A true ball mill is a porcelain jar partly filled with spherical or rounded cylindrical porcelain balls. Industrial versions are made of metal and have porcelain linings. Small scale operations most commonly employ ball mills for grinding glazes. The suspension is poured in, a lid secured, and it is rotated on a motorized rack, sometimes for many hours. The tumbling balls within grind particles smaller and smaller as they impact each other (and crush particles that happen to be at the points of contact). The creamier glaze that milling produces applies better, has more stable viscosity, fires more consistently and cleaner with less specks and imperfections (eg. pinholes and blisters), and melts better. Glazes can be overmilled, this can produce solubility, crawling, opacification and slurry issues (since certain materials in the glaze need to be kept above a certain particle size to behave correctly).
Potters and hobbyists are generally not aware of the importance of the ball mill to industrial ceramic ware production. For a small-scale stoneware operation it is possible to survive without one using a narrow range of glazes and techniques. But when production is ramped up consistency, reliability of the glaze appearance and defect free ware become paramount. Many materials in ceramics are simply not ground fine enough for glazes (they produce fired specks or defects related to expulsion of gases around larger particles); ball, native and slip clays are an example. In other materialsfine particles agglomerate into larger ones (e.g. barium carbonate, tin oxide, wollastonite). Others are supplied as a grain-type material rather than a powder and obviously have to be milled (eg. lithium carbonate, alumina hydrate). Engobes that must be sprayed, sink screened or even inkjet printed must be ball milled or nozzles will clog and screen will blind. Obviously, bottled engobes and glazes that potter's buy are ball milled when produced.
Amazingly, many industries routinely grind their body materials in ball mills (e.g. the insulator and even tile industries). One Kalemaden plant we visited in Canakkale, Turkey (one of the largest in the world) airfloats and mills local clays for all their products. They even collect their own flint rocks and break and mill them to round. Companies may be seeking residues of less than 0.1% on 325 mesh. Other benefits also ensue, including more plasticity, better fired maturity and strength. The benefits are not only very high quality and defect defect-free products, but better consistency. Typically a slurry of 65% clay and 35% water is made (only possible if deflocculated) and ball milled, then dewatered (using filter presses, spray driers, etc) to make powder or pellets. In addition, materials will melt or go into solution in the melting glaze significantly better or sooner if they are ground finer.
A small mill rack is $700-1300 US. However you can build your own (see the links here). href="https://www.digitalfire.com/gerstleyborate/ballmill/">https://www.digitalfire.com/gerstleyborate/ballmill/ Or you google the booklet "Thoroughly Modern Milling" by Steve Harrison (it is intended to assist the potter in building a ball mill with a roller mechanism to handle a jar in the 3 to 5 gallon range). The text describes how to assemble the parts illustrated in the detail drawings and briefly describes making your own jar and ball from porcelain clay body. A4 size, 6 pages of text and 6 x A3 pages of detail mechanical drawings. There is one color photograph.
If you are using a ball mill in your operation resist the temptation to think that using one is just a matter of throwing in some pebbles, pouring in the glaze, and turning it on for an hour or so.As a general rule you should mill for the same amount of time, fill the jar to the same level, use the same charge of pebbles and the range of sizes of the pebbles should be controlled (the pebbles wear down over time). There are many finer pointsto know about using ball mills and industry uses the term "mill practice" to embody them. Variation caused by poor mill practice can create a number of significant fired glaze faults and affect slurry and application properties. To learn more check the book 'Ceramics Glaze Technology'. You should be able to find a copy at one of the used ceramic book vendors or information online.
The last thing you want is for the steel balls you put into your ball mills to crumble, like stale cookies, on impact. You need to maintain peak pulverizing efficiency at all times. You deserve nothing less. And FOX has come up with your ball milling solution.
Each FOX Forged Steel grinding ball is solid from surface to center. This is a forged steel ball that is through hardened for superior strength. These sturdy, well built, shock-absorbing steel balls are the perfect hardness for ball mill grinding. Whether you are grinding raw materials in ball mills, or using planetary ball mills on a smaller scale to get a finer particle size, these are the forged steel balls for you.
A ball mill is a type of grinder used to grind and blend bulk material into QDs/nanosize using different sized balls. The working principle is simple; impact and attrition size reduction take place as the ball drops from near the top of a rotating hollow cylindrical shell. The nanostructure size can be varied by varying the number and size of balls, the material used for the balls, the material used for the surface of the cylinder, the rotation speed, and the choice of material to be milled. Ball mills are commonly used for crushing and grinding the materials into an extremely fine form. The ball mill contains a hollow cylindrical shell that rotates about its axis. This cylinder is filled with balls that are made of stainless steel or rubber to the material contained in it. Ball mills are classified as attritor, horizontal, planetary, high energy, or shaker.
Grinding elements in ball mills travel at different velocities. Therefore, collision force, direction and kinetic energy between two or more elements vary greatly within the ball charge. Frictional wear or rubbing forces act on the particles, as well as collision energy. These forces are derived from the rotational motion of the balls and movement of particles within the mill and contact zones of colliding balls.
By rotation of the mill body, due to friction between mill wall and balls, the latter rise in the direction of rotation till a helix angle does not exceed the angle of repose, whereupon, the balls roll down. Increasing of rotation rate leads to growth of the centrifugal force and the helix angle increases, correspondingly, till the component of weight strength of balls become larger than the centrifugal force. From this moment the balls are beginning to fall down, describing during falling certain parabolic curves (Figure 2.7). With the further increase of rotation rate, the centrifugal force may become so large that balls will turn together with the mill body without falling down. The critical speed n (rpm) when the balls are attached to the wall due to centrifugation:
where Dm is the mill diameter in meters. The optimum rotational speed is usually set at 6580% of the critical speed. These data are approximate and may not be valid for metal particles that tend to agglomerate by welding.
The degree of filling the mill with balls also influences productivity of the mill and milling efficiency. With excessive filling, the rising balls collide with falling ones. Generally, filling the mill by balls must not exceed 3035% of its volume.
The mill productivity also depends on many other factors: physical-chemical properties of feed material, filling of the mill by balls and their sizes, armor surface shape, speed of rotation, milling fineness and timely moving off of ground product.
where b.ap is the apparent density of the balls; l is the degree of filling of the mill by balls; n is revolutions per minute; 1, and 2 are coefficients of efficiency of electric engine and drive, respectively.
A feature of ball mills is their high specific energy consumption; a mill filled with balls, working idle, consumes approximately as much energy as at full-scale capacity, i.e. during grinding of material. Therefore, it is most disadvantageous to use a ball mill at less than full capacity.
Grinding elements in ball mills travel at different velocities. Therefore, collision force, direction, and kinetic energy between two or more elements vary greatly within the ball charge. Frictional wear or rubbing forces act on the particles as well as collision energy. These forces are derived from the rotational motion of the balls and the movement of particles within the mill and contact zones of colliding balls.
By the rotation of the mill body, due to friction between the mill wall and balls, the latter rise in the direction of rotation until a helix angle does not exceed the angle of repose, whereupon the balls roll down. Increasing the rotation rate leads to the growth of the centrifugal force and the helix angle increases, correspondingly, until the component of the weight strength of balls becomes larger than the centrifugal force. From this moment, the balls are beginning to fall down, describing certain parabolic curves during the fall (Fig. 2.10).
With the further increase of rotation rate, the centrifugal force may become so large that balls will turn together with the mill body without falling down. The critical speed n (rpm) when the balls remain attached to the wall with the aid of centrifugal force is:
where Dm is the mill diameter in meters. The optimum rotational speed is usually set at 65%80% of the critical speed. These data are approximate and may not be valid for metal particles that tend to agglomerate by welding.
where db.max is the maximum size of the feed (mm), is the compression strength (MPa), E is the modulus of elasticity (MPa), b is the density of material of balls (kg/m3), and D is the inner diameter of the mill body (m).
The degree of filling the mill with balls also influences the productivity of the mill and milling efficiency. With excessive filling, the rising balls collide with falling ones. Generally, filling the mill by balls must not exceed 30%35% of its volume.
The productivity of ball mills depends on the drum diameter and the relation of drum diameter and length. The optimum ratio between length L and diameter D, L:D, is usually accepted in the range 1.561.64. The mill productivity also depends on many other factors, including the physical-chemical properties of the feed material, the filling of the mill by balls and their sizes, the armor surface shape, the speed of rotation, the milling fineness, and the timely moving off of the ground product.
where D is the drum diameter, L is the drum length, b.ap is the apparent density of the balls, is the degree of filling of the mill by balls, n is the revolutions per minute, and 1, and 2 are coefficients of efficiency of electric engine and drive, respectively.
A feature of ball mills is their high specific energy consumption. A mill filled with balls, working idle, consumes approximately as much energy as at full-scale capacity, that is, during the grinding of material. Therefore, it is most disadvantageous to use a ball mill at less than full capacity.
Milling time in tumbler mills is longer to accomplish the same level of blending achieved in the attrition or vibratory mill, but the overall productivity is substantially greater. Tumbler mills usually are used to pulverize or flake metals, using a grinding aid or lubricant to prevent cold welding agglomeration and to minimize oxidation .
Cylindrical Ball Mills differ usually in steel drum design (Fig. 2.11), which is lined inside by armor slabs that have dissimilar sizes and form a rough inside surface. Due to such juts, the impact force of falling balls is strengthened. The initial material is fed into the mill by a screw feeder located in a hollow trunnion; the ground product is discharged through the opposite hollow trunnion.
Cylindrical screen ball mills have a drum with spiral curved plates with longitudinal slits between them. The ground product passes into these slits and then through a cylindrical sieve and is discharged via the unloading funnel of the mill body.
Conical Ball Mills differ in mill body construction, which is composed of two cones and a short cylindrical part located between them (Fig. 2.12). Such a ball mill body is expedient because efficiency is appreciably increased. Peripheral velocity along the conical drum scales down in the direction from the cylindrical part to the discharge outlet; the helix angle of balls is decreased and, consequently, so is their kinetic energy. The size of the disintegrated particles also decreases as the discharge outlet is approached and the energy used decreases. In a conical mill, most big balls take up a position in the deeper, cylindrical part of the body; thus, the size of the balls scales down in the direction of the discharge outlet.
