Increase the capacity of continuous annealing furnaces at Ovako
Viktor Dahlqvist December 17, 2012
School of industrial Engineering and Management
Department of Materials Science and Engineering
Royal Instityte of Technology
SE-100 44 Stockholm
Abstract
The capacity of soft annealing of low alloyed tubes at Ovako’s continuous anneal- ing furnaces have been evaluated by comparing how it is done today with information from published and internal articles in the subject. It was found that it is possible to reduce the cycle time by 30 % for one furnace, 55 % for one furnace and 72 % for two furnaces. Two separate full scale tests were made to assess whether the faster soft annealing procedure was feasible. The tests were performed without any recon- struction of the furnace and were made by continuously vary the speed of the batch inside the furnace. The temperature in the batch was measured and compared with results from computer simulations of the heating/cooling sequences. The computer simulations were performed in COMSOL.
The soft annealing was evaluated according to the SEP-1520 standard, which means evaluating the microstructure and hardness. The results show that the faster heat treatment could yield lower grades than today but still meet it’s requirements.
In order to achieve this increase a reconstruction of the furnaces is needed and the reconstruction is brifly treated in the report. Ideas to further increase the speed of the soft annealing procedure are also presented.
Keywords: soft annealing, spheroidization, heat treatment, OVAKO, capac- ity increase, productivity, continuous furnace, normalization, isothermal annealing, quench and tempering, tube mill, COMSOL, simulation.
Abstract
Kapaciteten för mjukglödgning av låglegerade rör i Ovakos kontinuerliga glödgn- ingsugnar har utvärderats genom att jämföra dagens körsätt med information från publicerade- och interna artiklar på området. Det har konstaterats att det finns möjligheter att reducera värmebehandlingstiden med 30 % för en ugn, 55 % för en ugn och 72 % för två ugnar. Två separata test i full produktionsskala gjordes för att se huruvida den snabbare cykeln var genomförbar och gav tillräckliga bra resultat.
Testet gjordes utan ombyggnation av ugnen genom att istället kontinuerligt variera hastigheten på lasten som var inuti ugnen. Temperaturen i lasten mättes under ex- perimentet och jämfördes med resultat från datorsimuleringar av samma ugn för att se hur väl ugnens värmning- och kylningskapacitet gick att simulera. Simuleringarna utfördes i COMSOL.
Mjukglödgningen utvärderades i enlighet med SEP-1520 standarden vilket bety- der att mikrostrukturen betygssätts och hårdheten testas. Resultaten från den utvärderingen visar att den snabbare värmebehandlingen ger något sämre resultat än vad som erhålls idag, men fortfarande inom kravspecifikation. För att åstad- komma denna kapacitetsökning krävs en ombyggnation av ugnen. Ombyggnationen avhandlas något i rapporten.
Nyckelord: Mjukglödgning, sfäroidisering, värmebehandling, OVAKO, kapacitet- sökning, produktivitet, kontinuerlig ugn, normalisering, etappglödgning, seghärd- ning, rörvalsning, COMSOL, simulering.
Preface
This master thesis is performed at Ovako AB, Tube & Ring Department located in Hofors, Sweden. The range of this work is 30 hp, which correspond to 20 weeks of study. The thesis is the final part of a master degree in materials design and engineering at the Materials Science and Engineering Department at the Royal Institute of Technology, Stockholm.
The main purpose of this work was to increase the capacity of the continuous annealing furnaces at Ovako. It has been an instructive project and I have learned that there can be quite big differences between the laboratory and the large scale production environment.
It is a good experience to do the thesis in the industry that I would recommend for everyone, even those going for an academic career.
Viktor Dahlqvist 2012-12-17
Table of content
Contents
1 Introduction 1
1.1 Project motivation . . . . 1
1.2 Ovako AB . . . . 1
1.3 Tube milling . . . . 1
1.4 Objectives . . . . 2
1.4.1 The heat treatment cycle . . . . 2
1.4.2 Charging/loading . . . . 2
1.4.3 New heat treatments . . . . 2
1.5 Limitations . . . . 2
2 Theoretical Background 4 2.1 Continuous furnaces . . . . 4
2.1.1 Loading the furnaces . . . . 5
2.2 Furnace atmosphere . . . . 5
2.3 Soft annealing/spheroidization . . . . 9
2.3.1 Soft annealing today . . . . 9
2.3.2 Soft annealing tomorrow . . . . 14
2.4 Furnaces’ current set up . . . . 15
2.4.1 Today’s capacity . . . . 17
2.5 Possible new heat treatments . . . . 18
2.5.1 Isothermal annealing . . . . 18
2.5.2 Normalization . . . . 19
2.5.3 Quench and tempering . . . . 20
3 Increased Capacity 22
3.1 Proposal 1 - equalize the rate . . . . 22
3.1.1 Requirements . . . . 23
3.2 Proposal 2 - maximize the rate . . . . 23
3.2.1 Requirements . . . . 24
4 Experimental procedure 29 4.1 The material . . . . 29
4.2 Simulating a faster heat treatment without reconstruction of furnace . . . 29
4.3 Computer simulation . . . . 30
4.4 Analysis of microstructure . . . . 32
5 Results 33 5.1 Loading of furnaces . . . . 33
5.2 Simulating a faster heat treatment without reconstruction of furnace . . . 33
5.3 Computer Simulation . . . . 37
6 Discussion 39 6.1 Simulating a faster heat treatment without reconstruction of furnace . . . 39
6.2 Computer Simulation . . . . 40
6.3 How to use the capacity increase . . . . 40
6.4 Possibilities for further capacity increase . . . . 41
7 Conclusions 43
8 Future work 44
9 Appendix 48
1 Introduction
1.1 Project motivation
Keeping Ovako in a leading position and competitive in Europe with long special steel products require constant research and development of new application areas. But it doesn’t matter how good the products are if they can’t be manufactured in a financially sustainable way. Therefore it is as important to maintain a constant process development and strive to always have lower production costs than the competitors.
