Advanced Fatigue from nCode

Robert Cawte

Principal Applications Engineer

HBM-nCode

1
 Contents

Introducing nCode ANSYS DesignLife

for ANSYSTM Workbench 13

Case Study 1 Lever Assembly Example

4 How to do Lifing analysis in Workbench
4 Design optimisation based on Life

Case Study 2 Wind Turbine Gearbox

4 What makes a good fatigue analysis?
4 Element types, mesh quality, stress convergence, support
for shells and solids in DesignLife

Case Study 3 Wind Turbine Main Shaft

4 Duty cycle loading in DesignLife
4 Multiaxial fatigue considerations
4 Selection of critical areas by damage criteria

2
What is nCode DesignLife ?

A fatigue analysis solver & interface for FE users

4Part of nCode GlyphWorks (signal processing)
To encourage test-FE collaboration
4Customised for Ansys users (DesignLife only)

DesignLife is both flexible and rigid !

4Flexibility
to set up the required analysis options
4Rigid options are saved in a locked process flow
Re-used to give consistent process results
DesignLife works from Ansys RST files
4and/or sits inside Workbench
3
Fatigue Life Prediction Process

3 Main Approaches:
Stress-Life (SN Analysis)
Strain-Life (EN or Crack Initiation Analysis)
Crack Growth (LEFM)
Material
Properties

Geometry Damage Fatigue

(FEA) Analysis Life

Load
History

4
Fatigue Life Prediction Process - Loading History

Fatigue loading is not just cycling a static loadcase

4Mostrealistic loading is variable amplitude
4Many are multi-channel
4Many are multi-event (and multi-channel) such as ISO/IEC 61400

5
Fatigue Life Prediction Process

Rainflow Histogram
Rainflow Cycle
Time Signal Counting 0.4

0.3
Cycles
(Count)
0.2

0.1

0
3 3
0 1 10 2 10
Range (MPa)

Fatigue cycles
3
2 10

3
1.5 10
Damage

LIFE 1 10
3

4
5 10

Palmgren- 0
3 3
0 1 10 2 10
Miners Range (MPa) 6

damage
summation Damage Histogram SN Analysis
Case study 1 Lever Assembly Example

Project schematic shows:

4Steady-state thermal analysis
4Static structural analysis
4Strain life constant amplitude analysis with
thermal correction

7
 Case study 1 Lever Assembly Example
Steady-state Thermal Analysis
Results

8
 Case study 1 Lever Assembly Example
Static Structural Analysis
Results

9
 Case study 1 Lever Assembly Example

nCode ANSYS DesignLife Fatigue

Results Automatic
hotspot
detection

Temperature
& Stress
results from
Workbench
Fatigue
solver

Bill of
Materials from
Workbench Export to ANSYS
Design Optimizer Preconfigured report output

10
 Case study 1 Lever Assembly Example

Full integration with parametric and optimisation capabilities of

Workbench

11
 Case Study 2 Wind Turbine Gearbox

Generalized Gearbox Schematic1

Ref 1. Improving Wind Turbine Gearbox Reliability

W. Musial and S. Butterfield National Renewable Energy Laboratory, B. McNiff McNiff Light Industry
Presented at the 2007 European Wind Energy Conference Milan, Italy , May 710, 2007
12
Case Study 2 Wind Turbine Gearbox Casing

1 million nodes
Linear or quad elements
Solids or skim with shells

13
Case Study 2 Unit loads applied at bearing positions

Typical 2-3MW rated Max torque ~<1MNm

Gearbox >1m diameter 750kN reaction / side (1kN nominal applied)
14
Case Study 2 Linear Static Superposition unit load results

15
 Case Study 2 Linear Static Superposition constant amplitude

Maximum load, 15 active loadcases (bearing reactions)

Available loadcases Selected for combination Scale factors

16
Case Study 2 Fatigue results

Stress results
Life results
Life results
Linear Quad
elements
elements
17
 Case Study 2 What makes a good FE-fatigue analysis?
Material fatigue properties
SN Fatigue Test Analysis

The regression curve represents

the life at which 50% of samples
have failed.
Design curve shown with 95%
Certainty of Survival and 95%

Stress Range S
Confidence
The more samples you test the
more confident you are about the
scatter Supplier A Test Data
Supplier A Regression fit
Compare batch/supplier variability 2 sigma range
2 sigma range
Supplier A Design curve
Supplier B Test Data

Number of Cycles to failure N

Supplier A
Supplier B
nCode Materials Testing /
Assurance Service now
4 2 0 2 4 6 8
available to ANSYS users!
18
Case Study 2 What makes a good FE-fatigue analysis?

