1

A cost-effective Cold Roll-Forming

FE model for industrial applications
Authors: Leonardo Marconi, Timothy Senart, Claudia Cofano

[email protected]
[email protected]
[email protected] For a better future
 Index
 Introduction
- CRM Group in a nutshell
- Cold Roll-forming process
- Research overview and objectives
 Experimental campaign
- Experiment set-up
- Outputs
 Numerical model
- Methodology
- Sensitivity and calibration
- Results
 Conclusions
 Future works

2
Introduction

 CRM Group in a nutshell

 Cold Roll-forming process

 Research overview and objectives

3
 CRM Group in a nutshell
Our Identity:
Founded in 1948 from the merger of independent & private research centres,
we focus on processes, products & solutions for the metallurgical industry
Our Mission:
Bringing innovation all over the world
Create value through innovation, transformation and generation of new markets, with more than 265 engineers
to empower metals in societal, environmental and economic challenges

o Digitalization

Expertise areas: Our industrial members

o Circular Economy
o Energy Transition
o Construction
o Digitalization
o Advanced Manufacturing
Testing, modeling and optimization of steel structures and manufacturing processes 4
 Cold Roll-forming process
Cold Roll Forming (CRF) is one of the most productive processes for manufacturing thin-walled products
Continuous bending process of a long metal sheet.
with constant cross section. It consists in a continuous bending operation of a long metal sheet. The sheet is
The gradually
sheet is formed
formedpassing
throughthrough
pairs of pairs of rotating
rotating rollsstations)
rolls (called (stations).
until the desired cross-sectional
Each stationismakes
configuration an incremental
obtained partCRF
(see Fig. 1). The of the bend.is widely used in aerospace, construction, automotive
process
and other industries with large production volumes [1]. Roll-forming processes gained high interest in the
industry“Forming
to form flower”
Ultra-High Strength Steels (UHSS). These steel grades permit to pursue advanced structural
Strength:
optimization and to achieve lighter design solutions. The extensive application of roll forming can be due to
its increased deformation capability and the limitations of UHSSowhen formed
High through
production traditional
volume [1] stamping
processes [2].
However, CRF remains a complex process and it is affected by Capability
odifferent to form
problems, UHSS
such as gradestorsion,
wave, [2]
Shape evolution at each station
twist or bow defects and elastic spring back [3]. Finite Element Analysis (FEA) is employed at the industrial
Problematics [3]:
level as an efficient and economical tool to examine the process. The aim of the analysis is to predict these
defects and optimize the roll-forming process.
o Geometrical imperfections
The objective of this research is to develop a Finite Element (FE) model using the commercial software LS-
DYNA to simulate the cold roll-forming process of a U-shapedoprofile Unbalanced
and thenspringback
to validate it through a
correlation with the experiment. In a future step, the outcome from this simulation will be used to better
understand the effects of the manufacturing process on the performance of the final product. The use of a
commercial FE software makes simple the process of integrating strain and stress fields due to
manufacturing into existent FE models.
Illustration of a Cold Roll-Forming line, from Urheber GmbH

5
Research goal and methodology
Aim:
Developing a cost-effective methodology to investigate cold roll-forming processes in LS-DYNA

Research phases :

To validate:
Test campaign o Process design parameter
o Model assumptions

Goals:
Numerical o Accuracy
model creation
o Pre-processing ease
o Computational efficiency
Cold roll forming FE model on LS Dyna
6
Experimental campaign

 Experiment set-up

 Outputs

 Observations

7
Experiment set-up
Pilot cold roll-forming line:
o Station: 9 (equally spaced 250mm)
o Blank speed: 4.2 m/min

Forming Flower

Final U-shaped profile

Symmetry

Initial flat sheet Cold roll-forming line at CRM

Group
Specimen:
o Dimensions: 0.63mm thick, 2000mm long, 360mm wide
o Material: S350GD grade (characterized by tensile tests EN ISO 6892)
Tools:
o 3D scan ( Cross section)
o Strain gauges ( Longitudinal strain)

Material characterization 8
Experimental output – Longitudinal strain
Central gauges for correlation
Limit of the elastic domain

Gauges disposition on the specimen

Design rule
satisfied

Longitudinal gauges for verification

Central gauges
verified .
Central gauges
𝜀𝜀50 ≠ 𝜀𝜀500 = 𝜀𝜀1000
for correlation

Longitudinal gauges
for verification
9
Experimental outputs – Cross-section

3D scanned final profile 3D scanned final profile

Cros s -section Springback

extrapolation

Deviation

Final cros s-section of the central section

10
Campaign conclusion
Experimental campaign outcomes: Observations:
Low eccentricity and displacement
1) Results for correlation
o Longitudinal strain
o Final cross section

Rolls displacement < 0.25mm

2) Observations (of which 0.12mm of eccentricity)

o Low roll eccentricity and displacement

No feeding misallignment
o Blank-line misalignment controlled

3) Open question

o Cross section deviation

11
Numerical model

 FE Methodology

 Guiding script

 Sensitivity and Calibration

 Results

12
Non-linear Finite Element Model
o Method  Non-linear static implicit FEM
Inputs:
o Material  Isotropic elasto-plastic law (Type 24)
1) Sheet properties
o Section  Fully integrated shell elements (Elform 16)
2) Forming line geometry
o Contact  FORMING_SURFACE_TO_SURFACE_MORTAR
3) Roll forming flower
o MPP domain  dec { region { parts 10 sx 1.e9 } sx 1.e9 }

