AMP²

Introduction

The value proposition of AMP² (Additive Manufacturing Parameter Predictor) is to reduce the risk and cost of metal AM parts by predicting physics-based AM process parameters. AMP² performs multi-scale, physics-based modeling of the metal AM process, providing clear insight for selecting part geometry-specific process parameters to attain relative density > 99.98% for as-built parts. For high-stress regions in structure-critical parts, the process parameters can be designed to eliminate surface defects and reduce internal defects by > 10x as compared to default parameters. AMP² can model the model AM deposition using most commercial machines, and the predicted physics-based parameters can be specified to produce the part during its build setup. AMP² has been used to generate process parameters to produce complex parts wtih Ti-6Al-4V, IN625, IN718, HA230, SS316, AlSi10Mg, and Scalmalloy. Presently, AMP² is being used to develop parameters for new generation alloys for which default parameters are unavailable.
 

Reduce Risk

Although Hot Isostatic Pressing (HIP) is commonly used to heal defects in AM parts, there is no substitute for high-quality as-built material, particularly for structure-critical parts since HIP cannot heal surface defects. Surface defects, if present in a high-stress region, can reduce dynamic material properties and degrade the structural factor of safety. Physics-based process parameters provide a scientific method to eliminate surface defects, reducing risk for AM parts.

 

Reduce Cost

It is generally feasible to produce ~99.6% dense parts using default parameters. A portion of the balance, ~0.4% pores, may be defects that occur in high-stress regions. It is feasible to use an experimental approach to correct the parameters, but the cost and time needed for quality assurance can be high, or uncertain. The AM part also may be oversized to compensate for the uncertainty, which is also undesirable.
 

Ease of Use

AMP² uses finite element and finite volume solvers for the physics-based modeling of thermal and fluid flow phenomena during AM deposition. These solvers use automated algorithms that make it simple to learn and use the software to produce higher quality parts. An advanced degree in engineering is not required to effectively use AMP²; a two-year associate degree is sufficient.

 

Value Proposition

It is expensive to qualify AM parts due to variability in mechanical properties, e.g., due to surface defects. Using physics-based process parameters to produce higher quality materials and enhanced mechanical properties is beneficial to meeting part performance requirements and reducing part qualification cost. It is also possible to use the enhanced mechanical properties lightweight the part, and create new possibilities to optimize its design.
 

Target Use Cases

1. Eliminate surface defects and increase part density: Physics-based modeling is used to design the process parameters to eliminate surface defects in the high-stress regions and to increase relative density for the part to > 99.8%
2. Attain higher dynamic mechanical properties: Higher quality build material, with no surface defects enables enhanced dynamic mechanical properties.
3. Increase build rate while maintaining target quality: Physics-based modeling can be used to increase the build rate (i.e., increase the bed drop), while producing high quality builds.
4. Weight reduction: Higher mechanical properties enable weight reduction for the part
5. Output solidification parameters and time-temperature plot for microstructure modeling.

 

Parameter Prediction Procedure

Typical parameter prediction procedure is a three-step process. This process optimizes AM parameters for a given part by accounting for the evolution of thermal conditions as a function of part geometry. Additionally, this work produces outputs a database of parameters optimized for geometric features of interest, which can be used to significantly reduce the time required to optimize parameters for additional parts. This database is validated using primitive deposition experiments.

1. Determine Baseline Parameters for Geometric Features: This is a one-time effort per machine, supplier, or material. This step is necessary if parameters are being developed to produce the part using a different machine, supplier, or material. This step also serves to establish baseline parameters for a known material to print multiple geometric features including overhangs, thin regions, and varying skin angles. The physics-based modeling is to predict parameters for simple shapes (e.g., cube, triangle, trapezoids with up-skin and down-skin sides). The predictions are validated using metallography and parameters are adjusted to minimize interior defects and to eliminate surface defects. The neighborhood of optimal parameters is evaluated using a matrix of melt pool simulations, accounting for the effect of hot spots, up-skin, and down-skin conditions. This step usually takes two deposition trials.
2. Verify Parameters for Critical Features: This is one-time effort per part family. Use the baseline parameters to produce critical features (e.g., at high-stress locations) for the part or part family of interest. Validate the minimization of interior defects and elimination of surface defects is using metallography. This step usually takes a single deposition trial.
3. Determine Geometry-Specific Parameters: The AM process for part is simulated using the estimated optimal process parameters. The evolution of thermal conditions and hot spots as a function of the part geometry is characterized. The process parameters are adjusted to account for the variability of thermal conditions. The part placement in the build plan is designed to mitigate defects that can be caused by spatter. For legacy materials, it is usually feasible to produce a high-quality part in the first deposition trial.

 

Transfer AM Process Between Systems

AM machines differ from one brand to another. The firmware for AM machines can be updated from time to time. AM suppliers change to respond to the business environment. World events may cause surge for rapid sustainment equipment. Technology advancement can lower costs. In all, it may become prudent to transfer or augment the production of AM part using a different machine or supplier. Such process transfer can be achieved using AMP² physics-based modeling by designing the new set of process parameters to emulate the behavior of current parameters, and thereby maintain part quality and verify part qualification.
 

Return on Investment (ROI)

The breakeven point for AMP² software license is estimated to equal the cost savings achieved for the process development for a single structure critical part.
 

Pipeline the Parameter Development

AM provides enormous opportunities to reduce the cycle time for producing low-volume parts, e.g., for rapid sustainment. There can be a backlog of such low-volume parts that could be produced using AM. However, the experimental approach currently used for parameter development is too slow to reduce the backlog. AMP² provides a physics-based method to pipeline the parameter development. The automated procedures in AMP² can be used to model the AM process for one part after another to create a pipeline of development for geometry-specific parameters to help reduce the backlog of parts.
 

Material Data Sets

Material data sets for thermophysical properties are used as input for AMP². The data sets are available for Ti-6Al-4V, IN625, IN718, HA230, SS316, AlSi10Mg, and Scalmalloy. Other material data sets, e.g., for new generation alloys for which default parameters are unavailable may be developed upon request.
 

Software

AMP² consists of software modules developed at Applied Optimization in Dayton, OH and open-source libraries.

• GUI: Input preparation and post processing
• ACME: Part-scale and layer-scale simulation
• ParaGen: Melt pool simulation software
• FSPGen: Feature-specific parameter generation software
• Support utilities

 

Recommended Hardware

• CPU: 11th or 10th gen Intel® Core™ i7
• Desktop PC for: GUI, FSPGen, and Support utilities
  • 64 GB Ram
  • Operating System: Ubuntu 22.04
• Desktop PC or Cluster for: ACME and Paragen
  • Operating System: Ubuntu 22.04, 64-bit

 

Disk Space

1. 10 GB for installation
2. Additional hard disk space for simulation results

 

Graphics

OpenGL 2.0 support
 

Licensing

AMP² is available as yearly single or multi-user license. A cluster is connected computers that act as a single system. A cluster license allows each user located within a single facility to simulate multiple AM parts in parallel based on the available compute capacity. AO can provide computer hardware with pre-installed software for stand-alone or network deployment. Cloud-based installations are not supported at present.
 

America Makes

Applied Optimization, Inc. is proud to be a member of America Makes since its beginning in 2012. We contribute to the AM Genome working group and Roadmap Advisory Group for the America Makes Technology Roadmap development.