# Calibration and Validation of a Cone Crusher Model with Industrial Data

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

## 1. Introduction

^{®}—Andritz Automation, Mimic

^{®}—Emerson, and DynSim

^{®}—Aveva) are based on Whiten’s model. Such dynamic simulation software is in high demand in digital twin applications, control system design, operator training, and flowsheet evaluation.

## 2. Literature Review

^{®}. Anticoi et al. [26] analyzed the influence of the type of material on the parameters of the breakage function in a milling circuit. The parameters were calibrated against data from piston press tests and single compression strength tests; the optimal parameters were obtained by minimizing the root mean square error between the measured and calculated values. Ambiguous results led the authors to conclude that numerical algorithms may not be the best option for calibration in the presented scenario. In a fashion similar to their previous study, Anticoi et al. [33] also studied the influence of the operating conditions of a pilot-scale HPGR on the breakage function parameters.

## 3. Description of the Cone Crusher Model

#### 3.1. Classification Function

#### 3.2. Breakage Function

## 4. Proposed Calibration Strategy

- $DS=\{1,\dots ,\phantom{\rule{4pt}{0ex}}{n}_{d}\}$: set of ${n}_{d}$ datasets;
- $CL=\{1,\dots ,\phantom{\rule{4pt}{0ex}}{n}_{f}\}$: set of ${n}_{f}$ classes used to describe the PSD of the material;
- $CS{S}_{l}$: closed side setting (CSS; mm) of the cone crusher for dataset l, $l=1,\dots ,{n}_{d}$;
- $TP{H}_{l}$: throughput (dry t/h) for dataset l, $l=1,\dots ,{n}_{d}$;
- $F{80}_{l}$: 80% passing size (mm) of the feed for dataset l, $l=1,\dots ,{n}_{d}$;
- $LLE{N}_{l}$: length of the face of the mantle liner (mm) for dataset l, $l=1,\dots ,{n}_{d}$;
- $LH{R}_{l}$: liner age (hours) for dataset l, $l=1,\dots ,{n}_{d}$;
- $E{T}_{l}$: eccentric throw (mm) for dataset l, $l=1,\dots ,{n}_{d}$;
- ${\widehat{P}}_{i,l}\left({\dots}_{1}\right)$, ${P}_{i,l}$: predicted and measured product PSDs, respectively, for $i=1,\dots ,{n}_{f}$ and $l=1,\dots ,{n}_{d}$;
- $\theta $: vector of decision variables;
- $LB$: lower bound on the decision variables;
- $UB$: upper bound on the decision variables.

**B**is equal to one. Constraints (1515) and (16) define bounds on ${K}_{1}$ and ${K}_{2}$ based on empirical recommendations [20]. Finally, constraint (17) defines bounds on ${\theta}_{1}$.

