Explain like I'm 5

One of my Facebook friends recently saw my NADDRG 2016 Fall Symposium talk titled, "X-ray Stress Measurements during Multi-axial Deformation."

He commented, "ELI5" (see title)

It's taken me over two weeks to get back to him, but this was my response:

Grab something by one end in one hand and the other in other hand. Maybe it's a piece of tape or strip of paper. Pull the ends apart from one another slowly and you're performing a uniaxial (single axis) tension experiment. As you deform the material, that change in length compared to its starting length is how much you strain the material by, and how much force you need to keep that strain is the stress.

Now let's take a sheet of paper. You're going to grab two sides of the sheet and someone else will take the other two. You guys will both slowly pull, so now you're stretching the sheet. This is multi-axial deformation now (in this case with two perpendicular axis). For uniaxial deformation, we have a pretty good idea of the math to relate the strain and stress, but this is not the case in multi-axial deformation.

To find the relationships between strain and stress in multi-axial deformation, we have ways to find the strain, but not so much the stress. To solve this, we use x-rays to look at what's happening at the atomic level. Materials are made out of atoms and those atoms are ordered. When we pull on a material, those atoms also move apart from one another. We can relate that separation of atomic distances to a stress, but we can't see those distances with our eyes.

X-rays are amazing. How we use x-rays here is not like medical imaging when a doctor wants to see your bones. We use x-ray diffraction to find the separation of those atomic distances. Think about when you shine light onto you hand in a room and look at the shadow behind it. The shadow has the same shape as your hand but is much larger than your actual hand. This is a simplified idea, but we use x-rays in the same way to magnify and observe those separations in atomic distances in the material during multi-axial deformation to get desired information about the stress.

How'd I do?



*Corrected for typos at 11:27 pm courtesy of my editor

Learning

For a while I wasn't happy with my post-doctoral position, and I should have been. I was in a position as a Materials Research Engineer, which means I still got to do materials-related research. I was in an environment where the pace of work was relaxed, enabling one to look at tangents when they came up (but ideally before you get too far and don't write up your previous results). Most importantly, as a post-doc at my work place, I had the freedom to determine my project. I dictated my research for the next two years.

Some of the problems I had was that I really wasn't doing any more materials research. I was researching on materials, but I felt much more like a mechanical engineer than materials scientist. When I went to the last two conferences I mentioned, I realized how much I missed microstructure-related science. Whether it was discussing improvements and novel techniques in microstructure characterization, or the details and complexity of microstructure evolutions during deformation, recrystallization, and grain growth. Additionally, because I felt outside my realm of expertise as well, I was easily frustrated when obstacles and set-backs occurred and I didn't know how to get around them.

It wasn't until I was catching up with my colleague at Carnegie Mellon University that my work attitude turned around. While she (another post-doc) was still getting to continue a similar line of work to her PhD, she reminded me that I was getting the chance to learn new techniques like digital image correlation and new fields like sheet metal forming.

This conversation reminded me of why most of us pursue a PhD in the first place. It's the opportunity to keep learning, whether to keep becoming an expert in a field you're in, which was what I thought needed to be happy, or branching out into a new area, which I am happy with where I'm at now.

Behind on updates again...

Back to back conferences though! I guess this is what they mean when they say, "If you love what you do you'll never work a day in your life." =P

Last week I was at the following learning a lot of 3D microstructure characterization and properties. The main takeaway I had is that grains are simply blobs in space, and if grain boundaries do exist, they are entirely flat boundaries. (This is a joke).

This week I'm focusing on recrystallization and grain growth. So far, the talks on the first day have been great and I look forward to the rest of the week.

Hopefully more updates will come after my two week work "vacation."

LMFIT and Peak-Fitting of XRD peaks

I recently added the Non-Linear Least-Squares Minimization and Curve-Fitting (LMFIT) package to my Python. Lately I've made an effort to move away from MATLAB and use Python and the packages available for it to take on some of my scientific problems.

