In the aerospace industry a move towards probabilistic design - - PowerPoint PPT Presentation

▶
in the aerospace industry
SMART_READER_LITE
LIVE PREVIEW

In the aerospace industry a move towards probabilistic design - - PowerPoint PPT Presentation

Mar 28, 2024 307 likes •1.22k views

In the aerospace industry a move towards probabilistic design practices has been recognized as a potential game-changer , as it is understood to reduce costs , risks and lead times , while increasing the confidence in analysis tools, and

slide-1
SLIDE 1
slide-2
SLIDE 2

”…a move towards probabilistic design practices has been recognized as a potential game-changer, as it is understood to reduce costs, risks and lead times, while increasing the confidence in analysis tools, and the quality and reliability of products“ Nasa report by Zang et al. (2002)

In the aerospace industry…

2

slide-3
SLIDE 3

A different simulation approach…

Deterministic design Nominal operation Worst-case scenarios Safety factors Easy to implement in simulation Probabilistic design Adds variation modeling Uses probability distributions, etc. Allows some probability of failure Difficult to implement in simulation

3

slide-4
SLIDE 4

Probabilistic design → Robust design → Geometry assurance

4

slide-5
SLIDE 5

Probabilistic design → Robust design → Geometry assurance

“Robust design is an engineering methodology that aims at minimizing the effects of variation without eliminating the variation itself. “ Phadke (1995)

5

slide-6
SLIDE 6

Probabilistic design → Robust design → Geometry assurance

“Geometry assurance is a set of activities aimed at reducing the effects of geometric variation and increasing the precision of functional attributes of products.” Söderberg et al. (2006)

6

slide-7
SLIDE 7

Part variation

slide-8
SLIDE 8

Part variation

→

Assembly variation

slide-9
SLIDE 9

Part variation

→

Assembly variation

slide-10
SLIDE 10

Part variation

→

Assembly variation → Functional variation

slide-11
SLIDE 11

11

slide-12
SLIDE 12

12

slide-13
SLIDE 13

13

slide-14
SLIDE 14

Aerodynamic loads Thermal loads Structural loads

slide-15
SLIDE 15

Fabricated assembly Cast, forged and sheet metal parts Welded together

15

slide-16
SLIDE 16

16

slide-17
SLIDE 17

17

slide-18
SLIDE 18

18

slide-19
SLIDE 19

19

slide-20
SLIDE 20

Conceptual Modeling Conceptual Modeling

20

(Adapted from: Oberkampf, 2002)

slide-21
SLIDE 21

Mathematical Modeling Conceptual Modeling

21

slide-22
SLIDE 22

Computer Programming Mathematical Modeling Conceptual Modeling

22

slide-23
SLIDE 23

Discretization & Algorithm Selection Computer Programming Mathematical Modeling Conceptual Modeling

23

slide-24
SLIDE 24

Discretization & Algorithm Selection Computer Programming Mathematical Modeling Conceptual Modeling Numerical Solution

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”vg/jBöÚöFõSö˚[b[∫OÃ>—÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹`€¢∂{ÁcŒfljoÔ∫t·“E ˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπÓzΩµ{f˜Èû7Œ›ÙΩ yÒˇ÷’û9=›ΩÛzo˜≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC›á ’?[˛‹ÿÔ‹j¿w†Û—‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯’Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚Ì¡w‹wÔ£flO‰| (ˇh˘±ıS–ß˚ìììˇòÛ¸c3-€ cHRM ⁄Ï›wú]eµˇÒœzûΩOõÙfiÅP“CTÏEØ•#≈vm\A§£"%4Ei" 5Ñ™¢`¡ÇTTÈ)Ñ@í…Ã){ÔÁYø?ˆôI@Ó˝›{•¨7ØyMrfrÍÊü Ôk≠Ô#™ä1∆cå1∆cå1/TŒficå1∆cå1∆ÛBfñ1∆cå1∆cåyA≥ Àcå1∆cå1∆º†YÄeå1∆cå1∆c^– ,¿2∆cå1∆cå1/h`cå1∆cå1∆ò4 ∞å1∆cå1∆cà öX∆cå1∆cå1ÊÕ,cå1∆cå1∆ÛÇñÿ[`å1∆cå1∆Û‚≤‡ùG∫ëC ít⁄îâ£&N6fÚËëS«é™OŸ|“∏´µ!√:˝[{fiÚµÉ⁄ˇìèôæ‚M" ¥~}í¸_ø^QU˚‘ç1∆cå1∆c^Äfº·ì≤È∆£”ßLª≈¶„∑ò1mÙ