For emptying, the conical mill is installed with a slope from bearing to one. In wet grinding, emptying is realized by the decantation principle, that is, by means of unloading through one of two trunnions.
With dry grinding, these mills often work in a closed cycle. A scheme of the conical ball mill supplied with an air separator is shown in Fig. 2.13. Air is fed to the mill by means of a fan. Carried off by air currents, the product arrives at the air separator, from which the coarse particles are returned by gravity via a tube into the mill. The finished product is trapped in a cyclone while the air is returned in the fan.
The ball mill is a tumbling mill that uses steel balls as the grinding media. The length of the cylindrical shell is usually 11.5 times the shell diameter (Figure 8.11). The feed can be dry, with less than 3% moisture to minimize ball coating, or slurry containing 2040% water by weight. Ball mills are employed in either primary or secondary grinding applications. In primary applications, they receive their feed from crushers, and in secondary applications, they receive their feed from rod mills, AG mills, or SAG mills.
Ball mills are filled up to 40% with steel balls (with 3080mm diameter), which effectively grind the ore. The material that is to be ground fills the voids between the balls. The tumbling balls capture the particles in ball/ball or ball/liner events and load them to the point of fracture.
When hard pebbles rather than steel balls are used for the grinding media, the mills are known as pebble mills. As mentioned earlier, pebble mills are widely used in the North American taconite iron ore operations. Since the weight of pebbles per unit volume is 3555% of that of steel balls, and as the power input is directly proportional to the volume weight of the grinding medium, the power input and capacity of pebble mills are correspondingly lower. Thus, in a given grinding circuit, for a certain feed rate, a pebble mill would be much larger than a ball mill, with correspondingly a higher capital cost. However, the increase in capital cost is justified economically by a reduction in operating cost attributed to the elimination of steel grinding media.
In general, ball mills can be operated either wet or dry and are capable of producing products in the order of 100m. This represents reduction ratios of as great as 100. Very large tonnages can be ground with these ball mills because they are very effective material handling devices. Ball mills are rated by power rather than capacity. Today, the largest ball mill in operation is 8.53m diameter and 13.41m long with a corresponding motor power of 22MW (Toromocho, private communications).
Modern ball mills consist of two chambers separated by a diaphragm. In the first chamber the steel-alloy balls (also described as charge balls or media) are about 90mm diameter. The mill liners are designed to lift the media as the mill rotates, so the comminution process in the first chamber is dominated by crushing. In the second chamber the ball diameters are of smaller diameter, between 60 and 15mm. In this chamber the lining is typically a classifying lining which sorts the media so that ball size reduces towards the discharge end of the mill. Here, comminution takes place in the rolling point-contact zone between each charge ball. An example of a two chamber ball mill is illustrated in Fig. 2.22.15
Much of the energy consumed by a ball mill generates heat. Water is injected into the second chamber of the mill to provide evaporative cooling. Air flow through the mill is one medium for cement transport but also removes water vapour and makes some contribution to cooling.
Grinding is an energy intensive process and grinding more finely than necessary wastes energy. Cement consists of clinker, gypsum and other components mostly more easily ground than clinker. To minimise over-grinding modern ball mills are fitted with dynamic separators (otherwise described as classifiers or more simply as separators). The working principle is that cement is removed from the mill before over-grinding has taken place. The cement is then separated into a fine fraction, which meets finished product requirements, and a coarse fraction which is returned to mill inlet. Recirculation factor, that is, the ratio of mill throughput to fresh feed is up to three. Beyond this, efficiency gains are minimal.
For more than 50years vertical mills have been the mill of choice for grinding raw materials into raw meal. More recently they have become widely used for cement production. They have lower specific energy consumption than ball mills and the separator, as in raw mills, is integral with the mill body.
In the Loesche mill, Fig. 2.23,16 two pairs of rollers are used. In each pair the first, smaller diameter, roller stabilises the bed prior to grinding which takes place under the larger roller. Manufacturers use different technologies for bed stabilisation.
Comminution in ball mills and vertical mills differs fundamentally. In a ball mill, size reduction takes place by impact and attrition. In a vertical mill the bed of material is subject to such a high pressure that individual particles within the bed are fractured, even though the particles are very much smaller than the bed thickness.
Early issues with vertical mills, such as narrower PSD and modified cement hydration characteristics compared with ball mills, have been resolved. One modification has been to install a hot gas generator so the gas temperature is high enough to partially dehydrate the gypsum.
For many decades the two-compartment ball mill in closed circuit with a high-efficiency separator has been the mill of choice. In the last decade vertical mills have taken an increasing share of the cement milling market, not least because the specific power consumption of vertical mills is about 30% less than that of ball mills and for finely ground cement less still. The vertical mill has a proven track record in grinding blastfurnace slag, where it has the additional advantage of being a much more effective drier of wet feedstock than a ball mill.
The vertical mill is more complex but its installation is more compact. The relative installed capital costs tend to be site specific. Historically the installed cost has tended to be slightly higher for the vertical mill.
Special graph paper is used with lglg(1/R(x)) on the abscissa and lg(x) on the ordinate axes. The higher the value of n, the narrower the particle size distribution. The position parameter is the particle size with the highest mass density distribution, the peak of the mass density distribution curve.
Vertical mills tend to produce cement with a higher value of n. Values of n normally lie between 0.8 and 1.2, dependent particularly on cement fineness. The position parameter is, of course, lower for more finely ground cements.
Separator efficiency is defined as specific power consumption reduction of the mill open-to-closed-circuit with the actual separator, compared with specific power consumption reduction of the mill open-to-closed-circuit with an ideal separator.
As shown in Fig. 2.24, circulating factor is defined as mill mass flow, that is, fresh feed plus separator returns. The maximum power reduction arising from use of an ideal separator increases non-linearly with circulation factor and is dependent on Rf, normally based on residues in the interval 3245m. The value of the comminution index, W, is also a function of Rf. The finer the cement, the lower Rf and the greater the maximum power reduction. At C = 2 most of maximum power reduction is achieved, but beyond C = 3 there is very little further reduction.
Separator particle separation performance is assessed using the Tromp curve, a graph of percentage separator feed to rejects against particle size range. An example is shown in Fig. 2.25. Data required is the PSD of separator feed material and of rejects and finished product streams. The bypass and slope provide a measure of separator performance.
The particle size is plotted on a logarithmic scale on the ordinate axis. The percentage is plotted on the abscissa either on a linear (as shown here) or on a Gaussian scale. The advantage of using the Gaussian scale is that the two parts of the graph can be approximated by two straight lines.
The measurement of PSD of a sample of cement is carried out using laser-based methodologies. It requires a skilled operator to achieve consistent results. Agglomeration will vary dependent on whether grinding aid is used. Different laser analysis methods may not give the same results, so for comparative purposes the same method must be used.
The ball mill is a cylindrical drum (or cylindrical conical) turning around its horizontal axis. It is partially filled with grinding bodies: cast iron or steel balls, or even flint (silica) or porcelain bearings. Spaces between balls or bearings are occupied by the load to be milled.
Following drum rotation, balls or bearings rise by rolling along the cylindrical wall and descending again in a cascade or cataract from a certain height. The output is then milled between two grinding bodies.
Ball mills could operate dry or even process a water suspension (almost always for ores). Dry, it is fed through a chute or a screw through the units opening. In a wet path, a system of scoops that turn with the mill is used and it plunges into a stationary tank.
Mechanochemical synthesis involves high-energy milling techniques and is generally carried out under controlled atmospheres. Nanocomposite powders of oxide, nonoxide, and mixed oxide/nonoxide materials can be prepared using this method. The major drawbacks of this synthesis method are: (1) discrete nanoparticles in the finest size range cannot be prepared; and (2) contamination of the product by the milling media.
More or less any ceramic composite powder can be synthesized by mechanical mixing of the constituent phases. The main factors that determine the properties of the resultant nanocomposite products are the type of raw materials, purity, the particle size, size distribution, and degree of agglomeration. Maintaining purity of the powders is essential for avoiding the formation of a secondary phase during sintering. Wet ball or attrition milling techniques can be used for the synthesis of homogeneous powder mixture. Al2O3/SiC composites are widely prepared by this conventional powder mixing route by using ball milling . However, the disadvantage in the milling step is that it may induce certain pollution derived from the milling media.
In this mechanical method of production of nanomaterials, which works on the principle of impact, the size reduction is achieved through the impact caused when the balls drop from the top of the chamber containing the source material.
A ball mill consists of a hollow cylindrical chamber (Fig. 6.2) which rotates about a horizontal axis, and the chamber is partially filled with small balls made of steel, tungsten carbide, zirconia, agate, alumina, or silicon nitride having diameter generally 10mm. The inner surface area of the chamber is lined with an abrasion-resistant material like manganese, steel, or rubber. The magnet, placed outside the chamber, provides the pulling force to the grinding material, and by changing the magnetic force, the milling energy can be varied as desired. The ball milling process is carried out for approximately 100150h to obtain uniform-sized fine powder. In high-energy ball milling, vacuum or a specific gaseous atmosphere is maintained inside the chamber. High-energy mills are classified into attrition ball mills, planetary ball mills, vibrating ball mills, and low-energy tumbling mills. In high-energy ball milling, formation of ceramic nano-reinforcement by in situ reaction is possible.
It is an inexpensive and easy process which enables industrial scale productivity. As grinding is done in a closed chamber, dust, or contamination from the surroundings is avoided. This technique can be used to prepare dry as well as wet nanopowders. Composition of the grinding material can be varied as desired. Even though this method has several advantages, there are some disadvantages. The major disadvantage is that the shape of the produced nanoparticles is not regular. Moreover, energy consumption is relatively high, which reduces the production efficiency. This technique is suitable for the fabrication of several nanocomposites, which include Co- and Cu-based nanomaterials, Ni-NiO nanocomposites, and nanocomposites of Ti,C .
Planetary ball mill was used to synthesize iron nanoparticles. The synthesized nanoparticles were subjected to the characterization studies by X-ray diffraction (XRD), and scanning electron microscopy (SEM) techniques using a SIEMENS-D5000 diffractometer and Hitachi S-4800. For the synthesis of iron nanoparticles, commercial iron powder having particles size of 10m was used. The iron powder was subjected to planetary ball milling for various period of time. The optimum time period for the synthesis of nanoparticles was observed to be 10h because after that time period, chances of contamination inclined and the particles size became almost constant so the powder was ball milled for 10h to synthesize nanoparticles . Fig. 12 shows the SEM image of the iron nanoparticles.
