·Table of Contents
·Materials Characterization and testing
Micromagnetic Stress Determination on Tailored Blanks
Hans Kurt Tönshoff, Thomas Friemuth, Ines Oberbeck and Harald Seegers
Institut für Fertigungstechnik und Spanende Werkzeugmaschinen
Schlosswenderstr. 5, 30159 Hannover
Welded metal sheets of different ferrous materials, so called tailored blanks, gain increasingly more importance in automotive industry. Microstructure, micro hardness and residual stress are important material properties for workpiece behaviour. The forming quality of tailored blanks is influenced by residual stress induced nearby the welding zone. Residual stress also plays an important role for the behaviour of welded components in practice. Stress occurs due to localised elastic expansion or contraction of the material lattice. Especially in welding processes this causes plastic deformations of the joint sheets. In later forming operations the forming quality is considerable influenced by these residual stress. Also premature failure is attributed to inadequate residual stress. Annealing processes have to be used right after welding to modify the residual stress state.
This paper deals with micromagnetic testing used as a non-destructive monitoring technique for determining residual stress nearby welding zones in ferromagnetic sheets. The difficulties which have to be bridged are the change in microstructure nearby the weldment joint during welding. This changes influence the micromagnetic parameters of the specimen like residual stress does. Hence, different parameters have to be taken into account to determine the residual stress and to eliminate the influence of microstructure changes. In this paper additional quantities of the micromagnetic signal were investigated concerning there dependence on different material properties. Based on this experiences a mathematical equation was developed for the determination of residual stress nearby a weldment joint by micromagnetic testing which is able to suppress the influence of microstructure. The results are in accordance with X-ray investigations and residual stress evaluated by micromagnetic testing.
Microstructure, micro hardness, and residual stresses are important material properties for workpiece behaviour of sheets in forming processes. Quality assurance of sheets require a fast detection of the material properties. Especially for tailored blanks, sheets welded out of different materials, a non-destructive testing method is necessary to guarantee a high quality standard.
For the scientific testing laboratory X-ray diffraction has proven its high standard to evaluate important material properties. On the one hand this method is known for a high accuracy, but on the other hand much time is used for the investigation. Furthermore, the workpiece often has to be destroyed because only cut-out segments are measurable. For 100% certification during manufacturing only a short time is available for determining material properties. To meet these requirements, a faster non-destructive detection technique is required. One possible testing method is the micromagnetic or Barkhausen noise analysis technique . Due to the sensitivity of the signal on different material properties it has to be correlated to laboratory methods like X-ray investigations and metallographic inspection .
Residual stress as one of the essential material properties plays an important role for the behaviour of welded components in practice. Localised elastic expansion or contraction of the material lattice, as it take place e.g. in welding, causes residual stresses. The changes usually occur as plastic deformations of the sheets. In later forming operations the forming quality is considerably influenced by these residual stresses. To change the residual stress state annealing processes can be used after welding.
Barkhausen noise analysis or micro magnetic testing is based on the behaviour of ferromagnetic materials in magnetic fields. In ferromagnetic materials domains with different local magnetisation directions are separated by Bloch walls. An increasing external magnetic field causes Bloch wall motions and rotations. As a result, the total magnetisation of the sheet is changing. The magnetisation process is characterised by a hysteresis. Due to the Bloch wall movements single turnarounds of Weiss' domains are required. Analysing the magnetic flux by a coil, this turnarounds are registered as electrical pulses. These pulses are the so called Barkhausen noise.
It was shown that the magnetic domain structure of ferromagnetic materials and also the Barkhausen noise, is influenced by residual stresses [3, 4]. Compressive residual stress reduces the intensity of the Barkhausen noise whereas tensile stress increases the max. Barkhausen noise level (Fig. 1) [5, 6]. In addition, the hardness and structure state of the steel sheet influences the Barkhausen noise. To separate the different material characteristics, different quantities have to be analysed.
