Yin R. Metallurgical Process Engineering
Подождите немного. Документ загружается.
138 Metallurgical Process Engineering
level structure controls and guides the development, even elimination or genera-
tion
of
the lower level structure.
2. Scale-level coupling and the division
of
steel manufacturing process. The
process system can be divided and recombined by means
of
separated procedures
and procedures' interrelation. As applied with conception
of
directed graph, the
principles
of
division and recombination should be determined:
(1) The process system is divided into some procedures, each can realize its
major function which have being in one procedure, but in some case, to share
properly by other procedure is not rejected.
(2) Subgraph
of
divided procedures must obey the relation
of
time order
of
pro-
cess system.
(3) Connection
of
some functions has arisen among a certain divided proce-
dures.
If
some functions are contradictory each other, the functions would be sepa-
rated into several adjustable parts, and assigned to proper procedures, the new pro-
cedure's division can be formed, it meets the non-antagonistic offunctions.
(4) If the probability
of
some function relations between divided procedures is
very little, these relations would be removed, and procedures would be divided
again, new relation between procedures could not form as closed circuit between
the adjacent procedures, that means the relation between two procedures not be-
ing circuit.
(5) The relation among inner systems in divided procedure and its implementa-
tion must be meet constraints and objectives
of
the divided procedure, i.e. the
divided procedures take their own constraints and objectives, and can be realize
with optimization.
So that, the multi-level structure
of
process system and the optimized divi-
sion
of
procedure in process system can be built up. The process
of
steel
manufacturing system also is expressed by directed graph as
D
={F,r},
where
F is all variables for set
of
functions
of
process system, including the common
function variables and also the special function variables in special process;
but
r is directed boundary
of
relations between functions set variables. The r
can express the time-sequence
of
relations, and also the acting direction
of
relations. The multi-level structure
of
the steel manufacturing process system
can be shown in Fig. 5.7 by using
of
the above expression on multi-level
structure, realizing method and applied principle. Where the lines and its di-
rections are relations among nodes and acting directions respectively, the
black points are nodes.
The first level structure shows function variables for child nodes
of
each sub-
system in each procedure
of
process system, and being divided level-by-level
constructs the
(k-l)
level structure by the above five principles
(l
)
~(
5). The
(k-l)
level structure contains seven procedures, which is subsystems
of
mini-
mum
strong connection. Where,
z,
(
k-I)
, z /
k-I)
, Z3(
k-l)
, Z4(
k-I),
z S(k-
ll
, Z6(
k-l)
and
Chapter
5 Multi-Factor Mass Flow Control for Metallurgical Manufacturing Process 139
Z7(
k-l)
represent the procedures
of
raw material preparation, blast furnace
ironmaking, BOF steelmaking, secondary refining, continuous casting, hot
rolling and deep finishing respectively. The
k'th level structure consists
of
three subsystems
of
minimum strong connection, where Z I(k-l
l,
Z2
(k-l)
and Z3(
k-l)
are subsystem for raw material supply and preparation, chemical metallurgy
including solidification, as well as metallurgical physical process respectively.
Finally, only the system
ZI (k
+l)
is on (k+ I) level
of
structure, that is the steel
manufacturing process system.
~
~
---------------
·0
~
·
..
.~.
zl
zl
~
I
I
I
o
o
o
I
I
~.~.~--~
-
-----;._.~
• •
z
Jk
-l)
zl
-l)
(k-l
).......... (k
-l)
(k
-l)
(k
-l)
Zz Z3 Z4 z
51
. .
--------
--
Z (k-l )
I
z[k)····································zjk)············ z? )
~
Fig.5.7
Multi-level structure in the steel manufacturing process system
Being understanding
of
the relations
of
level-scale in process as well as
optimizing division and recombination
of
procedures, it is very important
to
dynamic-order
running
of
process by further research on the scale-level
coupling in process. Some issues on formation and modalities
of
self-
organization in process system,
such
as self-reproduce,
self-growth
and so
on, have been discussed in 5.1.2C The time structure
of
self-organization,
e.g. periodic time
of
processing
is the simplest case
of
self-reproduce in
process system.
Evolution
of
process system shows order state in self-
organization
of
process running.
According
to the analysis on
variation
with time, the appeared
graph
is same as the original graph
after
a period.
This pattern
can
be called as self-reproduce.
Researching
on process sys-
tem
with
several
components
(procedures/devices), the state
of
process
system is unchanged on the process level, i.e. is kept initial
operating
state;
but
the state
of
component
is changed on the lower level.
Each
component
changes periodically, i.e. the process state is on self-reproduce (Fig. 5.8).
The stable state
of
self-organization in process system is
just
due to the
self-reproduce
of
components.
