« October 2004 | Main | December 2004 »
November 22, 2004
Sustainable Electrochemical Energy
If the goal is sustainable energy, where might we find it?
We all know about the possibilities and realities of solar, wind, and to a lesser extent tidal and geothermal power, but I thought there might be another untapped and under-researched area: junk yards.
Millions of tons of purified metals are discarded each year* in landfills, in the form of construction refuse and personal and industrial appliances.
These metals were expensive to mine and purify, and represent wealth to some degree, except for the fact that they are scattered about.
But in some developing countries, where scavenging objects is part of daily life for some people, and electrical energy is often in short supply, electrochemical energy from recycled materials might have promise.
EXPERIMENT I - Sandwiches
My first experimenting involved replicating Volta's pile - the first source of electrical energy, predating the mechanical generator by 20 years**.
I didn't actually realize Volta had done essentially the exact same experiment as I did, over 200 years ago, but I find myself often in that situation.
The experiment is simply to use two dissimilar metals, preferably as far apart as possible on the electropotential spectrum*** and separate them with an electrolyte - any water solution of salt, acid, or base****.
I used a copper penny and some aluminum foil as those are cheap and available and well-spaced electropotentially.
For the electrolyte I used a combination of cider vinegar, household bleach, and kosher salt in water.
A piece of paper towel is dampened with the solution and placed between the penny and foil, the electrodes in our simple pile.
The paper is necessary to separate the electrodes to prevent them from shorting out.
Right away, there is some potential and some current:
0.49mA
0.64v
Not enough to do much with, so the next step is to explore some more.
---
* fabricated statistic
** http://www.schoolscience.co.uk/content/3/physics/electric/
*** http://www.hilaroad.com/camp/projects/lemon/electric_potential.html
**** Almost any liquid will work, but the more 'stuff' in it, the better. Even saliva will work
------
EXPERIMENT II - Containers
I was looking for sources of good electrodes (aluminum, zinc, copper, nickel) and experiemnted with aluminum beverage cans, and zinc-coated steel cans*.
The test involved suspending a penny in an aluminum beer can and an emptied can of pineapple, both of which had been rinsed and filled with salt water.
In both cases, the results were about the same as Experiment I.
The zinc coating in the second can rapidly dissolved in the solution, and I chose not to pursue that.
The aluminum can also wasn't worth pursuing. I don't know if the can is glazed in any way at the factory, but even if not, having the container also be one of the electrodes isn't a good idea, as eventually a hole could appear, spilling acid.
---
Also at this time I began looking for a test application of the electrochemical energy source I hoped to build.
Lighting an LED is sort of the "Hello, World" of physical computing projects, and that would be a reasonable start.
Ultimately I hoped to perform electrolysis of water into hydrogen and oxygen**.
---
* Certain acidic foods, such as tomatoes and pineapple, are packaged in cans that are either lined with plastic, or galvanized with zinc.
** Another example of me being 20 decades too late. Davy used a voltaic pile to discover this phenomenon. I was barely able to reproduce it.
------
Experiment III - Series
Next I tried stacking the mini-piles.
The term 'battery' refers to having a bunch of piles, where a 'pile' is the literal pile of metal disks like those in experiment I.
A true pile would essentially be several of the penny-foil sandwiches arranged in series, one on top of the other.
2 sandwiches in series:
0.486mA
0.75v
2 sandwiches in parallel:
0.89mA
0.53v
Neither the voltage nor current doubled as I expected, though there was significant gain in the two positions.
I suspect that there was too much electrolyte and ions were flowing around all sides of the metals, not just where they met the membrane.
I tried several kinds of membrane, bits of cloth, paper, cardboard, etc. The best seemed to be thin paperboard (unpainted).
If some is good, more is better, so I stacked a bunch more, to the point where the pile began tipping, and I had to build some stands
Here are the results:
One Pile of about 20 pennies and bits of foil:
0.55mA
4.26v
Two piles in series:
0.46mA
7.9v
Two piles in parallel:
0.91mA
4.6v
Again, the expected doubling didn't quite happen, but it was close.
