Application of Computational Simulation tools in Welding Engineering Education

1
Application of Computational Simulation tools in
Welding Engineering Education

• Sudarsanam Suresh Babu, Ohio State University
• Babu.13_at_osu.edu
• Presented by S. P. Khurana, Edison Welding
  Institute

2
Acknowledgement

• Support from Dr. S. Gordon of Ohio Supercomputing
  Center, Columbus, Ohio for developing a course
  work with EWeldPredictor
• Mr. C. Conrardy and S. P. Khurana for providing
  access to E-WeldPredictor tool from Edison
  Welding Institute
• Department of Industrial and Welding Engineering
  students for participating in this exercise

3
Outline

• Motivation
• Role of Computational Tools
• Scope of these tools
• Challenges to deploy during teaching
• Current Approach Demonstrations
• Material Models
• Integrated Models
• Feed back from Students
• Summary Future Directions

4
Motivation How can we teach the complex
interaction between thermal, mechanical and
metallurgical processes during welding?
ArcLaser
Distortion in ship panel welding
Hardness gradients in longitudinal section of a
submerged arc pipe weld
Pictures courtesy of ORNL and EWI
Laser assisted arc welding

• Let us look at how we teach these subjects

5
Current curriculum focuses on process, material,
design and fitness for service in different
courses at different levels.

• Basic skills
• welding
• electronics
• Process technology
• arc welding
• lasers
• resistance welding
• brazing and soldering
• Robotics
• Sensors and control algorithms

• Materials science
• steels
• nonferrous alloys
• polymeric materials
• Design of structures
• strength of materials
• static and dynamics
• Inspection and quality assurance
• Problem solving

• But holistic view showing the interaction between
  these disciplines is difficult to articulate!

6
Solution We need an integrated welding process
model

• Each instructor can focus on a particular subject
  matter expertise.
• While students/users can explore the impact of
  these subjects on the overall weld!
• Vision by Prof. Kirkaldy and by other researchers
  is shown in the diagram
• Challenge is how to do it!

7
For integrated model we need the following
expertise

• Needed Expertise
• Weld Design (mechanical engineering)
• Weld pool (fluid dynamics)
• Thermal Cycles (heat transfer)
• Mechanical Loads (structural modeling)
• Microstructure (metallurgy expertise)
• Finite Element (mechanical engineering)
• Supercomputing (computer science)
• This is currently not possible to impart in the
  undergraduate level!
• So what do we do?

8
Outline

• Motivation
• Role of Computational Tools
• Scope of these tools
• Challenges to deploy during teaching
• Current Approach Demonstrations
• Material Models
• Integrated Models
• Feed back from Students
• Summary Future Directions

9
Simple material models are already deployed over
the Internet

• Low Alloy Steels
• Evaluating the minimum cooling rate to avoid
  martensite formation during weld cooling
• http//calculations.ewi.org/vjp/secure/TTTCCTPlots
  .asp
• Evaluate the fraction of ferrite, bainite and
  martensite in welds
• http//calculations.ewi.org/vjp/secure/AshbyModel.
  asp
• Stainless steels
• Calculate the ferrite number using WRC1992
• http//calculations.ewi.org/vjp/secure/FNPLots.asp
  
• Calculate the ferrite number as a function of
  cooling rate
• http//calculations.ewi.org/vjp/secure/FNCoolingRa
  te.asp

10
Material Model Demonstrations

• How about integrated model?

11
Integration of welding, finite element analyses,
supercomputer, and internet domain was achieved

• Collaboration between Ohio Supercomputing Center,
  Ohio State University and Edison Welding
  Institute
• Demonstration using internet will be done after
  this slide

12
Demonstration
Even over iphone!

• We need just a web browser and internet
  connection and simple computer (or pda) What can
  you do with this tool? Let us take a spin!

13
Weld Geometry Selection

• What are the process parameters?

14
Welding Procedure Selection

• How do we fill up the joint cavity?

15
Bead placement during multipass welds

• This can be changed by the user!

16
Peak temperature and hardness prediction
Peak Temperature
Hardness

• Use of 2.25Cr-1Mo steel leads to hard zone in the
  HAZ!

17
Residual Stress and Distortion Prediction
Red 413 MPa Blue 0 MPa

• Traverse Shrinkage Left 0.27 Right 0.08
• Angular Distortion Left 0.61 Right 0.63

18
Significance

• Students can understand the interactions between
  thermal, mechanical and metallurgical processes
  as a function of
• Base metal and weld metal composition
• Process parameters
• Joint geometry
• Designed for a welding engineering student to do
  what if evaluations quickly!
• Each instructor can focus one of the goal
  process or materials or properties or distortion
  true interdisciplinary teaching achieved!

19
Outline

• Motivation
• Role of Computational Tools
• Scope of these tools
• Challenges to deploy during teaching
• Current Approach Demonstrations
• Material Models
• Integrated Models
• Feed back from Students
• Summary Future Directions

20
Students without industrial experience

• Agreed about the simplicity of tool and used
  without running into any difficulties.
• But there was a problem
• Students used just blindly used the default input
  parameters and provided the results without much
  discussion of the output being generated by the
  E-WeldPredictor.

21
Students with industrial experience

• Agreed about the simplicity of tool and used
  without running into any difficulties.
• used the tool as what-if simulation tool and
  explored the input space beyond the default
  parameters.
• Goal is met!

22
Future directions Teaching Methodology

• For students without industrial experience
• Use the tool in conjunction with experimental
  work CAPSTONE Projects
• Introduce the tool with a class room exercise of
  solving a problem from beginning to the end

23
Future directions Expansion of Application Scope

• Expand the tools for other common joining
  processes
• Resistance spot welding
• Brazing
• Soldering
• Expand the tools for other emerging welding
  processes
• Friction stir welding
• Ultrasonic Welding
• