Tumbling Mill Research Articles

Experimental investigation of load behaviour of an industrial scale tumbling mill using noise and vibration signature techniques

Mill load (i.e. the load level of coal powder), which is critical in improving the production capacity and energy efficiency of pulverizing system in thermal power plant, has not been effectively monitored and controlled industrially. This paper investigates the load behaviour in an industrial scale tumbling mill under practical working conditions. A microphone and an accelerometer were installed to pick up mill noise and inlet trunnion vibration signals, respectively. By analyzing the sensitivity distributions of mill noise energy and mill vibration energy, characteristic power spectra (CPS) of mill noise and mill vibration were obtained. The CPS energy, centroid frequency and frequency domain variance of the mill noise and mill vibration were then investigated and compared under various working conditions. Experimental results show that the CPS energy of both mill noise and mill vibration can accurately represent the mill load. Moreover, the centroid frequency and frequency domain variance of mill noise can also be used to determine mill load. By combining these characteristics of mill noise and mill vibration, an improved estimation of mill load can be achieved.

The steel wheel abrasion test (SWAT): A tool to study wear, friction and ore breakage in the mining industry

Developed from the rubber wheel abrasion standard test, the steel wheel abrasion test (SWAT) illustrates a simple highly controllable testing environment that produces repeatable results for studies of material wear, friction in wear and ore breakage during wear. The wear and friction results can be tied to equipment models such as those of mill charge motion to predict wear of both steel media and liner. The ore breakage results have the potential to be correlated to typical ore breakage tests such as the lab tumbling mill or the drop weight test. In this paper, the methodology used to avoid “highly anomalous results due to differential friction” is presented. It is shown that the proposed methodology leads to repeatable results for sample wear, friction and ore breakage. Wear, friction and breakage results from a selection of ore and steel media tests are presented. Further, steel media wear estimates are presented and compared with industrial mill reality.

DEM Simulation of Abrasion of Nonspherical Particles in Tumbling Mill

Abrasion is a surface breakage event where irregular surfaces of particles are removed. Abrasion produces rounded particles by comminuting irregular edges. Extensive experimental work has been performed to study abrasion events and their process parameters. Very few studies have used the discrete element method (DEM) to study the abrasion behavior of particles. In this regard, the research work uses the DEM simulation of abrasion of nonspherical particles to study the process parameters in a tumbling mill environment. A 1.1 m diameter and 0.2 m length mill with square-shaped lifters of 30 mm cross section has been simulated. Particle trajectories, collision force, collision energy, and energy spectra are evaluated to understand the effect of surface roughness on abrasion. It was observed that as the asperities of the particle decreases, the rate of abrasion decreases. Based on the energy analysis, it was found that the normal collision force (average 0.4 kN) is larger than the tangential collision force (average 0.09 kN) for the considered case, and hence the former maybe more responsible for the abrasion of particles.

Discrete element method study of abrasion

Abrasion is a surface breakage event where irregular surfaces of particles are removed. In other words, in milling it is the process which produces sphere like particles by comminuting asperities. Extensive experimental work has been performed to study abrasion events and their process parameters. Very few studies have used the discrete element method (DEM) to study the abrasion behavior of particles. In this regard, this paper reports on using DEM simulation to study the abrasion in a small scale tumbling mill environment. DEM simulations have been carried out for different test mills ranging in diameter from 0.2 m to 1.1 m and all of the mills have 0.2 m length. Each mill has square shaped lifters of 30 mm cross-section. The outcome of the simulation are evaluated to investigate the abrasion and milling parameters, like, force, different energy losses, collision frequency, power consumption. The DEM results are compared with experimental results performed at the JKMRC laboratory.

Validating the instrumented ball outputs with simple trajectories

An instrumented ball capable of measuring a number of physical quantities within a highly dynamic environment, such as a tumbling mill, has been designed, built and tested. To ensure that the instrumented ball is operating as designed, it is made to follow a number of known trajectories. The physical quantities measured by the instrumented ball are consistent with the expected results along the trajectories. An example of the use of a properly functioning instrument is also shown.

