Balls casting simulation and practice for Hadfield Stell.

Marta, C. ; Doroftei, I. ; Prisacaru, G. 等

1. Introduction

The Hadfield steels have very diverse compositions, the manganese content ranging between 10% and 18%, generally over 1% C and 0.4-1% Si. They may also comprise alloy elements, i.e. 0,6-2.5% Cr, 0.9-1.2% Mo and 0.8-4% Ni. The sulphur content is limited to 0.05% and the phosphorous one to 0.07-0,11%. A very important issue for obtaining the austenitic manganese steels is the ratio [M.sub.n]/C[greater than or equal to]>10, where the Mn range is 11.50-14.0% and the C range is 1.05-1.40%, in order to obtain, after heat treatment, an austenitic structure and utilisation characteristics approaching the optimum levels. In order to obtain a fine austenitic structure and a fine dispersion of the manganese carbide, we need to correlate the carbon content with that of manganese. Consequently such a situation is obtained at a ratio Mn/C= 10 and Cr/C = 0.8... 1.92. This chemical composition has a very wide use in the casting of parts that must resist to the abrasive wear (in very hard conditions) and to corrosion (Sporea & Crainic 2005). If Mn/C>10, one favours the separation of the manganese carbide with large sizes and tending to separate at the limit of the austenite grains which is unfavourable to the shock wear resistance. The silicon is especially used as a dezoxiding and calming element and must be limited to 1% in order to avoid the favouring of the separation of carbon and then the occurrence of carbides. The alloying with chrome, molibden, nickel etc. aims at stabilising the carbides and improving the mechanical and resistance characteristics.

In the manganese steel, the chrome has the role of carbide stabiliser, through the formation of complex carbides and reduction of the risk of cracks occurrence at the knocking out of the cast parts. Moreover, the chrome contributes to the increase of the corrosion resistance of the manganese steel.

The manganese steels have good mechanical characteristics and a high resistance to wear if they have an austenitic structure as homogenous as possible,. In order to obtain it, the cast parts must be cooled very rapidly; we can accomplish that either by casting in shells, or through knocking out at 1200[degrees]C and intense cooling with water whose temperature is recommended to be 10[degrees]C, but no higher than 40[degrees]C.

If, after casting, we do not obtain a homogenous austenitic structure, the cast parts should be subjected to the heat treatment, i.e. the solution hardening, using rapid warming rates, especially in the temperature range 450-1050[degrees]C.

The literature is not generous on this subject. The research was done in mining exploitations in Romania, which demanded support for increasing the utilisation degree of the ore grinding mills and decreasing the exploitation costs. Moreover, one analysed the operation behaviour of balls made of manganese steel (Ivan&Mladen 2001) compared to that of the balls made of stainless steel, cast nickel alloy, hardened iron and pit irons highly alloyed with chromium. One found that the main factors influencing the level of the consumption of the grinding items are :

* Mechanical factors--quality of the material the balls are made of and their geometric shape, as well as the balls kinematics during the mill's operation;

* Technological factors--the quality and composition of the material subjected to grinding and also the filling degree and the type of grinding bodies in the mill load.

1.1. Mechanical factors

The observations made in these exploitations highlighted good operation behaviours of the balls made of pig iron, containing relatively little carbon--2.6% and nickel--0.42%, but over 23% chromium. Other grinding bodies used in the mining exploitations, executed by different companies through casting or forging and subjected or not to a hardening heat treatment, had a duration varying between 300 and 450 HB. One found experimentally that, as a result of internal strains and of the structures generated within the balls mass by a possible incorrectly conducted heat treatment, a large part of the balls had cracks starts, and other broke during operation.

Moreover, the "hidden" casting defects, such as airbells, porosities and inclusions, produced the same drawbacks in operation as those mentioned above. The use of the grinding bodies made of alloyed steels led to the increase of economic efficiency and to the increase of the mills occupation degree, i.e. the increase of revision intervals of the ball loading. Nevertheless, the presence of ridges, offaxings, contraction cavities and, in general, of geometrical shape imperfections through deviations from the spherical shape had a negative influence upon the grinding effect. In all situations, the above types of grinding bodies have high costs, which also increased the cost of the finite product.

In the big companies one uses very little the balls made of OAM, due to the fact that the presence of ridges, off-axings, contraction cavities and, in general, of imperfections of geometrical shape through deviations from the spherical shape had a negative influence upon the grinding effect.

From the above observations one also concluded that only a small part of the energy consumed during grinding is transformed into useful grinding work, the rest being consumed through frictions. The phenomenon can be explained by the fact that during the mill operation the balls engaged onto the surface of the lining, have, beside the ascendant motion, a rolling and sliding motion in the sense of the mill rotation. The friction phenomena that take place generate the premature wear and tear of the grinding bodies, thus explaining the high consumption of balls, signalled in the ore preparation industry.

Consequently, one analysed the casting manner and the exploitation behaviour of the balls made of austenitic manganese steel.

The main factors influencing the level of the grinding balls consumption are : the quality of the material they are made of, the geometrical shape, the heat treatments applied after casting, as well as the toughness of the materials subjected to grinding.

The casting defects such as airbells, porosities, inclusions and contraction cavities provoke the same bad effect in exploitation. The balls made of austenitic manganese steel are recommended for the grinding of the tough or very tough materials, considering their main property, i.e. the resistance to shock wear and tear due to their self-hammering capacity.

Two ball casting procedures were used--the mechanised conveyor casting and the atthe-ground casting in 6-ball moulds. The feeding was done through a central feeder, the feeding channels started from the central feeder towards the balls and was 10-mm long. The cavity was obtained at the level of the feeding channel, i.e. in the side part of the ball. In both casting variants the cast balls exhibited deviations from sphericity, as well as exterior and interior ridges and cavities. During grinding some of the balls used to break, because of the inadequate heat treatment, others, due to the cavity position (in the upper part) exhibited a non-uniform wear and tear or were even flattened. In both cases the explanation was the inappropriate elaboration and casting technology as well as the post-casting heat treatments, applied in an inadequate manner.

Despite these drawbacks in the case of the grinding of the tough or very tough materials one obtained smaller wears by using the balls made of manganese steel. In the case of the use of balls made of austenitic manganese steel, one remarked the advantage of the smaller wears and tears with a reduced specific consumption of grinding bodies, whereas the average productiveness increased due to a constant ball loading in the mill.

1.2. Technological factors

The main quality characteristics of the materials subjected to grinding, having decisively determined the balls' wear level, were their granulometry and chemical composition. A granulometry where the large material dimensions were predominant (over 25 mm at the entry) led, on the one hand, to the use of a large balls' load, which mechanically strained the mills' linings, and on the other hand created unfavourable conditions for the direct contact between the grinding bodies and lining. Consequently, all this favoured the phenomenon of load sliding in the sense opposite to the mill's rotation, increasing the wears and decreasing productiveness. The particularities of chemical compositions of the ores subjected to grinding significantly influenced the consumption of grinding bodies. The analysis of this aspect was effected based on the determinations of the chemical composition of the ores subjected to the grinding operations within two different companies. The higher ball consumption in the same types of mills was founds where the ore subjected to grinding was tougher, because of the high percentage of Si[O.sub.2].

The results obtained experimentally confirm also the conclusions of the consecrated theory of grinding in mills with balls : to each type of ore with given characteristics corresponds a certain level of consumption of the grinding bodies, for which an optimum cycle of loading is reached. Below this optimum level of the filling degree one witnesses the accenting of the phenomenon of ball's rolling in a sense opposite to the mill rotation, which leads to the decrease of productiveness and increase of wear and tear. Finally, in the case of the grinding of tough or very tough materials one obtained smaller wears by using the balls made of manganese steel. Consequently, we oriented our research on the improvement on the technology of OAM balls casting, for the grinding of tough corks in the ball mills, with diameters ranging between 60 and 130 mm. The high resistance to wear and tear, the low costs of the OAM balls compared to the cost of the balls made of stainless steel or tough pig irons with 25% Cr, the world shortage of iron-alloys encouraged the researches in the field.

The technical aspects, such as the steel contraction, of 2.5-3%, provoking large cavities and implicitly the influence of casting defects on the life duration of the balls were improved through simulations that led to the finding of certain solutions, i.e. that of concentrating the cavity to the centre of the ball (Marta 2006). The simulation was effected on 12-ball sand mould with central feeding gating. The results of the research can be applied to the entire range of OAM balls, with diameters of 60-130 mm, but it can be equally applied to other parts (Suratnam 1998). The simulation was done with the help of Magmasoft (MAGMA GIESSEREITECHNOLOGIE 2005).

2. The Magmasoft simulation programme

This simulation programme represents for the caster and designer an instrument by which he can rapidly test a great number of options and he can select the optimal combination for the improving and optimisation of the casting process. The traditional casting tests, expensive, time-consuming and causing manufacture delays, are much diminished. One avoids the cull cast pasts and assures the quality of the finite product from the very beginning of manufacture. MAGMASOFT allows, within the elaboration of a casting project, the creation of several versions containing diverse variants of the casting system, so that the improvement of the casting system is done only by means of modification of the geometry and parameters of the casting process, the analysis of results, their comparison and selection of the optimum solution. This principle is based on the idea that the improvement of the casting system can be done not only through a singular calculus, but also through several calculations. The use of the MAGMASOFT software imposes the following main compulsory stages : pre-processing of start initial data of the simulation, the calculus of the simulation with the selection of the adequate solver (simulator), post-processing with the presentation of the simulation results.

2.1. The pre-processing of the mould geometry

The first stage of simulation consists in modelling the part geometry and the casting network. In order to do this, one selects a "PRE-PROCESSOR" option from the working menu. The computerised simulation methods of the filling and solidification process require that the casting system should be made of 3D variable volumes. The realisation of the "casting system" (volume of the part, of the network, of the jet and feeder) are done according to the principle of volumes' superposing. MAGMASOFT allows the realisation of the pat geometry by several ways:

--one imports the geometry with the STL extension (rapid prototyping), done with the help of other drawing software, as CAD data files, or one builds geometries using the integrated MAGMASOFT geometry modelling;

In the pre-processor we find the display of the bi-dimensional construction views and a 3D view. An advantage of MAGMASOFT is that the standard geometries can be stocked in a data base and imported when necessary.

The important parameters for the display of results form post-processor must be already defined in the pre-processor.

[FIGURE 1 OMITTED]

Then one defines the components of the casting system according to the classes of materials, i.e.:

--Cast Alloy

--Sand Mould

--Feeder

--Gating and Inlet

[FIGURE 2 OMITTED]

This denomination allows components to be identified in the casting process, one does not name the materials, only the function of each component in the casting process. The definition of the components is done taking into account that the geometry of the part subjected to simulation, i.e. the 100-mm diameter sphere cast from the alloy to be defined as class of materials, roe precisely the G1X2[degrees]CrMn130 steel, in a sand mould, also defined as class of materials, based on theoretical studies regarding the simulation of casting in casting mixtures for balls, for 100-mm diameter spheres. The balls casting is done in 12 sand moulds fon balls, with central feeding and outlet gating for each set of 6 balls. For the simulation in which the casting system is symmetrical, like in our case, in order to reduce the simulation time, the software allows simulating only a quarter of the casting mould, according to Fig.2. The simulation is thus done for a quarter of the inlet, half of the feeder and 3 balls. The software permits the calculation of the interior volume of the mould, which is of 62853 cmc.

2.2. Simulation of solidification

In order to simulate solidification, one established control points for mould filling and part solidification, which brings simulation closer to reality. One can simulate the entire process or only one of the selected phases of the casting process. The simulation can be limited to the filling stage, to the cooling or solidification ones or only to the feeding phase. Defining the simulation parameters is based on the casters' practical experience and is specific to each type of steel and to the cast mould.

