ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR, MICH. SECOND PROGRESS REPORT TO MATERIALS LABORA TORY WRIGHT AIR DEVELOPMENT CENTER ON EFFECT OF PRIOR CREEP ON MECHANICAL PROPERTIES OF AIRCRAFT STRUCTURAL METALS by J. V. Gluck H. R. Voorhees J. W. Freeman Project 2498 Air Force Contract No. AF33(616).-336B Task No. 73605 June 20, 1956

SUMMARY This report, the second to be issued under Contract AF33(616)3368, Task No. 73605, covers the period from April 21, 1956 to June 20, 1956. The investigation is a study of the effects of prior creep on the shorttime mechanical properties of three aircraft sheet metals. The materials to be studied are: 2024-T86 aluminum alloy; C110M titanium alloy; and 17- 7PH (TH1050) precipitation hardening stainless steel. The aluminum alloy and the stainless steel have been received. The survey of creep-rupture properties for the stainless steel has been virtually completed, The effects on the room temperature tensile properties of exposures of up to 100 hours in the range between 600~ and 900~F have been evaluated for the stainless steel. The results indicate that the strength is increased for exposures of over 10 hours at 800~ and 900~F. A few completed tests cover the effects of stressed exposure for this material. An added strengthening effect appears to have taken place as a result of the creep strain, The fixture for tension-impact testing has been completed and its use and calibration are discussed.

.1 INTRODUCTION This second bi-monthly progress report issued under Air Force Contract No, AF33(616)-3368, Task No. 73605 covers the period from April 21, 1956 to June 20, 1956, This investigation is concerned with a study of the effects of elevated temperature creep-exposure on the mechanical properties of three aircraft sheet metal s The materials under consideration and the temperature ranges to be investigated follow~ 1, An aluminum alloy (2024-T86) from 350~ to 500~F 2, A titanium alloy (C.110M) from 650~ to 800~F 3, A precipitation hardening stainless steel (17-7PH, TH 1050) from 600~ to 900 ~F The exposure periods to be studied are from zero to 100 hours and the range of total deformation is from 0, 5 percent to 3. 0 percent. The combined effects of deformation, surface attack, and structural alterations are all possible contributors to changes in the strength and ductility of a metal. The net effect of these factors is considered a measure of the stability of the material and an aid in defining the operating limits for design purposeso The properties to be studied both before and after creep-exposure are the short-time tensile properties, tensiornimpact strength, short-time compressive properties, and the hardness. Where significant effects are noted, metallographic studies will be used to study them,

2 TESTING PROGRAM In the present investigation a study is being made of the effects of exposure on the mechanical properties of three aircraft sheet metals. The emphasis of this study is an evaluation of the effect of creep, although unstressed exposure is also under consideration. The testing program was discussed in detail in the First Progress Report (ref. 1). The discussion following summarizes the more important features of the program. The materials under investigation and the temperature ranges to be considered were listed in the Introduction. In addition, it was noted that time periods up to 100 hours and total deformations up to 3 percent were to be studied, This total deformation is all deformation occurring during application of the load and during creep of the specimen at the testing temperature and stress. Time periods for exposure were fixed at 10, 50, and 100 hours. The exposures were to be carried out at no stress and at those stresses resulting in 0.5, 1 0, 2.0, and 3. 0 percent total deformation in the three time intervals listed above, As a matter of policy it was decided that the time of exposure would be fixedd The tests were to be run at the stresses determined to give the indicated deformation in the indicated time period, however, it was expected that the actual amount of deformation might vary slightly from the nominally specified amount. By fixing the temperature, time,and stress it is necessary to accept the deformation obtained. Because of the large number of tests that were expected to be run, this procedure was adopted so that the problems of test scheduling, unit utilization, and laboratory administration could be made as simple as possible.

3 Three primary test temperatures were selected for each material. They are the following: a. 2024-T86 350~, 400~, 500~F b. C110M 650~, 700~, 800~F c. 17-7PH 600~, 800~, 900~F After the completion of the exposure period the following tests are to be carried out: lo Tensile tests at both room temperature and the exposure temperature. 2, Compression tests at the same temperatures. 3. Tension-impact tests at the same temperatures. 4. Hardness determinations at room temperature. Where deemed useful, metallographic examination will be made. The data are to be correlated with respect to the conditions of exposure. The bases for comparison are the properties of the unexposed material, established by a series of replicate tests designed to define the normal scatter of each material. The aluminum and titanium alloys are to be tested in the conditions as received from the manufacturers. The Cl10M titanium alloy is furnished as-hot rolled and annealed and the 2024-T86 aluminum alloy is furnished in the cold worked and aged condition. The stainless steel, 17-7PH, is to be tested in the TH1050 condition. This is a double aging treatment carried out at 1400~ and then 1050~F and is to be performed at the University. The aluminum and stainless steel alloys are to be tested in the direction crosswise to the sheet rolling direction, while the titanium alloy is to be tested in the direction parallel to the rolling direction.

4 Under the current contract, testing priority will be placed on evaluation of the aluminum alloy. Before final establishment is made of the temperatures to be investigated for the stainless steel, a survey will be made of the effects of prior creep to 2 percent total deformation in 100 hours on the room temperature tensile properties. The temperatures studied will be over the range from 600~F to 900~F. For the purpose of this survey, temperature increments of 50~F will be used in order to better establish the temperature of maximum effect. In addition to the above the feasibility of running notched tension-impact tests will be investigated. TEST MATERIALS To date, the aluminum alloy and the stainless steel have been received, w nile the titanium alloy remains on order with delivery promised about Septembe: 1, 1956o The specifications of the materials are as follows: Armco 17-7PH Stainless Steel Sixteen sheets of 17-7PH precipitation hardening stainless steel were received from the Armco Steel Corporation in the period covered by the First Progress Report. (ref 1). The material was supplied as 0. 064-inch thick sheets 120 inches long by 36 inches wide and was furnished in No. 2D finish and in Condition A. (Condition A consists of an annealing treatment carried out at 1925~F followed by air cooling). All the material was from Heat No. 55651. The certified chemical analysis furnished by Armco was within the nominal composition limits for this alloy. These limits are as follows:

5 Element Nominal Actual (heat 55651) Carbon 0. 09% Max 0. 072% Manganese 1.00 MTax- 0.55 Phosphorus 0.04 Max 0.018 Sulfur 0.03 Max 0.011 Silicon 1.00 Max 0.33 Chromium 16.00-18.00 17.03 Nickel 6.50-7.75 7.25 Aluminum 0.75-1.50 1.28 Iron Balance Balanc e The material is to be tested in the TH 1050 condition. The details of this treatment are as follows: (ref 2) 1. Condition A material heated in air at 1400~F for 1-1/2 hours 2. Air Cool 10 minutes (to approximately 500~F) 3, Quench in 600F water 4, Hold 8 12 hours at 60~F 5, Age at 1050~F for 1-1/2 hours, then air cool. 2024-T86 Aluminum Alloy Nineteen A1-clad sheets of the aluminum alloy, 2024-T86, were received from the Kaiser Aluminum and Chemical Corporation in the period covered by this report. The sheet dimensions were 0. 065-inches thick by 48 inches wide by 72 inches long. The chemical analysis and heat number have not yet been received, This material, formerly designated as Aluminum 24S, is a high strength heat treatable wrought alloy. The nominal composition limits for this material are the following:

6 Element Range (per cent) Copper 3.8-4.9 Manganese 0.3-0.9 Magnesium 1. 2-1.8 Silicon 0. 50 Max Iron 0.50 Max Chromium 0.25 Max Zinc 0. 10 Max Others 0. 10 Max Aluminum Balance The T86 condition of this material is a cold worked and aged condition and is carried out by the producer. It consists of the following steps: lo Solution Treatment: 910-930~F, quench in cold water 20 Cold Work: Approximately 5. 5% reduction 3. Agedo 370-380~F C 11OM Titanium Alloy The Cl10M titanium alloy, formerly designated RC 130A, remains on order from the Rem-Cru Titanium Corporation. The material is to be procured as 0O 064-inch sheet in the annealed condition. The alloy is a binary containing from 7-9 percent manganese, the balance being titanium. SPECIMEN PREPARATION The preparation of specimens under this program was discussed in some detail in the First Progress Report. (ref 3) A sampling procedure was set up in order to randomize the results with respect to both the sheet to sheet variations and the variations within an individual sheet. The sampling procedure for the 17-7PH material was illustrated and detail drawings were presented of the various test specimens.

7 Because of the different sheet size of the 2024-T86 aluminum alloy, the above procedure was slightly modified. The as-received size of the aluminum alloy was 48 by 72 inches. The specimens were specified to be taken in the direction crosswise to the rolling direction, i. e., in the 48-inch direction. Three sheets were arbitrarily selected for the initial tests of the scatter of normal properties and for obtaining the necessary time-deformation data to establish creep-exposure stresses. The three sheets were designated 2, 3, and 44 (Sheet I was laid aside since the clad coating was inadvertently scratched during handling). The two sampling schemes were designated Q and R. Scheme Q is that used for obtaining the specimen blanks for the scatter tests. Scheme R is similar and used for locating the creep tests specimen blanks. Each 72 inch long sheet was divided into 14 panels approximately 5 inches wide, The panels were labeled alphabetalically (omitting I and 0). The panels selected for each sampling scheme and the sampling schemes themselves are indicated in Figure 1. Since the First Progress Report was issued, design of the specimen for tension impact testing has been slightly modified. A drawing of the modified specimen is shown in Figure 2. The changes were made to reduce the over-all length to six inches and to change the fillet radius to 1 inch. This was done in order to reconcile the dimensions with a contemplated notched tensile-impact s pecimen.

0 TEST EQUIPMENT Stress -Strain Recording Equipment In the period covered by this progress report an automatic stress-strain recording system was installed on the Baldwin-Southwark tensile test machine. This system supersedes the use of the Martens optical extensometer system for tensile testing, although the Martens system continues to be used for creep tests, In addition, a strain pacer was also installed to permit greater accuracy and flexibility in conducting tensile tests at known strain rates. The recording system was made by the 00 S. Peters Company. The extension is measured by a linear variable transformer (or Microformer) which is part of a strain follower fastened to the gage section of the specimen. To permit the use of the strain recording equipment at elevated temperatures a transfer unit is under construction for transfer of motion outside the furntace. The transfer unit consists of a pair of specimen gripping screws fastened to a concentric rod and tube. The strain follower is to be attached to the bottom ends of the rod and tube. This device is also to be used for the measurement of extension in compression tests. It was originally hoped that this equipment would be ready for use during the period covered by this report, however, delay was encountered in obtaining tungsten carbide points for the specimen gripping screwso These are necessary because of the high hardness of the 17-7PH material. In addition, some difficulty was encountered with slack in the transfer unit, It is anticipated that these difficulties will be resolved shortly Since no compression tests have been run, the discussion of the fixture for this test will be deferred to a subsequent report.

9 Tension-Impact Test Equipment An existing Olsen impact testing machine was modified to permit carrying out tension-impact tests on sheet specimens. The modifications included the construction of a pair of specimen holding jaws that could be attached to the pendulum of the impact machine and the extension of the striking surfaces so that the impact occurred at the maximum downward point of the pendulum swing, In addition, it was necessary to re-calibrate the scale of the machine to obtain true values of impact strength under the test conditions employed. The design of the specimen holding jaws is shown in Figure 3 and a schematic representation of the test set-up is shown in Figure 4, The specimen holding jaws were modified from the design of Muhlenbruch (ref 4). The grip assembly was made to screw into a holding assembly fixed to the penduallm heado The specimen itself is held in the split grip sections. A cavity was machined into one side of the split grip to accommodate the specimen. The other half of the split grip has a plane surface and is fixed into position by an indexing pin located at the back end of the grip. The two halves of each grip are locked together by a snap ring, Stiffening rods fixed to the fr9nt holder and allowed to "float" in holes in the rear holder reduce the bending tendency of the assembly as the pendulum falls, The striker, which is screwed onto the rear holder provides the necessary tension impact loading as it hits the striking plates of the machine, Because of the location of the specimen holding assembly, it was necessary to add an auxiliary set of striking surfaces to the existing impact machine base. This modification is indicated in Figure 4. The location of the striking plates was fixed such that the distance L on Figure 4 was the same for the pendulum head and the striking surface in relation to the center of rotation of the pendulum

