Minerals Engineering 98 (2016) 232–239

Contents lists available at ScienceDirect

Minerals Engineering

journal homepage: www.elsevier .com/ locate/mineng

Ceramic bead behavior in ultra fine grinding mills

http://dx.doi.org/10.1016/j.mineng.2016.08.0160892-6875/� 2016 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: paul.hassal@saint-gobain.com (P. Hassall).

Paul Hassall a,⇑, Emmanuel Nonnet b, Ville Keikkala c, Tarja Komminaho d, Liisa Kotila d

a Saint-Gobain Zirpro, United Kingdomb Saint-Gobain Zirpro, FrancecOutotec Oy, Finlandd FQM Kevitsa Mining Oy, Finland

a r t i c l e i n f o

Article history:Received 30 May 2016Revised 29 July 2016Accepted 22 August 2016

Keywords:CeramicMedia

a b s t r a c t

A preliminary review of the developing morphology of ceramic beads during extended operations in highenergy stirred mills is presented.Ceramic beads are produced from various formulations and processes; these certainly affect the cost of

the media; but significantly, they impact the competence and performance.Mills operate with a ‘Working-Mix’ or ‘Seasoned-Charge’; invariably this means that worn beads of var-

ious sizes are present in the mill.These beads can have effect on ‘mill efficiency’, ‘overall media wear’ and particularly ‘mill wear’.This study reports preliminary findings of wear profiles of typical ceramic media and some of the

potential effects of extended use.� 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2322. Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233

2.1. Available beads type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332.2. Available ultrafine milling equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

3. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

3.1. Bead wear issue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2343.2. Effects of bead wear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2353.3. A case study: HIG mill installation at FQM Kevitsa Finland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236

3.3.1. Process presentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2363.3.2. Initial results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2373.3.3. Media considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238

4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

1. Introduction

Saint-Gobain Zirpro has been instrumental in the developmentof ceramic media for ultra-fine grinding applications in stirredmills. The history and experience are long, dating back to themid nineteen seventies. The company was in-fact established in1971 to refine Zirconium Oxide to meet the Saint-Gobain Glassrequirements for high performance refractories. The history of

the Glass division is somewhat longer and it celebrated its threehundred and fiftieth anniversary last year.

Zirconia (Zirconium Oxide) and Zircon (Zirconium Silicate)turned out to be significant raw materials necessary to manufac-ture high quality ceramic beads. Initially beads were produced bya fusion process; the operation required high temperatures andthe beads were formed in a molten state. The resulting product(ER120) was based on Zircon and had a density of 4.0 g/cc. Thebeads were round and non-abrasive and were ideal to replacethe glass and natural sand products used in the burgeoning stirredor bead mill applications. A comparative increase in density of over

P. Hassall et al. /Minerals Engineering 98 (2016) 232–239 233

50% was the important factor; greatly enhancing the productivityof the mills. Higher density media, such as steel (7.5 g/cc) forexample, were largely discounted due to increased abrasion ratesand contamination. Therefore ceramic technology was readilyadopted by major industries, including pigments, paints and agro-chemicals. The products proved to be extremely successful andremained at the forefront of the technology for over twenty years.Increased demands on milling technology were however inevitableand eventually faster, finer, more precise targets were expected.The industry responded with the development of new mill designswhich operated at higher speeds, with higher energy and higherthroughput rates. A new type of bead was required which wouldbe tough and could withstand the new operating conditions. Theresult was the evolution of sintered beads, initially based on thesame chemistry and having the same density as the fused products.These beads required low temperature forming before densifica-tion (sintering) at high temperature. The products proved to besuitable in many of the new applications, providing tougher beadswith extended bead lifetimes. Zirpro again developed a class lead-ing product (RIMAX) which gave exemplary performance in manyvaried fields for example cosmetics, inks and automotive coatings.Today the evolution has continued with the general acceptance ofhigh density (6.0 g/cc) stabilized Zirconia beads as the media ofchoice (Hassall and Nonnet, 2007). The higher density providesthe potential for superior and economic grinding and although ini-tially expensive, the controlled wear provides an overall cost effec-tive solution. Zirpro developed a premier product ‘ZIRMIL’ nowwidely adopted and widely used in the processing of the mostdemanding applications, such as pharmaceuticals and ceramics.

Mill design has also evolved to meet the ever increasingdemands for ever more specialized materials and applications. Atthe forefront of these developments are two extremely differentprojects; the first is nano grinding of electronic materials and thesecond the ultra-fine grinding of ore bodies in the mining industry.For nano grinding, beads sizes of approximately 100 lm arerequired and potentially bead densities increasing beyond of10 g/cc. For mining the environment is severe and beads mustwithstand high impacts from hard and large feed materials indilute slurries.