I've been using the Gaussian model and composite model feature of LMFIT to reproduce a fitted diffraction profile as close as possible to what the manufacturer's software provides me. In the software, the fitting procedure determines the position of 2theta to determine the d-spacing of the crystal structure by Bragg's law. The stress is then calculated based on the change in d-spacing or lattice strain (gross oversimplification). The software ignores the information that exists within the peak, such as peak width that can reflect a dislocation content or the peak shape that may influenced by the twin density. Hence my goal is to reproduce the fitting procedure and determine the peak characteristics such to hopefully find more information.

Below outlines my fun so far in trying to replicate what the manufacturer does for peak fitting. This is an example of a diffraction peak, that has been fitted with a Gaussian curve:

This is the output that the software reports:

One distinct difference right away is that two peaks are used to fit the diffraction profile. Obviously I'm already off to a bad start. For those curious, this is because both the K alpha1 and alpha2 are used during the measurement, so there is in fact two signals. This was my next attempt just using two Gaussian curves.

Not even close. Although the fit is actually better than the manufacturers, it has no physical reasoning. So I enforced some rules on the intensity of the two radiations as well as the expected separation of the two peaks and came out with this:

The constraints actually made my fit "worse" compared to when the summation of the two Gaussian curves were left unbounded, but at least now I'm little closer to what the manufacturer reports. The next constraint I added was to assume the interaction profile of both radiation types would be same same in terms of variance, which gave me this:

Then the end of the day rolled around and I called the quits =P

Overall though using LMFIT was pretty intuitive and seemed to offer a lot for optimized fitting routines. I'm hoping to use more of it in the future for my work.

CHiMaD Materials Design

Last week I participated in the CHiMaD Materials Design Workshop held on NIST. Greg Olsen gave a keynote talk to kick things off, introducing to us the concept of System Design Charts, specifically focused for materials science and engineering and focusing on the paradigm of processing, structure, properties (and sometimes performance). The idea is such design charts enable one to target the important areas of interest to investigate and focus on, but also being able to communicate better on what one is investigating. This makes sense from an industrial standpoint when one needs to talk with shareholders or clients of a consulting firm, but can also be applied academia so that members within a research group can keep track of what each person is working on specifically, while still seeing the big picture.

To be honest, there is nothing special with the System Design Charts. It's simply just another tool that one can use to organize their thoughts and convey their ideas on where to go next. It's a strategic planning tool (like SWOT or anything else) to keep groups focused on the important parts of the big picture, rather than investigating small parts of interest.

First, one lays out all the processing steps in a sequential manner. The processing steps will influence the structure of the material. This may be the literal crystal structure (if we're talking about metals), but also things such as phases, precipitates, or features like twins. The lines that connect the processing boxes and structure boxes are two-way, such that processing will directly influence structure, but the resultant structure can also influence subsequent processing. Properties sit on the side other structure, and highlight what structures are connected to what properties the most, again as a two-way connection. Optional is putting performance on the other side of properties.

Designing the chart this way, one attempts to solve the problem in determining how to optimize the processing to influence the properties of most interest, where the monetary interest lies. There's nothing special about this, except that the structure column clears up the historic blackboxes that we tried to directly correlate properties to processing. From what I understood, therein lies the materials design by considering the actual structure of the material and the impact it plays on the properties. As a materials science major, this is all something one learns in a first-year class. Looking from a bigger picture though, the tool tells you where money can be made, which for a scientist may not sit right.

Alternatively, it was introduced to us that such a concept can be applied to any science, which was the goal of the workshop. This was my attempt at a Monte-Carlo grain growth scheme (no money to be made here =P):

2016 DOE-AMR Vehicle Technologies

Last week I had the opportunity to participate as a reviewer for the Department of Energy (DOE) Annual Merit Review (AMR) in Vehicle Technologies, specifically propulsion materials. This past Friday, I finally submitted the two reviews assigned to me. 