13
Guiding method
Guiding [4-5] applied to node “N” between stations “ i ” and “ i+1 ”:
𝜋𝜋 𝑍𝑍 𝑛𝑛𝐸𝐸𝐸𝐸𝐸𝐸
o 𝑋𝑋𝑖𝑖(𝑍𝑍,𝑁𝑁) = 𝑋𝑋𝑖𝑖 (𝑁𝑁) + 𝑋𝑋𝑖𝑖+1 (𝑁𝑁) − 𝑋𝑋𝑖𝑖 𝑁𝑁
sin
2 ∆𝐿𝐿𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
Script to automatically create
𝑛𝑛𝐸𝐸𝐸𝐸𝐸𝐸
o 𝑌𝑌𝑖𝑖(𝑍𝑍,𝑁𝑁) = 𝑌𝑌𝑖𝑖 (𝑁𝑁) + 𝑌𝑌𝑖𝑖+1 (𝑁𝑁) − 𝑌𝑌𝑖𝑖 𝑁𝑁
sin
𝜋𝜋 𝑍𝑍
equations and boundaries
2 ∆𝐿𝐿𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

Local “Z” reference system

𝑿𝑿𝒊𝒊 (𝑵𝑵) , 𝒀𝒀𝒊𝒊 (𝑵𝑵) for each “N” and “ i ” are input through the “meshed forming flower”:

Discretized forming flower

..

Forming flower – Las t node Y interpolation path (Las t node for 𝒏𝒏𝑬𝑬𝑬𝑬𝑬𝑬 =2)
14
 Sensitivity analysis
High sensitivity on guiding shape
SHEET LENGTH: 100 mm

Strain varying over time

Longitudinal strain fringe (L=100mm, Analys is time: 15h, 18 CPU) Guiding shape
sens itivity

Sheet length converged

SHEET LENGTH: 500 mm

Steady state reached

Insensitive to guiding shape
Longitudinal strain fringe (L=500mm, Analys is time: 24h, 18 CPU) Sheet length sens itivity 15
(n=2)
Correlation status
1) Longitudinal strain
1 st 2 nd 3 rd 4 th 5 th 6 th 7 th 8 th 9 th o Accurate correlation
o Deviation after 7th station

2) Final cross-section
o Accurate prediction of spring back
Forming flower (assumed trajectory)
o Geometrical deviation not simulated
not followed at 8th and 9th station

Springback Forming flower

at 9th station

Deviation

16
 Calibration Phase
Sheet schematics Y trajectory of guided and central node

Central node coordinates

to guide the edges

New coordinate
used to guide

Forming flower evolution for 2 iterations Input modification

Iterative Procedure

8th station 9th station Original Modified

17
Results
1) Longitudinal strain
o Accurate correlation:
Strain field on the flange
1 st 2 nd 3 rd 4 th 5 th 6 th 7 th 8 th 9 th
Correlation at 2 points - Prediction of strain gradient
Flower flaw
- Better trend after 7th station
o Robust results
o Uncertainty range:
- Material properties
- Gauge misalignment (±2.5°)

2) Final cross-section

o Prediction of spring back

Cause of the
...
shape deviation o Simulation of the deviation
o Correction of forming flower

18
Conclusions

 Conclusions

 Future works

19
Conclusions
 Project aim:
To create a numerical methodology able to give fast and accurate data about the cold roll-forming process.

Forming process design  Numerical model creation Process check

Pros Cons

1) Results accuracy & robustness 1) Roll forming line stiffness

o Major outputs validated o No effects of roll-forming line flexibility
o Possibility to check the forming flower o Future work
o Small uncertainty range

2) Cost-Effectiveness 2) Different physics of the rolls

o Analysis time < 24h (18 CPU, Intel® Xeon® W-2195 2.30GHz) o No info about the rolls torque required
o Small portion of the blank modelled o No sensitivity about friction coefficient
o Highly automatized pre-processing

3) Easy to interface with other models

o LS-DYNA makes it easy VS. dedicated software
o Future work & final aim of this investigation
20
Future works
1) Validation of new profiles Semi-rigid line on LS-DYNA
2) “Semi-rigid” roll-forming line
To account for roll-forming line flexibility
Beams
Stiffness of the line:
- Axis modelled through beam elements
- Rolls’ stiffness modelled through springs
Discrete springs

3) Study effects of manufacturing over product performances

Mapping stress/strain state from manufacturing.
Convenient on LS-DYNA
Extraction of outputs at
mid-sheet and mapping

21
Thank you
www.crmgroup.be

[email protected]
[email protected]
[email protected] For a better future
 Literature

Liu, Huamin, Zhiqing Liu, and Zhongping Zou. "FEM Simulation of Cold Roll Forming in the Car's Anti-Collision Beam."
[1] Journal of Computational and Theoretical Nanoscience 9.9 (2012): 1472-1476.

Tsang, Kwun Sing, et al. "Industrial validation of strain in cold roll forming of UHSS." Procedia Manufacturing 15 (2018):
[2] 788-795.

Watari, Hisaki, and Hiroshi Ona. "Characteristic features of shape defects occurring in the cold roll forming of pre-
[3] notched products." Journal of Materials Processing Technology 80 (1998): 225-231.

Kiuchi M, Koudabashi T. “Automated design system of optimal roll profiles for cold roll forming.” Proceedings of the
[4] third international conference on rotary metalworking processes, Kyoto, Japan (1984): 423–36.

[5] G.T Halmos: “Roll forming handbook”, CRC Press, 2005

23
24

A cost-effective Cold Roll-Forming

FE model for industrial applications
Authors: Leonardo Marconi, Timothy Senart, Claudia Cofano

[email protected]
[email protected]
[email protected] For a better future