1. NSM LAB
Net Shape Manufacturing Laboratory
Development of oil canning index model for sheet metal forming products with large curvature
Speaker: M.S. Eng. Mohanraj Murugesan
Ph.D. Candidate/Researcher – Sogang University (South Korea)
정형가공연구실

2. Presentation Outline What’s oil-canning
Motivation and overall objective
Material properties estimation
Stamping process and indentation test results
Proposed oil canning index model
Model validation and verification
Summary
On-going research work

3. Test procedure to estimate minimum load to cause oil-canning.
What's oil-canning
Oil canning is a moderate deformation or an elastic buckling problem of sheet material & usually occurs at unsupported area of a body panel when a load is applied to the area. 
This terminology also refers to the popping sound made when pressure is applied to the deformed sheet forcing the deformation in the opposite direction.
Reasons:
Effect is extremely large when panel thickness is very small & panel width is too large.
Effect is increased when sheet metal has large curvature & low panel stiffness.
(stamping/deep drawing) Caused by unequal stretching of the material.
Typically caused by uneven stresses at the fastening points.
Test procedure to estimate minimum load to cause oil-canning.

4. Motivation Automobile Parts Roof Panels Home Appliances
Aircraft Components

5. Overall Objective Oil-canning experiment Identification of main
factors
Develop oil-canning prediction model
Sensitivity analysis
Modify material properties
Oil-canning prediction model
Model validation
Model verification
Automobile parts
Home Appliances
Applications
Aircraft Components
Roof Panels

6. Material Properties Estimation
A measure of planar anisotropy
A measure of normal anisotropy
𝑅 = 𝑅 0 +2 𝑅 45 + 𝑅 90 4
R0, R45 and R90 represents anisotropy values at directions, 0°, 45° and 90°, respectively.
Fig 1. Tensile test set-up
Fig 2. Stress-strain curve from tensile test
ε 𝑤 : strain at width direction
ε 𝑡 :strain at thickness direction
ε 𝑙 : strain at length direction
R=− ɛ 𝑤 ɛ 𝑡 + ɛ 𝑙
Lankford variable
Material R0
R45
R90 ΔR 𝑹
EGI steel
3.0414
1.8586
3.3977
1.3609
2.5391
GI steel
2.3767
2.7833
3.4373
0.1237
2.8451
Table 1. Material anisotropy values.

7. Stamping Process Initial plate size : 280 x 375 mm 2
Blank holding force : 250 N
Plate thickness : 0.6, 0.7, 0.8
Friction co-efficient : 0.1
Materials used : GI steel and EGI steel
Punch velocity : 13 m/s
Embossing depth : 4mm
Bead length : (a) 0mm (b) 7.854mm (c) mm
Fig 3. Stamping process experiment setup.
Fig 4. Stamping process FE model.

8. 3D scanning procedure Purpose: To measure the curvature of plate
Measurement of thickness distribution using the formed part
Verification with numerical model
(Curvature and thickness distribution comparison)
Curvature
Specimen
3D scanning camera
Fig 6. 3D Scanning Procedure
Fig 5. 3D scanning set-up with stamped part
Fig 7. Stamped part for scanning

9. Horizontal curvature ( 𝑚 −1 )9
Curvature measurements from 3D Scanning P1 P2 P3
0.6mm(EGI Steel)
case
Vertical bead force
Horizontal bead force
Horizontal curvature ( 𝑚 −1 )
Vertical curvature
( 𝑚 −1 )
Horizontal curvature 1
No bead
0.121
0.105
1.14
0.189
0.145
0.348 2
Low
0.173
0.122
0.217
0.069
0.17
0.77 3
High
0.408
0.788
2.82
0.055
0.316
0.682 4
0.388
0.902
3.68
0.44 5
0.79
0.09
0.109
0.065 6
0.039
0.832
0.175
0.05 7
0.06
0.089
0.764
0.235
0.072
0.076 8
0.054
0.939
0.161
0.056
0.015 9
0.058
1.06
0.176
0.014 P2 P3 P1
Horizontal
Vertical
0.7mm(GI Steel)
case
Vertical bead force
Horizontal bead force
Horizontal curvature
( 𝑚 −1 )
Vertical curvature 1
No bead
0.842
1.06
1.39
0.084
0.853
1.08 2
Low
0.019
0.441
0.278
0.013 3
High
0.647
0.796
1.38
0.53
0.804 4
0.576
1.79
0.608
0.955 5
0.053
0.592
0.07
0.045 6
0.02
1.13
0.153 7
0.081
0.437
0.33 8
0.269
0.133
1.55
0.083
0.293
0.285 9
0.098
1.04
* Curvature value = 0  this means the specimen is flat and there is no curvature

