April 14, 1999
MSC.Fatigue 8.0 Vibration
Fatigue Analyzer - FEVIB Update
Description
This FEVIB executable fixes the following problem(s):
1. Inability of FEVIB to recognize
user defined material S-N curves. Inability of the Design Optimization
option to list material S-N curves or recognize them from the database if specified.
(CSRs 34562, 34563)
2. +INF values written to output
FEF file if stress range is greater than that of the damage look up curve (S-N
curve). Zero life is now reported. (CSR 35454)
3. Global Analysis stationarity
check default is now set OFF. A new batch keyword exists for running in
batch: /SCO=Y (stationarity check ON, default is OFF). This
tends to speed up the global analysis by up to a factor of two. However
you will not be able to plot stationarity plots. You must rerun the analysis
interactively to do this and create an output file to plot in either MSC.Patran
or MSC.Fatigue Pre&Post or in FEVIB's PSD output mode.
4. Elemental based calculations
not working correctly in Design Optimization. Worst case element now recognized.
(CSR 35698)
5. Can't plot cycles histogram with
Dirlik method in FEVIBs Design Optimization mode. (CSR 35699)
6. Other minor changes to Design
Optimization mode:
- Nonsensical combinations of Parameter
Optimization/Sensitivity Analysis/Change Parameters options disallowed with
appropriate messages given if attempted
- Default value for Duration picked
up from environment (envi.usr file)
- Default Duration units set to
Seconds
- If user sets Duration in other
units, e.g., minutes, it is converted to seconds for subsequent analyses -
Duration only affects the z-axis of a cycles histogram plot
- If keyword $FLFUTSC is found
in environment, results form shows probability of exceeding UTS
Note that this new version of FEVIB
gives slightly different answers than the previously released version. This
is because of two reason:
1. The Kowalewski method has
been turned OFF by default. It used to be ON automatically but gives rather
poorer results. This is mainly used with mean stress. Now when you
plot a cycles histogram you will only get a single strip of cycles. Kowalewski
gives a more distributed hump rather than a single strip. You can still
use the Kowalewski method by setting the environment variable $FLFKOWA = ON
using MENM.
2. A frequency cut-off is now
implemented the same way as in MFLF. The answers should be very close
to those you would get using the single shot vibration fatigue analyzer MFLF.
Installation
1. Copy the correct compressed tar or zip file to your "<P3_HOME>" directory
depending on the platform of interest.
(i.e. --> c:\msc\patran80)
2. For UNIX, gunzip the file and
untar it. For example:
gunzip fevib_hpux.tar.gz
tar xvf fevib_hpux.tar
3. For NT, unzip the file into the
installation directory.
File Set
*****************************************************************************
mscfatigue_files/bin/fevib
mscfatigue_files/help/fevib.hqr
|
