-23
-
APPENDIX E - MEASUREMENT OF, AND CORRECTION FOR, TEMPERATURE
EFFECTS
If the ends of the structural member were free to expand or contract without restraint then
strain changes would take place without any change in stress. And in these situations the
strain gage would indeed show no change in reading. Conversely, if the ends of a steel
structural member were restrained by some semi-rigid medium, then any increase in
temperature of the structural member would result in a build-up of compressive strain in the
member, even though the actual strain would be tensile!
The magnitude of this temperature-induced compressive stress increase would be
measured accurately by the strain gage, because, while the member is restrained from
expansion, the vibrating wire is not restrained and the expansion of the wire would cause a
reduction in wire tension and a resulting decrease in the vibration frequency. This would be
indicated by a decrease in strain reading on the readout box, corresponding to an apparent
increase in compressive stress, which is, mirabile dictu, exactly equal to the temperature-
induced increase in compressive stress in the member.)
The temperature-induced stresses can be separated from the load-induced stresses by
reading both the strain and temperature of the strain gages at frequent intervals over a
period of time in which the external loading from construction activity can be assumed to be
constant. When these strain changes are plotted against the corresponding temperature
changes, the resulting graph shows a straight-line relationship the slope of which yields a
factor K
T
. This factor can be used to calculate the temperature-induced stress
σ
thermal
= K
T
(T
1
-T
0
)E……………………………………….E1
Which if desired can be subtracted from the observed apparent stress change
σ
apparent
= (R
1
-R
0
)BE………………………………………….E2
to give that part of the stress change due to construction activity loads only
σ
load
= [(R
1
-R
0
)B- K
T
(T
1
-T
0
)]E……………………………….E3
Note that the correction factor, K
T
, may change with time and with construction activity due
to the fact that the rigidity of the restraint may change. It would then be a good idea to
repeat the above procedure in order to calculate a new temperature correction factor.
If, for whatever reason, the actual strain of the steel member is required, that is, the change
of unit length that would be measured by, say, a dial gage attached to the surface, this is
given by the equation
μ
ε
actual
=(R
1
-R
0
)B + (T
1
-T
0
)x CF
1
……………………..…………………………..
E4
Where CF1 represents the coefficient of expansion of steel = +12.2 microstrain/
C.