140 Metallurgical Process Engineering
J(EAF)
+-
55-
+-
55
-11-
55-
+-
55
-11-
55-
+-
55
-11-
55-
+-
55
Ilf
fr
ansport)
ITT(
LF)
f--f--f---f--f---f--
I I I I I I 1 1
60
-+-60
-+-60
-+-60
-+-6
0-+-60
-+-60
-+-60
-+
1111111
V(Turret)
IV(Transport)
I--
- -
--++
-
-I+---+t
-
-I+---+t
-
-t+-
---t+
-
-++-
- - -
VT(e
e)
Time/min
Fig. 5.8 Self-reproduce for main procedure running in steel manufacturing process
(Electric furnace steelmaking)
Selfgrowth describes evolution with time for self-organization
of
process sys-
tem on level
of
whole process. The process keeps unchanged structure and func-
tions in self-growth, and then the input materials, energy and information will be
connected throughout and be constantly converted into products, by-products and
any emission by means
of
selfreproduce and interaction
of
components. Thus it
can be seen that the openings
of
process, incessant receiving set
of
inputs, self-
reproduce
of
component and nonlinear relations
of
components ' interactions are
the powers
of
selfgrowth in process system.
Therefore, the evolution
of
multi-factor mass flow can be regarded as the dif-
ferent level-scale coupling in the steel manufacturing process, that is:
• Multi-factor mass flow respectively realizes converting-coupling
of
many
factors on level (or scale)
of
each procedure/device along time axis;
• Multi-factor mass flow does self-reproduce on functions
of
these proce-
dures/devices, and then operates periodically along time axis, and realizes coordi-
nating-coupling
of
many factors on level (or scale)
of
components;
• Multi-factor mass flow is connected with time and space by self-reproduce
and interaction
of
procedures/devices, and then forms the flows
of
product, by-
product and emission, namely it realizes self-growth on process level. This is
evolving-coupling on level-scale
of
whole process (Fig. 5.9).
5.3.2 lnformationflow and model building
The concept and essentials
of
manufacturing process have been discussed on
physical view and emphasized that the basic essentials in their operation are as
process network, flow and program, also these meanings are described respec-
tively in chapter 5.1.1. The flow is the material subject in the process, the process
Chapter 5 Multi-Factor Mass Flow Control for Metallurgical Manufacturing Process 141
network includes node and connector, and is carrier
of
flow running and its time-
space boundary (framework), also the program is set
of
orders, rules, strategies
and paths, and mostly is reflected by information
of
characteristics in mass flow.
Process running
I
I
I
I
I
Procedure operation I I I I I i
I I I I I I
40 i
i::
Self-reproduce of
/~
,
I I I I procedure matter now of
~
-
-
-
-~
-
-
--@P)-
-
-
-~
-~
-~
--<@>
-~
~awrr
'
//
~
Processing time
Fig.5.9
Self-reproduce and self-growth (evolving-coupling on level or scale)
in steel manufacturin g process (BF-BOF route)
O
-Raw
material yard; I - Sinter(pellet); I'
-Co
ke oven;II- Blast furnace;
III
-Hot
metal pretreatment; IV- BOF; V- Secondary metallurgy;
VI
-C
ontinuous casting;
VIl
-R
eheating furnace;
VJll
-H
ot-rolling mill
About definition
of
information, N.Wiener (1948) said: "information is information,
not matter or energy". But in modem science it was proved that information funda-
mentally is a property
of
matter, and can not exist individually apart from the matter.
No any separated information aside from matter existed in objective world or to our
mind. Information is closely to matter and information is obtained, transferred, treated
and utilized without energy consumption. Any information must be carried, repre-
sented and fixed by a certain material form, and the material form is information car-
rier. Certainly, it can be known that the information is a very special property
of
matter,
and can not be recognized completely by natural science until now. Functions
of
in-
formation characterize the matter's composition, structure, state, nature, behavior,
performance, evolution trend and so on (Xu, 2000d) .
Therefore, if multi-factor mass flow does not run smoothly (includes the related
process network unfavorable), information consequently confuses, and it is difficult
to make the information flow to collect, to treat and to utilize easily. If model
of
information flow is not stable and dynamic-orderly, information technology could
not be applied in complex manufacturing process to effectively control the system.
Even information technology is constrainedly applied in some partial processes and
regions, it always gets halfthe result with twice the effort or is titular.
On the whole, in order to make an effective and valuable information flow in
manufacturing process, the operation behavior and the characteristics
of
flow
142 Metallurgical Process Engineering
must be understood firstly, and then the reliable process network (static structure
of
enterprise) should be re-constructed . Establishing physical model
of
the proc-
ess operation, the effective mathematical model can be formulated by information
technology with effective and valuable information flow, and then the dynamic-
order process can be controlled effectively. Fig. 5.10 shows the relationship be-
tween physical-mathematical model and integrated model.
