Investigation of long term exposure effects under stress on supersonic transport structural materials Phase I report, Jun. 1963 - Jul. 1964

N O R T H AMERICAN AVIATION, INC.

L O S A N G E L E S DIVISION I N T E R N A T I O N A L A I R P O R T LOS ANGELES 9, CALIFORNIA (PHASE I IEFOKC)

PREPARED BY

m

APPROVED BY, Research Lbboratories

No. o f Pages REVISIONS Date---.-----

PAGES AFFECTED REMARKS DATE REV. BY I 1

--

NA-64 -658 The appended report i s the result of studies carried out under Contract NASW '746 f o r NASA Headquarters, Washington D. C . It covers the first phase of L. 30,000 hours stress corrosion test and work done i n the period of June 1963 t o July 1964.

Investigatim of long term exposure effects under stress of T i t l e : supersonic transport structural materials.

Author: W. George Martin The report covers the first phase of a 30,000 hours exposure test.

Abstract:

Six materials - two titanium alloys, two F" steels and two super-

a l l o y s - were tested under stress i n five environmental conditions

comprising cambinations of salt and braze coatings and cozlstcmt and cyclic exposure a t 650 F. Braze coated titanium aUoys, salt coated T i t a n i u m 8A1-1V-LMo alloy and cyclically e,xposed AM 350 s t e e l were the only alloys subject t o stress corrosion failure w i t h i n the 15,000 hours exposure covered. I n t h i s period relative phase C-et occurredin the titaniun alloys without s i m f i c m t mechanical property changes and aging reactions occurred in both steels and the superalloys w i t h sppropriate mechanical propcrty changes. A hypothesis f o r corrosion reactions on t i t a n i u m alloys i n salt environments has been developed.

C orro s i on Stress Corrosion Titanium .4LLoys FH Steels Superalloys Long Term Exposure Brazinc U o y s Elevated Temperature Exposure INVESTIGATION OF Lx)NG 'J!ERM EXPOSURE E F F E C B UNDER STRESS ON SUPERSONIC " S P O R T STRUCTURAL ALLOYS Geo. k r t i n Prepared under C o n t r a c t No. NASw 746 by NORTH AMERICAN AVIA'ITON, INC.

Los A n g e l e s D i v i s i o n , Los A n g e l e s , C a l i f .

for NATIOrJI4L AERONAUTICS AND SPACE ADMII'?IS?RATION ABSTRACT The report covers the first phase of a 30,000 hours exposure t e s t . Six

materials - two t i t a n i u m a l l o y s , two PH steels and

two superalloys - were tested

under stress i n five environmental conditions c q r i s i n g cmbinations of salt and braze coatings and constant and cyclic exposure at 650 F. Braze coated t i t a n i u m alloys, salt coated T i t a n i u m 8Al-lV-lMo alloy and cyclically exposed AM 350 M s t e e l were the only alloys subject t o stress corrosion failure within the 15,000 hours exposure covered. I n this period relative phase changes occurred i n the titanium alloys without significant mechanical property changes a d aging reac- tions occurred i n both steels and the superalloys with appropriate mechanical property changes. A hypothesis for corrosion reactions on t i t a n i u m alloys i n salt environments has been developed.

ACKNOWLEDGEMENTS The work done i n this report was carried oiit i n the Research and Development Laboratories, North American Aviation, Inc., Los Angeles Division, which are under the general direction of N. KLinrmek. The electron microscopic work was carried out by G. C. Thomas, the microscopic work ‘by R . Brose and the X-ray dif- fraction studies by E. La.Rocca. Stress calculations and the equations and graphs plotted i n Appendix A were computed by L. Laclanan and R. M. A u t . The author i s also indebted t o .Mr. R. R a r i n g of NASA Headquarters f o r his advice and encourage- ment during the work.

ii TABU OF CONIIEPSTS Page No.

TI’I’IX PAGE i ABSTRACT ii ACH?a4IEDGEWIWS ii TABU OF COmNCS iii TqTST CIF FI-S iv LIST OF TABUS v i SUMMARY 1 II?rRomCT1oIq 2 SURVEY W PREVIOUS WORK ON STRESS CORROSION EX€lSCDEXCG PROCEDURE3 4 EXPOSURF: EZ’FECTS ON ME?JMUURGICAL STRUCTURE CRAmGEs IM MECHANICAL PROPERIIES CORROSION AND STRFSS COFN%ION EFFECTS 12

coNcLusIoNs rn R E C O ~ I O N 3 16

REFERENCES APFm4DIx A iii LIST OF FIGURES Page No.

Figures Dimensions of Exposure Spechens 44 2 Exposure Specimens in Test Frame 4 5 Exposure Test Frames in IXwnace System of Marking md Cutting Up of Exposure Specimens Dimensions of Nniature Tensile Specimens 4 8 6 Miniature Test Specimens Test R i g Components Miniature Test Specimen Test Rig Assembly with Extensometer 7 50 8 Microstructure of Titanium 6Al-4V Alloy Prior to Exposure 9 Microstructure of Titanium 6Al-4V Alloy After 1 0 , 0 0 0 and 15,000 Hours Exposure 10 Microstructure of Titanium 8Al-lV-lMo Alloy Prior To Exposure u: Mcrostructure of Titanium 8A.l-1V-lMoAlloy After 10,000 and 15,000 HOW E ~ ~ O S W 5 4 I 2 Microstructure of pH15-7M0 Steel Prior to Exposure Mcrostructure of PHl5-7Mo Steel After 10,000 and 15,000 Hours Exposure 14 Microstructure of AM 350 Steel Prior to Exposure 5 7 Microstructure of AM 350 Steel After 10,000 and 15,000 Hours Exposure 5 8 16 Microstructure of Inconel 7 1 8 Prior to Exposure Microstructure of Inconel 7 1 8 After 1 0 , 0 0 0 and 15,000 Hours Exposure 1 8 Microstructure of Rene' 41 Prior to Exposure Microstructure of Rene' 4 1 After 10,000 aad 15,000 Hours Exposure Mechanical Property Changes of Unstressed Titanium 6A1-4V Alloy 6 3 Mechanical Property Changes of Stressed Titanium W - 4 V Alloy 64 Mechanical Property Changes of Unstressed Titanium 8Al-lV-1MO W O Y 65 iv LIST O F FIGVRES (Continued) Figures Pa.ge No.

Mechanical Property Changes of Stressed Titanium 8Al-lV-lMo Alloy 66 Mechanical Property Changes of Unstressed PHl5-7Mo Steel 67 Mechanical Property Changes of Stressed P H 1 5 - m Steel Mechanical Property Changes of Unstressed AM 350 Steel Mechanical Property Changes of Stressed AM 350 Steel 28 Mechanicsl Property Changes of Unstressed Inconel 728 7 1 Mechanical Property Changes of Stressed Inconel 718 Mechanical Property Changes of Unstressed Rene' 41 lkchanical Property Changes of Stressed Rene' 41 3 1 Examples of Appearance of Exposure Specimens Swrmary of Stress Corrosion Failures Stress Corrosion Failures of T i t a n i u m 6Al-hV Alloy Stress Corrosion Failures of T i t a n i u m 6A1-4V Alloy Stress Corrosion Failures of T i t a n i u m 8Cu-lV-lMo A l l o y 36 79 Stress Corrosion Failures of T i t a n i u m 8Al-lV-1MO Alloy Stress Corrosion Failures of Tita.nium 8 U - l V - l M o Alloy Stress Corrosion Failures of T i t a n i u m 8Al-lV-lMo Alloy 40 Stress Corrosion Failures of AM 350 Steel 41 Stress Corrosion Failures of AM 350 S t e e l 84 Free Energy C h a n g e of Titanium Corrosion Reaction Electron Microprobe Photographs of Titanium 8Al-lV-1MO Alloy 44 Electron Microprobe Photographs of Titanium 6Al-4V Deflection and Moment Arm Versus Load Stif'fness Parameters 88 Moment Coefficients Versus Load Stiffness Paraneters f o r Inter- mediate Beam Positions 89 V LIST O F TABLES Tables Pa= no.

Analysis and Froperties of Candidate Materials I Summary of Heat Treatments 22 I1 I11 Summary of Surface Trea-nts 23 I v Smmaqy of Exposure Stress Levels V Stress Levels of Exposure Specimens of Titanium 6A1-4V U o y 25 VI Stress Levels of Exposure Speclmens of Titanium 8Al-lV-lMo Alloy VI1 Stress Levels of Exposure Speclmens of PHl5-7Mo Steel 27 VIII Stress Levels of Exposure Specimens of AM 350 Steel Ix Stress Levels of Exposure Specimens of Inconel 718 29 X Stress Levels of Exposure Specimens of Rem' 41 Mechanical Properties of T i t a n i u m 6Al-4V Alloy XI X I 1 Mechanicsl Properties of Titanium 8A1-1V-lMo Alloy Mechanical Properties of PHL5-740 Steel XI11 X I V Mechanical Properties of AM 350 S t e e l xv Mechanical Properties of Inconel 7 l . 8 X V I Mechanical Properties of R e n e t 41 XVII Surface Appearance of Exposure S p e c w n s After 2800 Hours XVIII Surface Appearance of Exposure Specimens &er 4700 H o u r s

x r x Surface Appearance of Exposure Specimens after 8900 Hours

4-0

xx Surface Appearance of Exposure Specimens after l5,OOO Hours

XXI

Analysis of S a l t C o a t i n g - X-Ray Diffraction Spacings

XXII Material Properties at 650 F XXIU Calculation of Effective Beam Lengths 43 v i This report describes t h e results of the first phase of an investigation of the metallurgical and mechanical property changes occurring i n six candidate materials for supersonic transport vehicles on exposure t o 650 F i n the stressed and unstressed condition under various surface environmental coatings.

The first phase consisted i n the examination of specimens exposed for 10,OOO and 15,000 hours and specimens which have failed by stress corrosion within that period.

six candidate alloys tested were two t i t a n i u m alloys, two precipitation The hardening s t e e l s and two superalloys. The specimens, i n t h e shape of cantilever s t r i p s loaded t o s t r e s s levels varying from 23s t o go$ of the yield stress, were exposed t o a temperature of 6509, Surface treatments consisted of salt coat- i n g s , braze coatings and salt coated braze coating and cyclic exposure a t 6509’ and a humidity cabinet A f t e r a period up t o 15,000 hours, most of the braze coated and braze coated salt coated t i t a n i u m alloys have fractured by stress corrosion. In addition, plus two out of six salt coated titanium 8-1-1 alloy specimens and two out of six of salt coated AM 350 M alloy specimens have failed. Four out of four AM 350 M alloy specimens exposed t o a cyclic environment of furnace temperature and humidity cabinet failed, a l l a f t e r approximtely 3,000 hours.