In the Ovako group there are a large number of annealing furnaces with similar set up and potential. It is believed that there is a potential for capacity increase in the combined system of furnaces that can be used for cust cutting and/or capacity increase, which in turn can be used for implementing new heat treatments. To clarify, this project have a potential for both process development and development of new application areas.
1.2 Ovako AB
Ovako produces engineering steel used in the bearing, transport and manufacturing indus- tries. Production comprises bars, tubes, rings and other pre-components in low-alloyed carbon steels. The market is mainly long products for demanding applications such as ball bearings and rock drills. Ovako has 11 production sites located in Sweden and Finland and customers are found mainly in Europe but also in North America and Asia.
Ovako has its roots in small steel manufacturers in Sweden, founded over 300 years ago. Since then, Ovako has been owned by SKF and Rautaruukki amongst others and most recently by Pampus Indistrie Beteiligungen until August 2010 when private equity investor Triton acquired the bar, bright bar and tube & ring business areas, excluding a wire division, forming the Ovako as it is today.
In 2011 Ovako generated sales of EUR 1121 million and had 3239 full time equivalent employees at year end. They have an annual production capacity of about 1.3 million tonnes and the current CEO is Tom Erixon. This thesis was carried out at the Tube and Ring - department located in Hofors, Sweden.
1.3 Tube milling
Ovako in Hofors has a scrap based steelmaking and melts the scrap in an electric arc furnace. After a series of treatments the liquid steel is cast as ingots and when completely solidified, transported to a rolling mill. In the rolling mill the ingots are rolled to round
bars and then cut into suitable lengths. The cut up round bars are put in a rotating kiln and heated to 1100-1250 oC in order to get soft enough for penetration. Penetration is made in the tube mill by having the round bar forced by cone formed rolls up over a plug.
The rolls press on the tube, which creates a tensile tension in the center that makes the penetration go easier. When the tubing is complete and the penetration plug is removed the wall thickness is formed in an assel mill. The next step is the reduction mill where the tube gets its final outer diameter. After reduction the tube is straightened in a two roller cross roll straightener before ending up on a cooling bed.
1.4 Objectives
The main objectives for this project was to increase the capacity of the continuous an- nealing furnaces. This were done by evaluating the heat treatment cycle time and tem- perature and the batch loading. Also possible new heat treatments were evaluated. The main outputs are listed below.
1.4.1 The heat treatment cycle
1. What quantified capacity increase can be achieved in the targeted 4 continuous annealing furnaces by minimizing the cycle time?
2. What changes to be done to achieve the maximum capacity increase?
1.4.2 Charging/loading
1. What quantified capacity increase can be achieved by optimize the charging?
2. What changes to be done to achieve the maximum capacity increase?
1.4.3 New heat treatments
1. What new heat treatment cycles can be added by utilizing the freed capacity?
1.5 Limitations
• The study was not focused on changing the temperatures in the temperature cycle since this require extensive evaluation.
• The furnaces chosen to be studied were furnace 9, 10, 13 and 14, which are contin- uous furnaces.
• The heat treatment studied was soft annealing since these furnaces almost exclu- sively perform that heat treatment.
• Ovako have many steel grades but this study focused on one steel grade only, namely Ovako 803, which is a low alloyed high carbon steel made mostly for bearings.
2 Theoretical Background
2.1 Continuous furnaces
At Ovako in Hofors the soft annealing is performed in four different continuous furnaces that all use a protective gas atmosphere. The furnaces are called furnace 9, 10, 13 and 14 in the report. In a continuous furnace the batch is loaded in one end of the furnace. The batch then moves along a conveyor that consists of rotating rolls in parallel, and comes out on the other end, where the batch is unloaded. A schematic 2D view from the side of a continuous furnace can be seen in figure 1.
The furnaces are 39 m or 47 m long and have 12 or 18 different temperature zones inside.
There are gates both before and after the actual furnace body, each about 15 m long.
Their function is to exchange the air atmosphere back and forth for a protective gas, in this case N2. In the gate by the furnace entrance an overpressure is made, which is kept throughout the furnace to avoid any other gas leaking into the furnace.
The temperature is continuously measured inside the furnace by thermocouples near the inside walls. Due to the location of the thermocouples the measured temperature does not necessarily reflect the temperature of the material being heat treated. The temperature data is stored in databases, which enables evaluation in retrospect. If the temperature difference between two zones is significant, a temperature barrier can be installed to make sure the heat stays in the warmer zone. The design of a temperature barrier vary but a screen of brick inside the furnace is one way to do it.
