Session 5: Modeling and Simulation
The last session focused on developing a predictive capability for nanostructured materials. Three presentations addressed current computation methods directed at process modeling for consolidation and additive processes and nanostructural modeling advances of the material structure at the atomistic through macro levels.
Thomas Gates of NASA Langley Research Center presented a talk entitled “Computational Materials at NASA Langley: Nano-structured Materials Modeling and Simulation.” He gave a comprehensive overview of the activities of the research team at NASA Langley involved in the materials modeling effort. The goal of the activities is to accommodate the shift from traditional materials such as polymers, ceramics, and composites to revolutionary materials that are nanostructured, functionalized, self-assembling, self-healing, and/or biologically inspired. The team is looking for a convergence of technologies to link the scales from atomistic to continuum. Activities include molecular dynamics modeling, micro- and millimechanics, constitutive models and continuum mechanics, fracture mechanics, and polycrystalline plasticity.
Dr. Gates cited a number of opportunities, barriers, and potential applications:
-
Designing novel lightweight, nanostructured materials using experimentally validated computation tools.
-
Linking analytical and numerical models.
-
Conducting trade-off studies in the selection of new materials.
-
In order to characterize structure, processing, and properties, one needs to know which devices and what level of fidelity will provide valid information.
-
Materials questions include quality, quantity, and scale-up.
-
Modeling issues that need to be addressed are how to cross the atomistic-continuum gap, deal with bond formation/cleavage explicitly, and model processing influences with needed sensitivity.
Stephen Ridder from NIST then gave his brief, “Modeling, Data Acquisition, and Process Control for Thermal Spray.” While he acknowledged that at this time not all work being done is with nanomaterials, the directions and advances will nonetheless be important to nanomanufacturing. The emphasis at NIST is on the development of new sensors for coating technologies that can be automated to give high production rates and reproducibility. Dr. Ridder classified thermal spray sensors/process control as follows: (1) torch-based, (2) spray plume-based, and (3) substrate-based. An example of a torch-based sensor is the Coriolis flowmeter to establish the powder flow rate at which
agglomeration breakup can be monitored. The Coriolis meter has a fast response, which may help to smooth out the powder feed.
Another sensor development, applicable to the spray plume region, is the imaging pyrometer. The two-color imaging pyrometer has advanced capabilities for noncontact thermometry suitable for nanoscale thermal spray. The ability to image the plume provides information such as plume alignment; changes in the spray pattern because of changes in the spray parameters; particle velocity, temperature, and size distributions without having to scan the plume; and powder feed-rate fluctuations (the relatively rapid frame rate can resolve fluctuations of 10 Hz or less). The large number of particles sensed in each image also speeds up the acquisition of statistical information concerning average particle velocity and temperature (for the entire particle field or for various regions in the plume).
An important issue that NIST still must examine is imaging at lower temperatures—how much of the plume is not registering because of cold material.
Finally, Dr. Ridder discussed a substrate-based sensor, a surface temperature pyrometer using an InGaAs photodetector. Substrate temperature is a key factor in splat formation and coating buildup. Substrate preheat affects coating adhesion in a poorly understood way (e.g., does substrate preheat boil off moisture, changing the substrate’s stress state?).
The opportunities, barriers, and potential applications identified by Dr. Ridder are as follows:
-
Controlling spray parameters to preserve nanostructures in coatings requires addressing feed problems such as periodic change in flow, feeder rates, and liquid precursor chemistry and concentrations.
-
Torch-based sensors need to consider arc parameters, powder mass flow-rate control, gas mass flow-rate control, and spray gun position control.
-
Spray plume-based sensing systems need to monitor and control particle spray pattern, particle size distribution, particle velocity and velocity distribution, and particle temperature and temperature gradients.
-
The accuracy with which the two-color imaging pyrometer can image the entire plume needs to be quantified. Cold material in the plume may prevent the full plume from being imaged.
-
Affordable real-time detectors are needed. An example is the InGaAs photodetector for surface temperature of the substrate.
-
Changing substrate conditions will require dynamic coupling of the plasma environment with the deposition to the substrate.
Walter Milligan of Michigan Technological University then completed the session with a discussion of modeling of consolidation processes in his presentation “Modeling of Nanopowder Consolidation.” Fe-base alloys and cryo-milled Al alloy powders have been created by ball milling. Ball milling can produce bulk quantities of powders economically with low capital equipment costs. Ceramics, intermetallics, and metal powders with particle size from 1 to 20 microns and grain sizes from 5 to 30 nm can be produced.
Consolidation via HIPing or inert environment rapid forging process can produce fully dense, 100-nm product.
Professor Milligan spoke about opportunities, barriers, and potential applications:
-
Modeling can provide insights into both cold consolidation, which generally does not produce full density, and hot consolidation, which results in grain growth.
-
The Arzt, Ashby, and Easterling HIP model considers two-stage densification with yielding, creep, and diffusion. To apply the model accurately, one must conduct mechanical tests on bulk specimens at temperatures and grain sizes of interest. One needs to know the values of the model parameters, such as creep and diffusion constants (obtainable by curve-fitting, guessing, and/or experimentation).
-
The mechanical behavior of metals with grain sizes from 50 nm to 1 micron is different from that of traditional metals.
In the session 5 follow-up question-and-answer period, Dr. Durham asked where nanotechnology needs to be in 10 years. Two major discussion points were raised:
-
Dr. Gates commented that a significant increase in properties is needed, and this can be accomplished through modeling, synthesis, and experiment. He said that a technical obstacle is the disconnect between molecular dynamics and continuum mechanics. He felt that interdisciplinary teams of modelers, physicists, chemists, materials scientists, and so forth are needed.
-
Dr. Rawers asked if any modeling exists to aid the development of ductile ceramics. Dr. Gates replied that he had no experience in ceramics. Professor Mayo replied that reported ductility at room temperature is overrated, though a transformation toughening approach might be useful.