Optimizedcontrol
~
[ralc
gic
s
Integrated model
Mood ofacti
\"
iry andparameter
--------------------------------_
!
~
~
g
~
~
~
-
Mathematical model
Proced
uredescription
Fig. 5.10 Relationship between physical-mathematical model and integrated model
The physical model
of
process system means the entire physical statement
based on understanding
of
operational essence
of
multi-factor mass flow and ra-
tional process network. The statement can be made with conceptual words, and
also made by formulation. Here, modeling can be adopted mainly with formulated
dynamic description.
The dynamic theory can provide the model to describe dynamic behavior
and evolution regularity, and provide the language for system analysis. There-
fore, it is suitable for describing integrity, stability, adaptability and evolution
in complex manufacturing process system. The characteristics
of
modeling are
as follows:
1. Research on the lower levels (some subsystems as procedure, device, work-
shop) in process system, relating a great number
of
its components;
Chapter
5 Multi-Factor Mass Flow Control for Metallurgical Manufacturing Process 143
2. Analysis and study on the partial connection among components (search for
the rules
of
partial connection);
3. Describing the relation between entire and partial process system qualita-
tively by attractor (general objective
of
process system);
4. Embodying the emergence
of
entire structure and function
of
manufacturing
process with interaction among components (subsystems as procedure, device,
workshop).
The steps
of
physical modeling
of
steel manufacturing process system are as
follows :
I. Dividing optimally the process system to form set
of
procedures;
2. Defining set
of
inputs, set
of
results, set
of
control and set
of
procedur
es'
re-
lations;
3. Constructing compact set and its influencing matrix by variables
of
set
of
functions and set
of
inputs;
4. Optimum select variables and values
of
compact set
of
procedures (include
cost matrix);
5. Combined the optimal compact set and set
of
procedures' relations to obtain,
the whole optimization on process system;
6. Rescinding the circuit between procedures during feedback arising in process,
namely circuit existence.
The schematic diagram
of
control in steel manufacturing process is shown in
Fig. 5.11.
-
- -.
~
I I I I
L L , L J
I
I
C
I~S
I
Fig. 5.11 Schematic diagram
of
control in steel manufacturing process
X- Set
of
procedures, the collection of procedures with different functions constructing the process; F-
set of functions, the collection of procedure functions realized under some constraints; R- Set
of
procedures'
relations, the collection
of
relations between adjacent procedures or among interacted procedures; I
-S
et
of
inputs, the collection of any inputting resources (incl. material, energy, information) into process or proce-
dure; C
-S
et of control, the collection
of
control strategies under a certain rule and constraint condition to
realize the procedure functions or optimal procedures' relations;
o-S
et of outputs, the collection
of
conver-
sion situations and performances realized by set of functions in someone procedure or even entire process
under control set C
The combination
of
process engineering with information technology will have
more fit description
of
the physical model
of
macro-scale process system, and
then improves the effectiveness and reliability
of
integrity control.
144 Metallurgical Process Engineering
The mathematical model
of
process can be established on the base
of
physical
model by mathematical analysis, its analysis method may be mechanism analysis
or measured data analysis in general.
The integrated model
of
process can be established by framework language
characterizing organization and behavior or by method
of
strategy, and made on
the base
of
physical and mathematical model under certain rules satisfying the
integrity, concept and reference
of
system.
5.4 Case Study
of
Multi-factor Mass Flow Control in Steel
Manufacturing Process
The essential target
of
multi-factor mass flow operation in steel manufacturing
process is minimum process dissipation. The minimum process dissipation means
few or none mass flow decay, smallest energy loss, least emission, briefprocess
time, shortest space route etc. Therefore, the evolution
of
steel manufacturing
process is trend to:
I. Continuation or quasi-continuation;
2. Integration(on time and space);
3. Dynamic-order and controllable.
Research or analysis on multi-factor mass flow in steel manufacturing process
could be made by the following steps.
1. The undulation in real running for procedures or devices in process system is
investigated, determined and analyzed. For example, in electric arc furnace
(EAF)-thin slab casting-rolling process the distribution graph
of
following pa-
rameters can be determined respectively.
• Distribution
of
tap-to-tap time
ofEAF
(Fig. 5.12);
4
oo
r-
- - - - - - - - - - - - - - - - - - - - - - - - - -
--,
300
~
]
'-
o 200
]
E
:::l
Z
100
329
Standard devialion=8.04
Mean valuc=62.8
Samples=1024.00
80.0 85.0 90.0 95.0 100.0
Tap-to-tap lime of EAF/min
Fig.5.12 Distribution of tap-to-tap time of EAF
Chapter
5 Multi-Factor Mass Flow Control for Metallurgical Manufacturing Process 145
• Distribution
of
tapping temperature in EAF (Fig. 5.13);
300
Standard dcvialion=9.72
Mean
value
~
163 7
.3
Samples»1068.00
78
248
76
74
7 68
'-~~
76
Standard deviation=28.07
70
[
56"
'2&f
I
:~,
Mean value=55.2
..,.:;.