Assessment of mill lifter bar deflection measurements using wavelets and discrete element methods

This paper shows how Partial Least Square Regression (PLS) methods can be used to model sensor data of spectral character. The modelling approach has been applied on a tumbling mill where a strain gauge sensor measures the deflection of a lifter bar when it hits the charge. The deflection of the lifter bar during every mill revolution gives rise to a characteristic signal profile that is shown to contain information on both the charge position and grinding performance. As a signal pre-processing method the discrete wavelet transform is used. It distinctly shows a capability of signal feature extraction where both time and frequency are of interest. Its well-known ability to achieve good data compression without loss of information is also demonstrated, a data reduction ratio of 20:1 is obtained here. Modelling results demonstrate that different operating conditions are well distinguishable from each other and by that the finding of proper operating regimes are highly feasible. Grinding parameters that are normally measured in the laboratory are now readily modelled from the on-line signal. A further objective of this paper is to link the experimentally obtained strain gauge sensor data with computational data from a discrete element mill model (DEM). This enables to visualise the charge motion and helps to interpret the complex phenomena that take place inside a grinding mill measured by the strain gauge sensor. The approach taken is to simulate the behaviour of a rubber lifter when it is exposed to forces from the grinding charge in a two-dimensional DEM mill model using a particle flow code. The deflection profile obtained from the DEM simulation shows a reasonably good correspondence to pilot mill measurements.

Multivariate projection and analysis of microstructural characteristics of mechanically activated hematite in different grinding mills

A statistical analysis was done to investigate the relationship between grinding variables and structural changes during mechanical activation of hematite concentrate. Experiments were carried out according to a statistical design by varying the grinding time, media surface and mill type. Several multivariate techniques are applied to interpret the grinding processes. The variance analysis revealed that the media surface and grinding time significantly influence the five main response variables at 95% confidence level. The use of multivariate analysis allows the projection of high-dimensional data to a low subdimensional subspace. An overview of principal component analysis (PCA) on 27 variables yielded a three component model explaining 89% and predicting 76% of the total variance. It was found that the observations belonging to low and high levels of media surfaces fall into two groups. Most of the microstructural characteristics such as microstrain, dislocation and amorphization and granulometric surface area, BET specific surface area, specific energy input, stored energy, portion of smaller particles and stress energy coincide with high level of media surface group. The variables crystallite size, peak intensity and mean particle size appear with lower media surface. The PLS-DA (partial least squares discrimination analysis) made it possible to discriminate the three types of mills. From the projection of dummy variables, it was concluded that the vibratory mill caused comparatively less structural changes in hematite than the other mills in spite of releasing higher stress energy. The planetary mill introduced relatively higher dislocation defects and generated higher lattice strain. The hematite ground in the tumbling and planetary mills had comparatively higher X-ray amorphization degree and subsequently higher excess energy than the hematite ground in the vibratory mill. The tumbling mill produced relatively lower specific surface than the others. It was concluded that the products of the tumbling mill represented higher defect concentration (amorphization) per unit surface area despite releasing lower stress energy level. From the PLS modeling of the five main response variables, it was found that the X-variables specific energy input and stress energy are the most influencing factors.

媒体径分布をもつ転動ミルの乾式・湿式粉砕における媒体摩耗

The wear generated in grinding balls having a ball size distribution in a tumbling mill was investigated using a laboratory-scale mill for both dry and wet grinding. The balls were divided into five diameters with the same mass fractions. The weight loss of grinding balls with a diameter Db, Mb appeared to have a non-linear relationship with the time of milling t in the ball mill. The kinetics of wear could be expressed as a power law of the type Mb=Atb, where the coefficient A is a function of the ball diameter Db, and the exponent b is a constant. The coefficient A obtained for each ball diameter, with the exception of the maximum diameter, followed the power law A ∝Dbn-3, based on the formulation of the weight loss of a grinding ball mb as mb∝Dbn. The exponent n indicates the ball wear mechanism in ball mill grinding. A value of n = 2 fits the surface theory which indicates that the intensity of the abrasive interactions in ball milling can be improved, while a value of n = 3 is in accord with the volume theory which shows that the intensity of the impactive interactions can be increased. In this study, the ball wear mechanism for dry milling changed from the volume theory to the surface theory with an increase of the milling time, while the mechanism of wet milling was consistent with the volume theory irrespective of the milling time.