As can be seen in figure 1, the material to be heat treated is inserted to the left and transported by rotating rolls into the furnace, where it gets heated. The orange coloured circles above and beneath the material inside the furnace are the combustion tubes that consist of SiC. Furnace 9 and 10 run solely on propane gas while furnace 13 and 14 run on propane gas in the first 10 zones and electricity in zones 11 to 18. The propane is combusted inside combustion tubes and the combustion gas then is transported out through the combustion tubes again in order to avoid contamination of the protective gas atmosphere inside the furnace. More information about protective gas atmosphere and why combustion gas can contaminate it is given in section 2.2. The last 8 zones in furnace 13 and 14 are heated by having electricity is sent through coils mounted on the furnace inside walls, which get heated due to electrical resistance.
The soft annealing cycle requires an intermediate cooling step, which is done in a ’cooling zone’, in the middle of the furnace. The cooling is done by letting room tempered air flow through the combustion tubes in that zone. The air will absorb heat and transport it out from the furnace by cooling the combustion tubes. This way, the protective gas atmosphere is unaffected.
In a large scale production the margin of error is higher compared to a controlled lab, especially for heat treatments. The furnaces at Ovako are built in the 60’s or 70’s, which was before the revolution of computerized control systems. Even though the furnaces have been continuously upgraded to have most functions computer controlled, there still have to be margins of error when trying to optimize the soft annealing.
This description of a continuous furnace regards the furnaces at Ovako in Hofors. There are many other designs of continuous furnaces.
Figure 1: Illustrates schematically how a continuous furnace is designed. The tube is inserted to the left and transported by a conveyor line consisting of rotating rolls into the furnace. The orange dots are combustion tubes that generates the heat. The tube then exits the furnace to the right. The gates can be seen as the smaller areas before and after the heating zone.
2.1.1 Loading the furnaces
The charging of the furnaces is done by operators on the floor by lifting the tubes with an overhead crane. The tubes are picked up from either the cooling bed or from temporary storage pallets and placed on a queue-bed, which is connected to the conveyor line that runs through the furnace.
The size of the charging is limited by a number of parameters. The width and height of the furnace, load capacity of conveyor rolls, load capacity of the overhead crane and of the heat treatment. A matlab program that calculates how many tubes fit per batch depending on these limitations has been written and can be found in appendix A.
2.2 Furnace atmosphere
When performing heat treatments on steel it can sometimes be important to have a spe- cific atmosphere inside the furnace [1, 2]. Specific atmosphere means that the composition of the gas inside the furnace is chosen to have specific thermodynamic properties. The
atmosphere can be either inert or active where inert means no reactions will take place while active means reactions will take place. When it comes to steel, inert atmospheres are used to maintain the chemical composition of the surface by not having any reactions take place. The most commonly used gas for this application is N2, but sometimes Ar or He is used, see table 1. However, it is difficult and expensive to upgrade Ovako’s furnaces to be completely impenetrable, which means that other gases, such as moist (water) or CO/CO2from the combustion gas, can leak into the furnace. An active gas can therefore be used to maintain a reducing/oxidizing state and/or a certain C activity.
A reducing state is required when oxidation of the surface of the steel is to be avoided.
Reducing state can be seen as that the atmosphere oxidize faster than the steel and thus, the steel will not be oxidized. This can be done by adding components to the atmosphere that tend to react with O2, for instance CO or H2.
Table 1: How different gases affect the properties of the atmosphere[2].
Neutral gas Active gas
Reducing Oxidizing Decarburizing Carburizing
Ar H2 H2O H2O CO
N2 CO CO2 CO2
He O2 O2 CnHm
Steel contains from low levels of C, up to two percent, and the mechanical properties of the steel are strongly connected to the C concentration[3], see figure 2. Some gases have the tendency to steal C from the surface (decarburize the surface) during heat treatment, why it is important to have an atmosphere with a certain C activity. By having the same C activity on the surface and in the atmosphere one can make sure that no decarburization will take place. If the atmosphere is internally equilibrated then its C activity is related to the concentration ratio of CO2/CO where a lower ratio increases the C activity [2], see eq 1.
ac= K·PCO2 PCO2
(1)
Figure 2: Illustrates hardness for steel with different C concentration after different types of heat treatments. The Y-axis to the left is HV (vickers hardness) and the right is HRC (rockwell hardness). The x-axis is carbon concentration in weight-percent. The three lines represent from the top, hardened, annealed and spheroidized heat treatments. [3]
Active gases can also be used to on purpose affect the chemical composition of the surface.
Sometimes different properties in the core and the surface are desired, which can be done by affecting the chemical composition of the surface during the heat treatment. Often it is desired to have a tough core and a very hard surface. This can be achieved by carburization, nitridization or carbonitridization, which means that the activity of C and/or N2is higher in the atmosphere than on the surface, which will lead to an increase of dissolved C and/or N in the surface, of a low carbon steel. This results in a core of a softer low C steel with a hard surface with high concentration of C and/or N.
The chemical properties of gases are highly temperature dependent, which can be related to the thermodynamic quantity Gibbs energy, see eq 4, where ∆Go, R and K are constants (at constant pressure and temperature). By analyzing the value of the Gibbs energy of the system one can determine what phases or components are stable, i.e. what reactions that are likely to take place in a system [3]. For instance if the piece that is to be heat treated is wet when placed in the furnace, the furnace atmosphere will be affected by the evaporation of water. Whether the gas then has oxidizing or reducing properties can be
determined by the Gibbs energy. The term Rln(K) stems from the entropy where R is the gas constant and K is defined by eq. 3 below. Its value at equilibrium is called the equilibrium constant and can be evaluated by setting up the reaction, see equation 2-3.
n1A + n2B⇔ n3C + n4D (2)
Introducing K;
K = anA1·anB2
anC3·anD4 (3)
where aij are the activities, we may write the change in Gibbs energy for reaction 2 as
∆G = ∆Go + RTln(K) (4)
At equilibrium ∆G = 0 and the equilibrium constant thus is exp(-∆Go/RT) .