## 5. Results and Discussion

#### 5.1. Evaluation of the FBS and CBS

#### 5.2. Calibration/Validation Using the FBS

#### 5.2.1. Scenario 1

#### 5.2.2. Scenario 2

#### 5.2.3. Scenario 3

## 6. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Napier-Munn, T.J.; Lynch, A. The modelling and computer simulation of mineral treatment processes—Current status and future trends. Miner. Eng.
**1992**, 5, 143–167. [Google Scholar] [CrossRef] - Frank, A.G.; Dalenogare, L.S.; Ayala, N.F. Industry 4.0 technologies: Implementation patterns in manufacturing companies. Int. J. Prod. Econ.
**2019**, 210, 15–26. [Google Scholar] [CrossRef] - Santos, M.S.; Pinto, T.V.; Júnior, Ê.L.; Cota, L.P.; Souza, M.J.; Euzébio, T.A. Simheuristic-based decision support system for efficiency improvement of an iron ore crusher circuit. Eng. Appl. Artif. Intell.
**2020**, 94, 103789. [Google Scholar] [CrossRef] - Nian, R.; Liu, J.; Huang, B. A review on reinforcement learning: Introduction and applications in industrial process control. Comput. Chem. Eng.
**2020**, 139, 106886. [Google Scholar] [CrossRef] - Qi, Q.; Tao, F. Digital twin and big data towards smart manufacturing and industry 4.0: 360 degree comparison. IEEE Access
**2018**, 6, 3585–3593. [Google Scholar] [CrossRef] - Whiten, W. The simulation of crushing plants with models developed using multiple spline regression. J. South. Afr. Inst. Min. Metall.
**1972**, 72, 257–264. [Google Scholar] - Lynch, A. Mineral Crushing and Grinding Circuits: Their Simulation, Optimization, Design, and Control; Elsevier Science Ltd.: Amsterdam, The Netherlands, 1977; Volume 1. [Google Scholar]
- Whiten, W.; Walter, G.; White, M. A breakage function suitable for crusher models. In Proceedings of the IV Tewkesbury Symposium, Melbourne, Australia, 12–14 February 1979; p. 19. [Google Scholar]
- Andersen, J.S. Development of a Cone Crusher Model. Master’s Thesis, School of Engineering, The University of Queensland, Brisbane, Australia, 1988. [Google Scholar]
- Evertsson, C.M. Modelling of flow in cone crushers. Miner. Eng.
**1999**, 12, 1479–1499. [Google Scholar] [CrossRef] - Herbst, J.; Oblad, A. Modern Control Theory Applied to Crushing Part 1: Development of a Dynamic Model for a Cone Crusher and Optimal Estimation of Crusher Operating Variables. IFAC Proc. Vol.
**1985**, 18, 301–307. [Google Scholar] [CrossRef] - Atta, K.T.; Johansson, A.; Gustafsson, T. Control oriented modeling of flow and size distribution in cone crushers. Miner. Eng.
**2014**, 56, 81–90. [Google Scholar] [CrossRef] - Atta, K.T.; Euzébio, T.; Ibarra, H.; Moreira, V.S.; Johansson, A. Extension, Validation, and Simulation of a Cone Crusher Model. IFAC-PapersOnLine
**2019**, 52, 1–6. [Google Scholar] [CrossRef] - Quist, J.; Evertsson, C.M. Cone crusher modelling and simulation using DEM. Miner. Eng.
**2016**, 85, 92–105. [Google Scholar] [CrossRef] - Cleary, P.; Sinnott, M.; Morrison, R.; Cummins, S.; Delaney, G. Analysis of cone crusher performance with changes in material properties and operating conditions using DEM. Miner. Eng.
**2017**, 100, 49–70. [Google Scholar] [CrossRef] - Sbárbaro, D.; Del Villar, R. Advanced Control and Supervision of Mineral Processing Plants; Springer Science & Business Media: Berlin/Heidelberg, Germany, 2010. [Google Scholar]
- Hulthén, E. Real-Time Optimization of Cone Crushers; Chalmers University of Technology: Göteborg, Sweden, 2010. [Google Scholar]
- Karelovic, P.; Putz, E.; Cipriano, A. Dynamic hybrid modeling and simulation of grinding–flotation circuits for the development of control strategies. Miner. Eng.
**2016**, 93, 65–79. [Google Scholar] [CrossRef] - Bhadani, K.; Asbjörnsson, G.; Hulthén, E.; Bengtsson, M.; Evertsson, M. Comparative study of optimization schemes in mineral processing Simulations. In Proceedings of the XXIX International Minerals Processing Congress, Moscow, Russia, 15–21 September 2018; pp. 464–473. [Google Scholar]
- King, R. Modeling and Simulation of Mineral Processing Systems; Elsevier: Amsterdam, The Netherlands, 2001. [Google Scholar]