The vibratory ball mill is another kind of high-energy ball mill that is used mainly for preparing amorphous alloys. The vials capacities in the vibratory mills are smaller (about 10 ml in volume) compared to the previous types of mills. In this mill, the charge of the powder and milling tools are agitated in three perpendicular directions (Fig. 1.6) at very high speed, as high as 1200 rpm.
Another type of the vibratory ball mill, which is used at the van der Waals-Zeeman Laboratory, consists of a stainless steel vial with a hardened steel bottom, and a single hardened steel ball of 6 cm in diameter (Fig. 1.7).
The mill is evacuated during milling to a pressure of 106 Torr, in order to avoid reactions with a gas atmosphere. Subsequently, this mill is suitable for mechanical alloying of some special systems that are highly reactive with the surrounding atmosphere, such as rare earth elements.
In spite of the traditional approaches used for gas-solid reaction at relatively high temperature, Calka etal. and El-Eskandarany etal. proposed a solid-state approach, the so-called reactive ball milling (RBM), used for preparations different families of meal nitrides and hydrides at ambient temperature. This mechanically induced gas-solid reaction can be successfully achieved, using either high- or low-energy ball-milling methods, as shown in Fig.9.5. However, high-energy ball mill is an efficient process for synthesizing nanocrystalline MgH2 powders using RBM technique, it may be difficult to scale up for matching the mass production required by industrial sector. Therefore, from a practical point of view, high-capacity low-energy milling, which can be easily scaled-up to produce large amount of MgH2 fine powders, may be more suitable for industrial mass production.
In both approaches but with different scale of time and milling efficiency, the starting Mg metal powders milled under hydrogen gas atmosphere are practicing to dramatic lattice imperfections such as twinning and dislocations. These defects are caused by plastics deformation coupled with shear and impact forces generated by the ball-milling media. The powders are, therefore, disintegrated into smaller particles with large surface area, where very clean or fresh oxygen-free active surfaces of the powders are created. Moreover, these defects, which are intensively located at the grain boundaries, lead to separate micro-scaled Mg grains into finer grains capable to getter hydrogen by the first atomically clean surfaces to form MgH2 nanopowders.
Fig.9.5 illustrates common lab scale procedure for preparing MgH2 powders, starting from pure Mg powders, using RBM via (1) high-energy and (2) low-energy ball milling. The starting material can be Mg-rods, in which they are processed via sever plastic deformation, using for example cold-rolling approach, as illustrated in Fig.9.5. The heavily deformed Mg-rods obtained after certain cold rolling passes can be snipped into small chips and then ball-milled under hydrogen gas to produce MgH2 powders.
Planetary ball mills are the most popular mills used in scientific research for synthesizing MgH2 nanopowders. In this type of mill, the ball-milling media have considerably high energy, because milling stock and balls come off the inner wall of the vial and the effective centrifugal force reaches up to 20 times gravitational acceleration. The centrifugal forces caused by the rotation of the supporting disc and autonomous turning of the vial act on the milling charge (balls and powders). Since the turning directions of the supporting disc and the vial are opposite, the centrifugal forces alternately are synchronized and opposite. Therefore, the milling media and the charged powders alternatively roll on the inner wall of the vial, and are lifted and thrown off across the bowl at high speed.
In the typical experimental procedure, a certain amount of the Mg (usually in the range between 3 and 10g based on the vials volume) is balanced inside an inert gas atmosphere (argon or helium) in a glove box and sealed together with certain number of balls (e.g., 2050 hardened steel balls) into a hardened steel vial (Fig.9.5A and B), using, for example, a gas-temperature-monitoring system (GST). With the GST system, it becomes possible to monitor the progress of the gas-solid reaction taking place during the RBM process, as shown in Fig.9.5C and D. The temperature and pressure changes in the system during milling can be also used to realize the completion of the reaction and the expected end product during the different stages of milling (Fig.9.5D). The ball-to-powder weight ratio is usually selected to be in the range between 10:1 and 50:1. The vial is then evacuated to the level of 103bar before introducing H2 gas to fill the vial with a pressure of 550bar (Fig.9.5B). The milling process is started by mounting the vial on a high-energy ball mill operated at ambient temperature (Fig.9.5C).
Tumbling mill is cylindrical shell (Fig.9.6AC) that rotates about a horizontal axis (Fig.9.6D). Hydrogen gas is pressurized into the vial (Fig.9.6C) together with Mg powders and ball-milling media, using ball-to-powder weight ratio in the range between 30:1 and 100:1. Mg powder particles meet the abrasive and impacting force (Fig.9.6E), which reduce the particle size and create fresh-powder surfaces (Fig.9.6F) ready to react with hydrogen milling atmosphere.
Figure 9.6. Photographs taken from KISR-EBRC/NAM Lab, Kuwait, show (A) the vial and milling media (balls) and (B) the setup performed to charge the vial with 50bar of hydrogen gas. The photograph in (C) presents the complete setup of GST (supplied by Evico-magnetic, Germany) system prior to start the RBM experiment for preparing of MgH2 powders, using Planetary Ball Mill P400 (provided by Retsch, Germany). GST system allows us to monitor the progress of RBM process, as indexed by temperature and pressure versus milling time (D).
The useful kinetic energy in tumbling mill can be applied to the Mg powder particles (Fig.9.7E) by the following means: (1) collision between the balls and the powders; (2) pressure loading of powders pinned between milling media or between the milling media and the liner; (3) impact of the falling milling media; (4) shear and abrasion caused by dragging of particles between moving milling media; and (5) shock-wave transmitted through crop load by falling milling media. One advantage of this type of mill is that large amount of the powders (100500g or more based on the mill capacity) can be fabricated for each milling run. Thus, it is suitable for pilot and/or industrial scale of MgH2 production. In addition, low-energy ball mill produces homogeneous and uniform powders when compared with the high-energy ball mill. Furthermore, such tumbling mills are cheaper than high-energy mills and operated simply with low-maintenance requirements. However, this kind of low-energy mill requires long-term milling time (more than 300h) to complete the gas-solid reaction and to obtain nanocrystalline MgH2 powders.
Figure 9.7. Photos taken from KISR-EBRC/NAM Lab, Kuwait, display setup of a lab-scale roller mill (1000m in volume) showing (A) the milling tools including the balls (milling media and vial), (B) charging Mg powders in the vial inside inert gas atmosphere glove box, (C) evacuation setup and pressurizing hydrogen gas in the vial, and (D) ball milling processed, using a roller mill. Schematic presentations show the ball positions and movement inside the vial of a tumbler mall mill at a dynamic mode is shown in (E), where a typical ball-powder-ball collusion for a low energy tumbling ball mill is presented in (F).
As the developer and manufacturer of industry-leading particle size reduction equipment, including Attritors (internally agitated ball mills) and DMQX horizontal media mills, Union Process is uniquely positioned to help you identify and source the correct grinding media for your application.
Union Process customers know they can rely on our extensive technical expertise and years of experience to ensure they get the right grinding media at the right time and the right price for their specific needs. Working in close consultation with our customers, our skilled technical service representatives reviewcustomer requirements like final particle size, physical compatibility and contamination concernsand then recommend media with the right characteristics, including:
metallic grinding medialike carbon steel, forged steel, stainless steel or chrome steel grinding balls are best for some applications, while others requirenon-metallic mediamade of alumina, ceramics, glass, silicon carbide, zirconium oxide or other specialized materials
Union Process is the source for the most up-to-date information on grinding balls and other media. Download our Grinding Media Literature (PDF) to view a detailed sheet, outlining factors to consider when selecting grinding media, along with specifications on the most common types of media.
Offering the optimal combination of grinding media knowledge and manufacturing expertise, Union Process takes your entire operation into account to identify the best grinding media to consistently generate the final particle size and shape required by your application, optimizing the cost effectiveness of your process and extending the life of your mills.
Backed by our long-standing commitment to customer satisfaction, we ensure quality manufacturing and reliable supply of grinding balls and media specifically engineered to meet the requirements of your most demanding milling applications.
Through-Hardened Carbon Steel Balls are magnetic and can be used in the food industry along with 440C stainless steel media. They are a low-cost media that are superior and recommended over case-hardened carbon steel media which have a soft core. They are packaged with no oil finishalways dry packedas they will rust in water.
Chrome Balls (steel type 52100) are through-hardened and tempered steel balls designed to achieve maximum strength and quality. Ball hardness is in the 6067 HRC range. They wear better than 440C stainless steel and through-hardened carbon steel. They are also recommended for applications where a through-hardened steel ball is needed in larger sizes ( and larger). They are sometimes packaged with a very light oil finish to reduce rust due to humidity.
440C Stainless Steel Balls are through-hardened and tempered throughout for maximum strength and quality. They are magnetic, and corrosion-resistant (generally rust-resistant). They are recommended for food applications and lighter colored slurries.
Forged Steel Balls are used for gold mining, cement factories, oil processing and large scale industrial applications. They are made by machine (standard) sizes 20mm75mm. They are manually made (hit by air hammer) sizes 75mm125mm. They have hardness 5563 HRC. All forged balls are through-hardened, and shipped in 55-gallon steel drums. They range in sizes from 20mm125mm. Standard lead time is 68 weeks FOB Akron, Ohio USA.
High Chrome Steel Balls can be used for many different applications. They are available in two grades1013% chrome (surface hardness 60 HRC, core hardness 58 HRC), and 1418% chrome (surface hardness 62 HRC, core hardness 60 HRC). They have a very rough black finish which quickly wears off during initial milling. After that, they have an excellent wear rate. They are available in sizes 6mm120mm. They are shipped in 55-gallon steel drums with standard lead time of 68 weeks FOB Akron, Ohio USA.