I was surprised that I was able to get nearly 8 volts from literally what was lying around the apartment.
However, voltage doesn't mean anything without current, and I had yet to generate even one milliamp.
At least 100mA is needed to do anything interesting, so I needed to either:
- use circuitry like those used in 'solar engines' to store and release energy from low-current sources such as solar panels
- make a bunch more piles until I had enough
- try something else (see below)
The circuitry was certainly possible, and may have been the right route to pursue.
In fact, I did try it, but I found that the wait time was hours before the LED would light, though it would stay on for about half a minute.
The piles wouldn't work out I realized, because I was two orders of magnitude below where I needed to be in terms of current.
Even if my apparati were completely efficient, I would need 100 piles like the ones I had made to achieve adequate current.
---
At this point I was also giving up on the electrolysis of water.
Although I had succeeded at generating lots of little bubbles, I was unable to store the hydrogen gas adequately.
So, in the spirit of of the rest of the project I decided that my ultimate draw on the power I generated would be a small electric motor I built a few years ago.
I tried running the motor with various batteries to see what the power consumption was.
The following 4 batteries all ran the motor, though the 9-volt battery was extremely fast and the AAA battery was just barely able to move it.
9-volt battery
~1A
8.78v
C-battery
122.0mA
1.38v
AA-battery
138mA
1.48v
AAA-battery
117.7mA
1.36v
Based on this, I figured I needed at least 1 volt of potential and 100 milliamps of current.
I was easily able to generate more than enough voltage, so current became my goal, and so I 'tried something else;'
------
Experiment IV - Parallel
What is a parallel connection but simply a larger one, one that increases the surface area of contact?
I read about how to make home-made capacitors, by simply using two sheets of aluminum foil separated by a piece of paper, which can then be rolled up*.
I thought about pounding pennies into sheets, but since 1982, American pennies have been made of zinc, with only a plating of copper.
Zinc and copper act as opposites in terms of electropotentiality, so any nick in a penny made since 1982 would cause electrons to flow between the zinc inside and the copper outside, rather than between the copper and the other electrode.
And smashing them flat would result in a sheet of zinc with bits of copper showing.
So I cheated and got some copper flashing (used in roofing houses).
The 20' roll has a tar-and-paper back which can only be removed by vigorous scraping**, but that's okay as it meant a built in barrier between the metals so I wouldn't have to add that.
I cut a piece of copper about 4" x 8" and laid on top of it a larger piece of paper towel wet with the same solution used in the piles.
On top of that, I put a piece of aluminum foil***, off-center.
I wrapped it up, with the positive terminal being where I had scraped tar off the copper, and the negative terminal where a bit of foil stuck out
The results pleased me:
0.58v
25.1mA
The same voltage as the previous experiments, but about 50 times the current.
So I made a bigger one, with about 3 feet of copper instead of 4 inches.
The results surprised me:
0.58v
46.2mA
Exactly the same voltage, but less than double the current, even though I had increased the area of contact by 9-fold.
So I tried an even bigger one, with 6 feet of copper, but got no results.
Unwrapping I saw that the paper towel had torn while I was rolling it, shorting out the insides.
So I cut it in half and made two more out of it, which had pretty much identical power readings to the other large one.
---
* http://www.scitoys.com/scitoys/scitoys/radio/homemade_radio.html
** I tried burning it off, but there must be some flame-retardant stuff in there - and flame-resistant material is hard to burn off!
*** It seems using the shiny-side of the foil against the paper worked best.
------
Experiment V - The Final Test
If my numbers were right, I only needed 4 of these batteries (2 parallel columns of 2 in series) to get the 1 volt/100mA I wanted to power the motor.
2 rolled batteries in parallel:
0.55v
44.1mA
2 in series:
1.1v
54mA
3 in parallel
0.53v
40.4mA
3 in series
1.56v
36.9mA
I wired 4 of them all up and connected them to the motor. I had just under the readings that I wanted, but with a nudge to the motor, it spun on its own for nearly a minute.
I tested the power readings again and found they were already deteriorating.
They seem to have a half-life of about 3 hours
SHELF-LIFE
All the piles degraded quickly after construction.
The penny-piles were completely inert after sitting overnight.
This partly due to the electrolyte evaporating, and re-wetting the membranes brought some back to life.
The penny-piles can be hydrated with a mister like those used to spray household cleansers, and in fact, the cleansers work well as electrolytes.
The flashing piles are more difficult to rehydrate, but drying is likely to be the issue since they dropped from their peak potentials and currents to about half or a third after only three hours.
In this case, I think the problem is small shorts within the rolls, at points where the membrane tore while rolling.
CONCLUSION
It is possible to create electrochemical energy sources using easily-found junk.
However, it is a true micropower source, similar to solar power, and the likely way to use this energy source is by storing the power in capacitors.
------
Epilogue - Can't Give up Yet
I unwrapped the first roll, the little one, and saw little cracks in the copper (perpendicular to the line of rolling) through which I saw the back of the tar paper.
Maybe I need an impervious membrane that separates the copper from the aluminum that is better.
So I unrolled it and rerolled it inside a plastic grocery bag (maybe we now have a use for those).
I also started using lemon juice in my electrolyte instead of vinegar because it smells better*.
The new results:
29mA
0.48v
Slightly higher amperage, slightly lower voltage, but it fluctuates so much it's hard to tell - really about the same.
It may be the plastic, or it may be that I rehydrated the paper towel.
Opening one of the big ones revealed foil with lots of pitting in it, and copper that was heavily tarnished (black, cupric oxide? this is a red/ox reaction after all)
Wrapping it up the same way, with plastic (but not rehydrating the paper) gave:
0.54v
wildly fluctuating current but peaking around 40, then settling back down to dead-level values.
The last thing to try is not wrapping it at all, just using a big sheet of copper (12 feet should work) with wet newspaper and a roll of foil over that.
Unrolling another big one and laying it flat brings the numbers back to where they were
The aluminum is significantly pitted - hold it up to the light and it looks like stars.
Both metals were irregularly corroded. In places the copper was turning green where it was exposed to air and electrolyte but no aluminum on the other side.
This tells me that really flat is important, and I wasn't able to keep it flat when rolling it.
It was harder to keep it flat when the rolls were bigger, which may explain why the bigger rolls didn't scale up as I expected.
---
* Bleach is stored in opaque containers because it breaks down in the presence of light. The stuff I used started to smell odd, not bad, but more like peroxide than chlorine.
Posted by mslaybau at 08:06 PM | Comments (0)
November 12, 2004
Review: Conqueror!
Game Design
November 11, 2004
Conqueror! (http://www.conquerorgame.com/ ) follows the tradition introduced to me via RISK, and continued with products such as Axis & Allies - games that simulate warfare by having players manage armies on a playing board that represents an actual map of the world - or in the case of Conqueror!, a map of Europe circa AD 1000. Unlike its predecessors, however, Conqueror! is not a board game but an online multiplayer (up to 16) game, currently in beta development, and currently free.
Making the game digital has allowed the authors to automate much of the more cumbersome tasks related to resource management. In games such as Axis & Allies, there is a practical cap on how complex, and thus how realistic, the game can be because of limitations related to the number of elements a human player can feasibly keep track of in a given turn. These limitations don't exist with computer-based games, so they are better able to simulate real-world, complex scenarios.
GAME PLAY
On each turn, each player can alter how each of his territories produce - developing resources (farms or culture), raising armies, levying taxes, or building defensive castles.
TURNS
Games traditionally have been turn-based because that was the only way to handle logistics in board games. Turn-based games tend to be slower while allowing more contemplation on the part of each player.
Real-time strategy games allow a much faster game at the expense of deliberated planning. Conqueror! has come up with a good compromise, having players each place orders over the course of a few minutes, the orders are then processed simultaneously by the system. This allows for more contemplation and planning while keeping the game faster than its offline brethren.
