7/28/2019 05-Tessmer
1/29
Institute of Structural Mechanics 1
Virtual Testing of Aircraft Structures, considering
Postcritical & Thermal Behavior
J. Temer, R. Degenhardt, A. Kling, R. Rolfes, T. Sprwitz, S. Waitz
(DLR Institute of Structural Mechanics, Braunschweig, Germany),
COMPOSIT Thematic Network,
Workshop on Modeling and Prediction of Composite Transport Structures
in Zaragoza, Spain, 30.06.2003
7/28/2019 05-Tessmer
2/29
Institute of Structural Mechanics 2
Reality
Experiment
Computer
Model
Model
Validation
ModelQualification
Model
Verification
Computer
Simulation
Analysis
Programming
Conceptual
ModelModel Validation:
Solve the rightequations
Model Verification:
Solve the equations right
Validation / Verification
7/28/2019 05-Tessmer
3/29
Institute of Structural Mechanics 3
Postcritical behavior of stiffened panels
Top plate
Clamping box
Specimen
Clamping Box
Displacement
pickup
Load distributor
Load cells
Drive plate
Panel in Buckling Test Facility Deformation pattern (ARAMIS)
7/28/2019 05-Tessmer
4/29
Institute of Structural Mechanics 4
Non linear FEM using ABAQUS/Standard
rough
estimate
FE-Model
Linear Eigenvalue Analysis
Buckling Load
Nonlinear Analysis
Newton-Raphson-Method + automatic / adaptivedamping to stabilize the analysis (*STATIC, STABILIZE)
scaled imperfektions
Postprocessing
(Load-Shortening-Curve, deformation of the structure, ...)
Buckling Modes
Real Structure
CFRP-Panel
Measured
Imperfections
rough
estimate
FE-Model
Linear Eigenvalue Analysis
Buckling Load
Nonlinear Analysis
Newton-Raphson-Method + automatic / adaptivedamping to stabilize the analysis (*STATIC, STABILIZE)
scaled imperfektions
Postprocessing
(Load-Shortening-Curve, deformation of the structure, ...)
Buckling Modes
Real Structure
CFRP-Panel
Measured
Imperfections
7/28/2019 05-Tessmer
5/29
Institute of Structural Mechanics 5
Experimental Validatition of computational Analysis is expensive & time consuming
Numerical Pre-Test Analysis
Pre-Test Analysis and Pre-Test Planing
Validation Experiments versus Phenomenological Experiments
Goal: Load Deformation curve showing:
- characterristic skin buckling, coupeled with axial stiffness reduction
- large load bearing capacity during postbuckling without structural failure
FEA investigation w.r.t.: panel geometrie
discretisationexperimental boundary conditions
initial imperfections
...
7/28/2019 05-Tessmer
6/29
Institute of Structural Mechanics 6
Pre-Test Analysis1. Influence of different STABILIZE parameters
2. Investigation of different failure criteria
0
10
20
30
40
50
60
70
80
0 0.5 1 1.5 2 2.5 3 3.5 4
Shortening [mm]
Load[k
N]
STABILIZE = 2e-4
STABILIZE = 2e-5
STABILIZE = 2e-6
STABILIZE = 2e-7
Nominal data
ABAQUS/Standard
Mesh I (3024 elements)
Tsai Hill
Azzi Tsai HillMaximum Stress
Tsai Wu
7/28/2019 05-Tessmer
7/29
Institute of Structural Mechanics 7
Pre-Test AnalysisConvergence study
0
10
20
30
40
50
60
0 0.5 1 1.5 2 2.5 3 3.5 4Shortening [mm]
Load[kN]
Mesh I: 3024 elements
Mesh II: 2*3024 elements
Mesh III: 16*3024 elements
Nominal data
ABAQUS/Standard
STABILIZE = 2.e-6
7/28/2019 05-Tessmer
8/29
Institute of Structural Mechanics 8
Pre-Test AnalysisInfluence of lateral boundary conditions
0
10
20
30
40
50
60
70
80
90
100
0 0.5 1 1.5 2 2.5 3 3.5 4
Shortening [mm]
Load
[kN]