10 arm, This arrangement permitted the impact blow to take place at the maximum downward point of the pendulum swing. The impact data can not be read directly from the scale of the machine, because of the added weight of the grips and because the rear portion of the grips falls from the head immediately after the impact blow. The scale of the machine can be used, however, in the sense that it indicates the height to which the pendulum rises on the up-swing. This is accomplished by an arm concentric with the pendulum arm and one-fifth of its length. This arm contacts a sine-curve rectifier that slides on a pair of vertical rods. The rectifier in turn contacts an indicating disk which provides a reference point on the scale of the machine. The height to which the indicating disk ii pushed above its rest position is directly proportional to the height of the pendulum head. A measurement of the scale showed that 0. 73 inches corresponded to 20 scale divisions. It was then possible to obtain an expression that would convert scale divisions to pendulum height, since a 5:1 ratio exists between the lengths of the pendulum arm and the indicator arm. The expression follows: H' = S (0. 73)(5) = (0. 0152)(S) where H' is the pendulum height in feet after impact, and S is the number of scale divisions traversed by the indicating disk. Using the terms indicated on Figure 4, the kinetic energy just before impact is E - WH - aF p where E = initial energy, ft-lb H = initial height of pendulum center of gravity,ft. W = pendulum weight, lb a = angle of rotation on downswing, degrees F = friction loss per degree

11 and the final energy just after impact is Ef = WH + PF where E = final energy, ft-lb H' = final pendulum height, ft p = angle of rotation on up-swing, degrees The difference between these terms is I, the energy absorbed by the impact, = Ep Ef = WH aF WH - PF I = W(H HIVt - F (a + 3) The calibratiLon of the machine consisted of the determination of the friction loss and the effective initial height of the pendulum head. Because of the irregular shape of the pendulum head it was necessary that the initial height of the center of gravity of the pendulum be calculated from the characteristics of the machine. The basis of the calibration was the fact that where no specimen was fixed onto the pendulum head, there would be no impact energy absorbed. Thus, the difference in heights would represent the friction loss of the machine. The equation then reduced to the following: I = = W(H - HI) F (a + P) W(H H') = F (a + 3) The pendulum was removed from the machine and weighed. This was found to be 57~ 5 pounds, The side plates and front specimen holder were found to weigh 5 pounds and the complete specimen holding assembly was found to weigh 7 7 pounds, Using these various pendulum weights, the assembly was allowed to fall from the initial height and the final height was determined,

12 Weight S (avg) H' Angle of Rotation (a + 3) W = 57.5 117 1.78 ft. 151~ W+ 5 = 62.5 122 1.85 1530 W + 7.7 = 65.2 123.5 1.88 154~ The angle of rotation was determined by laying out the various distances on a sheet of paper and measuring with a protractor. This procedure was approximate and possibly an average angle could have been used. From these data, a series of three equations could be written. Since only two unknowns were present it was possible to use the third equation as a check of the accuracy of the determinations. (57.5)(H - 1.78) = 151 F (62. 5)(H - 1. 85) = 153 F (65. 2)(H - 1. 88) = 154 F Solving these equations leads to values of F = 0. 40 ft-lbs per degree for the friction loss, and the effective initial height H = 2. 82 feet. With these values known, it was then possible to compute the equation for the actual testing conditions of 65. 2 pounds initial pendulum weight and 62. 5 pounds final weight. (This comes about since the rear grips fall from the assembly after impact. ) With the aid of a cord attached to the rear grips it was possible to simulate the operation of the equipment without a specimen being present. Thus, the rear grips were pulled from the assembly at the maximum downward position. In this case H! was found to correspond to 124 scale units. A value of 155~ was used for the angle of rotation and the calibrating equation was written as follows: I = O = W1H - W2H' - F (a + 3) I = (65. 2)(2. 82) - 62. 5 (.0152S) - 0.40 (155) I = 122 - 0. 95 S (1) where I is the impact energy absorbed inft-lbs S is the scale difference

13 This equation was then plotted so that values of impact energy could be read directly. TESTING PROCEDURES Tensile-Test Procedure The use of the automatic stress-strain recording equipment has changed tb some extent the tensile test procedure described in the First Progress Report, (ref 5) The principal difference is the fact that a constant and accurately known strain rate can be maintained throughout the test with the aid of the strain pacing equipment. At the request of Mr. E. Lo Horne of WADC, a strain rate of 0, 005 inches per inch per minute has been adopted for all tests. The data are obtained in the form of a continuous curve of load versus deformation. From this curve are calculated the 00 2 percent offset yield strength and the elastic modulus. In addition, the maximum load can be observed from the recorder trace. Tension-Impact Test Procedure The only tension impact tests run to date have been those at room temperature, Prior to assembling the specimen in the holding grips (see Figure 3) measurer ments are made of the shoulder to shoulder distance and the cross sectional area of the specimen gage section. The actual running of the test is very simple. The specimen grip assembly is screwed to the holder at the back of the pendulum head which has been previously raised to the fixed initial height. The latch holding the pendulum is released and the head allowed to fall between the striking surfaces fracturing the trailing specimen, The final pendulum height is obtained from the distance which

14 the indicating disc rises on the impact machine scale. This value is converted to energy absorbed using the equation (I), page 12. Measurements of the specimen elongation are made on the fractured specimen. Creep-Exposure Test Procedure Creep exposure tests run under load are conducted using the procedure discussed in the First Progress Report. (ref 6) For the tests carried out under no load, i.e., temperature exposure only, the same general scheme is used. That is, all the steps of a normal creep test are followed with the exception of the loading. The two duplicate specimens for each unstressed exposure are wired together and run in the same unit at the same time. The slight additional thickness of the material has no effect on the temperature distribution within the furnance. For tests in which more than a few per cent of creep are to be measured, excessive deformation of the gauge section makes impractical the direct attachment of extensometer to a uniform-area test length. Fastening the extensometer from the specimen shoulders allows deformation measurements to be made during the entire test, all the way to rupture. But in this case, due allowance must be made for the elastic and plastic deformations in the fillet and shank portions included between the points of extensometer attachment. Corrections for these deformations outside the uniform gauge length are covered in the next section. After sufficient time-deformation data were available from tests in which extensometers were fixed from the specimen shoulders, stressed exposure tests with limited total deformations could be run with extensometers attached to collars clamped directly onto the gage section of the specimen. This procedure which minimizes errors in computation of deformation has proved to be satisfactory in tests with 2% total deformation and can presumably be used in all creep-exposure tests of the present program.