10. Curvature measurements from 3D Scanning10
Curvature measurements from 3D Scanning P1 P2 P3
0.8mm(EGI Steel)
case
Vertical bead force
Horizontal bead force
Horizontal curvature
( 𝑚 −1 )
Vertical curvature 1
No bead
0.15
0.221
1.45
0.19
0.036
0.053 2
Low
0.052
0.038
0.888
0.232
0.051
0.04 3
High
0.241
0.247
1.95
0.23
0.25 4
0.451
0.466
1.36
0.348
0.422 5
0.071
0.9
0.203
0.081 6
0.011
0.666
0.158
0.033 7
0.084
0.805
0.17
0.028 8
0.019
1.16 9
0.022
1.09
0.212 P2 P3 P1
Horizontal
Vertical
* Curvature value = 0  this means the specimen is flat and there is no curvature

11. Indentation test procedure
Purpose: To determine whether oil canning phenomenon occurs or not  in the stamped part.
Obtain load-displacement curve. 
Choose three points where the curvature is most likely to occur in the stamped part.
Punch
Stamped part
Specimen holder
Fig 8. Indentation test using stamped part
Fig 11. Load-stroke curve for
0.7mm GI steel
Fig 10. Load-stroke curve for
0.7mm GI steel
Fig 9. Finite element model for numerical simulation

12. Indentation test results12
Indentation test results
Purpose: To determine whether oil canning exists for the particular test location and what is the minimum load to cause oil canning.
Bead force in vertical location high and no bead force in horizontal location.
Bead force in horizontal location low and no bead force in vertical location.
Oil canning at locations P1 and P3.
Bead force in vertical location low and no bead force in horizontal location.
Oil canning at location P2. P2 P3 P1
Horizontal
Vertical
0.6mm (EGI Steel)

13. Indentation test results13
Indentation test results
Bead force in vertical location high and no bead force in horizontal location.
Bead force in horizontal location low and no bead force in vertical location.
Oil canning at locations P1 and P3.
Oil canning at location P2.
No bead force in horizontal and vertical location. P2 P3 P1
Horizontal
0.7mm (GI Steel)
Vertical

14. Indentation test results14
Indentation test results
Bead force in vertical location high and no bead force in horizontal location.
Bead force in horizontal location low and no bead force in vertical location.
Oil canning at locations P1 and P3.
Low bead force in both horizontal and vertical location. P2 P3 P1
Horizontal
0.8mm (EGI Steel)
Vertical

15. Indentation test results
According to thickness (t), curvature (R) and residual stress ( 𝜎 1 , 𝜎 2 ) :
In the absence of initial stress, the smaller radius, R, (the greater curvature), the possibility of no oil-canning.
In the absence of initial stress, as the thickness increases, the buckling load also increases. The possibility of oil-canning is decreasing.
0.6mm (EGI Steel)
0.7mm (GI Steel)
Fig 12. Load-displacement curve from indentation test for GI and EGI steel

16. Indentation test results
If the R value is less than 1000, the oil canning effect is reduces and vice versa.
Here the thickness is kept constant as 0.6mm for all the experiments.
0.6mm (EGI Steel)
R = 6000 mm
R = 1000 mm
R = 600 mm