Fig 1: Influence of Residual stress on the Barkhausen noise signal|
The maximum amplitude of the Barkhausen noise Pmax is commonly used for residual stress determination. Other values like the peak position and the value of the maximum of the amplitude spectrum can be used for enhanced determination of material properties [7,8]. Especially for excluding changes in micro hardness or microstructure these values have to be analysed and to be taken into consideration for residual stress determination. It has to be pointed out that all parameters of this testing technique are influenced by the measurement setup. Important parameters are the excitation frequency fe, the excitation voltage ve and the used range of the analysing frequency fa.
X-ray residual stress investigations were carried out on a Seifert XRD PS two circle diffractometer. Cr Ka radiation was used with a 2 mm point collimator. The well known sin2y
method was applied with 7 y
angles between -45° and +45°. Peak positions were analysed by the centre of gravity method.
For Barkhausen noise analyses a commercial detector system of the American Stress Technologies Inc. (Detector: S1-163) was used. In this detector system the magnetisation unit and the receiver are integrated in one sensor. The Barkhausen noise was evaluated on a separate computer using different statistical programs.
By appliying different analysing frequencies, the penetration depth can be varied in a range of several mm up to the mm range depending on the magnetical properties of the used ferromagnetic material. For the magnetic properties of DC 05 (Standard No. 1.0312) and ZStE 340 (Standard No. 1.0548) steel sheets and the here used measurement setup the penetration depth is in the range of z = 2-28 mm.
As the residual stress sensitive Barkhausen noise values depend on various material properties, material calibration curves have to be determined for residual stress evaluation. Uniaxial tensile tests are commonly used to calibrate Barkhausen noise to residual stress measurements .
Calibration curves are obtained by measuring the level of Barkhausen noise during compression and tension in one or two directions. In this case only uniaxial tensile forces were applied. The calibration samples were 10 mm wide, 100 mm long, 1 mm thick and consisted for the highlighted diagrams of H 340 steel with a creep limit of 389 MPa. Equal tests were performed for the other used materials. The stepwise changing of the loads was controlled with strain gauges. Additionally X-ray stress measurements were carried out to correlate stress measured by X-rays sX-ray and stress evaluated by Barkhausen noise analysis sBN. Therefore, the tension test device was installed in the X-ray diffractometer and the calibration curve for both methods was surveyed simultaneously in the same run.
It has to be taken into account that the calibration curve is determined for a superposition of internal stress and external stress. This stress is also called total load. Based on this diagram a correlation between the laboratory method X-ray investigation and the much faster Barkhausen noise method is possible. But it has to be pointed out that for every different material and also in special cases for every charge, a calibration between the stress level and the Barkhausen noise signal is still necessary. This disadvantage is caused by changes in micro structure or alloying elements.
Different materials and different specimen dimensions were used to describe the influence of workpiece properties on the measured Barkhausen noise signal. In Fig. 2 the differences between the material DC 05 and H 340 are highlighted. Additionally sheets of DC 05 of different thickness were evaluated to describe the influence of sheet thickness on the signal level. The differences in the effective noise level between DC 05 and H 340 are about 0,15 V as shown in Fig. 2.
The distribution of the effective noise level versus the total load is nearly linear for all sheet thicknesses of DC 05. Due to this result, it can be stated that there is almost no influence of sheet thickness on the effective noise level of the Barkhausen noise signal. Furthermore, it can be stated that the influence of the charge properties is considerable lower than the influence of the used material.
Fig 2: Influence of different material properties on the effective noise level|
Also different analysing parameters were tested for optimising the measuring strategy. Best sensitivity was reached for an excitation frequency of fe = 300 Hz. The change in the effective noise level is then about 0,25V for a change of 100 MPa. The difference in the incline for different excitation frequencies of the effective noise levels are caused by the change of penetration depth resulting from the attenuation of the electromagnetic field in the ferromagnetic material (skin effect). A frequency of 300 Hz was used for the further investigations.
Welding of thin ferromagnetic sheets
An important field during manufacturing of tailored blanks is the monitoring of the welding process. The Barkhausen noise analysis presented in this paper was applied to laser beam welded sheets. Inert gas was used to avoid oxidation of the material in the welding zone. Laser beam welding was characterised by a small localised weld seam. For first investigations a CO2 laser Triagon 6000 manufactured by Wegmann Boasel was applied for welding. The used output power was 2 kW using a feed of f = 3,5 m/min.
