- Email: [email protected]
Analysis of forging defects for selected industrial die forging processes
EFA-02733; No of Pages 14 Engineering Failure Analysis xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Engineering Failure Analysis journal homepage: www.elsevier.com/locate/engfailanal
Analysis of forging defects for selected industrial die forging processes Marek Hawryluk ⁎, Joanna Jakubik Wroclaw University of Technology, The Chair of Metal Forming and Metrology, ul. Łukasiewicza 5, 50-371 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 4 May 2015 Received in revised form 22 October 2015 Accepted 2 November 2015 Available online xxxx Keywords: Die forging Finite element method Forging defects
a b s t r a c t The main goal of this paper is to identify defects in forgings in selected die forging processes. The major problem is the formation of underfills due to air pockets between the forging and the tool. In the literature there is no information about modeling of such defects using FEM software, therefore, attempts were made to build numerical simulations of analyzed processes. The high agreement of the numerical modeling results with the results of the macroscopic, microstructural and defectoscopic examinations confirmed the validity of the FEM modeling assumptions and justified the use of such IT tools for the analysis of industrial plastic working processes. © 2015 Published by Elsevier Ltd.
1. Introduction Due to the high competition between forging producers, in recent years besides the price (the main consideration) the quality of the offered forged products is a factor increasingly often taken into account when choosing a supplier. This particularly applies to customers from automotive and aircraft industries where the requirements as to the forging accuracy and quality are the highest. Die forging processes belong to one of the most difficult manufacturing techniques. Even though this technology has been mastered quite well, the correct manufacture of forgings with complicated shapes (connecting rods, worm gears, constant-velocity universal joints, turbines, levers, etc.) which satisfy the customers' high quality expectations, requires much experience from the designers, technologists and machine operators [1,2]. The implementation of new forging designs, the continuous optimization of the existing technologies and the large number of factors having a bearing on the correctness of the whole process and their mutual interactions make forging processes very difficult to analyze. In each of the stages in the forging process there is a risk that an error will occur, resulting in a flaw — a forging defect. For this reason several CAD/CAM/CAE tools (usually based on FEM and physical modeling) and special measuring-control systems are used to design and optimize the whole forging process [1–9]. 2. State of the art The design of preforms and slugs for forging processes is an important element of improving product quality and lowering the production costs due to the material lost as flash or to the losses connected with incorrectly manufactured parts. Most researchers and experienced forging engineers are inclined to agree that the most common forging defects (underfills, folds) are the result of the improper geometry and/or incorrect position of the preform or the slug on the die insert. Such errors are often due to the unavailability of a particular bar section from the steel works or the lack of proper equipment resources for slug preparation. In die forging processes the proper spacing of cross-sectional areas along the length of the straight axis of the preform (slug) ⁎ Corresponding author at: ul. Łukasiewicza 5, 50-371 Wrocław, Poland. E-mail address: [email protected] (M. Hawryluk).
http://dx.doi.org/10.1016/j.engfailanal.2015.11.008 1350-6307/© 2015 Published by Elsevier Ltd.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
2
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
and the preparation of the latter through forming is highly important for the proper filling of the cavity die by the material [10–11]. Other causes of forging defects include: a too low temperature of the billet, the use of too strong drafts, the improperly made tools, incompletely removed scale, and crude technology. The forge is responsible for most of the causes of defects, but there are also factors for which the forge is not directly responsible, however it can control and supervise them in order not to allow the quality of its products to deteriorate [6–9,11–12]. There are plenty of studies and papers on the selection, design and optimization of billet geometry, but only a few works are devoted to the application of numerical FEM modeling to the analysis of the causes of forging defects. The possibilities of using numerical FEM simulations were presented in [11], where, among other things, numerical analysis of the shaping of a forging out of an ingot, with modeled casting pores, were performed. Subsequently an experiment was carried out for the same machining parameters as in the simulations. The ways in which defects propagated in the numerical model and in the physical model were compared. Numerical FEM modeling is used mainly to determine the optimal shape and dimensions of the preform and the slug. This is required when the forging has a complicated shape, as in the case of turbine blades, toothed gears, forked forgings, etc. [7,8]. Examples of alternative techniques of designing preforms/slugs, based on conventional engineering methods, are sequential analysis techniques using radiosity, the upper-bound approach, the slip-line field method and physical modeling using soft modeling materials. For instance, in [10] it is proposed to use the backward tracing method to design the proper shape of a turbine blade. In [13] a sequential technique based on the upper-bound method was used for the analysis of a preform geometry, whereby a preform shape was obtained through the selection of proper tribological conditions. The technique used by the author is approximate, making it possible to estimate the yield stress needed for plastic working processes. The total deformation power in this method is the upper limit of the power expended by the external forces. In order to determine the limit loads one must know or adopt assumptions concerning stress fields, strain velocity fields, the yield criterion and the plastic flow law. Assuming that plastic strain zone VP consists of parts displacing relative to each other as stiff bodies inside of which there is uniform velocity field vk and some of surface SF is free of load, it follows from the principle of the power equilibrium of the internal and external forces that Z F i vi dS ≤ SV
XZ
r
k
k v dS:
ð1Þ
SL
By determining the upper and lower limits one can define the interval within which the actual force is contained. The sequential methods are less accurate, but they are considerably faster than FEM, and the special program procedures can directly interpret analytical results. Still another alternative method of designing the shape of a slug is the electric field method. In [14] various slug shapes were modeled using the theoretical electric field method and the results were optimized by means of artificial neural networks. An electric field is generated between two conductors with different voltages (Fig. 1). In order to generate an electric field the initial dimensions of the billet are appropriately rescaled (usually enlarged 2–3 times) so that the final outline of the forging is within the initial outline (a cylindrical preform is assumed). Depending on the voltage applied, equipotential lines of different shape are generated. The authors used artificial neural networks to select the optimal electric field line. In [15], using the finite volume method (FVM) and the parametric design method the authors developed a new procedure for designing an optimal slug for complicated forging shapes. The authors also used [16] a combination of artificial neural networks and genetic networks to optimize the initial injection parameters. A survey of literature on optimal slug design indicates that despite the theoretical bases used, most of the methods encounter difficulties, especially at large plastic deformations. Today forges most often use numerical software based on FVM and FEM to analyze the problem connected with the improper geometry and/or position of the preform. The producers of the current computing packages equip them with ever new functions enabling even better and more complete analyses of plastic working processes, making it possible, e.g., to detect defects in forgings and to analyze the durability of the tooling (Forge, QFORM, Simufact) [17–18]. Owing to such functions the user can
Fig. 1. Equipotential lines generated between two conductors [14].
