This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• Turbine Housing– Also called turbine scroll contains the radial turbine wheel and directs gas
into it.– Mostly made of cast iron alloys for turbocharger applications.– During operation these components experience severe thermo
mechanical (TM) load.– Severe TM loading can lead to cracks initiating and propagating with
accumulation of cycles. – Presence of a crack is not necessarily a failure unless it leads to loss of
functionality (gas leakage, wheel rub etc) or is externally visible.– Knowledge of how cracks grow (path) and how fast they grow (growth
rate) is important.– Methods/tools that can help predict path and growth rate of cracks:
• Help design products quantitatively as opposed to a ‘better than’ approach.
• Increase confidence during testing phase.• Lead to more reliable products.
Introduction
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
A typical accelerated test duty cycle.A typical free floating Turbine housing.
Typical housing failures via cracking.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• Several LEFM based, Mode I (planar) crack growth modeling codes and methods are popular and commercially available.
• However, due to the nature of loading and complexity of geometry above assumptions are not applicable to our case
• Large scale and even cross sectional yielding can occur in certain locations.– Cracks have been observed to exhibit mixed mode growth behavior (turning and
kinking)– Thermal loads can only be accounted for via an indirect approach in the above case.
• A conventional displacement (K) based approach is also an implementation challenge in 3D
– To capture the singularity in strain at the crack tip a focused/collapsed mesh is needed.
– Creating such a mesh at the crack front at each crack increment is near impossible in an automated manner.
• Developing an automated 3-d mixed mode crack growth code capable of dealing with elastic-plastic condition is a technical challenge !
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• J-integral is a path independent fracture mechanics parameter which predicts energy release rate and intensity of deformation at crack front (CF) for linear and nonlinear material behavior.
• W is strain energy and it is a point function. It varies from point to point within body of the component [13]
• ANSYS calculates J values using volume integral in terms of rings of elements at the crack front.
• The highest energy release rate is calculated for virtual crack extension angle for all the CF nodes and new crack front is extended in the direction of max energy value.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Evaluation of Crack growth vs Cycles : Constant Amplitude Approach
- During fatigue crack growth under LEFM assumptions, the relation between incremental crack growth and load cycles can be represented by Paris law.
- C and m are material properties and need to be evaluated (ASTM D647) .- However, under large scale yielding condition, J-integral approach needs to be directly implemented. This also involves non standard testing for estimating fcg material properties.
Crack growth calculation based on block technique:
N is cycles per Block
EKE
Kv
2
22 )1( −
J = Plane Strain.
Plane Stress
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• Sub model creation, crack feature insertion handled by CAD.
• Meshing, preprocessing, analysis and path/growth prediction handled by ANSYS.
• Logic and links implemented via in-house scripts (APDL and CAD package).
• Process is fully automated after initial set-up.• Can handle both elastic or elastic-plastic
analysis.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Predicted crack path (cruciform specimen under uniaxial loading).
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Gives more confidence to J integral approach implementation and mixed mode capability !
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Transition from surface crack to through crack automated using a c/s stress based criterion.
Validation runs: Quantitative.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
1. Differences in NASGRO and FE approaches are due to location of crack increment size.
2. NASGRO calculates next crack increment by considering only ∆a from free surface whereas ANSYS takes highest ∆a value from the crack front. This leads to more conservative results.
Validation runs: Quantitative.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• Global stress results used for crack path prediction.
Real life crack geometry.
Turbine Housing.
Center/bearing Housing.
V Band Connection.
Mounting base.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Actual crack path after tests.Predicted crack path.
Qualitative agreement between real life and analysis!
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
The developed methodology reduces the crack growth analysis time of test specimen by more than 90%
Results extraction and Crack growth Calculation for a
given load spectrum.
25 minutes
100 minutes
25 minutes45 minutes
2.5 minutes
Total cycle time for simulation of one crack increment
(excluding solution)
100
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• Elevated temperature fcg testing being currently conducted using M(t) specimen.
• In-house custom fcg test being developed for validation purposes.• Algorithms being further refined for robustness, accuracy and speed.• Further validation work is ongoing using literature cases.• Ansys Collaboration :• Bugs with CINT fixed in V11.0 SP1.• Thermal effects included in J calculation in V12.0 preview version.
– Tested and validated by HTT.• Need Ansys to implement J calculation using tetrahedral elements.• Need option for J value decomposition. (JI, JII, JIII)
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
• The J integral (energy) approach to fcg modeling was effectively used to develop a 3-d mixed mode capable process/methodology.
• J integral approach has several benefits over K based approach:– Ease of implementation as meshing requirments at cf are much
more relaxed.– Capable of dealing with large scale yielding problems.– Directly account for thermal loading.
• Automation via scripts make the process feasible in a production environment. ( ~ 500 designs analyzed/year)
• Some baseline validation runs show good promise for the methodology.• Qualitative validation using fcg material test data will pose its own
challenges. (non isothermal Vs isothermal, LEFM vs NLFM, closure effects etc)
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
1. H.A Richard, M Fulland and M sander, Theoretical Crack Path Prediction, Fatigue Fracture Engg. Mater Struct. 28, 3-12, 2005
2. Shih, C. F., B. Moran, and T. Nakamura, “Energy Release Rate Along a Three-Dimensional Crack Front in a Thermally Stressed Body,” International Journal of Fracture, vol. 30, pp. 79–102, 1986.
3. Daniel Bromberg, Guido Dhondt, Automatic crack-insertion for arbitrary crack growth, Engineering Fracture Mechanics, 9 January 2007 ( Publication is in Press)
4. A.F Liu, Structural Life Assessment Methods, ASM International5. L.P Pook, Crack Paths, WIT Press, 20026. L.P Pook, Metal Fatigue, What is , Why it matters, Springer,20077. H A. Richard, Computational simulation and experimental results on 3D crack growth
in a 3PB specimen with a inclined plane, Key Engineering Material, Vol 251-252 (2003) pp 85-90
8. ANSYS Manuals,9. NASGRO 5.1 Manuals10.J. H Kuang, Crack initiation direction for a crack under mixed mode loading, Inter.
Journal of Fracture, 73 33 -37, 199511.L. P Pook, H.A Richard, Crack paths, Engineering Fracture Mechanics, may, 200712.D.M Parks, Application of Domain Integral Methods using Tetrahedral elements to
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.