1 This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a nonexclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan(http://energy.gov/downloads/doe-public-accessplan). 1
2 ISABE-2015-20242 Residual Stresses in Additively Manufactured Inconel 718 Engine Mount Thomas R. Watkins, 1 Paris A. Cornwell, 1 Ryan R. Dehoff, 1 Vinod Nangia, 2 Donald G. Godfrey, 2 1 Oak Ridge National Laboratory, Oak Ridge, TN 37831 2 Honeywell Aerospace, Phoenix, AZ 85034 Abstract As part of a demonstration project, the residual stresses were determined using neutron diffraction in a selected region of an IN718 engine mount that was aged, heat treated and hot isostatically pressed (HIP ed). The residual stresses were low in magnitude, except normal to the main flange. Crudely, the larger compressive and tensile residual stresses were on the exterior and interior, respectively, of the region examined. This is opposite what is normally observed in asbuilt components and indicates the influence of the post build thermal treatments. Nomenclature d interplanar spacing from the planes experiencing stress E Young s modulus (200 GPa) ν Poisson s ratio (0.29) λ wavelength θ half the scattering or Bragg angle, 2θ σ residual stress ε residual strain Subscripts 0 stress-free ii 11, 22, 33 refer to the orthogonal directions Introduction Honeywell Aerospace is creating components using additive manufacturing (AM) processes, because of the savings in time and cost relative to casting processes. Because these processes are relatively new, testing, modeling and validation are required to assure this process can produce components that meet the design specifications. One of the manufacturing challenges is the presence of residual stresses in additively manufactured components. Residual stresses can cause part distortion and/or cracking and as such need to be measured, understood and accounted for in part design. Today, there are several techniques to determine the residual stress in a metal, both destructive and nondestructive [1]. Although there are limitations, the neutron diffraction technique is uniquely qualified to measure residual stresses within the bulk of a sample, non-destructively. While simple prismatic shapes manufactured by Direct Metal Laser Sintering (DMLS) and electron beam melting have been examined using neutron diffraction,[2] a residual stress determination in a complex component manufactured by DMLS was sought. This project seeks to demonstrate the feasibility of neutron residual stress measurements in complex parts and provide insight into Honeywell s DMLS process. 2
3 Experimental While some of the experimental procedures are given here, more detail of work done similarly can be found in reference [3]. The engine mount sample was supplied by Honeywell and was mounted onto the XYZ stage of the goniometer. The positions of key points on the external surface of the samples were verified first optically with a set of transit-levels and then by neutron edge scanning from the air into the sample. The position of the sample surface was finalized using a non-linear fit to the intensity-position scan. The IN718 engine mount is shown in Figure 1 with schematically (not to scale) indicated measurement locations around the area of interest, the flange. Neutron scattering experiments were carried out on the Second Generation Neutron Residual Stress Facility (NRSF2/HB-2B) at the High Flux Isotope Reactor (HFIR) at the Oak Ridge National Laboratory (ORNL). The neutron wavelength and gauge volume were 1.733 Å and 2x4x2 mm 3 (LxHxW). The (311) and (220) reflections were examined at nominal Bragg angles of 106 and 86 2θ, respectively. Reference cubes were EDM cut from a sister sample and taken to have zero stress. The reference cubes have nominal dimensions of 5 x 5 x 5 mm or less and were scanned to obtain the d 0 values. In order to remove any runto-run variability, a single global d 0 was not used for the stain calculations. That is, d 0 was measured for each measurement direction (X, Y and Z) and then used to calculate the strain for the same direction. The total average strain within the gauge volume is determined from the shift in the interplanar spacings of the planes within the diffracting grains relative to the interplanar spacing of a strainfree reference sample. The measured strain is an average of the strains in the large number of diffracting grains within the sampling volume. Bragg s law relates the angular position of the diffraction peak to the interplanar spacing as follows: d = λ/(2 sinθ). (1) The residual strain component is related to the shift in interplanar spacing by: ε = (d-d 0)/d 0. (2) Assuming negligible plastic deformation and the shear stresses are zero, Hooke s law is used to then transform the strain measurements to residual stresses by: [4, 5] σσ 11 = EEεε 11 (1+νν) + νννν(εε 11+εε 22 +εε 33 ) (1+νν)(1 2νν) σσ 22 = EEεε 22 (1+νν) + νννν(εε 11+εε 22 +εε 33 ) (1+νν)(1 2νν) σσ 33 = EEεε 33 (1+νν) + νννν(εε 11+εε 22 +εε 33 ) (1+νν)(1 2νν) (3) (4) (5) Additionally, x-ray diffraction was performed to determine the phases present in the engine mount. Results and Discussion X-ray diffraction (XRD) revealed that the sample was predominantly γ-fcc phase (a=3.59 Å, see Figure 2). The residual stress distribution was determined twice in the area of interest (see Figure 1) using two different crystallographic planes, (311) and (220), which showed two different distributions for the same stress state. While the (311) and (220) distributions are crudely tensile and compressive, respectively (see Figures 3-6), similarities between the contours exist. The differences could not be accounted for with crystallographic anisotropy alone, and might include effects due to large grain sizes 3
