FLD ELABORATION BY MEANS OF CONTACT-LESS OPTICAL SYSTEM FOR MEASURING DEFORMATION AND COMPARISON OF RESULTS WITH SIMULATION PROGRAMME PAM-STAMP 2G TVORBA DMP POMOCÍ BEZKONTAKTNÍHO OPTICKÉHO SYSTÉMU PRO MĚŘENÍ DEFORMACE A POROVNÁNÍ VÝSLEDKŮ SE SIMULAČNÍM PROGRAMEM PAM-STAMP 2G Jiří SOBOTKA a, Pavel SOLFRONK a, Pavel DOUBEK a, Michaela KOLNEROVÁ a a TECHNICAL UNIVERSITY OF LIBEREC, Studentská 2, 461 17 Liberec 1, Česká republika, jiri.sobotka@tul.cz Abstract Deformation measurement (but also e.g. digitalization of industrial parts) by means of contact-less optical systems during last years occupies very important position not only by monitoring material properties. In the case of measurement FLD was for example necessary to describe method how to find and evaluate limit forming state with the help of standard ISO 12004-2. Measuring FLD by means of optical methods is thus widely using not only in automotive industry, but also in truly many branches of whole engineering industry. Very important factor is e.g. also possibility to compare values of true strain measured by optical system and results from numeric simulation programme in this case was used widely using simulation programme for sheet forming from ESI company (France) PAM-STAMP 2G. This article has aim to present possibility to compare real values of true strain (logarithm) and results from simulation, their mapping and final graphical display of differences between these values. Key words: Photogrammetry, FLD (Forming Limit Diagram), Sheet forming, PAM-STAMP 2G 1. INTRODUCTION Numerical simulations are very important part of whole production process and not only in the branch of metal forming. But still their results are calculated by means of simulation software and thus sometimes they are quite far away from reality. So it s important to carry out validation of these results by means of systems which offer measurement of real strain values. These days validation of numerical simulations is quite often carry out by means of contact-less optical systems (in this paper system ARAMIS). In this paper comparison itself was made in the frame of FLD (Final Limit Diagram) measurement and as a testing material was used common zinc-coated deep-drawing material marked like DC 05B (electro-galvanized) with thickness 0,8 mm. In fig. 1 is shown basic stress-strain curve of this material for rolling directions 0 and basic mechanical properties are given in table 1. Fig. 1. Static tensile test of measured material
2. NUMERICAL SIMULATION OF SPECIMENS FOR FLD BY MEANS OF PAM-STAMP 2G Programme PAM-STAMP 2G (ESI company) is widely using software and serves for numerical simulation in sheet forming. In this paper was this software used as a part of FLD elaboration by means of stretch forming test using hemispherical punch. Such method is very close to Nakazima test [1], only shape of specimens is a little bit different. All results and images in this paper illustrate specimens for FLD evaluation with the lowest width 30 mm. That one used in PAM-STAMP 2G after deformation with major strain φ 1 distribution is shown in fig. 2 (right). In fig. 2 (left) is displayed whole lay-out for FLD measurement with hemispherical punch, die, blank holder and blank. Adjustment of stroke (progression) arose from depth of specimen (change of z-coordinate) measured by ARAMIS system - see label in fig. 4 (left). Fig. 2. Lay-out for FLD measurement (left) and major strain φ 1 distribution (right) from PAM-STAMP 2G As a testing material was chosen commonly using (especially in automotive industry) material marked like DC 05B which is deep-drawing material (producer Voest Apline). Basic mechanical properties from static tensile test of this material can be found in table 1 (stress-strain curve see fig. 1). Table 1 Mechanical properties of tested material DC 05B (deep-drawing material - Voest Apline) Mechanical properties Yield strength [MPa] Ultimate strength [MPa] Ductility [%] Homogenous ductility [%] Normal anisotropy [-] Rolling direction [ ] R p,02 R m A 80mm A g r α 0 155,7 282,4 44,5 26,2 1,974 45 162,3 194,2 44,2 25,7 1,789 90 158,9 281,6 45,1 25,8 2,358