Union Process is the one source for 90%, 94%, 99.5% and 99.9% alumina media. 90% alumina is available in satellites and rod/cylinders. 94% alumina balls have excellent wear resistance with higher impact strength to save running costs with less contamination. They have great wear rate generally better than 90% or 99.5%. 99.5% alumina balls have the highest alumina content for a moderate price. The .5% impurity is MgO that is added to inhibit grain growth during sintering in the kiln. There is less than 0.1% silica in the media. 99.9% alumina balls are made of very pure and reactive (expensive) raw materials. They are for high purity alumina applications where contamination is a factor.
Alumina Beads are specially formulated to be used in high-energy mills in which a high degree of fineness is required. They are used in various industrial fields such as inks, paints, advanced ceramics, mining, cosmetics and pharmaceutical industries. They are a perfectly spherical shape with high mechanical properties and high wear resistance at a moderate price.
Silicon Carbide Balls are very high-cost grinding media that are used for milling same materials (silicon carbide ball to mill silicon carbide materials) to avoid contamination. They are only available in 5mm,10mm, 15mm and 20mm sizes. They are a special order item.
Silicon Nitride Balls are very high-cost grinding media that are used for milling same materials (silicon nitride ball to mill silicon nitride materials) to avoid contamination. They are now available in 2mm and 3mm and sizes up to 25mm. They are a special order item.
Tungsten Carbide media is the hardest and densest (highest specific gravity) media and is available in both satellites and balls upon request. They are available in sizes ranging from 3/32 to 1 in diameter. They are a high-cost media and are a special order item.
Zirconium Oxide Balls (95% ZrO2) are the strongest, best wearing ceramic media for metal-free, pharmaceutical and food processing grinding. These balls have a white, shiny appearance. They are also available in 38 and cylinders. This 95% grade is high-cost.
Zirconium Oxide SatellitesCeria stabilized (rare earth)are a cheaper zirconium oxide alternative for metal-free applications. They are a brown, shiny ball media that come in size ranges in the smallest sizes (ex: 0.4 0.6mm), then at 6mm come in uniform sizes (6mm, 8mm, etc.) up to 31mm.
Zirconium Silicate Beads are available in fused 68% ZrO2 beads which are a standard reliable media at low cost, and sintered 58% ZrO2 beads which have high breakage resistance, are durable and cost effective. They are used to microgrind paints, inks, dyes, magnetic coatings, minerals, agrochemicals and ceramics.
NOTE: Grinding balls and media are sold on a per pound basis, but ATTRITORS and DMQX-Mills are loaded by volume. Therefore, the more dense the media, the more pounds of media required. For instance, a machine requiring only 40 lbs. of stainless steel may require up to 80 lbs. of tungsten carbide. Information contained herein is accurate and reliable to the best of our knowledge, but our suggestions and recommendations cannot be guaranteed because the conditions of use are beyond our control.
Our skilled technical service representatives also are available for consultation and are always ready to assist you with the selection of grinding media that best answer your particular needs. For assistance, contact Media Sales by calling (330) 929-3333 x228 or byemailing us.
Breakaway oxidation of alloy 304L at 600C was studied in four environments (O2+H2O+KCl, O2+H2O+SO2+KCl, H2+H2O+Ar, O2+K2CO3) for up to 168h. The resulting scales were investigated by FIB/SEM, SEM/EDX, STEM/EELS, STEM/EDS and oxidation was elucidated by thermodynamic calculations (Thermo-Calc). The initial thin protective scale broke down in all cases. After breakaway, the scale consisted of two layers, i.e. an inward growing spinel/reaction zone and an outward growing iron-rich layer. The general features and microstructure of the scales after breakaway were similar in all environments and were explained in terms of: (1) Different diffusivities of Cr3+ and Fe2+ in the spinel oxide. (2) The appearance of a miscibility gap in the FeCr and FeCrNi spinel oxides. (3) The equilibrium composition of the spinel (at low pO2 Ni is not present in the spinel).
Stainless steels are designed to form a dense slow-growing corundum-type Cr-rich ((CrxFe1x)2O3) scale in high temperature applications. In order to form such a protective scale the supply of chromium to the scale has to be sufficient. Depending on the corrosive environment and the alloy microstructure, different concentrations of Cr in the alloy are necessary for corrosion protection. The corundum-type Cr rich ((CrxFe1x)2O3) scale is known to provide protection against oxidation and corrosion. A great number of publications concerning the oxidation properties of steels and stainless steels are summarized by Kofstad , Birks and Meier  and Young .
In demanding high temperature applications the ((CrxFe1x)2O3) scale may lose its protective properties resulting in breakaway oxidation. The rate of further oxidation then relies on the properties of the scale formed after breakaway. Breakaway oxidation of stainless steels in various environments and at different temperatures have been reported to result in a similar microstructure consisting of an outward growing Fe-rich oxide and an inward growing Fe, Cr, (Ni) spinel oxide, see e.g. . The microstructure of the inward growing spinel scale has been shown to be complex, including both fully oxidized regions and regions of internal oxidation as reported for FeCr model alloys at 600 to 900C [4, 5, 11, 1517]. A similar behavior has been reported for commercial alloys in different corrosive environments and temperatures, see e.g. [9, 11, 16]. Different authors have attributed this characteristic breakaway behavior to a variety of causes, which are more or less specific for the type of environments studied. However, the similarities in the scale structures observed after breakaway hints that there is an underlying mechanism which is generally applicable and that does not require a specific combination of corrosive factors. The present study is designed to study that underlying mechanism using different corrosive factors to trigger breakaway oxidation in Fe, Cr, (Ni) alloys. A comparative investigation of the microstructural evolution post-breakaway oxidation is therefore performed. Breakaway oxidation is induced in alloy 304L by exposure to different corrosive environments at the same temperature and then the microstructural evolution in the different environments is studied in detail. The four corrosive environments selected were:
It has been suggested that, at high oxygen activity, water vapour and alkali cause breakaway oxidation of stainless steels in a similar way. Thus, the breakdown of the initial protective oxide in the presence of water vapor/alkali was explained by the formation of chromium(VI) oxide hydroxide/alkali chromate in a reaction with the protective scale temperatures [4, 8, 18, 19]. Addition of small amounts of sulfur dioxide has been shown to prolong the incubation time for breakaway oxidation in the presence of KCl and to slow down scale growth after breakaway . At low oxygen activities, evaporation of chromium (VI) oxide hydroxide is negligible and the accelerated oxidation of stainless steels in H2+H2O environment must have other causes. Various explanations have been suggested for the effect of water vapor at low oxygen activity [16, 2123]. In addition, one environment was selected that does not include water vapor or chlorine-containing species (K2CO3+O2). The investigation was performed on the commercial austenitic stainless steel 304L (two grades 18Cr10Ni and 18Cr8Ni) in the four different environments giving rise to breakdown of the ((CrxFe1x)2O3) scale.
The aim of the present paper is to investigate the oxidation of the stainless steel 304L after breakdown of the protective scale. The different corrosive environments give the possibility to study the propagation after breakaway oxidation, independent of the breakdown mechanism. The work involves a detailed microstructural investigation of the oxide scales formed after breakaway in different corrosive environments as well as thermodynamic equilibrium calculations. The following characterization techniques were employed: Scanning electron microscopy coupled with energy dispersive X-ray spectroscopy (SEM/EDX) in order to acquire an overall understanding of the microstructure (in the micrometer range) with good statistics so that characteristic and representative features of the microstructure could be detected and selected for further analysis. Scanning transmission electron microscopy coupled with energy dispersive X-ray spectroscopy (STEM/EDX) was used in order to investigate the scale microstructure and its chemical composition at nanometer range resolution. Scanning transmission electron microscopy coupled with electron energy loss spectroscopy (STEM/EELS) was used to detect trace elements due to the high detection sensitivity of this technique.
The investigated materials are the austenitic stainless steels 304L (18Cr10Ni) and (18Cr8Ni), see Table1 for detailed compositions. The material was cut into coupons (15152mm3) and grinded on 320 grit SiC with deionized water and then polished with 9, 3, and 1m diamond solution on a polishing cloth. The polished samples were degreased and cleaned in acetone and ethanol using an ultrasonic bath. Salt (KCl or K2CO3) was applied by spraying the samples with a saturated solution of the desired salt in water, water/ethanol or acetone. The amount of salt corresponded to a potassium content of 1.35molK+/cm2 (0.10mg/cm2 in the case of KCl). The samples were dried with air and stored in a desiccator prior to exposure. Mass change was measured after exposure using a six decimal Sartorius balance and several samples were included for every exposure time in the mass gain curve, i.e. the spread between samples gives the error bar.
All exposures in high oxygen partial pressure were isothermal and performed in silica tube furnaces, see  for experimental setup. The triplicate samples of 304L (18Cr10Ni) were mounted on a sample holder and positioned parallel to the direction of gas flow. Temperature was kept at 600C (1C) and the exposures lasted 1, 24, 72 or 168h. The exposure gas consisted of 5%O2+95%N2 or 5%O2+40%H2O+55%N2 and the flow rate was 3.2cm/min. The dry gas was humidified to the desired water concentration. Flow rates were controlled with Bios DC2 Flow Calibrators. SO2 was added directly to the furnace 20cm upstream from the samples, at a temperature of about 300C. The flow was controlled by a digital mass flow regulator, to reach a concentration of 300ppm of SO2, and N2 was used as a carrier gas.
The exposures performed in low oxygen partial pressure were isothermal and performed in tabular furnaces on 304L (18Cr8Ni). The gas flow was set at 0.8cm/min using digital mass flow regulators. An alumina tube was set in to the furnace and samples were place in a sample holder standing parallel to the flow. A peristaltic pump was calibrated precisely and used as water supplier. The exposure gas composition was 10%H220%H2OAr while the furnace was isothermally kept at 600C (1C) giving an oxygen activity of about 1024. The exposures were performed for 1, 24, 72 or 168h. Different exposure processes were tested (using oxygen and humidity sensors) to achieve the optimum process to get rid of the impurity traces available in the environment and thus provide the intended low oxygen activity .
Crystalline corrosion products were analyzed with grazing-incidence X-ray diffraction (GI-XRD) using a Siemens D5000 powder diffractometer, equipped with grazing incidence beam attachment and a Gbel mirror. Cu K radiation was used and the angle of incidence was 110. The measurement range was 20<2<70.