Covered width = 25.0 mm
Covered width = 12.5 mm
Only lateral nodes fixed
Nominal data
ABAQUS/Standard
STABILIZE = 2.e-6
Mesh I (3024 elements)
Edge support
Filler
Gliding plane
Test panel
Detail
7/28/2019 05-Tessmer
9/29
Institute of Structural Mechanics 9
Pre-Test AnalysisInfluence of imperfections
0
10
20
30
40
50
60
0 0.5 1 1.5 2 2.5 3 3.5 4
Shortening [mm]
Load[kN]
No imperfections
Mode 1, 2% skin thickness
Mode 1, 10% skin thickness
Mode 1, 100% skin thickness
Nominal data
STABILIZE = 2.e-6
Mesh I (3024 elements)
7/28/2019 05-Tessmer
10/29
Institute of Structural Mechanics 10
Pre-Test AnalysisABAQUS/Standard vs. ABAQUS/Explicit
0
10
20
30
40
50
60
0 0.5 1 1.5 2 2.5 3 3.5 4
Shortening [mm]
Load
[kN]
ABAQUS/Standard, STABILIZE = 2e-6
ABAQUS/Explicit, v=10mm/s, no damping
ABAQUS/Explicit, v=10mm/s, with damping
Nominal data
Mesh I (3024 elements)
7/28/2019 05-Tessmer
11/29
Institute of Structural Mechanics 11
Measurement of geometrical Imperfections (1)
Optical 3D - digitalization
ATOS Sensor strip sequenz
Measurment of real radius vs. nominal radius (ca. 6% deviation)
Measurement of initial imperfection
7/28/2019 05-Tessmer
12/29
Institute of Structural Mechanics 12
Measurement of geometrical Imperfections
Fringe plot w.r.t.perfect shell
Data points (ASCII-Format):
1093.6441 211.5491 1.6805
1093.1718 211.1541 1.6764
1093.8102 210.6003 1.6764
1093.0879 210.3660 1.6910
Modification of perfect
FE geometry
Application of initial
imperfection
7/28/2019 05-Tessmer
13/29
Institute of Structural Mechanics 13
Results of FEA using ABAQUS/Standard
0
0,5
1
1,5
2
2,5
0 0,5 1 1,5 2 2,5 3 3,5 4
Skalierte Verschiebung
SkalierteL
ast
ABAQUS/Standard ohne Imperfektionen
ABAQUS/Standard mit Imperfektionen
Lokal skin buckling
Global unsymmetric Buckle
Global symmetric
Buckle
7/28/2019 05-Tessmer
14/29
Institute of Structural Mechanics 14
Validation of FEA (Animation)
7/28/2019 05-Tessmer
15/29
Institute of Structural Mechanics 15
Validation of computational results
Globale Ebene
0
0,5
1
1,5
2
2,5
0 0,5 1 1,5 2 2,5 3 3,5 4
Skalierte Verschiebung
Skalierte
Last
Experiment (1)
Experiment (2)
ABAQUS/Standard mit Imperfektionen
7/28/2019 05-Tessmer
16/29
Institute of Structural Mechanics 16
-10
-5
0
5
10
15
20
25
0 0,5 1 1,5 2 2,5 3
Skalierte Verschiebung
Radialverschiebu
ng[mm]
Wegaufnehmer W88
Knoten 40696; entspricht W88 PositionWegaufnehmer W89
Knoten 46666; entspricht W89 Position
Wegaufnehmer W90
Knoten 52636; entspricht W90 Position
Wegaufnehmer W91
Knoten 58606; entspricht W91 Position
Validation of computational results
Lokale Ebene
W88W89
W90W91
7/28/2019 05-Tessmer
17/29
Institute of Structural Mechanics 17
directSun Radiation
thermalRadiation
Reflected
Sun Radiation
Cross SectionFuselage
InsideOutside
Time
Temperature
Taxiing Stop Take Off
Thermal Problem
FMLs in future aircraft structures
7/28/2019 05-Tessmer
18/29
Institute of Structural Mechanics 18
Modeling (on panel level)
7/28/2019 05-Tessmer
19/29
Institute of Structural Mechanics 19
345
350
355
360
365
370
375
380
-2,70
Thickness-Coordinate of skin (mm)
Temperatur(K)
layered skin
smeared homogenous skin
Model Verification
Convergence study with 6 different
discretisations
(1 to 36 elements in thickness direction)
Homogenisation of smeared layers with:
_ ,
_ ,
/ ( / ) ,
( ) / .
