Abstract
Early detection of problem conditions such as defects and cracks can be very difficult without damaging the actual material or requiring disassembly. The areas being inspected can be quite large and the inspection can be very time consuming. What technicians need is technology that is sensitive to a full range of possible defects and failure modes that also minimize subjective operator decisions and time on the ground. NDI needs are pervasive and cross-service, being required in a cross section of ships, airplanes and ground systems. This technology forum will discuss multiple NDI technologies, novel applications, and emerging trends.
Agenda
1300-1309: Welcome – Greg Kilchenstein (OSD-MR) Presentation
1309-1310: Administrative Notes – Debbie Lilu (NCMS)
1310-1330: Blade Scanner – Ryan Glembocki (FRCE NDI Engineering)
1330-1350: C-130 Propeller NDI Development / Assisted Defect Analysis / Augmented Reality for Nondestructive Evaluation – Eric Lindgren & Ryan Mooers (AFRL) Presentation
1350-1410: Underwater Inspections Using Drones and Robots– Craig Glennie (UH) Presentation
1410-1430: Large Standoff, Large-Area Thermography (LASLAT) Non-Destructive Inspection of Composite Structures – Maria Beemer (Thermal Wave Imaging) Presentation
1430-1450: Acoustic Weld Project – Anu Gupta (Microsoft)
1450-1500: Wrap-Up – Greg Kilchenstein (OSD-MR) Presentation
Minutes
Event: On 27 October 2020, the Joint Technology Exchange Group (JTEG), in coordination with the National Center for Manufacturing Sciences (NCMS), hosted a virtual forum on “Emerging Non-Destructive Inspection (NDI) Capabilities”.
Purpose: The purpose of this forum was to examine and share information on multiple NDI technologies, novel applications, and emerging trends. The forum provided descriptions of processes employed by the military Services as well as programs from both academia and industry.
Welcome: Greg Kilchenstein (OSD-MR) welcomed everyone to the forum and thanked the presenters and all the listeners for their attendance. He also stated how important inspections are in driving sustainment costs and readiness and how NDI innovations have a significant impact on the DoD maintenance community.
Administrative: This was an open forum. The presentations, along with questions and answers, were conducted through Adobe Connect. Four of the five presentations were also available online at the JTEG website at https://jteg.ncms.org/ . A separate audio line was used. We had over 100 participants from across DOD, industry, and academia join in the forum.
Blade Scanner – Ryan Glembocki (FRCE NDI Engineering) described an inspection technology using a fully automated process that can perform the H-1 Y/Z 25-hour inspection of the main rotor blade while decreasing inspection time from 4 hours per blade to 1 hour per blade, and providing the capability to digitally archive and track blade defects for trending and predictive maintenance. He described the components and a few concerns with scan length and data capture overload, and provided an overview of the evaluation plan. Depot standup is currently delayed due to equipment calibration issues, but they plan to enter into a CRADA to further refine the technology.
C-130 Propeller NDI Development / Assisted Defect Analysis / Augmented Reality for Nondestructive Evaluation –Ryan Mooers (AFRL) provided an informative brief on C130 inspection development to include field and depot level Eddy current inspections, automated Eddy current scanner, and a fast field level on-wing ultrasonic inspection. He described extensive use of 3-D printing for rapid prototyping and construction, and added that the automated Eddy current system provided a 2.5-5 times improvement in detection capability compared to standard pencil probes. Eric Lindgren (AFRL) discussed emerging nondestructive evaluation (NDE) for integrity management to include intelligence augmentation, and briefed augmented reality (AR) for NDI. The objectives of AR for NDI are refresh (critical but infrequent inspections), training (understanding the inspection), and guidance (higher fidelity coaching), all of which support the goal to enhance aircraft availability.
Underwater Inspections Using Drones and Robots – Craig Glennie (University of Houston) described the underwater inspection research ongoing at UH in three primary areas: 1) sensors and sensing systems, 2) localization in complex and cluttered environments, and 3)
Advanced Data Analysis (AI/ML). Underwater robotic sensors included a bio-inspired fish and bolted connection inspection. He demonstrated localization between robotic operating vehicles using triaxial coil antennas, and described a framework for automating inspections using machine learning. He concluded with a discussion of an ongoing U.S. Army Corps of Engineers project involving inspections and modeling of miter gates of navigation locks, with a key idea involving deep learning for image interpretation.