Good Surface Stress Results!!!

Fatigue cracks usually initiate at free surfaces
Fatigue damage increases exponentially with stress

10% error in stress 100% error in life!

Shell models
FE Accuracy/convergence Low mesh density
Stresses calculated at
1. Nodal displacement
Gauss points and OK for load path &
extrapolated to node 2. Nodal forces & moments
natural modes
Node has separate stress
result from each element 3. Element Gauss point stresses
Most FEA uses average 4. Node at element stresses Required for
nodal stress 5. Nodal averaged stresses Fatigue
High mesh density

Recommendations
Best choice = Gauss point stresses or Node at element
Best stress accuracy & check for convergence
Use nodal averaged only if you are certain of convergence
19
Case Study 2 What makes a good FE-fatigue analysis?

Good Surface Stress Results!!!

Fatigue cracks usually initiate at free surfaces
Fatigue damage increases exponentially with stress

10% error in stress 100% error in life!

Solid models
Stresses calculated at Gauss points but Gauss points z
x
are not on the surface!

Option 1 skin surface with membrane or thin shells y

Shells resolve stresses to surface plane
Use element stresses from the shells as before

Option 2 Use surface node results

Transform stresses to surface plane
Use Node at element (or Nodal averaged stresses)
on surface nodes only
Node at element are preferred!
Preferred
20
Case Study 2 Fatigue results

Node Node-on-element fatigue life results

Linear Quadratic
204611 Shell skin 18,000 to 100,000 Shell skin 2,676 to 126,000
Solid 8,800 to 82,000 Solid 17,800 to 44,000

Preferred solution
Model not fully refined but demonstrates the importance of:-
Linear elements converge very slowly (and over-stiff)
Shell skins on solid models help resolve surface stresses but do not
improve accuracy of stress results
Shell skins also prevent stress gradient fatigue techniques
A well refined quadratic mesh is recommended.
Node-on-element* fatigue results demonstrate the degree of convergence
21
Case Study 2 a good FE-fatigue analysis refined model fatigue results
Case study 3 Analysis of main drive shaft

23
Case Study 3 Linear Static Superposition

24
Case Study 3 Duty cycle load definition

5 years
10 years
4 years
1 year
20 (in 20 years)

20 year target life

Duty cycle breaks life down
into discrete events which
are repeated
Only need to analyse short
event histories and multiply
damage using Miners rule

25
Case Study 3 Fatigue results from hot-spot detection runs

Combined
Damage

26
Case Study 3 Fatigue using full multiaxial analysis

27
Modular structure of ANSYS nCode DesignLife

Package/Module Description
ANSYS nCode DesignLife Standard Base package including Stress-Life, Strain-Life
and Dang Van analyzers

ANSYS nCode DesignLife Modules/add-ons

ANSYS nCode DesignLife Vibration Adds ability to do vibration fatigue

analysis. Simulate swept sine and PSD loadings

ANSYS nCode DesignLife Accelerated Testing Signal processing package complementary to

vibration option. Design accelerated virtual and
physical vibration tests.

ANSYS nCode DesignLife Welds Fatigue life prediction for seam welds and spot
welds.

ANSYS nCode DesignLife Parallelization DesignLife is multi-threaded and licensed per core
with one core included in the base package.

28
Capabilities

ANSYS nCode DesignLife has an extensive scope of fatigue capabilities

Stress-Life (single, multi-curve, Haigh diagrams)

Strain-Life (automated multi-axial corrections)

Multi-axial safety factor (Dang Van) Finite Element results supported

4 Static (linear superposition)
Seam welds and spot welds 4 Transient

High temperature fatigue 4 Modal

4 Frequency Response
Vibration fatigue (shaker simulation)
4 Linear & Non-linear
Multiple runs in a single analysis

Complete duty cycles / flight spectrums

Multi-processor enabled for fast results

Use Python for proprietary or custom methods

29
 Conclusion

Introducing nCode ANSYS DesignLife

for ANSYSTM Workbench 13.0

Case Study 1 Lever Assembly Example

4 How to do Lifing analysis in Workbench
4 Design optimisation based on Life

Case Study 2 Wind Turbine Gearbox

4 What makes a good fatigue analysis?
4 Element types, mesh quality, stress convergence, support
for shells and solids in DesignLife

Case Study 3 Wind Turbine Main Shaft

4 Duty cycle loading in DesignLife
4 Multiaxial fatigue considerations
4 Selection of critical areas by damage criteria

30
 Robert Cawte
HBM United Kingdom
Thanks Tel: +44 (0) 121 733 1837
[email protected]

31