18. - 20. 5. 2011, Brno, Czech Republic, EU 3. MEASUREMENT OF SPECIMENS FOR FLD BY MEANS OF ARAMIS SYSTEM Using contact-less optical systems for deformation measurement made a great progress during last years. They represent very good solution e.g. for measurement deformation behavior both of widely used materials and newly developed materials (e.g. TRIP, TWIP, ). There is couple of cameras (fig. 3 left) to be able to measure 3D, strong lighting device, T-box for controlling signal for cameras and PC for data scanning and results evaluation. In fig. 3 (right) is shown calibration desk which contains so-called coded points and own deformation measurement is thus subsequent based on proper calibration. In principle this system measures displacements (x,y,z) of so-called facets [pixels] from observed area and from such displacement calculates mainly strain distribution (if interest also e.g. strain rate, velocity, acceleration, ). For more detail description see internet page about GOM company (producer) or e.g. [2]. Fig. 3. Couple of cameras (left) and calibration desk (right) for measurement by system ARAMIS In the following figure (fig. 4 left) is shown major strain φ1 distribution for specimen with width 30 mm from ARAMIS measurement. Such stage represents image right before the crack opening. Subsequent FLD evaluation for all specimens with different widths is supposed to be carrying out according international standard ISO 12004-2 - more details see [2]. There is also shown section 0 (see label) and values are plotted in the fig. 5 together with that ones measured by means of programme PAM-STAMP 2G. Final FLD for material DC 05B can be found in fig. 4 (right) which illustrates FLC of DC 05B for all states of strain mφ. Fig. 4. Major strain φ1 distribution (left) and FLD for deep-drawing material DC 05B (right)
Fig. 5. Major strain φ 1 distribution along specimen from PAM-STAMP 2G (/) and ARAMIS system (/) Fig. 5 illustrates comparison of major strain φ 1 distribution along specimen measured by both used systems. It can be clearly seen that in the programme PAM-STAMP 2G was used quite rough mesh and due to that there are huge steps in the major strain φ 1 distribution. Such comparison is made only along used section (along axis of specimens) and not on the whole surface (like mapping is) but these results exhibited good matching namely if it is taken into account that simulation was carried out without any precious adjustment. Fig. 6 shows situation after importing into programme S-VIEW (grey part is from PAM STAMP 2G; colored from ARAMIS) which is part of ARAMIS system and serves for basic overview of measured values and also for mapping. It can be seen that both parts are truly only imported cause there is no positioning. Thus next step before mapping itself was to positioned both parts and ensure them same orientation and position Fig. 6. Preparing for mapping situation just after importing of both parts
Not only due to such positioning was necessary to carry out registration of these two results before mapping itself. It was made by so-called manual pre-registration and best-fit registration. Fig. 7 (left) shows situation after manual pre-registration. Such type of registration rests in choosing different points on first part and to these points are allocated corresponding points in the other part. This method doesn t exhibit high accuracy and thus is important after manual pre-registration to carry out best-fit registration (fig. 7 right). Fig. 7. Situation after manual pre-registration (left) and best-fit registration (right) Distance differences after registration are shown in fig. 8. It displays differences from the position point of view and gives a rapid overview about accuracy of whole mapping. Fig. 8. Comparison of distance differences between PAM-STAMP 2G and ARAMIS system
4. CONCLUSION This paper describes possibility how to compare results from numerical simulations of metal forming (in this case PAM-STAMP 2G) and modern contact-less optical system ARAMIS system from GOM company. Comparison was carried out by means of mapping with the help of programme S-VIEW (delivered as a part of ARAMIS system). Fig. 9 shows final comparison of major strain φ 1 distribution from PAM-STAMP 2G (grey surface) and ARAMIS system. It is possible to see that differences between FEA and ARAMIS aren t significantly high and rather around 0. Whole measured differences lie within interval: -0,25; 0,10. From the comparison point of view is really necessary firstly to take into account that in this case was not in the programme PAM-STAMP 2G calculated with spring-back which can greatly influences final comparison. This is also one of the important recommendations for further research to find out influence of spring-back on the position of FLC in FLD for different types of materials. Fig. 9. Final comparison of strain distribution differences between PAM-STAMP 2G and ARAMIS system This paper was written in support of grant project GAČR 101/09/1996 and research project MSM 4674788501. LITERATURE [1] HOSFORD W.F., CADDELL R.M.: Metal Forming Mechanics and Metallurgy, Cambridge University Press, New York, 2007, ISBN 978-0-521-88121-0 [2] SOBOTKA, J.: Formability of TWIP Materials and its Evaluation by Optical System, Tribun EU s.r.o., Liberec, 2010, ISBN 978-80-7372-681-2