Semester in review, and what I did this summer

The classes I took this Spring semester were:

Materials Processing-- The manufacture and modification of the most important engineering materials: metals, plastics, ceramics, and composite materials. Topics covered included casting, forging and other shaping processes, welding (there's a whole lot to it; you'd be surprised) and other joining processes, machining, and a bunch of other stuff. It was a pretty tough class. I pulled several grueling all-nighters for it. I thought I was going to end up with a C but I got a A- somehow. It was the first A- I was extremely happy to receive. I will spare you pictures of my quizzes and assignments-- I trust you will take my word when I say it was hard.

The most interesting thing I learned in that class is that the idea that glass is NOT a "slow moving liquid" that deforms over decades. People made the induction from old houses having deformed windows and spread the commonly-held misconception that it had done so over many years. In fact, older glassmaking techniques had no way of ensuring consistency and flatness of produced glass; if a thicker end resulted, they would just use common sense and install that end at the bottom so it was less prone to breaking.

Materials Processing Lab-- Deformation of metals and plastics, like shearing, bending, upsetting (in layman's terms, "squishing") and wire-drawing. We would look at the microstructure of the deformed metal most of the time. There was a lot more to it than what I just described, but that's probably all you want to know.

This is what martensite looks like on a microscopic scale. Martensite is formed by quenching mid-to-high carbon steel from a high temperature. It is very strong (in terms of elasticity) but very brittle. For this reason, not all steels are "weldable" because high-carbon steels can form martensite at the weld as it cools and make the weld prone to failure, which is a very bad thing.

Mechatronics-- This class was a mess. It was a crash course in Electrical Engineering. We had 2 quizzes per week (one for reading comprehension, one for homework comprehension) and the tests were hit-or-miss. I made a (on an objective grading scale) failing grade on one, and a 95 on another. The final was a very anticlimactic one-- it was just another test. Off the top of my head, major topics covered were basic DC circuits, power engineering, signal analysis, transistors, op-amps, AC circuits, magnetism and magnetic circuits, and motors (there are about a dozen or so different types of electric motors, and they're all complicated to even understand how they work).

 This is an animation of how a 5-5-5 timer operates. It's a cheap, "simple" chip that has an only function of toggling on and off by itself. I still could never make sense of how it actually works.

Mechatronics Lab-- The most ridiculous lab I've ever taken (and probably will take) in my college career. The way it was administrated was very... unique. I've had labs in the past with very time-consuming and painful-to-write lab reports, but there were no reports in this lab. Instead, most of the work went into prelab assignments, which, in every other lab class, are not so bad. But these regularly took many hours to complete, even with the help of a friend.The lab section itself was more of an intelligence test than anything. So many things could go wrong with building your circuit, and for me and many others, they almost always did. Murphy's Law was blatantly obvious in that lab. Circuits are confusing, and there is no way to prepare for it. There was no way to practice; you just had to think quick on your feet. I made a previous post about this class.

Engineering Statistics-- At first, this class was pretty neat. It was my first introduction to probability theory and it was interesting stuff. However, as the class progressed, the material progressively became more difficult, abstract, and boring. Thankfully, I had a great TA, and his discussion sections were much better than the seasoned professor's lectures. The tests were ridiculous-- they were split up into two separate tests. One was quantitative problem solving during the discussion section, and the other was pure theory, in the form of essays, during class time. There were EIGHT essays per test. Luckily, I was exempt from the final (receiving the good news was probably the happiest moment of my semester), which turned out to have SIXTEEN essay questions to be answered over the course of 3 hours. I was saved from that painful experience, and not to mention the task of studying for it.

Confidence intervals and null hypothesis error.

Machine Elements-- This course was pretty eye-opening as far as structural design is concerned. We covered failure and fatigue of metal of structural components of machines, and we looked at many important elements like shafts, bearings, gears, springs, and bolts. Each and every one of them are much more complicated than you'd think, and it really made me appreciate how much consideration goes into what seems like the simplest of things.

Planetary gears are pretty crazy. And hard to analyze.

I also took a machining certification class near the end of the semester. I learned how to run a bandsaw, mill, and lathe.

This summer, I worked ~20 hours a week doing "research" here at UT, but it's more of a design project than anything else. It's called research because it's such a creative, open-ended process. I, another student, and my Machine Elements professor are designing a da Vinci-inspired walking lion which must also be able to turn and sit, that is almost purely mechanical (no robotics) for a guy called Shaun Whitehead and a French company called Dassault Systemes, to showcase their Catia software. It will probably end up in a French museum when they build the actual product from our design. However, the design is not such a straightfoward process. It's challenging to say the least, but I like the job even though it's not easy. I'd much rather do a mentally demanding job than a physically mandating job, or a repetitive job of either kind.

Here's what we have to show so far:

Digital Logic Circuits

One of the classes I'm taking this semester is called Mechatronics. The area of study is becoming extremely common in modern technology; an easy example of this is robotics. In fact, I just recently signed up for a robotics class in my last semester here at UT. It's really fascinating stuff but the inner workings are incredibly complex.

