared to unalloyed titanium The high-molybdenum alloys offer a unique combination of high strength, low density, and superior corrosion resistance

Inhibitor Additions. Various oxidizing species can effectively inhibit the corrosion of titanium in reducing acid environments when present in very small concentrations Typical potent inhibitors for titanium alloys in aggressive reducing acids are listed in Table 14 Many of these inhibitors are effective at levels as low as 20 to 100 ppm, depending

on acid concentration and temperature If not normally present in a given corrosive acid stream, minute additions of a process-compatible inhibitive species may be considered to protect titanium components These can be especially practical when process streams are recycled

Minute additions of water may be required to maintain titanium alloy passivity in certain anhydrous environments This has been highly effective in certain anhydrous organic compounds, absolute methanol, red fuming nitric acid, dry hydrogen or chlorine gas, and nitrogen tetroxide

Precious Metal Surface Treatments. Precious metals such a platinum and palladium have been ion plated, ion implanted, or thermal diffused into titanium alloy surface to achieve improved resistance to reducing acids (Ref 235) This approach has not been used commercially for industrial components because of high cost, coating application limitations, and the limitations (mechanical and corrosion damage) normally associated with very thin surface films However, ion plated platinum or gold surface films impart significant improvements in titanium alloy oxidation resistance

of temperatures up to 650 °C (1200 °F) (Ref 236, 237)

Thermal Oxidation. Protective thermal oxide films can form when titanium is heated in air at temperatures of 600 to

800 °C (1110 to 1470 °F) for 2 to 10 min The rutile TiO2 film formed measurably improves resistance to dilute reducing acids as well as absorption of hydrogen under cathodic charging (Ref 47) or gaseous hydrogen conditions Corrosion studies in hot, dilute HCl solutions have confirmed its superior protective benefits as compared to as-pickled, polished, or anodized surfaces on unalloyed titanium (Ref 47, 238) Corrosion and hydrogen uptake resistance was afforded by thermal oxidation in molten urea at 200 °C (390 °F) (Ref 238) Enhanced protection from dry chlorine attack can also be expected Like anodizing, thermal oxidation offers no improvements in titanium resistance in highly alkaline or oxidizing aqueous media

Although the thermal oxide has proved to be protective in relatively short-term tests in dilute reducing acids, long-term performance has not been fully demonstrated Mechanical damage and plastic strain of thermally oxidized components must be avoided for effective protection The oxide has been successfully applied on tubing and small components, but may be impractical for large components or where component distortion may occur during heating

Anodic Protection. Titanium alloys can be effectively protected in reducing acid media by impressed anodic (direct current) potentials Sustained impressed potentials in the range of + 1 to +4 V versus the standard by hydrogen electrode (SHE) are usually adequate to measure full passivation of titanium in many acids, as indicated in Table 16 Limited use of anodic protection by impressed currents has been made in concentrated H2SO4 and H3PO4 solutions in which a very wide range of impressed potentials can be applied The added cost of impressed current systems, challenges with protecting complex component geometries, and stray current problems have inhibited its application Also, titanium surfaces exposed to alternating wet/dry or vapor-phase conditions are not protected by this method

Other Surface Treatments. Surface films of titanium nitrides and carbides are highly resistant to reducing acids Studies have shown that the dense adherent nitride films produced by reactive plasma ion plating provided superior protection in deaerated H2SO4 solutions when compared to several other film-forming methods (Ref 239) Methods of applying nitride surface films to titanium include ion implantation (Ref 240), ion plating, sputter deposition, or thermal diffusion (nitrogen gas or molten cyanide bath) Because of the cost and limitations of film application and the inherent thin film performance limitations, these films are generally not used for corrosion resistance only The improved water resistance offered by these hard films is generally the primary incentive

The crevice corrosion resistance of titanium alloys can be enhanced by the following strategies:

• Alloying titanium

• Precious metal surface treatments

• Other metallic coatings

• Thermal oxidation

• Noble alloy contact

• Surface pickling (for smeared surface iron)

Alloying. The crevice corrosion resistance of titanium alloys tends to parallel general corrosion resistance in reducing acids Thus, alloying with certain precious metals, such as palladium (Ref 131, 226, 241), nickel (Ref 23, 133, 226, 231, and 232), and/or molybdenum (Ref 13, 138), also significantly improves the resistance of titanium to crevice corrosion (Ref 242)

The commercially available titanium alloys that exhibit superior crevice corrosion resistance include grades 7, 11, and 12

In addition, high-strength titanium alloys containing at least 4 wt% Mo offer an excellent combination of improved crevice corrosion resistance, low density, and high strength (Ref 13) Proper alloy selection depends largely on pH, temperature, and other conditions, as discussed previously in this article

Selection of a more resistant titanium alloy is generally preferred from a long-term reliability and cost standpoint over surface treatment options for less resistant titanium alloys On the other hand, localized surface treatments may be cost-effective when very heavy alloy wall sections are involved or when it is necessary to upgrade existing equipment components

Surface films of many precious metals and/or their oxides offer significant improvements in the crevice corrosion resistance of titanium In Fact, these treatments can offer crevice resistance approaching that of the grade 7 or 11 titanium alloy The most common precious metals applied to titanium surfaces are palladium, platinum, and ruthenium, and their oxides Thermal diffusion coatings of palladium (Ref 235), platinum, and ruthenium are readily applied by firing in an air furnace after coating titanium with modified metal chloride solutions (Ref 243) The resultant coating is typically a mixture of titanium and precious metal oxides, depending on firing temperature (Ref 235) Thermal palladium surface treatments have been successfully used in oil refinery tubular exchangers (Ref 244)