In the nano application Zirpro has launched a derivative of theZIRMIL range. Advanced ceramic technology and process engineer-ing have enabled the production of 100 and 200 lm beads inindustrial quantities at economic price levels. The material is cur-rently under evaluation in extended customer trials. For mining acomposite material MINERAX has been developed and successfullylaunched into the industry (Fig. 1). It is a tough composite material

Fig. 1. MINERAX ceramic beads (75% Alumina 25% Zircon).

with a classic density of 3.9 g/cc and fully competent in all ultra-fine mining applications.

2. Materials and methods

2.1. Available beads type

Minerax is one of a number of ceramic beads available to themining industry; many ceramic producers have launched productsto meet the industry’s high volume demand. The range of chemis-tries, formulations and technologies employed are rather broad. In-fact the only consistent factor is probably the size ranges offered;usually 2–3, 3–4, 4–5 and 5–6 mm. Although the formulationscan vary significantly, fundamentally they are all based on threebasic oxides, Silicon (SiO2), Aluminum (Al2O3) and Zirconium(ZrO2); Silica, Alumina and Zirconia respectively. Non on the mediaare based on the pure forms, either due to suitability, ease of pro-cessing or cost. For example pure Silica bead is available but not atall suitable for the application. Pure Alumina product is possiblebut is difficult to process and would be prohibitively expensive.Zirconia is a special case where a stabilizing element is requiredto fix the material in its high temperature form. In the case of ZIR-MIL, the addition of 5% Yttria produces a premier grade of product,very high performance but with a rather high cost. An alternativeto Yttria is Ceria and similar stabilization can be achieved at a moreaffordable rate. Zirpro have developed an additional product forthe mining industry ‘CERMIL’. It finds application in the miningindustry in specific mill types where high bead densities arerequired.

The general range of media for the mining industry is actuallymore affected by the choice of available and cost effective rawmaterials; for example kaolin, Alumina and Zircon. From thesethe majority of formulations can be achieved. The raw materialchoice can also affect subsequent processing possibilities as thenatural plasticity of kaolin can be exploited; alternatively addi-tional organic plasticizers need to be added. The range of formula-tions and chemistry has significant effect on the nature andbehavior of the bead. Density is the primary consideration(Becker et al., 2001) with products ranging from a low of 2.7 g/ccto a high of 6.2 g/cc. predominantly, however a mid-range of den-sities is preferred, in the range of 3.6–4.0 g/cc; this seems to suitthe majority of mining applications. The physical properties ofthe beads are also linked to original formulation; factors such ashardness, toughness and strength are fundamentally affected. Apoint of consideration here might be an analogy with a ‘brick wall’;the individual grains of the bead (the bricks) have extremely goodphysical characteristic but the actual performance is more depen-dent on the binding agent (the mortar); it is often the consistencyand make-up of grain boundaries of a bead which dictate itsperformance.

The processing of ceramic materials these days is more of anadvanced engineering operation, but there certainly remains anelement of the artisan or craftsman. An example of this is the rela-tionship between the forming process and the final behavior andproperties of the product. Ceramics are said to have a ‘memory’and this refers to the fact that forces and stresses experienced informing become manifest in the final structure of the product(Ford, 1967). For beads there are two main forming processes;dripping and granulation. The dripping process involves forminga plasticized paste which is extruded through fine forming nozzle.The process provides a high degree of homogeneity in the formedbead. This is important in subsequent firing procedures as theuniformity assist in grain growth and densification at hightemperature. The main drawback of the process is the lack of uni-formity in the final shape, beads often tend to be ‘tear drop’ shapedwhich is a potential problem when the bead is in use. The other

234 P. Hassall et al. /Minerals Engineering 98 (2016) 232–239

process is granulation; here a seed is progressively coated with aplasticized mixture of the raw materials. The process eliminatesthe issue of the droplet shape and tends to produce overall veryround beads. The main issue is control of process and the ‘memory’factor. If the layers are built too quickly then internal stress candevelop; these become exacerbated during firing and can resultin delamination (see Figs. 2 and 3). Both processes have theirown merits, fundamental to both is the management of raw mate-rials (particularly grain size) and tight production and quality con-trol procedures. The final stage in the manufacture of a bead is the‘firing’ or ‘densification’ step (Kingery et al., 1991). This too is crit-ical and its optimization is integral to the successful performanceof the bead. The firing must be controlled and accurate, suited tothe specific materials employed. The old adage to improve perfor-mance, simply ‘fire at higher temperature’ does not work for beads.At the temperatures and firing cycles most commonly used (1250–1450 �C) firing higher for a particular formula potentially wouldresult in high glass formation and tendency for the bead to beprone to spalling. There are obviously many options to producebeads but there are only two final criteria for their design; theymust be ultimately competent and provide economic operation.

