Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed

Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.

AM Discussion National Academy of Sciences

Aeronautics and Space Engineering Board

R. Allen Roach

SAND 2016-10171 PE

Sandia National Security Mission Areas

Top row: Critical to our national security, these three mission areas leverage, enhance, and advance our capabilities.

Middle row: Strongly interdependent with NW, these three mission areas are essential to sustaining Sandia’s ability to fulfill its NW core mission.

Bottom row: Our core mission, nuclear weapons (NW), is enabled by a strong scientific and engineering foundation.

2

Sandia National Security Mission Areas

Strong Nuclear Weapon Mission Area driver supporting future agile & affordable capabilities foundation for uncertain and unknown future

Strongly supported as a corporate level objective by Sandia, NNSA, and DOE

Multiple $M’s of projects are being funded at Sandia to explore and define Additive Manufacturing (AM) in support of our mission

Examples

of Metal,

Ceramic

AM Tools

Examples

of Polymer

AM Tools

Direct Write LENS®*

2

the top surface of the powder bed, the energy from the beam spot is absorbed by the exposed powder

causing that powder to melt. This small molten area is often described as the melt-pool. Individual

powder particles are fused together when the melt-pool re-solidifies. After one layer is completed, the

build platform is lowered by the prescribed layer thickness, and a new layer of powder from the dispenser

platform is swept over the build platform, filling the resulting gap and allowing a new layer to be built.

Figure 1 depicts one such process that uses a laser beam as the energy source. When a part build is

completed, it is fully buried within the powder in the build platform.

Figure 1. Components of the build chamber: (a) photograph showing the positions of the build platform,

powder dispenser platform, and recoating blade, and (b) schematic depicting the process of recoating and

spreading a new layer of powder over the previously fused layers of the part

There are several different types of PBF commercial systems that can produce either polymer or metal

parts. Today, most of the commercially available metal-based AM systems are PBF processes [3]. Some

varieties/ variations of PBF processes use low power lasers to bind powder particles by only melting the

surface of the powder particles (called selective laser sintering or SLS) or a binder coating the powder

particles. These processes produce green parts that require further post-processing to infiltrate and sinter

the parts to make them fully dense. Another class of PBF processes uses high power energy beams to

fully melt the powder particles, which then fuse together to the previous layer(s) when the molten

material cools, e.g., selective laser melting (SLM), direct metal laser sintering (DMLS), or electron-beam

melting (EBM). Repeating this process, layer-by-layer, directly results in a part with near 100% density,

even in metals. These processes are of primary interest to this study. General specifications for metal-

based PBF systems can be seen in the Appendix.

3 Literature Review The central idea to the review strategy followed in this report is to identify the correlations between

process parameters, process signatures, and product qualities to exploit these relationships in the

monitoring and control solutions. AM process parameters are the ‘inputs’ and primarily determine the

rate of energy delivered to the surface of the powder and how that energy interacts with material. We

categorize process parameters into either controllable (i.e., possible to continuously modify), such as laser

power and scan speed, or predefined (i.e., set at the beginning of each build) material properties, such as

Build

Platform

Dispenser

Platform

Recoating

Blade

(a) (b)

Build PlatformDispenser Platform

Recoating

Blade

= Powder

= Part

= Supports

Laser

f-θ lens

Powder Bed

PolyJet MakerBot

Q1 - How is additive manufacturing used in your field/application area today?

AM has become a well-accepted tool for the design, prototyping and production of Tooling, Gages, Fixtures, Molds and Mass Mocks

Significant cost-avoidance already achieved in production at Sandia (NG’s) and KCNSC

Current Uses of AM at Sandia & KCNSC

Exploratory efforts underway to evaluate when and if AM parts are adequate for use in weapons components

Today’s Goal - Integrate AM early with design projects as we continue developing background and expertise

Continued growth in use of AM for Tooling, Gages, Fixtures, Molds and Mass Mocks

First use in design and manufacture of weapons components

Continued expansion of R&D in AM

In 15 years new Qualification paradigm centered around AM

Q2 - How do you expect additive manufacturing to be used in 5 years?

Answers to Questions: • Q3 - Why have you chosen to move into additive manufacturing,

and what technical capacities are you focused on? • Q4 - What do you believe the major challenges are to more effective use of additive manufacturing? Overview of Sandia project titled:

Strong NW Mission Area driver supporting future agile & affordable capabilities foundation for uncertain and unknown future

Goal: Combine promise of additive manufacturing with deep materials & process understanding to revolutionize design, manufacturing, & qualification paradigms

Materials, designs, and ultimately components are Born Qualified/Born Certified

Born Qualified is a 15-Year Vision and this project is a 3-Year step in that direction

Born Qualified also provides the long term ability for materials and components to be “Born to be Surveilled”

Born Qualified Overview & Vision

Why Additive Manufacturing (AM) as driver for design, manufacturing, and qualification revolution?

