Simulation and advanced control of

transient behaviour in gyratory

cone crushers
Pekka Itävuo*, Antti Jaatinen**, Matti Vilkko*

*Tampere University of Technology, Finland

**Metso Automation, Finland
 Agenda
• Introduction
• Dynamic model of a gyratory cone crusher
• Simulation results
• Conclusions
 Introduction
• Gyratory cone crushers are commonly used comminution equipment in mines and
quarries
– Gyratory cone crusher in this work refers to the secondary and tertiary hydrocone type crushes
• Disturbances like changes in feed material size & hardness and wear of crushing
liners cause variation in the crusher output
– As a result, the process is never in a steady state although modelling is often carried out
as if it were
– Desired performance is only achieved if the effect of occurring disturbances is fully
compensated
• Crushers are typically operated with constant closed side setting (CSS) or constant
load
• None of the currently existing cone crusher models is truly useful for analytic
control system design purposes
– Modelling in crushers is biased towards process dimensioning and equipment design i.e.
steady-state models
 Introduction
• Main components of a • Operating variables of a
gyratory cone crusher gyratory cone crusher

Closed side setting

CVs have nonlinear &

dynamic relationship
with associated MVs
& DVs

Nominal stroke length

Eccentric speed
 Dynamic model of a gyratory
cone crusher
• Modelling principle
– Combine the existing static nonlinear performance
models with simple linear dynamics
– Hammerstein type system:
 Dynamic model of a gyratory
cone crusher
 Feed material +
 Material properties

 Moisture (M)
Feed grading (FG)
w Crushability (LA)
Feed rate (Q)

 Specific Gravity (SG)

Feed flakiness index Linear dynamics
64

63.5  (FFI)
CA

63

62.5

15 450
20 400
Feed hopper

350
25 300 Material volume in hopper [m3]
CSS
250
ES model
Linear dynamics Static nonlinearities


Static nonlinearities
(Log-transformed




(ANFIS)
Linear regression)

Product size 

Product Size [mm]
distribution model


Piston
CSS dynamics y
u model
Engine
Flow model
Crusher
dynamics model
Power model RMS Power [kW] 
 ES dynamics
model

Stroke
Cavity Capacity model Capacity [t/h] 

geometry
Linear dynamics

y = f (u (t ), w(t ))
Linear dynamics




Crusher constants
 Simulation results
-Controllability of product size distribution-
• Metso Nordberg Disturbance variable Min Max
GP200M tertiary Specific Gravity 2.6 2.8
gyratory cone Crushability, LA-value 17 20
Maximum effect
crusher Feed moisture [%] 0.1 3.0
Feed Flakiness Index 10 34 of disturbances
– 25 mm stroke
Feed Grading [mm] 5/32 5/104
length Feed Rate [% of max] 75% 100%
– ~150 t/h capacity
• Feed material, Maximum effect of
Granite manipulated variables
– Variations based
on realistic data on
quarry operation
• Influence of MVs Manipulated variable Min Max
exceed the maximum Closed side setting (CSS) 10 30
range of generated Eccentric speed (ES) 250 450
disturbances
 Simulation results
-Product size distribution control-
DV: Change in feed size & moisture CV: Product Size
• Metso Nordberg GP200M
Input Sequence Metso GP200 Product size
4 15

[%], Feed size/10 [mm]

Moisture Constant CSS
Feed size Constant Size

tertiary gyratory cone 3.5

Feed size 24 17 36 24
14.5

14

Moisture [%], Feed equivalent size / 10 [mm]

crusher

size [mm]
3
13.5

Product Equivalent Size [mm]

2.5

– 25 mm stroke length 2
13

12.5

– ~150 t/h capacity

Product
12
1.5

• Feed material, Granite 11.5

Moisture
1
Moisture 0.5 3 0.1 0.5 3 11

– LA 18.5, SG 2.7 [t/m³] 0.5

10.5

• Nonlinear CSS actuator 0

0 100 200
time [s]
300 400 500
10
0 100 200
time [s]
300 400 500

– Pump up MV: Closed side setting (CSS)

Metso GP200 CSS Crusher Power
Metso GP200 Power Draw
23 120

– On/off valve down [mm]

22
Constant CSS
Constant Size
Constant CSS
Constant Size

110

• Control objective:

wer Draw[kW] [kW]

21
d Side Ssetting

100

– Maintain the desired

etting [mm]

Crusher PoPower
20

product size 19 90
side

Crusher
• Controller:
Close

18
Closed

80

17

– PID with Ziegler Nichols 70

tuning 16
large CSS changes needed for steady product size constant power ≠ constant product size
15 60
0 100 200 300 400 500 0 100 200 300 400 500
time [s] time [s]
 Conclusions
• Current operating modes (constant CSS or
constant load) cannot provide steady product
size
• The effect of every reasonable process
disturbance can be cancelled by using
advanced feedback control system
• Further investigation and full-scale tests are
needed in order to validate these findings