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
3
significantly shorten the time needed to implement a new project and reduce errors in tooling design. Obviously, the conventional methods of designing the charge are still used, particularly in the older forges, but even there IT tools are beginning to be used. One of such functions available in Forge 2011 is the detection of laps (the folds function) [18]. During the simulation of the forging process some areas of the element being deformed may come in contact with one another. Initially these are lines on the surface of the forging, which in the course of simulation widen and penetrate inside, precisely showing the size and depth of occurrence of the defect. Thanks to remeshing based on the adaptive superimposition of a finite element mesh, consisting in the automatic densification of the mesh in places with a complicated geometry and small corner radii and in zones in which elements come in contact with one another, the location folds is more accurate. The defects (laps) in the forging being shaped are visible as a cloud of red points (dots) in the postprocessor. The development of folds is estimated in each computing step. Moreover, the fact that the lines of material flow in the forging are taken into account contributes to a more accurate analysis of the causes of such defects. Forge 2011 also enables the detection of air pockets (the trapp function), i.e. empty spaces between the forging and the tool, in which air has been trapped. When an air pocket is detected, the computing solver calculates the pressure in it on the basis of the closed space volume and treats it as a boundary condition during simulation, in this way influencing the flow of the material and the filling of the cavity die. Initially, the pressure is calculated under the assumption that the temperature inside the close space changes in range T0–T1: 0
1
BT C B C P ¼ P0 B 1 C @T 0 A
ð2Þ
where P0 — the atmospheric pressure, T1 — the temperature of the slug, T0 — the external temperature. Also closed space volume V0 and coolant volume Vlubtrapped in the air pocket are determined. In the next step new pocket volume V1 and the internal pressure as: P¼
C V1
ð3Þ
where: C = P ⋅ (V0 − Vlub) are computed. Since the computations have an explicit character it may happen that the closed space will disappear between successive computing steps and the pressure will assume an infinite value. In order to avoid such situations default maximum pressure Pmax and default minimum pocket volume Vmin are assumed in the program. If computed volume V1 reaches a value lower than Vmin, Eq. (2) is omitted and the pressure amounts to Pmax. Air pockets can cause cavity die underfilling and premature degradation of the tools as a result of increased pressure in such places. The aim of this research was to assess the possibility of using numerical modeling results in the analysis of forging defects in selected die forging processes. 3. Range of studies The operations of forging lever and yoke elements complicated in their shape conducted in Forge Jawor were subjected to analysis. Fig. 2 shows a lever forging after the trimming operation (Fig. 2a) and an exemplary position of the preform in the
Fig. 2. a) Lever forging after trimming, b) bottom die insert with schematically shown preform position. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
4
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 3. a) Yoke forging after trimming, b) bottom die insert for preliminary operation.
bottom die insert (Fig. 2b, the yellow points are the places on which the slug, in the form of a round bar, rests). Fig. 3 shows the correct finished yoke forging after the trimming operation (Fig. 3a) and the contour cavity die in the bottom die insert (Fig. 3b). First, preliminary macroscopic, defectoscopic and microstructural examinations were carried out to detect and identify flaws in the forgings. Then numerical thermomechanical models were built for the two forging processes and numerical FEM simulations were run using the Forge 2011 computing package. Thanks to the simulations of the yoke and lever forging processes for different billet positions the forming processes could be analyzed more thoroughly and the causes of defects in the forgings could be determined through the use of the program's trapp and folds functions.
4. Identification of defects in analyzed forgings 4.1. Lever forging The macroscopic examinations of the lever forging showed numerous defects in the form of folds and cavity die underfills. A photograph of the forging with a marked underfill and a place with a lap are shown in Figs 4 and 5. A considerable underfill occurs in the lever pin (Fig. 4), where a deep and narrow cavity die in the insert considerably hinders the inflow of the material. Underfills also occur in the forging head. In addition, an extensive lap, caused by the improper flow of the material, appears in the lever foot (Fig. 5). Structural examinations of the lap in the plane perpendicular to the crack (the plane marked in Fig. 4b) showed the structure of the forging to be typical of hypoeutectoid steel, i.e. pearlitic–ferritic. Banding in the vicinity of the lap and numerous nonmetallic precipitates, which become visible in a dark field of view (Fig. 6), were observed. A preliminary EDX analysis showed that they were Fe2O3 and Fe3O4 oxides probably originating from the surface of the
Fig. 4. Underfill in lever forging.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
5
Fig. 5. Defects in lever forging: a) lap in lever foot and b) defect magnification (plane of sample cut for metallographic examinations).
forging. The occurrence of oxides and bits of scale coming from the forging surface in this place can cause further cracking and damage to the element.
4.2. Yoke forging The macroscopic examinations of the yoke forging (Fig. 7) revealed that the most common defects for this element are folds (the area marked red) and underfills (the area marked blue). During the observation of the forging process in the Jawor Forge it was noticed that the excess of the coolant (which has not evaporated from the surface of the tools) collects in the pin area (area 2 in Fig. 7) in the yoke's back part, preventing cavity die filling and increasing pressure in this area and so accelerating microcracking and producing the Rebinder effect [19]. This is a disadvantageous phenomenon since it significantly lowers the resistance to mechanical cracking (especially in the corners) of the forging tools (Fig. 8).
Fig. 6. Microstructure in lap area in lever foot: a) structure of hypoeutectoid steel — pearlite + ferrite, b) magnified lap area — visible banding and c) fine precipitates — dark field of view.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
6
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 7. Yoke forging with marked defects: 1) lap (area marked red) and 2) underfill (area marked blue). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 8a shows a set of (top and bottom) inserts for yoke forging in preliminary forging with marked and magnified (Fig. 8b) area in which cracks develop. As one can see, for a similar number of forgings the crack in the bottom die (Fig. 8c) is much deeper than the one in the top die (Fig. 8d). This can be due to the occurrence of greater thermal loads and the incomplete evaporation of
Fig. 8. Crack in corner of preliminary inserts: a) inserts for preliminary forging, b) magnified area where cracks occur, c) crack in bottom of preliminary insert and d) crack in top of preliminary insert.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
7
Fig. 9. a) Lap in yoke forging, area no. 1 from Fig. 6 and b) microstructure of material in yoke fork.
the coolant in the bottom die, which has been observed many times and confirmed by the Jawor Forge engineers and by infrared camera inspections. Also the lap area was subjected to metallographic examinations. Fig. 9 shows the place from which a sample for metallographic examinations was taken and the plane in which the material structure was examined. Similarly as in the case of the lever, the structure of the yoke is a typical hypoeutectoid steel structure. Banding can be observed in the area where folds occur, but no oxides were found to be present as in the case of the lever forging. The lap forming in this place disqualifies the forging from further production and service.