2.2. Available ultrafine milling equipment

The mining industry has a requirement to treat fine feed mate-rials (50–250 lm), a target difficult or ineffective for traditionalmilling processes such as SAG or Ball milling. Even formoremodern

Fig. 2. Granulated bead cross section with poor process control.

Fig. 3. Dripped beads with poor shape control.

machinery such as High Pressure Grinding Rolls (HPGR) it is at theextreme of their capabilities. The favored route has been the use of‘stirred’ or ‘agitated’ bead mills. Just as there is a great variation inthe beads available, there is also a considerable choice inmachineryfor ultra-fine grinding in mining applications. There are four mainequipment’s available as follows, Stirred Mill Detritors (SMD man-ufactured byMetsoMinerals), Isamills (manufactured by Glencore),VXPmill (manufactured by FLSMIDTH) and HIGmill (manufacturedby Outotec). Essentially the fundamentality’s are the same; there isa vessel and an internal stirring mechanism designed to agitate thebeads. The stirred beads are accelerated to a level where inter beadcollisions have sufficient energy to break the target slurry particlesand thus affect grind. Practically this operational feature is poten-tially all that the mills have in common. Mills can have a verticalor horizontal configuration. They can use discs or stirring bars.The size of the mill can less than 100 L or reach as high as45,000 L. Power can range from a few Kilo-Watts to a few Mega-Watts. The speed of the agitators can be as low as 4 m/s and as highas 22 m/s. The power profile (Watts per unit mill volume) can varysignificantly as can the throughput profile (liters flow per unit millvolume). All these factors have a profound effect on the perfor-mance and efficiency of the bead employed. Certainly it is not a caseof one size (bead) fits all! Bead selection needs to be appropriate tothe application and to the mill type proposed. Optimization of beadis a serious question and should be determined by laboratory andpilot testing prior to implementation. In-correct choice can affectthe overall efficiency and potentially the viability of the process.Beads are an important factor andmust be closelymatched to oper-ating conditions (Curry and Clermont, 2005).

3. Results and discussion

3.1. Bead wear issue

The other effect of optimizing bead would be to provide con-trolled bead wear profiles. Initially beads are selected to fit purposeand provide economic grind. However the media will wear and theway in which they wear will have impact on the process efficiency.Firstly if the conditions in the mill are imbalanced then severe beadwear can result. In such circumstances we can expect beads to de-laminate, split, crush and spall. Some example of beads after severewear are giving in Figs. 4–6. All are extremely undesirable resultingin rapid loss of mill efficiency and increased mill wear (Kotze,2012).

Less severe but equally detrimental is the generation of flat orfacetted beads in the mill charge (Figs. 7 and 8). These

Fig. 4. Delaminated bead.

0

10

20

30

40

50

60

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4

Wei

ght %

Bead size (mm)

1.6-2.0 mm new beads

After 1000h

Fig. 9. Bead size evolution due to wear. Example of mineral application.

Fig. 7. Flat beads.

Fig. 6. Split bead.

Fig. 5. Spalled bead. Fig. 8. Facetted bead.

P. Hassall et al. /Minerals Engineering 98 (2016) 232–239 235

phenomenon results from preferential wear of the media. This typeof wear is somewhat due to the initial shape of the bead but alsodue in great part to the operational conditions of the mill. It isgreatly exacerbated by limited or restrictive bead flow in the mill‘dead-spots’. These ‘dead-spots’ can be due to hydraulic packingor simply due to less agitated parts of the mill chamber. The wornbeads generated tend to remain in the mill, but their contribution

to grinding is extremely limited. In severe cases the efficiency ofthe mill is absolutely compromised.

Not with-standing these dramatic forms of wear, good beads ingood mill conditions will also ultimately wear (Gallimore, 2010).

3.2. Effects of bead wear

Mills operate with a ‘working mix’ or ‘seasoned charge’. It israther this ‘working mix’ rather than the initial bead size whichdetermines mill performance. Original bead size selection musttake this factor into consideration and provide a suitable ‘workinglife’ of bead. This ‘working life’ refers to beads which contribute togrinding. Once beads fall below a certain size they cannot be accel-erated sufficiently and cannot affect grind. These beads merelydilute the grind and absorb energy. In many industries where con-tinuous mill operations are standard it is estimated that up to 30%of input energy can be lost by generating high levels of media fines.As an example size distribution of initial beads are compared tobeads size after 1000 h for a mineral milling application in Fig. 9.This shows that only approximately 25% of the used beads remainin the initial media size range; greatly compromising overallefficiency.

