Testing the warmth of a blanket.

March 9, 2009 2:03 PM

How can I empirically test the relative warmth of a blanket?

For science (and for my very strange job), I need to test three different blankets for "warmth".

My plan: to put three different people inside of three separate refrigerators (like cola refrigerators - the ones you see in convenience stores with the glass doors). My test subjects will be wearing nothing but boxer shorts and one of three test blankets. The refrigerators will probably be set to about 40 degrees Fahrenheit.

I am taking all safety precautions and won't be keeping my subjects inside the refrigerators for more than 30 minutes. Don't worry, there's going to be PLENTY of people on hand to ensure this is done safely.

How best can I test the "warmth" of these blankets? I'm thinking that if each subject wears a large pool thermometer around their neck, I'll get a reading from the amount of heat that exists between the person's body and their blanket.

Am I on the right track? Or will each thermometer read approximately 98.6 (because of its proximity to the body) after 30 minutes of testing?

posted by plasticbugs to science & nature (23 comments total)

Why do you need people at all? Why not just three thermometers?
posted by grumblebee at 2:06 PM on March 9

I think you need to remove the human variable from the experiment. Substitute a space heater or something that you can calibrate exactly?
posted by otherwordlyglow at 2:08 PM on March 9

Why do you need test subjects? Why not use three identical jugs of warm water, for example, and wrap them similarly with different blankets. Take a temperature reading at various intervals to see how well the blankets insulate.
posted by JJ86 at 2:08 PM on March 9 [4 favorites has favorites]

Or wrap a jar with ice cubes in it in each blanket. Make sure you have an equal number and weight of ice cubes in each jar.

Putting people in fridges is always a bad idea. Depending on a thermometer around someone's neck is imprecise because results will depend mostly on how far from the skin the thermometer rides.
posted by musofire at 2:09 PM on March 9

I think you absolutely need test-subjects, because the relative warmth of a blanket is hugely subjective. How cozy is it? How easily can one wrap it about oneself? Does it reach my toes?

I do, however, think you need to use the same subject over and over again on different days. Some people are simply furnaces and some aren't (you know what I mean; you've cuddled around a bit, I'm sure), and this could reflect unfairly on certain blankets.

Either way, this is not going to be science.
posted by tapesonthefloor at 2:12 PM on March 9

Why do you need people at all? Why not just three thermometers?

Because a blanket needs a body to conserve body heat - otherwise the thermometer would just show the temperature of the inside of the refrigerator after 30 minutes.

I could use hot water bottles instead, as that would probably be the easiest way to measure "heat loss". Wrap water bottles in blankets, put water bottles in refrigerator, wait 30 minutes, see what the water bottle temperature is after 30 minutes.

HOWEVER, that wouldn't be fun - this set-up is for a demonstration in front of an audience. I need to use people.
posted by plasticbugs at 2:12 PM on March 9

Well as long as you have all three people test all three blankets and evaluate them, thats a basically sound strategy. Small sample size though, especially since the metrics will vary by subject (people do have different body temperatures, adapt to temperature differently, etc). Also, "warm" is a relative concept. Retention of heat is measurable. You need to define your terms better. If you are really just looking for what people consider to be "warm", you dont even need to measure anything. All you need is a survey.
posted by elendil71 at 2:13 PM on March 9 [1 favorite has favorites]

If the refrigerators are sealed well enough to keep the cold inside, don't you risk suffocating your subjects? (I'm pretty sure there's a law requiring fridges to open from the inside because people used to get trapped in them and suffocate.)

From a rigorousness standpoint, your experiment also fails, since people's body temperatures and organic responses to cold vary widely.
posted by decathecting at 2:14 PM on March 9

Could you use something like blocks of ice. Wrap them in the different blankets, and measure melt.

OR

Can't you find a thermal imaging camera that can measure the body temperature of your test subjects?
posted by KokuRyu at 2:16 PM on March 9

On preview, what elendil71 said. If you must use live test subjects, you need to do three rotations of this experiment, with each test subject testing each blanket. Further, each round of testing must begin with the blankets and test subjects at the same initial temperatures as the other rounds of this experiment.

Also, as has been pointed out, the sample size is too small to be statistically meaningful.