5. Modeling and numerical simulations For the purpose of a more in-depth analysis of the causes of defects, numerical simulations were carried out using the finite element method and the Forge 2011 program made by Transvalor. 3D models of the tools (the die inserts were modeled as elements with heat exchange) and the preforms were built. The technological parameters of the process were selected on the basis of the operation sheets. The rate of shift of the top die was adopted consistently with the kinematic parameters of the crank press and depended on the angular position of the crank. The ambient temperature and the temperature of the forging were assumed to amount to 30 °C and 1150 °C, respectively. The temperatures of the tools were measured by means of a pyrometer and a thermal imaging camera and amounted to about 250 °C. A friction model, which took into account the cooling-lubricating agent (an aqueous solution of graphite), was used. Carbon steel C45 and hot-work tool steel 1.2344 were used for respectively the forged material and the die inserts. The material specifications, i.e. thermal expansion, specific heat, thermal conductivity were taken from the Materials Forming Properties Database [20]. The stress–strain dependences for the proper rate of straining and the Young modulus–temperature dependence were determined through a torsion test carried out in a plastometer. The studies covered the temperatures: 650 °C, 750 °C, 850 °C, 1000 °C and 1150 °C and the strain rates: 0.1 s−1, 1 s−1 and 10 s−1. The temperatures and the strain rates were selected on the basis of an analysis of the industrial processes of forging the yoke and
Fig. 10. Comparison of forged lever shape with numerical model.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
8
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 11. Forging shapes for different values of opening between tools: a) forging moved 28 mm away from die end — lap at process end, revealed by folds function, b) forging moved 10 mm away — no folds.
the lever and the preliminary FEM simulations. The coefficients of heat exchange between the billet and the tools and with the environment were assumed to amount to respectively 30 W/mm2 K and 0.35 W/mm2 K. Hansel-Spittel equation was selected in order to determine the flow stress approximation of function:
m1T m2
σ ¼ Ae
ε
m4
• m3
eε ε
ð4Þ
where: ε equivalent strain, ε_ strain rate, s−1, T temperature, °C A, m1, m2, m3, m4 coefficients determined on the basis of plastometric tests of a material investigated. Three-dimensional simulations of forging were carried out using tetrahedral elements. Due to high deformations encountered during simulation it was necessary to regularly remesh the part. To ensure an optimal and well adopted mesh two basic criteria for initiating the remesh were used. The remeshing was launched according to a specific defined period of time and increased deformation of elements. Moreover finite element mesh of the parts has been refined according to the curvature of the die at the place of contact.
Fig. 12. Intensity of strain after preliminary forging: a) lever with lap (distance from cavity die end — 28 mm) and b) lever without lap (distance from cavity die end — 10 mm).
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
9
Fig. 13. Rate of flow during preliminary forging: a) lever with lap, b) lever without lap.
5.1. Analysis of simulation results of lever element preliminary forging In order to verify the numerical modeling assumptions the shape of the forging obtained from the real forging process conducted in Jawor Forge plc was compared with that of the numerical model obtained from the simulation (Fig. 10). The slight differences in the shape of the flash are due to the different positions of the slug on the bottom die insert during the actual forging in the Forge. A preliminary analysis of the FEM simulation results revealed that the position of the initial material has a significant influence on the proper filling of the cavity die. In the first case (Fig. 11a), when the billet is moved 28 mm away from the bottom insert cavity die, a lap appeared in the lever foot. This is caused by the curling of the material in the final stage of preliminary forging. Through the next numerical simulations, in which the distance from the end of the cavity was changed at every 2 mm, the optimal preform position was selected whereby folds no longer appeared in the lever foot. The flow of the material was significantly improved when the slug was positioned at a distance of 10 mm from the end of the bottom insert cavity die (Fig. 11b). FEM simulations showed that the further shifting of the preform towards the end of the insert would result in an underfill in the upper part of the forging (the lever head). Fig. 11a shows the stages in the appearance and growth of folds in the tested element, revealed by the folds function. The material flow shown in Fig. 11 as well as the intensity of strain (Fig. 12) and the rate of material flow (Fig. 13) for the different billet positions confirm the risk that a lap may appear in the lower part of the lever. Numerical simulations were run with the active trapp function (air pockets detection). As a result, the trapped air and lubricant volume and the pressure prevailing in this space were calculated from Eq. (3). The external pressure of 1 atm (1013 hPa) and the average lubricant thickness of 0.5 mm were assumed for the computations. An analysis of the simulation results shows that as
Fig. 14. Air pockets after lever preliminary operation, revealed thanks to trapp function.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
10
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 15. Comparison of yoke forging shape with numerical model.
the pressure in the closed space increases, it becomes more difficult to fill the die insert, which may result in underfills in the forging (Fig. 4). In the analyzed forging process the tools are cooled and lubricated with graphite spray mist fed through nozzles. It has been found that also graphite spray can collect in the places where there are surface air pockets (Fig. 14), producing the Rebinder effect [19]. Therefore it is important to reveal such zones during the design of technological process since then special attention can be given to the places in the cavity die to ensure that the manufactured product meets the accuracy and quality requirements. If air pockets appear it may be necessary to correct the shape of the tools to improve the material flow, especially in narrow and deep places in the cavity die.
5.2. Analysis of simulation results for yoke preliminary operation The other analyzed process was the hot forging of a yoke. FEM simulations with the trapp and folds functions active were run to check the correctness of the technological process with regard to the appearance of defects in the forgings. Fig. 15 shows a comparison of the forging shape with the numerical model. The intensity of strain and the temperature distribution in the forging after the preliminary operation are shown in Figs 16 and 17. The numerical simulations (with the folds function used) showed that the side flash and the front flash overlap each other whereby the risk that folds will occur at the end of the yoke fork ends arises (Fig. 18). This was confirmed by defectoscopic examinations of a randomly selected forging (Fig. 19).
Fig. 16. Intensity of strain in yoke forging after preliminary operation.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
11
Fig. 17. Temperature distribution on yoke surface after preliminary operation.
Also modeling with the use of the trapp function for detecting unfilled areas (air pockets) in the cavity die was carried out for preliminary forging. Fig. 20a shows the places where air pockets occur while Fig. 20b shows the pressures (exceeding 600 MPa in the final phase of preliminary forging) in these places. In the industrial forging process also the excess of the lubricating agent which has not evaporated from the forging surface may collect in such places.
5.3. Optimization of yoke forging process Also research on the use of FEM for optimizing the shape of the preform/slug through a change in the geometry of die inserts is being conducted [21]. Its aim is to minimize slug volume in the industrial forging process and to reduce the risk that the above defects will arise in yoke forging. In the first stage of the research, numerical FEM simulations were run for the existing process. The simulation results for the preliminary operation showed that the large dimensions of the flash are due to the premature flow of the material out of the cavity die (Fig. 21). The cause of the incorrect filling of the cavity die was the improper preparation of the slug for this operation. Then it was shown that if the shape of the slug for preliminary forging is modified it is possible to reduce the amount of the initial material and obtain full cavity die filling. For this purpose the authors proposed a modification of the tools for the preceding flattening operation, consisting in adding an oblong projection to the existing top upsetting insert and adding walls blocking the excessive lateral outflow of the material, to the bottom upsetting insert (Fig. 22). The proper spacing of the walls, their inclinations and the location of the base threshold were found to be critical for the forging after this operation to perfectly fit into the cavity die of the initial inserts. Similar optimization was performed for the top flattening insert by fitting the cross sections of the shaping projection along the length of the forging.
Fig. 18. Place where folds occur during preliminary forging of yoke (use of folds function).