I The mechanical t e s t s indicate that a l l alloys except the steels are affected by the braze alloys. The t i t a n i u m alloys are strongly affected. Steels and superalloys undergo aging reactions during the exposure period, which affects both the strength levels and t h e ductility. Stress during exposure did not appear t o affect any of the changes i n mechanical properties observed. No gross changes i n metallurgical structure of any of the alloys could be observed. X-ray i n the relative amounts of diffraction studies, however, do indicate changes alpha and beta phase a f t e r exposure of t h e t i t a n i u m alloys.

of a typical intercrystalline nature i n both Stress corrosion fractures are the t i t a n i u m alloys and the AM 350 alloys. Neither microscopic nor electron microscopic replica studies have so far indicated any structural change directly related t o the corrosion mechanism. However, examination of fractured t i t a n i u m alloy specimens by means of a n electron microprobe indicate evidence of segre- gation of heavy alloying constituents near t h e crack.

The corrosion products on titanium specimens exposed under coatings of natural and a r t i f i c i a l sea salt have been examined by X-ray diffraction and there appears t o be tentative evidence for the existence of N a O H formed during the exposure process, Thermodynamic calculations show the f e a s i b i l i t y of sev- eral reactions resulting i n t h i s product.

The Institutions responsible for m t e r i a l selection for hi-speed a i r c r a f t designed for long t i m e service, such as the National Aeronautical and Space Administration, the Federal Aviation Agency and the Aerospace Industry,have for considerable t i m s been keenly aware of the problems involved i n t h e selection of mterials f o r such a i r c r a f t . Materials required for such a i r c r a f t are either radically new families which have t o be developed and evaluated or they consist of known m t e r i a l s exposed t o a new type of environment. Materials l i k e l y t o be of prime importance are titanium alloys, hi-strength s t e e l s and superalloys.

Most of those materials must be heat treated t o develop suitable properties.

However, such properties can only be obtained i n w h a t are basically meta-stable m t a l l u r g l c a l structures. It is a matter of concern whether long ti= exposure t o elevated temperatures w i t h or without stress would lead t o changes of such meta-stable structures and therefore t o changes i n t h e wchanical properties.

I n addition t o t h i s standard type of information,such as the s t r e n g t h of the various temperature levels, creep data and fatigue information, a very high depee of assurance is required that the m t e r i a l s chosen w i l l not be subject t o a sudden type of failure such as stress corrosion. The environments likely t o lead t o s t r e s s corrosion are braze coatings, which may have been used for joining and particularly the possibility of sea salt incrustation covering the external surfaces of the aircraft. It is e s t i m t e d that the type of a i r c r a f t considered here may be exposed on the external surfaces t o temperatures up t o 650 F during the service l i f e i n excess of 30,000 hours. Complex interactions between the coatings, the temperature and any structural changes appearing i n t h e m t e r i a l after long time elevated temperature exposure can therefore be expected.

A general program t o determine the likelihood of a s t r e s s corrosion f a i l u r e i n appropriate candidate materials was started by North American Aviation, Inc., i n 1962. The present program consists of a comprehensive evaluation of speci- mens w i t h a failed by stress corrosion or remained exposed t o periods of 10,OOO and l5,ooO hours. This program was aimed a t t h e determination of any possible degradation i n mechanical properties,at a search for evidence of possible changes i n t h e metallurgical structures and a t an evaluation of a possible reaction involved i n s t r e s s erosion. The specimens exposed under t h e program organized by North American Aviation, Inc., were t o form the basis of information. Specific t e s t s carried out include notched and unnotched tensile t e s t s , microscopic examination, examination of the metallurgical structure by electron-microscope replication, X-ray diffraction studies of possible phase changes, and also examination of surface products with a view t o a better under- standing of corrosion and stress corrosion mchanisms. The entire program is planned t o investigate the effects on materials of a t o t a l exposure time of 30,OOO hours .

The second phase This report covere the first phase of the investigation.

w i l l deal with 20,OOO hour exposure effects and a f i n a l t h i r d phase w i l l deal with materials after 30,OOO exposure. The design of the experiment, that 1s the number and type of specimens chosen, is such t h a t a t the completion Of 30,000 hour i n v e s t i g a t i o n , all test results w i l l be available i n duplicate.

SURVEY OF F'REVIOUS WORK OH s m s s CORROSIOB The susceptibility of t i t a n i u m alloys t o stress corrosion cracking when i n intimate contact w i t h sodium chloride a t elevated temperatures has been known for many years. The limits of the corrosion reaction were not known nor was the corrosion mechanism established, although many theories have been proposed.

Evidence (reference 1) had established that t i t a n i u m alloys are subject t o stress corrosion cracking when in intimate contact w i t h sodium chloride at temp- eratures above 5 0 0 F . No service failures were reported ( t o mid-1957) which were attributed t o t h i s type of corrosion. Laboratory t e s t s had shown that various coatings (oxide films, anodic films, aluminum and nickel metallic coatings) would mitigate t h i s type of corrosion. The limits of the corrosion reaction were not known nor was the corrosion mechanism established,although various theories have been proposed. Further studies were recommended t o identify the corrosion product, establish differences between types of titanium alloys, and the effect of s a l t concentration and thickness of salt coatings. A theory of the corrosion mechanism was advanced which suggests t h a t t i t a n i u m i n the presence of oxygen and a reducible chloride forms Tic%. Sodium chloride was established as a crack nucleating agent. It was shown that moving air across the specimen surfaces dur- ing corrosion exposure increases the resistance of the material t o s t r e s s corro- sion. Glass bead peening and a sodium hydroxide anodizing t r e a t m n t were demon-

I O

strated t o afford protection against stress corrosion of T i 6A1-4V alloy.

A p a r t from salt, three other environments were found t o stress corrode t i t a n - ium; (reference 2) these were molten cadmium, red fuming n i t r i c acid (RFNA), and hydrochloric acid formed b y the decomposition of a chlorinated diphenyl compound i n air a t 600 F. The molten cadmium corrosion occurred on a T i 4Mo-4Al alloy i n contact w i t h a cadmium-plated bolt a t 600-750 F. Stress corrosion cracking of T i 91-2.4 Sn alloy was found t o take place i n the presence of halides (trichloro- ethylene) during heat treatment at temperatures of 11% F and 1500 F for 16 hours (reference 3 ) . Severe cracking, other than stress corrosion cracking, was also found t o occur w i t h s t r e s s present if a surface oxide cuating was present. This e f f e c t accentuated any difficulty encountered by halide contamination.

T1 6Al-kV and T I . 8Al-lMo-lV were incapable of withstanding an exposure of D i s - 25,000 psi a t 650 F f o r lo00 hours, but T i 8-1-1 did not f a i l at 450 F .

cusses possible corrosion mechanisms involving gaseous chlorine attach and also galvanic corrosion. Tests conducted t o i n v e s t i g a t e electro-chemical corrosion mchamism. salt environ- Exposure of notch titanium specimens i n a 650 F sea ment prior t o stressing at 25 KSI was found t o prolong specimen l i f e (reference 4). Stress corrosion cracks were found i n T i 8Al-lMo-lV specimens i n which Bea s a l t had been packed i n t o a notch consisting of a 1/16 inch diamter hole in the center of a sheet specimn and the specimens then exposed a t 650 F and stresses of 25 and 63 X S I for 50 and 200 hours, respectively, and a t 800 F and 25 K S I for

100 hours (reference 5) . Mterials Research Laboratory reported evidence that

Ti 6A1-4V alloy s t r e s s corrosion i n hot s a l t is electrolytic i n nature, with the

I )

titanium becoming anodic t o chloride ions i n a postulated t h i n film of eutectic '3 J or low-melting salts. In t h i s hypothesis free chlorine does not enter directly i n t o the corrosion mechanism and is not essential t o it. Titanium alloys T i 6A1-4V and T i 8 A l - M o - l V exhibit stress corrosion failures a t 700 F and above, but not a t 600 F (reference 6).

As fax as other SST alloys are concerned, data are less conflicting t h a n those for t i t a n i u m .

The superalloys (Inconel W and cobalt-base v-36 alloy) were unaffected by heavy coatings of dry sea salt when exposed a t b , O O O p s i for lo00 hours a t 650 F and 850 F (Inconel W only). AM 350, C R m w a s also unaffected a t 650 F (reference 4) Similar specimens of AM 350 CR and PHl5-W RH1050 did not exhibit any crack- ing a f t e r exposures at 650 F and 800 F and stresses of 40 and 70 KSI for up t o lo00 hours, and a t 650 F and 100 KSI stress for lo00 hours (reference 5). Doughs Aircraft Company t e s t s show that AM 350 SCT 850 is quite susceptible t o stress corrosion under beach exposure and s a l t spray testing.

PHl5-7Mo R H l l O O is appar- ently not susceptible t o stress corrosion testing a t room temperature but becomes susceptible a f t e r 10oO hours exposure a t 650 F.

Superalloys have not sham any stresa corrosion defects through lo00 hour alternate Fcrmersion and 100 hour beach exposure. Lockheed Aircraft Corporation reported that AM 350 SCT 850 w i t h an i n mrine environ- "optimum chemistry" showed g o d resistance t o s t r e s s corrosion mnt, but that AM 350 SCT 850 w i t h a n "unfavorable chemistry" was considered r e l a t - ively susceptible (reference 6).

Six materials were chosen f o r investigation.

Two titanium alloys, Titanium 6A-4V and T i 8Al-lMo-lV appeared t o be the mst likely titanium alloy candidate m t e r i a l . PHl5-m s t e e l and AM 350 s t e e l were selected t o represent the whole family of age hardening meta-stable austentite stainless steels. Superalloys selected were Rene' 41 and Inconel n8. A n a l y s i s of the various alloys as well as their heat t r e a t condition and mechanical properties a s deterinined by standard tensile specinren are s h a m i n Table 1.

It should be noted that T i 8A1-1V-lMo used for the experiment had been subjected t o a single solution treatment because a t the time of commencement of the investigation, the advantages of the duplex annealing process of t h a t alloy had n o t yet been discovered.

Materials selected were i n the form of sheet ranging i n thickness from -020 t o .OW inches. From these sheets, s t r i p s approximtely 12 inches long and a n inch and one-half wide were cut i n I 2 inch length i n the longitudinal rolling direction. The details of the heat treatment used is shown i n Table 11. A l l brazing and heat treatment i n t i t a n i u m alloys was carried out i n r e t o r t s f i l l e d w i t h argon i n order t o minimize contamination. Subsequent t o heat treatment, the specimens were machined t o dimensions s h a m i n Figure 1. Two types of S P C i - mens were prepared. One type of specimen was exposed i n the plain sheet form while another specimen was notched i n a direction a t right angle of the Principal axis a t two points: one point of s l o t t i n g was close t o the support area while the other point was near t h e loading points. A l l specimens were suPPOrted a t one end in a s t a i n l e s s s t e e l frame, clamped between mica-strips for insulation, and loaded on the other end t o produce a cantilever type specimen.