Disruptive technology that allows simultaneous creation of part and material

Inherently flexible and agile

Ability to create near-net shape parts

Ability to tightly control and monitor manufacturing process at the voxel level

AM is ideal for low volume, high value, high consequence, complex parts

Paradigm Change

Common challenge:

“qualification and validation of materials and processes is an important dimension for integrating 3D printing into production. Removing this barrier is a recurrent

theme” Mick Mahler, DARPA 8/2/2016

Drive Qualification revolution by

Predicting Performance Probabilistically

Tightly controlling and monitoring process at voxel level

Accelerated cycles of learning

Integrate validated, predictive capability with real-time and ex-situ diagnostic tools to create the Capability Base to realize UQ driven qualification of design and process

Utilize Capability Base and Diagnostic artifacts to verify materials and process assurance

Qualification Approach

At the end of 3-years, test with 3 Exemplars to evaluate progress and future investment needs

PerformanceParameter

Requirement

Notional example of margins

and uncertainty

quantification for every

part/material

Overview of our Approach Using Metal Powder Bed Example

Complex Data:

ü Regression

Classification

ü Density Estimation

ü Statistical

Estimation

ü Dim reduction

MeasurePredict

Property Aware

Processing

Powder

bed

Densified

Structure

AM

Process

Alinstante

Properties

Materials

Models

Process

Models

Exemplar

Models

Exemplar

Performance

Performance

Predictions

Quantify &

Optimize

In-Situ

Measurements

100 µm

shear lip

lack of fusion

voids

fracture across

print layers

failure at 2% elongation, Vendor 1 H900

Accelerate Qualification

For New Designs

and Materials, BQ

will accelerate

time from TRL 1-9

and MRL 1-9

Accelerate

Process

Development

Reduce

Development

Builds

Accelerate

Testing

Simplify Production

Capability

Assessment Accelerate

Material

Assessment

R&D Technology Maturation, Design

Definition, & Production Engineering

Impact on Sandia

Goal - After 6 years of investment, demonstrate concepts and approaches to reduce timelines for:

R&D from 3-5 to 2-4 years

Tech maturation from 12-15 to 8-10 years

Cost savings could easily approach 25% with reduction in development builds and testing

Funding for the 15-year vision would allow us to:

Reach ultimate goal to be “Born” qualified

Take tech maturation and qualification beyond current process

Complications

Not all components will use AM (nor should they)

New qualification paradigm will need to cover all manufacturing including traditional

Already working in close collaboration with Sandia assessment and analysis groups

Also communicating and collaborating with non-Sandia organizations to go beyond NW

AM has had a significant impact already with bright future

Significant investment planned in the next 5-10 years

Entering year 2 of 15-year Born Qualified vision to change design and qualification paradigm

Collaboration with universities, industry, and government agencies will be key to our success

AM at Sandia Closing Remarks

Backups

Ceramic and Metal exemplars each selected to provide unique material, design, and process challenges

Applications with distinct opportunity for enhanced functionality using AM

Applications that have modest performance requirements

Applications that have reliability which can be assessed by measuring a limited number of relevant performance metrics / properties

Consider AM shortcomings, such as dimensional tolerances or surface finish which could dilute the focus from the goals

Exemplar Selection Principles

Sandia Core AM Capabilities are world-class

FY16 Highlights

Ceramic insulator ring process evaluation and down-selection was completed with build of initial coupons for evaluation

Multi-material printing demonstration

Direct Laser Melting of 99.5% Al2O3

Fabricated an initial set of valve housings using powder bed platform and performed dimensional analysis to obtain initial data on part relaxation and residual stress

Teaming with collaborators for Neutron Diffraction measurements at LANSCE to look at defect structures

AM Process Development

Develop innovative, real-time diagnostic tools of AM

FY16 Highlights

Testbed capability needs and design #1 complete, construction underway

Leveraging IR techniques developed in LDRD “Born Certified Additive Manufacturing: Predicting the Performance and Reliability of Laser Engineered Materials”

Project with MIT initiated for mm-wave capability originally developed for plasma diagnostics

Calibration strategy developed for IR temperature measurements

Science Challenges

Development of innovative experimental techniques for “Alinstante” high-throughput, real-time measurements, in tandem with detailed, lower throughput, measurements to efficiently establish the structure, process, and property relationships of AM materials

FY16 Highlights Designed and built a 1st generation diagnostic artifact that is easily inspected to

quantify printability limits and material / build properties Characterization techniques and tools identified Filed a Technical Advance and obtained approval for a patent filing regarding a non-

destructive characterization technique for additively manufactured materials Filed a Technical Advance regarding a concept for a flexible automated robotic

workcell for high-throughput materials and manufacturing process evaluation

Science Challenges

ProcessMaterial

ChemicalAnalysis(2weeks)

MetallographicPrep(6weeks)

X-rayphaseanalysis(4weeks)

Machinetestcoupons(6weeks)

Performmaterialtests(6weeks)

Analyzedata&feedback(4weeks)

Typical Multi-week

Cycle

Ability to relate microstructure to bulk measurable properties to Translate AM process results to material properties & ultimately predict component performance

FY16 Highlights

Phase Field solidification modeling: multiple grains/orientations, solid-liquid and solid-solid interfaces, far-field heat reservoirs

Random walk model of electrical conductivity applied to 3D grain structures, working on comparison to experiments on AM cylinders