2.2.1 Selection of the materials form the database

In the database we find all the materials included which can be selected according to the needs. The components of the casting system defined according to the class of materials is the following in our case:

* cast alloy--the steel of the make T120CrMn130;

* sand mould, made of sand containing 98% Si[O.sub.2]

2.2.2. Defining the heat transfer

In order to calculate the processes of heat transfer, one will have to know the heat transfer between the classes and groups of materials, separately. This heat transfer coefficient is the extent to which at a certain moment the heat is transferred from a domain to another. The heat transfer is also selected from the data base and establishes also the type of heat transfer existing between the particular groups of materials. In the case of the sand mould, the thermal conductivity is very small. During solidification, the HTC (heat transfer coefficient) decreases due to the air space forming between the part and the mould wall. The MAGMA database uses a heat transfer coefficient between the part and the sand wall with the value ranging between 400 and 1000 W/[m.sup.2] K. In our case, the value of the heat transfer coefficient was established at a value of 500 W/[m.sup.2] K, value recommended by the MAGMASOFT database. In order to simulate the filling it is necessary to introduce the filling parameters of the casting mould. The filling process can strongly influence the following solidification profess. Then one introduces the solidification parameters. After having introduced the parameters of filling and solidification, one presses START command, having the possibility to visualise the partial results, to stop the process at the desired moment, to modify certain parameters and to continue the filling and solidification simulation.

2.3. Post-processing

After the end of simulation, the software automatically calculates the criteria selected at defined moments. The results allow the identification of defects in the itnerior of the cast part and the analysis of the solidification behaviour.

We continue by presenting some of the most important criteria, which can lead to the analysis of the filling and solidification process and of the positions of the casting defects.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

[FIGURE 13 OMITTED]

In the casting mould one installed 23 temperature control points, defined in the pre-processor. These control points (thermocouples) allow the determination of the temperature variation during solidification of steel in the casing mould. The initial temperature of casting was 1440[degrees]C, thus a 60[degrees]C overheating. According to this plotting, the knock out time of the parts from the mould can be done after around 20 minutes. The total solidification time is of 76 minutes. This plot can offer a lot of information about the temperature variation of the casting ensemble during solidification. One can also obtain plottings representing only the temperature variation in the mould filling, the variation of the filling rate or the variation of pressure in the casting mould.

3. Practical tests for the casting of the balls from 100-mm diameter spheres

[FIGURE 14 OMITTED]

As shown in the introduction, the main factors influencing the level of the grinding balls consumption are: the quality of the materials they are made of, the geometrical shape, the heat treatments applied after casting, as well as the toughness of the material subjected to grinding. The casting defects such as airbells, porosities, inclusions and contraction cavities provoke the same drawbacks in exploitation; in both casting variants used prior to the improvement of technology the cast balls had deviations from sphericity, exhibiting exterior and interior ridges and cavities. During grinding some balls used to break, because of the inadequate heat treatment, and others, due to the cavity position (at the upper part) exhibited a non-uniform wear and tear and even used to flatten. In both cases the explanation was the inappropriate elaboration and casting technology, as well as the heat treatments inappropriately applied after casting. Fig. 14 below presents a section in a ball cast with the old casting technology.

As one can remark from the photograph, the casting defect is placed on the vertical diameter of the part, the defect height being of approximately 65 mm, and the maximum diameter along the horizontal axis is of 20 mm. We continue by presenting the results of the casting technology improved according to simulation. In order to underline and establish the importance of the MAGMASOFT simulation system, we performed casting for 100-mm balls in conditions similar to those from the MAGMASOFT software. After the casting process, the balls were sectioned and they exhibited a cavity (Figure 15) similar to that obtained through simulation (Figure 16), in the case of the GX120CrMnl30 steel. The cavity has the following dimensions: along the vertical diameter--15 mm, whereas along the horizontal diameter--approx. 10 mm. The maximum cavity depth is of around 7 mm. The volume of the cavity determined also by liquid filling is of 5.5 [cm.sup.3]. The cavity volume is estimated due to the fact that one works on sectioned parts. In relation to the volume of the part, the cavity volume represents only 1.051% of the total volume of the part. One cast 1500 balls with the help of the newly designed technology. In order to verify the results, one made also measurements of the balls weight. The weight variations ranged between 4.370 kg and 4.470 kg, difference of 100 g, which confirms the validity of the casting technology. Another imposed issued refers to the sphericity of the cast balls. One made measurements along 3 diameters and the deviations ranged between +2 and -2 mm.