This canti- lever type specimen was chosen i n preference t o the usual constant s t r a i n U-type stress corrosion specimen. It Xa6 considered possible that on the long exposure periods some relaxation i n a constant strain type specimen may take place and thus change the stress level. Another advantage of the cantilever type specimen is that a large number of specimens can be accomodated in a limited space. A t e s t involving direct tensile loading would have required an extensive and expen- sive set up, i n order t o produce the required stress levels.

Cantilever specimen allows direct comparison of s t r e s s corrosion effects and corrosion effects, a s one end of t h e specimen is under a condition of maximum s t r e s s and the other end is virtually unstressed. The assembly of the specimens i n a test frame is indi- cated i n Figure 2 , and Figure 3 shows the loaded frame placed i n a f”urnace.

A t o t a l of six frames each accommodating 24 specimens were available. The heating device chosen w m an air circulation type furance equipped w i t h dual control.

The a i r circulation f’urnace i s run ConstaEtly, except for two periods of break- downs. During those breakdown periods, the specimens remained untouched and loaded, but a t room temperature.

Each of the six materials tested was exposed with a variety of five different surface conditions or surface treatments. These conditions are summarized i n Table 111. Prior t o a l l surface treatments s t r i c t attention was paid t o cleanli- ness of specimen surfaces. A l l surface grease and s t a i n s had been carefully removed by a degreasing and pickling treatment and no handling of specimns a f t e r t h i s treatment w i t h bare hands was permitted. The surface treatments were applied as follows : Exposure i n as received condition except for surface cleaning and pickling treatments.

Specimens prepared as above, and subsequently coated by brushing with a suspension i n water of synthetic sea salt comprising 6 This parts sodium chloride and one part magnesium chloride.

suspension was brushed on and a f t e r drying resulted i n an even coating of approximtely 1/32 of an inch thickness.

Specimens prepared as (a) above and one coated with a coating .001to .003 inches thick of a braze material considered suit- able and likely at the time of the commencement of t e s t . Braze coatings were selected from the following brazing alloys : Titanium Alloys : Dynabraze B. (94.q Silver, 5$ Aluminum, 0 . 2 $ Manganese) Steel and Superalloys : Prembraze 130 (72$ Gold, 6% Chromium, 22$ Bickel) Specimens prepared as (c) above but coated subsequently w i t h a s a l t coating as under (b) Specinens prepared as under (a) above but exposed cyclically by (e) mintaining them i n the furnace a t exposure temperature for a fortnight, then removing the frame into a humidity cabinet for exposure i n water saturated air a t 100 F for a fortnight, followed by return t o the furnace for a fortnight’s exposure and so on.

Suspension weights were machined from stainless s t e e l and connected through I n the specimens by mans of stainless s t e e l wire and smll insulating bead.

t h i s Iwnner, accidental electric contact between weights frame and different specimens was minimized. During exposure i n the furnace the a i r circulation caused a small oscillators movement of a l l specimens. This was not considered significant, as t h i s small mvernent extending t o perhaps .Ow inches on either of the equilibrium conditions would not produce significant changes i n the stress levels. The frame taken out for cyclic treatments a t fortnightly intervals was handled as gently as possible, but a certain amount of joggling could not be avoided. Furthermore, it w a s found t h a t a f t e r the two years exposures, the speci- mens became coated w i t h a certain a m o u n t of dust, particularly brick dust from furnace flues. These vibrations, jolts, and dust coatings therefore do constitute a n unknown environmental factor. However, it is n o t considered t h a t t h i s factor was very s i g i f i c a n t . In addition t o the regular thermocouples forming part of the furnace equipment and uperating controlling mchanisms, temperature checks were carried out a t three separate instances. Fur the purpose of these tempera- ture checks, six thermocouples were distributed a t various points inside the f u r - nace and the temperatures measured by mans of a potentiomter. I n a l l cases, the thermal temperature variation i n t h e furnace was found t o be plus 0 minus 2 0 A s l i g h t tempera- degrees as that indicated by the regular f’urnace thermometer.

ture drop was indicated near t h e furnace door, t h i s leak could not be sealed Cam- pletely. Furnace atmosphere can thus be assumed t o be f a i r l y s t a t i c , although a certain admixture of fresh air did take place.

The size of the weight loading the specimens was selected such t h a t two series of stress levels w a s obtained.

In the plain specimens the lllaxislum s t r e s s level was 25 t o 30 percent approxirnztely, i n the notched specimens a stress level approaching the yield stress was achieved, theoretically a t least, a t the bottom of the notch. Stress levels are summarized i n Table IV for each type of m t e r i a l , and for each type of exposure, both for the notched and the unnotched specimens.

Actually stress levels i n individual specimens had t o be calculated w i n g the theory of beam with large deflections. The general method of calculation and equations involved are given i n A p p e n d i x A. A computed program was developed t o allow t h e calculation for each specimen for each level arm length and for each specimen thickness. The results of these calculations are given i n Tables V t o X for each specimen. The Tables also show the s t r e s s concentration factors which were assumed for the notched specimn. Appendix A also gives a method of calcu- lation of stress levels at position intermdiate between the point of load support and the point of specimen support.

A t fracture or a t any indication of the specimen obtaining a permanent bent both the specimen and portion held w i t h i n the clamp were removed from the furnace and stored for further examination. These specimns together with One specimen for each material and s e t of surface conditions removed a f t e r 10,OO~ a n d 15,000 hours were then cut up for further examination. Details of method Of sectioning of specimens is s h a m i n Figure 4. The pieces of both ends are reserved for microscopic, electro-microscopic and X-ray investigations. The section adhcent t o the edge of the supporting beam and adjacent t o the hold carrying the weight was used for notched and unnotched tensile specinen testing. Each specimen pro- vided a t o t a l of two notched and t w o unnotched tensile test specimns.

As the original t e s t s t r i p s were a l l cut i n the longitudinal rolling direction, the ten- s i l e specimens therefore represent the transverse properties.

Considerable trouble was experienced i n the design of a suitable tensile specimen. This was due t o the fact that the t o t a l tensile specimen lengths were limited by the one and one-half inch width of the exposure s t r i p , while extenso- m t e r w i t h t h e shortest available guage length required a one-half inch length.

One purpose of cutting tensile specimens i n the transverse direction as indicated was t o assure t h a t the stress distribution across the tensile specimen would be Had the t.ensile q e c i m e n s been cut the same direction as the v i r b a l l y constalzt.

s t r i p lengths, then the stress during exposure would have varied appreciably across the tensile specimen gauge length. A rider of specimen design configurations and specimen holding grips were t r i e d and discarded a f t e r it was found that specimens either tended t o s l i p or break i n the grip. It was essential t o use a pinned type tensile specimen, because the rough surface,after the exposure, made friction grips quite unreliable. The f i n a l specimen configurations are shown i n Figure 5 , and the types of grips are sham i n Figure 6. A complete s e t up showing specimen, grips, and extensomter is shown i n Figure 7.

EXPOSURE EFFECTS ON ME;TALLWICICAL S'IRWTURE Metallurgical Examinations Cross sections of a11 specimens were examined just prior t o exposure and a f t e r 10,OOO and l5,ooO hour exposure or failure. During examination particular atten- I n the case of specimens which tion was paid t o the top surface of the specimens.

had f a i l e d due t o stress corrosion failure, sections were also taken i n the plane of the specimen across t h e cracked zone. After completion of microscopic examina- of two stage r e p l s tion, electron microscopic e x a n a t i o n was carried out by means cas. Replicas were prepared from the etched nicro-specimen surface i n the usual mnner. A collodion replica was mde of the surface which w a s then shadowed with carbon. After dissolving away the collodion, the carbon copy was placed on the A l l microscopic specimen holder of a H i t a d x i HU-11 electronmicroscope and examined.

examinations were carried out a t a mgnification of 500, and electron-microscopic examinations were reproduced at mgnifications of 2,500 and 15,000 times.

Titanium 6A1-4V Alloy The micro- The structure of the alloy prior t o exposure is shown i n Figure 8.

There is very l i t t l e structure shows the typical alpha beta phase distribution.

Some of the grain boundaries near the surface

evidence of surface contamination. 0

I in photomicrograph are somewhat heavier t h a n i n the body of the material. The structure is resolved further i n theelectronmicroscopes shown below the photo- micrograph. 10,OOO and 15,000 hours is shown The metallurgical structure a f t e r i n Figure 9. There is definite evidence on t h i s photomicrograph of the grain s i z e and the grain shapes. I n the 15,000 hour case some coagulation of the l i g h t etching phase i n t o larger g r a i n s has resulted. This observation is borne out by the electron micrographs which quite distinctly show coagulation of the l i g h t etching phase after the l5,OoO hour exposure.

Titanium 8A1-U4o-l.A Alloy Figure 10 indicates t h e structure prior t o exposure. Comparing t h i s structure t o the structure of a n uncoated specimen exposed t o l5,OOO hours, sharn i n Figure 11, it appears that there is no significant change i n the structure, which con- sists primarily of the alpha phase.

pHl.5-7Mo Steel Neither photosnicmgraphs nor electron-micrographs indicate any apparent changes i n the metallurgical structure. Figure 12 shows the structure prior t o The exposure. Figure 13 structures a f t e r 10,OOO and l5,OOO hours of exposure.

There distribution of austenite and Martensite phases appears t o be unchanged.

is no indication of any widening of g r a i n boundaries nor is there any indication of the appearance or disappearance of any precipitates. There is, however, a i n the h r t e n s i t e structure on exposure. A slight indication of possible change high magnification electron-ndcrograph shows a f a i r l y coarse Martensite structure prior t o exposure which after 10,OOO and l5,ooO hours is progressively refined.

AM 350 M Steel Figure 14 shows AM 350 s t e e l exhibited definite change i n the f i n e structure.

unexposed structure which can compare t o Figure 15 showing the l5,OOO hours struc- ture. A most significant change is a loss of fine structure i n the Mnrtensite g r a i n s . There is no evidence of austenite stringers i n any way disappearing or interfering with stress corrosion cracking. Electron-micrograph indicates same absorption of intergrannular precipitates after exposure.

Inconel 718 Alloy No major changes i n the structure of Inconel 718 alloy a f t e r exposme are indicated, as can be seen by comparison of Figures 16 and 17. There is,however, a slight indication of the absorption of sow of the intergrannular precipitates both from the photomicrographs and electro-micrographs and also an indication Of a coarsening of the precipitation hardening phase. There is some indication Of

a

of precipitate coagulating i n the grain boundary areas, particulary as indicated by the electron-micrographs Rene’ 41 Alloy Microstructurally Rene’ 4 1 shows no g r o s s structural changes a f t e r exposure up t o l5,OOO hours. The surface structure effect i n t h i s alloy, where the g r a i n boundaries tend t o disappear near a f’ree surface, is typical of t h i s alloy and found i n a l l specians. Electromicrograph show an absorption of a precipitate phase a f t e r exposure both in 2,500 and 15,000 magnification electron-micrographs.