Microstructural characterization of hematite during wet and dry millings using Rietveld and XRD line profile analyses

The effects of extended milling in a stirred media mill and a tumbling mill on the structural changes in hematite have been examined using a combination of particle size analysis, BET surface areas, X-ray diffraction (XRD), thermal (TG and DSC) and FTIR measurements. Rietveld's whole profile fitting based on crystal structure refinement and the Warren–Averbach's method of X-ray line profile analysis were applied to monitor the microstructural evolution of the hematite phase. It is found that the BET surface area, X-ray amorphization phase content and XRD line breadths increase over the specific energy input. The use of stirred media mill exhibits larger surface areas, smaller particle sizes, more XRD line broadening and subsequently greater structural distortions compared to the tumbling mill for a given energy input; although the X-ray amorphous phase content remains unaffected by the grinding environments. The maximum X-ray amorphization degree of about 80 and 95% were calculated at a specific energy of 22,400 and 82,000 kJ/kg in the tumbling and stirred media mills respectively. The maximum specific BET surface area in the stirred media and tumbling milling increases to about 72.5 and 6.8 m 2/g after 82,000 and 22,400 kJ/kg energy consumption respectively. As a result of structural refinement during milling, the surface-weighted crystallite size in ground hematite are 17 and 4 nm after consuming 22,400 and 82,000 kJ/kg in the tumbling and stirred media mills respectively, corresponding to the volume-weighted crystallite size of 17 and 11 nm. For the same energy consumption in the mills, in turn, the root mean square strain, 〈 ɛ L = 10 nm 2 〉 1 / 2 , increases to about 4.4 × 10 − 3 and 4.9 × 10 − 3 . The results of the two applied methods are compared and discussed in details. In addition, thermogravimetric analysis of the wet ground samples reveals that the TG curves represent two weight loss steps. A weight loss is observed around 100 °C in the sample which is attributed to the removal of adsorbed water due to the wet milling operations. A strong weight loss step starting from 100 to 400 °C is attributed to surface/bulk dehydroxilation of iron hydroxides. The weight loss increases with extending of the milling. In contrast, the dry milled samples yield negligible weight losses comparing with the wet ground samples.

Study of the influence of liner wear on the load behaviour of an industrial dry tumbling mill using the Discrete Element Method (DEM)

Liner profiles influence load behaviors in tumbling mills. Generally, as liners are wearing, tumbling mills performance slightly increases then decreases till they are replaced preferably when the cost associated to the drop of performance is higher than the cost of relining or when liners reach a critical thickness susceptible to breakage. In this paper, we use the DEM to model the load behaviour of an industrial dry tumbling mill used for the grinding of coal at Eskom Kendal power station in South Africa. We study the influence of the liner wear and liner modification on the load behaviour in terms of position of shoulder and toe, power draw, and impact energy spectra. The results found improve our understanding on the grinding process in that particular mill as a function of liner wear and liner modifications. The results increase also our confidence in using the DEM as a tool to design tumbling mills since predicted results are in agreement with observation at the plant.

Production of Titanium Nitride by Carbothermic Reduction of the Anatase and Rutile Forms of Titanium Dioxide

Two types of titanium oxide were used, rutile (>99.9% TiO2) and anatase (>99% TiO2). Both samples were mixed with graphite in accordance with the required stoichiometry, and then milled in a tumbling mill for 50 hours in an argon atmosphere to ensure thorough mixing. The mill vial was loaded with five 25.4 mm diameter stainless steel balls giving a powder to ball mass ratio of 1:43. After milling, samples were heated to 1400°C in an alumina crucible at 20°C min-1 under a flowing nitrogen (100 mL min-1) atmosphere in a thermogravimetric analyser (TGA). Srilankite was detected in the as-milled anatase sample but the anatase to rutile transformation was not completed during milling. After heating to 800°C most of the anatase had transformed to rutile. Reduction of anatase started just below 900°C whilst rutile underwent reduction below 800°C. TGA results showed that the anatase reduction was more complex than the rutile reduction with several stages evident between 880 and 1000°C in the anatase sample whilst only two steps were observed for rutile. The initial identified products were Ti5O9 and Ti4O7 prior to TiN in anatase sample but in rutile sample only Ti4O7 was detected. Reduction was completed in rutile sample before 1180°C whilst in anatase completed at 1230°C. TiN was the final product in both systems after heating to 1400°C. These results are discussed in light of recent work demonstrating the different reductions paths of rutile and anatase.

INVESTIGATING THE DIFFERENCES IN CHARGE DYNAMICS DUE TO A VARIATION OF THE INSTRUMENTED BALL PROPERTIES

The validation of charge mill motion simulations is limited by the lack of information from inside a tumbling mill. To this end, an electronic system capable of measuring a number of physical quantities within a highly dynamic environment has been designed, built and tested. Once protected, the system can be used to record physical quantities within a ball mill. Different types of protective shells may be used to protect the electronics. When comparing the measurements recorded using two different shells, some differences in the behaviour of the instrumented ball were observed.

