Ask A Scientist General Science Archive

name         P. B.
status       educator
age          20s

Question -   Water freezing question: If you placed water in a U 
shaped container made of glass pipe and filled it full of water but left 
the top of the U open to atmosphere see below diagram. Then placed the 
unit in a very cold location so it would freeze. Would the horizontal 
part of the U glass tube break because of the freezing of the ice. It has 
been said that it  would not break because the ice would relive its 
pressure via the open part of the U. I think it would break the 
horizontal glass pipe because ice creates a force in all directions and 
the horizontal piece would break. Can you help?
|    |                          |    |
|    |                          |    |
|    |       +                 |    |
|    |_______________|    |
|________________ ____|
              +
------------------------------------------------------------------------
P.B.,

First of all, this URL from our own site may help you to reinforce why ice 
expands when it freezes:

http://newton.dep.anl.gov/askasci/chem99/chem99245.htm

Second of all, I am also of the opinion that none of this glass apparatus 
would break.  I say this for the same reason you stated.  If the entire 
apparatus was frozen all uniformly with time (all water reached 32 °F at 
the same) the water would simply start to build its crystal lattice.  If 
this was to be conducted as a practical experiment, this glass tube would 
be found in the freezer, unharmed, and with a slightly higher ice crystal 
height at the top of your glass tube.  More specifically, I believe that 
the shape of the top of the tube (frozen water portion) would be conical 
and pointing upward.  WHY?

When you place this 75 °F water inside its 75 °F glass tube inside the 
freezer.  A temperature gradient is present.  All this means is that you 
have 75 °F water inside a freezer that is more than likely being 
maintained at -20 F°.  As the water cools it will cool with a temperature 
profile that looks something like this (temperature on Y-axis and tube 
diameter on X-axis)


  |       ..             |  --- 75.0 °F
  |      .  .    |  --- 74.9 °F
  |     .    .   |  --- 74.8 °F
  |    .      .  |  --- 74.7 °F
  |   .        . |  --- 74.6 °F
  | ..          ..       |  --- 74.5 °F
  |                      |

** THESE OF COURSE ARE VERY ROUGH FIGURES AND IS ONLY MEANT TO GENERICALLY /
QUALITATIVELY DESCRIBE THE TEMPERATURE PROFILE ALONG THE AXIS OF THE GLASS
TUBE **

This poor mans plot represents one slice, or cross-section of the glass 
tube and also at one slice or point in time while the glass - water setup 
starts to drop in temperature.  Now, provided that the apparatus is not 
agitated at all this temperature profile is correct.

** THIS IS MY MAIN ARGUMENT THAT, IN FACT, THE GLASS WILL NOT BREAK BECAUSE
THE WATER (LIQUID) AT THE INSIDE EDGES OF THE GLASS TUBE WILL FREEZE FIRST.
THIS WILL CAUSE THE UNFROZEN WATER IN THE MIDDLE TO BE PUSHED UP TO MAKE
ROOM FOR THE LESS DENSE WATER (SOLID) TO BE FORMED.

*** SIDE NOTE:  GENERICALLY SPEAKING, THE DENSITY OF H(2)O INCREASES WITH
DECREASING TEMPERATURE.  HOWEVER, FEW PEOPLE KNOW, OR EVEN CARE TO KNOW THAT
WATER BEGINS TO BECOME LESS DENSE AT AROUND 3.98 °C.  PLEASE TAKE A LOOK AT
THIS URL FOR A PICTORAL:

http://ess.geology.ufl.edu/ess/Notes/Water/dens_temp.jpeg

DARIN WAGNER
=====================================================
Very interesting question. I am not a structural engineer, so I do not have
an authoritative answer, but we can consider some thought experiments.
Assumptions: ice is less dense than water, there is no air in the water,
there are no "seeds" to initiate ice formation except at the water / air
interface (i.e. no "plugs" form in the base of the "U" tube.
     Very strong glass tubing, or elastic tubing. You are correct that the
force produced by freezing acts in all directions; however, movement or
growth of the ice will occur along the path of least resistance -
so in an infinitely strong "U" tube ice would form at the surface of the
arms and the level would rise in the
"U" tube. If this motion is not allowed, tremendous forces would develop as
the water freezes down the
arms of the "U" tube since water is essentially incompressible. If the tube
is elastic, the diameter of the tube would just bulge to relieve any
increased pressure.
An interesting variation would be to wrap half the "U" tube with a heating
tape to keep the left hand arm of the "U" tube from freezing. Then as the
ice forms in the right arm of the "U" tube the expansion of the ice would
push the left hand arm up. The distance multiplied by the cross section
would be the difference in the density of the ice and water in the left arm
of the "U" tube.

Vince Calder
====================================================
P. B. -

This sounds like an excellent experiment to try, especially if you have a
couple of students to do it with you.

My guess is that the answer would be a function of the length of the
horizontal portion of the tube and the diameter of the tube. This would be a
testable hypothesis.

Larry Krengel
=====================================================