There also appears t o be a n absorption of g r a i n boundary precipitates on pro- longed exposure.

X-Ray Diffraction Analysis Samples of a l l specimens were examined by X-ray diffraction in the unexposed In the case of the titan- condition and a f t e r 10,OOO and 15,000 hours exposure.

i u m 6A1-4V alloy, diffraction patterns indicate that the beta phase is retained up t o 10,OOO hours and then tends t o dlminish. I n t i t a n i u m &Al-lV-lMo alloy, however, a p p r o x h t e l y 50$ of the original amount of beta phase is removed i n t h e first 1 0 , O O O hours, a f t e r which the amount of beta tends t o stabilize. In PHl5- 7Mo s t e e l the amount of retained austenite tends t o diminish progressively, but small does not disappear a f t e r 15,000 hours exposure. AM 350 M s t e e l contains a amount of retained austenite only, which appears t o be stable. I’?o changes were noticed i n the superalloys. The diffraction patterns obtained allow only a qualitative comparison of the various amounts of phases existing. Work was commenced on the determination of the actual c e l l sizes, from which more absolute quantitative data could be obtained, but t h i s work has not been completed.

C W G E S IN ME-CAL FRoPmms Mechanical properties as exemplified by the ultimate strength, the yiel s t r e n g t h and the elongation for the case of unnotched tensile t e s t specimens and by t h e ultimate tensile strength i n the case of notched tensile specimns were determined on all s p e c m n s removed from the t e s t a f t e r 10,ooO hours and a f t e r l5,OOO hours and on a l l specimens which fractured prior t o the l5,OOO hour period, Each exposure specimen thus yielded two tensile i n a manner described above, one f’romthe stressed and one fromthe unstressed portion of exposure specimens, specimen. Data are therefore given for the unnotched and notched tensile pruper- For t h e purpose of analysis, each t i e s f o r t h e stressed and unstressed condition.

material will be considered separately.

When considering the tensile data, the fact that a l l spechens tested are cut i n the direction transverse t o the direction of rolling and the curve of t h e A p a r t from the anisotropy due t o rolling direct- specimen must be borne i n mind.

ion, the test direction employed eliminates the effect on strength of any minute Due t o the bend curvature of the exposure specimen edge cracks due t o corrosion.

such cracks are most likely t o be i n the direction transverse t o the length of the exposure specimens, i.e. parallel t o the direction of tensile load i n the tensile test specimens. A clear distinction can therefore be mde between the effects of uncontrolled and un-masurable corrosion surface deterioration a.nd true nnsterial property changes. It is f e l t that i n materials subject t o excess- ive corrosion, stress corrosion failures would occur w i t h i n the t e s t period and thus point up the need for a protective coating.

A comparison has been made between the results obtained from the miniature specimerseqloyed and standard tensile t e s t specimens. As the data i n tables show, there i s very l i t t l e difference i n t h e t e s t results, and the t e s t results obtained on the miniature specimens can therefore be considered t o be quite representative.

In the case of braze coated specimens, no special allowance has been made for the lower strength of the thin layers of braze alloy. Braze alloy thickness is of the order of 0.001 inch t o 0.002 inch maximum and a l l braze coated specimens have a thickness of 0.040 t o 0.50 inches. Braze alloy strength is of the order

of 30,OOO psi. The t o t a l error introduced by - not considering the f a c t that braze

coated specimens do in f a c t represent a composite beam i n tension i s therefore of the order of approximately 2$ only.

!titanium 6A1-4V Alloy i n Table X I and shown diagamatically i n Figures 20 Test results are given and 21. There does not appear t o be any significant difference i n the propert- ies of the material exposed i n the stressed and the unstressed condition.

Strength levels of materials exposed without braze coatings do not vary signifi- cantly except for the case of 10,OOO hours exposure of the stressed specimens, which appear t o have a lower strength i n the as treated surface condition. There is an indication of a n increase i n the d u c t i l i t y after exposure as shown by the elongation. Braze coated specimens, exposed both with and without a salt coating, have a significantly lower strength and ductility, although there is considerable scatter of data.

The notched/unnotched tensile strength r a t i o appears t o be l i t t l e affected by exposure time and remains above unity f o r the type of notched specimen used.

Titanium 8Al-lV-LMo Alloy This alloy exhibits a stress behavior similar t o the other titanium alloy above, except that the salt coated specimens, too, exhibit a loss i n strength on exposure. Ductility losses on exposure appear t o be less pronounced t h a n those of t i t a n i u m 6A1-4V alloy. The enibrittling effect of braze coating is Consider- able and appears t o increase with t i m .

Both notched and unnotched tensile T e s t strength are similar,indicating no significant change i n notch toughness- data are given i n Table XI1 and are summrized i n Figures 22 and 23.

I

PHl5-7Mo Steel Relevant data are s h a m i n Table X I 1 1 and illustrated i n Figures 24 and 25.

The mechanical property determinations of the alloy indicate quite clearly that the aging process is continuing during the first l5,ooO hours exposure.

Both stressed and unstressed specimens exhibit similar behavior.

The unnotched t e n - s i l e properties show that there appears t o be an aging peak after approximately 10,OOO hours exposure. However, the f a c t that both the cyclically and the con- tinuously exposed specimens show similar properties indicates that the aging peak occurs somwhere prior t o the 10,000 hour exposure level, because the cyc- l i c a l l y exposed specimens actually only s p e n t 5,000 hours a t elevated temperature.

The aging peak is accompanied by a loss i n d u c t i l i t y and a loss i n notched t e n - s i l e strength, as compared w i t h the ultimate strength of the unnotched specimns.

This decrease i n the notched/unnotched tensile r a t i o is improved after prolonged c q c s u r e t o 15,OW h m m , hut does not reach the r a t i o of the unexposed material.

The effect of the aging process on mechanical properties can be considered t o be quite significant from a design point of view. There is also a considerable spread of t e s t results, especially i n the case of notched tensile date after 10,OOO hours, for t h e different surface treatments.

AM 350 M Steel The results of the t e s t data are shown i n Table XIV and i n Figures 26 and 2 7 .

Like t h e P H l 5 - m steel, t h i s mterial too undergoes an overaging process during the 15,000 hours exposure period. The spread of the t e s t results, however, is wide and it i s more d i f f i c u l t t o draw definite conclusions a t t h i s stage from the available data. It appears that the aging peak occurs somwhere before 10,000 hours i n the braze coated specimns, but is i n excess of l5,OOO hours i n the other specimens. Notched tensile data, too, indicate a braze coating effect. The possi- b i l i t y of a component of t h e braze alloy diffusing into the s t e e l and changing This m t e r i a l did t h e precipitation reaction can therefore not be excluded.

exhibit stress corrosion failures and the test data on the failed specimens are sorraewhat lower than on the spechens which have not failed, particularly with regard t o ductility. Notched/unnotched tensile ratios are around unity i n a l l cases, except for the case of braze coated specinens after 10,OOO hours exposure, where they are significantly below unity.

Inconel 718 The test data given i n Table XV and drawn i n Figures 28 and 29 show that t h i s nnzterial undergoes an aging process, without, however, reaching a strength peak.

The aging process appears t o be stress insensitive. There is some spread of data, particularly a t the 10,000 hour level, with the braze coated specimen persistently

showing the lowest results . The notched/unnotched tensile r a t i o remains around

unity for a l l conditions and exposure periods.

Rene' 41 Several observations T e s t data are shown i n Table XVI and Figures 30 and 31.

appear t o be significant. The material undergoes an aging reaction w i t h a peak strength somewhere before the 10,OOO period. This aging reaction is not connected w i t h a reduction i n ductility or notched strength, both of which appear t o increase on the average. There is no stress sensitivity i n any pro- perty. The aging reaction appears t o be affected considerably by the braze coating. The notched/unnotched tensile r a t i o is considerably below unity f o r a l l cases of surface treatment and exposures, being worst a t the aging peak and approaching unity a f t e r prolonged exposure.

CORROSION AlJD SlRESS CQRROSION EFFECTS General Surface Corrosion Effects Specimens exposed i n the furnace atmosphere were examined several times each Specimen frames (except the one carrying speci- week for fractures and bending.

E n s undergoing cyclic exposure) were removed from the f’urnace for a thorough examination at five intervals only, a f t e r 2800 hours, 4,700 hours, 8,900 hours, 10,OOO hours and l5,ooO hours exposure. !I!he surface appearance of the various specimens f o r the different surface treatments is sumolarized i n Tables XVII t o XX. The surface appearances a f t e r 10,OOO hours were substantially similar to the appearance after 8900 hours exposure and details on examination a f t e r 10,OOO hours have therefore n o t been reproduced separately.

A photograph of typical surface appearances is shown i n Figure 32. The photo- graph indicates the main observations rnade on the various surfaces: Titanium alloys without braze coatings form white areas on a pre- dominantly blackground. These areas are either i n spots or stringers.

Braze coatings on titanium alloys tends t o flake off completely a f t e r even a few thousand hours exposure. The surface beneath the flakes i s coarse crystalline i n appearance and very rough.

Precipitation headening steels are attacked i f salt coatings are present. Particularly heavy was the attack on AM 350 M exposed under cyclic conditions.

Uncoated specimns suffer discoloration only.

Superalloys exhibited the greatest resistance t o corrosion under all conditions.

Stress Corrosion Fractures O f t h e six materials exposed, only three m t e r i a l s have so far shown any These are the two t i t a n i u m a l l o y s and AM 350. evidence of s t r e s s corrosion.

I n t i t a n i u m 6~1-4V Figure 33 summarizes the t o t a l number of failures found.

No failures have alloy only the braze coated specimens were prone t o failure.

been found in s a l t coated or cyclically exposed specimens of t h i s alloy.

Titanium 8-1-1 alloy is likewise prone t o s t r e s s corrosion failure after braze

l o

coating, with or without s a l t ; and failures of t h i s alloy have also been found i n two s a l t coated specimens.

Failure of AM 350 s t e e l a f t e r cyclic exposure occurred following significantly closely-related failure periods.

O f four speci- mens exposed, all failed a f t e r 3,000 hours + lo$. None of the other materials

-

have so f a r failed i n s t r e s s corrosion.

- Stress Corrosion Mechanism

I n addition t o a study of direct stress corrosion failure data and the changes i n mechanical and metallurgical properties on exposure, the present study also aims a t an attempt t o obtain more information on t h e stress corrosion mechanisms involved. To further t h i s aim, both vertical and horizontal sections thrcagh cracked specimens were prepared and examined by means of both the con- ventional microscope and the electron microscope, using replica techniques for the latter.

These studies w e r e supported by X-ray diffraction studies of t h e corrosion products, particularly on t i t a n i u m alloys, and by electron microprdbe studies of the areas immediately adjacent t o the cracks.

