On Sun, Jul 20, 2003 at 10:12:11PM +0000, Robert J. Chassell wrote: > Thank you. Am I right in thinking that for air, this is the heat > conductivity for still air, and not the heat transfer capabilities of > moving air?
It is the thermal conductivity of air, which is almost the same whether the air is still or moving (it results from excited gas molecules bumping into adjacent molecules and exciting them in turn). However, if the air is moving, heat can be transferred by the moving gas molecules -- the molecules themselves have energy, and when the molecules move, they carry that energy which can excite other molecules at a different location, thus transferring heat. That could be called a convection current. Since the thermal conductivity of air is so low, in most cases more heat can be transferred by moving air than by conduction through the air. > I can tell you by personal experience that a cold wind can chill a > human body fast. Cold water will conduct heat out of a body much more efficiently than cold air can conduct heat away. Of course, if the air is moving, it will be more efficient at taking the heat away from the body as compared to still air. A windbreaker shell covering a fluffy layer is thus most effective (when dry) at keeping you warm -- the outer layer stops the wind from carrying heat away, and the fluffy layer traps air, which has a very low thermal conductivity, to slow the transfer of body heat by conduction. Heat transfer by natural and forced air convection are complicated processes. I'm not aware of any first principle calculations that are useful for modeling heat transfer between surfaces and air. There are some phenomenological formulas, however: natural convection ------------------ W/m^2 = h (delta_T)^(5/4) h ~ 1.3-2.5 W/m^2 K^(5/4) (value depends on shape and orientation of surfaces) forced air convection --------------------- W/m^2 = eta delta_T eta ~ 5-60 W/m^2 K eta depends on airspeed and slightly on delta_T Here's an example comparing heat transfer from a body to either (windy) air, or to cold water. The normal human skin temperature around the torso is about 33C, which is 306 Kelvin. We'll look at heat transfer from the skin to either a moderate wind at 7C, or to still water at 7C (280K). Assume the surface area of the skin is 2 m^2 (rough estimate assuming all the skin is exposed) and that half of the skin is in a wind shadow (opposite side of body from the wind). Then for forced air convection, assuming eta=50 for this windspeed and delta T, the heat transferred per second is: 50*1*(306 - 280) = 1300 W Compared to radiation cooling, which would be about 150W, this is quite large. If we model the body as being made up of water, with a mass of 80kg, then the heat capacity of the body is 4216*80 J/K = 337000J/K. Dividing the heat transfer rate 1300 J/s, by the heat capacity 337000 J/K, gives a temperature drop of about 0.004 K/s, or about 0.23 degree Celsius per minute. This assumes zero cooling by evaporation of sweat. This is just a rough estimate, since the core of the body is somewhat insulated from the skin, the real problem is quite a bit more complex. But this should be roughly correct. I found a paper http://www.arbetslivsinstitutet.se/publikationer/pdf_ah/2003-04.pdf that gives 1319 W/m^2 for wind chill in 4 m/s wind. For a body in 7C water, to make the math simple I again have to make some approximations. In addition to the ones already mentioned, I am going to assume that the distance the heat must travel through the body and water as it moves away from the body is 4cm (I chose 4cm because all the water within 4cm of the body weighs about 80kg, same mass as assumed for the body). Since the thermal conductivity of water is 0.6 W/m K, then the heat transfer rate is 0.6*2/.04 * (306-280) = 780 W which translates to a temperature drop of 0.14 deg C per minute. I found a link http://www.heat.uk.net/cooling2.htm that gives body cooling by ice water bath immersion of 0.2 C per min, so this estimate isn't too bad. If the water were moving (or the person were moving in the water), I wouldn't be surprised if the actual rate could be more than twice that calculated above. So, standing naked in a moderate wind at 7C and being submersed in cold water at 7C will result in roughly the same body chilling by convection and conduction. Of course, few people stand naked on a windy cold day, and clothing makes a big difference in keeping a layer of warm air trapped near the body, drastically reducing the wind cooling. In contrast, clothing won't help much in the cold water (unless you are wearing something like a neoprene wetsuit, which keeps the water away from the skin). > Hence the practical effect in any environment in which air can move, > such as the interior of a spinning space habitat, is that air has a > much higher possible heat transfer capacity than might be indicated by > its still air heat conductivity value. Yes. > Are you saying that there will be no air movements in the spinning > space habitat? No, just the opposite. If there weren't air currents, then no equilibrium