A typical diagrammatic section of a welding zone is shown in Fig. 3. The resulting micro structure depends on the maximum temperature reached during the welding process.
Fig 3: Typical changes in micro structure during welding|
Different quantities of the induced Barkhausen noise signal were analysed on their sensitivity on different material properties. Best correlation of one single parameter to the residual stress state is found for the maximum Barkhausen noise level. Other values like FWHM and peak shift are more significant for the material properties of the sheet. In the middle of the welding seam these values show a significant change. The reason is the sensitivity of the Barkhausen noise analysis on every material change even in case of recrystallization, stress relief, annealing or just heat induced dislocation movements. Because of the sensitivity of FWHM and peak shift, it is possible to determine the influence of the microstructure on the maximum noise level and to eliminate this influence for calculating residual stress.
By using such a compensation a well correlated residual stress state is found by Barkhausen noise analysis (Fig. 4). In this figure s
BN represents the stress values calculated out of the Barkhausen noise data. s
X-ray represents the measured stress obtained by X-ray residual stress determination. A good accordance between the stress determined by X-ray analysis s
X-ray and determined by Barkhausen noise analysis s
BN is achieved.
In particular the residual stress s
BN is evaluated by Barkhausen noise analysis can be determined by the following equation:
Fig 4: Comparison of residual stress evaluated by X-ray and Barkhausen noise analysis|
Whereas sBN is the determined residual stress value, a, b, c, d are the coefficients of calibration measurements, Pmax is the maximum of the Barkhausen noise amplitude, Dt is the peak shift and Imax the maximum of the number of impulses versus the amplitude. It has to be noticed that a constant offset represented by the coefficient d is needed for this equation.
In this paper the Barkhausen noise technique for the evaluation of residual stress of ferromagnetic sheets is presented. Characteristic values obtained by the Barkhausen noise signal allow residual stress evaluation in very short time. Laboratory reference tests using X-ray diffraction are necessary to correlate the values of the Barkhausen noise to residual stress. For unwelded and laser beam welded sheets, tension tests or sample welds can be used for calibration. Therefore, the application of correlation curves allows to determine the residual stress state of welded steel sheets by Barkhausen noise data.
The presented research work is supported by the German Research council (DFG) in a Special Research Department (SFB) 362 named "Fertigen in Feinblech" (manufacturing in sheets).
- Heyn, H., Decker I., Wohlfahrt H.: Tools and requirements for in-process quality control in laser beam welding. Proceedings of SPIE: Laser Material Processing: Industrial and Micorelectronics Applications, Vol. 2207, 1994
- Oberbeck-Spintig, I.: Einsatz der Barkhausenrauschanalyse zur zerstörungsfreien Eigenschaftsprüfung geschweißter Feinbleche, Diss., Universität Hannover, 2000
- Tiito, S.: Magnetoelastic Testing of Uniaxial and Biaxial Stresses. Int. Conf. Residual Stresses, Soc. Francaise de Metallurgie, Nancy, 1988.
- Theiner, W.A.: Micromagnetic techniques. in Structural and Residual Stress Analysis by Nondestructive Methods, ed. V. Hauk, Elsevier Science B.V., 1997
- Karpuschewski, B.: Mikromagnetische Randzonenanalyse geschliffener einsatzgehärteter Bauteile. Dr.-Ing. Dissertation, University of Hannover, 1995.
- Tönshoff, H. K.; Karpuschewski, B.; Oberbeck, I,: Residual stress determination of ferromagnetic sheets. Proc. SheMet'96, Enschede 1996, pp. 347-355.
- Tönshoff, H. K.; Karpuschewski, B.: Quality Control of Surface Integrity with Micromagnetic Techniques. Prod. Eng., Annals of the German Academic Society for Production Engineering, Vol. II/2 1995, pp. 203-206
- Tönshoff H. K., Karpuschewski B., Oberbeck-Spintig I., Seegers H.: "Non Destructive Stress Determination on Tailored Blanks", Production Engineering Vol. VI/2 (1999), pp. 143-146