At Ovako in Hofors, most furnaces have no controlled atmosphere, which means they run in air. Although they have four furnaces where N2 serves as neutral gas and cracked methanol serves as active gas to steer the carbon potential upwards. These four furnaces are used for soft annealing of low alloyed high carbon steels that easily get decarburized and oxidized. Equations 5-8 describes how to calculate the ratio of P2CO/PCO2 required to not decarburize the surface depending on the amount of carbon in the steel, Xc. The reactions are assumed to be in Equilibrium (∆G = 0).
2CO⇔ CO2+ C (5)
∆G = ∆Go+ RT·ln(aC·PCO2
PCO2 ) = 0 (6)
aC= γC·XC = exp� -∆Go RT
� PCO2 PCO2
(7)
For low C concentrations, Henry’s law can be assumed to be valid, which means that γC → γoC (becomes constant).
XC= exp�-∆Go
RT
� γoC
PCO2 PCO2
= Const·PCO2 PCO2
(8)
Equations 9-11 describes how to calculate the partial pressure of oxygen inside the furnace depending on the ratio of (PCO2/PCO)2.
CO2⇔ CO +1
2O2 (9)
∆G = ∆Go+ RT·ln(
�PO2·PCO
PCO2
) = 0 (10)
PO2 = exp� -2·∆Go RT
� �PCO2
PCO
�2
(11)
Equations 12-14 describes how to calculate what partial pressure of oxygen is required to avoid oxidation of the surface.
F e +1
2O2⇔ F eO (12)
∆G = ∆Go+ RT·ln( aF eO
aF e·� PO2
) = 0 (13)
If Fe and FeO are assumed to be pure solids, one may set aF eO = aF e= 1.
�1 PO2
= exp� -∆Go RT
�
= Const (14)
By having a lower PO2 than the equilibrium reaction above states, the reaction in eq 12 tend to go to the left, which means that oxidation is avoided.
It is also possible to do this by put in values in eq 4, and if ∆G > 0 the reaction will not happen, if ∆G = 0, the reaction is in equilibrium and if ∆G < 0 the reaction will be spontaneous until ∆G becomes zero.
2.3 Soft annealing/spheroidization
2.3.1 Soft annealing today
Soft annealing is a way of heat treating steel to make it softer [1, 2, 4, 5] . Since soft annealing increases machinability it is a common heat treatment at steel production sites
because steels often are to be worked on in some way after leaving the steel plant. After a material is machined and the shape is near end-product it is often heat treated again to get mechanical properties more suitable for a specific application. For instance, Ovako produce rings for ball-bearings. Ball-bearings need to have a carved groom to keep the balls in position and that is done by machining. However, ball-bearings need to be hard and tough in order to have a long lifetime but such a material is very difficult to machine.
Therefore Ovako soften the rings and the ball-bearing producer harden them, all by heat treatment.
Soft annealing is performed on tool steels and low alloyed high carbon (roughly > 0.35 wt% C) steels. If the steel have a low carbon concentration it will get too soft after a soft annealing, which actually decreases machinability because the material only will smear out instead of breaking into small chips during machining.
During soft annealing of steels with a pearlitic structure, the cementite lamellae will spheroidize, hence the process is sometimes called spheroidization. The spheroidized cementite will be embedded in a soft ferritic matrix, which is the key to get a softer material. Figure 3 illustrates schematically the microstructural development during soft annealing. At Ovako in Hofors, the spheroidization procedure is mainly performed on high carbon steels.
Figure 3: SEM micrographs of pearlite (a) before and after spheroidization at 700 °C for (b) 10 min, (c) 60 min and (d) 360 min.[8]
Spheroidization can be performed in numerous ways [1, 2, 4, 6, 7]. However, the basic idea is to heat the material to just below or above the austenitization temperature, depending on the C content, for about an hour and then cool the material down to about 650 oC with a cooling rate of 10 oC/h. Due to the slow cooling rate spheroidization is a time consuming process and can easily become a bottle neck in the production.
The combination of high temperature and long heat treatment times can make decarbur- ization become a problem, which was the case at Ovako in Hofors for about 20 years ago.
The eutectoid point in the phase diagram for Fe - C, see figure 5, is at about 0.8 wt-% C, which means that if the depleted surface contain less than 0.8 wt-% C it will transform into austenite/ferrite instead of austenite/cementite during heating. Therefore no cemen- tite nuclei will be present at top temperature in the surface region and if there is no nuclei present, the surface material will transform into hard lamellar pearlite during cooling. A
new heating cycle was then introduced[9], that differs somewhat from the conventional cycle. This is the cycle used today and it is illustrated schematically in figure 4.
