Open Access
Brennan, Marissa C
Graduate Program:
Materials Science and Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
August 16, 2018
Committee Members:
  • Todd Palmer, Thesis Advisor
  • Allison Michelle Beese, Committee Member
  • Clifford Jesse Lissenden III, Committee Member
  • Jayme Scot Keist, Committee Member
  • Laser Ultrasonics
  • Ti-6Al-4V
  • Inconel 718
  • 17-4PH stainless steel
  • In situ
  • Directed Energy Deposition
  • Additive Manufacturing
Directed energy deposition (DED) processes commonly form lack of fusion defects when adjacent passes do not achieve consistent overlap. Insufficient fusion is rooted to the behavior of the molten pool which is influenced by high temperature thermo-physical material properties. Internal defects have the tendency to be detrimental to mechanical properties. Thus, expensive and time-consuming methods such as post-processing (HIP) are required to repair defects in additively manufactured (AM) parts, otherwise components are scrapped. Efforts to minimize post-processing has been redirected towards developing in situ monitoring techniques based on nondestructive evaluation (NDE) methods to detect defects during processing. Knowledge of defect location and size can then be used as a guide for correcting defects prior to removing the build from the chamber. While thermal and optical techniques have been extensively integrated for in situ measurements, information can only be obtained for the exposed build layer. However, techniques based on ultrasound are capable of detecting subsurface features. Noncontact, laser ultrasonics offers the unique capability of operating in-line and within close proximity to the DED head which makes it attractive for in situ monitoring of the build process. An assessment was completed for choosing laser ultrasonics as feasible tool for in situ monitoring. The feasibility of the laser ultrasonic system was evaluated by measuring the minimum defect diameter and maximum defect depth that could be detected for DED builds processed with intentional lack of fusion conditions. Defect measurements were then cross-correlated to optical microscopy and x-ray computed tomography (CT) which have been extensively used as acceptable techniques for characterizing defects in AM components. Intentional defects were formed upon selecting processing parameters to control melt pool shape and size. Lack of fusion defects were produced by altering hatch spacing, powder flow, and power to replicate defects typically formed during processing for titanium, nickel superalloys, and precipitation hardened martensitic grade stainless steel alloys. X-ray CT results reported the majority of defects were between 0.50 mm and 0.75 mm in diameter, despite the hatch spacing or material. In addition, sphericity values between 0.35 and 0.50 were reported for the majority of defects formed. A general decreasing power law trend was observed upon plotting degree of sphericity as a function of defect diameter, despite the processing conditions and material selected. This concluded the lack of fusion defects formed were non-spherical and elongated in shape. A closer look at the volume percent porosity reported the highest percentages for builds processed with 3.56 mm hatch spacing and disrupted powder flow. A comparison of material conditions, however, indicated while Ti-6Al-4V showed several instances of lack of fusion between passes, Inconel® 718 and 17-4PH stainless steel builds featured most lack of fusion near the ends of the deposition passes. Parabolic, scattered skimming waves defined in laser ultrasonic B-scan data identified lack of fusion defects along passes from a machined Ti-6Al-4V AM sample. These same signals were detected at processing speeds, however, the signal to noise ratio appeared to be significantly increased. In addition, an artificial EDM hole of 0.80 mm in diameter and approximately 1.50 mm below the surface of an as-deposited Ti-6Al-4V build was also identified as a scattered Rayleigh wave. This was used to compare the signal of a single lack of fusion defect, approximately 1.11 mm in diameter and 1.10 mm below the surface, reported for a Ti-6Al-4V AM sample with an as-built surface condition. Additionally, a preliminary assessment of using laser ultrasonics as a characterization tool was evaluated by measuring the attenuation across a polished Ti-6Al-4V AM and wrought titanium sample. Results indicated a higher attenuation in the as-deposited sample versus the wrought sample. Higher attenuation was thought to be influenced by the large the prior-β grains which averaged a grain width of 264.3 ± 79.8 µm as compared to the average grain diameter of 3.37 ± 5.08 µm calculated in the rolling direction of the wrought titanium. However, since the wavelength selected for measurements were between 100 μm - 6000 µm, feature sizes less than 50 µm were not accounted for in the data. Other factors such as melt pool boundaries and/or defects present in the material, likewise, may have contributed to some of the signal attenuation reported.