The phenomenon cannot be overcome, it is a factor present inall mills in all industries. The only recourse is to control the issueas much as possible. In some applications it is possible to replacethe entire charge periodically. The opportunity for this is indeedrare and certainly not applicable in the mining industry. The onlyalternative is to use more durable beads with extended medialifetimes. The effect here is to reduce the speed at which finesare generated and consequently maintain a higher proportion ofcontributing beads for as long as possible. Consistently the

236 P. Hassall et al. /Minerals Engineering 98 (2016) 232–239

‘working mix’ for more durable beads is proportionately larger insize and achieved grinds significantly superior (Hassall, 2008).

3.3. A case study: HIG mill installation at FQM Kevitsa Finland

3.3.1. Process presentationThe FQM Kevitsa, Nickel-Copper-PGE mineral deposit is located

in Finnish Lapland. Production at the mine started in summer2012. The site produces copper and nickel concentrates with gold,platinum, palladium and cobalt by-products. Primary and

Fig. 10. Kevitsa flotation circuit flow sheet.

Fig. 11. Basic flowsheet for Outote

secondary mill cyclone overflows are selectively floated in threephases as illustrated in Fig. 10. In recent times, higher millthroughputs with coarser grades highlighted the need to enhanceCu regrinding to improve mineral liberation (Cu-Ni separation).This requirement led to the installation in February 2015, of aHIG700 mill to Cu flotation circuit. The HIGmill feed is copperrougher scavenger concentrate and product guided to first coppercleaning phase.

Cyclone underflows report to a HIG feed tank and the overflowgoes to the HIG mill product tank. The feed slurry is pumped

HIG mill is circled (Lehto et al., 2016).

c HIGmill (Lehto et al., 2016).

Fig. 12. Grinding philosophy of the Outotec HIGmill (Lehto et al., 2016).

0

20

40

60

80

100

120

0 20 40 60 80 100 120

Pass

ing

%

Par�cle Size (μm)

FeedProduct

Fig. 13. Feed and product particle size distribution after HIG mill in Kevista mineoperations.

Fig. 15. Media general state after six weeks commissioning.

P. Hassall et al. /Minerals Engineering 98 (2016) 232–239 237

through the mill and overflows into the product tank. Samplers(PSI samplers – Outotec) are located before and after the HIGmill.A PSI500 analyzer (online particle size analyzer – Outotec) is con-nected to these samplers and it monitors continually the feed andproduct sizes (Fig. 11).

The HIG mill is configured with stationary discs on the liner androtating discs attached to the drive shaft (Fig. 12). Slurry is pumped

0

10

20

30

40

50

60

<25 25-45 45-75

Dist

ribut

ion

%

Particle size (μm)

Fig. 14. Cu distribution and grade in Copper

through the feed inlet, located at the bottom of the mill. Centrifugalaction forces coarse particles and media to the outer shell. Gravitykeeps the media compact during the operation and most of thegrinding is done by attrition. All particles have to pass throughthe entirety of the mill and face multiple grinding stages.

HIGmill operation is being controlled by a preset value of theSpecific Grinding Energy (SGE) related to dry tons. Flow and den-sity meters measure the slurry characteristics in real time. The millspeed is constantly being changed with the VSD to meet the chang-ing flow. All HIGmills are equipped with a VSD to provide controlto the operation and to avoid overgrinding.

Minerax grinding media is introduced into the mill from a feedtank via the HIGmill feed pump. A screw feeder adds media onceevery hour, depending on the kWh used in the mill. This ensuresthat the grinding media level remains constant and that the bestpossible grind efficiency is maintained.

Ore feed size to the mill (F80) vary between 80 and 120 lm;however despite the large variance, P80 can be maintained at thedesired level due to the adaptable and controlled mill operation.Currently the P80 is at 35 lm. Installation of the HIGmill hasresulted in significant increases in tonnages processed throughthe primary and secondary grinding stages while always retainingthe correct particle size for flotation.

3.3.2. Initial resultsHIGmill feed and product particle size distributions are shown

in Fig. 13. Achieved grind (P80) was 32 lm compared to target at30–35 lm. SGE of 13–15 kW/t was used to control the grind.

>75

Grad

e %

Size fraction Before

Size fraction After

Cu Grade Before

Cu Grade After

Concentrate before and after HIG mill.

Type Amount (Wt%)Spherical beads 91%Broken beads 0.4%Facetted beads 8.9%

Fig. 17. Bead shape and condition after six months operation.