Wide cross sections in the millimeter range were prepared using two broad ion beam (BIB) instruments: a Hitachi E-3500 and a Gatan Illion system. The instruments were equipped with argon ion guns for sputtering which were operated at 6kV for 35h. Protective masks were positioned on the top surface of the samples in order to shield them from ion beam damage as well as to render a smooth cross section. Furthermore, to enhance the surface finish and to avoid projecting the surface topography on to the final cross-section (curtaining effect), a thin glass sheet was applied to the samples top surface using super glue adhesive. After the glue was cured, samples were dry cut using a low speed diamond saw before mounting in the broad ion beam (BIB) instrument.
An FEI Quanta 200 Environmental Scanning Electron Microscope (SEM) equipped with a field emission gun (FEG) was used for analytical scanning electron microscopy. The following SEM operational conditions were employed: accelerating voltage of 1015kV and 10mm working distance. Both secondary electron (SE) and backscattered electron (BSE) imaging modes were used. Chemical analysis was carried out using an 80mm2 X-MaxN Oxford Instruments silicon drift detector (SDD) and Oxford Instruments INCA Energy Dispersive X-ray (EDX) software. For consistency all the EDX results are presented in atomic percentages of the total cation content.
Thin foil specimens for subsequent transmission electron microscopy analysis were prepared using an in-situ lift-out technique  in an FEI Strata 235 DualBeam and an FEI Versa 3D DualBeam focused ion beam/scanning electron microscope (FIB/SEM) instruments both equipped with OmniProbe micromanipulators. Thin cross-sectional specimens were extracted from the samples using the micromanipulator and then welded onto a copper grid for further ion milling. The samples were first investigated with top view SEM imaging using the SEM column of the FIB/SEM to locate region of interest. Then a 2m wide and 4m thick platinum strip was deposited on the area of interest; first, a 100nm thick protective layer of platinum was deposited using electron beam, and then this strip was thickened to 4m with ion-assisted platinum deposition. This was done to prevent Ga ion beam damage and milling of the oxide layer during subsequent milling of the lift-out specimen. Thereafter a specimen, about 10m in depth below the protected region, was lifted out from the sample and transferred onto a copper grid using the micromanipulator and milled down to electron transparency (~100nm). Finally, Ar ion milling was used to further reduce the amount of Ga ion damage from the FIB process in some of the samples employing a Fischione 1010 argon mill instrument operating at 1kV, 3mA, incident angle of 6 and rocking angle of 25 for about 15min.
The analytical scanning transmission electron microscopy (STEM) investigation was carried out using a Phillips CM200 FEG microscope and an FEI Titan 80-300 microscope equipped with a Gatan 866 GIF Tridiem energy filter. Both TEMs were equipped with Oxford Instruments Inca X-Sight energy dispersive X-ray (EDX) system, which was used for elemental mapping, linescans and chemical quantification while only the FEI Titan 80-300 was employed for electron energy loss spectroscopy (EELS) analysis on one of the samples using a Gatan 866 GIF Tridiem energy filter. The CM200 microscope was operated at 200kV while the FEI Titan 80300 was operated at 300kV. The EELS technique was only used for qualitative identification of elements while EDX was employed for quantitative analysis. The sample with the longest exposure time after breakaway and highest oxygen activity was investigated with EELS for trace amounts of oxygen in the alloy matrix ahead of the oxidation front. Since no oxygen was detected in the alloy matrix in this sample, it was considered that the possibility of finding oxygen in other samples with shorter exposure time and lower oxygen activity was low. Hence, EELS investigation was not carried out on these samples. Due to the technical difficulties in quantifying light elements (e.g. oxygen) using EDX in thin foil specimens  elemental composition of the oxide scales are presented in terms of percentages of the total cation content. The diffraction patterns were indexed by direct measurement and not ratios.
Thermodynamic calculations were performed using the THERMO-CALC program [28, 29] with the database TCFE7. The software was used in order to calculate the equilibrium composition of the different oxide phases formed by an alloy with the 304L composition at 600C, as a function of O2 activity.
The focus of this paper is the microstructural evolution of the oxide scale after breakaway. However, the general features of the oxidation kinetics and surface morphology in the different environments are described first and summarized in Table2.
Figure1a, b shows mass gain curves for 304L (both 18Cr10Ni and 18Cr8Ni) exposed in different environments including dry O2 and O2+40% H2O (from ). The mass gains recorded in O2+H2O, O2+H2O+KCl, O2+H2O+SO2+KCl, H2+H2O+Ar, O2+K2CO3 all indicate breakaway oxidation. The only environment giving a slow-growing scale is dry oxygen where a 200 nm Cr-rich (Fe,Cr)2O3 protective oxide was detected after 168h, in accordance with the minimal mass gain . Both 304L grades studied behaved in the same way in dry O2 (only 18Cr10Ni shown). In, contrast, exposure in O2+H2O environment resulted in a strongly corroded surface with island-like features . With time, the oxide islands grow in size and increase in number, the oxidation kinetics during the first 168h being linear. The addition of small amounts of KCl (O2+H2O+KCl) resulted in significantly faster corrosion in the initial stages. The effect of KCl on mass gain is initially smaller in the presence of 300ppm SO2. However, SO2 has little effect on the slope of the mass gain curve between 24 and 168h. The mass gain curve in K2CO3+O2 also indicates breakaway oxidation. The highest mass gains after 168h were recorded in 10%H2+20%H2O+Ar. A comparison of all the investigated samples shows that oxidation is initially fastest in the presence of KCl, while the 10%H2+20%H2O+Ar environment caused the fastest corrosion between 24 and 168h.
a, b Mass gain of alloy 304L (18Cr10Ni) in dry O2 coated with K2CO3, in the presence and absence of 300ppm SO2 in 5% O2+40% H2O+KCl at 600C and from 304L (18Cr8Ni) exposed in 10%H220%H2OAr. Reference exposures in dry 5% O2 and 5% O2+40% H2O at 600C are also included from 
The effect of KCl on the corrosion of 304L (1810) at 600C has been reported in several papers, see e.g. , and the present plan view analysis corresponds closely to previous results. Figure2 shows BSE images of samples exposed in the presence of KCl in 5%O2+40%H2O environment for 1 and 24h. After 1h, the partly unreacted salt particles have become surrounded by an oxide rim. Besides the salt particle, two additional morphologies are visible at this stage. The dark area (lower left) corresponds to thick oxide while the brighter area is covered by a thin oxide. The dark, 12m size, spots on the oxide surface were identified as potassium chromate (K2CrO4) . After 24h the entire surface was covered by thick, Fe-rich scale and the former salt crystals were replaced by irregularly shaped oxide agglomerations. Between the oxide agglomerations the oxide scale exhibited an island-like morphology that reflects the substrate grain structure. According to SEM/EDX and XRD analyses the outer part of the thicker oxide consisted of almost pure hematite. The surface morphology after 168h was similar (not shown).
BSE plan view images of 304L (18Cr10Ni) samples exposed for 1 and 24h at 600C in 5% O2+40% H2O+KCl. After 1h exposure some thin protective scale with K2CrO4 particles could be observed together with thicker scale and unreacted KCl. Longer exposure times gives a surface morphology consisting of iron oxide and former KCl particles
Figure3 shows BSE images of samples exposed for 24 and 168h. SEM/EDX analysis showed that the small (<10m) KCl particles were fully converted to potassium sulphate (K2SO4) after 24h. The larger KCl particles were only partly converted after 24h as shown by the detection of chlorine. In contrast to the corresponding exposure in the absence of SO2, large parts of the surface were still covered by thin oxide after 24h, in agreement with the relatively low mass gain (Fig.1a). According to SEM/EDX and XRD analyses, the outer part of the thick oxide consisted of almost pure hematite while the thin oxide was chromium-rich. After 168h the KCl was completely reacted, no chlorine being detected, the former salt particles consisting of a K2SO4-rich centre surrounded by a Fe-rich oxide rim (SEM/EDX). In the vicinity of the former KCl particles, dark 28m diameter K2SO4 particles appeared. After 168h the entire surface was covered by a uniform thick Fe-rich oxide scale (hematite according to XRD and SEM/EDX) containing large K2SO4 particles, see Fig.3.
BSE plan view images of 304L (18Cr10Ni) samples exposed for 1, 24 and 168h at 600C in 5% O2+40% H2O+300ppm SO2 KCl. After 24h exposure most of the surface is covered by a thicker iron rich oxide while only some regions with thin scale could be observed. Longer exposure times gives a surface morphology consisting of iron oxide and former KCl particles
Figure4 shows BSE images of samples exposed in 10%H2+20%H2O+Ar for 1 and 24h. After 1h most of the surface is covered by thick scale. However, there are small islands covered by thin oxide in the scale. The size of the regions covered by thin oxide corresponds to the typical alloy grain size. At this stage the scale morphology on the alloy grains was different from that at the alloy grain boundaries (Fig.4). After 24 and 168h the entire surface was covered by magnetite (SEM/EDX and XRD). The grain size of the surface oxide had grown compared to 1hs exposure.
BSE plan view of the 304L (18Cr8Ni) samples exposed at 600C in the environment containing 10%H220%H2OAr. After 1h exposure thin protective scale could be observed on small regions together with thicker scale iron rich scale. Longer exposure times gives a surface morphology consisting of iron oxide
According to the oxidation kinetics and the examination of the surface morphology, the thin protective scale suffered breakdown in all environments studied except dry O2. However, the incubation time to breakaway differed (Table2). While the breakdown was initially local in all cases, with time the entire surface became covered by thick rapidly growing scale.
After 1h, two distinctive scale morphologies were observed (Fig.2). Previously it was shown that the thick oxide developed under these exposure conditions consists of a 12m thick duplex oxide scale , the outer oxide layer being made up of almost pure hematite while the inward growing oxide consists of spinel-type oxide with a typical composition of 40wt% Fe, 40wt% Cr and 20wt% Ni, (cationic) i.e. (Fe,Cr,Ni)3O4. It was reported that the thin oxide consists of an approximately 100nm Cr-rich oxide . After 24h the entire surface was covered by thick oxide scale (not shown). The present results show that the oxide formed after 24h was similar to the thick oxide observed after 1h. Previously it was shown that the outer part of the scale consists of almost pure hematite while the inner part consists of FeCrNi spinel .