The hypereutectoid carbon steels at Ovako are soft annealed by initial partial austeniti- zation just above the A1 line in the Fe - C phase diagram, see figure 5. The main reason for this is to dissolve the extensive amount of carbide network that form during the slow cooling that follow the tube milling, i.e. the as-rolled structure consist of thick carbides that need to be dissolved. At this temperature, austenite and cementite are stable, which leads to initial spheroidization of cementite and dissolution of carbon into the austenite.
The holding time is 1-1.5 h. The material is then cooled down to 690 oC in 1 h in order to get a fully pearlitic structure at the surface. Then the material is heated again, this time to 770 oC in order to spheroidize the pearlite at the surface as well. Finally the material is cooled down to 690 again with a cooling rate of about 20oC/hto get further spheroidization.
Figure 4: A schematic illustration of the soft annealing cycle used at Ovako.
Ovako follow the SEP-1520 standard for soft annealing results, The analysis refers to as- sessing hardness and three different microstructural properties by the SEP 1520 standard.
The properties to be measured are carbide size (CG), pearlite amount (PA) and carbide
network (CN) where CG measures the size of the spheroidized cementite, PA measures the remaining amount of pearlite and CN measures the remaining carbide network in the sample.
The SEP 1520 standard includes pictures of microstructures that are graded regarding CG, PA and CN. CG is graded from 2.0 to 2.5 where 2.0 have the smallest grains and 2.5 the largest. PA is graded from 3.0 to 3.3 where 3.0 are abscense of pearlite and 3.3 have a substantial amount. CN is graded from 4.0 to 4.6 where 4.0 are abscence of carbide network and 4.6 are substantial amount. The microstructure to be evaluated are then graded by using a light optical microscope (LOM), to find the most corresponding pictures to the microstructure. The average result of today can be seen in table 2.
Table 2: The average results of Hardness, Carbide size, Pearlite amount and Carbide Network today.
Hardness [HV] CG PA CN
Value 190 2.3 3.0 4.0
The times and temperatures in the soft annealing procedure is not carved in stone but an effect of the specification by the customer. Some customers have high demands that may require longer times while others have lower demands. In the production however, it is more difficult to have specific times and temperatures for each specification than to have a specific time and temperature for each steel grade, which might lead to overdoing the soft annealing in some cases, but the bar have to be set to meet the highest demands.
However, as the quality of the steel gets better and better, the properties of the machining tools get better, which tend to decrease the demands on the soft annealing. Therefore it is important to evaluate the time and temperature of soft annealing now and then to avoid overdoing and thereby be as cost effective as possible. It is also very important that the process is stable over time so that the soft annealed material have the same properties regardless when soft annealed due to the customers optimizing their processes to fit a narrow range in the material’s properties.
Figure 5: Phase diagram of iron-carbon.[14]
2.3.2 Soft annealing tomorrow
Numerous ideas of how to spheroidize carbon steels more effectively are presented in the literature [1, 4, 7, 10, 11, 12]. The most promising results applicable to bulk production are achieved by either manipulating the as-rolled structure or by using multiple heating- cooling cycles.
The results from [1] show that the as-rolled structure, i.e. the initial structure, have a high influence on the soft annealing results. It reveals that the finer the pearlite the better results, where a fully martensitic structure gave the best results. The underlying causes of this are not discussed in the report.
The results from [7, 12] show that by using a cyclic heat treatment the soft annealing times can be reduced. A cyclic heat treatment means that the temperature cycles around the austenitization temperature several times, schematic illustrations of cyclic heat treatments can be seen in figure 6. Saha et al [7] concludes, regard using a cyclic heating-cooling cycle, for a 0.6 wt-% C steel:
1. “The fragmentation of lamellae is augmented by the dissolution of cementite through atomic diffusion from preferred sites of ‘lamellar fault’ in cementite to the adja- cent austenite during short-duration holding above Ac3 temperature. The non- equilibrium forced air cooling generates more lamellar fault regions that act as the potential sites for spheroidization.“
2. “With increasing number of heat treatment cycles the proportion of ferrite and spheroidized cementite increases, the proportion of lamellar pearlite decreases and microconstituents (pearlite and ferrite) become finer.”
3. “After 8 heat treatment cycles the microstructure mostly contains spheroidized ce- mentite particles in very fine ferrite matrix ... with an excellent combination of strength and ductility properties.”
Figure 6: A schematic illustration of potentially faster soft annealing. Red line cycles between two temperatures while the blue line continuously decreases temperature with each cycle. At present it is unknown which is most effective.
2.4 Furnaces’ current set up
The schemes for soft annealing in furnace 9, 10, 13 and 14 can be seen in table 3-5. Notice that the specifications only consider the actual furnace body and not the gates in front of and after. This means that the soft annealing procedure takes approximately three hours longer in reality than what is stated in the tables, however the actual heat treatment is made within the time specified in the tables.
Table 3: The specifications of furnace 9 and how it is run during soft annealing.
Furnace 9
Zone Temperature [oC] Zone length [m] Length [m] Time at 4 m/h [h]
1 810 3.4 3.4 0.839
2 820 3.1 6.5 1.62
3 820 2.2 8.6 2.16
4 820 2.2 10.8 2.7
5 670 4.3 15.1 3.78
6 800 3.6 18.7 4.68
7 800 3.6 22.3 5.58
8 750 3.6 25.9 6.48
9 720 2.9 28.8 7.2
10 710 2.9 31.7 7.92
11 700 2.9 34.6 8.64
12 690 4.3 38.9 9.72
Table 4: The specifications of furnace 10 and how it is run during soft annealing.
