ETL 1110-2-365
31 Aug 94
and time is contained in ABAQUS through the
dimension per unit length per unit of temperature
UMAT subroutine and must be calibrated for each
change. The coefficient of thermal expansion is
project's selected mixture.
determined according to CRD-C 39 (USAEWES
1949h). The value of this property is strongly influ-
(d) Autogenous shrinkage. Autogenous shrink-
enced by the type and quantity of coarse aggregate in
age is a decrease in volume of a concrete specimen
the mixture. The coefficient of thermal expansion for
or member due to hydration of the cementitious mate-
mineral aggregates varies from less than 2 to over
8 10-6 in./in./F, while the coefficient of thermal
rials without the concrete gaining or loosing moisture.
expansion for cement paste may vary from 6 to
This shrinkage is also referred to as "sealed length
12 10-6 in./in./F. Typical values for the coefficient
change." This type of volume change occurs in the
of thermal expansion for mass concrete range from 4
interior of a large mass of concrete. For small
to 6 10-6 in./in./F.
volumes of concrete, such as structural concrete
members, the magnitude of autogenous shrinkage is
negligible compared to drying shrinkage and thus is
(f) Tensile capacity. The cracking resistance of
usually not distinguished from drying shrinkage.
concrete is determined by a combination of limiting
However, for large mass concrete structures, auto-
tensile strain and tensile stress. These properties are
genous shrinkage can be a significant factor. Auto-
time and rate of loading dependent. They are deter-
genous shrinkage occurs over a much longer time
mined according to CRD-C 71 (USAEWES 1949i).
than drying shrinkage, the localized phenomenon that
Results from the slow load tests as defined in the test
affects only a thin layer of concrete near the surface.
method are required to define the failure envelope for
Autogenous shrinkage tends to increase at higher
the concrete cracking criteria. Refer to para-
temperatures, with higher cement contents, and with
graph A-6d, Cracking criteria, and Annex 2 of
finer cements. This property is modeled as a function
Appendix A for use of the test results. Typical
of time in the analyses. Typical values for mass
values for tensile strain capacity of mass concrete for
concrete vary significantly depending on the test
a slow load test for a specimen loaded at 7 days and
procedure utilized. Melvin Price Locks and Dam
failing at between 75 and 150 days range from 75 to
values were about 300, 350, and 400 millionths at
150 microns.
10, 25, and 50 days, respectively. Olmsted Lock
values were about 20, 30, and 45 millionths at 10, 25,
(3) Physical property. Density is defined as
and 50 days, respectively. A curve of shrinkage
mass per unit volume. It is determined according to
versus time will be input into ABAQUS through the
CRD-C 23 (USAEWES 1949j). Typical values of
UMAT subroutine. No standard test method exists
density for mass concrete range from 140 to
160 lb/ft3.
for determining the autogenous shrinkage of a con-
crete mixture. However, recent experience at WES
for the Olmsted Locks Project has shown that prop-
c. Foundation properties. The thermal and
erly prepared and instrumented sealed creep cylinder
physical properties of the foundation are dependent
specimens with no load applied can be used to mea-
on the type of soil or rock, the moisture content, the
sure autogenous shrinkage. Method CRD-C 54,
presence of piles, and any discontinuities in the foun-
(USAEWES 1949g) describes the preparation of
dation. In situ properties may vary significantly from
creep test specimens. Autogenous shrinkage speci-
those obtained from laboratory testing of small sam-
mens must by completely wrapped in a moisture
ples obtained from borings or test pits. Exact thermal
retentive membrane to minimize loss of water to the
properties are not necessary for the foundation mate-
surroundings. This can be verified by periodic deter-
rials and adequate values for use in a NISA may be
minations of the mass of the autogenous shrinkage
obtained from Jumikis (1977) or Kersten (1949).
specimens to determine if the specimen mass is
Likewise, exact mechanical properties are not
remaining constant or varying with time. WES Tech-
required, and adequate values can be estimated from
nical Report SL-91-9 (Hammons et al. 1991) docu-
foundation test data or from Hunt (1986). The struc-
ments the test procedure and results obtained for the
tural and geotechnical engineers should jointly select
Olmsted Locks project.
foundation properties for ABAQUS input based on
any in situ properties available and varied based on
(e) Coefficient of thermal expansion. The coef-
information from the above referenced texts and past
ficient of thermal expansion is the change in linear
experience.
A-13
Previous Page
