Here, the variables m and n indicate the number of half sine waves
in the buckling mode. Considering the equilibrium boundary condition and also the small radial tensile stress in y-direction, the initial value of the buckling stress st,cr can be calculated by solving the differential equation as follows:
In relation to wrinkling, m indicates the number of wrinkles and n is the number of half sine waves for each wrinkle and can be set to 1. Hence, to predict the condition of the drawn part only the number of possible wrinkles should have to be calculated. Senior [9] used the energy method to predict the number of wrinkles for deep drawing without blankholder force. In the areas of the flange, which are not supported by the blankholder, there are three energy components at the onset of buckling: the energy due
to bending into a half-wave segment of the wrinkled flange, the energy due to the tangential compressive forces and also the energy due to the lateral loading of the flange surface. Considering these energy components, it is possible to determine the condition of minimum tangential compressive stress at the onset of buckling. This reveals that the number of waves m into which the flange would wrinkle is given by:
Here, rm is the mean flange radius, r0 and ri are the initial radius of the flange and inner radius of drawing die, respectively. The critical stress calculated from Eq. (11) can be used as comparative criterion to determine the maximum allowable tangential stress for the free end part of the sheet metal.
3.3. Stability analysis – bottom crack
Since the stress state in the cylindrical wall of a drawn part is very similar to the uniaxial stress state, it is possible to use it as a simple failure criterion by comparing it with the tensile strength of the material. Considering Rm as the ultimate tensile strength of the sheet, the maximum transferable punch force Fbc can be estimated as follows:
Here, a0 and CR are the cross sectional area of the drawn part and crack factor for the correction, respectively. To identify the crack factor the model from Doege is used [10]. He showed that the parameter range can vary from 0.9 to 1.5 as a function of the
material properties (anisotropy value and strain hardening rate) and the punch edge radius.
4. Results and discussion
In order to verify the complete analytical model, numerical simulations as well as experimental analysis are done. The numerical results are based on a 3-D FEM-simulation (simufact.- forming v13.1) using solid-shell elements with five integration layers over the thickness and a reduced integration scheme. The maximum element size used here is 1 mm. Due to its wide distribution in deep drawing applications the mild steel DC04 (1.0338) is chosen for the investigation. Of course, other materials like aluminium wrought alloys like AA5754 can also be utilised [11]. For the numerical and experimental analysis a macro structured drawing die of tool steel 1.2379 with an inner radius of RI = 50 mm, a die edge radius of RD = 10 mm and a wavelength of l = 8 mm is considered for the rotational symmetric geometries. Here, the immersion depth can be adjusted using
distance rings from d = 0.0 to 0.4 mm. Furthermore, rectangular macro structured deep drawing tools (85 mm 85 mm) with a corner radius of RC = 20 mm are considered. The die edge radius and wavelength of these tools are also RM = 10 mm and l = 8 mm, respectively. Due to other stress conditions, the immersion depth can vary from d = 0 to 1.0 mm.
Fig. 4 shows typical examples of a stable and an unstable process for rotational symmetric and rectangular parts in order to demonstrate the feasibility of the process.
Obviously, on the wall of each drawn part are small optical surface defects parallel to the die edge. These defects are caused by the contact condition on the tool surface and are typical for a deep drawing process.