Figure5a shows a BSE image of a BIB cross-section trough the scale after 168h. Similar to the situation after 24h , the scale is layered, consisting of an outward growing and an inward- growing part. There is a characteristic variation in scale thickness, from about 3m at the alloy grain boundaries to about 10m at the grain centers. The thicker scale corresponds to island-like features in the outward-grown scale and craterlike features in the inward-grown scale. The outward-grown scale consisted of hematite (SEM/EDX and XRD). The inward-growing oxide exhibited two distinct microstructures; see below. Figure5b shows a BSE image at higher magnification of the region indicated in Fig.5a. This morphology was observed on about 10% of the inward-growing scale (interpreting SEM/BSE and SEM/EDX from the several mm wide BIB cross-section) after 168h and consists of small dark regions (labeled oxide in Fig.5b) surrounded by brighter regions, (labeled reaction zone). The dark regions have an oxygen content of 5560at.% while the reaction zone has an oxygen content of 3040at.% (SEM/EDX, the iron oxide in the upper part of the scale was used as a standard for the quantification of oxygen). It may be noted that the overall composition of the inward-growing part of the scale in Fig.5b, with about 40% Fe, 30% Cr and 20% Ni (cations), is iron-deficient compared to the alloy. A TEM sample was ion milled from a region similar to the area labeled 5c in Fig.5b, using the FIB/SEM, showing the reaction zone and the oxide. Figure5c shows a scanning transmission electron microscope (STEM) high angle annular dark field (HAADF) image of the microstructure at higher magnification. The image shows a reaction zone, consisting of a mixture of oxide and remaining metal, including a darker fully oxidized region with voids (top right). The oxide inclusions in the reaction zone had an average size of the about 100nm. From the TEM/EDX analysis it was concluded that the particles consisted of FeCr oxide with about 45% Cr and 55% Fe (cationic). The remaining metal contained Fe and Ni. The composition varied, the Ni content reaching up to 50%. Small amounts of Si and Mn were detected in the oxide while small amounts of Mo were found in the remaining metal. The typical composition and size of the oxide particles in the reaction zone (=the internal oxide) is summarized in Table3. It is challenging to measure the composition of particles having sizes similar to the thin foil thickness and the data in Table3 are therefore based on interpreting the EDX analysis of a large number of particles (>100 particles). In the lower left of the image in Fig.5c there is a compact, fully oxidized region, which was connected to an alloy grain boundary. It contained about 60% Cr and 40% Fe (cationic) together with 12at.% Mn and Si. Both the oxide precipitates in the reaction zone and the oxide in the fully oxidized grain boundary region were identified as spinel structure (M3O4) by indexing convergent beam electron diffraction (CBED) patterns, see Fig.5c (i, ii). The metal just below the reaction zone (lower right corner in Fig.5c) was investigated by STEM/EELS and no oxygen signal was detected.
a BSE image of a BIB cross-section through the oxide scale and into the alloy on the 304L (18Cr10Ni) samples exposed for 168h at 600C in 5%O2+40%H2O+KCl. A magnified image of the marked region close to the right hand side of the micrograph is shown in (b) and the SEM/EDX elemental maps from the central marked region is shown in (d). b BSE image of the marked area in (a). The inward growing oxide scale can be divided into darker areas (fully oxidized regions) and brighter areas, i.e. mixed oxide and metal (reaction zone). A TEM sample was taken from a similar region as indicated in the marked region. c HAADF image of a region similar to the marked region in (b). The inward growing oxide scale can be divided into a region of mixed oxide and metal, i.e. a reaction zone, and darker fully oxidized regions. At the alloy grain boundary a FeCr oxide could be found. CBED patterns from the oxides are inseted in the image. The oxide at the grain boundary was indexed as a spinel structure (M3O4) with the zone axis  (i). The small oxide precipitates were also indexed as a spinel structure (M3O4) with the zone axis  (ii). d SEM/EDX elemental maps from the marked area in (a) of the sample exposed for 168h in 5%O2+40%H2O+KCl
Figure5d shows SEM/EDX elemental maps of the area labeled 5d in Fig.5a. The morphology of the crater region in 5d was typical of about 90% of the inward-growing scale after 168h. At the scale/alloy interface there is a Cr-rich oxide similar to the grain boundary oxide in Fig.5c. Above it there is a region with variable Ni concentration. This type of region was analyzed by SEM/EDX and was found to have an oxygen content of 5560at.%. The SEM/EDX has insufficient resolution to identify the complex microstructure seen with TEM. However, the difference in contrast and the SEM/EDX data indicate that the area corresponds to a fully oxidized reaction zone, i.e., it consists of a mixture of internal oxide (=FeCr spinel oxide) precipitates and fully oxidized remaining FeNi metal (compare Fig.5c). There were no indications of chlorine in the scale after breakaway .
Both plan view imaging and mass gain kinetics (see above) indicate that the oxide scale was still thin and protective after 1h exposure. While this scale was not investigated in detail, ion milled BIB cross-sections were prepared from samples exposed for 24 and 168h. After 24h the scale in the ion-milled cross-section was in the range 0.12m (not shown), representing both the remaining thin scale and the thicker scale shown in Fig.3. The thick scale was duplex with an iron-rich outer part (about 75% of the scale thickness) and a FeCrNi-rich inner part. At the alloy grain boundaries the inward- and outward-growing parts of the scale had about the same thickness. Furthermore, about 5at.% S were detected beneath the inward growing oxide, at the metal/oxide interface. At this stage (24h) there were no signs of formation of a reaction zone/internal oxidation in the millimetre wide cross-section. Figure6a shows a BSE image of an ion-milled cross-section from a sample exposed for 168h. The scale was typically 24m thick and the microstructure and composition of the scale were similar to the thick scale observed after 24h. However, in the alloy grain boundary regions the oxide penetrated deeper into the alloy (reaching 45m) and some regions could be observed that were interpreted as reaction zones (internal oxidation). Figure6b, c shows such a region at higher magnification. The upper section of the inward growing scale is fully oxidized and very similar to the scale found after 24h. It has a typical composition with about 60% Fe, 3035% Cr, up to 5% Ni and small amounts of Si, Mn and Mo. The HAADF image indicates small variations in composition and/or void formation in parts of this layer and void formation at the oxide/reaction zone interface. Similar to Fig.5c (KCl in the absence of SO2), the reaction zone consists of a mixture of internal oxide particles and remaining metal. Also, the size and composition of the remaining metal and the oxide are very similar to that observed in the corresponding exposure in the absence of SO2. Thus, the internal oxide consisted of FeCr oxide with about 45% Cr and 55% Fe (cationic) and the remaining FeNi metal in-between the oxide particles contained up to 50% Ni. The composition of the reaction zone is summarized in Table3. In addition, sulphur enrichments were observed at the alloy grain boundaries.
a BSE image of a BIB cross-section of the 304L (18Cr10Ni) sample exposed for 168h in presence of SO2 at 600C in 5%O2+40% H2O+KCl. The marked area is shown in higher magnification in (b, c). b, c HAADF image of a region similar to the marked region in (a). The inward growing oxide scale can be divided into a region of mixed oxide and metal, i.e. a reaction zone, and darker fully oxidized regions. c The marked region in (b) shown with higher magnification
The plan view images and mass gain kinetics indicate that the oxide scale has lost much of its protective properties already after 1h in this environment. The initial stages of breakaway and first steps of propagation have been investigated in detail elsewhere [30, 31], showing an outward growing magnetite layer above a Cr rich remnant of the initially formed scale after 1h. Further down into the scale an inward growing reaction zone was observed . After 24h the scale had grown to about 8 m and was made up of an outward growing magnetite layer (about 50% of the scale thickness) and an inward growing part consisting of both spinel oxide and reaction zone regions (internal oxidation) . The spinel oxide regions were found both immediately below the original steel surface and at alloy grain boundaries . The present investigation after 72h exposure shows all the features previously reported after 24h exposure, see BSE image of a mechanically polished cross-section in Fig.7a. At this stage the scale has grown to about 12m and exhibits an outward growing magnetite layer. The inward-growing part of the scale is dominated by regions that are not fully oxidized according to SEM/EDX (3040at.% O), being interpreted as reaction zones (=internal oxidation). The microstructure of the inward growing scale was investigated with the TEM both in the middle of an alloy grain and at an alloy grain boundary (marked regions in Fig.7a). Figure7b shows a HAADF image of a reaction zone located in the middle of an alloy grain at higher magnification. The top part of the inward growing scale is fully oxidized but it is very thin. The fully oxidized part consisted of iron chromium oxide (about 70% Fe and 30% Cr). However, the composition varies and in some positions the initial Cr rich oxide could be observed above a region consisting of Ni-rich metal (up to 70% Ni and 30% Fe) at the former metal/oxide interface. The reaction zone consisted of a mixture of internal oxide particles and remaining metal. While the internal oxide precipitates were smaller (about 30100nm) than in the high PO2 environments (compare Fig.5c), the composition of both the remaining metal and the oxide precipitate was similar to that observed in the other environments. The smaller size made the analysis even more difficult but the analysis showed that the oxide precipitates consisted of FeCr oxide with about 4550% Cr and 5055% Fe (cationic). The composition of the remaining FeNi rich metal varied, the Ni content being somewhat lower than observed on the 18Cr10Ni alloy. Close to the alloy grain boundary, a fully oxidized region could be observed, see HAADF image in Fig.7c. It contained approximately 35% Fe and 65% Cr and was surrounded by Ni rich metal with up to 70% Ni and 30% Fe.
a BSE image of a polished cross-section of the 304L (18Cr8Ni) sample exposed for 72h in 10%H220%H2OAr. The marked areas are shown in higher magnification in b, c. b HAADF image of a region similar to the marked area in (a). The inward growing oxide scale can be divided into a region of mixed oxide and metal, i.e. a reaction zone, and darker fully oxidized regions. c The marked region in (a) covering an alloy grain boundary region shown with higher magnification
Samples coated with small amounts of K2CO3 were exposed for 24 and 168h in dry oxygen. The analysis was performed on the 168h exposure in order to be able to study the microstructure formed without water vapor/chlorine and no plan view or detailed investigations were performed after shorter times. Figure8 shows a BSE image of a BIB cross-section ion milled through the oxide scale and into the alloy. The BSE image shows that the scale microstructure is very similar to that observed in the other environments after breakaway (compare Figs.5, 6). Thus, the scale consists of an outward growing layer consisting of iron oxide and an inward growing more complex oxide (FeCrNi oxide). In the top part of the outward growing iron oxide layer small amounts of potassium were observed (SEM/EDX). The outward- growing and inward-growing scale layers had similar thickness, the total scale thickness ranging from about 3 to 6m. The BSE image indicates that the inward growing scale has a complicated microstructure with alternating bright and dark areas, being very similar to samples exposed in the other environments, indicating the formation of reaction zones. The complicated microstructure was verified by SEM/EDX, which showed that the bright parts in the inward growing FeCrNi oxide were high in Ni even though the limited resolution made it impossible to measure the composition accurately.