Large Standoff, Large-Area Thermography (LASLAT) Non-Destructive Inspection of Composite Structures – Maria Beemer (Thermal Wave Imaging) described LASLAT, which uses thermographic NDI to provide large area coverage and effective composite diagnostics, and provided a progress update. They are currently wrapping up a Phase II SBIR with FRC-SW with updated components and controls, and the development of temperature controlled thermography (TCT) which has several advantages to include better depth penetration. Additionally, they are working on two more projects: “ProjectIR” which uses a single LASLAT thermal projector and manual adjustments, and EagleIR which provides real time, automated surface and subsurface FOD detection. Maria finished with descriptions of the LASLAT demonstration on the RQ-4 Global Hawk fuselage, and the A320 rudder repair identification.
Acoustic Weld Project – Dawn White (Perisense) discussed a demonstration designed to show that welding defects can be discovered using acoustic sensors and artificial intelligence. Anu Gupta (Microsoft) described the experiment as the study of three types of defects: (1) Lack of shielding gas – introduced by gradually reducing the flow of shielding gas from 50cfh to 0cfh; (2) Slag inclusion – introduced by laying down a short bead of the weld prior to main welding; and (3) Lack of fusion – introduced by creating a void by previous passes. Sound generated by the welding process was recorded. Resulting soundtracks included both good weld conditions and welds with defects. A neural network was created and an ML model developed to detect weld quality. The demonstration has shown that welding defects can be discovered using acoustic sensors and AI, and also, that they were able to detect limits to which audio sensors and analytics can detect defects.
Q&A – A Q&A occurred after each briefer finished their presentation. Questions and answers will be posted on the JTEG website with these minutes.
Closing Comments: Greg Kilchenstein thanked the presenters for their contributions and all the work being done to support NDI efforts across the DoD, academia, and industry. He suggested continuing the information exchange beyond the forum and the importance of collaboration within the DoD maintenance community.
Action Items:
- All cleared briefing slides were posted to the JTEG website at https://jteg.ncms.org/ prior to the forum start.
- Obtain Distribution “A” level slides for the remaining brief.
Next JTEG Meeting: The next scheduled JTEG virtual forum is 24 November 2020, 1:00 – 3:00 pm EST. The topic is “Welding Inspection, Operations, and Training”.
POC this action is Ray Langlais, rlanglais@lmi.org, (571) 633-8019
Forum Q&A
Blade Scanner – Ryan Glembocki (FRCE NDI Engineering)
Q1. What is the blade material?
A1. Multiple materials. Majority is skin over composite, an aluminum spar, and the edge is fiberglass.
Q2. What is the NDI technology that this scanner is based on?
A2. Resident ultrasound
Q3. Is anyone else using or evaluating a version of this blade scanner?
A3. No
Q4. What NDI modality are you using? Phased Array Ultrasound? Eddy Current?
A4. Single element resident ultrasound. Four arms
Q5. What defects is this system able to detect? Any size or depth limits of detection?
A5. Goal is up to ¼ inch depth all skin delams.
Q6. How do you plan to determine the overall detection capability for the system with so many different material combinations?
A6. Use scrap blades and induce defects. Multiple different blade scanners.
Q7. What defects/modes is the scanner looking for? Corrosion, delamination? etc
A7. Delaminations and disbonds
Q8. What is the resolution of the technoloogy (layer thickness)
A8. Layer thickness index 100 thousands between each w. ¼ inch transducer.
C-130 Propeller NDI Development / Assisted Defect Analysis / Augmented Reality for Nondestructive Evaluation – Eric Lindgren & Ryan Mooers (AFRL)
Q1. How do you anticipate field testing quality control of inspections? May you please speak to utilizing your Eddy Current procedure and ways you intend on implementing consistent inspection processes across all depot facilities?