15 Calculation of Effective Gage Length for Extension Measurements Creep test data for the preliminary studies of total deformation were made with extensometers located between collars attached to pin holes on the shanks of the specimens. Thus, the deformation recorded included not only the extension in the gage section but also in the fillets and shanks. Consequently, it was necessary to calculate an effective gage length for these specimens that would take into account the deformation of the fillets and shanks, Two different corrections were needed, depending on whether elastic or plastic defo.Ermation was being measured. In general, the elastic correction was used for loading deformations or short time tensile tests, while the plastic correction was used for creep testso The effective gage length for the elastic case was calculated from the specimen geometry alone, A sketch of a typical specimen is presented in Figure 5. This drawing defines the various terms used in the calculation of the effective gage length, The measurements made on a typical specimen were the widths at equal increments of distance from the fillet shoulder to the actual gage section, The distance "increments used were 0. 050 inches and the widths are tabulated in column (b) of Table 1 In the elastic case the strain in any section is inversely proportional to the area, so that the efective gage length for each short interval was simply the interval length multiplied by the ratio of the relative area of that interval to the area in the gage section, For a sheet specimen, the relative areas (columnn g) were merely the interval widths, The ratio between this width and the gage section width was multiplied by the interval length (column h) and summed to give the total contribution of the fillet. This was doubled to account for both fillets. The equivalent shank length was the total shank length multiplied by the ratio between the gage section width and the shank width. The shank length is the pin to pin

16 distance minus the shoulder to shoulder distance. The effective gage length then was calculated by subtracting the actual fillet lengths from the shoulder to shoulder distance and adding the effective lengths of the fillets and shanks. This is indicated in Table 1 and was equal to SS + 0.45" for the particular specimen design under consideration. For the calculation of the plastic case, it was necessary to have information on the stress-creep rate relationships for the material at the test conditions. This information was obtained from the test data themselves. Examination of the rough time-elongation data showed that the minimum deformation rate occurred at about 25 percent of the rupture life. A calculation was made of the minimum creep rate in scale units (AL + AR) per hour at the time corresponding to 25 percent of the rupture life for that particular stress. A plot of stress versus creep rate was made from these data and is presented in Figure 6. For the case illustrated in Table 1 a gage section stress of 110,000 psi was considered for a test at 800~F. From the specimen geometry, the stress was then calculated at each previously measured interval in the fillet. The stresses are tabulated in column (c). The creep rate at 800~F corresponding to each stress was obtained from Figure 6 and recorded in column (d) and finally the average creep rate was calculated for each 0. 050 inch interval of fillet length, The effective length for each 0. 050-inch interval was then calculated by multiplying the length by the ratio of the creep at that particular stress to the creep rate in the gage section (i. e., at 110, 000 psi). This calculation is indicated in column (f) and was then doubled to account for both fillets. The equivalent length of the shank was computed by multiplying the shank length by the ratio of the shank creep rate to the gage section creep rate. The effective gage length for creep at 800~F was then expressed in terms of the shoulder to shoulder

17 distance by first subtracting the actual fillet lengths and then adding the calculated effective lengths of the fillets and shanks. The value obtained for 800~F was SS - 0 905". The values for 600~ and 900~F obtained in a similar manner are as follows: EGL600oF = SS - 1.01" EGL = SS - 0. 83" RESULTS AND DISCUSSION Tensile Test Results - 17-7PH Stainless Steel The results of ten tensile tests run at room temperature on samples of the 17-7PH precipitation hardening stainless steel in the TH1050 condition are summarized in Table Six of the tests were run on samples from sheet 1 and two each were run on samples from sheets 2 and 3, The data indicate that the samples from sheet 1 were somewhat stronger and less ductile than the samples from sheets 2 and 3. There was little difference between the hardnesses of the samples from all three sheets.'The average properties were calculated for each sheet and for all ten tests. Since six specimens were run from sheet 1 and only two each from sheets 2 and 3, a second set of averages, that of the average of the sheet averages, was calculatedo The reason for computing this average was to reduce the bias that would result from the simple average of all ten tests. The basis for this was that an average of all sheets should not be weihted 60 percent with values from sheet 1, These values show good agreement with those given by Armco Steel Corporation (ref 2) as "typical"' and "minimum" properties and also with the results noted by Brisbane (ref 7) for "high strength" heats of 17-7PH.,

18 The variation in properties appears to be principally concerned with sheet to sheet differences. The variations within the sheets areno larger than the variation between sheets. Creep-Rupture Tests of 17-7PH Alloy The results of creep-rupture tests run on the TH1050 condition of the 17-7PH stainless steel at 600, 800, and 900~F are summarized in Table 3. The primary purpose of these tests was to establish curves of time -versus total deformation to aid in the selection of stresses for the stressed exposure tests. The tests were allowed to run to rupture since the scatter in rupture properties is another measure of the inherent variability of the material. The data obtained from these tests are presented in graphical form in Figure 7 as curves of rupture and total deformation time versus stress for the three temperatures under consideration. The rupture times and fracture elongation show good consistency at each of the test temperatures. Hardness measurements were also taken on the specimens after the completion of testing and show that a moderate increase in hardness resulted from the exposure to time, stress, and temperature. Six random hardness impressions were taken on each sample, three in the shanks and three in the gage section. No significant differences in hardness were noted between the shanks and gage section, A statistical analysis of the hardness date indicated that the hardness change at all test temperatures was significant with respect to the as-treated (base) condition, but that there was no significant difference between the average hardness changes for 800~ and 900~F exposures, while the difference between the 600~F exposures and the other two temperatures was on the borderline of significanceo In this analysis, the effects of exposure time were neglected — the basis of comparison was exposed versus unexposed specimens hardness.

