INFLUENCE OF TECHNOLOGICAL PARAMETERS ON THE SPRINGBACK ANGLE OF HIGH-STRENGTH STEELS

The most common problem in area of sheet metal forming technology is a problem of achieving accurate and repeatable shapes of drawn and bent parts. This phenomenon is caused by the elastic springback. Springback can be defined as an elastically-driven change of shape of the deformed part upon removal of external loads. Several technological parameters influence amount of springback, between these belong friction coefficient, blankholder force, different geometry of tools, etc. In this paper is presented this topic, and also numerical simulation of this process. For an experimental process were used two high-strength steels. Steel with TRIP effect and dual phase steel DP 600. Numerical simulation was performed in the static implicit code Autoform. Results were compared and discussed.


Introduction
The most frequent defects that occur in the case of parts made by cold plastic forming of metal sheets are caused by the springback phenomenon.The main effects of springback on the formed parts are as follows: the alteration of the part form and dimensions after tools removing and also the change of the state of stresses and strains in the deformed material [1,2].In the past, handy tables [3,4] or graphs [5] were the traditional ways used for prediction of springback.Nowadays, numerical simulation of springback is frequently used tool to the springback prediction.Although it seems that this method is nowadays accurate enough, there were observed lot of problems associated with this method, especially when AHS or UHS steels are used.Current problems associated with numerical simulation of springback are detail described in several publications [6,7].The determination, elimination or avoidance of the defects generated by springback requires the analysis and knowledge of its specific causes and its relation with different factors of influence [8].Some theoretical and experimental research studies have considered the influence of blankholder force (BHF) and die geometry upon the springback and side wall curl [9], whereas, other studies have ignored the above effect [10].Some studies were devoted to the influence of the tool geometry to the springback amount [11], material properties [12], die gap [13], die radius [14], bending radius [15,16] etc.Also friction condition on the springback can be considered from the same point of view with blankholder force since higher friction condition increases the tension applied to the sheet and decreases the required bending moment for a given curvature [17], therefore, friction condition influence the springback.Following remarks were drawn on this topic [2]:  Springback in the U-drawn part increased with the punch radius, depending on the blankholder force. Side wall curl decreased with the die radius but depending greatly on blankholder force too. The high value of friction reduces springback and side wall curl by increasing tension in the wall.This is also dependent on blankholder force. Springback ratio increases as the normal anisotropic value increases, or as the strain hardening exponent decreases.Influence of friction coefficient (f) and blankholder force to the springback amount were investigated in the U bending with stretching process.This process was also modelled in the static implicit finite element code Autoform.A yield function that can describe orthotropic properties of the material has been proposed by the Hill'48 as an extension of the von Mises criterion [18].Hill'48 yield criterion was set in the combination with hardening curve defined by Hollomon [19].In this paper, high strength steels DP 600 and TRIP steel were used, because these steels are very often used in the automotive industry.Their usage is associated with the great problems of springback [20,21].

Experimental materials and methods
Aim of this experiment was to investigate influence of different blankholder force and different friction condition on the amount of springback.Blanks were cut in direction parallel with rolling direction with dimensions 150 x 20 mm.Dual phase steel DP 600 with 1.00 mm thickness and steel with TRIP effect RAK 40/70 with 0.75 mm thickness were used.Mechanical properties of used steels are in the Table 1 and Table 2. R.D. in tables means rolling direction.Geometry of used tools is illustrated in the Fig. 1.Gap between punch and die was greater than thickness of the sheet, therefore there were provided bending without ironing.Radius of the die and punch edge is R2.Two different friction coefficients were used in order to determine influence of the friction coefficient to the springback amount.In the Table 3 is presented combination of different technological parameters of this process.Various combination presented in the Table 3 were combined with blanks made of TRIP steel and DP steel.The effect of BHF on springback was investigated at two different conditions, i.e. 50N and 10 kN.The effect of friction condition on springback was investigated at two boundary conditions, where f was 0.2 and 0.05.The value of friction adequate to the value 0.2 was reached by the drawing without lubrication and the value 0.05 was reached by the using of teflon film.First of all, results of U-drawn parts of TRIP steel will be presented.