EDUCATIONAL VALUE
Realism has little, if anything, necessarily to do with how fun a game is, but in this case the realism is important (to me anyway). I found this game to be educational, in a limited way, in how I was unable to not learn some geography from playing. The geography was not part of the instructions, and never explicitly forced onto the players, but was just a natural consequence of playing the game. This type of education is the model I hope to use in educational games I develop, where no memorization - or even conscious thought - is required to absorb the information available.
MULTIPLAY VS AI
The developers have clearly done a lot of work in developing the various artificial intelligences assigned to nations not controlled by human players. Some are unpredictably aggressive, and some (according to the developers' notes) even have built in 'revenge' algorithms so that they are more likely to attack players that have attacked them.
However, playing versus humans is very different than playing against AIs. Even discounting the common trend of 12-year-old boys (or adults regressing in anonymity to being 12 years old again) being obnoxious, human opponent are so much better at anticipating other players' moves that they require a much more defensive strategy to defeat.
Human players also add the optional element of alliances - which typically seem to consist of working together to beat the AIs, then fighting each other (perhaps a representation of what the War of the Robots will look like).
ADDICTION
The central fun of this game comes from what I call the 'lawnmower principle', the premise of which is that it is inherently pleasurable to see the effects of one's labor - to 'wipe the slate clean'. Conqueror! is more than mowing the lawn, however, because of the complexities involved in doing so; on any given turn there are between 5 and 10 decisions that have to be made; decisions that affect both this turn and future ones.
These decisions are typical of many games of this type - whether to build resources or attack.
But what makes the game addictive is what I will call 'scaling depth' - the fact that a new player can pay attention to only the more coarse details (whether one is being attacked or not, for example) and still take some ground (although likely lose eventually) while more experienced players know which of the more subtle details (which of one's territories are most suited to raising armies versus building farms, or how large of a front one wants to maintain, for example) can lead to victory.
This scaling depth is what keeps players coming back. A game such as Tetris is addictive because every game ends with the player thinking about how he would do a little better next time, and Conqueror! has the same quality.
SECOND-GUESSING
I’m making this its own section since second-guessing is so critical to multiplayer strategy games. The second-guessing in Conqueror! is quite good, and perhaps the dominant quality of the game, at least when playing human opponents. When playing AIs, it’s much easier to figure out what their driving algorithm is, and act accordingly.
USER INVOLVEMENT IN GAME DESIGN PROCESS
Conqueror! is a game-in-progress, having gone through perhaps 200 iterative builds (to date, the build number is 86, but the developers typically release a, b, and c sub-builds) in the past couple of years.
They have an active forum (http://conquerorgame.com/forum/index.php?f=6) where players can report bugs, plan games, and request features. For armchair game developers such as myself (I'm currently lifting myself out of the La-Z-Boy) this is a valuable window into the process of creating games.
After playing a few times, I started flooding the list under different names, with concerns over the same issue: that the end-game is too easy. The positive feedback is so strong (as it is with most War games) that once a player has conquered about a quarter of the board, he's unstoppable. My suggestion was that a better simulation of Empire would be to have more rebellions in territories far removed from the capital, that boarding soldiers should be more expensive when far from the capital, and a few other related ideas.
I was surprised and excited when I saw all of my ideas implemented in the next build, although in some ways, the game was better before my ‘fixes’ were added.
Still, it is the only environment I know where players can affect the conditions of the game so directly - even democratically, since my forum posts received numerous follow-ups with both supporting and refuting arguments.
IMPROVEMENTS OVER SIMILAR GAMES
There are many, but two interesting ones are devotion level within a territory and influence over adjacent ones. Having a variable for how happy citizens are is not new, perhaps first introduced in the Civilization series, but in Conqueror! the value is tied to resource production and likelihood of rebellion in a compelling way.