Anyway, this class, the coolest-sounding one I've ever taken, is not living up to its name. It's essentially a crash course in Electrical Engineering; no mechanics to speak of. It's interesting but it's tough, it's not taught very well, the textbook sucks, and I'm not good at it.

The lab is a bit more interesting. Thankfully there are no lab reports, but prelabs are insanely time-consuming and sometimes depressingly frustrating (seriously). I have spent over 6 hours on one or two before, even working with another friend. It wouldn't be so bad if the lecture actually lined up with the lab, but we get a crappy 50 minute lab lesson every week with an entirely new concept each time, and the prelabs require inside-and-out knowledge of that material.

Regardless, this week's lab was actually really neat, and almost fun if it weren't so challenging. It was a mock digital logic circuit design for a vending machine. I shouldn't have to tell you that it's way simpler than a real vending machine, but it was still an interesting mental exercise.

Pictured below is the final page of the prelab. Believe it or not, this is a simple logic circuit. It's NOWHERE near as complex as the electronics in, say, a remote control.

I'll explain what this is supposed to be later in the post.

 This is the closest thing to rocket surgery I've ever done.

Now do you understand why I have little patience for people that don't know how to USE electronics?

If you want to make the most of your day and learn something interesting, keep on reading.

This is the first thing I've ever done that actually used binary (1s and 0s; 1=true and 0=false). Binary and regular numbers can be interconverted, and it's actually easy to learn. The one weird thing is that it reads from right to left. The rightmost digit has a value of 1 and doubles every time as you move one to the left (<- 32 16 8 4 2 1). Every real number is a combination of 1s and 0s. A 1 means you count the number, and a 0 means you ignore it. So 001 is 1, 010 is 2, and 100 is 4. Logically, 011 is 3, 110 is 6, and 111 is 7.

 Hopefully this shirt makes sense to you now.

Okay, so why the binary lesson? Because binary is really freakin' important. It's used in almost every electronic device you own. It's worthwhile to understand.

Let me relate it to something you use all the time:

A 1 or a 0 is called a bit (b). 8 bits are called a byte (B). 1000 bytes make a kilobyte (KB), 1000 kilobytes make a megabyte (MB), and 1000 megabytes make a terabyte (TB). You've also heard the terms of kilobit and megabit, and those are usually used in telling you transfer or connection speeds (a DSL connection is typically 1.5 megabits per second, or Mbps, for download speed. That means your computer can receive 1.5 million 1s and 0s per second. The best part is that this is SLOW to some people. 30 megabit cable is not uncommon.)

Everything on a memory card or hard drive is just a lot of 1s and 0s that are bunched together in groups of 8s. Pictures are all 1s and 0s. Songs are all 1s and 0s. Games are all 1s and 0s. You get the idea. If you ever wonder why your computer says your hard drive or memory card have less storage space than what the box said, that's because computers take bytes in groups of 1024 instead of 1000. So don't worry; you're not getting ripped off-- you're getting all the bits as advertised.

 These things can go up to at least a terabyte-- that's a trillion 1s and 0s!

So, back to the vending machine problem.
There are 4 buttons (A, B, C, and D) on the machine: 2 for coin input, and 2 for snack selection.

The buttons are like binary: 1 for a pressed button, and 0 for unpressed. 4 buttons means 4 bits. They could be anything from 0000 to 1111; the bits are labeled ABCD in the circuit diagram.

If you separate the buttons by function, you get two sets of 2 bits: with 2 binary digits, you can have a total of 4 combinations: 00 (combo #1), 01 (#2), 10 (#3), or 11 (#4) and you can assign a meaning to each combination.

For the coins, 00 (#1) is no coin, 01 (#2) is a nickel, 10 (#3) is a dime, and 11 (#4) is a quarter.
For the snacks, 00 is no snack, 01 is a snack that costs $0.05, 10 costs $0.10, and 11 costs $0.25.

The vending machine has to decide whether or not to give you the snack you selected based on the coin you put in, what change (if any) to give you, and to return your money if you don't get a snack. This stuff makes intuitive sense but we had to convert it all into binary because that's how circuits work.

The two circuits we built in the lab collectively output the value of the first coin in the coin return (a second coin only appears if you buy the 5-cent snack or the 10-cent snack with a quarter).

So, either you will get no coin (00), a nickel (01), a dime (10), or a quarter (11) back.

Since each individual circuit only outputs either a 1 or 0, we had to make two during the lab: one for the bit on the right, and one for the bit on the left (e.g., a 0 and a 1, respectively, for a dime as the first coin in the return). The first circuit is the one drawn on the paper, and the second one is the picture of the actual circuit.

Again, this is not even close to the complexity of computers and cell phones or even a real vending machine. Don't take electronics for granted. Appreciate how mind-blowingly complex they are.