[king] frost in concrete STRUCTURES 117 



with the velocity of the air or water in contact with the surface, while 

 the remaining constants will depend on the amount of water contained 

 in the medium. 



Section 3. Determination of the Thermal Constants of 



Porous Media 



For many practical purposes the required constants may be 

 obtained with sufficient accuracy by an extremely simple test. Two 

 identical slabs of the water-soaked medium are prepared about 30 cm. 

 square and from 1 to 3 cm. thick. Each of the slabs is wetted over 

 one face, the wet surface placed in contact and the two exposed to a 

 temperature below freezing. In this way an "ice-sandwich" is made 

 up. When the water contained in the medium has become entirely 

 solidified, the "ice-sandwich" is brought to a uniform temperature 

 slightly below freezing. It is then brought indoors and hung up 

 vertically so that both faces are equally exposed to air at a temperature 

 d, several degrees above freezing (ordinary room temperature is suit- 

 able). Thawing begins at the outer surface, penetrates towards the 

 interior and finally reaches the ice-layer cementing the two slabs 

 which fall apart in a time t given by 



X (h^ + Jx) = (KJLpJ -dt (3) 



where x is the thickness of the slabs. 



If the ice-sandwich is immersed in a tank of water at room 

 temperature 6', kept constantly stirred so that hy, is negligible com- 

 pared to X, the time t' taken for the slabs to come apart is given by 



J x^ = (K JLp^) ~dt" (4) 



the surface conductivity being so good that h^ may be neglected 

 compared to x. From (3) and (4) we are enabled to calculate separ- 

 ately (L,^p^/K^,) and /î^ in terms of easily observed quantities. 



The corresponding constants for a frozen slab may most easily 

 be obtained by placing the thawed out slab (reduced to a temperature 

 slightly above freezing) on the surface of a large sheet of freezing water 

 out of doors, so that one face is exposed to the atmosphere. The 

 slab freezes through and ice finally begins to accumulate on the outer 

 surface: by rapidly weighing the slab at two known intervals from the 

 commencement of the test we ma}^ calculate the thickness of ice 

 formed and from the equations of the form (1) and (2) governing the 

 process we may determine (Lpy/Ky) and hp 



Section 4. Results of Tests 

 Four samples of wood (soft pine) were tested by the above method 

 with the following results. Specimens I. and II. were creosoted pine: 