The results of t h e metallographic investigations are shown i n Figures 34 t o 41. Very strong evidence of surface corrosion is exhibited i n the braze coated specimens. Figure 34 shows the cross section through two braze coated specimens of t i t a n i u m 6 A l - 4 V alloy which failed a f t e r 7,124 hours and l5,OOO hours expos- ure respectively. A photomicrograph showsevidence of a n interdiffusion zone between the braze coating and complete disintegration of t h e grains below t h a t braze coating diff'usion interface. O f particular interest is the electron micro- of t h e specimen which failed a f t e r 7,124 hours. This replica shows scope replica the crack preceding through the alpha-beta grain boundary i n most casegbut there are at l e a s t two incidences where the crack traverses a g r a i n . The shape of the alpha and beta grains appear t o be somwhat altered, possibly due t o the results of a diff'usion reaction. It should be noted t h a t i n t h i s electron microscope replica the border on the top l e f t hand corner is not the specimen surface, but the shadow of the specimen holding grid i n the electron microscope. Figure 35 is a microstructure of the specimen which failed a f t e r 15,000 hours exposure, sec- tioned i n a plane parallel t o the specimen's surface. The fracture is a very typical stress corrosion branch-type of fracture proceeding intergranularly, thrU the material. The bottom right hand corner of the specimen indicates cracked zones which are not connected t o the surface.

Titanium 8Al-lV-lMo shows a similar effect t o the Titanium 6A1-4V alloy.

Figures 36 t o 38 show the tendency of cracks t o mve through the heavy inter- granular precipitation zone, although some examples indicate t h a t short cuts of Figure 39 shows the structure of a s a l t the cracks through grains are possible.

coated specimen of titanium 8A1-LV-No alloy, which had not failed a f t e r 10,OOO hours. It is interesting t o note that the structure of t h i s specimen does not indicate any heavy and continuous grain boundary precipitates.

The third mterial prone t o stress corrosion failure w a s AM 350 M steel.

Figure 40 shows the heavy grain boundaries formed and incipient cracks a t a Figure 41, the structure of a failed specimen, section through the surface.

however, does not Indicate any material structural cbmges which may be respons- ible f o r such failures.

Very interesting data were indicated by the X-ray diffraction examination of the corrosion products on t i t a n i u m alloys.

It had generally been assumed that the Culprit i n the stress corrosion attack of salt on such alloys is the forma- t i o n of chlorine o r t i t a n i u m chlorides. However, the presence of such chemical products has never been proven.

OLlr X-ray diffraction results show the complete absence of any spectral lines due t o the presence of t i t a n i u m chlorides or sodium titanate, which has also been suggested aa a by-product of chloride reactions. One specinen of t i t a n i u m 8-1-1 alloy, exposed f o r 15,000 hours w i t h a coating of synthetic sea salt, did indicate positive evidence of the eldstence of N a C 1 and M@, and also showed an additional five lines which have been tenta- tively identifiedwith NaOH.

Huwever, the diffraction pattern was diffuse, which was probably due t o particle size distribution. The intensity and spacing of the lines obtained on t h i s sample are sunaPsrized in Table XXI. The M g O lines m e probably derived from dust contamination originating f r o m the furnace bricks.

Another sam(P1e of the corrosion products of a spechen exposedto less than 5000 hours wae boiled i n distilled w a t e r i n an effort to el.lminate the diffuse pattern.

In this material a positive identification of anatase (a form of TiOg) was made, and again three weak lines corresponding t o NaOH were found. Now, anatase is the t i t a n i u m oxide which is preferentially precipitated from W n e solutions in preference t o the more c m o n t i t a n i u m oxide, rutile.

The thermodyaamics of possible chemical reactions of the ingredients present which could result i n the formation of NaOH was then studied. One example is the reaction:

2 N a C 1 + T i + 1/2 O2 + K O = Tic% + NaOH

The free energy change of that reaction is shown i n Figure 42.

It must be stressed that the hypothesis that NaOH is actually formed r e l i e s on a very f e w prelhirmry data, and requires further study f o r the positive identification of the surface reaction.

The possible presence of NaOH could result i n the formation of a number of campounds which are liquid a t temperatures of 650 F and s l i g h t l y beluw. The proof of such a reaction could lead t o the establishment of a minimum temperature of stress corrosion susceptibiuty, corresponding t o the lowest melting point of the reaction products.

Electron Microprobe Analy sis

I

In order t o gain further insight i n t o the stress corrosion mechanism Of t i t a n - iu, electron mlcroprobe analysis was carried out on sections cut a t right angles t o the crack in titanium alloy s t r i p s which had failed by stress corrosion. Fig- ure 43 shuws the results of the examination of titanium 8-1-1 alloy, both i n the unexposed condition and also af'ter exposure under a salt coating t o failure which occurred af'ter 4OOO hours a t 650 F.

In the sample current image, dark areas indicate concentration of elements w i t h high atoanic numbers. In the back Scatter images such areas ehow up light.

It can be seen that e a o s u r e resulted in a distinct coarsening of the size of the areas containing elements with high atosnic fiunibers, in t h i s case molybdenum.

The central areas i n t h e photos relating t o the exposed speckens indicate the crack.

There is a dlstinct concentration of elements w i t h high atanic nmbers (again molybdenum) in the area adjoining the crack. Analysis of the camposition of the light and dark areas gave the follow- i n g results: Backscatter Imsge Dark Areas L i g h t Areas 0 6 5 4 6 3 9 25$ 0.86$ 1.04$ 8.5* 7.n$ These analyses are estimatedto be accurate t o about 2 percent of the w u n t of the elmemt report. The segregation of molybdenum in local areas w a s confinned by a randautraverse taken over a distance of about 100 microns, analyzing the Mo K wline w i t h a l i t h i u m fluoride crystal.

I n titanium 6Al-4V alloy, somewhat similar e w n t segregation phencmena could be observed relating t o concentration changes i n vlagadium and aluminum.

Backscatter electron image photographs of the a l l o y are sham i n Figure 44. The s-le taken f o r this investigation (No. AD3K) was a specimen which had been both braze coated and salt coated and had failed by stress corrosion. The electron microprobe measurements were carried out on a transverse section.

Corrected analysis results were as f o l l w s for the various structural areas: B r i g h t Areas i n Matrix 5.73 6.41 Dark Areas i n Matrix 3.12 8.07 S m a l l Crack Area 3.80 10.36 B r i g h t Area i n Vicinity of S m a l l Crack 4.79 8.33 B r i g h t Area in Network Region 2-55 8.57 D a r k Area in Network Region 3.39 5.40 T e s t s w e r e also carried aut t o detect the possible presence of s i l v e r ( f r o m the brazing a l l o y ) and sodium (from the salt coating). O f the latter, the lower limit of detectability is around. 2 percent and none but a possible trace near the edge could be found. Silver was detected qualitatively near the edge and in various locations within the cracks. There was therefore, within the limits of detectability, no evidence of the diffusion of either of these elements into the titanium a l l o y .

CONCLUSIONS AND R E C O M M E N D A T I O N S The report presents i n t e r i m results obtained during the first half of a 30,000 hours exposure test. M o s t conclusions must therefore be considered t e n t - ative only and subject t o correction and amplification a f t e r completion of t h e test.

1 . Of the candidate materials t e s t (titanium 6A1-4V, titanium 8A1-1 A 1 -

Mo, PH15-7Mo steel, UM 350 M steel, Inconel 718, Rene' 41) only the t i t a n i u m alloys and AM 350 Mare subject t o stress corrosion failure i n the presence of s a l t on exposure a t 650 F and stress levels above approxirnately 255 of the yield strength.

Silver braze coating on t i t a n i u m alloys cause rapid surface deter- 2.

ioration and there i s a complete loss of adhesion between the braze coating and t h e parent material. Gold base brazing alloy does not appear t o affect t h e corrosion behavior of ferrous and superalloys.

A l l alloys appear t o undergo slight changes i n the metallurgical 3 - structure, which are reflected i n the mechanical property changes and possibly also i n the stress corrosion behavior.

4. Mechanical properties are changed i n precipitation hardening alloys on exposure. The change is most pronounced i n precipitation harden- ing steels, least pronounced i n the superalloys.

The results on examination of surface films and microprobe tests 5 .

allow the establishment of a very tentative hypothesis of factors affecting the stress corrosion mechanism of t i t a n i u m alloys. It appears t h a t t h i s mechanism is related t o the formation of NaOH from salt coatings and segregation phenomena i n the alloys. Hypo- thetically, these compositional changes produce local potential differences, which, under the possible presence of a liquid phase containing NaOH, are capable of propagating stress corrosion crack- ing.

Future work under t h i s program w i l l be concentrated m i n l y on changes i n rrrctallurgical structure and mechanical properties. It is strongly recommended t h a t programs be initiated supplementary t o t h i s to: 1 . Investigate f'urther the hypothetical stress corrosion mechanism f o r t i t a n i u m alloys suggested here.

Such an investigation would lead not only t o a better understanding of the mechanism, but also t o the establishment of guidelines for the development Of alloys of improved stress corrosion resistance.

2. Investigate aging effects of various heat treatments on F % s t e e l s and superalloys t o reduce the effects of elevated temperatwe exposure on mechanical properties.

Progress Report on the Salt Corrosion of 1. Titanium MetaYurgical Laboratory: TML Report No. 88, Titanium Aunys at Elevated Temperature and Stress.

Battelle Memorial Institute, ?ToveItiber 20, 1957.

Memrand.um on Notes on the Stress Cor- 2 . Titanium lvletallurgical Laboratory: rosion of Titanium. Battelle bkmorial Institute, July 27, 1956.

3. E . Brawn: Stress Corrosion Cracking of Ti-5Al-2.5 Sn. DMIC Memorandmu 60, August 4,1960.

4. Elevated Temperature Stress Corrosion of R. L . mrchner and E . J. Ripling: High Strength Sheet Materials in the Presence of Stress Concentrators.

lkterials Research Th%oratory, Inc., Quarterly Progress Reports Nos. 1 , 2 , 4 and 9 for NASA Contract MSr-50, November 1961, March 1 to May 31, 1962 and June 1 to August 31, 1963.

Research on the Basic Xature of Stress Corrosion for V a z i a u s 5 . F. A . Crossley: WA2 Technical Docu- Structural Alloys at Room and Elevated Temperatures.

mentary Report No. ASD-TR-61-7l3, Part 11, Februa,ry 1963.

Minutes of E i g h t h Meet- of NASA Special Camittee on PIaterials Research for 6 .

Supersonic Transports. April 4-5, 1963.

K . E . Bisshopp and D. C . Druclcer: Large Deflections of Cantilever Beams.

7 .

Quarterly of applied mathematics, Vol. 3 , No. 3, l & 5 .

8 . R. E . Peterson: Stress Concentration Design Factors, John Wiley & Son, 1959.

APPENDIX A SlRESSES AND DEFLECTIONS I N CANTILFVW B W Introduction Due t o the small bending stiffness of mny of the cantilever beams relative t o the loading, it was necessary t o resort t o a large deflection theory for t h i s analysis. The material properties a t 650°F l i s t e d on Table X X I I were used f o r t h i s analysis.

Large Deflection Theory Figure 45 shows the relation between momnt arm and m a x i m deflection versus a The nnaximum decrease i n lever arm for speci- load stiffness index (reference 7 ).

mens was found t o be about a$, i.e. from Figure - L*-D was found t o be about .800.

L * Effective Length of Beams shows a typical loaded beam. Where P is the applied load, L is the Figure length of the beam from load p o i n t t o angle support, and L, is the distance inside the angle t o the fixed point of the beam. Since temperature the t o t a l length of the cantilever beam can be expressed as

L* = (Lo + L) (1 + O A T ) or = L* Equation (1)

m-L

where : L* c t o t a l length of beam ( i n .)

o ( = coefficient of thermal expansion a t 6500 (in/in*)

AT = (6509 - moF)

For five t e s t beams t h e b deflection readings a t the load point were recorded.

For each beam, the following procedure w a s then used t o obtain the value of Lo.

A value of L * was assumed and the corresponding value of Pz *2/B was calculated.

where : B = E1 bending stiffness (# i n 2 )

I = bt3 moment of inertia ( i n 4 1

I 2 t = beam thickness (in) b = beam width ( i n ) V = Poisson's r a t i o

Using t h i s value of Pz*2/B and Figure ls5 , the corresponding value of /L* was

found. This value was compared t o ( b masured]/L*. If the two values were different a new value of L * was assumd and t h e process was repeated until b /L*

from Figure 45 equaled (b measured) / L * . Then, by the use of Equation ( l ) , Lo

was calculated. Table gives the values of L , for t h e beams on which the deflec- tions were meaured. Based on Table;C?ZJJ%t was decided t o make Lo equal t o .3 for a l l t e s t beams. Equation (1) then becomes:

L* = ( . 3 + L) (1 + 5 8 0 0 ' ) - Equation (2)

Un-Notched Beams Using these corrections the maximum deflections and stresses f o r the un-notched beams here calculated. These calculations are based upon the t o t a l thickness of beams and do not take i n t o account the material properties effects of those beams coated w i t h braze alloy. If t h e actual s t r e s s in t h e braze alloy and parent mater- i a l are wanted, the following procedure can be used. From the tables, the nominal maximum s t r e s s of the specimen can be determined by the following equations:

6 , = G i (1 + a13 Equation (3)

braze alloy nominal t

E2 + 3d + a (d)2 + (d)3

-

-

z t t

Gnrax = G -

x - t x Ea Equation (4) parent material braze allay t+d E i where : E 1 P Young's rnodulus of braze alloys E2 sc Young's modulus of parent material t = t o t a l thickness of parent material d = t o t a l thickness of braze alloy Notched Specimns A similar s e t of calculations was carried out t o determine the maximum deflection at the load point and the maximum stress a t the interior notch for the notched beams. The corresponding stress concentration factor K.t is also given (Reference

8 1. These calculations are also based on the nondnal thickneso of the speciEns

and exact values of stresses can again be obtained w i t h the use of Equations (3) and (4).

Determination of Stresses i n Un-Notched Beams a t Intermediate Locations The stresses a t any point intermediate between the support and the point of load i9 application i n an un-notched beam can be determined from t h e following procedure: a . Obtain the mimum bending s t r e s s and load-stiffness index (PL*/B) from the appropriate table.

b. From Figure 46, read the factor ( L -4- X ) / ( L - P ) for the appropriate

beam position and load stiffness index. Intermdiate values of S/L must be interpolated from t h e curves given.

The desired stress is the beam's xmximum stress multiplied by t h e factor C .

obtained from Figure 4 . 6 .

Ly L u v) v) N N

z

0 00 a

c c

a a c

v) c v) ey I Z v

a m

v) ey cy

Ir

I -.

I ,I&

c

I 6

~

a

*o

z

a E

c h

-

rir

z

Lu c

a

c

I u

5= P,

*

SUMMARY O F HEAT TREATMENTS Titanium 6 U l - 4 V Braze a t l725F air cool unbrazed specimn: Age a t LOOOF for 4 hours Titanium 8-1-1 Braze a t l725F a i r cool unbrazed specimen : no heat treatment PH 15-7 MO Braze a t lgoOF (brazed specimen only) cool t o r o o m temperature heat t o 1730 F

COO^ t o - lOOF, hold 4 hours

age a t lO75F for 1 hour a i r cool t o room temperature AM 350M Braze a t LgOOF (brazed specimen only) cool t o room temperature heat t o lnOF, cool t o - 100 F, age a t 850 F for 3 hours a i r cool t o room temperature Rene’ 41 Braze a t 1950 F (Brazed specimen only) a i r cool t o r o o m temperature age a t 1400 F f o r 6 hours a i r cool t o room temperature Inconel 7l8 Braze a t 1900 F (Brazed specimens only) air cool t o room temperature stress relieve a t 1600 F for 4 hours a i r cool t o room temperature age a t 1325 F f o r 16 hours air cool t o room tenperature EXPOSURE TREATMENT COATING TEMPERATURE H NONE B 650" F SYNTHETIC SEA SALT BRAZE COATING TITANIUM ALLOYS: DYNABRAZE B C 650°F (94.8% Ag, 5% A I , 0.2% Mn) OTHER ALLOYS: PREMABRAZE 128 (72% Au, 6% Cr, 22% N i l BRAZE COATING AS ABOVE D 650" F PLUS SYNTHETIC SEA SALT ALTERNATING FORTN I GHTLY 650°F AND E SYNTHETIC SEA SALT HUMIDITY CAB I NET AT 100°F Table I11 Summary of Surface and Exposure Treatments NQTCHED S PEC IMENS I N C O 718 RENE' 4 1 Ti 6-4 \M 350 M T i 8-1-1 PH15-7 M a

-

12.41105.5 126.81147.4 56.2166.6 140.61168.1 29.31160.3 M A X I MUM S PEC I M E N 44.6158.0 STRESS ( K P S I ) 85.61100 65.5174.8 % FTY (STD TMT) 34.0138.4 41.5146.6 72186 77.3196 42.2143. 7 96 FTY (BRAZED TMT) 35.8138 UNNOTCHED SPECIMENS 48.8157.4 53. 7161.3 50.0159.1 34.0140.8 33138.9 27.5131.4 30135.4 24.1128.9 I I

TABU3 V - ~aximum S t r e s s levels

Test lkterial: Titanium 6A1-4V SPECIMEN MAX. STRESS ENVIRORMENT NUMBER p s i K t A A 1 Continuous exposure i n 35,900 c i r c u l a t i n g a i r a t 650 F AA2 35,200 u 3 35,- AA4 35,300 a 5 5 9 , m 1 895 A A ~ 5 9 , m 1.895 m 7 33,900 Coated with synthetic sea AB1 33,200 salt and then continuous A B 2 34,200 exposure i n c i r c u l a t i n g air AB3 33,400 AB4 at 650 F A B 5 52,500 1.945 AB^ 52,200 1.945 A B 7 33,700 C o a t w i t h brazing alloy A C 1 28,100 28,600 and then continuous exposure AC2 i n c i r c u l a t i n g air a t 650 F AC3 28,900 48,700 1.970 ACS AC6 48,400 1 975 Ccmted with brazing alloy AD1 plus synthetic sea salt and AD2 g ; ; : then continuous exposure i n A D 3 30,200 c i r c u l a t i n g air a t 650 F 45,- 1.992 AD5 AD6 47,800 1.975 Coated with synthetic sea salt, AE1 33,600 A E 2 then a l t e r n a t i n g 14 day expos- 32,500 ure i n humidity cabinet at 100 F AE3 3 4 , m and i n c i r c u l a t i n g air a t 650 F AE4 33,000 AE7 33,200 A 1 3 8 33,700

TABLE V I - Maximum S t r e s s Levels

Test Material: Titanium 8AL-1V-lI40 SPECIMEN MAX. s m s s EMrIRONMENT rouMBER ps i Kt Continuous exposure i n circu- BA1 30,300 l a t i n g air a t 650 F BA2 28,900 B A 3 30,900 1.88 BA5 67,600 1.88 B A ~ 67,400 BB1 26,400 Coated w i t h synthetic sea s a l t and t h e n continuous exposure BB2 27,700 28,500 i n c i r c u l a t i n g air a t 650 F m 3 1.885 61,200 BB5 1.89 BB6 60,700 Coat w i t h brazing alloy a n d BC 1 25,800 then continuous exposure i n BC2 31,000 c i r c u l a t i n g aiF a t 650 F BC3 21,900 1.895 BC 5 5 9 , m BC6 1.895 59,900 Coated w i t h brazing a l l o y plus BD1 23,800 synthetic sea s a l t and then BD2 22,000 continuoua exposure i n circu- 24,800 B D 3 l a t i n g air a t 650 F 60,400 1.09 BD5 BD6 1.91 57,600 Coated with synthetic sea salt BE1

27, mo

t h e n a l t e r n a t i n g 14 day exposure BE2 27,000 in humidity cabinet a t 100 F and RE3 28,900 i n c i r c u l a t i n g a i r a t 650 F BE4 27,mo

!UBI& V I 1 - Wimum Stress Levels

Test k t e r i a l : PHl5-7M[) SPECIMEN M u . STRESS EWIRONME2lT m E R psi K t Continuous exposure i n cir- C A 1 59,700 cuhtinp: air a t 650 F CA2 58,800 58,200 C A 3 CA4

57, wo

C A 5 165 ,m 1.735

C A 6 165,200 1.735 58,500 C A 7 Coated with synthetic sea salt cB1 55,700 and then continuous exposure i n cB2 55,600 circulating a i r a t 650 F C B 3 58,200 C B 4 58,000 1 . 7 6 CB5 154,200 c ~ 6 155,900 1.755 CB7 5 8 , w

Coat w i t h brazinn alloy and cc1 54,100

then continuous exposure i n cc2

52,800 circulating air a t 650 F cc3 55,500 cc4 53 ,OOo 146,500 1.78

cc 5

cc6 140,300 1.79 cc7 53 ,m Coated with brazing alloy plus C D 1 52,400 synthetic sea salt and then CD2 52,600 continuous exposure i n circu- 0 3 53,700 lating air a t 650 F C D 4 52,300 146,300 1.781 0 5 C D 6 146,400 1 791 C D 7 53,200 CE1 Coated w i t h synthetic sea salt, 57,900 then alternating 14 das expos- CE2 57,060 ure i n humidity cabinet a t 100 F CE3 57,100 and i n circulating air a t 650 F CE4 57,000 CE7 57,500 CE8 @,OOo