Furnace 10
Zone Temperature [oC] Zone length [m] Length [m] Time at 4 m/h [h]
1 810 4.4 4.4 1.1
2 820 2.2 6.6 1.64
3 820 2.2 8.8 2.18
4 670 4.3 13.1 3.26
5 740 2.9 16.0 3.98
6 800 5.8 21.8 5.42
7 800 5.8 27.6 6.86
8 750 2.9 30.5 7.58
9 720 5.0 35.5 8.84
10 710 5.0 40.5 10.1
11 700 4.3 44.8 11.18
12 690 3.0 47.8 11.92
Table 5: The specifications of both furnace 13 and 14 since they are identical, and how they are run during soft annealing.
Furnace 13 and 14
Zone Temperature [oC] Zone length [m] Length [m] Time at 3.6 m/h [h]
1 815 3 3 0.83
2 820 3.4 6.4 1.78
3 820 1.7 8.1 2.25
4 820 4 12.1 3.36
5 670 3.2 15.3 4.25
6 750 1.3 16.6 4.61
7 770 5.7 22.3 6.19
8 770 2 24.3 6.75
9 735 1.3 25.6 7.11
10 730 2.5 28.1 7.81
11 725 1.7 29.8 8.28
12 720 2.4 32.2 8.94
13 - 2.3 34.5 9.58
14 - 2.5 37 10.28
15 - 2.4 39.4 10.94
16 - 2.3 41.7 11.58
17 - 2.4 44.1 12.25
18 - 2.9 47.1 13.06
2.4.1 Today’s capacity
The capacity will be calculated by multiplying the maximum weight allowed per length (kg/m), which is set by the conveyor rolls’ mechanical properties, with the speed of the furnace (m/h). The outcome is [kg/h] and the capacities can be seen in table 6.
Table 6: Displays the maximum capacity of the furnaces.
Furnace 9 10 13 14
Maximum weight [kg/m] 750 750 750 750
Speed [m/h] 4 4 3.6 3.6
Capacity [kg/h] 3000 3000 2700 2700
2.5 Possible new heat treatments
2.5.1 Isothermal annealing
Low alloyed, often case hardened, low carbon steels (< 0.3 wt% C) are frequently used in the automotive industry for instance for cogwheels or other transmission details[2]. These applications require material with high homogenity and good machinability, which can be achieved with isothermal annealing.
Isothermal annealing means initial austenitization above the A1line, followed by quench- ing or at least fast cooling down to the transition temperature below the A1line where the material is maintained for several hours. Isothermal annealing leads to a ferrite-pearlite structured material with good chip breaking properties, which is good for machinability.
A schematic curve of isothermal annealing can be seen in figure 7.
A continuous furnace suits well for isothermal annealing since it requires a rapid cooling between two high temperatures. Isothermal annealing can be done at Ovako today, how- ever the furnace it is done in have no cooling abilities. It is a continuous furnace that is similar to furnaces 9, 10, 13 and 14 but have only two zones; one high temperature (HT) zone and one low temperature (LT) zone. The material is austenitized in the HT zone and then brought out to air cool before placed in the LT zone to have phase transformation take place. This makes the procedure very operator dependent, which tend to create big variations in the process.
The furnace doing isothermal annealing today has no ability to control the atmosphere.
This is a problem since austenitization means high temperatures, sometimes up to around 960oC, which means the surface get decarburized and oxidized fast.
Figure 7: A schematic illustration of an isothermal annealing.
2.5.2 Normalization
Normalization is a heat treatment performed mainly on carbon steels and low alloyed steels[2]. The idea is to get smaller grains and a more homogeneous grain size, which makes the material harder and enable for more homogeneous properties. Material that has been worked on will not have a homogeneous mictrostructure, but will consist of both large and small grains and can have a variation of structures such as carbide precipitates and bainite. A normalization heat treatment removes these inconveniences and thereby lead to better mechanical properties as well as increased machinability.
Initially the material is austenitized in 800 - 920 oC with a short holding time. This generates newly formed homogeneous austenite grains that dissolves carbides and other unwanted structures. The austenitization is followed by a controlled cooling to room temperature. The cooling rate will determine the size of the newly formed ferrite-pearlite grains and also the thickness of the lamellae in the pearlite. A schematic illustration of a normalization procedure can be seen in figure 8.
A continuous furnace might be considered overqualified for normalization, but it is good to have the ability to vary the cooling rate depending on the specification of the customer.
It is also good to have the ability to use a protective gas atmosphere in order to not affect the chemical composition of the surface.
Normalization is done at Ovako today, in the same furnace as the isothermal annealing and it thus yields the same difficulties and limitations.
Figure 8: A schematic illustration of a normalization heat treatment.
2.5.3 Quench and tempering
It is of great interest for steel producers to find applications for their products where no intermediary is needed. The intermediaries are is most often subcontracting engineering companies performing machining operations. If the intermediary can be cut out there is more profit to gain. In order to do so the steel producer must be able to perform finaliz- ing heat treatments and since most applications require materials with good mechanical properties (high hardness, toughness etc.) the heat treatment must be hardening. Quench and tempering is such a hardening heat treatment.
There are several ways to harden steels, but for low alloyed steels the alternatives are fewer where the easiest and most common is to create a martensitic or bainitic structure.