0%

5%

10%

15%

20%

25%

30%

Wei

ght %

Top mill Middle mill Bottom mill

Fig. 16. Media size distribution and general state after six months operation.

238 P. Hassall et al. /Minerals Engineering 98 (2016) 232–239

Copper grade and total Cu recovery increased after HIG millinstallation as presented by Lehto et al. elsewhere (Lehto et al.,2016). Benefits of improved Cu/Ni separation are illustrated byan increase in Cu deported to all the size fractions below 75 lm(floatable size range), see Fig. 14. Correspondingly the Cu gradesin the detailed size fractions also increased.

3.3.3. Media considerationsCommissioning of any new equipment is intrinsically compli-

cated; with various parameters being defined and numerous oper-ational conditions experienced. It is a particularly arduous forstirred mills in terms of the milling media. Start/stop operations,with extremes of speed and throughput can have a detrimentaleffect on the charge. It was therefore important to check the beadcondition immediately after the commissioning period. An auditwas carried out specifically looking for damaged beads and anyexcessive media wear. The basic results are detailed in Figs. 15and 16.

The main observation was that the working mix was alreadyestablished, with the majority of the bead falling into a 2.5–3.15 mm range. The bead significantly remained round and pol-ished, with no broken or severe abrasion evident. Beads were col-lected from both the top and bottom of the mill; the wear profilefrom each was seen to be controlled with no perceived issues toreport. There were no fines observed at this stage, potentially asthe running time of the mill was limited.

The next audit was taken after a period of 6 months where amore ‘steady state’ of mill operation had prevailed. Again sampleswere taken from various locations in the mill. The results aredetailed in Fig. 16.

Again the working mix profile was well detailed and remainedconsistent across the mill. The charge remained predominantly in

the 2.5–3.2 mm range, although there was an increased proportionin the 2.0–2.2 mm sizing. A very high proportion of the mediaremained round and polished, providing good media flow andrestricting mill wear. An increase in damaged beads was observedas detailed in Fig. 17.

The only identifiable issue was facetted beads; which wereapproaching 10% of the charge. The limited severity of the facetsand the relatively low levels; were not expected to have any detri-mental effect on continued operation or ultimately mill efficiency.In-fact no perceivable decline in achieved grind or efficiency wereevident.

Again there were few fine beads observed below 1.8 mm, andafter 6 months there should have been sufficient time for the beadprofile develop fully. The phenomenon is not yet explained as therewere no broken beads evident within the mill and no solid beadsescaping the mill. The issue needs further investigation.

The mill has continued to run successfully. Further audits willhelp to develop a broader understanding of the evolution of the‘working mix’ and its long term effects on HIGmill operations.

4. Conclusion

The observed results are extremely encouraging; reflecting theaccurate modelling and testing carried out prior to installation.The mill achieved target grind and provided an efficient and costeffective process. The correct choice of bead type and bead sizewas a fundamental contributor to the level of success. This opti-mization in media selection provided a competent, problem freeand assured mill operation. Ultimately the bead wear was con-trolled and delivered a working mix which was efficient in grindingtarget particles. There were no observed media issues and conse-quently mill wear was also minimized.

P. Hassall et al. /Minerals Engineering 98 (2016) 232–239 239

References

Becker, M., Kwade, A., Schwedes, J., 2001. Stress intensity in stirred media mills andits effect on specific energy requirement. Int. J. Miner. Process. 61, 189–208.

Curry, D.C., Clermont, B., 2005. Improving the efficiency of fine grinding—developments in ceramic media technology. Paper Presented at the RandolConference, 2005.

Ford, W.F., 1967. The Effect of Heat on Ceramics. Maclaren, London.Gallimore, M., 2010. Fluidized mill media selection considerations. Paper Presented

in Comminution ’10, Cape Town, 2010.Hassall, P., 2008. Milling Media Review: Part 1. Paints and Coatings Industry, April 1,

2008.

Hassall, P., Nonnet, E., 2007. Optimized media solution for horizontal mills. In:Grinding and Dispersing with Stirred Media Mills, Research and Application 5.Colloquium, Braunschweig, 24–25 October 2007.

Kotze, H., 2012. Selecting ceramic media: the theory. Paper Presented inComminution ’12, Cape Town, 2012.

Kingery, W.D., Bowen, H.K., Uhlmann, D.R., 1991. Introduction to Ceramics, seconded. J. Wiley, Singapore.

Lehto, H., Musku, B., Keikkala, V., Kurki, P., Paz, A., 2016. Developments in stirredmedia milling testwork and industrial scale performance of Outotec Higmill.Paper Presented in Comminution ’16, Cape Town, 2016.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具