Finally, it would help to have a frigid, blanket-less person as a control.
posted by mosk at 2:23 PM on March 9 [1 favorite has favorites]

3 different people will definitely lose heat at different rates, based on body size (a taller person has more surface area, and it's less likely that the blanket will cover all of them), and subcutaneous body fat, and body hair, and how thick their head hair is, and the actual size of their heat-losing heads, and how good the circulation is to their extremities, and whether they breathe through their mouth or nose, and gosh, who knows, probably a huge whack of other factors. So this won't be scientifically accurate, although it will be funny to watch as long as you don't suffocate your subjects.

You could put them in Hypercolour-type underwear for even more of an amusing visual.

Maybe along with the people, put in other blanket-wrapped warm objects, to test how the blanket insulates on a more objective level? A glass bottle of coffee or hot soup, a tupperware of microwaved Jell-o, or-- ooh!

A bottle of hot gravy, or melted butter, or olive oil- something that congeals disgustingly when chilled!

...each wrapped in a piece of the same blanket, and their temperatures measured before and after the 30 minutes. That way you keep the funny visual of the cold dude, but add in a layer of objectivity.
posted by pseudostrabismus at 2:25 PM on March 9

These are all good points. I'm considering giving each test subject a hot water bottle to take in there with them. Then I can measure the temperature of the water bottle in addition to their subjective thoughts regarding how well the blanket covered them, the feel of the material against the skin, cozyness, etc.

I'm obviously not going to be able to write this up in a scientific journal as these test conditions are not very ideal. However, I want to judge these blankets as well as possible, even if we do make the test a little silly in order to "play it up" for audience.
posted by plasticbugs at 2:28 PM on March 9

Second KokuRyu's suggestion of thermal imaging. Various TV shows I've seen that have done similar experiments used thermal imaging, and other than looking high-tech it does give the viewer a really good idea of how the body temperature changes.

One thing worth pointing out is that the human body is not a constant temperature throughout. The kinds of body parts that suffer from frostbite the most (fingers, toes, noses, etc.) are the ones that are most susceptible to cold conditions, so they would give you the lowest temperature readings and the biggest swings between normal and freezer temperature.
posted by burnmp3s at 2:30 PM on March 9

Not a bad idea. Have the water bottles at identical temperatures, and instruct the subjects to cuddle them lovingly in the fuzzy folds of each blanket. Then, simply, test the exit temperature of each bottle.

I like, I like. I also must insist that this hits YouTube eventually. With or without the hypercolour underwear.
posted by tapesonthefloor at 2:31 PM on March 9

You should also consider the fiber weave of the blankets, meaning either you either need to ensure there is minimal air movement, or you need to quantify the effect of each. A breeze will blow through a cotton blanket. A fleece blanket doesn't fare much better. A wind-stop fleece (generally having some sort of a neoprene or spandex weave) will perform much better.

An other considerations: blankets wrapped tightly against the skin have a very different transfer than blankets wrapped loosely. In other words, by creating a pocket of space between the blanket and the fixed heatsource core, you will improve the heat retention.

So I'd go:
(#1) thermocouple in the middle of the water jug
(#2) thermocouple on the outside surface of the jug
(#3) thermocouple 1-2 inches off the edge of the jug
(#4) thermocouple on the inside of the blanket
(#5) thermocouple on the outside surface of the blanket
(#6) thermocouple 2-3 inches off the edge of the blanket

Personally I'd conduct this experiment at 0mph wind, 5mph wind, 15mph wind, and 25mph wind, and varry the temperature of the water jug at 96.6, 98.6, and 100.6 degrees. I'd be sure to try it with each blanket, and then with no blanket. I'd then repeat the experiment at least twice.

Mathwise, I'd look up: DOE, Factorial Design, and GLM for empirically proving the performance of a given blanket.
posted by Nanukthedog at 2:34 PM on March 9 [1 favorite has favorites]

Because a blanket needs a body to conserve body heat - otherwise the thermometer would just show the temperature of the inside of the refrigerator after 30 minutes.

Not really true. The object of a blanket is to insulate, and the better the insulating value of the blanket, the longer the warm object inside (body or jug of water) stays warm, the better the insulating value of the blanket. Non-electric blankets don't create warmth, they only conserve it. So what you want to measure is how well a blanket conserves heat. You don't need an actual body in order to do this.