Martensite is created by initial austenitization followed by quenching while bainite is created by initial austenitization followed by rapid cooling. Often these structures are so hard that the material becomes brittle. To get a higher toughness the material is tempered after quenching in about 550-700oC for a few hours. This procedure enables a hardness of around 250 HB - 350 HB[2]. A schematic illustration of a quench and tempering procedure can be seen in figure 9.
Quench and tempering are sometimes used on semi-finished products that are to be machined. Materials that have been quench and tempered often undergo some kind of case-hardening, for instance nitridization, carburization or the combination nitrocarbur- ization.
Today, quench and tempering can not be done at Ovako in Hofors, but it can be done at Ovako Bright Bar at Hällefors.
Figure 9: A schematic illustration of quench and tempering.
3 Increased Capacity
Evaluation of todays set-up and running scheme enabled finding a potential for capacity increase. In this report, two ideas for capacity increase will be presented. The first proposal will increase the capacity by making all 4 furnaces run equally fast compared to their length. The second proposal sets as high speed as possible and have greater potential to increase the capacity substantially.
3.1 Proposal 1 - equalize the rate
As mentioned earlier furnace 9 is shorter than furnace 10, 13 and 14. However, furnace 9 and 10 run of the same speed while 13 and 14 run on lower speed. The soft annealing procedure require only a certain time in the furnace and since furnace 9 is shorter the time for each batch in that furnace are shorter than in the other furnaces. Since all furnaces have the same heating/cooling capabilities, logically it should be possible to run faster in furnace 10, 13 and 14 compared to furnace 9.
The new speed can be calculated by:
• Furnace 9 runs in 4 m/h and is 38.9 m long. This makes a total soft annealing time of 38.9 m /4 m/h = 9.7 h.
• Furnace 10 run in 4 m/h and is 47.7 m long. This makes a total soft annealing time of 47.7 m /4 m/h = 12 h.
• Furnace 13 and 14 runs in 3.6 m/h and are 47.1 m long. This makes a total soft annealing time of 47.1 m /3.6 m/h = 13 h.
If it is assumed that furnaces 10, 13 and 14 require only as long time as in furnace 9 then:
• Furnace 10 can be run in 47.7 m /9.7 h = 4.9175 m/h ≈ 5 m/h
• Furnace 13 and 14 can be run in 47.1 m /9.7 h = 4.86 m/h ≈ 5 m/h If this would work, then the speed increases are:
• For furnace 10: 5 m/h /4 m/h = 1.25, i.e. 25 % increase.
• For furnaces 13 and 14: 5 m/h /3.6 m/h = 1.39, i.e. 39 % increase.
And the total maximum capacity increaes are:
• For furnace 10: (5 m/h · 750 kg/m) - (4 m/h · 750 kg/m) = 750 kg/h
• For furnace 13 and 14: (5 m/h · 750 kg/m) - (3.6 m/h · 750 kg/m) = 1050 kg/h
• Total capacity increase: 750 kg/h + 2·1050 kg/h = 2850 kg/h
• Percentage increase: 2850 kg/h /(2·3000 kg/h + 2·2700 kg/h) = 0.25 = 25 % increase
3.1.1 Requirements
In order to achieve this increase the furnaces have to be reconstructed. The main thing is that the cooling zone have to be moved backwards approximately 4-5 m in furnace 10 and about 1 m in furnace 13 and 14, in order for the initial austenitization to be complete. The new zone lengths, temperatures and times can be seen in tables 7-9 and an illustration of the heating cycle can be seen in figure 10.
Moving the cooling zone and remake zone lengths means moving combustions tubes and cooling tubes. This means new holes have to be drilled in the furnace. The exact number of new holes and the details of the new combustion/cooling tube set-up require must be evaluated further. Also the last 6 zones that run on electricity in furnace 13 and 14 have to be started up.
3.2 Proposal 2 - maximize the rate
Different test results combined with how the furnace are run today are used to optimize the cycle in proposal 2, which set to be as fast as possible.
Soft annealing requires, according to [15]:
• 1.5 h heating time to reach 820oC(from measurements made in furnace 13)
• 1.5 h holding time at 820 oC (to dissolve carbide networks)
• 2 h cooling after second heating, from 770 oC to 690 oC. This makes 40 oC/h (report states maximum 50 oC/h)
According to current set-up, soft annealing requires:
• 1 h at the intermediate cooling step, from 820oC to 670oC.
• 1.5 h heating after intermediate cooling step, from 670oCto 770oC. Holding time at 770oCincluded.
This gives a theoretical cycle of total 7.5 h. The speed can then be calculated by dividing the furnace length by the cycle time, 39 m /7.5 h = 5.2 m/h for furnace 9, 47.7 m /7.5 h
= 6.36 m/h for furnace 10 and 47.1 m /7.5 h = 6.28 m/h for furnace 13 and 14. 6.2 m/h is chosen for furnace 10, 13 and 14.
If this speed is used, then the speed increases are:
• For furnace 9: 5.2 m/h /4m/h = 1.3, i.e. 30 % increase
• For furnace 10: 6.2 m/h /4 m/h = 1.55, i.e. 55 % increase.
• For furnaces 13 and 14: 6.2 m/h /3.6 m/h = 1.72, i.e. 72 % increase.