With a combination of procrastinating and not being 100% efficient, two single reviews took me the entire afternoon. In my sample size of two, the second one was "easier" than the first one. I was talking to my lab director, and I mentioned that I did not think two reviews would take me so long.

The overall experience was interesting. The presentations that I attended were allotted 30 minutes each, with the presentation to be about 20 minutes and 10 minutes for questions and answers, with priority given to the reviewers. Unsurprisingly, some presenters still clearly went over the the 20 minute mark, like in any other conference. Although this was not like any other conference. For one, there was no conference fee. Secondly, the contents of the presentation were organized in a way that made the review process challenging for me as a first-timer.

Reviews needed to cover their approach, their technical content, their collaborations, and their future work. Often collaborations are quickly glanced over and rated on a scale of 1-4 but almost seemed to be more of a check-box. The approach was where the objective of the project was generalized such that anybody could understand, and since most of these talks dealt with Integrated Computational Materials Engineering (ICME), typically presented a flow diagram as well. The technical content focused on what was achieved that year, and finally the future work covered what will be done for the next year, which at times seemed relevant to the technical content presented or other times seemed out of place before you realized you needed to look at the big picture.

Therein lied the challenge for me being a first-time review. I was reviewing projects that were both somewhere half-way complete, so in my head I couldn't see the complete picture at times of what had been accomplished before. And even in the technical content, it was a quick overview glance at the most major achievements, without digging deeper into the science at times. This made the merit of the future work even harder to judge. Lastly, I approach science with mostly a positive mindset (until I learn more and become a pessimist), so I think all approaches are generally valid and very cool. Then to score and comment each one of these categories on an arbitrary scale from 1-4...

Reviewing a project based a 20 minute presentation was unlike a reviewing a paper where you did have all the details. And as a scientist, details are important to us. I started making the effort to look up papers had been published, but most of this could not be found since it was all recent work. It took me awhile before I finally conceded and took a step back and judge the review based on what was presented to me (as well as last year's presentation. I realized I had to judge the project, and not necessarily the science (although it is a large part of it). I gave my input on the things I was impressed with, but also on what I thought could be use some more work or consideration. I hope in the future if I do this again, I'll also be able to give input on what I found was lacking (which comes with years of knowledge in the field...)

It was an experience to sit on the other side and be presented to, rather than making the presentations for my advisor as I had once done during graduate school. 

NADDRG 2016 Spring

Last Thursday I went to the North America Deep Drawing Research Group Spring Symposium.

It was the first "conference" I've gone to since I've started my new position. I've placed conference in quotations since it was a single day of talks starting from 8:30am and ending at 6:00pm (although originally planned for 5:00pm). In that essence, it was like a conference in which some talks ran a little long, and there were always too many questions asked. Neither of which is bad, as the content was always interesting (at least in my opinion) and the questions brought up discussions between the audience. Furthermore, since there is only one talk at a time, there is no scramble from one room to the next like at the typical conference.

The speakers were well divided between industry and academia (and no one from a National Lab). The industry talks turned out to be equally as interesting as the academic ones. Sales pitches are not allowed, so the talks are typically focused on addressing an engineering problem or challenge without any superfluous information. Furthermore, the talks are designed to be either updates or works in progress (this is attributes for the large discussion aspect from the audience).

Two things that stood out for from these talks is the lack of microstructure characterization and discussion of crystal plasticity. It was a reminder that I was in a room of mostly mechanical engineers and not materials scientists. However during one of my side discussions with an engineer at Aleris, he expressed his surprise on just how things at the microscale influence the macroscale behavior. (Context: They found that special grain boundaries played a major role in crash-worthiness in aluminum. However I didn't get the chance to ask if these were Sigma3 boundaries or not, which would've been very interesting...).

My takeaway from all of this was the (micro)structure to property and performance relationships is equally important during forming/processing operations, and there is a large room for the development of these understandings.