19 A similar analysis, taking into account the effects of time, temperature, and deformation, will be made on the data from the regular exposure tests, Curves of stress versus total deformation were established at each of the test temperatures from the available data. At 600~F, none of the test stresses used was low enough to give less than 0. 5 percent loading deformation. Consequently, additional tests will be required at this temperature and one or two low stress tests remain to be run at the two higher temperature, In order to fit curves to the available time-deformation data it was necessary'o allow for slight breaks in the curves. In general, these curves tend to follow the same slopes as the rupture curves at the same temperature, A check of the data was possible by a comparison with the results of Hanlon, Salvaggi, and Guarnieri (ref 8) for tests at 800~F. Figure 8 compares the two investigations The agreement appears to be good. The stresses estimated from experimental data at 600~, 800~ and 900~F for the ac....evernent of specified total deformations in the time periods required in this investigation are presented in Table 4, The figures in parentheses were obtained by extrapolation of the total deformation curves. The stresses required for deformations at intermediate temperatures will be estimated by interpolation from these data. Effect of Re- Machining on Tensile Test Results In the First Progress Report (ref 9) the statement was made that after creep exposure the edges of the gage sections of the tensile test specimens would be remachined, This was done for two reasons. First, it was desired that the final tests measure the properties of the sheet material itself and not the particular specimen's edge effects. Second, it was desired that the practice for subsequent tensile test specimens conform to that adopted for the tension-impact and

20 compression specimens in that these specimens were designed to be machined from the gage section of an exposed creep specimen. Consequently, the creep-exposure specimens were originally machined to a 0. 530 inch wide gage section. Following the exposure, the specimens were ground to a final gage section width of 0. 500 inch prior to tensile testing. The question did arise as to what effect, if any, the subsequent remachining operation would have on the tensile properties of an exposed specimen. Advantage was taken of a specimen that had been originally intended for exposure at 700~F and 120, 000 psi to reach 2 percent total deformation in 100 hours. This specimen 2R-T2 was given the standard remachining operation after exposure. Another specimen 3Q-T6 was given the identical exposure and then tensile tested without remachining; i. e., with 0. 530 inch wide gage section. The results of these tests are as follows: Total Ult. Tensile 0. 2% Yield Elong. Red. Area E x Hardr S-icimen Remachined Def-% psi psi %o-2 in. % 10 s R"C' (2R-T2) 2 54 219,000 218,000 2.2 13.7 30.6 44.0 Not Remachined (3Q-T6) 1.95 220,000 219,000 3.5 13.8 29.6 44.2 The result of these two tests indicates that the remachining operation has no significant effect on the tensile properties. This conclusion should be checked further at other conditions. Exposure Test Results —17-7PH Alloy Results of twenty-one tests of the effects of exposure to elevated temperature on the room temperature tensile properties of the 17-7PH alloy in the TH1050 condition are reported in Table 5. Eighteen of these tests were intended to study the effects of time and temperature alone. In addition, three tests have been completed that add the factor of creep deformation to the evaluation.

21 The tests after unstressed exposure were run for 10, 50 and 100 hours at temperatures of 600~, 800', or 900 F, Two specimens from different sheets were run for each exposure condition and the results are reported as averages, The average of the sheet averages of the as-treated material was taken as the base condition for compari.:-g propertieso Statistical analysis of the scatter obtained in the as-treated material indicated that changes in tensile and yield strength of the order c: 12,000 to.15, 000 psi would be significa.,-:rt A plot of the average effects of the unstressed exposure conditions on. the tensile str eegth, elongation, and hardness is presented in Figure 9 In co:rnjunction with Table 5 these results indicate that the unstressed exposure at 600'F had little effect o: the tensile properties of this material,, The 10=hour exposure at 800' or 900'F also had little effect, however, the 50 and 100 hours exposures at 800 and 900'F did have a significant effect on the room temperature properties0 The maxima..um efftect at 9003F appears to have been reached at a time less than 100 hours, At 800F the time to reach a maximum may be somewhat over 100 hours, although the rate of increase in strength appears to have dimLinished after the initir l rise for an exposure of 50 hours, The hardness data confirrn the trend indicated by the tensile strength0 The data covering the effects of stressed exposure are as yet sparseo As agreed in a conference with a representative of the Materials Laborat ory, WADC it is inteded to survey the results of exposure to 2 percent total deformation in 100 hours 4fo this rnaterial at 50'@F temperature intervals between 600~ and 900 F. One test has been completed at 600 F and the two tests at 800 F are complete. In addition, two tests were peformed at 700 F to check the effects of remachining of tensile specimenso These tests are not reported in Table 5 since the stress used (which was interpolated from the survey total deformation data) resulted in higher deforrnation than was desired, These test results were noted on page 20,

22 The test result from the 600'F stressed exposure indicates that the effect of creep deformation may change properties appreciably where unstressed exposure to temperature did not. Of particular interest is the loss of ductility, the increase in the tensile strength, and the increase in the yield strength relative to the tensile strengtho At 800'F the effect of stressed exposure to 2 percent total deformation was similar but not quite as marked. The 700~F stressed exposure also resulted in increased strength over the base properties and an increase in the yield strength relative to the tensile strength, Further discussion of these results will be deferred until the data are more complete. Tension-Impact Test Results -17-7PH Alloy The results of six tension-impact tests run at room temperature on the TH1050 condition of the.17-7PH stainless steel are reported in Table 6, These results inuclude two specimens each from the three sheets sampled for the determination of the scatter in normal properties of the as -treated material. Two to three more tests from different locations in the same sheets remain to be run, The data reported are the energy absorbed on impact and the elongation of the specimen after fracture. The agreement between the duplicate specimens in each sheet is good, however, there appear to be some scatter in properties between the sheets themselves, The additional tests to be run will enable an estimate to be made of the sheet-to-sheet scatter, Tensile Test Results -2024-T86 Table 7 summarizes results of three room temperature tensile tests run to data on the 2024-T86 aluminuam alloy.