17. Fig 13. Radius of curvature estimation
Proposed Oil Canning Index Model
For oil canning damage value estimation, the empirical model can be expressed as follows:
A00 to A20 are the material constants.
t*, σ* represents the geometric parameters and the residual stresses, respectively.
Based on the geometric factors, we propose a mathematical model for the risk of oil canning effect as follows:
𝑓( 𝑡 ∗ , 𝜎 ∗ )= 𝐴 00 + 𝐴 10 𝑡 ∗ + 𝐴 01 × 𝜎 ∗ + 𝐴 20 (𝑡 ∗ ) 2 + 𝐴 11 𝑡 ∗ 𝜎 ∗ + 𝐴 𝜎 ∗ 2
R : Radius of curvature
w: Panel width
T : thickness
ν : Poisson’s ratio
k : width to length ratio
E : Young’s modulus
If OC ≥ 1, Oil Canning
If OC < 1, No Oil Canning
Fig 13. Radius of curvature estimation

18. Proposed Oil Canning Index Model
“OCPF (Oil Canning Prediction Factor)” model
“OCDV (Oil Canning Damage Value)” model
𝑓( 𝑡 ∗ , 𝜎 ∗ )= 𝐴 00 + 𝐴 10 𝑡 ∗ + 𝐴 01 × 𝜎 ∗ + 𝐴 20 (𝑡 ∗ ) 2 + 𝐴 11 𝑡 ∗ 𝜎 ∗ + 𝐴 𝜎 ∗ 2
𝑡 ∗ = thickness radius × length width
𝜎 ∗ = 𝜎 1 + 𝜎 2 + 𝜎 1 𝜎 2
𝒇 𝒕 ∗ , 𝝈 ∗ =𝟏𝟗.𝟗−𝟕𝟑.𝟒 𝒕 ∗ +𝟐.𝟏𝟗 𝝈 ∗ +𝟏𝟖𝟖 (𝒕 ∗ ) 𝟐 −𝟐.𝟗𝟒 𝒕 ∗ 𝝈 ∗ +𝟎.𝟗𝟐 𝝈 ∗ 𝟐
𝜎 ∗
𝑡 ∗
Fig 14. Oil Canning Prediction Factor
Fig 15. Proposed Oil Canning Model using Experimental Data

19. Model verification using experimental results
Pexperiment : Buckling load from experiment
Pprediciton : Buckling load from prediction
𝒇 𝒕 ∗ , 𝝈 ∗ =𝟏𝟗.𝟗−𝟕𝟑.𝟒 𝒕 ∗ +𝟐.𝟏𝟗 𝝈 ∗ +𝟏𝟖𝟖 (𝒕 ∗ ) 𝟐 −𝟐.𝟗𝟒 𝒕 ∗ 𝝈 ∗ +𝟎.𝟗𝟐 𝝈 ∗ 𝟐 t
case
point
curvature
width
length
𝝈 𝟏
𝝈 𝟐
Pexperiment
Pprediction
Error (%)

20. Summary
Anisotropic material properties was estimated from simple tensile test experiments.
Stamping experiment was conducted and then curvature radius was measured from 3D scanning procedure.
Further the load-displacement curve was obtained from the indentation test.
From the load-stroke curve, the minimum load to cause the oil canning was estimated.
Surrogate modeling method was utilized to develop the oil-canning prediction model for future predictions.
Later the model verifications were done using enormous experimental data.
Overall proposed oil-canning model showed quite good agreement with the experimental data.
We found that the panel thickness, the curvature radius and the residual stresses plays major role.

21. Challenging Tasks and Future Work
Further empirical model implementation is required.
The identification of main factors are quite sensitive.
So we are planning to do more experiments using the real specimens such as home appliances or automobile parts or aircraft components or roof panels.
As well as we decided to develop a oil canning prediction model for each applications separately.

22. Net Shape Manufacturing Laboratory
Thank you !
정형가공연구실