out plane i i i normal i i
in plane i plane i ii i
k t t k
k k t t
7/28/2019 05-Tessmer
20/29
Institute of Structural Mechanics 20
Infrared Radiator
Isolation (optional)
Skin
Water
Frame
Stringer
Experimental Test in THERMEX B test site
7/28/2019 05-Tessmer
21/29
Institute of Structural Mechanics 21
20
30
40
50
60
70
0 2000 4000 6000 8000 10000
Zeit (sec)
Temperatur(C) MP43/TE43
MP43_RF1
MP34/TE04
MP34_RF1
MP24/TE09
MP24_RF1
Validation
Experiment vs. Computation
7/28/2019 05-Tessmer
22/29
Institute of Structural Mechanics 22
Modeling of large fuselage structure
7/28/2019 05-Tessmer
23/29
Institute of Structural Mechanics 23
2D Finite Elements for Thermal Analysis of FMLs
Motivation
Reduction of modeling effort by using 2D geometrical models
Reduction of CPU-time
Compatible temperature field for thermo-mechanical calculations
Scientific Challenge
3D temperature field description based on 2D geometry
Shape functions in z-direction
2D-3D-Coupling (Connection to local 3D meshes)
7/28/2019 05-Tessmer
24/29
Institute of Structural Mechanics 24
3D temperature distribution by 2D finite elements
Layerwise Thermal Lamination Theories
Idealisation as layered structure
Homogenisation of layers
including heat conduction,
radiation, convection
z
t
t
z +d-z
z
N
z=0
1
1
k
k kk
b
1
k
2
Linear Layered
Theory (LLT)
Quadratic
Layered Theory
(QLT)
T
T
T
0
0 , z
( k )
( k )
( z )
Face sheet
Honeycombcore
Hybrid composite structures
Composite structures
Sandwich structures
Aluminum sheet
Fiber/resin
7/28/2019 05-Tessmer
25/29
Institute of Structural Mechanics 25
Assumptions and Prerequisites
perfect thermal contact at interfaces
monolithic conduction within layers
no internal heat sources
temperature independent material
properties
convection
conduction
radiation
radiation
equivalent thermal conductivity
conductionconvection
homogenisation
N
1
2
k
b
t1
tk
z
z = 0
z1
zk + d
k
-zk
approximation by layered construction
7/28/2019 05-Tessmer
26/29
Institute of Structural Mechanics 26
FE-Formulation
(Weak form for heat conduction)
( ) 0ddd W+G+W WGW
vTvnqTKv &rcgradgrad TT
Boundary conditions
- Convection(Robin)
- Heat flux density(Neumann)
q
q T Tc c - a Wandb g
q nT
cq q +
FE-Formulation
Weak form for heat conduction
7/28/2019 05-Tessmer
27/29
Institute of Structural Mechanics 27
FE-Formulation
(Weak form for heat conduction)
h r a h r h r
a h
T T
zA
c
T T T T
zA
cT T
z A c z A
T q
SKS RR RR
R
d d d d d
d
zz z zzz
+ +
-
G
G
G
G
b g
$K SKS
S K S
z
z+
T
z
kT
k k
z
z
k
N
z
z
k
k
d
dbg bgbge j
1
1
Composite-heat-conduction-matrix Composite-heat-capacity-matrix
$
( )
C RR
R R
z
z+
c z
c z
T
z
k k
k T k
z
z
k
N
k
k
d
dr c h bg1
1
FE-Formulation
Weak form for heat conduction
7/28/2019 05-Tessmer
28/29
Institute of Structural Mechanics 28
Example: 3D vs. 2D FEA
Number of 3D-elements: 2
Number of 3D-elements: 36
Number of 2D-elements: 1
GLARE skin with 3D finite elements (Nastran)
345
350
355
360
365
370
375
380
-2,70
Thickness-Coordinate of skin (mm)
Temperatur(
K)
layered skin
smeared homogenous skin
Thermal Lamination Theory (QLT)
7/28/2019 05-Tessmer
29/29
Institute of Structural Mechanics 29
Summary
Experimental data basis for validation of non linear FEA w.r.t. postbuckling of
stiffened shells under axial loading
Investigation of sensitivity w.r.t. to different modeling parameters
Excellent agreement between experimental and computational results deep into
elastic postcritical regime (global & local)
Reliable thermal analysis of FML structures by verification of discretisation throughfine 3D model on panel level
Validation of thermal panel model by experiments in THERMEX B test site
Application of thermal model to large fuselage structures
Description of fast 2D Finite-Element-Formulation
Question: How many experiments are needed for validation
of a specified parameter space?