23 One specimen from each sheet has been tested. Seven more specimens will be tested; two each from sheets 2 and 4, and three from sheet 3. The three specimens so far tested show good agreement in properties. The variation in reduction of area is not too significant and is probably due to the difficulty in obtaining such measurements on sheet materials. Further discussion of the tensile properties of this material will be deferred until the results of the other seven tests are available. FUTURE WORK During the next two-month period efforts will be made to complete the extensometer fixture for elevated-temperature tension and compression testing. Priority will be given to determination of scatter in room-temperature tensile properties of the 2024-T86 aluminum alloy, after which tests can be started to cover effects of unstressed exposure and to determine the proper stresses to give the desired total deformation values, Until cr, ep frames are needed for the aluminum alloy tests, creep exposures to give 2 percent total deformation at 100 hours will be continued for 17-7PHo Tension-impact tests with both alloys are scheduled to be initiated during the coming work period. Compression tests will be delayed until the completion of the extensometer fixture.

24 REFERENCES 1. Gluck, Voorhees, Freeman, "First Progress Report to Materials Laboratory, WADC," on Contract AF 33(616)-3368, "Effect of Prior Creep on Mechanical Properties of Aircraft Structural Metals" - page 2 - 4. 2. Armco Steel Corp., Product Data Bulletin on Armco 17-7PH Steel, March 1, 1954. 3. ref.i pages 8 - 9. 4. Muhlenbruch, "A Tension-Impact Test for Sheet Materials" A.S. T. M. Bulletin No. 196, page 43, February 1954. 5. ref 1 - page 11. 6. ref 1 - page 10. 7. Brisbane "Mechanical Properties of 17-7PH Stainless Steel" WADC Technical Note 56 - 169, page 7, April 12, 1956. 8. Hanlon, Salvaggi, Guarnieri "Intermittant Stressing and Heating of Aircraft Structural Metals" Bi-monthly Progress Report to Air Research and Development Command, Contract AF 33(616)-2226, page 7, June 30, 1955, 9o ref 1 - page 4.

Table 1 Example Calculation of Effective Gage Lengths for Strip Specimens Gage Length for Creep at 800~F Elastic Gage Length (a) (b) (c) (d) (c) (f) (g) (h) Distance from Spec. Widths Stress Creep Rate Avg. Creep Avg. Creep (e) Rel. Area Area in GS Length of Shoulder inches inches psi AL+ AR/hr Rate per. 050" Creep at Center of Interval (g) Interval per.050" 0 1.0050 55,400 0. 025 1. 005.0255 0. 050 0. 9236 60,300 0. 066 0. 0455 0.001.924.0274 0. 100 0.8466 65, 800 0. 095 0.0805.002.847.0299 0.150 0.7790 71,500 0.210 0.1525.004.779.0326 0. 200 0.7210 77,300 0. 470 0. 340.008.721.0351 0.250 0.6707 83,100 1. 1 0.785.019.671.0377 0.300 0.6278 88,800 2.2 1.65.039.628.0403 0. 350 0. 5916 94,200 4.8 3.50. 083. 592. 0428 0.400 0.5622 99,000 9.5 7. 15.170.562.0450 0.450 0. 5387 103,500 17.0 13.25.316.539. 0470 0.500 0.5208 107, 000 27.0 22.0.524.521.0485 0.550 0.5094 109,400 38.0 32. 5.774.509.0498 Sum 1.t'40- Sum.4616 =.46 Center 0. 5064 110,000 42. 0 --- 0.506 PLASTIC CASE - 800~F ELASTIC CASE Effective Length 2 fillets = 2(1. 940)(. 050) Effective Length 2 fillets = 2(. 46)=. 92" = 0. 194" Shank Length = pin-pin minus shoulder-shoulder Shank Length = 1. 27" = 4.82 - 3.55 = 1.27 Equivalent Length of Shank: Equivalent Shank Length: 1.27 (Creep Rate in Shank) = 1.27 (.025) = 0008" (Gage Section Area) = 1. (.506 (Creep Rate at Center) ('-) 1.27 ( Shank Area ) (1I Total Equivalent Length: Shank plus Fillets Total Equivalent Length: Shank plus Fillets.0008 + 0. 194 = 0. 195" =.92 +. 63 = 1. 55" EFFECTIVE GAGE LENGTH = SHOULDER TO SHOULDER - ACTUAL FILLET + (EFFECTIVE SHANK PLUS FILLETS) EGL 80OF = SS - 2(. 55) +.195 EGL = SS - 2(. 55) + 1. 55 = SS -.905" = SS +. 45" Similarly EGL =SS- 1.01" (600OF) EGL = SS- 0.83" (900 ~F)

Table 2 Room Temperature Tensile Data 17-7PH Alloy (TH 1050 Condition) E 6 Spec. Ult. Tensile 0. 2% Yield Strength Elongation Reduction of x 10 Hardness Loc. (psi) (psi) (% in 2 in) Area (%) psi (R'"C") 1C-T1 212,000 208,000 4.2 16.9 29.0 43.9 1L-T1 209,000 196,000 6.0 13,.9 28.8 42.8 1L-T2 214,000 196,000 5.5 14,.3 29.0 43.5 1L-T3 216,000 210,000 5.6 9.8 28.7 43.8 1L-T4 215,000 208,000 4.5 13.4 28.6 44.9 1U-T2 211,000 207,000 4.0 17.9 28.6 43.4 Average 212,833 204,160 5.0 14.4 28.8 43.7 2J-T3 197,000 186,000 9.0 19.4 27.6 43.4 2N-T3 187,000 173,500 8.8 21.7 29.3 42.3 Average 192,000 179,750 8.9 20.6 28.4 42.8 3G-T1 201,000 195,500 7.8 18.7 28.2 45.1 3Q-T1 202,000 195,000 7.2 20.8 29.2 44.3 Average 201,500 195,250 7.5 19.8 28.7 44.7 Avg. of 206,400 197,500 6.3 16.7 28.7 43.7 10 Tests Avg. of 203,100 193,050 7.1 18.2 28.6 43.7 Sheet Avg. Properties Reported by Armco Steel Corp. (ref 2) Typical 200,000 185,000 9.0 43 Minimum 180,000 150,000 6.0