A territory’s influence is based on the citizen’s devotion to the player, the number of armies in the territory, the level of culture that has been achieved, and the amount of money in the coffer. A territory is influenced by all adjacent ones and at times of invasion or rebellion, the influence can determine how much the soldiers resist being invaded, and the differences in influence are great enough, can even lead to secession to another player without having to attack.
Overall, Conqueror! is an enjoyable game that you can have 30 or more hours of fun with before getting bored. Considering that it's free (as long as it's still in development) that's a bargain.
KRAMER SCALE:
Based on Wolfgang Kramer's list of issues to consider in evaluating games, I'm rating Conqueror! on a series of 5-point scales. Kramer's list is imperfect, but good as a starting point when evaluating any game.
ORIGINALITY **** Good
War simulations are thousands of years old, but all game genres are well-established so I can't criticize this game for producing yet another "improvement on Risk".
The originality in Conqueror! is in how the developers took lessons from previous games such as Axis & Allies and those by Avalon Hill and combined them with some fresh approaches, specifically regarding how to exploit the advantages that computer games have over board games in terms of allowing lots of behind-the-scenes calculations.
FRESHNESS AND REPLAYABILITY *** Fair
Conqueror! has a replayability value of about 30 games, which is pretty good, but doesn't compare with truly addictive games such as Tetris. There are more problems in this regard when playing just AIs than there are when playing human opponents because the second-guessing is so much easier with AIs.
But, either way, there are a small number of overall strategies: build then attack vs. attack while building, protect your base vs. extending out in order to flank other territories, etc.
Also, regardless of the type of opponent, there is still a significant wait in between turns and lots of required clicking in order to maintain production of resources, both of which lead to boredom once you've figured out all the intricacies of the game.
SURPRISE ** Poor
The AIs in Conqueror! are sometimes quite good, acting in ways that mimic the best human players. In fact, human players, though more challenging overall than the AIs, are often more predictable than the AIs, which clearly have some randomness built into their algorithms. But, reacting to randomness is much less satisfying than reacting to a brilliant but unexpected move, as is often found in other strategy games.
EQUAL OPPORTUNITY **** Good
Creating true equality in games in games (as in life) is very difficult, particularly for turn-based ones. But Conqueror! does a good job in evening out the usual bias that favors the first turn.
The real-world map is what creates the most inequality in this game. Countries that begin with few borders (Morocco borders only Tangiers) have an early advantage in that they can invest less in defense. However, these same countries have a disadvantage in the later game when they need to expand to the opposite side of the board. Countries that begin in the middle have a more difficult time early on, with attackers on all sides, but then have it easier later as they have equal distances to cross to reach the edge of the playing area in any direction.
WINNING CHANCES * Bad (but doesn't matter)
This rule is different for board games compared to computer games. Risk-style war simulations require that nations can be obliterated during the course of the game. Most computer games allow players to quit and start over when they're killed. Conqueror! allows player to sign in as another country when killed. This itself can be a problem for other players when a defeated player returns.
NO "KINGMAKER EFFECT" **** Good
Heavy positive feedback ensures that losing countries dwindle in strength before they're crushed, so there is no kingmaker effect in that regard. However, a 'feature' of Conqueror! is that when a player quits, his nation is taken over by an AI, which can be frustrating for the remaining players.
I typically quit a game once I've reached 'Emperor' level, at which point I usually control about a quarter of the board. The AI that takes over for me usually acts in erratic ways, which are magnified because the resource base I left with it is so strong.
NO EARLY ELIMINATION ** Poor
Again, war games require that some players can be eliminated, but Conqueror! does have a flaw in that some games begin with such unfair disadvantages (multiple strong, aggressive, adjacent opponents) that survival is impossible. You can simply quit and start over, but that's no fun.
REASONABLE WAITING TIMES *** Fair
When setting up a game, you can choose how long to set the timer (15-second increments between 15 seconds and an hour) which is very useful, but some turns require little decision-making while others require a lot, and being locked into the time can be frustrating, especially when trying to realize a particular aggressive strategy. Turn-based games have to include some wait-time, but when playing Conqueror! I yearned for the dynamically flexible wait time in Chess, with periods of contemplation punctuated by rapid turns of aggression. ‘Rush-games’, with only 15 or 30 seconds per turn can be quite challenging, and give the AIs a strong advantage.