Discussion
In the first part, TRIP steel will be discussed.In all cases, numerical simulation underestimated results of real experiment.This may be caused by the transformation of the material structure itself, where transformation of the residual austenite cannot be described simple in numerical simulation.As was illustrated in the previous chapter, it was proved, that in the case of lower blankholder force, lower coefficient of friction decreases amount of springback.On the other hand, in the case of greater blankholder force, greater friction coefficient decreases amount of springback.A high coefficient of friction decreases springback by increasing tension in wall of the U-shape part.From the results implies, that the smallest springback occurred in the case where the friction coefficient was 0.2 and blankholder force was set to the 10 kN.As can be seen in the results illustrated in the Fig. 4 -Fig.7, differences between results of numerical simulation varied in the small scale.In the area B, there was straight die wall.Up to the last case where the friction coefficient was equal to the value 0.2 and BHF equal to the 10 kN, this straight die wall caused gain of springback amount.
Greater consistency between results of numerical simulation and real experiment happened in the case of DP steel.Results were similar to the results of TRIP steel in terms of friction and holding force influence.If the holding force is set to the 50 N, smaller coefficient of friction decrease amount of springback.If the holding force was set to the greater value, smaller greater friction coefficient caused smaller springback.Higher friction helps to prevent relative sliding between the sheet and tool surfaces and thus avoids scratching the blank surface.Smallest springback occurred in the case of 10 kN holding force with friction equal to the value 0.2.In this case, in area of straight wall of die a smaller springback occurred.Smaller amount of springback was observed in the case of using DP steel.Since TRIP steel has a higher strength, greater springback occurs.Numerical simulation of springback of high-strength steels springback is very sensitive to the parameters set in the numerical simulation software.The size of the shell elements should be not smaller than 0.75 of the blank thickness, but it is recommended to set size of elements maximally only to fulfil the condition of minimum 10 elements per radius of the tool.Velocity of the tool should be small, in order to avoid dynamical effects.Mesh of the tools should have value 10 mm.It is recommended to use material models which take in to the account mixed kinematic hardening models such as Chaboche and Yoshida.These material models are crucial in springback modelling, especially when non-conventional steels are used.

Conclusion
Based on results presented in this paper could be noted that the forming of high strength steels is associated with springback problems largely.Friction condition and different blankholder force influence the amount of springback significantly.In the case of lower blankholder force, lower coefficient of friction decreases amount of springback.Greater blankholder force in combination with greater friction coefficient decreases amount of springback.Numerical simulation was in most cases not accurate enough.This problem may be caused by the using conventional material models, which did not take into the account variable Young modulus, Bauschinger effect, kinematic hardening model etc.This entails future work which includes new mechanical test such is cyclic tension-compression test which is necessary to perform on the sheets in order to achieve values of different parameters for definition of mixed hardening kinematic models.

Fig. 1 2
Fig. 1 Geometry of used tool Fig. 2 Dimmensions of the die

Fig. 3
Fig. 3 Schematic view of parameters for springback measurement

Fig. 4 U
Fig. 4 U-drawn part after springback: a) for DP steel, b) for TRIP steel

and Fig. 18 .
Different friction coefficient and blankholder force influence of springback amount significantly.

Fig. 17 Fig. 18
Fig.17 Section profiles of U-shape parts for DP600 Fig.18Section profiles of U-shape parts for TRIP

Table 1
Mechanical properties of the TRIP steel

Table 2
Mechanical properties of the DP steel

Table 3
Combination of the technological parameters

Table 4
Parameters of numerical simulation Experiment was realized on the hydraulic press ZD 40.Data were digitalized and measured using MATLAB software.All obtained results will be described in the following chapter.