BSE image of an ion milled cross-section of the 304L (18Cr10Ni) sample exposed for 168h at 600C in O2+K2CO3. The scale can be divided into an outward growing part consisting of iron oxide and an inward growing more complex oxide (FeCrNi oxide)
It is well known that stainless steels, e.g. alloy 304L, form a thin protective oxide scale in dry oxygen at 600C. The scale is well-characterized, consisting of a ((CrxFe1x)2O3) inner layer with about 70% Cr (cation%) and an iron-rich outer part [17, 3234]. In more corrosive environments containing, e.g. water vapor, alkali and/or chlorine the ((CrxFe1x)2O3) scale may lose its protective properties resulting in breakaway oxidation, see e.g., [8, 9, 35]. Breakaway tends to be triggered by Cr depletion of the scale and of the substrate. Hence, due to faster diffusion of Cr at alloy grain boundaries, breakaway in austenitic steels usually starts at the center of alloy grains while the oxide on alloy grain boundary regions is more resistant, see e.g. [8, 9]. This general trend is in accordance with the oxidation kinetics and surface morphologies observed at high oxygen activity in this study, see e.g. Figs.2 and 5a, while no clear difference between the alloy grains and grain boundaries was observed in H2O+H2 environment see Fig.4. The thin protective scale broke down in all environments investigated but the incubation time differed, see Fig.1 and Table2. The results show that the breakdown is initially localized in all cases, see Figs.2, 3 and 4. However, with longer exposure time the entire surface becomes covered with a thick, fast growing scale. The incubation time (an interpretation of the plan view images since the localized behavior makes it impossible to use the mass gain data) and the growth rate subsequent to breakaway are summarized in Table2.
Adding water vapour or alkali to an O2 containing environment is known to accelerate the corrosion of chromia-forming steels. Previously, the vaporization of chromium(VI) oxide hydroxide was shown to cause breakaway oxidation of FeCr model alloys and commercial stainless steels in environments containing O2+H2O at high temperatures [4, 18, 19]. In an analogous manner, the breakdown of the initial protective oxide in the presence of alkali  was explained by the formation of alkali chromate in a reaction with the protective scale. Addition of small amounts of sulfur dioxide has been shown to prolong the incubation time for breakaway oxidation in the presence of KCl and to slow down scale growth after breakaway . Both chromium (VI) oxide hydroxide volatilization and alkali chromate formation cause breakaway oxidation by depleting the oxide scale and the substrate in chromium. In this study, the breakdown of the protective scale in O2+H2O environment can be explained by chromium (VI) oxide hydroxide volatilization, see Figs.1, 2 and 3. At low oxygen activities, evaporation of chromium (VI) oxide hydroxide is negligible and the accelerated oxidation of stainless steels in H2+H2O environment must have other causes. Various explanations have been suggested for the effect of water vapor at low oxygen activity [16, 2123]. Tveten et al.  suggested that exposure to water vapor results in uptake of protons which influence the transport properties of the oxide. Essuman et al.  argued that hydrogen dissolution into the alloy enhances the inward diffusion of oxygen in the alloy, promoting internal oxidation and hindering the formation of an external chromia scale. Recently, a new mechanism was proposed explaining the corrosiveness of H2+H2O environments towards chromia-forming alloys. It was argued that the low oxygen activity causes the Cr-rich (Fe,Cr)2O3 scale to transform into faster growing Cr2O3. This causes a deeper Cr depletion of the substrate that allows Fe at the alloy/scale interface to oxidize . A study of K2CO3-coated steels in oxidizing environments without water vapor or chlorine also shows breakaway oxidation. This was explained by the formation of alkali chromate in a reaction with the protective scale . The different observations concerning mechanisms, incubation times and growth rate after breakaway oxidation in this work are summarized in Table2.
The oxide scale formed directly after breakdown of the slow-growing (CrxFe1x)2O3 scale has been characterized in a number of papers [5, 79, 15, 35] in the temperature range 500900C and involving several different environments and alloys. The oxide scale typically consists of an inward-growing layer consisting of spinel oxide+reaction zones and an outward-growing iron-rich layer, the two layers being separated by the original metal/oxide interface. This is in good agreement with the present observations after long exposure times, see Figs.5a, 6a, 7a and 8. The characteristic morphology may be explained by the different diffusivities of Cr3+ and Fe2+ in the spinel phase [37, 38]. Thus, Cr3+ ions diffuse several orders of magnitude slower than Fe2+ and O2 in the spinel, resulting in an outward growing hematite/magnetite layer and an inward growing spinel.
The outer layer consisted of almost pure hematite in the high oxygen activity environments and magnetite in H2+H2O, see Figs.5a, 6a, 7a, and 8. The prevalence of magnetite in contact with H2+H2O (oxygen activity about 1024 ) is expected. The hematite layer morphology was similar in all the high oxygen environments, the grain size being much smaller than for magnetite, see Figs.2, 3, 4 and . This is in agreement with earlier studies on oxidation of iron at 500C in H2O(g) and in dry air . It has been reported that the growth rate of the iron oxide layer in high oxygen activity at 500600C is accelerated by water vapor because of changes in the transport properties of the hematite layer . This may explain why the scale formed after breakaway oxidation grew slower in O2+K2CO3 environment than in the O2+H2O+KCl exposure (Figs.1, 5a, 8). In contrast, addition of small amounts of SO2 has been reported to decrease the growth rate of iron oxide scales due to formation of sulphate on top of the hematite layer, blocking the surface sites necessary for O2 reduction . Accordingly, the growth rate after breakaway oxidation in the O2+H2O+KCl+SO2 exposure was slower than in the O2+H2O+KCl exposure (compare Figs.5a, 6a; Table2). This implies that SO2 not only delays the breakdown of the protective scale by converting the corrosive KCl to non-reactive K2SO4, it also slows down the growth of the iron-rich scale after breakaway .
While the outward growing scale consisted of nearly pure iron oxide in all cases, the inward growing scale had a much more complex microstructure/composition, see Figs.5, 6, and 7. Also, the detailed investigation showed that the main microstructural features were similar, irrespective of environment/alloy grade. Thus, in all cases the inward growing scale formed after breakaway consisted of fully oxidized regions on one hand, and reaction zones consisting of mixtures of nanosize oxide particles and remaining metal, on the other (Figs.5, 6, 7, and ). This similarity suggests that the mechanisms of formation and growth of the inward-growing scale are also similar.
An examination of the inward-growing scale formed in the three high PO2 environments that caused breakaway oxidation (O2+H2O, O2+KCl+H2O and O2+KCl+H2O+SO2) shows that the new scale formed immediately below the original metal/oxide interface is always fully oxidized, consisting of FeCrNi spinel (see e.g., Figure6a). The cationic composition of this oxide layer corresponds to the alloy composition, taking into account the removal of iron by outward diffusion to form the iron oxide at the top of the scale. These observations are in good agreement with earlier studies on FeCr alloys and stainless steels exposed in O2+H2O  and it may be concluded that the formation of a fully oxidized inward growing layer is the first stage in the process of breakaway oxidation in these cases.
The observation that this oxide layer forms before the onset of internal oxidation and that it is in contact with the alloy where internal oxidation occurs, implies that it plays a seminal role for internal oxidation. Thus, it is suggested to be involved in the process of oxygen dissolution into the alloy, which is considered to be the first step in internal oxidation . A comparison of this first inward oxide layer formed in the three high PO2 environments shows that it is considerably thicker in O2+H2O  and O2+KCl+H2O exposures than in O2+H2O+KCl+SO2 environment (compare Figs.5a, 6; Table2). The observation that the presence of SO2 resulted in a thinner inward growing oxide is in accordance with the behavior of iron at 500C where traces of SO2 in O2 reportedly resulted in a reduced inward flux of oxygen through the scale .
As noted above, internal oxidation appears to be the second step in the breakaway process in the high PO2 environments. Table3 shows that the composition of the internal oxide is similar in all environments studied and on both 304L grades. Also, the phase and chemical composition of the internal oxide is consistent with earlier observations on internal oxidation of FeCr model alloys  and commercial alloys [9, 10] (Table3). According to the FeCrO phase diagram at 600C, reported in , the FeCr spinel oxide exhibits partial solid solubility, a miscibility gap appearing between 9 and 43 at.% Cr (cationic). The high Cr boundary of the miscibility gap thus more or less coincides with the composition of the internal oxide precipitates in 304L, see Table3. The situation for FeCrNi alloys is analogous, a miscibility gap appearing in the FeCrNi spinel oxide (see below, i.e., Figure9).