TABU VI11 - Maximum Stress Levels

Test Material: AM 350 M S m N MAX. STRESS ENVTROPMEWJ! P s i K t DA1 56,800 DA2 57,400 D A 3 57,500 Continuous exposure in circu- D A 4

57, mo

1.745 l a t i n g a i r a t 650 F M5 159,200 D A ~ 1 745 157,900 54,500 m 7 Coated w i t h synthetic sea salt DB1 54,500 and then continuous exposure DB2 53,600 in circulating air a t 650 F 56,600 D B 3 D B 4

57, mo

1.762 151,500 D B 5 DB6 1 ~ , 7 0 0 1-77 D B 7 54,500 Dc1 48,800 Dc2 49,900 C o a t w th brazing a l l o y and E 3 49,700 Dc4 then continuaus exposure i n 50,400 1.8 circulating air a t 650 F Dc5 u7,200 1.814 ~ c 6 129,600 Dc7 49,800 DD1 50,400 Coated w i t h brazing alloy plus DD2 49,- synthetic sea s a l t and then con- m 3 %- tinuo- exposure i n circulating DD4 50,800 1.814 a i r a t 650 F DD5 130,200 1.80 D D 6 136,200 D D 7 50,800 Coated with synthetic sea salt, DE1 55,- t h e n alternating 14 day exposure DE2 55,600 i n humidity cabinet at 100 F 55,- D E 3 and i n circulating a i r a t 650 F D E 4 *,goo

TABLE - Maximum Stress Levels

Test Material: Inconel 718 SPECIMEN MAX. STRESS ENVIROlJMENT NUMBER p s i Kt Continuaus exposure i n circu- U l 39,600 l a t i n g air at 650 F En2 39,000 40,000 EA3 E A 4 36,600 1 . ~ 5 EA5 105,500 EA^ 1.885 105,500 EA7 39,400 Coated with synthetic sea E 3 3 1 37,400 s a l t and then continuous EB2 36,800 exposure i n c i r c u l a t i n g air E 3 3 3 38,800 a t 650 F EB4 39,400 EB5 1.895 99,200 m 6 107,600 1 875 E37 39,400 C o a t with brazing alloy and EC1 33,800 then continuaus exposure i n EC2 33,300 c i r c u l a t i n g air a t 650 F E 3 35,400 EC4 34,800 EC5 101,900 1.895 EC6 98,400 1 915 EC7 35,400 ED1 Coated w i t h brazing a l l o y 35,400 plus synthetic sea salt and ED2 34,900 then continuous exposure i n 35,200 ED3 c i r c u l a t i n g air at 650 F Fa4 34,900 ED5 94,400 1.92 ED6 1.92 94,800 ED7 35,400 Coated with synthetic 8ea EE1 39,400 salt, t h e n a l t e r n a t i n g 14 EE2 37,900 day exposure in humidity E33 37,800

m

cabinet a t 100 F and i n 38,600 c i r c u l a t i n g air at 650 F =7 39,300

TABLE X - Maximum Stress Levels

Test Material: Rene' 4 1 SPECIMEN MAX. STRESS ENVIRONMENT NUMBER p s i Kt.

F A 1 Continuous exposure i n circu-

53, mo

lating air a t 650 F FA2 53,- 54,100 FA3 FA4 53,000 FA5 1.780 147,600 FA6 146,600 1.780 FA7 53,100 Coated w i t h synthetic 8ea salt

and then continuous exposure ;:E

i n circulating a i r a t 6fjo F 52,900 53,600 1 . 7 8 0 145,700 150,600 1 772 53,100 Coat w i t h brazing alloy and F C 1 47,800 then continuous exposure i n Fc2 47,500 circulating air a t 6% F 48,400 FC3 PC4 48,000 Fc5 1.820 13~,800 FC6 1.83 128,400 48,800 Fc7 Coated w i t h brazing alloy FD1 48,800 plus synthetic sea salt and FD2 4 8 , w then continuaus exposure i n 47,600 FD3 circulating air a t 650 F FD4 48,800 1.820 130,400 FD5 1.830 FD6

133 , 500

48,800 FD7 Coated w i t h synthetic sea salt, 54,500 then a l t e r n a t i n g 14 day exposure FE2 5 3 % ) in humidity cabinet a t 100 F and FT3

in circulating a i r a t 650 F :;;E FE4

m 7 52,000 FE8 54,200 .

R

7”

k

b

$4 td) Sl d -6- -6- ? 4 m I I

I I

o m

L-

a

2 0 J ..I I .

. .- I TABU XXI ANALYSIS OF SALT COATING FROM 8-1-1 TITAKCUM ALU;>Y, EXPOSED To 650 F FOR 15,000 HOURS LINE INTENSITY SPACING NaC 1 NaOH Mgo 1 W 3 -255 3 -258 2 w s 2.805 2.82 2.85

-

vvw 3 2 .o9G 2.16 vs 1 0994 2.03 1 993 VVW

5 - 1.70

1.693 6 1.626 1.628

-

f 1.48 W 1.409 1.410

-

f 1.29 f

1.27 - 1.27

1 1 m 1.260 1.261 12 dif f ,w 1.151 1.1515 W 0 -997 0 09969 W 0 0953 0 09533 db It 0 4 4 0 1 5 0.9401 16 d b l t o .891 0.8917 d b l t 0.849 17 o .8503

-

18 d b l t 0 783

-

db lt 0.781 M A T E R I A L PROPERTIES Temperature = 6509

G E a in/in/*F

-

- k t e r i a l -

1.2912 x 106 5. x 10-6

6-4 T i t . 14.1 x lo6 5.4 x lom 3 w

FTU = 160

1.419 x lo6

5.4 x 10-6 3 w

2.385 x lo6

. 308 6.15 x loo6

PH15-7Mo 25.9 x 106 10.1 x 106 F'IU = 200

2.373 x lo6

. 318 6.9 x 10-6

25.6 x 106 9.9 x i o6 A M 3 5 0

m = m

7.9 x 10-6 2.541 x 106 INCONEL 27.7 x 106 10.2 x 1 0 ' 303 F ! J ! u = 190 2 . 6 4 1 x 106

. 318 7.05 x loe6

REME' 41 28.5 x lo6 10.9 x 10

FTU = 194

* Estirmted

TABLE ]MI11 CALCULATIONS OF m T I v E BEAM LEBGT&s

&am b (measured) L* E+%

L L,

C A 2 3 375 If 9.74" 1 . 1887 9.7" .00538~~

DA3 3.625" i o . 13 1 .2334 9.8" .28962

FA1 3 .mtl 10 . 16 0957

9.7311 38862 " BA2 4.500" 9.31" 1.9329 9.07" . 2 w 3 3.152" 10.05 1.0302 9.76" .&93 Thickness ltl Titanium 6 A I . - 4 V 0 . 0 5 0 11 T i t a n i u m 8A1-1v-U.lo 0,04811 and O.02Ot1 PH 15-7 M o S t e e l 0.03911 AM 350 M Steel 0.04.0'1 Inconel 718 0,042" Rene 4 l 0.05011 O.O1O1l 5 0,00111 ; rC, 1 . 6 to 1.8 Depth of notch

c 0

-Id T

i

Gt

I 1

b\&l cftl.c

4k 4k

Notched Specimens Figtux 1; Dimemions Figtux 1; Dimemions of 44 44 Ekposure Specimens In Test Fram 4 ; Each Specincn is identified by a four syrabol code, as follows: First syxbol: Specimm I h t e r i a l A T i t a n i u m 6 AI.-4 V a l l o y B Titaniun: 8A1-1V-U.10 a l l o y C P H 15-7 1 - 1 0 Steel D AI4 350 M Stcel E Inconel 718 F Rene l+l Second oyrfibol : Surface Treatmnt A None B SiLt Coating C Braze Coating D Salt plus Braze Coating E Cyclic Expo- T h i r d symbol : Spcchien S e r i a l Kunber f o r e x h n d c r i a l wti surface treaknent Fourth symbol : Specimen Position Code Letter as illustrclted i r . the diagram below.

L '-\ /'-, I-' b-\ c4 -\ / I 1 # I ' 1 I I : I

: : ; {

; I ! ! 0

: : I , I I I I I t I

-) L, ,'

,*J \,p; \, 4; Systcil: of !hrking and Cutting-up o f Exposure Specimens a d other Test Specimens.

t . O & mE3: 1. Holes to be on centre l i n e of notch root width + . 0 0 2 .

Notch radius 0,002 mrudnnan . $J 6 . 0

2 , or Notches to be made with li&t finishing cuts 3.

light grinding and must have contour shown.

Tool Chatter o r other Tool marks v K I . l be cause f o r 4 .

reject. Do NOT buff.

Machine surface of notch 5 .

6. Notch and reduced section t o be symrietricd about centre l i n e f .0015.

Notched and Un-notched Ifiniatube Tensile F’ieure 5 ; specimens.

I" Figure 7 Miniature T e s t Specimens T e s t R i g - Assembly with Extensometer

I

Etched x 500 Two stage replica x 2500 Figure 8 .=Microstructure of T i t a n i u m 6A14V Alloy prior t o exposure to test environment.

.

K c x K ci LL c \ -m K w a, k a , I , >, ,

Y

-Q r n Etched x 500 Two stage replica x 2500 TWO stage r e p l i c a x ~ , O O O Figure 10 .-F?icrostructure of T i t a n i u m 8~1-&-1v Alloy prior t o exposure t o t e s t environment.

15,000 HOUR EXPOSURE NO FAILURE Etched x so0 ' e TWO stage r e p l i c a x 15,000 Figure 11- Ylcrostructure of Titanium 8~J.-ll$0-1~ A l l o y a f t e r eXpoSUre Fn circulating air a t 650% (Specimen BA2) Two stage replica x 2500 Two stage replica x 15,000 Figure 12- Microstructure of pHl+7Mo S t e e l p r i o r t o exposure to t e s t environment.

, , _.

8 8

I n n U K K

s a

a , k I

. ' '"?

* ,, , \ -

E *

5 6 Two stage replica x 2500 Two stage replica x 15,000 Frlgure 1 4 . - Microstructure of AM 350 M Steel prior to exposure to t e s t environment.

K c- t Etched x so0 .- Two stage replica x 2500 h r o stage replica x 15;,ooO

Figure16 .- Microstructure of Inconel 718 Alloy

p r i o r to exposure t o t e s t environment.

H l n x d cd V a m cu x H cd d d E4 a k Two stage replica x 2500 ' h a stage replica x 15,000 Figure I & - Microstructure of Renet Alloy p r i o r t o exposure to test environment, rr\ 0

m -

cu -Ln x x d M Id cd 0 .e- 0 d .

d

d

a, E k

B

8 =

-Ln N x

e

16C - v

v

-

S O VA

- v

3 . 4 0

-

120- 1 6 0 , - SPECIMEN CONDITION EXFYXURE ON REMOVAL FROM

" I

01 I I I I I I 0 10 20 25 30 T I h I N TEST (lo00 hours)

Figure a- T i t a n i u m 6Al-hV Alloy Ilizstres-d Bcposure

Tensile Test Results.

