← Back to NASA Technology Projects
Completed TRL 1 (started at 1, targeting 3)
This project addresses the NASA needs for Advanced Manufacturing: Structures and Materials, focusing on novel process monitoring approaches to assist in efficient manufacturing verification. Particularly this proposed effort directly targets improving the robustness of the friction stir welding (FSW) process. The FSW process has been baselined for the fabrication of the fuel tanks and dry structures, such as the launch vehicle stage adaptor (LVSA), for the Space Launch System (SLS). As methods to reduce fabrication time and cost are implemented on the SLS program, the robustness of the FSW process must not be compromised. Correlation of the controllable process parameters with the output parameters can provide the basis for in-situ monitoring and ultimately in-situ process control. Digital Manufacturing refers to the analysis of the data captured from the instrumentation installed on the FSW equipment, combined with the theory, simulation, visualization, and analysis of the process. Thus, this novel process modeling approach provides the basis for achieving in-situ process control in addition to in- situ inspection for ensuring a robust FSW is produced. Ultimately this will reduce cost by decreasing development time and eliminating post-process inspection.
Currently post weld, ex-situ inspection techniques are used to verify the weld quality. However, in spite of passing non-destructive evaluation (NDE) testing, random variations are sometimes observed during destructive post weld testing of the FSW joint strength. Radisavljevic, et al., has reported a 13% variation in the strength of aluminum alloy FSWs which were not due to perceptible defects. In both these studies all specimens were verified defect free by NDE and displayed acceptable fracture surfaces in destructive testing. As random strength variations are difficult, if not impossible to detect, they can ultimately compromise the robustness of the process. As the tool and process parameters control of the regularity of the metal flow during the FSW process, it is proposed that interruptions in this regularity are responsible for the random variations observed in FSW strength. Interruptions in the regularity of the material flow are theorized to result from variations in the through thickness workpiece panel temperature distribution. Since the flow stress of the weld metal is influenced by the temperature fields, the force measurements are considered the most sensitive output parameter for monitoring the regularity of the process.
This proposal will collect data during FSWs of varying thickness panels and parse the collected data into subfiles that can be spatially identified with specific regions of the FSW. Big data analytics will be applied to these subfiles to reduce the size of the output data parameter files. The FSWs made in support of this proposed effort will be destructively characterized to identify data characteristics that correlate with any detected low strength regions of the FSW panels.
FSW is a solid-state joining process invented by The Welding Institute (TWI) in 1991 that avoids many of the metallurgical issues associated with traditional fusion welding of heavily alloyed metals. Since the metal is not melted, the reduced temperature gradients result in less residual stress buildup, distortion, and crack formation. Figure 1a illustrates the control parameters in the FSW process of rotation, travel and force, that result in the response parameters of X, Y, and Z forces in addition to the tool torque shown in Figure 1b. The tool consists of a threaded pin that plunges into the material and a shoulder that contacts the top surface of the workpieces and serves to constrain the material beneath. The material in the joint area is softened via frictional heating from contact of the shoulder and deformation heating as the pin shears the workpiece. As the pin tool translates along the weld seam, the thermally softened material is subjected to severe plastic deformation, which mechanically mixes the material as it enters the stir zone and is driven through the workpiece thickness prior to exiting in the wake. Researchers have speculated that weld quality can be interpreted from the force output.
While the FSW processes have been used extensively in the aerospace industry for the joining of various 2xxx series aluminum alloys, optimizing the process parameters remains highly empirical. This is additionally complicated by a dependency of the process parameters on the workpiece material, thickness, and tool design. Therefore, due to the heavy reliance on brute force experimental characterization, development of a FSW process schedule has remained a time intensive process. This empirical based approach lacks a physical understanding of the complex relationships between process variables (i.e., plunge force, traverse speed, pin tool geometry, applied torque, sheet thickness) on the material flow and heating in the weld zone, and weld strength. Thus it is difficult to diagnose the root causes of defective welds or provide guidance for adjustment of process variables to alleviate problems. Lack of a fundamental understanding becomes even more critical as current and future NASA missions require rapid assessment of the impact of workpiece materials and dimensions and tooling design on the complex, difficult to monitor physical processes occurring during the FSW.
The goal of this proposed study is to understand the source of FSW strength variations by correlating the physics of the FSW process with the process parameter output to establish criteria for acceptable material flow and hence processing conditions. Data will be analyzed for variations in material flow resulting in low strength as a function of process parameters changes in addition to variations along the length of a weld panel. As the material flow is influenced by the tool design and resulting workpiece temperature, this study proposes to investigate the process parameter output signals for consistency in the material flow process. These data files will be parsed into subfiles and analyzed for changes in the response. Any sections identified with differences will be correlated with the metallographically characterized images and weld strength. Once verified on this proposed effort, the developed methodology can be used ex-situ to identify regions of suspected compromised strength during the FSW. This validated digital platform would also provide the basis for in-situ process control by recognizing the onset of flow instabilities and altering the process parameters to correct on-the-fly during the FSW process.