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
12
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 19. Lap in yoke, revealed by defectoscopic examinations.
Fig. 20. Air pockets (trapp function) in final stage of preliminary forging of yoke: a) size of area exposed to underfilling (red) and b) distribution of pressure on forging surface. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
As a result, a better fit of the slug to the insert cavity die for the initial operation was achieved. After the analysis of the FEM simulation results technological trials of forging this element with a 10% reduction in the billet volume were carried out. The change in the geometry of the flattening tools not only contributed to the better filling of the cavity die, but also, through the
Fig. 21. Model of material deformation in preliminary operation [21].
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
13
Fig. 22. Modified tools for flattening operation.
change in the way of material flow, reduced the risk of the occurrence of folds in the yoke fork. Macroscopic and defectoscopic examinations confirmed the correctness of the changes made and the numerical FEM modeling assumptions (Fig. 23). 6. Conclusion The improper position of the preform (the lever forging simulation) and the improper flow of the material (the yoke forging simulation) have been identified as the causes of defects. The results of the structural examinations and the numerical simulations have a utilitarian character and indicate a need for identifying the causes of folds and underfills in the cavity die in hot forging processes. By eliminating forging defects or limiting their occurrence to a minimum one can significantly reduce the production costs. The identification of the causes of defects is needed to properly design the technological process and eliminate the defects. In order to properly design the forging process so that it enables the manufacturing of a series of repeatable forgings free of defects one must select optimal process parameters, properly design and manufacture (which includes the choice of a material and its thermal treatment) the tools and optimize the shape of the preform and the slug. The number and complexity of the factors having a bearing on the correctness of the forging process make their assessment difficult. Numerical modeling proves to be highly valuable for identifying heterogeneous deformations (e.g. folds) complicated in their shape, which often pass unnoticed during regular visual inspections in the forge. Thanks to FEM-based computing package FORGE 2011 one can fully analyze the industrial process of die forging and determine several parameters (the way of flow, the distribution of strains and strain rate and the thermal field) essential for the industrial forging process, which are difficult to determine experimentally or analytically.
Fig. 23. Comparison of forging shape a) before and b) after modification of tools.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
14
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
It has been demonstrated that the position of the preform has a significant influence on the proper filling of the cavity die and so on the elimination of defects in lever die forgings. Thanks to the special program (folds and trapp) functions places where defects are likely to occur have been indicated. By changing the geometry of the preform and the slug in the flattening and preliminary forging operations the amount of the billet has been minimized. The numerical modeling results were found to be in good agreement with the results of the macroscopic, microstructural and defectoscopic examinations. This confirms the validity of the adopted FEM modeling assumptions and justifies the use of such IT tools for the analysis of industrial plastic working processes.
References [1] J. Tomczak, Z. Pater, T. Bulzak, Designing of screw impressions in the helical rolling of balls, Arch. Civ. Mech. Eng. 14 (1) (2014) 104–113. [2] Z. Gronostajski, M. Kaszuba, M. Hawryluk, M. Zwierzchowski, A review of the degradation mechanisms of the hot forging tools, Arch. Civ. Mech. Eng. 14 (4) (2014) 528–539. [3] Z. Gronostajski, M. Hawryluk, M. Kaszuba, A. Niechajowicz, S. Polak, S. Walczak, D. Jablonski, Die profile optimization for forging constant velocity joint casings, Arch. Metall. Mater. 56 (2) (2008) 551–558. [4] D. Pociecha, B. Boryczko, J. Osika, M. Mroczkowski, Analysis of tube deformation process in a new pilger cold rolling process, Arch. Civ. Mech. Eng. 14 (3) (2014) 376–382. [5] V. Vazquez, T. Altan, Die design for flashless forging of complex parts, J. Mater. Process. Technol. 98 (2000) 81–89. [6] A. Kocańda, in: A. Piela, F. Grosman, J. Kusiak, M. Pietrzyk (Eds.),Określenie trwałości narzędzia w obróbce plastycznej metali 2003, pp. 213–256. [7] M. Kopernik, A. Milenin, Numerical modeling of substrate effect on determination of elastic and plastic properties of TiN nanocoating in nanoindentation test, Arch. Civ. Mech. Eng. 14 (2) (2014) 269–277. [8] G. Winiarski, A. Gontarz, Z. Pater, A new process for the forming of a triangular flange in hollow shafts from Ti6Al4V alloy, Arch. Civ. Mech. Eng. 15 (4) (2015) 911–916. [9] Gronostajski Z, Hawryluk M, Kaszuba M, Sadowski P, Walczak, S, Jablonski D. Measuring & control system in industry die forging process. Eksploatacja i niezawodnosc — Maint. Reliab. 3 (2011); 62–69. [10] J.J. Park, N. Rebelo, S. Kobayashi, A new approach to preform design in metal forming with the finite element method, Mach. Tool Des. Res. 23 (1) (1983) 71–79. [11] G. Banaszek, A. Stefanik, Theoretical and laboratory modelling of the closure of metallurgical defects during forming of forging, J. Mater. Process. Technol. 177 (1– 3) (2006) 238–242. [12] A.J. Wilby, D.P. Neale, Defects introduced into metal during fabrication and service, Adv. Mater. Sci. Eng. 3 (2012). [13] A.N. Bramley, Upper bound elemental technique (UBET) and UBET in 3D and tetrahedral upper bound analysis (TEUBA). Fast method for forging simulation and preform design, Int. J. Mater. Prod. Technol. 116 (2001) 80–83. [14] S.R. Lee, Y.K. Lee, C.H. Park, D.Y. Yang, A new method of preform design in hot forging by using electric field theory, Int. J. Mech. Sci. 44 (2002) 773–792. [15] M. Sedighi, S. Tokmechi, A new approach to preform design in forging process of complex parts, J. Mater. Process. Technol. 97 (1–3) (2005) 314–324. [16] S. Mok, C.K. Kwong, W.S. Lau, A hybrid neural network and genetic algorithm approach to the determination of injection moulding, Int. J. Adv. Manuf. Technol. 18 (2001) 404–409. [17] Simufact forming reference manual 11.0. [18] Forge 2011 Documentation — Datafile Forge 3v75. [19] L. Dobrzański, A. Dobrzańska-Danikiewicz, Zmiany struktury i własności powierzchni materiałów inżynierskich w wyniku eksploatacji, in: L. Dobrzański, Dobrzańska-Danikiewicz (Eds.),Obróbka powierzchni materiałów inżynierskich, Gliwice, vol. 5, 2011. [20] http://www.ita-tech.cz/en/products-services/materials-properties/material-database-of-forming-properties, (2015.10.12) [21] Z. Gronostajski, M. Hawryluk, J. Jakubik, M. Kaszuba, P. Kondracki, A. Niechajowicz, et al., Problemy jakościowe w procesie kucia na goraco odkuwki typu widlak, Hut. Wiad. Hutn. 81 (1) (2014) 68–71.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
Contents lists available at ScienceDirect
Engineering Failure Analysis journal homepage: www.elsevier.com/locate/engfailanal
Analysis of forging defects for selected industrial die forging processes Marek Hawryluk ⁎, Joanna Jakubik Wroclaw University of Technology, The Chair of Metal Forming and Metrology, ul. Łukasiewicza 5, 50-371 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 4 May 2015 Received in revised form 22 October 2015 Accepted 2 November 2015 Available online xxxx Keywords: Die forging Finite element method Forging defects
a b s t r a c t The main goal of this paper is to identify defects in forgings in selected die forging processes. The major problem is the formation of underfills due to air pockets between the forging and the tool. In the literature there is no information about modeling of such defects using FEM software, therefore, attempts were made to build numerical simulations of analyzed processes. The high agreement of the numerical modeling results with the results of the macroscopic, microstructural and defectoscopic examinations confirmed the validity of the FEM modeling assumptions and justified the use of such IT tools for the analysis of industrial plastic working processes. © 2015 Published by Elsevier Ltd.
