The Naked Scientists

The Naked Scientists Forum

Author Topic: Thermometers  (Read 8077 times)

paul.fr

  • Guest
Thermometers
« on: 25/03/2007 22:44:50 »
How accurate are alcohol/spirit filled Thermometers compared to mercury ones? plus, how do "they" (for want of a better word) calibrate them?

is it x amount of fluid in a thermometer of y volume?

hope you understand the question, major struggle to put thought to paper here.


 

Offline Karen W.

  • Moderator
  • Naked Science Forum GOD!
  • *****
  • Posts: 31653
  • Thanked: 5 times
  • "come fly with me"
    • View Profile
Thermometers
« Reply #1 on: 25/03/2007 22:51:05 »
Thats a good question.. Hope to here an answer.
 

another_someone

  • Guest
Thermometers
« Reply #2 on: 26/03/2007 04:54:15 »
I would have thought there are at least three advantages to mercury thermometers (but I do not know how correct or otherwise my opinion might be):

1) mercury, being a metal, has better heat conductivity than alcohol, so will be at a more even temperature along its length.

2) mercury is easier to see (although the addition of dyes to the alcohol will mitigate this, and is probably a necessity for there proper use).

3) mercury is less volatile, and will have a lower vapour pressure in the vacuum above the liquid, thus losing less of the liquid with increasing temperature.

For the fairly imprecise use usually made of these thermometers, these differences are probably not that significant (the poor visibility of the working fluid is probably the greatest everyday problem with alcohol).
 

Offline lightarrow

  • Neilep Level Member
  • ******
  • Posts: 4586
  • Thanked: 7 times
    • View Profile
Thermometers
« Reply #3 on: 26/03/2007 23:01:11 »
How accurate are alcohol/spirit filled Thermometers compared to mercury ones? plus, how do "they" (for want of a better word) calibrate them?
is it x amount of fluid in a thermometer of y volume?
hope you understand the question, major struggle to put thought to paper here.

Exactly I don't know which substance there is inside a laboratory's thermometers for low temperatures I have, but it's certainly not Hg (mercury) because the liquid is red and presumably less viscous than Hg, since it moves faster with temperature changes.
So, since it's for Lab's use, I think it should be quite precise...

However thay say that there could be even ≈1.5C difference in reading between an alcohol and a Hg thermom.

To calibrate a thermometre, you first have to know the use of it and so its temp range. Then you find (at least) two phase transitions with well known temperature, for example ice/water transition and water boiling. You sign "0" to the level reached by the liquid when the therm. is put in a water/ice solution at equilibrium, then you sign "100" at boiling water and then you divide the range in 100 small lines...

This works as long as the volum of liquid inside the bulb is much greter than the volum in the column (so that what counts is only bulb's temp. and not also column's temp.) and as long as ΔV/V is proportional to ΔT.

If the proportionality is not perfect, you have to use more than two fixed points, and linearize among these.
« Last Edit: 26/03/2007 23:04:15 by lightarrow »
 

paul.fr

  • Guest
Thermometers
« Reply #4 on: 26/03/2007 23:56:00 »
To calibrate a thermometre, you first have to know the use of it and so its temp range. Then you find (at least) two phase transitions with well known temperature, for example ice/water transition and water boiling. You sign "0" to the level reached by the liquid when the therm. is put in a water/ice solution at equilibrium, then you sign "100" at boiling water and then you divide the range in 100 small lines...

I never thought of that  [:I] , Thanks Lightarrow
 

Offline lightarrow

  • Neilep Level Member
  • ******
  • Posts: 4586
  • Thanked: 7 times
    • View Profile
Thermometers
« Reply #5 on: 27/03/2007 15:23:30 »
I never thought of that  [:I] , Thanks Lightarrow

You're welcome. If you are going to use this procedure, you have to find an insulated container, put there distilled water and ice (from distilled water) and stir them gently. If the ice melts completely after some minutes, then you add ice, until there still is ice and water after stirring. At this moment the water temperature is exactly 0C.

The boiling point is achieved after having made the (distilled) water boil (at 1 atm) at low flame for some minutes, to remove all air and CO2 bubbles. You have to put your bulb's thermometer just over the water's surface and then sign 100C.
 

Online Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8663
  • Thanked: 42 times
    • View Profile
Thermometers
« Reply #6 on: 27/03/2007 20:17:53 »
Mercury isn't used at low temperatures, not least because it freezes at about -39C. Alcohol (coloured with dye) is used for most but, for really low temperatures pentane (melting at -130C) gets used. For some aplications gallium is used because it has a very long liquid range, it melts at about 30C but doesn't boil until you reach about 2400C.
You can use mercury for temperatures up to about 500C even though this is well above the normal boiling point- the trick is to fill the top of the thermometer with nitrogen, the gas pressure raises the boiling point.

If you took 2 thermometers , one with alcohol and one with mercury and marked both of them at the ice point and the steam point then divided the scales into 100 they would aggree at 0C and also at 100 C. Unfortunately, they wouldn't quite agree anywhere in between because the expansion of the 2 liquids isn't perfectly linear.
To get the best precision you need to find something more linear to fill in the points in between. One of the common ways is to use the electrical resistance of a piece of platinum wire but (assuming they havent changed things since I last checked) the ultimate standard is the constant volume gas thermometer using helium, and extrapolated to zero pressure. Unfortunately, that equipment is a total pain in the neck toset up. So what people do is use that to measure the temperatures of some phase transitions like the freezing point of gallium or the boiling point of sulphur. On a day to day basis these phase changes (the so called fixed points) are much more simple to set up  and you can check any other sort of thermometer against them.
The thermometers we usually use in the lab read to 0.5C, the special calibrated ones read to 0.1 By doing some clever tricks you can get a mercury thermometer to read temperature changes to 0.001C (It's called a Beckmann thermometer if anyone's interested).
My fairly expensive digital (Platinum resistance) thermometer reads to 0.01C but it's only accurate to abou 0.2. I really need to recalibrate it.
 

The Naked Scientists Forum

Thermometers
« Reply #6 on: 27/03/2007 20:17:53 »

 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums