MONITORING

OF METALLURGICAL
PROCESSES DURING
HOT ROLLING

ITA Ltd., Ostrava, Czech republic

Metallurgical Monitoring
What does it mean?

METALLURGICAL MONITORING
prediction of final mechanical properties of hot rolled steel
products coupled with process data being received on-line
from the rolling mill control system!
Metallurgical Monitoring
Three possible levels of implementation

It was
• off-line implementation developed
Off-line tool for technologists that enables prediction of yet
final! mechanical
properties of rolled production based on predefined and manually
entered rolling technology parameters.
• on-line monitoring It is being
Similar functionality as off-line level but developed
implemented directly to the
Level2 of the control system. It enables automatical
for flats ! loading of rolling
technology parameters and passive monitoring of final mechanical
properties of whole production.
• on-line control It is
Level2 application that based on actual rolling conditions sets up cooling
planned for
strategy to receive requested or improved final mechanical properties of
the future!
rolled product.
 Metallurgical Monitoring
Physical based temperature & metallurgical modules

METARoll Module
Input Data • grain evolution
Technology&Process • recrystallization
Steel Properties • residual strain
• precipitation

Temperature Module The last Pass Data

THERM 1D FEM • residual strain of austenite
Time – Temperature • austenite grain size
curve for a single bar

METACool Module Output Data

Temperature Module • microstructure
• CCT Diagram
COIL 2D FEM in every Pass
• austenite decomposition
Time – Temperature • structure shares
• mechanical properties
curve for a bar in coil • mechanical properties
Metallurgical Monitoring
Steel groups and limits of chemical composition
C Mn Si Cr Ni Mo V W Ti Nb B Al N
GROUP
min max min max min max min max min max min max max max max max max max max

C < 0.06 0.001 0.06 0.05 0.70 0.02 0.30 0 0.10 0 0.10 0 0.10 0.05 - 0.05 0.10 0.004 0.05 0.015

C < 0.20 0.06 0.25 0.20 1.50 0.05 1.50 0 0.40 0 0.25 0 0.10 0.10 - 0.05 0.10 0.004 0.05 0.015

C < 0.50 0.20 0.55 0.20 1.50 0.05 1.50 0 0.40 0 0.25 0 0.10 0.10 - 0.05 0.10 0.004 0.05 0.015

C < 0.95 0.50 1.15 0.20 1.50 0.05 1.50 0 0.40 0 0.25 0 0.10 0.10 - 0.05 - 0.004 0.05 0.015

Mn 0.10 0.50 1.50 2.00 0.05 1.50 0 2.0 0 0.25 0 0.10 0.25 - - 0.10 0.004 0.05 0.015

Cr C < 0.6 0.15 0.65 0.20 1.20 0.05 0.40 0 1.50 0 0.25 0 0.10 0.40 - - - 0.004 0.05 0.015

Cr C > 0.6 0.60 1.15 0.20 1.20 0.05 0.40 0 1.50 0 0.25 0 0.10 0.40 - - - 0.004 0.05 0.015

Mo 0.15 0.55 0.40 1.20 0.05 0.40 0 0.40 0 0.25 0 0.40 - - - - 0.004 0.05 0.015

Cr-Mo 0.15 0.65 0.40 1.00 0.05 0.40 0 1.50 0 0.25 0 0.40 - - - - - 0.05 0.015

Cr-Si 0.45 0.70 0.20 2.00 1.00 2.00 0 1.50 0 0.25 0 0.10 0.10 - - - - 0.05 0.015

Ni-Cr-Mo 0.15 0.65 0.40 1.00 0.05 0.40 0 1.50 0 3.50 0 0.40 - - - - - 0.05 0.015
Metallurgical Monitoring – Theoretical Background
FEM Temperature calculations

• Therm 1D -> nonstationary heat

transfer FE analysis for 1D axisymmetric
bodies (linear 2-node ring)
• Coil 2D -> nonstationary heat transfer
FE analysis for 2D plane axisymmetric
bodies (linear 4-node quadrangle)
• Model of thermal properties ->
thermal conductivity, density and specific
heat depend on the coil temperature and
its tightening
• Heat transfer -> time and temperature
dependent heat transfer coefficient and
ambient temperature
 Metallurgical Monitoring – Theoretical Background
FEM Temperature calculations
Time - Temperature curves - Head
39mm , 3.00 m/s, 160x160 mm / 12 m
T emperature [°C] Induction surface temperature
1211.6
reheating core temperature
5.00 mm below surface
ST AND#1 ENTRY 1151.6°C
ST AND#4 EXIT 1086.2°C
1100.0
slow cooling ST AND#5 ENTRY 1076.4°C
ST AND#13 ENT RY 1048.1°C
on cooling bed ST AND#18 EXIT 1026.0°C
SIZING ENT RY 795.8°C
or in coil
1000.0
SIZING EXIT 837.9°C
CB/GARRET T ENT RY 766.2°C
900.0

800.0

700.0
heat
transfer heat
generation
600.0
into rolls
500.0
due plastic
deformation
accelerated
386.1
0.00 water cooling
80.00 160.0 240.0 320.0 400.0 480.0 541.1
T i me [s]
 Metallurgical Monitoring – Theoretical Background
MetaROLL Module – Metallurgy during rolling

Activation energy of recrystalization

Chemical composition
Zenner-Hollomon parameter

Process parameters
Increasing of the activation energy
in every pass:
in case of precipitation
• temperatures
• strains
• strain rates
Recrystallized part Xrex for given
temperature, effective strain and time
available between neighbouring passes

! pure physical approach ! Residual strain

Austenitic grain after every pass - Head average grai n si ze

FIrest = FIef *(1 - Xrex)

da sd, 15.00 m /s, 15 0x12 5 m m / 1 4 m grai n after growth
grai n after re crystal l i zati on
Grai n si ze [um ] T em perature [°C]
average rol l i ng te m perature
40 0.0 average i nte 11 00.0
rsta nd te m perature

10 00.0
35 0.0

90 0.0

30 0.0

Size of recrystallized
80 0.0

25 0.0 70 0.0

20 0.0
60 0.0
grain of austenite
50 0.0

15 0.0 40 0.0

30 0.0
10 0.0

Growth of recrystallized grain

20 0.0

50 .00
10 0.0

0.00
1/1 1/2 1/3 1/4 1/5 2/1 2/2 2/3 2/4 2/5 2/6 3/1 3/2 4/1 4/2 4/3 4/4
0.00
between particular passes
M i l l / Pass
Metallurgical Monitoring – Theoretical Background
MetaCOOL Module – Metallurgy after rolling - Step 1

Temperature [°C]
1000,0
Temperature [°C]

CCT Diagram
900,0

800,0
1000,0 GRAIN, FIREST
Prediction 700,0
900,0

600,0 800,0
500,0 Carbon&alloyed Steels
700,0 Time Coordinates of Noses
400,0

600,0
Structure Shares
300,0 10000000

1000000
200,0 500,0
Calculation 100,0
100000

10000 ferrite strat

0,010 400,0
0,10 1,00 10,00 100,0 1000,0 10000

from literarture [s]

pearlite start
1000 Time [s]
perlite finish
300,0 100
bainite start
10 bainite finish

200,0 1

0,1

Mechanical Properties 100,0 0,01

0,001

Calculation
1,00 10,00 100,0 0,001 0,01
1000,0
0,1 1
10000
10 100
100000
1000
1000000
10000 100000 1000000
computed [s]

Linear multiparametric regression Time [s]

S(i) = exp (Bo + (B(i) . C(i))

T(i) = Ao
DS + (A(i) . C(i)) 1/3
Sx Sox AS exp((-BS CS CEQ) FIR) DI exp(ES MIKRO) FIR )) exp( FS B)
Metallurgical Monitoring – Theoretical Background
MetaCOOL Module – Metallurgy after rolling - Step 2

CCT Diagram
Prediction Avrami equation
for pearlite and bainite
transformation
n
Structure Shares X (t, T) (1 exp( k t )) X
i
Calculation

Mechanical Properties Koistinen-Marburger equation for martensite

Calculation
transformation
Xm(T) (1 exp( b (T T)n )) X
Ms
Metallurgical Monitoring – Theoretical Background
MetaCOOL Module – Metallurgy after rolling - Step 3

Linear regression
HV = CO + %Fe * (C1i * ci) + %Pe * (C2i *ci) +
CCT Diagram %Ba * C3i * ci) + %Ma * (C4i * ci)
Prediction Vicker's hardness of tool steels
for complete interval of chemical compositions and
for different cooling rates

1000

900

experimental values of Vicker's hardness

800

from anisothermal TTT diagrams

Structure Shares 700

Calculation
600

500

400 900 tempering 100°C

300
tempering 200°C
800 tempering 300°C

HV hardnessfrom experiments
200
tempering 400°C
100
100 200 300 400 500 600 700 700 800 900 tempering
1000 500°C

Mechanical Properties
Vicker's hardness calculated by TTSteel

600

Calculation
500

400

300
300 400 500 600 700 800 900
HV hardness from regression analysis
Metallurgical Monitoring – Off-line implementation
SPP software – Strip Properties Predictor
Metallurgical Monitoring – Off-line implementation
DLPP software – Danieli Long Products Properties Predictor
 Off-line implementation
DLPP software – Verification for WIRE ROD rolling
( 64 pcs) C Mn Si Cr Ni Ti B
various
diameters 0.403 0.64 0.26 0.06 0.02 0.002 0.0002

650

600
Výpočet [MPa]

550

500

450

800
400
400 450 500 550 600 650
Měřeni [MPa]
750
Výpočet [MPa]

700

650
Yield Stress Ultimate
600
Tensile
600 650 700 750 800 Strength
Měřeni [MPa]
 Off-line implementation
DLPP software – Verification for BAR rolling

C Mn Si Ti
0.065 0.6 0.2 0.15

Finální mechanické vlastnosti - střední Fe ri t

Perl i t
T e pl ota [°C] Bai ni t
M arte nzi t
10 00.0
prům ěrná tepl ota
po vrchová te pl o ta

90 0.0

80 0.0

70 0.0

60 0.0

50 0.0

40 0.0

F i 30
n ál0.0
n í vl astn o sti
Diameter Re Rm % %
Teplot y [ ° C ]
Ac 3 = 890
40 mm (MPa) (MPa) of Ferrite of Pearlite
Ac 1 = 745
20 0.0
Ar3 = 782
Ar1 = 647
St ruk t urní podí ly [ %] measured 284 396 97 3
Feri t = 85.7
Perli t
10 0.0 = 14.3
Bai ni t = 0.0
Martenzi t = 0.0 calculated 305 414 86 14
Aus teni t = 0.0
Mec hanic k é v las t nos t i
0.00
HV = 166
0.001 = 161
HB 0.010 0.10 1.00 10 .00 10 0.0 10 00.0 10 000
Re = 310 MPa
Rm = 494 MPa Čas [s]
On-line Monitoring
An example for Hot Strip Mill - Strip Observer Monitor

Strip Observer Monitor predicts final mechanical

properties of hot rolled strip in coil in one point of
interest from process data being received on-line from
the rolling mill control system.

thickness rolling direction

Point of interest
(L; W; T) length

width
On-line Monitoring
Strip Observer Monitor – Basic chart
SpOb Monitor
Superior Level The L2 delivers predicted (Setup) and
measured process data in the point of interest
visualisation
User selects requested coil off-
on-line after finishing production ofofeach coil!
results
line from the list of coils available
Predicted strip temperatures except
in the SpOb Database!
temperatures when SpObstrip is cooled in coil are
delivered as wellPredictor
but it is not strictly required.
L2, LCCS All temperatures can be calculated!
coil
Measured Data The Quality Control Data from the L3 are
selection
Predicted Data
imported off-line when some significant
SpOb SpOb
portion
Database of this data is at disposal. The user
Corrections
L3 decides about running the SpOb Corrector
Quality Control to update corrections of the SpOb Predictor!
Data
SpOb
Corrector

learning
On-line Monitoring
Strip Observer Monitor – Cooling of strip in coil

TemperatureEdgeL (i) strip segment (i) Model of cooling in coil

supposes the final strip is
divided lengthwise into uniform
TemperatureCentre (i) Point of interest W
segments and average
temperatures and coiling times
TemperatureEdgeR (i) L are know for each segment!

Plane strip segments are coiled step Point of interest

Coil Layer (i)
by step and create axisymmetric coil
layers

2D axisymmetric FEM Model

Cooling of the final coil after coiling is

calculated by the same temperature
module but for fixed coil dimensions.
On-line Monitoring
Strip Observer Monitor – SW Implementation
 On-line Monitoring
Data Interface – Strip Data

Chemical composition
double C, Mn, Si, Cr,V, Ti, Nb, Al, N,
chemCompos Mo, Ni, B, W, P, S, Cu, As, Sn, Co;
SPOB_CHEM_COMPOS and initial grain size of austenite of steel
chemCompos; double Grain;
double dThickness; dThickness Strip thickness in mm.
double dLength; dWidth Strip width in mm.
double dWidth; dLength Strip length in mm.
double dLengthCoord; dLengthCoord Relative coordinate of the point of interest along strip length.
double dWidthCoord;
dWidthCoord Relative coordinate of the point of interest along strip width.
double dThicknessCoord;
dThicknessCoord Relative coordinate of the point of interest along strip width.
int nLayer;
nLayer Number of strip lengthwise segments.
double *dTau;
*dTau Coiling times of each strip segment in sec.
double *dTempLeftEdge;
Lengthwise averaged temperature of the left edge of each strip
double *dTempCentre; *dTempLeftEdge
segment before coiling in C.
double *dTempRightEdge; *dTempCentre
Lengthwise averaged temperature in the centre of each strip
Double dInDiameter; segment before coiling in C.
Lengthwise averaged temperature of the right edge of each strip
double dOutDiameter; * dTempRightEdge
segment before coiling in C.
dInDiameter Inner diameter of the coil in mm.
dOutDiameter Outer diameter of the coil in mm.
On-line Monitoring
Data Interface – Rolling Technology Data

nPass Number of passes.

int nPass; *pFI Logarithmic strain in each pass.
double *dFI;
*pRate Strain rate in each pass in s-1.
double *dRate;
double *dTVA; *pTempTVA Rolling temperature in each pass in C.
SPOB_CURVE dTPC; dTPC Temperature curve between passes.
double MTemp_Furnace; MTemp_Furnace Measured temperature in the Stossofen1 in C.
double MTemp_MH1; MTemp_MH1 Measured temperature in the Messhaus1 in C.
double MTemp_MH2;
MTemp_MH2 Measured temperature in the Messhaus2 in C.
double MTemp_MH3;
double MTemp_MH4; MTemp_MH3 Measured temperature in the Messhaus3 in C.
double MTemp_MH5; Measured temperature in the Messhaus4 (exit
MTemp_MH4
rolling temperature) in C.
Measured temperature in the Messhaus5 (coiling
MTemp_MH5
temperature) in C.
On-line Monitoring
Special module 1
Beginning of austenite transformation during rolling:
int SpOb_CalcTransf(
const SPOB_MAIN_INPUT *pMainInput,
const int nResultCount,
SPOB_TRANSFORMATION_RESULT *pResults,
int *pPassNumber);
Input data structure containing information
const SPOB_MAIN_INPUT
about rolling process parameters in the point
*pMainInput
of interest
const int
Number of passes
nResultCount
Output data structures containing information
SPOB_TRANSFORMATION_RESULT
about austenite transformation in the point
*pResults
of interest in each pass
Index of the pass where austenite
int
transformation in the point of interest
*pPassNumber
started
Metallurgical Monitoring
Special module 2

Full automatic regime of metallurgical calculations:

int SpOb_AutomaticCalc(
const SPOB_MAIN_INPUT *pMainInput,
const SPOB_COIL_INPUT *pCoilInput,
SPOB_PROPERTIES_RESULT *pResult);

Input data structure containing information

const SPOB_MAIN_INPUT
about rolling process parameters in the point
*pMainInput
of interest
const SPOB_COIL_INPUT Input data structure containing information
*pCoilInput about cooling of the strip in coil
Output data structure containing calculated
SPOB_PROPERTIES_RESULT
microstructure shares and mechanical
*pResult
properties of strip in the point of interest
Metallurgical Monitoring
Summary
• the software SPP and DLPP has been developed for off-line computer
simulation of metallurgical processes in hot rolled strips, bars and wire
rods during rolling and after subsequent cooling!
• the SpOb software has been developed for prediction of final
mechanical properties of hot rolled strips coupled with process data
being received on-line from the rolling mill control system!
• based on specified steel chemistry and rolling technology these
software tools predict microstructure parameters of deformed
austenite after rolling, especially grain size, recrystallized fraction and
retained strain!
• based on specified steel chemistry, microstructure of deformed
austenite and cooling strategy these software tools predicts secondary
structure shares and corresponding mechanical properties of final
product!
 thanks
for your
attention!