It was discussion day again at English Aerospace Summer Camp. Today was somehow more difficult than the first day – they seem to have regressed? I guess the topics my coteacher chose were also a bit harder – nanomaterials, high speed rail, and 3D printing. These guys are rising sophomores and have really only taken introductory math and physics classes, so one challenge is that we’re teaching him both concepts and vocabulary. With my labmates, we can describe things like “isotropy” or “eigenvalue” in some rough combination of Chinese and English, and eventually we figure out the right English or Chinese word to go with the concept in our minds. But when homogeneity came up today during class, my coteacher and I kept trying to get them to connect this English word with the concept in their minds, only to realize that the concept wasn’t in their minds yet, so we had to put it there.
We got the most class participation when talking about high speed rail. We asked them how they get home, and got a wide variety of answers, from a few hours on bullet trains to the guy from Xinjiang, who takes a 39-hour train to Urumqi and then another overnight train to his hometown. All of a sudden Jilin and Xiamen don’t seem so far away!
In the afternoon, I started the computational part of my project. I use a commercial finite element software package called Abaqus which . . . I should probably explain what finite elements are. In mechanics, remember, we study how things react to forces acting on them. This is easy for a simple system like, say, a cantilevered beam with a point load. Think of a diving board, supported at one end with a person standing on the other end – there’s a simple equation that will tell you how much the board will deflect given the stiffness of the board and the wait of a person. For more complex systems – say, a plane flying through turbulence – there are no such easy equations.
Instead, we “cut” the object in question into tiny pieces – finite elements, if you will, where “finite” means it is not infinitesimally small. This way, you end up with a bunch of simple shapes, like hexahedrons (cubes) or tetrahedrons (a shape with four triangular faces). Meshing – the practice of numerically cutting objects into appropriate pieces – is somewhat of an art.
As you can see, if the pieces are too big, you’re not really modeling the object you want to model, but if the pieces are too small the system gets harder/slower/more expensive to calculate. In the picture above, #1 and #2 are too coarse, but depending on the application, any one of #3-6 could be appropriate.
Once the object has been reduced to a large number of really simple objects, the problem has essentially become a large number of simple problems. They’re all connected, or “coupled”, of course – if one of the elements moves, it will drag its neighbors along with it. Because of this, we have to solve the equations simultaneously. This is done using linear algebra, which is the grown-up way to solve those problems you had to do in algebra, where you were given equations like
0 = x + y + z
3 = 2x + y
7 = y – 4z
and were told to figure out what x, y, and z are. Except in finite element problems, we have several equations to solve for each element, and there may be hundreds of thousands of elements in a simulation.
Thank goodness for computers, amiright? I use Abaqus, which is a [very expensive] commercial software package specially designed to solve problems like this. It’s got a nice friendly looking interface where you can essentially draw an object, mesh it, apply loads, and solve for deformation – to see how it will look after loading. For simple problems, that’s essentially all there is to it.
My problems aren’t usually very simple, so there’s also some problem solving involved, figuring out a workaround so I can get the program to do what I want. Some people who do finite element analysis (FEA) write their own code, which gives you more control over the calculations, but then there are other demons to contend with. I do a little bit of both, using a feature of Abaqus called user subroutines; essentially I can write a little bit of code for one part of the calculations, and plug it into Abaqus’ own code. This is one way around the unfortunate fact that Abaqus does not natively handle materials that grow. Another way is to use thermal expansion instead – materials tend to expand as temperature rises, an effect that Abaqus does model – although this only works for very limited types of growth.
Anyway, today I started working with Abaqus. I figured out that I have to use this thermal expansion workaround and with the help of my labmates got a linear perturbation buckle analysis running. Don’t worry, I won’t bore you with these details, but it is one of the things I had wanted to learn about here, so I’m happy about it. More exciting developments to come . . .
I went to dinner with GuoYang and two other guys I haven’t interacted with much, and we talked about living situations on campus. I recently found out that two of my labmates are married, and tonight I found out that they don’t live together. “Of course!,” GuoYang said when I was surprised. Haha, nothing about that is “of course” in the US – at Stanford we have housing for singles, couples, and even families. One of the guys lives in an experimental mixed housing option, where Chinese and foreigners live together. This led me to ask why Chinese and foreigners are always separated – it seems like the powers that be are worried that we’ll corrupt the Chinese students. This was something that I found very frustrating when I was at XiaDa, because I was there to learn Chinese but contact with Chinese students was extremely limited – we had our own dorms, our own classroom buildings, our own cafeteria.
As of tonight, I’ve been in Beijing for a month. There have been the best of times and the worst of times. Week 3 was bad – between the air quality, my almost-constant nausea, research delays, and limited internet, I found myself wanting to be anywhere but Beijing. I may or may not have described Beijing in my private journal using the following words:
On the best days, it’s a concrete jungle; on the worst, it’s an apocalyptic wasteland. There are many places in the world where you can’t drink the water, but this is the first place I’ve lived where you can’t breathe the air. If Xiamen was the beginning of The Lorax, this is the end.
Those words seem a little extreme now. Perhaps still an accurate description of the city, but no longer an accurate description of my feelings about the city. A few blue sky days helped, as did whatever Paris Baguette did to make my stomach feel better. But also things like discovering the roof of our building, making actual progress on my project, finding a Coco on my way home from work, and the hundred small indications that my labmates are becoming friends. Beijing is still my least favorite place that I’ve lived in China, but honestly when you’re up against Xiamen and the farm, you have to be content with 3rd place. I’m looking forward to another few weeks here, and I know the goodbyes will be hard.
Today I learned:
The word for “hooligan”. I was telling GuoYang why I prefer WeChat over QQ, and I said it made my computer slow. He said those programs are called 流氓, or hooligan programs, because they install a bunch of other things without asking. The Chinese seem to have a high tolerance for these programs – I have all sorts of things floating around my phone’s home screen now because every app I downloaded wants to help me make my phone faster. Thanks but no thanks?
You’re not supposed to put chopsticks directly on the table. Usually in the cafeteria, one person will get the chopsticks for everyone, and today I did it. I put them on the table, and when ZhaoYan came by with his food, he invented a pretext for going back up to the front so could grab a new bunch of chopsticks. This makes perfect sense, as people spit food out on the tables and they’re just wiped down with a rag occasionally, but I didn’t really think about it and besides, it’s sometimes hard to tell what kinds of uncleanliness are acceptable in different cultures, right? Like in the US we wear our shoes indoors, and in China there’s never soap outside of bathrooms. But, now I know!