SBIR-STTR Award

This work develops a simulation framework and demonstrates a methodology to integrate nondestructive evaluation (NDE) and stress analysis to cost effectively assess the ability to detect critical defects as determined by an assessment using a simulation environment of damage growth based on material properties, part shape and local stress fields. Integrating emerging simulation tools in NDE, stress analysis and damage evolution we demonstrate a powerful approach to lifing assessment using a damage tolerance approach, thus providing a key component in the emerging additive manufacturing processes. Integrating XRSIM (NDE simulation), DARWIN (damage evolution) and NESSUS (optimization tools), for the first time, demonstrates a means that accurately captures the relationship between critical AM design cycle parameters with defect morphology in NDE and damage evolution. Key for this integration is developing quantitative and part appropriate probability of detection curves needed for accurate lifing analysis yielding a significant cost reduction by reducing the need to rely on extensive experimental POD curves. A plan to address verification, validation and uncertainty quantification in the use of models and simulations for rapid qualification is provided. The Honeywell, Southwest Research Institute and NDE Technologies partnership in this project positions us to further dovetail with ongoing DARPA AM programs.

Benefit:
This project proposes to enable rapid qualification of AM processes through the development of a methodology that accurately captures the relationships between key manufacturing parameters and the resulting product, including location-specific microstructure, defects, material properties, inspectability, and damage tolerance. This methodology will, in the short term, integrate predictive models for non-destructive evaluation (NDE), stress analysis, and damage tolerance (DT) simulations, by leveraging our expertise and existing products XRSIM, NESSUS and DARWIN. In the longer term, the methodology will also broaden the integration to include simulations of the manufacturing process and microstructure-property relationships. Significant design cycle time savings are possible by identifying the best process parameters needed for an optimized design, allowing for a faster response time to engineering changes and requirements that will enable the full range of AM benefits. An additional benefit is for simulation methods to generate NDE POD curves. We are planning to demonstrate several methods, 1. Directly from XRSIm, 2. Using XRSIM to supply Monte Carlo simulation or response curves. These results combined with Nessus generate POD curves. Cost savings are huge in that a demonstrated means to generate meaningful POD curves if only from a reduction of the sample costs. Specific cost savings to additive manufacturing users will come from a reduction in the number of rejected parts. When a virtual inspection is performed, lots of virtual parts can be scrapped at first. As the process is adjusted the number of virtual part rejections reaches an acceptable level. If the model has been properly calibrated and V & V"d, this will lead to fewer scrapped parts. The cost reduction is directly proportional to the reduction in rejected parts. Two major components of the proposed integrated modeling environment, DARWIN and XRSIM, are already successful commercial products in certain markets. Their integration and further development here will result in significant expansion of their commercial potential, including for example, selling XRSIM to current DARWIN customers, and selling DARWIN to current XRSIM customers. Exposure to other, particularly non-aerospace, industries is expected as these industries begin to implement DMLS and other AM processes. Since these two software programs are both mature and supported by existing marketing and sales organizations, expansion into these new markets will be expedited.

Keywords:
additive manufacturi

In this proposed project, a novel framework will be further developed and demonstrated to link physics-based simulations of non-destructive evaluation (NDE) methods with probabilistic fracture mechanics (PFM) analysis to support rapid qualification of advanced materials processes such as additive manufacturing (AM). In particular, the X-ray and CT simulation software XRSIM and SimCT will be linked with the PFM software DARWIN and the general-purpose probabilistic analysis software NESSUS to create a virtual reliability environment (VRE). XRSIM/SimCT will be used to determine full-field, location-specific POD curves for significant manufacturing defects that DARWIN will incorporate in predictions of fracture risk for the component in service. A prototype framework was successfully developed and demonstrated in a Phase I project. In the proposed Phase II project, the computer programs XRSIM/SimCT and DARWIN/NESSUS and their interfaces will be enhanced to substantially improve automation and efficiency of the computations, in support of improved efficiency for physical inspection plans and improved reliability for the component. Model credibility will be established through revision and execution of the verification and validation plan developed in Phase I. The integrated VRE, which will have broad applicability beyond AM, will be licensed commercially.

Benefit:
AM processes promise a long list of benefits to both legacy systems (low cost low volume parts) and new program development (lower cost prototypes early in the design cycle), but without a way to qualify parts in a highly variable piece to piece manufacturing process there is no way to trust the product. The VRE proposed here will provide a major step forward in establishing quality and hence usability in AM parts. However, the VRE isnt just tied to AM processes, but can accept process input from any other manufacturing method (either model or empirically generated). This provides a much broader list of potential applications and benefits. Specific cost savings to additive manufacturing users will come from a reduction in the number of rejected parts. When a virtual inspection is performed, lots of virtual parts can be scrapped at first. As the process is adjusted the number of virtual part rejections reaches an acceptable level. If the model has been properly calibrated and verified/validated, this will lead to fewer scrapped parts. The cost reduction is directly proportional to the reduction in rejected parts. There is also a significant design cycle time savings to be had in identifying the best process parameters needed for an optimized design, allowing for a faster response time to engineering changes and requirements. The ability to inexpensively get POD information, as provided from this projects VRE, is a huge cost savings for industry and the ability to target or zone inspections based on accurate POD and DT probabilities is a risk reducer as well. Just being able to examine an existing inspection for adequacy has huge advantages. The broader manufacturing perspective, massive investments are currently being made across the board in ICME methods for improved simulation of materials development and manufacturing processes. These simulation systems will all eventually need to carry through to address component reliability if they truly support the cradle-to-grave design, manufacturing, and life management system. Any defect-sensitive and safety-critical application will need to address NDE and POD considerations as well as fracture risk. Lastly, the proposed project will result in significant enhancements to commercially mature software and the creation of an integrated VRE that will also be licensed commercially.