1. Introduction Due to the high competition between forging producers, in recent years besides the price (the main consideration) the quality of the offered forged products is a factor increasingly often taken into account when choosing a supplier. This particularly applies to customers from automotive and aircraft industries where the requirements as to the forging accuracy and quality are the highest. Die forging processes belong to one of the most difficult manufacturing techniques. Even though this technology has been mastered quite well, the correct manufacture of forgings with complicated shapes (connecting rods, worm gears, constant-velocity universal joints, turbines, levers, etc.) which satisfy the customers' high quality expectations, requires much experience from the designers, technologists and machine operators [1,2]. The implementation of new forging designs, the continuous optimization of the existing technologies and the large number of factors having a bearing on the correctness of the whole process and their mutual interactions make forging processes very difficult to analyze. In each of the stages in the forging process there is a risk that an error will occur, resulting in a flaw — a forging defect. For this reason several CAD/CAM/CAE tools (usually based on FEM and physical modeling) and special measuring-control systems are used to design and optimize the whole forging process [1–9]. 2. State of the art The design of preforms and slugs for forging processes is an important element of improving product quality and lowering the production costs due to the material lost as flash or to the losses connected with incorrectly manufactured parts. Most researchers and experienced forging engineers are inclined to agree that the most common forging defects (underfills, folds) are the result of the improper geometry and/or incorrect position of the preform or the slug on the die insert. Such errors are often due to the unavailability of a particular bar section from the steel works or the lack of proper equipment resources for slug preparation. In die forging processes the proper spacing of cross-sectional areas along the length of the straight axis of the preform (slug) ⁎ Corresponding author at: ul. Łukasiewicza 5, 50-371 Wrocław, Poland. E-mail address: [email protected] (M. Hawryluk).
http://dx.doi.org/10.1016/j.engfailanal.2015.11.008 1350-6307/© 2015 Published by Elsevier Ltd.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
2
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
and the preparation of the latter through forming is highly important for the proper filling of the cavity die by the material [10–11]. Other causes of forging defects include: a too low temperature of the billet, the use of too strong drafts, the improperly made tools, incompletely removed scale, and crude technology. The forge is responsible for most of the causes of defects, but there are also factors for which the forge is not directly responsible, however it can control and supervise them in order not to allow the quality of its products to deteriorate [6–9,11–12]. There are plenty of studies and papers on the selection, design and optimization of billet geometry, but only a few works are devoted to the application of numerical FEM modeling to the analysis of the causes of forging defects. The possibilities of using numerical FEM simulations were presented in [11], where, among other things, numerical analysis of the shaping of a forging out of an ingot, with modeled casting pores, were performed. Subsequently an experiment was carried out for the same machining parameters as in the simulations. The ways in which defects propagated in the numerical model and in the physical model were compared. Numerical FEM modeling is used mainly to determine the optimal shape and dimensions of the preform and the slug. This is required when the forging has a complicated shape, as in the case of turbine blades, toothed gears, forked forgings, etc. [7,8]. Examples of alternative techniques of designing preforms/slugs, based on conventional engineering methods, are sequential analysis techniques using radiosity, the upper-bound approach, the slip-line field method and physical modeling using soft modeling materials. For instance, in [10] it is proposed to use the backward tracing method to design the proper shape of a turbine blade. In [13] a sequential technique based on the upper-bound method was used for the analysis of a preform geometry, whereby a preform shape was obtained through the selection of proper tribological conditions. The technique used by the author is approximate, making it possible to estimate the yield stress needed for plastic working processes. The total deformation power in this method is the upper limit of the power expended by the external forces. In order to determine the limit loads one must know or adopt assumptions concerning stress fields, strain velocity fields, the yield criterion and the plastic flow law. Assuming that plastic strain zone VP consists of parts displacing relative to each other as stiff bodies inside of which there is uniform velocity field vk and some of surface SF is free of load, it follows from the principle of the power equilibrium of the internal and external forces that Z F i vi dS ≤ SV
XZ
r
k
k v dS:
ð1Þ
SL
By determining the upper and lower limits one can define the interval within which the actual force is contained. The sequential methods are less accurate, but they are considerably faster than FEM, and the special program procedures can directly interpret analytical results. Still another alternative method of designing the shape of a slug is the electric field method. In [14] various slug shapes were modeled using the theoretical electric field method and the results were optimized by means of artificial neural networks. An electric field is generated between two conductors with different voltages (Fig. 1). In order to generate an electric field the initial dimensions of the billet are appropriately rescaled (usually enlarged 2–3 times) so that the final outline of the forging is within the initial outline (a cylindrical preform is assumed). Depending on the voltage applied, equipotential lines of different shape are generated. The authors used artificial neural networks to select the optimal electric field line. In [15], using the finite volume method (FVM) and the parametric design method the authors developed a new procedure for designing an optimal slug for complicated forging shapes. The authors also used [16] a combination of artificial neural networks and genetic networks to optimize the initial injection parameters. A survey of literature on optimal slug design indicates that despite the theoretical bases used, most of the methods encounter difficulties, especially at large plastic deformations. Today forges most often use numerical software based on FVM and FEM to analyze the problem connected with the improper geometry and/or position of the preform. The producers of the current computing packages equip them with ever new functions enabling even better and more complete analyses of plastic working processes, making it possible, e.g., to detect defects in forgings and to analyze the durability of the tooling (Forge, QFORM, Simufact) [17–18]. Owing to such functions the user can
Fig. 1. Equipotential lines generated between two conductors [14].