A1. It is only at one depot facility – Robbins AFB
Q2. What is the sensitivity (min flaw size) for the on-wing UT inspection?
A2. We have developed the minimums. We know it will be a rather large crack to identify and are still looking into it.
Q3. Could other (non USAF) activities participate in training and/or schoolhouse knowledge sharing for VAMRAM? I have inspectors and NDE experts that would benefit from this progressive initiative!
A3. Excellent question. It would have to be coordinated with the program office. We are open to it.
Q4. You mentioned a Developer’s Guide. Where will that be posted?
A4. It is a DTIC Document.
Q5. Are there any formal collaboration venues for NDI engineers and practitioners across DoD? How does USAF share with other Services?
A5. A community of interest exists. We also share amongst the program offices – ALC working.
Underwater Inspections Using Drones and Robots – Craig Glennie (UH)
Q1. What are some of the high thermal materials that are being used in the subsea applications?
A1. That’s not my area of expertise. Contact me and I will connect you to the proper people.
Q2. Could you use the magnetic field to detect non-comformaties in metal or corrosion?
A2. I don’t think it is sensitive enough to do that.
Q3. Is there any value in subsea touch probing for higher accuracy as done with divers?
A3. Well, it can be done just as easily with divers. The difference is that we want to be using the robots in areas that don’t want to send a diver.
Q4. Are you using the synthetically rendered data to train a computer vision model for underwater corrosion detection?
A4. No. The synthetic model is just for visualization afterwards. It is a tool if we have problems getting data.
Q5. And what are your thoughts about using swarms to scale a solution to stitch together the full inspection picture faster?
A5. That is definitely how we would like to go. i.e. we could inspect a large tank inside and outside at the same time.
Q6. Are you using any of the magnetic induction in combination with multi-beam sonar to increase accuracy of inspections?
A6. We haven’t done it yet.
Large Standoff, Large-Area Thermography (LASLAT) Non-Destructive Inspection of Composite Structures – Maria Beemer (Thermal Wave Imaging)
Q1. Maria, are you working with FRC?
A1. Covered later in the brief. FRC-E
Q2. At the Naval Surface Warfare Center, Port Hueneme Division, we have one of your large units. It becomes especially challenging trying to do thermography testing of combat system composite structures such as radomes and antennas. Some radomes are over 3″ thick, antennas are a complex layup of fiberglass, wires, RF reflective surfaces, cores, etc.. What has been TWI’s experience with combat and communication system Navy composite structures? This is something NSWC PHD will be pursuing now that we have a unit.
A2. We have experience inspecting a number of different radomes/antennas. Success of the inspection depends both the makeup (material stack and geometry) of the target and on the inspection requirements (e.g. defect type and size). Once those are known, we can determine whether an inspection is feasible and, if so, provide appropriate inspection parameters. Please contact us to discuss the details of your specific inspection.
Q3. Are you integrating AI and ML tools to assist the inspectors focus in on potential trouble areas?
A3. We have ML tools integrated into many of our systems used in manufacturing. In the maintenance space, both targets and inspection requirements are more variable. Thus, we have focused our ML efforts into providing a physics-based interpretation of that data that can be used for ML, customized to the need of the customer.
Q4. Does TWI offer any hands-on training courses?
A4. Yes, we offer hands-on training both at customer facilities and in-house at TWI. We also partner with Snell Infrared to provide IR NDT training courses.
Q5. What type of heat source would you use for carbon-carbon material?
A5. Choice of heat source varies depending on the material properties (i.e. thickness, thermal diffusivity) and on the inspection requirements. Typically when we inspect carbon-carbon materials, they are fairly thin, which requires that we use a truncated flash source (https://blog.thermalwave.com/advanced-flash-technology). However, depending on thickness and requirements, other sources could potentially be appropriate.
Acoustic Weld Project – Anu Gupta (Microsoft)
Q1. Have you used the system to detect levels of corrosion under military CARC painted structures?
A1. The short answer is no, but that’s not to say it isn’t worth an experiment. Probably need to dig in a bit on the use case.