Table 3 Rupture and Total Deformation Data 17-7PH Alloy (TH 1050 Condition) Time to Reach Indicated Total Deformation (hours) Spec. Test Temp Stress Rupture Time Elongation Reduction of Hardness after Loading 0. 5% 1. 0% 2.0% 3.0% Loc. (~F) (psi) (hours) (% in 1 in) Area (%) test (R"C") Def. % IK-T5 600 180,000 0.1 4.0 13.1 44.4 -- --- --- --- --- IK-T2 180,000 on load 4.0 0 42.3 -- --- 1U-T1 175,000 15.1 10.0 17.5 45.5.80 ---.05 1.6 4. 1 IC-T3 170,000 11.9 7.0 15.0 44.7.75 ---.4 2.5 5.7 IG-T1 170,000 42.6 9.0 20. 5 45.8. 72 --- 4.0 4.6 12.0 1L-T6 165,000 98.8 12.5 20.5 44.3. 69 -- 0.6 17.5 32.0 IC-T5 160,000 661.2 16.5 26.0 47. 7.64 --- 8.5 71.0 152.0 ZE-T1 150,000 in progress. 59 --- 46.0 330.0 2E-T5 125,000 in progress approx. 30 approx 1000 1G-T5 800 105,000 37.3 22.0 37.0 46. 3. 47. 05 0.3 0.8 1.3 3A-T6 100,000 61+3 43.0 44.0 4b. 8.41 0.1 1.9 6.4 10.2 IU-T3 100,000 107T5 32.0 44.0 46.0.41 0.1 2.3 8.5 16.0 1C-TI 95,000 179. 1 26.0 36.5 47.2. 38 0.4 4.0 14.0 25.5 1U-T4 90,000 555.6 37. 5 48.5 48.2. 35 0.5 8.0 29.0 52.0 IQ-T?5 70,000 in progress. 26 7.0 90.0 850.0 --- 1U-T5 900 75,000 19.8 31.0 46.5 46.9.32 0.10 1.0 2.9 4.6 1K-T4 70,000 29.1 28.0 49.0 46.1.29 0.20 1.2 3.4 5.7 1K-T6 70,000 56.7 38.5 52.8 45.9.30 0.25 1.4 4.6 7.7 1L-T5 70,000 24.5 29.8 48.6 45.6. 31 0.15 0.8 2.4 4.3 1C-T2 70,000 28.4 32.2 47.2 46. 8.30 0.17 1. 1 3.2 5.2 1K-T3 60,000 152.2 33.0 65.0 45.7.26 0.8 4.4 13.5 24.4 1G-T4 55,000 323.6 43.0 51.0 46.8.24 1.5 5.0 23.0 45.5 1C-T2 50,000 699.1 40.0 49.5 46.4.21 4.0 17.0 76.0 134.0

Table 4 E stimated Stresses for Given Total Deformations 17-7PH Alloy (TH 1050 Condition) Stress (psi) to reach stated total deformation in following time periods Temp Total Deformation Time - hours (OF) (To) 10 hrs 50 hrs 100 hrs 600 0.5 -- -- -- 1.0 159,000 150,000 (146,000) 2.0 167,000 161,000 157,000 3.0 170,000 164,000 161,000 800 0.5 (73,000) (63,000) ( 59,000) 1.0 88,000 (76,000) ( 70,000) 2.0 98,000 86,000 81,000 3.0 102,000 90,000 85,000 900 0.5 (42,000) -- -- 1.0 53,000 (40,000) 46,000 2.0 63,000 51,000 49,000 3.0 68,000 56,000 51,000 ( ) extrapolated

Table 5 Effect of Exposure on Room Temperature Tensile Properties 17-7PH Alloy (TH 1050 Condition) Spec. Temp Stress Time Total Def. Ult. Tensile 0.2% offset Elongation Reduction of Area E x 10 Hardness Loc. (~F) (psi) (hr) (%) (psi) Yield (psi) (% - 2 in) (%) psi R"C" 2E-T4 600 None 10 None 192,000 183,000 8.5 20.0 30.2 40.9 3P-T3 " " " " 2 09,000 203,000 6.5 19.7 30.4 43.4 Average 200,500 193,000 7.5 19.8 30.3 42.2 1P-T21 600 None 50 None 193,000 186,000 10.5 18.3 30.4 42.0 3L-T5 " " " " 1 87,000 180,000 9.0 22.6 31.2 41.6 Average 190,000 183,000 9.8 20.5 30.8 41.8 2N-T4 600 None 100 None 192,500 184,200 8.0 18.4 29.442.5 3A-T1 " " " " 2 12,000 206,000 7.5 19.4 28.8 45.8 Average 197,250 195,100 7.8 18.9 29. 144.2 2N-TI 600 157,000 100.5 1.89 223,000 (223,000) 1.5 13.7 29.4 45.3 3H-TI 800 None 10 None 201,000 195,000 7.5 18.7 30.8 43.8 2R-T3 " " " " 192,000 187,000 5.0 16.9 29.8 41.9 Average 196,500 191,000 6.2 17.8 30.3 42.8 2S-T2 800 None 50 None 218,000 213,000 4.9 15.1 31.2 43.6 1Q-T23 " " " "220,000 215,000 6.0 17.9 30.4 45.5 Average 219,000 214,000 5.4 16.5 30.8 44.6 3G-T6 800 None 100 None 222,000 216,000 4.0 17.9 29.8 47.4 2N-T2 " " " " 222,500 216,000 5.0 17.3 30.2 48.2 Average 222,250 216,000 4.5 17.6 30.0 47.8 3P-T1 800 81,000 102.6 2.12 232,000 229,000 3.5 12.1 29.9 46.7 2S-T6 " " 102.1 1.88 227,000 222,000 4.2 15.8 30.2 46.7 Average 2.0 229,500 225,500 3.8 14.0 30.1 46.7 IP-T23 900 None 10 None 205,500 200,000 6.5 17.2 30.2 43.1 3A-T5 " " " " 206,500 201,000 5.0 18.2 30.8 44.2 Average 206,000 200,500 5.8 17.7 30.5 43.6 2S-TI 900 None 50 None 230,000 224,000 3.5 14.1 30.9 46.2 3H-T6 " " " " 234,000 228,500 5.0 13.7 30.6 47.0 Average 232,000 226,500 4.2 13.9 30.8 46.6 3P-T6 900 None 100 None 222,000 215,000 4.5 8.4 30.4 46.2 2N-T5 " " " " 220,000 213,000 6.0 17.1 78.8 47.0 Average 221,000 214,000 5.2 12.8 29.6 46.6 As treated properties - average of sheet averages Sheets 1, 2, and 3 203,100 193,050 7.1 18.2 28.6 43.7