NOTCHED TENSILE TEST UNNcficHED TENSIIE TEST

r SPECIMENS Kt - 6

SPECIMENS n

8 170

d t

110 llllllr 110 I I I I I I I ~~

o 5 io 15 20 25 30 o 5 io 15 20 25 30 TIME I N TEST (1000 hours) TIME I N TEST (1000 hours) n . d 150 l i W SPECIMEN CONDI"I0N EXPOSURE ON RENOVAL FROM

B 120

SPECIMEN ENVIRONMENP

A a

B TIME I N TEST (1000 hours)

idi 30 r

C

E 3 I

D

A

E D I I I I 1 I O 5 10 15 20 25 30 TIME I N TESI (lo00 hours) Fig- 2L- T i t a n i u m 6 A 1 4 ~ V Alloy StmSSed GxpOSUre T e n s i l e Test Resulta.

NCITCHED TENSILE TEST c SPECIMENS K t ' 6

-

U

160 t

!

lloo

5 10 20 25 30 4 ; i o ;5 i o ;5 30

TIME I N TEST (lo00 hours) TIME I N TEST (1000 hours) 160r SPECIMEN CONDITION MPCEiURE ON REMOVAL FROM SPECINEN ExmlfL- Figure 2 2 . - Titanium 8Al-lNo-lV Alloy Unstressed Exposure Tensile Test Results.

UNNcrrcHED TENSIIE TESP NOTCHED TENSIIS TEST SPECIMENS K t = 6 SrnIMENS 190r

8 18

16 d-

B

v

-

130 V

A

120- I I I I I I 10 1 20 25 30 T&E I N TEST' (1000 hours ) SPECIMEN CONDITION

M W G m I OM REMOVAL FROM

B

mITIo"T P=YEiz UNBROKEN

A

Q

7 V

A 1 - 1

I I I I I I l d B

E - 0 m 0 5 10 15 20 2 5 30

TIME IN TEST (1000 hours) C

v--v v

H D

&--* A

3- 2

$8

E-3

I I I I I 20 2 30

TIME I N s ( l 0 0 0 bok)

Figure 2 3 . - T i t a n i u m 8Al-lHo-lV Alloy Strearred EXposUre I Tensile Test Results W 230

[ 220-

t n

a 210-

H

E 4 2 0 0 -

E-c 180 1 9 r

l7 0 5 1 0 20 2 5 30

TIME I N TEST (1000 hours)

sml" C O N D I T I ~

EI[pCtjuRE ON RpiovBL FROM SPECIMD?

ENVIRONMENT A B

ub 2 0 25 30

& 10 T IN TEST (1000 hours)

C D E Figure 24.- Fiis-7Ho Steel Unstressed Exposure Tens- Test Results.

UNNOTCHED TENSXU TEST

ft

l7'b 2 i o 2 i o is G

TIME I N TEST (lo00 hours) A

o---o Q

B

O--O m

C

o-----o v

s d

D &--* A

3 3

2 0 3(1

E D----O I0 88

0 5 10 15 20 2 5 30

E . ? 0

TIME I N TEST (1000 hours)

Figure 25 .= HIl5-7Ho Steel Stresmd Exposure

Tensile Test Results.

NOTCHED TENSIm TEST SPECIMENS K t ' 6

23 P

I I I I t I I 170i

5 X I 15 2 0 25 30

TIME I N TESP (1000 hours) SPECIMEN CONDITION EKPC6URE ON RplOVkL FROM SPECIMEN M P O G U R E EST EwlRO"EN!C UNBROKEN BROKEN I I I t I I 10 20 25 30

T&IE I N TEY!'!'(lOOO hours)

20 3 0 [

0 5 ID $ 20 2 5 30

TIME I N TEST (1000 hours) Figure 26.- A I ! 350 Steel Unstressed Ekposure Tensile Test Result.s.

230 t

1 I I I I I lo 20 25 30 TI& I N TES'!?(lOOO hours) SPECPIEN C O N D I T I r n EXPOSURE ON REMDVAL FROM SPECIMEN EXPOSURE TEST ENVIRONMENT TIME I N TEST (1000 hours) - Figure 27.- AX 350 Steel Str888ed EYCposUre Tensile Teat Results.

UNNOTCHED TENSILE TEST NUI!CHED TENSILE TEST SPECIMENS SPECIMENS % = 6

/&p

-4

160 l 7 O I

4 0 ""I

I I I I I I I I I I I l l a O 10 20 25 3 3

uoL T&E IN 10 TES?(lOOO 20 hours) 25 30

TIME I N TES?(1000 hours) 160r n I SrnIMEN C O N D r r I O ~ EXPOGWIE ON REMOVAL FROM fn 1 2 0 ' SPECIMEN EST ENVIRONMENT El UNBROKEN El 110- k I I I 1 I I

rn

v

A

lot

I I I I I l o 20 25 30 O : , TIME I N "E$(1000 hours) Figure 28.- Inconel 718 Alloy b t r e s s e d Exposure Tensile Test Results.

UNNUEHED T-IU TEST SPECIMENS n !i

8 l8

e 17

I 1 I I I 1 I 1 I I I I 10 15 20 25 30 10 20 25 3 3 T&E IN T E & ! (1000 hours) T h IN TEST (1000 hours) n I h SPECIMEN CONDITION EXK6URE ONREMOVALFROM I I I 10 I I 20 2 10 1 l'& IN TEST5 (1000 ~O'U~S 1 '

I C

I I I I I

5 10 15 20 2 5 h

TIME I N TEST (1000 hours) Figure 29.- Inconel 718 Allay Stressed S s t p O s ~ Tensile Test Results.

TEST

170 ' " " i

I 1 I I I I I I I I I

10 20 25 30 '9

0 10 20 25 40 5

T I & ! I N "FS?(lOOO hours) TIME IN Tg(1000 hours) 170 r SPECIMEN CONDI!I!ION M P O G U R E ON REMOVAL FROM EKPo6URE TEST

a

a

k 110 I I I I I I I

0 10 15 20 25 30

2 T IN TEST (1000 hours)

E $ 301-

0 I I I I I 10 20 2 5 3b

' T l h IN TZ(1000 hours)

Figure 30.- Renet 41 Alloy Unstressed Exposure Tens& Test Results.

UNNOTCHED TENSILE TEST

n 4 247 SPECIMENS

" 8 17

166 I I I I I I 1 6 6 0 5 lo 15 20 25 30

0 5 10 15 20 25 3 0

TIME I N TEST (1000 hours) TIME IN TEST (1000 hours) SPECIMEN C O N D r r I o N

I I ON W O V A L FROM

I I I 1 1 I

lido 5 I D l5 2 0 25 30

Figure 3 . - Renet kl Alloy Stresaad Exposure

Tensile Test Results.

SURFACE SPECIMEN NOTCHED (N) OR TOTAL FAILURE TREATMENT, MATE R I A L TIME HOURS NO. UNNOTCHED (U) EX PO S U RE SPECIMENS 5516 A C 6 T I TA N I UM BRAZE 7124 A C 3 6A I -4V 650°F AC 5 9740 A C 2 15,439 A D 5 BRAZE PLUS SALT A D 3 550°F 15,480 AD 2 U TITANIUM SALT 2640 BB 2 I U 650°F 8AI-1V-1 MO 3980 BB 1 BRAZE 650°F 10,796 13,652 BRAZE PLUS BD 3 U SALT 650°F 10,796 DB 5 N SALT A M 350 M 650°F 15,463 DB 6 N 2880 DE 4 SALT DE 2 ALTERNATE 3290 DE 3 i50"F & 100°F 3360 DE 1 I I ~ i g u r e 33 Summary o Stress Corrosion Failures 'Ln tu X t d a 0, k a co J rn

E

Etched x 500 F2gure 35 o- Microstructure of Braze and Salt Coated T i t a n i u m 6A1-4V Alloy notch s p e c b n A D S showing surface crack. Specimen failed after 15,OOO hours exposure in circulating air a t 650%.

9,308 HOUR EXPOSURE FAIIED Two stage replica x 2500 Two stage replica x s , O o O

Figure36 .- Microstructure of Braee Coated T i t a n i u m 8 ~ b ~ f 0 - 1 ~ Alloy

after exposure i n circulating air a t 650%. (Specimen ~ 2 ) 9,644 HOUR EXPOSURE FAILED Etched x 500 I_ ' ' . I \ y y J . ; .

Two stage replica x 2500 Two stage replica x 15,000 Figure 37.- Microstructure of Braze and Salt Coated Titanium 8~1-1Mo-lV Alloy after exposure in circulating air at 6509. (Specimen m3) 3,980 HOUR EXPOSURE FAILED Tu0 stage replica x 2500 Two stage replica x 15;,000 Figure 38.- Fticrostructure of Salt Coated T i t a n i u m ~AI.-~.KO-~V Alloy after exposure in circulating air a t 650OF. (Specimen BB1) 1 0 , O O O HOUR MPOSURE NO FAILURE /' .

Two stage replica T w o stage replica x 15,000 Figure 39.- Microstructure of Salt Coated Titanium 8~1-1Mo-lV UaJr after alternating 4 day exposure i n circulating a i r a t 650OF and in humidity cabinet at lOO*, (Specimen BE1) a2 Etched x 500

Figure 4o .- Microstructure of S a l t Coated AM 350 M Steel notch

specimen DE36 a t fracture i n notch showing old and new crack areas. Specimen failed after 15,OOO hours exposure i n circulating air a t 6 5 0 9 .

10,796 HOUR EXPOGURE FAILED Tu0 stage replica x 2500 h r o stage replica x l5,0oO Figure 41.- Microstructure of S a l t Coated AM 350 M Steel after exposure in circulating air at 650oFO (Spechen D B ~ ) 2 N a C l . C T i S & O2 = T i ( = L 2 t 2 N a OH Change in Free & e r g of a Figure 42; possible %res s Corrosion Reaction.

Backscatter Image X888 Unexposed Alloy v- Backscatter Image X888 Salt Coated Specimen # BBlK Failure Time 3980 Hours S q l e Current Image X 8 8 8 Specimen as above Figure 43 Electron Microprobe Analysis of Titanium 8-1-1 U a Y %.-all Crack near Large Crack

3 2 ~ k - s ~ ZL tw hs x 886

# aD3K a e c t r o n 1.!3.croprobe iimlysis of T i t a n i u m 6Al-4V Figure @+:

b

I r, 008 1 . 0 0 0 . 2 Deflection andMooneolt -Am veFsu8 figure 6

Load - Stiffness Parameters

P s/L 0 . 7 P L * / B

Figum & ; Moment Coefficieplts vemw Load - Stifmess

Parmeters for intennediate Eeam Locations.

July, 1964, North American Aviation, h c .

Los Angeles Division, Los Angeles, C a l i f .