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
3
significantly shorten the time needed to implement a new project and reduce errors in tooling design. Obviously, the conventional methods of designing the charge are still used, particularly in the older forges, but even there IT tools are beginning to be used. One of such functions available in Forge 2011 is the detection of laps (the folds function) [18]. During the simulation of the forging process some areas of the element being deformed may come in contact with one another. Initially these are lines on the surface of the forging, which in the course of simulation widen and penetrate inside, precisely showing the size and depth of occurrence of the defect. Thanks to remeshing based on the adaptive superimposition of a finite element mesh, consisting in the automatic densification of the mesh in places with a complicated geometry and small corner radii and in zones in which elements come in contact with one another, the location folds is more accurate. The defects (laps) in the forging being shaped are visible as a cloud of red points (dots) in the postprocessor. The development of folds is estimated in each computing step. Moreover, the fact that the lines of material flow in the forging are taken into account contributes to a more accurate analysis of the causes of such defects. Forge 2011 also enables the detection of air pockets (the trapp function), i.e. empty spaces between the forging and the tool, in which air has been trapped. When an air pocket is detected, the computing solver calculates the pressure in it on the basis of the closed space volume and treats it as a boundary condition during simulation, in this way influencing the flow of the material and the filling of the cavity die. Initially, the pressure is calculated under the assumption that the temperature inside the close space changes in range T0–T1: 0
1
BT C B C P ¼ P0 B 1 C @T 0 A
ð2Þ
where P0 — the atmospheric pressure, T1 — the temperature of the slug, T0 — the external temperature. Also closed space volume V0 and coolant volume Vlubtrapped in the air pocket are determined. In the next step new pocket volume V1 and the internal pressure as: P¼
C V1
ð3Þ
where: C = P ⋅ (V0 − Vlub) are computed. Since the computations have an explicit character it may happen that the closed space will disappear between successive computing steps and the pressure will assume an infinite value. In order to avoid such situations default maximum pressure Pmax and default minimum pocket volume Vmin are assumed in the program. If computed volume V1 reaches a value lower than Vmin, Eq. (2) is omitted and the pressure amounts to Pmax. Air pockets can cause cavity die underfilling and premature degradation of the tools as a result of increased pressure in such places. The aim of this research was to assess the possibility of using numerical modeling results in the analysis of forging defects in selected die forging processes. 3. Range of studies The operations of forging lever and yoke elements complicated in their shape conducted in Forge Jawor were subjected to analysis. Fig. 2 shows a lever forging after the trimming operation (Fig. 2a) and an exemplary position of the preform in the
Fig. 2. a) Lever forging after trimming, b) bottom die insert with schematically shown preform position. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
4
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 3. a) Yoke forging after trimming, b) bottom die insert for preliminary operation.
bottom die insert (Fig. 2b, the yellow points are the places on which the slug, in the form of a round bar, rests). Fig. 3 shows the correct finished yoke forging after the trimming operation (Fig. 3a) and the contour cavity die in the bottom die insert (Fig. 3b). First, preliminary macroscopic, defectoscopic and microstructural examinations were carried out to detect and identify flaws in the forgings. Then numerical thermomechanical models were built for the two forging processes and numerical FEM simulations were run using the Forge 2011 computing package. Thanks to the simulations of the yoke and lever forging processes for different billet positions the forming processes could be analyzed more thoroughly and the causes of defects in the forgings could be determined through the use of the program's trapp and folds functions.
4. Identification of defects in analyzed forgings 4.1. Lever forging The macroscopic examinations of the lever forging showed numerous defects in the form of folds and cavity die underfills. A photograph of the forging with a marked underfill and a place with a lap are shown in Figs 4 and 5. A considerable underfill occurs in the lever pin (Fig. 4), where a deep and narrow cavity die in the insert considerably hinders the inflow of the material. Underfills also occur in the forging head. In addition, an extensive lap, caused by the improper flow of the material, appears in the lever foot (Fig. 5). Structural examinations of the lap in the plane perpendicular to the crack (the plane marked in Fig. 4b) showed the structure of the forging to be typical of hypoeutectoid steel, i.e. pearlitic–ferritic. Banding in the vicinity of the lap and numerous nonmetallic precipitates, which become visible in a dark field of view (Fig. 6), were observed. A preliminary EDX analysis showed that they were Fe2O3 and Fe3O4 oxides probably originating from the surface of the
Fig. 4. Underfill in lever forging.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
5
Fig. 5. Defects in lever forging: a) lap in lever foot and b) defect magnification (plane of sample cut for metallographic examinations).
forging. The occurrence of oxides and bits of scale coming from the forging surface in this place can cause further cracking and damage to the element.
4.2. Yoke forging The macroscopic examinations of the yoke forging (Fig. 7) revealed that the most common defects for this element are folds (the area marked red) and underfills (the area marked blue). During the observation of the forging process in the Jawor Forge it was noticed that the excess of the coolant (which has not evaporated from the surface of the tools) collects in the pin area (area 2 in Fig. 7) in the yoke's back part, preventing cavity die filling and increasing pressure in this area and so accelerating microcracking and producing the Rebinder effect [19]. This is a disadvantageous phenomenon since it significantly lowers the resistance to mechanical cracking (especially in the corners) of the forging tools (Fig. 8).
Fig. 6. Microstructure in lap area in lever foot: a) structure of hypoeutectoid steel — pearlite + ferrite, b) magnified lap area — visible banding and c) fine precipitates — dark field of view.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
6
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 7. Yoke forging with marked defects: 1) lap (area marked red) and 2) underfill (area marked blue). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 8a shows a set of (top and bottom) inserts for yoke forging in preliminary forging with marked and magnified (Fig. 8b) area in which cracks develop. As one can see, for a similar number of forgings the crack in the bottom die (Fig. 8c) is much deeper than the one in the top die (Fig. 8d). This can be due to the occurrence of greater thermal loads and the incomplete evaporation of
Fig. 8. Crack in corner of preliminary inserts: a) inserts for preliminary forging, b) magnified area where cracks occur, c) crack in bottom of preliminary insert and d) crack in top of preliminary insert.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
7
Fig. 9. a) Lap in yoke forging, area no. 1 from Fig. 6 and b) microstructure of material in yoke fork.
the coolant in the bottom die, which has been observed many times and confirmed by the Jawor Forge engineers and by infrared camera inspections. Also the lap area was subjected to metallographic examinations. Fig. 9 shows the place from which a sample for metallographic examinations was taken and the plane in which the material structure was examined. Similarly as in the case of the lever, the structure of the yoke is a typical hypoeutectoid steel structure. Banding can be observed in the area where folds occur, but no oxides were found to be present as in the case of the lever forging. The lap forming in this place disqualifies the forging from further production and service.
