|
Integration of Metallurgical
Model into Mill Level 2
Integration with
metallurgical model can
greatly improve the
quality of the Level 2
model. Traditional Level
2 model was designed
without metallurgical
concept in mind. Such a
design, however, is
facing increasing
challenge due to the
application of the new
rolling technologies
such as the low
temperature rolling, hold
between rolling
passes including the two-piece
rolling
and the use of microalloy for the
steel. In the low
temperature rolling the recrystallization
often cannot be
completed and this
causes a retained strain
from pass to pass.
A slight modification of
the Level 2 model with
metallurgical principle would
create great benefit. In
many times, the
modification is only
necessary for the input
data files or inside a
single function in a
single file. The resume
pass after the hold may
have over 40% force
error, as in the case of
a client. Modification
for such a system is
necessary. Quite often,
in the finish pass with
the draft 10% or lower,
some formula is actually
not valid. This cause
force error, and thus a
wrong draft schedule
would be created and bad
shape of the steel is
produced.
What to Integrate
Metal Data, with its
rich experiences in
integrating
metallurgical models
into mill Level 2
systems, has identified
various issues in the
present Level 2 models
and a long list of
metallurgical parameters
to be integrated into
current Level 2 systems.
Contact us via
bli68@qq.com
for a list of
metallurgical parameters
and related models
specific to your Level 2
model and mill practice.
How to Integrate
Metallurgical models to
be integrated into the
existing Level 2 system
don't have to be a very
complicated one.
Integration in various
levels of complicity
could be performed.
Special attention should
be paid to protect the
existing Level 2 system
from any possible
damage.
- Modification of existing
Level 2 model, preferably, the modification of only a
function or only a file.
- Attaching a metallurgical
model package, e.g., the one to predict the retained strain,
into an existing Level 2 system. If the metallurgical model
to be added into an existing system is pretty complicated,
it is preferred to create an extra software package to be
attached into the existing Level 2 system.
- Modification of an input
file or an input database, without the change of actual
source code.
Existing Level 2 model can
be improved with the
metallurgical effects being
integrated into Level 2
parameters. For example, the
effect of the retained
strain may be integrated
into the strain coefficient
C3. In a consulting project for a steel plate
reversing mill, 6000 sets of new coefficients for the flow stress were designed by integrating retained stain into
those coefficients,
and the flow stress for
the resume pass was
modified. In this way, the
metallurgical effects
were added into the
existing Level 2 system
with minimal change of
source code, because
most changes were for
the data in the
database. Due to the
inclusion of the
metallurgical effects,
much higher prediction
accuracy was achieved
with the force error
below 5% in most
passes for the
previously troubled grades
and sizes.
Model Development
Metallurgical models are usually very complicated and are
very cost-intensive to develop. Currently available
metallurgical models, mainly off-line ones, don't have
sufficient prediction accuracy and thus are not widely
accepted. Microstructure model development from Metal Data
integrates our metallurgical model into the steel mill online
systems (Level 2 systems) to utilize accurate rolling process
parameters. In addition, we have advanced learning
technologies, so in all the trials, the neural
network learning will be used, and the expert system and fuzzy
logic rules will be applied to specify the value boundaries.
More detailed, such a system consists of followings:
- An expert system, which consists of logics, data and
influence factors on the mechanical, thermal and
metallurgical parameters depending on rolling and thermal
processes. Past mill-experiences can be programmed as a
portion of the expert system.
- A hybrid intelligent learning system, based on rolling
process models, in which the neural network provides the
correcting factors for the model coefficients, and fuzzy
logic rules and expert system provide guidelines (upper and
lower boundaries, etc.) for the learning.
Related Resources
Metal
Data
Work List on Level 2 and Mill Modeling
Metal
Data Recent
Publications
Technical papers published
in the February and March of
2008. Those
publications are primarily
on Level 2, Level 2 model
and process automation.
Metal Data has dozens of
research reports and some
software applications
available at no charge for the consulting clients (only for the projects in the
name of Metal Data and approved by
Dr. Benjamin Li).
|