24

slide-25
SLIDE 25

Solution Representation Numerical Solution Discretization & Algorithm Selection Computer Programming Mathematical Modeling Conceptual Modeling

25

slide-26
SLIDE 26

Research Question II: How can geometry assurance methods be implemented in multidisciplinary simulations in industrial settings? Research Question III: What role should geometry assurance play in the early phases of aerospace component design? Research Question I: What barriers to implementing geometry assurance practices can be identified in the aero engine industry?

26

slide-27
SLIDE 27

Research Question I: What barriers to implementing geometry assurance practices can be identified in the aero engine industry?

27

slide-28
SLIDE 28

Barrier #1:

28

slide-29
SLIDE 29

Model Form Error Barrier #1:

29

slide-30
SLIDE 30

30

Model Form Error

slide-31
SLIDE 31

Barrier #2:

31

slide-32
SLIDE 32

Discretisation Error Barrier #2:

32

slide-33
SLIDE 33

33

Discretisation Error

slide-34
SLIDE 34

NX ANSYS

34

Discretisation Error

slide-35
SLIDE 35

NX

ANSYS

35

Discretisation Error

slide-36
SLIDE 36

Barrier #3:

36

slide-37
SLIDE 37

Backwards Incompatibility Barrier #3:

37

slide-38
SLIDE 38

?

Deterministic Probabilistic

38

Backwards Incompatibility

slide-39
SLIDE 39

Barrier #4:

39

slide-40
SLIDE 40

Forward Applicability Barrier #4:

40

slide-41
SLIDE 41

?

41

Forward Applicability

slide-42
SLIDE 42

Research Question II: How can geometry assurance methods be implemented in multidisciplinary simulations in industrial settings?

42

slide-43
SLIDE 43

#1: The parametric point method

43

slide-44
SLIDE 44

44

slide-45
SLIDE 45

45

slide-46
SLIDE 46

46

slide-47
SLIDE 47

47

slide-48
SLIDE 48

48

slide-49
SLIDE 49

49

slide-50
SLIDE 50

50

slide-51
SLIDE 51

51

slide-52
SLIDE 52

52

slide-53
SLIDE 53

Reconstructing the geometry…

53

slide-54
SLIDE 54

Models

1st gen -Papers A and B 2nd gen -Paper C 3rd gen - Papers D-G

54

slide-55
SLIDE 55

#2: Integrated simulation environment

55

slide-56
SLIDE 56

#1 #2 #3

56

NX C#

Simulation workflow

slide-57
SLIDE 57

NX

#1 #2 #3 #4

57

C#

Simulation workflow

NX

slide-58
SLIDE 58

NX

#1 #2 #3 #4 #5

58

NX C#

Simulation workflow

slide-59
SLIDE 59

NX

#2 #3 #4 #5 #6

59

NX C#

Simulation workflow

slide-60
SLIDE 60

NX

#2 #3 #4 #5 #6

Ansys

60

NX C#

Simulation workflow

slide-61
SLIDE 61

#2 #3 #4 #5 #6

Ansys

61

NX NX C#

Simulation workflow

slide-62
SLIDE 62

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹`€ ¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo˜ ≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC›á ’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯’ Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚Ì ¡w‹wÔ£flO‰|

NX

#2 #3 #4 #5 #6

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”- vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹` €¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo ˜≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC ›á’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯ ’Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚ Ì¡w‹wÔ£flO‰|

Ansys Ansys

62

NX NX C#

Simulation workflow

slide-63
SLIDE 63

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹`€ ¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo˜ ≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC›á ’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯’ Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚Ì ¡w‹wÔ£flO‰|

NX

#2 #3 #4 #5 #6

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”- vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹` €¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo ˜≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC ›á’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯ ’Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚ Ì¡w‹wÔ£flO‰|