- Richter, C.; Rößler, T.; Kunze, G.; Katterfeld, A.; Will, F. Development of a standard calibration procedure for the DEM parameters of cohesionless bulk materials—Part II: Efficient optimization-based calibration. Powder Technol.
**2020**, 360, 967–976. [Google Scholar] [CrossRef] - Karr, C.L.; Yeager, D. Calibrating computer models of mineral processing equipment using genetic algorithms. Miner. Eng.
**1995**, 8, 989–998. [Google Scholar] [CrossRef] - Seppälä, P.; Sorsa, A.; Paavola, M.; Ruuska, J.; Remes, A.; Kumar, H.; Lamberg, P.; Leiviskä, K. Development and calibration of a dynamic flotation circuit model. Miner. Eng.
**2016**, 96, 168–176. [Google Scholar] [CrossRef] - Klimpel, R.; Austin, L. The back-calculation of specific rates of breakage from continuous mill data. Powder Technol.
**1984**, 38, 77–91. [Google Scholar] [CrossRef] - Yianatos, J.; Vallejos, P.; Grau, R.; Yañez, A. New approach for flotation process modelling and simulation. Miner. Eng.
**2020**, 156, 106482. [Google Scholar] [CrossRef] - Anticoi, H.; Guasch, E.; Hamid, S.A.; Oliva, J.; Alfonso, P.; Garcia-Valles, M.; Bascompta, M.; Sanmiquel, L.; Escobet, T.; Argelaguet, R.; et al. Breakage function for HPGR: Mineral and mechanical characterization of tantalum and tungsten ores. Minerals
**2018**, 8, 170. [Google Scholar] [CrossRef][Green Version] - Perez, E.; Bouchard, J.; Poulin, É. Development and calibration of an autogenous/semi-autogenous mill simulation model for process control applications. In Proceedings of the 58th Conference of Metallurgists Hosting the International Copper Conference 2019, Vancouver, BC, Canada, 18–21 August 2019. [Google Scholar]
- Esnault, V.P.B.; Zhou, H.; Heitzmann, D. New population balance model for predicting particle size evolution in compression grinding. Miner. Eng.
**2015**, 73, 7–15. [Google Scholar] [CrossRef] - Torres, M.; Casali, A. A novel approach for the modelling of high-pressure grinding rolls. Miner. Eng.
**2009**, 22, 1137–1146. [Google Scholar] [CrossRef] - Hasanzadeh, V.; Farzanegan, A. Robust HPGR model calibration using genetic algorithms. Miner. Eng.
**2011**, 24, 424–432. [Google Scholar] [CrossRef] - Daniel, M.J.; Morrell, S. HPGR model verification and scale-up. Miner. Eng.
**2004**, 17, 1149–1161. [Google Scholar] [CrossRef] - Campos, T.M.; Bueno, G.; Barrios, G.K.; Tavares, L.M. Pressing iron ore concentrate in a pilot-scale HPGR. Part 2: Modeling and simulation. Miner. Eng.
**2019**, 140, 105876. [Google Scholar] [CrossRef] - Anticoi, H.; Guasch, E.; Oliva, J.; Alfonso, P.; Bascompta, M.; Sanmiquel, L. High-pressure grinding rolls: Model validation and function parameters dependency on process conditions. J. Mater. Res. Technol.
**2019**, 8, 5476–5489. [Google Scholar] [CrossRef] - Machado Leite, M. Kinetic models for the simulation of crushing circuits. Miner. Eng.
**1990**, 3, 165–180. [Google Scholar] [CrossRef] - King, R. Simulation—The modern cost-effective way to solve crusher circuit processing problems. Int. J. Miner. Process.
**1990**, 29, 249–265. [Google Scholar] [CrossRef] - Chehreghani, S.; Noaparast, M.; Rezai, B.; Shafaei, S.Z. Bonded-particle model calibration using response surface methodology. Particuology
**2017**, 32, 141–152. [Google Scholar] [CrossRef] - Yamashita, A.S.; Thivierge, A.; Euzébio, T.A. A review of modeling and control strategies for cone crushers in the mineral processing and quarrying industries. Miner. Eng.
**2021**, 170, 107036. [Google Scholar] [CrossRef] - Napier-Munn, T.J.; Morrell, S.; Morrison, R.D.; Kojovic, T. Mineral Comminution Circuits: Their Operation and Optimization; JKMRC Monograph Series in Mining and Mineral Processing; Julius Kruttschnitt Mineral Research Centre, University of Queensland: St. Lucia, Australia, 1996. [Google Scholar]
- Boggs, P.T.; Tolle, J.W. Sequential Quadratic Programming. Acta Numer.
**1995**, 4, 1–51. [Google Scholar] [CrossRef][Green Version] - Sivasubramani, S.; Swarup, K.S. Hybrid DE–SQP algorithm for non-convex short term hydrothermal scheduling problem. Energy Convers. Manag.
**2011**, 52, 757–761. [Google Scholar] [CrossRef]

