TECHNOLOGY AREA(S): Air Platform, Ground/Sea Vehicles, Materials/Processes

ACQUISITION PROGRAM: 2019 Quality Made FNC

OBJECTIVE: To develop feed-forward control (FFC) hardware, algorithms, and multi-physics-based models to allow real-time tracking of powder bed layer variability and corresponding laser processing compensation to improve the quality of laser fusion-based metal additive manufacturing (AM) parts.

DESCRIPTION: Additive Manufacturing (AM) technologies continue to draw increased engineering interest, with technical advances in multiple fronts including hardware, software, and design processes.� AM is finding new applications areas with even a few documented operational demonstrations of fracture critical components, but this is still the exception rather than the rule.� Additively manufactured parts still require several trial and error runs with post-processing heat treatments and machining to optimize the build, reduce residual stresses, and meet tolerances.� AM still lacks a stable process that can produce consistent, defect-free parts.

One of the reasons for the lack of a stable process is the inherent variabilities at all steps (pre-, in-, and post-processing) of the AM process which tend to produce parts with inconsistent tolerances, mechanical properties, and defects.� These variabilities can be broadly classified into random (or aleatoric, fluctuating, statistical) and systematic (predictable, quasi-static, deterministic), primarily depending on the time or length scale at which they occur, and our ability to track and compensate for them.� Depending on the nature of the variability (random or systematic), different approaches can be developed to minimize their deleterious effects on part quality.� Of the many parameters that could be monitored and controlled, this SBIR topic seeks to develop innovative FFC hardware, software, and multi-physics models with the aim to compensate for the systematic physical property variabilities of the powder bed layer (temperature, mass, absorptivity, heat capacity, thermal conductivity). These variabilities result from the random distribution of the powder particles geometry, the thermal evolution of the part and powder inside the work volume, as well as the systematic layer thickness variability that develops during the AM process caused by splatter, molten particle ejects, and denudation processes, as well as from the previous layer surface roughness.� Not being able to anticipate and compensate for these systematic changes in the powder bed tends to produce an inconsistent melt-pool shape and temperature distribution, which leads to non-uniform microstructure and defects such as lack of fusion, keyholes, Marangoni flow surface ripples, porosity, balling, and surface roughness.

Feed-forward control can be performed at multiple levels such as at the track, layer, or part levels.� In general, the closer one probes the powder from the melt-pool, the more useful the information will be for purposes of adjusting the processing parameters and compensating for powder property variability.� At the same time, the closer one probes from the melt-pool, the less time there will be available to process the information.� A balance between the amount of information collected, the data processing time, the system response time, and the proximity to the melt-pool needs to be achieved in order to reliably obtain efficient FFC of the AM process.� This SBIR topic will consider all approaches to feed-forward control, but real-time approaches that aim to capture current, relevant information ahead of the melt-pool in the shortest amount of time possible will be favored over those approaches that focus on a track, static layer, or static build method, respectively.

PHASE I: During Phase I, the contractor will define and develop a concept for a FFC system including the hardware, the software, and multi-physics models for real-time tracking and compensation of the powder bed layer physical property variability towards the production of quality AM parts in laser powder bed fusion-based metal AM systems.� The Principal Investigator (PI) will also describe how to prepare powder bed test articles with a range of well-defined parameter variables for the purpose of model development, system verification, and eventually for technology validation.� The metal powders of interest to the Navy are Ti64, 316L SS, or Inconel 625.� During Phase I, the PI will continue to refine the models, improve the hardware, and expand the number of validation tests. The design created in Phase I will result in plans to build a prototype unit in Phase II.

PHASE II: During Phase II, the contractor will complete the purchase of all the components necessary for the development of a feed-forward control system and will start assembling the prototype design.� The PI will also develop a strategy for integrating the FFC system into an existing AM unit, unless the PI is developing a completely new AM system with the FFC already integrated into the design.� It is highly recommended that the PI team with an OEM of metal powder-based AM systems if the PI does not have access to AM equipment.� As part of the final validation, the contractor will fabricate the test articles defined in Phase I and measure the degree of improvement in part quality.

PHASE III DUAL USE APPLICATIONS: If Phase II is successful, the company will be expected to support the Navy in transitioning the FFC metal AM system for Navy use.� Working with the Navy, the company will integrate the FFC Metal AM system onto a Navy platform for evaluation to determine its effectiveness.� The OEM involved during Phase II will be part of the transition team.� Phase III will include defining the FFC system and test coupons for qualification, testing the coupons, and establishing facilities capable of achieving full-scale production capability of Navy-qualified parts.� The small business will also focus on identifying potential commercialization opportunities.

REFERENCES:

1.� Nassar, A. R., Keist, J. S., Reutzel, E. W., and Spurgeon, T. J. �Intra-layer closed-loop control of build plan during directed energy additive manufacturing of Ti�6Al�4V�. Additive Manufacturing 6 (2015) 39�52. https://edisciplinas.usp.br/mod/resource/view.php?id=241938

2.� Hu, D. and Kovacevic, R. �Sensing, modeling and control for laser-based additive manufacturing�. International Journal of Machine Tools & Manufacture 43 (2003) 51�60. http://www.sciencedirect.com/science/article/pii/S0890695502001633

3.� Everton, S. K., Hirsch, M., Stravroulakis, P., Leach, R. K., and Clare, A. T. �Review of in-situ process monitoring and in-situ metrology for metal additive manufacturing�, Materials and Design 95 (2016) 431�445. http://www.sciencedirect.com/science/article/pii/S0264127516300995

4.� Spears, T. G. and Gold, S. A. �In-process sensing in selective laser melting (SLM) additive manufacturing�. Integrating Materials and Manufacturing Innovation, 2016 (a Springer Open Journal) DOI 10.1186/s40192-016-0045-4.
https://link.springer.com/article/10.1186/s40192-016-0045-4

KEYWORDS: Additive Manufacturing; Feed-Forward Control; Feedback Control; Reliability; Multi-Physics Models