Samples=1087.00
52 52
I \!'
49
~
49
46
1\
1*,
~
rr
e
.
ie'
42
39
l
~:,
,~~
11
I
~
I',"
"5
26
~,
I
~
\
,
- 22
r
I:F
I¥i
I
i'
.
G£
91
0
..
·';
h'
I
:
~
,.,1".~
I
~
;:;
I
~
'
.
tm"TI
~
,..,."
I{
,.,;;
1r
.:;\[ ':P:f .'"
~
200
~
..c
'-
o
]
E
:>
z 100
1645.0 1655.0 1665.0 1675.0 1685.0
1640.0 1650.0 1660.0 1670.0 1680.0 1690.0
Tapping temperature in EAF/"C
Fig.5
.13 Distribution
of
tapping temperature in EAF
• Distribution
of
time-interval from EAF tapping to LF arrival (Fig. 5.14);
80
60
o
5.0 15.0 25.0 35.0 45.0 55.0 65.0 75.0 85.0 95.0 105.0 115.0 125.0 135.0 145.0
Time-interval from EAF tapping to I.F/min
Fig.5.14 Distribution
of
time-interval from EAF tapping to LF arrival
20
~
e
..c
'-
o
] 40
E
:>
z
• Distribution
of
temperature drop value from EAF tapping to LF arrival (Fig.
5.15);
• Distribution
of
refining time in LF (Fig. 5.16);
• Distribution
of
temperature variation in LF refining (Fig. 5.17);
• Distribution
of
casting time
of
one heat
of
steel on thin slab caster (Fig. 5.18).
• Distribution
of
slab surface temperature change from entering to exiting tun-
nel reheating furnace (Fig. 5.19);
• Distribution
of
time from slab exiting tunnel furnace to biting by rolling mill
(always about
8
~
2
0
minutes);
• Distribution
of
time from slab entering stand F1 to leaving stand F6
of
rolling
mill (Fig. 5.20);
146 Metallurgical Process Engineering
200.---- - - - - - - - - - - - - - - - - - - - - - - - - - -
-...,
:!l
~
.r::
'-
o 100
]
E
"
Z
167
Standard
dc
vi
a
l i
on
=2~
.1
0
Mean valuc
-7
8.5
Samplcs=I066.00
o
~.BI
•••
~
~
O
.
o
50.0 60.0 70,0 80,0 90,0 100,0 110,0 120,0 130,0
I
~O
.O
150.0
Temp. drop
of
lOAF
to LF/"C
Fig.5.15
Distribution
of
temperature drop value from EAF tapping to LF arrival
300r-----------------------------,
200
j
'-
o
l!
E
"
Z 100
214 209
161
11
6
1
<{
~
(
1
':
~
' 1
~
1
82
50.0 60.0
Fig. 5.16
Standard deviationelS.Sl
Meanvaluc=60.5
Samplcs=I007.00
120
Standard
dC
"\'iation=20,46
101
Mcan valuc-36 .7
100
Samplcs
-I0
77.00
91
82
86
~
80
79
~
...
0
]
60
E
"
Z
40
33
20
6
0 10,0
20.0 30.0 80.0
90,0
100.0
Fig. 5.17 Distribution
of
temperature variation in LF refining
Chapter
5 Multi-Factor Mass Flow Control for Metallurgical Manufacturing Process 147
500
tandard devialion=7.38
Mean \'alue-66.4
Samples=1577.oo
3&4
400
~
~
300
'-
0
~
E
200
::J
Z
100
0
50.0 53.9
84.5
Casting time
of
one heal ste
eb
min
Fig. 5.18 Distribution
of
casting time
of
one heat steel on thin slab caster
Standard deviation=18.97
952
Mean value=146.1
-
Samples=1577.00
~
~
n
-
1000
1200
o
200
1l 800
"
-;;
'0
~
600
E
::>
Z 400
130.0 135.0 140.0
Ih
O 150.0
1
~5
.
0
160.0 165.0 170.0
Ih
O 180.0 I&YO
Slab temp. change in tunnel furnacel'C
Fig. 5.19 Distribution
of
slab surface temperature change from entering to exiting tunnel
furnace (as few grade in statistical samples only on low carbon steel
and middle carbon steel, so deviat ion is more obvious)
500
400
~
'0..
""
300
<;
~
'-
0
~
200
E
::J
Z
100
Standard dcviatione l
f,74
Mean value=137.4
Sampk's=1577.oo
213
10 1
100.9 107.7 114.5
453
169.1
Time from FI
10 F6Is
Fig.5.20
Distribution of time from slab entering stand FI to leaving stand F6
of
rolling mill