5. Modeling and numerical simulations For the purpose of a more in-depth analysis of the causes of defects, numerical simulations were carried out using the finite element method and the Forge 2011 program made by Transvalor. 3D models of the tools (the die inserts were modeled as elements with heat exchange) and the preforms were built. The technological parameters of the process were selected on the basis of the operation sheets. The rate of shift of the top die was adopted consistently with the kinematic parameters of the crank press and depended on the angular position of the crank. The ambient temperature and the temperature of the forging were assumed to amount to 30 °C and 1150 °C, respectively. The temperatures of the tools were measured by means of a pyrometer and a thermal imaging camera and amounted to about 250 °C. A friction model, which took into account the cooling-lubricating agent (an aqueous solution of graphite), was used. Carbon steel C45 and hot-work tool steel 1.2344 were used for respectively the forged material and the die inserts. The material specifications, i.e. thermal expansion, specific heat, thermal conductivity were taken from the Materials Forming Properties Database [20]. The stress–strain dependences for the proper rate of straining and the Young modulus–temperature dependence were determined through a torsion test carried out in a plastometer. The studies covered the temperatures: 650 °C, 750 °C, 850 °C, 1000 °C and 1150 °C and the strain rates: 0.1 s−1, 1 s−1 and 10 s−1. The temperatures and the strain rates were selected on the basis of an analysis of the industrial processes of forging the yoke and
Fig. 10. Comparison of forged lever shape with numerical model.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
8
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 11. Forging shapes for different values of opening between tools: a) forging moved 28 mm away from die end — lap at process end, revealed by folds function, b) forging moved 10 mm away — no folds.
the lever and the preliminary FEM simulations. The coefficients of heat exchange between the billet and the tools and with the environment were assumed to amount to respectively 30 W/mm2 K and 0.35 W/mm2 K. Hansel-Spittel equation was selected in order to determine the flow stress approximation of function:
m1T m2
σ ¼ Ae
ε
m4
• m3
eε ε
ð4Þ
where: ε equivalent strain, ε_ strain rate, s−1, T temperature, °C A, m1, m2, m3, m4 coefficients determined on the basis of plastometric tests of a material investigated. Three-dimensional simulations of forging were carried out using tetrahedral elements. Due to high deformations encountered during simulation it was necessary to regularly remesh the part. To ensure an optimal and well adopted mesh two basic criteria for initiating the remesh were used. The remeshing was launched according to a specific defined period of time and increased deformation of elements. Moreover finite element mesh of the parts has been refined according to the curvature of the die at the place of contact.
Fig. 12. Intensity of strain after preliminary forging: a) lever with lap (distance from cavity die end — 28 mm) and b) lever without lap (distance from cavity die end — 10 mm).
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
9
Fig. 13. Rate of flow during preliminary forging: a) lever with lap, b) lever without lap.
5.1. Analysis of simulation results of lever element preliminary forging In order to verify the numerical modeling assumptions the shape of the forging obtained from the real forging process conducted in Jawor Forge plc was compared with that of the numerical model obtained from the simulation (Fig. 10). The slight differences in the shape of the flash are due to the different positions of the slug on the bottom die insert during the actual forging in the Forge. A preliminary analysis of the FEM simulation results revealed that the position of the initial material has a significant influence on the proper filling of the cavity die. In the first case (Fig. 11a), when the billet is moved 28 mm away from the bottom insert cavity die, a lap appeared in the lever foot. This is caused by the curling of the material in the final stage of preliminary forging. Through the next numerical simulations, in which the distance from the end of the cavity was changed at every 2 mm, the optimal preform position was selected whereby folds no longer appeared in the lever foot. The flow of the material was significantly improved when the slug was positioned at a distance of 10 mm from the end of the bottom insert cavity die (Fig. 11b). FEM simulations showed that the further shifting of the preform towards the end of the insert would result in an underfill in the upper part of the forging (the lever head). Fig. 11a shows the stages in the appearance and growth of folds in the tested element, revealed by the folds function. The material flow shown in Fig. 11 as well as the intensity of strain (Fig. 12) and the rate of material flow (Fig. 13) for the different billet positions confirm the risk that a lap may appear in the lower part of the lever. Numerical simulations were run with the active trapp function (air pockets detection). As a result, the trapped air and lubricant volume and the pressure prevailing in this space were calculated from Eq. (3). The external pressure of 1 atm (1013 hPa) and the average lubricant thickness of 0.5 mm were assumed for the computations. An analysis of the simulation results shows that as
Fig. 14. Air pockets after lever preliminary operation, revealed thanks to trapp function.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
10
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 15. Comparison of yoke forging shape with numerical model.
the pressure in the closed space increases, it becomes more difficult to fill the die insert, which may result in underfills in the forging (Fig. 4). In the analyzed forging process the tools are cooled and lubricated with graphite spray mist fed through nozzles. It has been found that also graphite spray can collect in the places where there are surface air pockets (Fig. 14), producing the Rebinder effect [19]. Therefore it is important to reveal such zones during the design of technological process since then special attention can be given to the places in the cavity die to ensure that the manufactured product meets the accuracy and quality requirements. If air pockets appear it may be necessary to correct the shape of the tools to improve the material flow, especially in narrow and deep places in the cavity die.
5.2. Analysis of simulation results for yoke preliminary operation The other analyzed process was the hot forging of a yoke. FEM simulations with the trapp and folds functions active were run to check the correctness of the technological process with regard to the appearance of defects in the forgings. Fig. 15 shows a comparison of the forging shape with the numerical model. The intensity of strain and the temperature distribution in the forging after the preliminary operation are shown in Figs 16 and 17. The numerical simulations (with the folds function used) showed that the side flash and the front flash overlap each other whereby the risk that folds will occur at the end of the yoke fork ends arises (Fig. 18). This was confirmed by defectoscopic examinations of a randomly selected forging (Fig. 19).
Fig. 16. Intensity of strain in yoke forging after preliminary operation.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
11
Fig. 17. Temperature distribution on yoke surface after preliminary operation.
Also modeling with the use of the trapp function for detecting unfilled areas (air pockets) in the cavity die was carried out for preliminary forging. Fig. 20a shows the places where air pockets occur while Fig. 20b shows the pressures (exceeding 600 MPa in the final phase of preliminary forging) in these places. In the industrial forging process also the excess of the lubricating agent which has not evaporated from the forging surface may collect in such places.
5.3. Optimization of yoke forging process Also research on the use of FEM for optimizing the shape of the preform/slug through a change in the geometry of die inserts is being conducted [21]. Its aim is to minimize slug volume in the industrial forging process and to reduce the risk that the above defects will arise in yoke forging. In the first stage of the research, numerical FEM simulations were run for the existing process. The simulation results for the preliminary operation showed that the large dimensions of the flash are due to the premature flow of the material out of the cavity die (Fig. 21). The cause of the incorrect filling of the cavity die was the improper preparation of the slug for this operation. Then it was shown that if the shape of the slug for preliminary forging is modified it is possible to reduce the amount of the initial material and obtain full cavity die filling. For this purpose the authors proposed a modification of the tools for the preceding flattening operation, consisting in adding an oblong projection to the existing top upsetting insert and adding walls blocking the excessive lateral outflow of the material, to the bottom upsetting insert (Fig. 22). The proper spacing of the walls, their inclinations and the location of the base threshold were found to be critical for the forging after this operation to perfectly fit into the cavity die of the initial inserts. Similar optimization was performed for the top flattening insert by fitting the cross sections of the shaping projection along the length of the forging.
Fig. 18. Place where folds occur during preliminary forging of yoke (use of folds function).
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
12
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
Fig. 19. Lap in yoke, revealed by defectoscopic examinations.