**Figure 1.**Schematic representation of a cone crusher, extracted from Yamashita et al. [37].

**Figure 4.**Measured and predicted PSD curves to compare the effect of the number of degrees of freedom of the breakage function on the model accuracy for dataset 1.

**Figure 5.**Measured and predicted PSD curves in Scenario 1 for the Serra Leste datasets. (

**a**) PSD predicted using the model calibrated on datasets 1, 2 and 3. Validation with Dataset 1. (

**b**) PSD predicted using the model calibrated on datasets 1, 2 and 3. Validation with Dataset 2. (

**c**) PSD predicted using the model calibrated on datasets 1, 2 and 3. Validation with Dataset 3.

**Figure 6.**Measured and predicted PSD curves in Scenario 1 for the S11D datasets. (

**a**) PSD predicted using the model calibrated on datasets 4 to 7. Dataset 4 is used for validation. (

**b**) PSD predicted using the model calibrated on datasets 4 to 7. Dataset 5 is used for validation. (

**c**) PSD predicted using the model calibrated on datasets 4 to 7. Dataset 6 is used for validation. (

**d**) PSD predicted using the model calibrated on datasets 4 to 7. Dataset 7 is used for validation.

**Figure 7.**Measured and predicted PSD curves in Scenario 2 for the Serra Leste datasets. (

**a**) PSD predicted using the model calibrated on datasets 2 and 3. Validation with Dataset 1. (

**b**) PSD predicted using the model calibrated on datasets 1 and 3. Validation with Dataset 2. (

**c**) PSD predicted using the model calibrated on datasets 1 and 2. Validation with Dataset 3.

**Figure 8.**Measured and predicted PSD curves in Scenario 2 for the S11D datasets. (

**a**) PSD predicted using the model calibrated on datasets 4 and 7. Validation with Dataset 5. (

**b**) PSD predicted using the model calibrated on datasets 4 and 5. Validation with Dataset 6. (

**c**) PSD predicted using the model calibrated on datasets 5 and 6. Validation with Dataset 7.

**Figure 9.**Measured PSDs for CSS = 35 mm, 38 mm and 41 mm. Predicted PSDs for CSS = 37 mm, 39 mm and 40 mm.

Iron Mine | Cone Crusher | Dataset | $\mathbf{CSS}$ (mm) | $\mathbf{TPH}$ (t/h) | $\mathit{F}80$ (mm) |
---|---|---|---|---|---|

1 | 35 | 883 | 102.36 | ||

Serra Leste | Metso HP400 | 2 | 38 | 986 | 102.36 |

3 | 41 | 998 | 102.36 | ||

4 | 28 | 368 | 45.35 | ||

S11D | Sandvik CH660 | 5 | 28 | 368 | 73.03 |

6 | 28 | 368 | 30.89 | ||

7 | 28 | 368 | 27.79 |

FBS | CBS | |||
---|---|---|---|---|

Dataset(s) Used for Calibration | SSE | Time [s] | SSE | Time [s] |

1 | $6.04\times {10}^{-5}$ | 20.72 | 98.66 | 1.43 |

2 | $3.01\times {10}^{-6}$ | 13.32 | 72.16 | 2.90 |

3 | $2.21\times {10}^{-6}$ | 12.92 | 76.10 | 1.53 |

1 and 2 | 22.51 | 180.39 | 407.95 | 1.94 |

1 and 3 | 1.26 | 79.04 | 308.75 | 1.51 |

2 and 3 | $1.28\times {10}^{-2}$ | 97.63 | 155.85 | 1.60 |

1, 2 and 3 | 21.23 | 58.07 | 755.82 | 3.76 |

Mean (StDev) | 6.43 (±10.56) | 66.01 (±60.50) | 267.90 (±250.22) | 2.10 (±0.89) |

FBS | CBS | |||
---|---|---|---|---|

Dataset(s) Used for Calibration | SSE | Time [s] | SSE | Time [s] |

5 and 7 | 10.15 | 30.94 | 244.83 | 10.45 |

4 and 7 | 7.67 | 22.29 | 258.50 | 3.23 |

4 and 5 | $4.94\times {10}^{-2}$ | 56.52 | 70.15 | 2.99 |

4, 5, 6 and 7 | 479.51 | 42.11 | 1064.00 | 2.50 |

Mean (StDev) | 124.34 (±236.82) | 37.96 (±14.79) | 409.37 (±444.76) | 4.79 (±3.78) |

Iron Mine | ${\mathit{\alpha}}_{0}$ | ${\mathit{\alpha}}_{1}$ | ${\mathit{\alpha}}_{2}$ | ${\mathit{\alpha}}_{3}$ | ${\mathit{\beta}}_{0}$ | ${\mathit{\beta}}_{1}$ | ${\mathit{\beta}}_{2}$ | ${\mathit{\beta}}_{3}$ | ${\mathit{\gamma}}_{0}$ |
---|---|---|---|---|---|---|---|---|---|

Serra Leste (1 to 3) | 0 | 0.905 | 0 | 0 | 0 | 0.096 | 0.095 | 0 | 3.000 |

S11D (4 to 7) | 0.155 | 0.441 | 0 | 0.053 | 0.047 | 1.698 | 0 | 0 | 3.000 |

Iron Mine | ${\mathit{\alpha}}_{0}$ | ${\mathit{\alpha}}_{1}$ | ${\mathit{\alpha}}_{2}$ | ${\mathit{\alpha}}_{3}$ | ${\mathit{\beta}}_{0}$ | ${\mathit{\beta}}_{1}$ | ${\mathit{\beta}}_{2}$ | ${\mathit{\beta}}_{3}$ | ${\mathit{\gamma}}_{0}$ | $\Phi $ | $\mathit{\delta}$ | $\mathit{\sigma}$ |
---|---|---|---|---|---|---|---|---|---|---|---|---|