Kinetic Monte Carlo models improved to do more complex geometries and scan patterns

Developing capability to atomistically predict thermal conductivity in powder beds to improve KMC and continuum models

Science Challenges

Official Use Only

GC Proposal FY16-0089 6

1. Understand how AM material properties correlate to processing conditions and microstructures

2. Collect data to inform, calibrate, and validate computational modeling efforts

3. Inform exemplar activities with realistic property sets that enable next generation component

development for the future stockpile

In conventional manufactured materials, procurement and engineering design hinge on the establishment

of a material specification sheet. These spec sheets, whether they are ASTM standards, Sandia specs, or

manufacturer specs, build on time-consuming and costly property measurements to establish practical

material expectations. For example, the Mil-5 Handbook requires many hundreds of repeated tensile tests

to establish a single value for the specification of a yield strength or percent-elongation. The creation of a

spec sheet for a new alloy typically costs manufacturers years and millions of dollars. To succeed in AM

material assessment, where the properties might vary from voxel-to-voxel or build-to-build, we must

modify our existing materials assessment paradigm. Conventional materials assessment requires time

scales on the order of months to machine test coupons, prepare for metallographic microstructural

analysis, and chemical analysis. What components of materials assessment can be automated and

centralized? Is it feasible to generate a material property data sheet in a manner of hours rather than

months? This challenge we call “Properties Al Instante”, a paradigm under which we will utilize high-

throughput, real-time measurements, in tandem with more detailed, lower throughput, measurements to

efficiently establish the structure, process, and property relationships of AM materials. The following

outlines this paradigm.

“Properties Al Instante”: Rapid High-Throughput Screening of Stochastic Properties

The quantities of interest in material assessment can be categorized into structure (composition, phase

content, defect content) and properties (thermal, electrical chemical, mechanical, magnetic, etc.). Some

quantities are amenable to high-throughput automation and integration (hardness, chemistry, and

electrical conductivity) whereas others (ductility, grain structure, long-term corrosion resistance, thermal

diffusivity) are not. Properties al instante is required to link the limited information available in-situ,

during processing, to the full structure-processing-property-performance relationships of real components,

see Figure 6. For the 3-year BQGC, we will develop a high-throughput assessment platform that

quantifies the easiest-to-integrate physical quantities that can be

used as a screening tool to identify locations of interest for detailed

structure-property measurements. In the following, we outline

approaches for high-throughput harvesting of structural, thermal,

and chemical information that will be leveraged for these purposes.

Structural: Rapid macroscopic hardness evaluation and ultrasonic

resonance provides high-throughput estimation of porosity, cracking

and microstructure. For dielectric materials, high-throughput

evaluation of the dielectric constant and acoustic resonance can

similarly indicate presence or absence of porosity and cracks. To

obtain this information, we will develop a tool with the capability to

perform several measurements including: roughness measurement,

hardness testing, electrical contact resistance testing, acoustic

modulus measurements, acoustic flaw detection, and scratch/wear

testing. Augmentation with x-ray fluorescence, diffraction, or tomography will add chemical, phase, or

defect information respectively. A possible platform for this tool is a Coordinate Measuring Machine

(CMM), already a critical tool to assess part dimensions and warping in AM parts [3].

Thermal: In-situ thermometry techniques will be developed to address the gaps between machine process

input parameters, material structure, and properties. Measuring the temperature history provides input to

not only process models but is also a critical parameter defining material properties and must therefore be

Performance*

Structure*

Processing*Property*

Materials*Science**and*

Metrology*

Design

*and*

Born*Qualified*Components*

Engineering*

Figure 6: Structure-Processing-

Property-Performance relating

performance to material science

Ability to incorporate material & process variability in AM process and exemplar models to predict performance probabilistically, calibrate models, and optimize design

FY16 Highlights Initiated various modeling efforts to characterize LENS and molecular dynamics of

powder beds coupled to thermal models Multiscale material modeling to understand difference between Type 1 and Type 2

residual stress (macroscale vs microscale) Devoted considerable coordination efforts to form interdisciplinary teams across the

entire project that to date have resulted in deeper understanding of AM process, numerical modeling at different scales, mathematical analysis to achieve optimal designs, quantification of uncertainties, and the ultimate integration

Science Challenges

Intelligent data collection & analysis of diverse sources (experiments, diagnostics, models) which requires filtering, selecting, sampling, & generating data to provide maximal information to create robust solutions in the face of uncertainties

Science Challenges

AM

Integrated AM experiments

RT metrology

Exemplars

Robust optimal control

New design concepts

Qualification requirements

Risk averse optimization

Characterization

AM modeling

Micro-meso material modeling

Validation

Quantification of uncertainties

Optimal experimental design

Design optimization

Macro modeling

Qualified: ceramic ring

valve housing

weak link

FY16 Highlights

Initiated development of foundational capabilities for PDE-constrained optimization under uncertainty, with capability established for phase-modeling, solidification, multiscale phenomena, and residual stress inversion

Completed initial data characterization of powder-bed AM tensile bars; statistical analysis of multiple attributes is now being performed