ETL 1110-2-365
31 Aug 94
investigation. Contours should be relatively smooth
(c) Crack potential contours. This type of plot,
within a lift and should not abruptly change direc-
Figure A-5, is only available through the use of
tions. All contour plots should be checked to ensure
ANACAP-U and is used in conjunction with crack
they reflect symmetry where a symmetrical boundary
location plots. Contours reflect the largest percent of
condition has been applied to the model. Plotting
the cracking criteria reached at a specified time for
software should use results at element integration
all principal strain directions. This type of plot is
points, not nodal results averaged from locations in
useful in assessing the potential for cracking in the
different lifts. Although each lift consists of con-
structure for a given loading and material property
crete, they are in essence different materials, since the
combination. These plots are also useful for the
age of the concrete is different from one lift to the
evaluation of insulation requirements for winter pro-
next. Therefore, averaged results tend to smooth
tection. Once a crack has formed, the largest percent
contours and lessen their magnitudes.
of the cracking criteria for the remaining uncracked
principal strain directions is plotted in subsequent
(a) Stress/strain contours. Contours of stresses
contours for locations along the crack. This typically
and strains in the global coordinate directions along
results in a drastic change of crack potential in sub-
with contours of principal stresses and strains should
sequent contours for locations where a crack has
be used in formulating and supporting conclusions
previously formed. Therefore, crack potential contour
drawn about the response of the structure. Contours
plots must be used in conjunction with crack location
near piles, lift interfaces, and reentrant corners typi-
plots to correctly assess the cracking potential of the
cally lack smoothness due to abrupt changes in stiff-
structure.
ness in nearby materials and stress concentrations.
Such abrupt changes may indicate errors in input or
(2) Time-history plots. Time-history plots of
modeling, or areas where the FE mesh should be
temperature, stress, strain, and principal stress results
refined. Mesh refinement should be considered only
at a single location or multiple points across a section
if contours lack smoothness in critical areas. Due to
of significance are useful in showing the response of
changing orientation, principal stress and strain plots
that location throughout the time of the analysis.
typically are not as smooth as stress and strain plots
These are useful in determining the critical material
in the x, y, or z directions. A contour plot of maxi-
property combination when multiple analyses are
mum principal stress, s11, is shown in Figure A-3.
performed. A time-history plot of horizontal stress is
The ABAQUS computer program outputs and plots
shown in Figure A-6. To assist reviewers and per-
total strain which includes strain due to free expan-
sons unfamiliar with the model, a locator section is
sion, not just the net strain (the portion of the total
provided to show the location in the model where the
strain that produces stress). Strains due to free ther-
results are presented. Selection of locations for pre-
mal expansion do not produce stress or contribute to
sentation of time-history results may be determined
cracking of concrete and should not be included in
from contour plots, from the determination of loca-
strain plots. Therefore, total strains are not useful in
tions of maximum values of results, or from locations
interpreting results, and the strain plotting capabilities
of particular interest. These may be places where
of ABAQUS should not be used. Net strain contour
similar structures have experienced problems, places
plots can be obtained by using the ANACAP-U soft-
where previous NISA's have presented results, or
ware (ANATECH Research Corp 1992).
places which help explain the overall response of the
structure.
(b) Temperature contours. Temperature contours
should be smooth throughout a lift and across lift
(3) Section plots. Plots of results (i.e., stress,
interfaces. Temperature contours should never
temperature, net strain) across a specified section or
abruptly intersect free surfaces of the model where
location at a specific time are useful in determining
film coefficients are applied, except for locations
the behavior of the section or location. Stress results,
where a very low film coefficient is used to model an
as seen in Figure A-7, typically show distortions at
enclosed void. This indicates the application of an
lift interfaces. These distortions arise from the differ-
incorrect thermal boundary condition. A temperature
ent load history and different material properties
contour is shown in Figure A-4.
resulting from different ages of concrete in adjacent
A-23
Previous Page