And the total maximum capacity increaes are:
• For furnace 9: (5.2 m/h · 750 kg/m) - (4 m/h · 750 kg/m) = 900 kg/h
• For furnace 10: (6.2 m/h · 750 kg/m) - (4 m/h · 750 kg/m) = 1650 kg/h
• For furnace 13 and 14: (6.2 m/h · 750 kg/m) - (3.6 m/h · 750 kg/m) = 1950 kg/h
• Total capacity increase: 900 kg/h + 1650 kg/h + 2·1950 kg/h = 6450 kg/h
• Percentage increase: 6450 kg/h /(2·3000 kg/h + 2·2700 kg/h) = 0.57 = 57 % increase
3.2.1 Requirements
In order to achieve this increase the furnaces have to be reconstructed. The main thing is that the cooling zone have to be moved backwards approximately 10 m in furnace 10 and about 6 m in furnace 13 and 14, in order for the initial austenitization to be complete. The new zone lengths, temperatures and times can be seen in tables 10-12 and an illustration of the heating cycle can be seen in figure 11.
Moving the cooling zone and remake zone lengths means moving combustions tubes and cooling tubes. This means new holes have to be drilled in the furnace. Exactly how many must holes that must be drilled and how the new combustion/cooling tube set-up should be must be further evaluated. Also the last 6 zones that run on electricity in furnace 13 and 14 have to be started up.
Table 7: Specifications of how the reconstruction will affect the zoning in furnace 10.
Zone Length [m]4 m/h Length [m]5 m/h Length [m]6.2 m/h
1 4.4 4.4 4.4
2 2.2 2.2 2.2
3 2.2 2.2 2.2
4 4.3 6.3 4.3
5 2.9 5.5 5.5
6 5.8 3.2 6.2
7 5.8 4.8 5.8
8 2.9 2.9 2.9
9 5.0 4.0 2.0
10 5.0 5.0 5.0
11 4.3 4.3 4.3
12 3.0 3.0 3.0
Table 8: Specifications of how the reconstruction will affect the times per zone in furnace 10.
Zone T ime [h]4 m/h T ime [h]5 m/h T ime [h]6.2 m/h
1 1.10 0.88 0.71
2 1.64 1.31 1.06
3 2.18 1.75 1.41
4 3.26 3.01 2.10
5 3.98 4.11 2.99
6 5.42 4.75 3.99
7 6.86 5.71 4.93
8 7.58 6.29 5.39
9 8.84 7.10 5.72
10 10.10 8.11 6.52
11 11.18 8.97 7.22
12 11.92 9.55 7.70
Table 9: Specifications of how the reconstruction will affect the tem- perature in furnace 10.
Zone 4 m/hT emp.[o C] 5 m/hT emp.[o C] 6.2 m/hT emp.[o C]
1 810 810 810
2 820 820 820
3 820 820 820
4 670 820 820
5 740 670 820
6 800 800 670
7 800 800 800
8 750 760 800
9 720 725 780
10 710 710 730
11 700 700 700
12 690 690 690
Table 10: Specifications of how the reconstruction will affect the zoning in furnace 13 and 14.
Zone Length [m]3.6 m/h Length [m]5 m/h Length [m]6.2 m/h
1 3 3 3
2 3.4 3.4 3.4
3 1.7 1.7 1.7
4 4 4 4
5 3.2 1.2 2.5
6 1.3 5.3 4
7 5.7 2.7 3
8 2 2 3
9 1.3 2.3 1.3
10 2.5 2.5 2.5
11 1.7 1.7 2.2
12 2.4 2.4 1.9
13 2.3 2.3 2.8
14 2.5 2.5 2
15 2.4 2.4 2
16 2.3 2.3 2
17 2.4 2.4 2
18 2.9 2.9 2
Table 11: Specifications of how the reconstruction will affect the times per zone in furnace 13 and 14.
Zone T ime [h]3.6 m/h T ime [h]5 m/h T ime [h]6.2 m/h
1 0.83 0.60 0.48
2 1.78 1.28 1.03
3 2.25 1.62 1.31
4 3.36 2.42 1.95
5 4.25 2.66 2.35
6 4.61 3.72 3.00
7 6.19 4.26 3.48
8 6.75 4.66 3.97
9 7.11 5.12 4.45
10 7.81 5.62 4.85
11 8.28 5.96 5.21
12 8.94 6.44 5.52
13 9.58 6.90 5.97
14 10.28 7.40 6.29
15 10.94 7.88 6.61
16 11.58 8.34 6.94
17 12.25 8.82 7.26
18 13.06 9.42 7.58
Table 12: Specifications of how the reconstruction will affect the tem- perature in furnace 13 and 14.
Zone T emp. [3.6 m/hoC] T emp. [5 m/hoC] T emp. [6.2 m/hoC]
1 815 820 815
2 820 820 820
3 820 820 820
4 820 820 820
5 670 820 820
6 750 670 820
7 770 800 670
8 770 800 670
9 735 800 800
10 730 780 800
11 725 760 800
12 720 745 800
13 - 730 760
14 - 720 735
15 - 710 720
16 - 705 710
17 - 700 700
18 - 690 690
!
Figure 10: Illustration of the soft annealing cycles at different speeds in furnace 10 com- pared to furnace 9 with speed of 4 m/h.
Figure 11: Illustration of the soft annealing cycles at different speeds in furnace 13 and 14 compared to furnace 9 with speed of 4 m/h.