Ansys Ansys

63

NX NX C#

Simulation workflow

slide-64
SLIDE 64

#2 #3 #4 #5 #6

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”- vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹` €¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo ˜≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC ›á’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯ ’Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚ Ì¡w‹wÔ£flO‰|

Ansys Ansys

64

Uçú∆‚#pDy‰È˜ fl˜- :⁄vå{¨·”vg/jBöÚöFõSö˚[b[∫OÃ>— ÷Ífiz¸G€- ú4<YyJÛT…i⁄ÈÇ”ìgÚœåùïù}~.˘‹`€ ¢∂{ÁcŒfljoÔ∫t·“EˇãÁ;º;Œ\Ú∏tÚ €ÂW∏WöØ:_mÍtÍ<˛ì”O«ªúªöÆπ\kπ ÓzΩµ{f˜Èû7Œ›ÙΩyÒˇ÷’û9=›ΩÛzo˜ ≈˜ıfl›~r'˝ŒÀªŸw'Ó≠ºOº_Ù@ÌAŸC›á ’?[˛‹ÿÔ‹j¿w†Û— ‹G˜ÖÉœ˛ëıèCèôèÀÜ ÜÎû8>99‚?r˝È¸ßCœdœ&û˛¢˛ÀÆ/~¯’ Î◊Œ—ò— °óÚóìøm|•˝Í¿ÎØ€∆¬∆-æ…x31^ÙV˚Ì ¡w‹wÔ£flO‰|

NX NX NX C#

CFD-post CFD-post CFD-post

Simulation workflow

slide-65
SLIDE 65

65

slide-66
SLIDE 66

66

slide-67
SLIDE 67

67

slide-68
SLIDE 68

68

slide-69
SLIDE 69

69

slide-70
SLIDE 70

70

slide-71
SLIDE 71

71

slide-72
SLIDE 72

Left Right Nominal Max Thermal Stress

72

slide-73
SLIDE 73

Material fatigue

73

de Havilland Comet 1 crash, Jan 10th 1954

slide-74
SLIDE 74

Genetic algorithms (Papers D, E and G)

Genetic algorithms (GA) are search procedures that mimic the mechanics of natural selection and genetics (Goldberg 2013).

74

slide-75
SLIDE 75

?

  • 11 vanes
  • 10! = 3,628,800 different combinations!
  • 10 minutes per fatigue life simulation ≈ 70 years of simulations

A combinatory problem

75

slide-76
SLIDE 76

Paper D

76

slide-77
SLIDE 77

Results

77

slide-78
SLIDE 78

Let’s change a design parameter!

Paper E

78

slide-79
SLIDE 79

Blend radius (mm)

MPa Max von Mises stress

79

slide-80
SLIDE 80

Paper G

80

slide-81
SLIDE 81

Research Question III: What role should geometry assurance play in the early phases of aerospace component design?

81

slide-82
SLIDE 82

Source: Oberkampf, Error and Uncertainty in Modeling and Simulation, 2002

82

slide-83
SLIDE 83

83

slide-84
SLIDE 84

84

slide-85
SLIDE 85

Conclusions

85

slide-86
SLIDE 86

Conclusions –Research Questions I & II

Identified barriers Proposed solutions Model form error Parametric point method (PPM) Discretisation error Backwards incompatibility COTS Software independence and redundancy Forward applicability

86

NX

ANSYS

?

slide-87
SLIDE 87
  • Geometric variation has the most major impact within thermal

fatigue analysis.

  • The case studies showed that simulation results were heavily

affected by geometric variation in parts and assemblies.

  • Welding exacerbates the effects of geometric variation.
  • Genetic algorithms can be used to mitigate these effects.

Conclusions –Research Question III

87

slide-88
SLIDE 88
  • We should improve the quality and technological maturity of the multidisciplinary simulation

environment, including its breadth as well its depth.

  • To make this research more comprehensive, more researchers should be invited to contribute

within their respective fields of expertise.

  • The weakest links of the simulation chain are currently those activities that stray from the

established and validated industrial design practices. Currently, these tools and methods lie within TRL 3-6. They need to be further matured.

Future work

88

slide-89
SLIDE 89

Thank you!

89