Other methods of applying these precious metals on titanium surfaces include electroplating, brush plating, and ion implantation Electroplating or brush plating provides precious metal layer thicknesses of the order of 0.01 mm (0.4 mils) Brush plating is a special technique for achieving localized electroplating of surfaces (Ref 245) Palladium implantation generates palladium-rich surface layers of the order of 0.5 m (0.02 mil) or less These surface layers have been shown to

be effective in hot, concentrated chloride brines (Ref 246)

Other Metallic Coatings. The application of certain metals and their oxides within titanium alloy crevices can effectively inhibit crevice corrosion initiation in hot chloride media These metals include nickel and copper, their oxides,

and the Fe2O3 and MoO3 oxides (Ref 247, 248) In finely powdered form, these materials can be painted on creviced titanium surfaces in slurry form Alternative, these powders can be formulated into sealants, such as silicone or ethylene propylene diene monomer (EPDM) rubber For example, a 5 wt% addition of nickel metal/NiO powder (50/50) blended into EPDM gaskets for chlor-alkali cell anode components is known to prevent the crevice corrosion of unalloyed titanium in hot, low-pH, chlorine-saturated NaCl brines It has been shown that unalloyed titanium anodized in a molybdate solution also provides crevice corrosion resistant metal surfaces (Ref 248)

Thermal Oxidation. Crevice corrosion resistance is afforded from thermal oxide films formed when titanium alloys are heated in air to 500 to 800 °C (930 to 1470 °F) for 2 to 10 min Increasing the temperature (or time) within this range results in thicker, more protective oxide films In hot NaCl brines, thermally oxidized titanium has proved its superior resistance to anodized or as-pickled titanium in both metal-to-metal and metal-to-gasket crevices (Ref 47, 129) Avoidance of mechanical damage or plastic metal strain is necessary for good protection Thermal oxide films may exhibit limitations at higher temperatures and in low-pH brine, and they are better considered for situations in which only borderline crevice conditions exist for unalloyed titanium

Noble Alloy Contact. Crevice corrosion of unalloyed titanium or other titanium alloys can be averted in metal-to-metal crevice situations If a more noble alloy is one member of the metal-to-metal crevice, the titanium metal surface can be anodically protected by the galvanic couple achieved in the crevice More noble metals include the precious metals, more resistant titanium alloys such as grade 7 or 12, or copper alloys For example, it is well known that unalloyed titanium tubes that are roll expanded into copper alloy tubesheets resist tube joint crevice corrosion in high-temperature seawater (Ref 41, 44) Similarly, the more noble grade 7 titanium alloy will protect unalloyed titanium with which it is in direct contact; this provides resistance similar to that of grade 7 titanium For example, a grade 7-grade 2 titanium sheet crevice will resist crevice attack in boiling chlorine-saturated NaCl brines, in which a grade 2-grade 2 joint would undergo crevice corrosion This principle also applies to treated titanium surfaces in contact with untreated unalloyed titanium

Surface Pickling. If the presence of galled or smeared surface iron is suspected on titanium equipment, it should removed to prevent possible pitting or hydrogen uptake in titanium exposed to hot chloride brines This surface iron contamination can be removed with a light ( 5 min) pickle in near-ambient temperature 35 vol% HNO3-5 vol% HF or 12-1 HNO3-HF solution, followed by water flushing This procedure will remove less than 0.03 mm (1.2 mils) from the titanium alloy surface Studies have shown that pickling in dilute HNO3-HF solutions is much more effective in removing surface iron contamination than either anodizing (Ref 47) or pure HNO3 exposures

Appendix 1: General Corrosion Data for Unalloyed Titanium

This appendix is a compilation of general corrosion rate values for unalloyed titanium (ASTM grades 1 to 4) These values were derived from various published sources (Ref 8, 26, 67, 80, 85, 86, 109, and 127) and from unpublished in-house laboratory tests These data should be used only as a guideline for alloy performance Rates may very depending on changes in medium chemistry, temperature, length of exposure, and other factors Also, total alloy suitability cannot be assumed from these values alone, because other forms of corrosion, such as localized attack, may be limiting The text should be consulted to assess overall alloy suitability more thoroughly for a given set of environmental conditions In

complex, variable, and/or dynamic environments, in situ testing may provide more reliable data In the following table,

temperatures are given only in centigrade, and corrosion rates are reported only in millimeters per year

Adipic acid 67 240 nil

Aluminum sulfate

Ammonium bicarbonate 50 100 nil

liquor

Aniline hydrochloride

3:1 80 0.86

Barium chloride

Benzene (traces of HCl)

Boric acid

Bromine gas, dry 21 Dissolves rapidly

possible)

99 Boiling 0.005

Carbon tetrachloride

Chlorine gas, wet

Chloracetic acid

Cyclohexylamine 100 Room nil

Liquid -196 0.011

Fluorine, HF free

Hydrochloric acid, aerated

Hydrochloric acid + 4% FeCl3 + 4% MgCl2 + Cl2 saturated 19 82 0.46

pH 1 5 66 0.152

Magnesium chloride

Magnesium hydroxide Saturated Room nil

<About 2% H 2 O Room Ignition sensitive

Nitric acid, red fuming

>About 2% H 2 O Room Not ignition

Nitric acid + 170 g/L NaNO3 and 2.9 g/L NaCl 27.4 Boiling 0.483-2.92

Perchloryl fluoride + 1% H2O