Table 6 Room Temperature Tension Impact Test Data 17-7PH Alloy (TH 1050 Condition) Spec. Energy Elongation Loc, Absorbed (% - 2 in) ft-lb IC-M5 35 2.5 1C-M6 37 4.0 Average 36 3.2 2J-M1 48 4.0 2J-M2 41 3.0 Average 44.5 3.5 3L-M2 52 5.6 3L-M5 52 5.6 Average 52 5.6 Grand Average 44.1 4.1 Table 7 Room Temperature Tensile Data 20 24-T86 Aluminum Alloy 6 Spec. Ult. Tensile 0. 2% Yield Strength Elongation Reduction of E x 10 Loc. (psi) (psi) (%i in 2 in) Area (%) psi 2P-T5 75,000 70,100 10.2 12.9 10.7 3E-T1 74,500 69,400 8.0 19.0 10.4 4A-T5 75,200 70,000 8.0 8.6 10.7

SHEET 2 SHEET 3 SHEET 4 PANEL NO. R 0 A R B Q R C R D Q E R F R 0Q G R H Q______ J R_ K Q L R _ M N _________ _________ _R P 0 20 SHEET SAMPLING SCHEME 2024-T86 SCALE,........ INCHES T ll I C I T I T TENSILE 21 X22 )M22 X2 I T2 | M 2 C COMPRESSION X 33 -331 T3 JC 3 X3 M TENSION-IMPACT M44 T4 X44 M4 X4 r4 X EXTRA T5 I M5 X5 PANEL SAMPLING SCHEME Q TIl I XI |T I ALL BLANKS X22 T2 X2 I INCH WIDE X 33 T 3 X 3 X44 T4 1 X4 T55 |X5 | T5 PANEL SAMPLING SCHEME R 0 10 SHEET SIZE 48 K 72. 065 -INCHES SCALE....... I PANEL SIZE 48" 5-5-1/2 INCHES INCHES SAMPLE CODE -(EXAMPLE) 4C-T21,.E.,SIEET 4 LENGTH ONLY PANELC -TENSILE SPEC. NO.2 Figure 1. - Sampling Procedure for s-;eets of &'C24-T86 alurnin.urrL alloy

I 1 RADIUS O 21-1/4\'3/" _ 2 314 1-1/4" SPECIMEN IS FULL SHEET THICKNESS -.064 INCHES Figure;. - Tension-impact.pecim.e. --.s mciified from previous design

Figure 3. - Tension-impact specimen and gripping assembly. The plates at left fit over the impact machine pendulum head

INDICATOR AND SCALE " i. j.! A. 5 \I * --- L EXISTING BASE OF | IMPACT MACHINE ADDED STRIKING PLATES AND SUPPORTS Figure 4. - Geometry of Olsen impact machine as modified for tension-impact testing

SCALE UNITS, AL+AR, MEASURED OVER THIS LENGTH?. EFFECTIVE GAGE LENGTH SHANK FILLET I GAGE SECTION i0 ~~,0 j'1 I" RADIUS SHOULDER TO SHOULDER, SS:3.55' PIN TO PIN. PP-4.82" Figure 5. - Gage section of typical strip specimen for illustration of calculation of effective gage length

175 1 600 ~F 150 125 0 - 900 ~ 100 800F 75O 50 25 o L I __I,I II I I I I I I I I III I --— 1-.01 0.1 1.0 10.0 100.0 Creep rate - scale units (AL + AR)/hour at 25% of rupture life Figure 6. - Stress versus relative creep rate at 600~, 800~, 900~F for 17-7PH alloy in TH 1050 condition

180 (4. 0) 170 - (9 600~F Rupture (16.5) ~12 8 160 c odee 150 0- rupture time Z ( ) elongation in rupture % 3 * 0.5% ") 600~F Total Deformation 140 A 1 % X 2% (1/ Total Deformation + 3% J 130 120 0 o 110 (22) 800~F Rupture l2) 100 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~(37, 5 90 0, 5 G/o 3 Lyo 800~F Total Deformation 1 "ib 2 %0 80 x + ~~~~~(31) 60 3 3) 50. 40 0. 1 1.0 10.0 100.0 Time - hours Figure 7. - Time for rupture and specified total deformations versus stress for 17-7PH alloy in TH 1050 condition

I~~~~~ I^^ Il I I li 110 10 30 _ 0 uptueture 90 80 P' 70 00 60 - 50 40 Solid Lines - Data of Hanlon, Salvaggi, Guarnieri (ref 8) 30 0 Rupture * 0.5 % A 1 % f Total Deformation University of Michigar Data Points 20 + % J10 0.1 1.0 10.0 100.0 1000.0 Time - hours Figure 8. - Comparison of University of Mic1igan data points with rupture and total deformation curves of Hanlon, et al for 17-7PH at 800~F

Tensile Strength 240 - - 900 ~F 220 800 — 200 180 160 1-4, 1 -i0 z 50 ~1 ~~00 ~Hardness 5 900 ~F 80 0- 2.. - z 8001n 600~F 60 40 o 0 Elongation o 40 - __ 600~0F -3 10 _ ~u) 800~F _ _ 0 ~F 20 — 0 ____________________________________________________,___0 0 10 20 30 40 50 60 70 80 90 100 0 Exp-osure Time - hours Figure 9. - Effect of unstressed exposure at indicated conditions on tensile strength, elongation, and hardness of 17-7PH alloy in TH 1050 condition

UNIVERSITY OF MICHIGAN 3 9015 03126 6037111111111 3 9015 03126 6037