Experimental analysis of charge dynamics in tumbling mills by vibration signature technique

Up until now, real time identification of the dynamics of the charge in a tumbling mill has not been accomplished. This paper examines the possibility of correlating the vibration signature of tumbling mills to characterize the motion of the charge and the state of grinding. Vibration signals were picked up using accelerometers mounted directly on the mill shaft of a 90-cm diameter mill. The time domain signals were transformed to frequency domain by using fast Fourier transform (FFT). The Fourier spectra in the frequency domain were methodically interpreted and correlated to establish the prevailing mode of the charge motion under any operating condition. The grinding behavior under dry as well as wet grinding conditions were analyzed by following the variations in the vibration signature as a function of speed of the mill, volumetric filling, powder loading, and time of grinding. Experimental results clearly show that the dominant peak in the FFT spectra is quite sensitive to the variations in any mill operating parameter. This feature has been employed to detect undesirable operating conditions such as surging, mill over-load, etc. Finally, it is demonstrated that by proper interpretation of the vibration signature of the mill, it is possible to predict the charge dynamics and establish the state of grinding.

Comparative study of microstructural characteristics and stored energy of mechanically activated hematite in different grinding environments

Hematite concentrate was mechanically treated using different milling machines and experimental conditions in air atmosphere. The changes in phase constitution, particles size, specific surface area, lattice parameters and X-ray amorphous phase fraction of activated hematite were determined. It was found that the agglomeration of the particles take place during extended milling with accessible pores for Nitrogen gas. The higher media surface brought about the largest specific surface area whatever milling devices used. After 9 h of grinding with higher media surface, the maximum and minimum specific surface area resulted from the grinding in the tumbling and vibratory mills, accounting for 6.83 m 2/g and 18.42 m 2/g, respectively. For the same grinding condition, tumbling mill produced the lowest X-ray amorphous phase. The maximum X-ray amorphous material estimated around 85% from the grinding in the planetary mill with higher media surface for 9 h of milling. Structural changes were followed by XRD line broadening analysis (LPA) using the integral breadth method and Warren–Averbach approach. From the Williamson–Hall plots, it was understood that strain and size contributions exist simultaneously in the milled samples. Besides, the physical broadening increases as milling time and media surface increase regardless of milling types. Besides, it was found that hematite crystal is ‘soft’ between (024) and other crystallographic directions. From the Warren–Averbach approach, it was observed that the higher grinding media surface and prolonged milling favor the generation of small crystallite, higher microstrain, limited crystallite length and subsequently uniform activation of hematite. After 9 h of milling with higher media surface in tumbling, vibratory and planetary mills, the surface weighted crystallite size reached 17.3, 12.2 and 5.6 nm respectively. The maximum lattice strain, < ε L=10 nm 2> 1/2, in the grinding with tumbling, vibratory and planetary mills was found about 4.44 × 10 − 3 , 3.95 × 10 − 3 and 5.23 × 10 − 3 , respectively. The maximum dislocation density accounted for 46.3 × 10 14 m/m 3 in the planetary milling with higher media surface after 9 h of milling. The evaluation of energy contributions of structural defects suggested that the energy contribution of the amorphization was dominant and amounted to 92–98% of the overall stored energy in hematite, depending on milling conditions. Finally, for a given stress energy, the products of tumbling mill represent higher reactivity potential.

Determining the steel media abrasive wear as a function of applied force and friction

In the ongoing effort to develop a tumbling mill total steel media wear model, a decoupled abrasive, corrosive and impact wear model is used in conjunction with a mill DEM charge motion model. In order to determine the contribution of abrasive wear to total steel media wear in mills as a function of charge motion, it is necessary to determine abrasive wear as a function of applied force and slippage speed. To this end, a modified abrasion wheel testing apparatus was developed and used under dry conditions to determine the relationship between abrasive wear, applied force and friction. Four different ores from North American mines were examined where the test specimens used with each abrasive were cut from grinding media supplied by the corresponding mines. The results showed an initial expected increase in wear, as measured by the volume (mass) loss, with increasing applied force. The mass loss however became constant beyond a certain applied force. The friction remained relatively constant for each ore/media combination over the test range. Further abrasion wear research will include testing of ore under wet conditions.

