Blast furnace load simulation using EDEM

Guilherme Defendi – Vale

Anderson Baltazar – Vale

Paulo de Freitas Nogueira - Vale

Daniel Schiochet Nasato - ESSS

SUMMARY

• Overview - Vale

• Introduction

• Objectives

• Conclusion

Objectives

• Use EDEM to predict the material behavior when loading the

blast furnace;

• Analyze the voidage distribution in the blast furnace;

• Use EDEM to predict compressive forces;

• Show potential of EDEM to simulate blast furnaces.

Geometry dynamics

• Geometry dynamics:

3 opening (batches)

• Linear translation z:

-75mm/s

• Total opening time:

10 s per batch

Operating conditions

Informed by Vale:• Batch 1 (Coke):

– Coke: 31.2 ton

– Coke density (bulk): 500 kg/m³

• Batch 2 (Metal):

– Sinter: 70.59 ton

– Pellet: 26.91 ton

• Batch 3 (Metal):

– Sinter : 31.07 ton

– Pellet: 20.93 ton

– Metal density (bulk): 2300 kg/m³

Calculated by ESSS• Material density (absolute):

– Coke: 1250 kg/m³

• Considering a maximum

packing of 40%

• 500 kg/m³ /0.4 =1250 kg/m³

– Metal: 5750 kg/m³

Plate Position

(displacement):– Batch 1: 300mm

– Batch 2: 400mm

– Batch 3: 200mm

Particle Size

• Informed by Vale:

• For pellet was used two fixed value from 8mm and 5mm

Pellet Size (mm) 18 8 5

Withheld (%) 0,0 97,96 2,04

Coke Size (mm) 40 37,5 30 25

Withheld (%) 69,2 12,47 14,87 3,46

Sinter Size (mm) 50 25 15 10 5

Withheld (%) 11,77 34,12 26,56 13,92 13,63

0

10

20

30

40

50

60

70

80

25 27 29 31 33 35 37 39

COKE

0,00

5,00

10,00

15,00

20,00

25,00

30,00

35,00

40,00

5 10 15 20 25 30 35 40 45 50

SINTER

Data calibration – Coeff. of Restitution

• Experiments were performed at Vale, in order

to obtain correct materials data;

– Coefficient of restitution, static and rolling friction

were calibrated at ESSS.

• Restitution parameters calibrated based on

free fall test.

Initial Hight (cm) Return Hight (cm) Coeff of Restitution

40 0,400 0,100

40 1,270 0,178

40 1,200 0,173

40 1,000 0,158

40 1,140 0,169

40 1,300 0,180

pellet/pellet

sinter/pellet

sinter/coke

sinter/sinter

coke/pellet

coke/coke

Data calibration – Static Friction

• Parameters were calibrated based on normal/shear force

test.

y = 0,6574x + 8,4531

y = 0,2222x + 10,054

y = 0,6484x + 43,417

0,00

10,00

20,00

30,00

40,00

50,00

60,00

70,00

0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00

Shear Force(N)

Normal Force (N)

Sinter Pellet Coke Sinter Pellet Coke

Coke Sinter Pellet Sinter Pellet Coke

9,81 14,24 36,92 12,90 12,54 49,38 13,17

19,62 22,68 45,37 20,46 13,79 56,94 16,28

29,43 27,13 56,50 25,36 16,90 62,10 18,68

Shear Force (N)Normal

Force (N)

Data calibration – Rolling Friction

• Based on repose angle, simulations were

performed to reproduce the same angle on

EDEM;

• Friction coefficient were adjusted to

reproduce the repose angle.

Parameter coke/coke coke/pellet coke/sinter coke/steel pellet/pellet pellet/sinter pellet/steel sinter/sinter sinter/steel

Coeff. of Restitution 0.18 0.1 0.17 0.3 0.18 0.16 0.3 0.18 0.3

Coeff. of Static Friction 0.58 0.62 0.65 0.5 0.64 0.9 0.5 0.52 0.5

Coeff. of Rolling Friction 0.4 0.35 0.38 0.25 0.25 0.33 0.25 0.35 0.25

Material Repose angle (degree)

Coke 35 to 38

Sinter 29 to 33

Pellet 25 to 26

Simplifications

• Due to real model size, it was simulated only a slice of the

real geometry;

• Batch 1 (Coke): – Coke: 174.2 Kg

• Batch 2 (Metal): – Sinter: 394 Kg

– Pellet: 150.2 Kg

• Batch 3 (Metal): – Sinter : 173.4 Kg

– Pellet: 116.8 Kg

Initial conditions – particle creation

• It was defined a virtual area in the top of the blast furnace

to create particles.

• Particles were created in a dynamic way.

• Approximated 500 thousand particles simulated.

Results

Voidage Distribution

• Sample points in the material bed – voidage distribution

Material Distribution

Coke distribution Metal distribution

Material Distribution

Sinter distribution Pellet distribution

Compressive Force

• Compressive force can be seen in the particle bed

Compressive Force

• Compressive force profile similar to coke distribution profile

Coke distribution profileCompressive force profile

Conclusions

• EDEM has demonstrated the ability to reproduce real

phenomena from a large and complex model;

• A furnace model was simulated in about 160 hours:

– More than 500.000 Particles Simulated;

– Physical results based on real Physics.

• EDEM is able to generate quantitative results for blast

furnaces;

– Particle size distribution, compressive force, total force, voidage.