4 (>0.1 mm) resulting from the HIPing. Large grains can skew the results by shifting the center of gravity of the gauge volume causing peak shift. This would be erroneously interpreted as a residual stress. In the literature, neutron data taken with using the (311) planes is favored given their approximate linear response to lattice strain when plastic deformation begins [6] and minimal sensitivity to intergranular or type 2 strains [7]. Further, the multiplicity of the (311) plane is twice that of the (220) plane (24 vs. 12 for a cubic material). Therefore the (311) results were taken as more representative. Figures 3 and 4 show a region of high compression towards the exterior of the mount and high tension in the interior, in the direction normal to the flange. This is opposite of what is normally observed in additively manufactured samples in the asbuilt condition,[7] suggesting that the heat treatment and HIP ing had an additional benefit. The stress differences the directions along, Y, and parallel, Z, to the flange radius are small relative to those normal. Finally, mechanical testing of this part was performed by Honeywell, and its performance was found to be satisfactory [8]. Summary and Conclusion It was demonstrated that the residual stress state in a select region of a complex part could be determined using neutron diffraction. The flange region of an IN718 engine mount, that was aged, heat treated and HIP ed, exhibited higher levels of residual stress perpendicular to the flange. In addition to densification, the heat treatments had an additional benefit of placing the exterior in a more compressive state than the interior, which is opposite what is typically observed in the as-built state. References 1. Watkins, T.R., G.S. Schajer, and M.J. Lance, Chapter 1.07 Residual Stress Measurements, in Comprehensive Materials Processing, C.J. VanTyne, Editor. 2014, Elsevier Ltd. NY: New York. p. 113 134. 2. Sochalski-Kolbus, L.M., et al., Comparison of Residual Stresses in Inconel 718 Simple Parts Made by Electron Beam Melting and Direct Laser Metal Sintering. Metallurgical and Materials Transactions A-Physical Metallurgy and Materials Science, 2015. 46A(3): p. 1419-1432. 3. Johnson, E.M., et al., A Benchmark Study on Casting Residual Stress. Metallurgical and Materials Transactions a-physical Metallurgy and Materials Science, 2012. 43A(5): p. 1487-1496. 4. Noyan, I.C. and J.B. Cohen, Residual Stress: Measurement by Diffraction and Interpretation. MRE: Materials Research and Engineering, ed. B.I.a.N.J. Grant. 1987, New York: Springer-Verlag. 5. Hutchings, M.T., et al., Introduction to the Characterization of Residual Stress by Neutron Diffraction. 2005, Boca Raton, FL: CRC Press. 6. Moat, R.J., et al., Residual stresses in laser direct metal deposited Waspaloy. Materials Science and Engineering: A, 2011. 528(6): p. 2288-2298. 7. Rangaswamy, P., et al., Residual stresses in LENS components using neutron diffraction and contour method. Materials Science and Engineering: A, 2005. 399(1-2): p. 72-83. 8. Purkhiser, M. 9NOV2011, Honeywell internal report. Acknowledgements Research efforts for TRW and RRD were sponsored by the U.S. 4
5 Department of Energy, Office of Energy Efficiency and Renewable Energy, Advanced Manufacturing Office, under contract DE-AC05-00OR22725 with UTBattelle, LLC. Research effort for PAC and neutron time at the Second Generation Neutron Residual Stress Facility (NRSF2/HB-2B) at the High Flux Isotope Reactor was sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy. XRD equipment purchased by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Vehicle Technologies Program, through the Oak Ridge National Laboratory s High Temperature Materials Laboratory User Program. 5
6 (A) (B) (C) (D) Figure 1 Inconel 718 engine mount with the measurement locations schematically shown. (A) Sample mounted on neutron goniometer. Stress-free cubes affixed to flange tip (0,0) with incident and receiving slits seen in the upper right and left corners, respectively. (B) Another view showing the stress-free cubes affixed to flange tip (note arrow) with incident and receiving snouts, which hold the slits seen in the upper center and left, respectively. (C) The -Z Y plane view. Note lines 1 and 2. (D) The X Y plane view, which corresponds to the horizontal plane in Figures 4-7.
Figure 2 The XRD pattern of the aged, heat treated and HIPed IN718 engine mount (Co Radiation, λ = 1.788965 Å). 7
8 Figure 3 The residual stress distribution for the (311) plane, line 1 (see Figure 1C). The X = direction normal to the flange, Y = direction along the flange, Z = direction parallel to the flange radius. Figure 4 The residual stress distribution for the (311) plane, line 2 (see Figure 1C). The X = direction normal to the flange, Y = direction along the flange, Z = direction parallel to the flange radius.
9 Figure 5 The residual stress distribution of the (220) plane, line 1. The X = direction normal to the flange, Y = direction along the flange, Z = direction parallel to the flange radius. Figure 6 The residual stress distribution of the (220) plane, line 2. The X = direction normal to the flange, Y = direction along the flange, Z = direction parallel to the flange radius.