would be possible. > Certainly, air at the top of a mountain I went to Friday did not gain > temperature from heat conducted through the rock that day from the > warm valley floor. How do you know that? > That air at the top of the mountain was cool. Did you measure the temperature of the rock at the surface and, say, 1 meter below the surface? Did you measure the air temperature near the rock and far away (miles) from the rock? When in sunlight, the rock will be heated by sunlight much more than the air (the atmosphere only absorbs about 25% of sunlight incident on the top of the atmosphere) and will thus be warmer than the air. Therefore the rock will transfer heat to the air as long as it remains warmer than the air, which will be for quite a while since the rock has a much larger mass and higher heat capacity than the air. Haven't you ever felt the outside of a brick or stone building after sundown, when it has been in the sun during the day? It is quite warm compared to the air temperature. > The rocks at the top of the mountain were cool, too, they could not > transfer much heat to the air. If the rock was even 1 degree warmer than the air, then heat will transfer from rock to air, and since the mass of the rock is so much greater than the air, it can make a significant difference in the air temperature. You even noticed this yourself -- you said that you see thermals above fields more than forests. The same goes for rock. > > I would expect ... that with sufficiently good insulation, the > > surface of the end caps would come to the same temperature as > > the air fairly quickly and would neither contribute nor take > > much heat from the air. > > Same temperature as which air? The air on the axis is colder than the > air at the rim. > > I meant the rock and the air at the same distances from the axis. If the lapse rate is 8 C / km (very rough approx), then at 5km height the air would be 40 C colder. If we assume the habitat has an inside length of 6km, then let's look at 3km of air being heated by one endcap. The volume is about 235 km^3. Very roughly, the air mass is 200e9kg. Air has a specific heat of about 1000 J/kg K, so the heat capacity of the entire air mass is about 2e14 J/K. If we put 1m of fiberglass insulation on the endcaps, thermal conductivity 0.03 W/m K, and assume the asteroid itself is all at the same temperature, then very roughly the heat flow to the air will be .03/1*pi*5^2*1e6*40 = 9e7 W, which results in a temperature change in the air of 0.04 C per day due to heat transfer from the endcaps. Quite low. However, the thermal conductivity of the air is comparable to the thermal conductivity of the fiberglass insulation, so it is unrealistic to assume that the entire 3km length of air is heated by thermal conduction from the endcap. If that 3km length of air is all at the same temperature, there would need to be quite a bit of air movement along the axis. If we looked at air 3m from the endcap instead of 3km, it could be heated by 40 C per day if it was not moving. Quite high. So perhaps 1m of insulation is insufficient. Make it 100m thick, and the air 3m from the endcap (if still) can only be heated by 0.4C per day. Probably not enough to create strong air currents. > Yes, but what is the rate of heat flow from the interior of the rock > through its surface into the air? How would it compare to heat > transfered by moving masses of air? What wind speed would it produce? How could this be answered without a simulation? Do you know of a simple model for calculating conduction, convection, and wind speeds? > Would the heat from the sides flow to the end caps and then be > radiated into space? Would the air by the end caps get chilled by the > heat being radiated into space? Or is the insulating quality of rock > and dirt good enough to make that fairly irrelevant? If the walls are 1km thick and have an average thermal conductivity of 80 W/m K, and if the asteroid has the same temperature/depth profile at all points on the surface, then radiation and conduction must be equal and we get (assuming emissivity of 1) (300K - T0) * 80 / 1000 = 5.67e-8 * (T0^4 - 3^4) assuming the outside temperature is 3K. The solution is T0=125K, which results in heat radiation of 14 W/m^2. The sunlight reaching the top of the Earth's atmosphere is about 1400 W/m^2. Even if only about one quarter of this is absorbed by the habitat (due to shadowed areas and imperfect absorption), it still seems the habitat would have more of a problem staying cool than staying warm, if it orbited at 1 A.U. from the sun. The habitat would need either to be in a further orbit (about 5 A.U.'s), or to have high reflectivity on its outer surface (99%), or to have lots of radiative thermal panels that stick out only on the shadowed side, enough to increase the effective surface area by about 100 (but that might be difficult since it rotates they would have to go in and out). -- "Erik Reuter" <[EMAIL PROTECTED]> http://www.erikreuter.net/ _______________________________________________ http://www.mccmedia.com/mailman/listinfo/brin-l