Serra Leste (1 to 3) | 0 | 1.375 | 0.018 | 0 | 0 | 3.060 | 0 | 0 | 3.000 | 1.000 | 0.506 | 211.441 |

S11D (4 to 7) | 0.001 | 0.425 | 0 | 0.076 | 0.004 | 0.704 | 0.063 | 0.177 | 3.000 | 1.000 | 0.400 | 875.255 |

Datasets | ${\mathit{\alpha}}_{0}$ | ${\mathit{\alpha}}_{1}$ | ${\mathit{\alpha}}_{2}$ | ${\mathit{\alpha}}_{3}$ | ${\mathit{\beta}}_{0}$ | ${\mathit{\beta}}_{1}$ | ${\mathit{\beta}}_{2}$ | ${\mathit{\beta}}_{3}$ | ${\mathit{\gamma}}_{0}$ |
---|---|---|---|---|---|---|---|---|---|

2 and 3 | 0.431 | 0.249 | 0 | 0.096 | 0 | 2.507 | 0 | 0 | 1.695 |

1 and 3 | 0 | 7.481 | 0.277 | 0 | 0 | 2.451 | 0 | 0 | 2.636 |

1 and 2 | 0 | 0.950 | 0 | 0 | 0 | 0 | 0.082 | 0 | 3.000 |

4 and 7 | 0.021 | 0.391 | 0 | 0.109 | 0.017 | 1.309 | 0.030 | 0 | 3.000 |

4 and 5 | 0.193 | 0.486 | 0 | 0.005 | 0.033 | 1.703 | 0 | 0 | 3.000 |

5 and 6 | 0.387 | 0.454 | 0 | 0.029 | 0.206 | 1.693 | 0 | 0 | 3.000 |

Datasets | ${\mathit{\alpha}}_{0}$ | ${\mathit{\alpha}}_{1}$ | ${\mathit{\alpha}}_{2}$ | ${\mathit{\alpha}}_{3}$ | ${\mathit{\beta}}_{0}$ | ${\mathit{\beta}}_{1}$ | ${\mathit{\beta}}_{2}$ | ${\mathit{\beta}}_{3}$ | ${\mathit{\gamma}}_{0}$ | $\Phi $ | $\mathit{\delta}$ | $\mathit{\sigma}$ |
---|---|---|---|---|---|---|---|---|---|---|---|---|

2 and 3 | 0.024 | 2.127 | 0.303 | 2.380 | 0 | 3.500 | 0 | 0 | 2.275 | 1.000 | 0.534 | 75.929 |

1 and 3 | 0.022 | 29.596 | 1.410 | 2.276 | 0 | 3.218 | 0 | 0 | 3.000 | 1.000 | 0.514 | 237.607 |

1 and 2 | 0 | 0.500 | 0 | 0 | 0 | 0 | 0.067 | 0 | 2.688 | 0.016 | 0.487 | 242.818 |

4 and 7 | 0 | 0.642 | 0.011 | 0 | 0.047 | 0.096 | 0.093 | 0.379 | 3.000 | 1.000 | 0.343 | 577.232 |

4 and 5 | 0 | 0.858 | 0.051 | 0.192 | 0 | 0.397 | 0.094 | 0.076 | 3.000 | 1.000 | 0.408 | 764.263 |

5 and 6 | 0 | 0.903 | 0.035 | 0.053 | 0.001 | 0.813 | 0.109 | 0. | 3.000 | 1.000 | 0.417 | 644.004 |

CSS [mm] | *35 | 37 | *38 | 39 | 40 | *41 |
---|---|---|---|---|---|---|

$\widehat{P}80$ [mm] | 34.66 | 40.84 | 46.95 | 48.51 | 50.00 | 52.39 |

$P80$ [mm] | 33.78 | - | 47.16 | - | - | 53.90 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Duarte, R.A.; Yamashita, A.S.; da Silva, M.T.; Cota, L.P.; Euzébio, T.A.M. Calibration and Validation of a Cone Crusher Model with Industrial Data. *Minerals* **2021**, *11*, 1256.
https://doi.org/10.3390/min11111256

**AMA Style**

Duarte RA, Yamashita AS, da Silva MT, Cota LP, Euzébio TAM. Calibration and Validation of a Cone Crusher Model with Industrial Data. *Minerals*. 2021; 11(11):1256.
https://doi.org/10.3390/min11111256

**Chicago/Turabian Style**

Duarte, Robson A., André S. Yamashita, Moisés T. da Silva, Luciano P. Cota, and Thiago A. M. Euzébio. 2021. "Calibration and Validation of a Cone Crusher Model with Industrial Data" *Minerals* 11, no. 11: 1256.
https://doi.org/10.3390/min11111256