1. Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects
PhD thesis
Samir Lemeš
Supervisor: prof. dr. Karl Kuzman
Ljubljana, 2010

2. Acknowledgments Slovenian Science and Education Foundation Ad-futura
Slovenian Tool and Die Center Tecos Celje
Forming Laboratory at the Faculty of mechanical engineering in Ljubljana
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

3. Presentation Outline Introduction Motivation
Determination of product characteristics
3D scanning
Deformation analysis
3D FEA
Measurement uncertainty
Statistical analysis
Algorithm for automated measurement process
Conclusions
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

4. Introduction
Validation is defined as a process for assessing simulation uncertainty US by using benchmark experimental data and, when conditions permit, estimating the sign and magnitude of the simulation error δS.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

5. Introduction
The validation and verification process is standardized by ASME.
The determination of computational uncertainty uses metrology approach.
It is necessary to identify and quantify the error sources, to model their relationships, perform simulation calculations, and finally to quantify the output uncertainty.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

6. Introduction
Finite Element Method (FEM): proven engineering tool used in design phase of product lifecycle
3D scanning: a novel method for dimensional quality control and measurement
Idea: to use FEM in quality control, in order to simulate clamping process
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

7. Product lifecycle Introduction Idea Design Tooling FEM FEM Recycling
Manufacturing
Redesign
Maintenance
Quality control
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

8. Introduction Sheet metal products often exhibit springback behaviour
The springback is usualy reduced in tooling phase: tool is designed to compensate springback
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

9. Introduction
Components containing (springback) defects are not necessarily unusable
When they are used in assemblies, they are additionally deformed
Therefore, the quality control requires clamping / fixtures
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

10. Introduction
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

11. Introduction
It is justifiable to simulate clamping process by means of numerical methods (FEM).
The main objective: To determine whether it is possible to measure the geometry of thin-walled products, deformed due to residual stresses, using numerical simulations of clamping process
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

12. Introduction
H0: Numerical simulation of clamping of sheet-metal products with elastic springback has the same measurement uncertainty as 3D scanning of physically clamped products, when clamping is simulated with appropriate boundary conditions which describe accurately the behaviour of the physical clamping.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

13. Introduction
Previous researches include: springback, tolerance analysis and synthesis, process optimization for sheet-metal production, reverse engineering, optical 3D measuring for quality control of sheet metal products, measurement uncertainty, and finally use of FEM in quality control of sheet metal products.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

14. Motivation
In order to avoid fixtures during dimensional measurement of elastic products, it is planned to digitise products as they are (without fixture), and then to simulate clamping.
To estimate if this method can be used practically, it is important to test the procedure on a real product.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

15. Motivation This research consists of the following phases:
Determination of product characteristics
3D scanning of products as they are (without clamping)
3D scanning of products clamped by means of rigid fixture
Processing of scanning results
Checking reverse engineering accuracy
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

16. Motivation FEM simulation of clamping process
Static strain analysis based on measured deformations
Identificate dominating influence factors
Measurement uncertainty analysis
Statistical analysis of results and hypothesis testing
Algorithm for automation-ready procedure, with well defined influence factors and measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

17. Determination of product characteristics
Oil filter housing, manufactured by Mann+Hummel Tešanj, B&H
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

18. Determination of product characteristics
The circular cross-section diameter (Ø92-0,2) measured 10 mm from top of the housing will be used for control by 3D scanner and FEA.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

19. Determination of product characteristics
Since anisotropy is proven to be among the most influencing factors in deep drawing, the samples from 5 rolls were tested for anisotropy.
5 sets of 15 specimens, taken from different rolls, at 0, 45, 90°
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

20. Determination of product characteristics
To simulate real conditions (when housing is assembled with other filter components), a rigid clamping assembly was constructed.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

21. 3D scanning
5 sets of 5 filter housings were prepared for 3D scanning: each set manufactured from different sheet-metal roll.
Each sample was scanned with and without clamping.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

22. 3D scanning
Simplified 2D analysis: manually manufactured saddle brackets, with excessive springback
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

23. 3D scanning
Digitized data is usable only for visualisation purposes; 3D scanned entities do not have any geometry usable for finite element meshing
The data conversion errors were estimated
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

24. Deformation analysis a) free, b) clamped, c) ideal contour
Deviations between the free and the clamped contours, are used to define boundary conditions for simulations
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

25. Deformation analysis Equivalent hoop stresses in chosen cross-section
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

26. Deformation analysis
To visualise deviations from nominal cylindrical shape, the cross-section is presented in Cartesian coordinate system:
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

27. Deformation analysis
Cross-section radii of unclamped (free) and clamped part
Rotational profile fitting
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

28. Deformation analysis
Interpolation of vectors defining clamped and unclamped profiles
Pairs of correlated points can be used to define FEM displacement restraints
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

29. Deformation analysis
Fortran program developed for RMS based rotational profile fitting
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

30. Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

31. 3D finite element analysis
Objective: to validate if numerical simulation can be used to compensate springback-caused deformation of thin-walled products when these products are digitised with 3D scanner.
For validation, the same product was scanned in both deformed (clamped) and undeformed (free) state.
Clamping was then simulated by FEM.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

32. 3D finite element analysis
Impossible to define boundary conditions that correspond 100% to the actual model
The contact between the model and the fixed surface is realised across circular edges.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

33. 3D finite element analysis
The influence of boundary conditions on FEM results
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

34. 3D finite element analysis
The influence of boundary conditions on FEM results
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

35. 3D finite element analysis
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

36. 3D finite element analysis
Conversion of FEM results to CAD model
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

37. 3D finite element analysis
Dimensional deviations derived from 2D FEM simulation results
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