[FIGURE 15 OMITTED]

[FIGURE 16 OMITTED]

4. Conclusion

By using the simulation software, the traditional casting tests are much diminished. The cull rejected cast parts are avoided and the quality of the finite product is assured from the manufacture start. While elaborating a casting project one can elaborate several versions containing different conditions of the casting system, so that the improvement of the casting system is done only by modifying the geometry and parameters of the casting process, the analysis of the results, their comparison and the selection and the optimum solution, avoiding thus the doubling of work and saving time and money. The use of the technology created based on simulation does not always trigger the obtaining of perfect and cheapest parts. The simulation does not take into account the technology of the alloy elaboration, the treatments in the ladle after elaboration, the times of keeping in the mould, the technologies of extraction from the mould, and the modality of applying the corresponding heat treatment, all these being the parameters of the manufacture process. This software allows very important factors in the case of the OAM casting, such as: the temperature at the beginning of casting and casting times, the mould temperature, the materials used, the heat transfer coefficients, to be integrated into the pre-processor, and they decisively influence the results. The simulation method using MAGMASOFT and focused on the idea of placing the contraction cavity at the centre of the balls is original, and together with the obtaining of holes more reduced in value is a consequence of the casting technology used appropriately in order to obtain a spherical shape, which triggers a uniform wear of the sphere surface and an increase of the exploitation duration. The results of the simulation are valid only for the manganese austenitic steels with the following chemical composition:: C = 1.05 - 1.35, Cr = 1.50--2.50, the ratio Mn//C = 10. Moreover, one effected simulations of the behaviour of balls with interior contraction cavities, compared to the full balls.

DOI: 10.2507/daaam.scibook.2009.44

5. References

Ivan, St., & Ghiua, M. (2001). Symposium, Western University, On the duration of exploitation of wear-resistant steel of the T110Mnl30 Type.pag.400-405, ISSN 1453-7394, 9-12 November, 2006,Timisoara

Marta, C.; (2006). Theoretical and experimental studies on the casting of steel balls with high contents of manganese for ore-grinding crushers), D. P. Publishing House, ISBN code 973-30-1522-9, Bucharest

Sporea, I.; Crainic, N. & Mladen, M, (2005). On casting of austenitic manganese steel in wear-resistant parts, Annals of the University of Oradea, Mechanics fascicle, pag.479-486, ISSN 1011-2855, 12-15, May, 2005, Oradea.

Suratman R.; (1998) Alloy Design and Casting Practice of Hadfield's Manganese Steel, Metallurgical Science and Technology, Vol. 1, 1990, pag. 822-840, 1216 July, 1998, Hawaii

*** MAGMA GIESSEREITECHNOLOGIE GMBH, Copyright 2000, 2001 2002;

This Publication has to be referred as: Marta, C[onstantin]; Doroftei, I[oan]; Prisacaru, G[heorghe]; Hamat, C[odruta]; Suciu, L[enuta] & Zgardea, E[manuel] (2009). Balls Casting Simulation and Practice for Hadfield Stell, Chapter 44 in DAAAM International Scientific Book 2009, pp. 427-442, B. Katalinic (Ed.), Published by DAAAM International, ISBN 978-3-901509-69-8, ISSN 1726-9687, Vienna, Austria

Authors' data: Prof. PhD Marta, C[onstantin]*; Prof. PhD Doroftei, I[oan]**; Prof. PhD Prisacaru G[heorghe]**; Prof. PhD, Hamat C[odruta]*; Lecturer PhD Suciu, L[enuta]*; Eng. Zgardea E[manuel]*, *University "Eftimie Murgu" Resita, Romania, **University Iasi, Romania, maco@uem.ro, ioan_doroftei@yahoo.com, prisacara_ghe2004 @yahoo.com, codruta.hamat@yahoo.com, ilesuciu@yahoo.com, emanuel@uem.ro

Tab. 1. Analysis criteria for silidification and simulation results

Criterion                                  Measuring unit

GRADIENT-HEAT GRADIENT                     [degree]C/mm
COOLRATE-COOLING RATE                      [degree]C/s
LIQTOSOL-PASSAGE FROM LIQUID TO SOLID           S
SOLTIME-SOLIDIFICATION TIME                     cm
FEEDMOD-HEAT MODULE                             s
HOTSPOT-HOT SPOTS                               s
FEEDING -QUALITY OF THE MOULD FILLING           %
POROSITY-POSITION OF THE CASTING DEFECTS        %