CREATIVE CONTROL **** Good
There is some randomness in the determination of losses during combat, but overall Conqueror! offers good creative control. The only drawback is occasional, when the erratic behavior of some AIs effectively punishes a player for no reason.
UNIFORMITY and QUALITY OF COMPONENTS ***** Excellent
Verisimilitude is not necessary for a game, but I really enjoyed the quality of the map and graphics in general - making the game far more enjoyable than if I were to 'play' an Excel spreadsheet with all the algorithms as embedded equations in different cells.
TARGET GROUPS AND CONSISTENCY OF RULES *** Fair
This is one of Kramer's most debatable rules, and Conqueror! is clearly an exception. It is a strategy game that relies on randomness to some degree. However, the randomness is well-enough contained that the luck factor enhances rather than detracts from the game. I'd be curious to play this game without any chance factor, though.
TENSION **** Good
This varies greatly between games against AIs and those against human opponents.
Overall, though, there is a good mix of paranoid building phases, sudden reactionary defense phases, and planned aggressive phases.
LEARNING AND MASTERING A GAME *** Fair
The initial learning curve is steep (assuming time on the Y axis and learning on the X axis) to the effect that simply understanding all the rules requires playing about 3 games.
Unfortunately, mastering the subtleties takes only several games more, without a broad, pleasant band of discovery in between those two moments. Beyond that, the game is still enjoyable as players can experiment with different strategies.
COMPLEXITY AND INFLUENCE *** Fair
Similar to Csikszentmihalyi's (phew!) 'flow channel', Kramer’s graph of complexity and influence shows how a game should scale along with the players' abilities.
Conqueror! is fairly rigid in this regard, being a little too difficult for new players and too simple for experienced ones.
Posted by mslaybau at 03:11 PM | Comments (0)
November 05, 2004
White Flour, White Rice
I just realized a possible reason why Asian cultures historically separated out the bran and germ from rice, and why Europeans did the same with their wheat.
It wasn't for flavor or appearance (though those are the reasons now) but because bran has fat in it, which turns rancid after a few weeks, precluding the possibility of storing the grain without refrigeration.
Posted by mslaybau at 10:43 PM | Comments (0)
Chocolate-Orange Beverage
Chocolate and Orange is one of the less-appreciated yet more delicious flavor combinations.
Many agree on its gustatory appeal, but other than a few brands of orange-flavored chocolate candies, there is very little on the market to excite my tongue the way I wish it to be.
Ever eager to experiment, I wondered the following:
"What happens when you mix hot chocolate with Tang?"
I could have mixed Tang with chocolate milk, but from experience I know that mixing acidic beverages with milk often curdles it, leaving it unpalatable and gross.
So, I opted for a hot beverage and put a spoonful of store-brand hot chocolate powder in a mug, along with a spoonful of Tang powder (regular kind, not the new tropical fruit kind) and added hot water.
The Tang disolved right away, leaving a hunk of poo-looking chocolate crud at the bottom that had to be smashed and stirred in. The final color was rather reddish, like apple cider.
I let it cool and took a sip, which was surprisingly good, and definitely of the taste I had sought. It was too sweet, but that's because I added half the amount of water I would have added had I been making a mug each of hot Tang and hot chocolate.
The only error I made was in mixing equal parts orange and chocolate. One must think about balance in flavor mixtures. Which is the base? Which is the accent?
With peanut butter-jelly, an equal mix is appropriate.
With apple-cinnamon, apple is clearly the base, with cinnamon as the accent.
With chocolate-mint, the mint accents.
With chocolate-peanut butter, all sorts of mixtures seem to work, with either party dominating - like a good marriage.
Chocolate-orange, however, is not like that. Orange must be the accent.
So in my next cup of chocolate Tang, I'll only use half as much orange.
Posted by mslaybau at 10:12 PM | Comments (3)