A microscale look at tumbling mill scale-up using high fidelity simulation

Tumbling mill scale-up has had largely empirical roots. Energy size and energy size distribution relationships have been derived from relatively large data bases combined with a sprinkling of fundamental science. High fidelity simulation (HFS), which is based on multi-physics models, allows one to examine the interactions that occur in a mill amongst media, particles, wear parts and slurry at the microscale level. Thus, fundamental questions concerning particle capture, energy utilization and fragment creation as they depend on “equipment size” can be addressed. In this paper, the process is taken from the single particle breakage level, through laboratory grinding experiments up to full-scale commercial mill operation. The basis for certain empirical relationships is explained in fundamental terms. A look at the future of scale-up based on the use of such tools is presented.

Feature Extraction of a Strain-Gauge Signal Using Wavelets for Monitoring of a Tumbling Mill

Techniques that measure the force acting on a lifter bar, when it hits the charge inside a tumbling mill, have got an increased interest because of its direct physical relation to the behavior of the grinding charge. The possibility to combine it with discrete element modeling (DEM), which gives an opportunity to visualize the charge motion, opens new possibilities to understand the complex phenomena that takes place inside a grinding mill. In this work, a method for the measurement of the apparent filling level in a pilot ball mill, the Metso Continuous Charge Measurement system (CCM) has been used. The technique uses a strain gauge sensor, mounted on a flat steel spring, which is fitted into a recess underneath a lifter bar. Deflection of the lifter bar during every mill revolution will then give rise to a characteristic signal pattern depending on different operating conditions.In this work, which should be treated as an introduction, we show how the discrete wavelet transform can be used in multivariate calibration. It will be shown that by using the fast wavelet transform on individual signals as a pre-processing method in regression modeling on CCM measurements, good compression is achieved with almost no loss of information. The predictive ability and diagnostics of the data compressed regression model is almost the same as for the uncompressed.

Ball mill simulation in wet grinding using a tumbling mill and its correlation to grinding rate

A method for simulating the motion of balls in tumbling ball mill under wet condition is investigated. The simulation method is based on the three-dimensional discrete element method (DEM) and takes into account the effects of the presence of suspension, i.e., drag force and buoyancy. The impact energy on balls' collision, which enables us to provide useful information for predicting the grinding rate, was calculated from the simulation work. Actual ball motion in the suspension in the mill was observed through a transparent mill lid made of acrylic acid resin at different solid concentrations and the suspension was made of water with powder samples such as gibbsite, limestone and talc. Grinding tests of the gibbsite sample using two laboratory tumbling ball mills made of stainless steel and using alumina with balls made of steel, alumina and glass were conducted and the grinding rate was determined by measuring the 50% particle size of the product. The specific impact energy calculated from the simulation work correlates well with the grinding rate of the sample, irrespective of the grinding conditions, so that the energy is the key factor controlling the grinding phenomena in wet milling.

Contributions to the experimental validation of the discrete element method applied to tumbling mills

Accurate 3D experimental particle trajectory data, acquired from a laboratory tumbling mill using bi‐planar X‐ray filming, are used to validate the discrete element method (DEM). Novel numerical characterisation techniques are presented that provide a basis for comparing the experimental and simulated charge behaviour. These techniques are based on fundamental conservation principles, and provide robust, new interpretations of charge behaviour that are free of operator bias. Two‐ and three‐dimensional DEM simulations of the experimental tumbling mill are performed, and the relative merits of each discussed. The results indicate that in its current form DEM can simulate some of the salient features of the tumbling mill charge, however, comparison with the experiment indicate that the technique requires refinement to adequately simulate all aspects of the system.

Discrete element modelling of lifter stresses in tumbling mill

The discrete element code PFC3D has been used to model lifter stresses within a large tumbling mill. The intensity of the induced stresses (shear and normal) is directly proportional to the intensity of lifting action and liner/lifter wear. Results show that, for the modelled case, the magnitude of the stresses decreases as the number of lifters increase. Hence, longer intervals between relining can be expected for a mill with a larger number of lifters. However, it appears that beyond a critical number, a further increase in the number of lifters will not result in significant further reduction of stress. Distribution of the impact energy is also affected by the number of lifters. With very few lifters, the dominant form of energy consumption will be low intensity abrasion events. With larger numbers of lifters, high intensity impacts will be more frequent. Net power draw will be at a minimum for very small numbers of active lifters. As the number of lifters increases, power draw increases and eventually reaches a relatively stable value. Further work will be required to investigate the effect of lifter shape on the induced stresses. It appears to be possible to determine the number and shape of lifters that will result in the optimal power draw, while simultaneously minimising liner/lifter wear.