|MadSci Network: Engineering
Most thermometers depend on the differential thermal expansion rates of materials to measure temperatures. That is, we all know that materials expand in heat and contract in the cold; what's less obvious is that different materials do so at different rates. The rate is called "coefficient of thermal expansion". I describe three different designs below that reflect this principle and classic thermometer designs. My recommendation is to go with design #3 as being the simplest in concept and construction. Thermometer #1 -------------- The difference in expansion rates of different metals is the principle is used in mechanical dial thermometers. Consider two long strips of dissimilar metals are welded together along their lengths. As the temperature increases, one of the metal strips want to expand more than the other, so the assembly bends towards the side that isn't expanding as fast. In a dial thermometer, the bimetal strip is coiled into a spring that expands as the temperature rises and contracts as the temperature falls. One end of the sprig is fixed to the thermometer's case, and the other to the pointer. If you were to find two long, thin strips of dissimilar metal (say, brass and aluminum), these could be fastened together along their lengths using screws and nuts every few inches, or perhaps by gluing them with a strong glue such as epoxy. One would take note of the amount of curvature in the strips to detect temperature changes. Thermometer #2 -------------- Another approach is to take advantage of the different expansion rates of a liquid and solid. Galileo used this idea and Archimede's principle to invent his thermometer, the aptly named "Galileo's thermometer". Liquids such as water often expand much more than solids such as glass or metal. Objects float in water if the weight of water they displace is more than their own weight; if the weight of water displaced is less than the object's weight, it will sink. The weight of an object does not change with temperature. The density (weight per unit volume) of a liquid changes with temperature because although the total weight is constant, the total volume changes. For water above 4 deg C, but below boiling, water expands at the rate of about 0.2 cc per kg per deg C. That is, 1.000 liter of water at 25 deg C will occupy about 1.002 liters (that is, one liter plus 2 cc more) at 35 deg C. An object that just *barely* floated at 25 deg C will sink at 35 deg C, if it doesn't expand very much (its temperature is increasing, too). But glass doesn't expand very much compared to water. What Galileo did is have a glass blower make some sealed, air-filled glass balls. Galileo then carefully added weights to these balls so that each had a specific temperature at which could no longer float. That is, balls with a lot of weight would sink when the temperature was low, while balls with less weight would float until the temperature got higher. The balls were marked with the target temperature at which they would sink. Galileo's thermometers can often be found in scientific gift shops (search the Web for "Galileo +thermometer" to see illustrations). I have one at home and it works quite nicely. You could try reproducing this invention using water and any kind of rigid floating object. For example, small bottles in which you put a little sand (but mostly air) might work. If you have access to a chemistry lab (or better yet, the chemist!) with accurate scales and volume measuring equipment (you need to measure the displacement volume of the bottles accurately), you could assemble the weighted bottles in fairly straightforward fashion. One would need to reference a table of water density versus temperature, with the objective to weighting the different bottles to get to the same average densities for different water temperatures. The scales would be used to measure the mass of each bottle and deterrmining how much sand to add. Alternatively, one could get there experimentally (as did Galileo) by floating the bottles in tubs of water at different temperatures until just the right amount of sand is in each (and checking that there's not too much or too little by seeing if they float or sink in tubs colder or hotter than the target temperature of each bottle). This design (#2) is probably the fussiest of the three to get working right. Thermometer #3 -------------- The simplest approach of all (sorry I made you wait 'til the end) is to reproduce the bulb thermometer on a grand scale. The bulb thermometer has a reservoir of liquid (alcohol or mercury, typically) that feeds a very narrow channel inside a glass column. As the liquid expands with temperature, the level inside the columns rises; as the temperature decreases, the level sinks. Why not do this with a glass jug, a stopper, some clear, thin plastic tubing, a funnel, a long thin stick, a small stick to plug a hole in the stopper (see below), and some duct tape? (No good science experiment can be done without duct tape; however, any kind of reasonably waterproof adhesive tape will do for your purposes). Make two holes in the stopper that snugly fit the plastic tubing. Fill the jug with colored water, and insert the stopper. Use some of the plastic tubing to connect the funnel's outlet to the jug (via the hole in the stopper). Insert a different piece of plastic tubing in the other hold in the stopper part way into the jug. Use the tape to affix the stick to the jug so that the stick stands vertically, then tape the second piece of tubing to the stick. The glass jug is the reservoir of your thermometer and the tubing taped to the stick is the "glass column". All that's left to do is to use the funnel to force in a bit more liquid to the jug to force out all air and to get some liquid partway up the other tube (the one taped to the stick). Remember that "water seeks its own level". Once you've done that, remove pull out the tubing connected to the funnel and plug its hole through the stopper (this is where the "plug stick" comes in). This will take some practice and deft application of fingers, as the water partway up the column will want to come spurting out of the hole formerly occupied by the filling tube. You need to use a glass jug, as opposed to plastic, to ensure rigidity, so that as the water expands and contracts it must rise and fall up and down the thermometer's column. The other trick to making this work is to force out all the air bubbles. Again, you want to make sure that the water has no place to go when it expands except up the column. A gallon jug (about 4 liters) will work nicely if thin tubing (say, 1/8" to 1/4" diameter, that is, 3 to 6 mm diameter) can be procured. I'm thinking of the clear plastic tubing that's often used by hobbyists who keep aquariums to supply air to their water filters from an external air pump. A pet shop or well stocked hardware store should also be able to fix you up with some of this tubing (it's quite inexpensive). With a 4 liter reservoir and 5 mm diameter tubing, you should get a rise of about 2 cm of liquid per deg C temperature rise. But remember, it will take time for this amount of water to change temperature, so don't expect rapid response times. You should easily see temperature changes through the course of a day, though. This apparatus will be a bit messy to set up, but probably the easiest and most accessible to your students in terms of concept and materials. Have fun! Steve Czarnecki P.S. A simple "dinner table" experiment to demonstrate the thermal expansion of gasses: Take an empty glass soda bottle ("a Coke bottle" or a wine bottle), wet the top of the rim with saliva, and place a small coin over the opening. A U.S. dime works well with one of the old 16 oz. returnable Coke bottles. The saliva acts to help seal the air inside the bottle. Carefully wrap your hands around the bottle as it stands on a table top. Continue holding the bottle, being careful not to shake or move it. Watch the coin over the opening very carefully. After five minutes or so, you should see it pop up a bit and come back to rest, with the cycle repeating a few times over the next few minutes. What's happening is that your hands are warming the air inside the bottle, which expands, forcing the coin to flip up a bit and allow the excess pressure to escape. It takes time to get this going because you have to warm up the glass bottle and then the air inside. Note that a glass bottle is essential to making this work (the air must have no where to go except to dislodge the coin). Use the smallest (i.e., lightest) coin that will just cover the bottle opening.
Try the links in the MadSci Library for more information on Engineering.