To test warmth on clothing, the industry standard has long been the Copper Man test, which involves using a life size thermal mannequin to test heat transfer rates.
Here's another study that used thermal mannequin's to simulate newborns in order to do some testing about heat stress.
posted by anastasiav at 2:35 PM on March 9 [1 favorite has favorites]

Personally I'd conduct this experiment at 0mph wind, 5mph wind, 15mph wind, and 25mph wind

These would be good variables for outerwear, but I'm not sure when you'd encounter 25 mph wind in the bedroom..... Your point about how close the blanket is to the body is a good one however. This would also relate to the "drape" of the fabric -- a very stiff wind-block fleece would perform differently than a very supple one.
posted by anastasiav at 2:39 PM on March 9

I thought it must be for a demonstration when you mentioned the glass fridges. Sounds like fun!

I suggest you have a thermometer inside the blanket, next to the subject's skin, and another one outside the blanket. The difference between them should give you some idea of how much body heat is making it through the blanket. Maybe even have more than one thermometer pair, at different spots on the subject's body. You should also try to make sure each subject is "wearing" their blanket in the same configuration, since the amount of air inside the blanket will affect its insulating properties.

I agree that each subject should test all three blankets, if possible--obviously these results won't be very rigorous however you slice it, but I think that would help a bit. I also think that you should give them nice wool socks to wear, as it's often difficult to get a blanket to cover your feet effectively if you are, say, sitting in a refrigerator.
posted by fermion at 2:39 PM on March 9

Well, I'm no expert, but ...

I'm under the vague impression that humans are warm blooded, and will therefore try to maintain themselves at 98.6 degrees F or so, and so one hopes that the temperature right next to the skin will be fairly constant. Indeed, if the body temperature drops by even a couple of degrees, it appears that bad things may happen.

I'm also under the vague impression that the thermal conductivity of common materials does not vary that much. A great deal of the insulating properties of jackets and the like, I am told, comes from air space -- a great big puffy jacket is warmer because air gets trapped in the pockets and provides resistance to heat transfer, not because it is filled with cotton or whatever. Therefore choice of blanket may have a great effect on your results: if they are all of roughly similar thickness, you may not find much of a difference at all.

I am just worried that your demonstration will have either disappointing results (half nude people in refrigerators aside), or, uhm, unexpectedly exciting results.
posted by Comrade_robot at 2:46 PM on March 9

Get a fridge box or some other containment device. Blanket is put in the middle in some standardized manner. Control variables such as room temp, drafts, and other scientific stuff.

The air on the bottom should be precisely 90F (or whatever, as long as it's consistent). An incandescant lamp warms air nicely.
The air on the top should be precisely 50F (again w/ the consistency)

Heat wants to rise. How fast does the upper chamber get to 65F? The blanket that conserves heat the best, and keeps the upper chamber from increasing temp, is the warmest.

Yay, science! I'm not sure when you'd encounter 25 mph wind in the bedroom, either, but there's another fun experiment!
posted by theora55 at 3:01 PM on March 9

Well after reading what other folks have offered, and your comments, I'm going to suggest that since this is clearly a public "for-fun" demonstration and not something to be published in Nature, why not just dispense with all the measuring and work on making more of a visual demonstration? You can still follow the scientific method while not resorting to complicated devices. Devices do not measure "cozyness". Explain its semi-scientific, try and get three subjects that are approximately the same size and weight, wrap them up as identically as possible, attach temp gauges if you like, and make sure they either fill out a survey of your criteria after or arrange that you can talk to them while the experiment is going on (more fun). A truly scientific experiment the likes of which you described would take hours, if not days. Make it fun and educational by explaining the scientific method and your thoughts on its design, using your experiment as a laymans example, rather than complicate it with a bunch of McGuyver-esque techniques that could not possibly withstand the scrutiny of real science.
posted by elendil71 at 3:19 PM on March 9

Explain its semi-scientific, try and get three subjects that are approximately the same size and weight, wrap them up as identically as possible, attach temp gauges if you like, and make sure they either fill out a survey of your criteria after or arrange that you can talk to them while the experiment is going on

This is exactly what I think we'll be aiming for. We'll be talking with the participants throughout the test and I'll be personally checking each blanket for thickness, body coverage, etc. (before our staged event) to be factored into our final results.

Thanks again for all the great answers! I knew I could count on MeFi.
posted by plasticbugs at 3:33 PM on March 9

You might be interested in EN 13537 which is the European standard used for measuring the warmth of sleeping bags. They do it by measuring not one but 4 temperature ranges - and they use a mannequin.
posted by rongorongo at 4:39 PM on March 9