Fig. 20. Air pockets (trapp function) in final stage of preliminary forging of yoke: a) size of area exposed to underfilling (red) and b) distribution of pressure on forging surface. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
As a result, a better fit of the slug to the insert cavity die for the initial operation was achieved. After the analysis of the FEM simulation results technological trials of forging this element with a 10% reduction in the billet volume were carried out. The change in the geometry of the flattening tools not only contributed to the better filling of the cavity die, but also, through the
Fig. 21. Model of material deformation in preliminary operation [21].
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
13
Fig. 22. Modified tools for flattening operation.
change in the way of material flow, reduced the risk of the occurrence of folds in the yoke fork. Macroscopic and defectoscopic examinations confirmed the correctness of the changes made and the numerical FEM modeling assumptions (Fig. 23). 6. Conclusion The improper position of the preform (the lever forging simulation) and the improper flow of the material (the yoke forging simulation) have been identified as the causes of defects. The results of the structural examinations and the numerical simulations have a utilitarian character and indicate a need for identifying the causes of folds and underfills in the cavity die in hot forging processes. By eliminating forging defects or limiting their occurrence to a minimum one can significantly reduce the production costs. The identification of the causes of defects is needed to properly design the technological process and eliminate the defects. In order to properly design the forging process so that it enables the manufacturing of a series of repeatable forgings free of defects one must select optimal process parameters, properly design and manufacture (which includes the choice of a material and its thermal treatment) the tools and optimize the shape of the preform and the slug. The number and complexity of the factors having a bearing on the correctness of the forging process make their assessment difficult. Numerical modeling proves to be highly valuable for identifying heterogeneous deformations (e.g. folds) complicated in their shape, which often pass unnoticed during regular visual inspections in the forge. Thanks to FEM-based computing package FORGE 2011 one can fully analyze the industrial process of die forging and determine several parameters (the way of flow, the distribution of strains and strain rate and the thermal field) essential for the industrial forging process, which are difficult to determine experimentally or analytically.
Fig. 23. Comparison of forging shape a) before and b) after modification of tools.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008
14
M. Hawryluk, J. Jakubik / Engineering Failure Analysis xxx (2015) xxx–xxx
It has been demonstrated that the position of the preform has a significant influence on the proper filling of the cavity die and so on the elimination of defects in lever die forgings. Thanks to the special program (folds and trapp) functions places where defects are likely to occur have been indicated. By changing the geometry of the preform and the slug in the flattening and preliminary forging operations the amount of the billet has been minimized. The numerical modeling results were found to be in good agreement with the results of the macroscopic, microstructural and defectoscopic examinations. This confirms the validity of the adopted FEM modeling assumptions and justifies the use of such IT tools for the analysis of industrial plastic working processes.
References [1] J. Tomczak, Z. Pater, T. Bulzak, Designing of screw impressions in the helical rolling of balls, Arch. Civ. Mech. Eng. 14 (1) (2014) 104–113. [2] Z. Gronostajski, M. Kaszuba, M. Hawryluk, M. Zwierzchowski, A review of the degradation mechanisms of the hot forging tools, Arch. Civ. Mech. Eng. 14 (4) (2014) 528–539. [3] Z. Gronostajski, M. Hawryluk, M. Kaszuba, A. Niechajowicz, S. Polak, S. Walczak, D. Jablonski, Die profile optimization for forging constant velocity joint casings, Arch. Metall. Mater. 56 (2) (2008) 551–558. [4] D. Pociecha, B. Boryczko, J. Osika, M. Mroczkowski, Analysis of tube deformation process in a new pilger cold rolling process, Arch. Civ. Mech. Eng. 14 (3) (2014) 376–382. [5] V. Vazquez, T. Altan, Die design for flashless forging of complex parts, J. Mater. Process. Technol. 98 (2000) 81–89. [6] A. Kocańda, in: A. Piela, F. Grosman, J. Kusiak, M. Pietrzyk (Eds.),Określenie trwałości narzędzia w obróbce plastycznej metali 2003, pp. 213–256. [7] M. Kopernik, A. Milenin, Numerical modeling of substrate effect on determination of elastic and plastic properties of TiN nanocoating in nanoindentation test, Arch. Civ. Mech. Eng. 14 (2) (2014) 269–277. [8] G. Winiarski, A. Gontarz, Z. Pater, A new process for the forming of a triangular flange in hollow shafts from Ti6Al4V alloy, Arch. Civ. Mech. Eng. 15 (4) (2015) 911–916. [9] Gronostajski Z, Hawryluk M, Kaszuba M, Sadowski P, Walczak, S, Jablonski D. Measuring & control system in industry die forging process. Eksploatacja i niezawodnosc — Maint. Reliab. 3 (2011); 62–69. [10] J.J. Park, N. Rebelo, S. Kobayashi, A new approach to preform design in metal forming with the finite element method, Mach. Tool Des. Res. 23 (1) (1983) 71–79. [11] G. Banaszek, A. Stefanik, Theoretical and laboratory modelling of the closure of metallurgical defects during forming of forging, J. Mater. Process. Technol. 177 (1– 3) (2006) 238–242. [12] A.J. Wilby, D.P. Neale, Defects introduced into metal during fabrication and service, Adv. Mater. Sci. Eng. 3 (2012). [13] A.N. Bramley, Upper bound elemental technique (UBET) and UBET in 3D and tetrahedral upper bound analysis (TEUBA). Fast method for forging simulation and preform design, Int. J. Mater. Prod. Technol. 116 (2001) 80–83. [14] S.R. Lee, Y.K. Lee, C.H. Park, D.Y. Yang, A new method of preform design in hot forging by using electric field theory, Int. J. Mech. Sci. 44 (2002) 773–792. [15] M. Sedighi, S. Tokmechi, A new approach to preform design in forging process of complex parts, J. Mater. Process. Technol. 97 (1–3) (2005) 314–324. [16] S. Mok, C.K. Kwong, W.S. Lau, A hybrid neural network and genetic algorithm approach to the determination of injection moulding, Int. J. Adv. Manuf. Technol. 18 (2001) 404–409. [17] Simufact forming reference manual 11.0. [18] Forge 2011 Documentation — Datafile Forge 3v75. [19] L. Dobrzański, A. Dobrzańska-Danikiewicz, Zmiany struktury i własności powierzchni materiałów inżynierskich w wyniku eksploatacji, in: L. Dobrzański, Dobrzańska-Danikiewicz (Eds.),Obróbka powierzchni materiałów inżynierskich, Gliwice, vol. 5, 2011. [20] http://www.ita-tech.cz/en/products-services/materials-properties/material-database-of-forming-properties, (2015.10.12) [21] Z. Gronostajski, M. Hawryluk, J. Jakubik, M. Kaszuba, P. Kondracki, A. Niechajowicz, et al., Problemy jakościowe w procesie kucia na goraco odkuwki typu widlak, Hut. Wiad. Hutn. 81 (1) (2014) 68–71.
Please cite this article as: M. Hawryluk, J. Jakubik, Analysis of forging defects for selected industrial die forging processes, Engineering Failure Analysis (2015), http://dx.doi.org/10.1016/j.engfailanal.2015.11.008