Through close collaboration with engineers at the NASA-MSFC, the PI on this proposal has been involved with the development and verification of analog relationships for guiding process parameter selection and tool design in FSWed aluminum alloys. While process parameter space can be predicted using a kinematic model that influence of other variables can reduce this parameter space. Figure 3a illustrates how a balance between process forces and hardness of the resulting FSW can be used to identify a “sweet spot” for robust joints. Hardness variations suggest that the heat input affects the resulting weld quality in an age hardenable alloy such as 2219. Since the heat input is also related to the solid metal flow streams, it may be possible to use the process parameter output of force (X, Y, and Z) along with torque to monitor the process.
Various published studies have suggested that the FSW process is enhanced by oscillations which can be monitored by the process forces. The X and Y force relationship supports the hypothesis that a critical pumping action is required to produce fully consolidated FSWs. As the work piece thickness is increased, the input of heat varies from shoulder dominated friction with thinner panels to deformational heating dominated with thicker panels. Understanding how this variation in heat input affects the material flow and hence resulting weld quality will be evaluated in this proposed study through analysis of process output parameters and metallographic characterization and testing of the resulting welds. This information is critical to further efforts to capture the physics of the process through numerical modeling efforts.
Technical challenges to ensuring reliable production of high quality robust FSWs: In support of the joining activities for the SLS components, high quality, repeatable welds must be reliability produced. Utilizing monitoring of process parameters can help to reduce time and cost by ultimately eliminating post weld processing steps. While use of FSW has been reliability demonstrated in thin (< 8mm) panels of Al alloys for improved strength, the thicker workpieces (> 16mm) used in support of SLS must also be verified as consistently reliable. Coupling the experimental results with modeling predictions is needed to insure the rapid implementation of advances in solid state joining critical to support the advanced fabrication needs of the NASA.
This study will improve our understanding of the material flow during the FSW process by investigating methods to monitor in-situ. Perturbations in the material flow are speculated to be responsible for low strengths encountered due either to volumetric defects or metallurgical changes. As many parameters affect the material flow, it is postulated that perturbations in the process would present variations in the output parameters which can be correlated with the resulting quality of the weld. Data obtained in this study will be used to expand the current kinematic model of the FSW process and validate the prediction. This will serve as a basis for understanding how metal flow events lead to the development of compromised weldments. Understanding the progression of signals associated with abnormal flow will establish the reaction time for ultimately correcting these issues on-the-fly during the FSW process.
To overcome the current technical challenges, this proposal seeks funds to build off previous research on the FSW process undertaken by Dr. Schneider’s team in collaboration with the NASA-MSFC weld engineers. Dr. Schneider’s team is well versed in both the FSW process operation and resulting weldment properties. Correlating the output parameters with weld variations is the next step in continuing this successful collaboration in advancing the modeling and parameter predictions. The main technical challenge will be to correlate the process parameter outputs with the onset of flow instabilities that ultimately result in compromised weld integrity. Metallographic characterization and mechanical testing will be used to evaluate the resulting joint quality. This data will be correlated with the forces and torque measured during the process. This will provide the basis for understanding how variations in the process output parameters develop into sections with reduced weld strength.
The main objective of this proposed study is to conduct a series of FSWs in AA2219 of varying panel thickness as summarized in Table I and record the output parameters. The nominal schedule for each panel thickness will be evaluated using the kinematic model predictions. Discussion with the NASA welding engineers will be used to select optimized processing parameters for a given tool that have been used to produce robust FSWs in AA2219. While a tool rotation variation of plus/minus 50 RPM is targeted, final selection of parameters will be in conjunction with the NASA welding engineer recommendations.
This study is anticipated to help assemble a FSW Handbook being developed by the NASA- MSFC welding group. This handbook documents weld schedules and tool designs to assist in determining possible trends for future design applications. As slight variations in tool rotation have been observed to change the weld quality, a change of + 50 RPM is proposed to evaluate potential changes in material flow. This will result in a total of 15 weldments in bead-on-plate panels that are nominally 610mm (24 in) long. All weldments will be repeated 3 times.
These FSW panels will be cut into strips corresponding to the length of a parsed data file. Representative sections will be characterized by metallographic preparation of transverse samples for optical imaging. The phase composition of the mounted samples will be characterized using x-ray diffraction analysis (XRD). The metallographic mounts will be subjected to hardness testing for evaluating the mechanical properties in different weld nugget regions which will be correlated with the bulk tensile property data.
Force and torque data measured during each weld will be parsed into subfiles and examined for variations along the length of the weld in addition to changes in tool rotation. Findings will be used to modify the existing kinematic model for the flow process by modifying to account for the variations in the material flow resulting from variations in panel thickness.
The objectives of this study will be met by the following three tasks:
The goal of this proposed study is to contribute to the robustness of the FSW process by understanding and controlling a random source of weld strength variation by mining the parsed digital data files collected. The specific objectives are (1) to test the hypothesis that variations in transverse hardness distribution mappings cause variations in weld strength by providing easier or more difficult channels for shear deformation and (2) to relate observed weld hardness mappings to weld structural theory with an eye to controlling the distribution of hardness for optimal weld properties.
Listed on TechPort itself — the most direct way to ask about this specific project.
This is early/mid-stage (TRL 1) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.
None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.