38. 3D finite element analysis
FEM limitation: node-based loads and displacement restraints
Displacement boundary conditions based on extracted CAD feature
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

39. 3D finite element analysis
Surface-based boundary conditions with equivalent radius
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

40. 3D finite element analysis
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

41. Measurement uncertainty
Parameter, associated with the result of a measurement, that characterizes the dispersion of the values that could reasonably be attributed to the measurand [ISO GUM, 1995]
A measurand Y is not measured directly, but is determined through a functional relationship Y = f (X1, X2, ..., XN)
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

42. Measurement uncertainty
The combined standard uncertainty for N uncorrelated uncertainty components is calculated as:
ucomb: combined standard uncertainty
ui: standard uncertainties of components
ci: sensitivity coefficients of components
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

43. Measurement uncertainty
Type A: uncertainty estimates obtained from a number of observations estimated by statistical analysis expressed by standard deviations.
Type B: non-statistical uncertainty estimates from: past measurements, calibration certificates, manufacturer’s specifications, calculations, published information, common sense.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

44. Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

45. Measurement uncertainty
Influence factors
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

46. Measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

47. Measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

48. Measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

49. Measurement uncertainty
Graphical comparison of uncertainty contributions
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

50. Measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

51. Measurement uncertainty
The largest contribution to the measurement uncertainty in both cases comes from the scanning errors (u3).
To eliminate the influence of manufacturing errors, the samples were measured on a 3D CMM.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

52. Measurement uncertainty
Inversed equation for combined standard uncertainty:
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

53. Measurement uncertainty
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

54. Statistical analysis
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

55. Statistical analysis
H0: "There is no difference between measurement uncertainty of simulated and physical clamping", is tested by a two-sample t-test for equal means.
The calculated value of t-test variable (-1) is smaller than the critical value (2,131847); the null hypothesis can be accepted.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

56. Statistical analysis
ANOVA single factor analysis was used to evaluate the variation of material properties between different rolls of sheet metal used.
H0: "The yield stress Rp02 has the same value for all 5 sheet metal rolls“.
F variable (2,066392) is smaller than the critical value (2,502656); the null hypothesis can be accepted.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

57. Statistical analysis
H0: "The deformation strengthening exponent (n) has the same value for all 5 sheet metal rolls".
F variable (10,74722) is larger than the critical value (2,502656); the null hypothesis can be rejected.
The deformation strengthening exponent (n) differs significantly between the 5 rolls of sheet metal.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

58. Statistical analysis
H0: "The anisotropy r10 has the same value for all 5 sheet metal rolls“.
F variable (0,256751) is smaller than the critical value (2,502656); the null hypothesis can be accepted.
The anisotropy r10 does not differ significantly between the 5 rolls of sheet metal used in the experiment.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

59. Algorithm for automated measurement process
The starting algorithm is based on presumption that scanned part has some extractable features (holes, grooves, edges, etc.) which can be used to define boundary conditions.
The analysis performed in this dissertation showed that it is not always possible to define such boundary conditions.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

60. Algorithm for automated measurement process
Statistical analysis proved that virtual fixation can have the same measurement uncertainty as the real clamping, but only with careful and proper boundary condition settings.
The difference between the ideal and the scanned contour in undeformed position presents the major contribution to measurement uncertainty.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

61. Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

62. Algorithm for automated measurement process
The most important steps:
The real material properties
Model orientation
Part imperfections
Carefully chosen coordinate systems
Defined material side
Defining constitutive models and solution parameters
Converting FEM results into CAD data
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

63. Conclusions
This research gives a modest contribution to automated inspection and quality control of thinwalled, flexible parts, trying to combine the two emerging technologies into a new, hybrid technology.
The combination of 3D scanning and computer simulations requires careful choice of many adjustable parameters.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

64. Conclusions
The numerical simulation of clamping of sheet-metal products with elastic springback has practically the same measurement uncertainty as 3D scanning of physically clamped products, when clamping is simulated with appropriate boundary conditions which describe accurately the behaviour of the physical clamping.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

65. Conclusions
A new computer software was developed for alignment of circular contours obtained by 3D scanning.
The iterative algorithm is based on angular division of contours, and calculation of RMS.
It is proved that Hausdorff distance is not related to stress/strain state in deformed sheet metal objects
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

66. Conclusions
The influence factors onto 3D scanning and numerical simulation processes are identified and analysed.
It is shown that major contribution to measurement uncertainty comes from scanning method.
The deviations of parts due to manufacturing technology are the second largest influence factor.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

67. Conclusions
Material properties are confirmed as less significant factor in numerical simulations, in terms of contribution to numerical errors, when variation in material properties is compared with variations of other influence factors.
The boundary conditions are identified as major source of shape deviation in numerical simulations.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

68. Conclusions
A novel algorithm for automated inspection of springback-deformed components was developed.
The algorithm defines the key steps which have to be performed in order to get the correct simulation results.
Only when all these steps are followed carefully, the simulated results describe the physical behaviour correctly.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

69. Conclusions
I claimed in the main hypothesis that numerical simulation of clamping of sheet-metal products with elastic springback has the same measurement uncertainty as 3D scanning of physically clamped products, when clamping is simulated with appropriate boundary conditions which describe accurately the behaviour of the physical clamping.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects

70. Conclusions
I also anticipated that it is possible to measure the geometry of thin-walled products, which are deformed as a result of residual stresses, using numerical simulations of clamping process.
The results of research, statistical analysis and the comparison of experimental and simulated data confirmed the hypothesis.
Validation of Numerical Simulations by Digital Scanning of 3D Sheet Metal Objects