In comparison to the high PO2 environments, the H2+H2O environment is more effective in promoting the formation of a reaction zone (internal oxidation). Also, the reaction zone either appears immediately below the old metal/oxide interface or is separated from it by only a very thin (50100nm) oxide layer, see Fig.7a, b. It may be noted that in H2+H2O environment, this oxide layer consisted of FeCr oxide and not FeCrNi spinel as in the high PO2 environments. Figure9 shows the calculated equilibrium composition of the spinel oxide formed by an alloy with the 304L composition at 600C, as a function of O2 activity (note that only spinel oxide is shown). It shows that an oxygen activity >1024 is necessary for the Ni-containing spinel to be thermodynamically favored. In contrast, the equilibrium oxygen activity at the scale/alloy interface is certainly far below that value (e.g., the Cr2O3/Cr equilibrium oxygen activity is about 1035). This implies that, at the start of the breakaway event in the high PO2 environments, the oxygen activity at the interface between the thin protective oxide and the alloy is much higher than the oxide/stainless steel equilibrium. Conversely, the observation that Ni2+ did not enter the corresponding first inward oxide layer in H2+H2O environment implies that the oxygen activity at the interface in question is closer to the equilibrium value in that case. It may be noted that in a corroding system, the oxygen activities must deviate from the equilibrium values. Thus, the scale/alloy interface must have higher oxygen potential than the metal/oxide equilibrium while the oxygen activity at the oxide/gas interface has to be lower than that of the gas. The observations on the composition of the first inward oxide layer imply that the oxygen activity at the interface between the thin protective oxide and the alloy is higher (and hence farther from equilibrium) in the O2-containing experiments than in H2+H2O. Indeed, this is expected because the oxygen activity in the H2+H2O gas is rather low at about 1024 compared to about 1 in the high O2 activity experiments. The observation that the spinel oxide precipitated in the reaction zone did not contain Ni implies that the oxygen activity in that part of the alloy was <1024, regardless of the gas composition. The different stages of propagation following breakaway oxidation of Fe, Cr, (Ni) alloys in both high and low PO2 are summarized in Fig.10.
A schematic drawing summarizing the different stages of propagation following breakaway oxidation of Fe, Cr, (Ni) alloys in both high and low PO2. First stage of propagation after breakaway: (high PO2 environments) the new scale formed immediately below the original metal/oxide interface is always fully oxidized, consisting of FeCrNi spinel. (low PO2 environments) Ni2+ did not enter the corresponding first inward oxide layer and internal oxidation is promoted. After longer exposure times the composition of the internal oxide (RZ) is similar in all environments studied and on both 304L grades
In the high PO2 exposures, the reaction zones tend to become fully oxidized with time (O2+KCl+H2O, see Fig.5d), implying that the oxygen activity has increased so as to enable the remaining FeNi metal to oxidize. As expected, this did not happen in the H2+H2O environment (see Fig.7ac) due to the low oxygen potential in the gas. It may be noted that the fully oxidized reaction zone in Fig.5d is separated from the alloy by a layer of Cr-rich oxide situated on top of a layer of Ni rich metal. This new external oxide beneath the fully oxidized reaction zone consists of FeCr spinel oxide.
The composition of the spinel oxide precipitate in the reaction zone is close to the high Cr boundary of the miscibility gap. As noted above, it is challenging to measure the composition of these oxide inclusions and the occurrence of precipitates with higher Cr/Fe ratios cannot be ruled out. Also, the composition of the precipitates may differ from the equilibrium value because of e.g. the kinetics of nucleation and growth. Anyway, the composition measured (about 45% Cr and 55% Fe) implies that the oxygen activity in the reaction zone is high in comparison to the metal/oxide equilibrium value (see Fig.9), as expected for an alloy suffering internal oxidation. The alloy beneath the reaction zone was investigated by STEM/EELS to search for traces of oxygen in the alloy, ahead of the reaction zone. However, despite a detection level of about 1ppm in STEM/EELS  no signal from oxygen was detected.
The internal oxide particles were somewhat smaller in the H2+H2O experiment than in the other environments/alloys, see Table3. This may be related to the lower Ni content in the alloy or to the environment, see below. SEM/EDX analysis of the scale cross-sections showed that the inward growing scale, consisting of spinel oxide areas plus reaction zones, was iron-depleted compared to the alloy composition. This is because the inward-growing scale supplies iron to the outward growing iron oxide. (Note that fully oxidized areas and reaction zones cannot be resolved by SEM/EDX).
To summarize, the characteristic features in the complex microstructures observed on 304L type stainless steel after breakaway oxidation are governed by certain properties of the FeCr(Ni)O system and are more or less independent of the corrosive environment. Thus, the different diffusivities of Cr3+ and Fe2+ in the spinel phase result in a layered scale with outward growing iron oxide and inward growing spinel. Also, the thermodynamics of the FeCr spinel oxide system, including its miscibility gap; (a) explain the composition of the internal oxide particles; (b) explain why Ni enters the inward-growing spinel in H2+H2O environment but not in the high PO2 experiments.
As described above, the microstructure of the inward growing scale formed after breakaway includes regions termed reaction zones having similar microstructure in all environments, see Figs.5, 6, and 7. In H2+H2O environment these regions remain as mixed oxide/metal also after long exposure times while they tend to become fully oxidized in the high oxygen activity exposures, see Figs.5d, 7ac and . The characteristic microstructure of the reaction zones may influence the properties/oxidation kinetics in the following ways:
The remaining Fe/Ni metal between the oxide precipitates may form voids when acting as a source of ions for the outward-growing scale. Indications of this have been observed on FeCr model alloys  and can be seen in Figs.6b and 7b. Void formation may cause spallation.
The incubation time to breakaway and the growth rate after breakaway oxidation varies in the different environments, see Fig.1a, b and Table2. The fastest oxidation rate after breakaway was observed in the H2+H2O experiment, see Fig.1b. The absence of hematite on the surface in combination with the microstructure of the inward growing scale may explain this. Diffusion through magnetite is known to be faster than through hematite [44, 45]. Also, the microstructural investigation showed that the ratio: reaction zone/fully oxidized spinel oxide, was higher in H2+H2O (compare Figs.5a, 7c). The inward growing scale developed in H2+H2O therefore provides little protection.
The results imply that the general features and microstructure of the scales formed after breakaway are not directly linked to the environment that causes breakaway. Instead they are governed by the general characteristics of FeCr(Ni) alloys and oxides, i.e. the thermodynamic properties of the FeCrNiO system (Fig.9 and ) and the diffusivity of cations in the spinel [37, 38]. However, the corrosive species that trigger breakaway oxidation of stainless steels may influence the rate of oxidation after breakaway. As noted above, depending on the gas environment, the outward growing iron oxide scale forms hematite or magnetite, influencing the transport properties. Also, the properties and growth of iron oxide can influence the microstructure of the inward growing scale, e.g. as in the case of SO2 which slows down the inward flux of oxide ions, causing the reaction zones to form closer to the original metal surface. The results indicate that the presence of KCl decreased the time to breakaway because of the reaction of potassium ions with chromia in the protective scale and that it did not strongly influence the propagation stage, see Fig.1a. While it is well known that Cl is very corrosive in these environments and temperatures  significant levels of Cl were lacking in the scale after breakaway . It has been suggested that the prolonged incubation time in the presence of SO2 is explained by the conversion of reactive KCl to non-reactive potassium sulfate, resulting in the formation of less potassium chromate .
Water vapor is present in three of the four environments and is known to influence the oxidation in several ways as mentioned in the introduction. Thus, the presence of H2O may increase the growth rate of the iron oxide formed after breakaway oxidation as described above. Also, H2O has been suggested to influence the solubility and/or diffusivity of oxygen in the alloy, causing increased internal oxidation [11, 13, 17]. This study shows that, at high oxygen activity, the formation of a reaction zone (internal oxidation) does not occur directly after breakaway but is preceded by the formation of an inward-growing FeCrNi spinel oxide see Figs.5, 6, and 7. However, in H2+H2O environment it is difficult to interpret the sequence of events because internal oxidation occurred very close to the original interface, see Fig.7b.
To elucidate the role of H2O on scale microstructure in the propagation step, breakaway was induced on samples by coating with K2CO3 and exposing in dry O2 at 600C for 24 and 168h. The microstructural investigation showed that, while the scale was relatively thin it exhibited all the characteristic features of the breakaway scale in high PO2 environment as described above, compare Figs.5a, 6a, and 8. This shows that the microstructure after breakaway oxidation is not influenced by the presence of H2O in a major way, supporting the conclusion that the post-breakaway scale microstructure is governed by the inherent properties of the alloy and the FeCrNiO system as discussed above ([37, 38] and Fig.9).
In all environments the scale formed after breakaway oxidation consisted of two layers separated by the original metal/oxide interface, i.e. an inward growing spinel/reaction zone and an outward growing iron rich scale.
The general features and microstructure of the scales formed after breakaway are not directly linked to the environment/breakaway mechanism. Instead they are general for the FeCr(Ni) system governed by:
The oxide precipitates in the internally oxidized metal consisted of iron chromium spinel oxide containing about 45 % Cr and 55 % Fe (cationic). The composition implies that the oxygen activity in the alloy where precipitation occurred was high in comparison to the alloy/oxide equilibrium. The formation of reaction zones may influence the properties/kinetics in the following ways:
This work was carried out within the High Temperature Corrosion Centre (HTC) at Chalmers University of Technology. The authors are grateful to Dr. Samuel Hallstrm and Dr. Lina Kjellqvist at Thermo-Calc Software AB for discussing the Thermo-Calc calculations.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
Jonsson, T., Karlsson, S., Hooshyar, H. et al. Oxidation After Breakdown of the Chromium-Rich Scale on Stainless Steels at High Temperature: Internal Oxidation. Oxid Met 85, 509536 (2016). https://doi.org/10.1007/s11085-016-9610-7
Our automatic production line for the grinding cylpebs is the unique. With stable quality, high production efficiency, high hardness, wear-resistant, the volumetric hardness of the grinding cylpebs is between 60-63HRC,the breakage is less than 0.5%. The organization of the grinding cylpebs is compact, the hardness is constant from the inner to the surface. Now has extensively used in the cement industry, the wear rate is about 30g-60g per Ton cement.
Grinding Cylpebs are made from low-alloy chilled cast iron. The molten metal leaves the furnace at approximately 1500 C and is transferred to a continuous casting machine where the selected size Cylpebs are created; by changing the moulds the full range of cylindrical media can be manufactured via one simple process. The Cylpebs are demoulded while still red hot and placed in a cooling section for several hours to relieve internal stress. Solidification takes place in seconds and is formed from the external surface inward to the centre of the media. It has been claimed that this manufacturing process contributes to the cost effectiveness of the media, by being more efficient and requiring less energy than the conventional forging method.
Because of their cylindrical geometry, Cylpebs have greater surface area and higher bulk density compared with balls of similar mass and size. Cylpebs of equal diameter and length have 14.5% greater surface area than balls of the same mass, and 9% higher bulk density than steel balls, or 12% higher than cast balls. As a result, for a given charge volume, about 25% more grinding media surface area is available for size reduction when charged with Cylpebs, but the mill would also draw more power.