50 Mr. A. Malloek .
On a Direct Method of Measuring [ Apr. 19 , " On a Direct Method of Measuring the Coefficient of Volume Elasticity of Metals .
" By A. Mallock , F.R.S. Received April 19 , \#151 ; Read June 2 , 1904 .
For most hard materials the coefficient of volume elasticity is usually calculated from measurements of Young 's modulus or of the coefficient of rigidity , either of which , when Poisson 's ratio is known , suffices for its determination .
Although , however , the total alteration of volume for a given pressure can be calculated from the coefficient thus obtained , it is only for isotropic material that the alteration of dimensions in any given direction can be inferred from.it .
The following direct method of measuring the coefficient of volume elasticity is of some interest , as it allows of the linear contraction or extension in any given direction , caused in a substance by fluid pressure , to be measured independently of other changes of form .
When a long circular cylinder is acted on by internal fluid pressure , if the walls are very thin compared to the diameter of the cylinder , the stress in the material parallel to the axis is just half the stress at right angles to the axis in the tangent plane .
The conditions of strain and stress in the walls of the cylinder can be expressed in terms of the volume elasticity and rigidity of the material as follows:\#151 ; Consider a small cube of the material , with edges parallel to the axis ( X ) , radius ( Y ) , and tangent of circular section ( Z ) of the cylinder respectively .
Let K be that part of the stress which produces alteration of volume , and N** and N^z the parts which produce shear in the planes XZ and YZ .
The total force acting parallel to Y , .
, in the direction of the radius , vanishes in comparison with the forces at right angles to it when the walls of the cylinder are thin , for while the radial force is O at one surface of the cylinder , and equal per unit area to the fluid pressure ( P ) on the other , in the two directions at right angles to the radius the stress is of the order Pr/ t ( = 2P ' ) , where r is the radius and t the thickness of the wall .
Hence we have the following relations between K , N , and P ' : K + Nsz + Nyz = 2P ' ... ... ... ... ... ... ( 1 ) , K-Nasz - P ... ... . .
, ... ... ... . .
( 2 ) , K \#151 ; N^z .= O ... ... ... ... ... ... ... ... ( 3)\#187 ; whence K = P ' ... ... ... ... ... ... ..\#151 ; ( ^)* That is , the alteration in the length of a cylinder caused by fluid pressure depends solely on the coefficient of volume elasticity.* Since K is that part of the applied force which goes to produce alteration of volume , we have , if k be the coefficient of volume * It may be shown that this result is independent of the thickness of the walls of the circular cylindrical tube .
1904.1 the Coefficient Volume Elasticity of Metals .
elasticity ( k = kx+ Ky + Kz ) and ex , ey , the alterations of dimensions caused in the cube by volume expansion in X , Y , Z , In isotropic materials kx = kv = kz and ex = = eZi so that k \#151 ; Pr2/ ( Me .
The ease with which minute variations of length can be measured in the case of rods or tubes allows of a very accurate determination of that component of K which refers to stress parallel to the axis of the tube .
For this purpose it is merely necessary to so mount a suitable length of tube of the material to be experimented on that it can be subjected to strain by internal fluid pressure , and the variation of length caused by that pressure be observed .
I have used this method with steel , copper , and brass tubes in order to see whether annealing altered the value of k. The arrangement employed is shown in fig. ( 1 ) .
Fig. 1 .
The tube AB is closed by caps at each end .
Through the cap at B a small pipe connected with a pressure pump and gauge is introduced .
The cap at A has a small hole in it , which can be ' closed with a plug after the tube has been filled with water .
The tube is embraced at C and I ) by gymbal rings fig. ( 2 ) where the steel points EE ' are closed on the tube , and the pivots FF ' rest at ( D)\#187 ; fig* ( 1 ) on a fixed support , and at C , fig. ( 1 ) on a rocking support G. This rocking support carries a small reflecting prism H , whose upper surface is parallel to the axis of the tube , and passes through the axis FF ' of the gymbal ring .
The movement of a mark on this surface of the prism was measured by the microscope hi , fig. ( 2 ) a |^-inch objective being used .
The only practical difficulty met with was the variation of temperature of the room during the experiments .
From this cause it was generally found that the zero was slowly altering , Fig. 2 .
52 .
Coefficient of Volume Elasticity of Metals .
but the temperature effect can be readily eliminated from the observations .
, , - .
m In the experiments , the results of which are recorded below , the maximum fluid pressure in the tube was 400 lbs. per square inch .
The longitudinal extension was measured at 100 , 200 , 300 , and 400 lbs. pressure , , and the extensions when corrected for temperature and plotted in terms of pressure lay on a straight line within the limits of error .
Table of Results .
Condition .
Length between supports in inches .
Outside , diam. in inches .
Thickness of wall in inches .
Total extension at 400 lbs. per square inch .
9kx in C.GLS .
units . .
K from experiments.* Seamless Steel Tube .
Hard 60 0*75 0-0190 2 *8 x 10~3 18 *4 x 10u 18*41 x 10u Annealed 2-75 " 18-2 " Solid Drawn Brass Tube .
\#166 ; : : * \#166 ; " .
ly- v .
V Hard 50 0*415 9-0185 2 -12 x 10"3 1105 x 10 " * Different samples from 10*85 .
x 1011 to 1002 Annealed 2-09 " 1075 " Solid Drawn Copper Tube .
Hard 50 0*4485 0-0382 5 -62 x 10"4 23 x 10u 16 -84 x 10u Annealed 7*1 " 16-2 " * From Everett 's " C.G-.-S .
System of Units .
" The value of 3k* is given in the table to facilitate comparison with other measures , as 3k* is the value k would have if the material had the .same properties in all directions as it has in that of the axis of the .
tub* .
' j \#166 ; .
r the case of copper is remarkable , for in its hard drawn state it extends less in the direction of the drawing than a similar tube of steel would when equally strained . . . .
;

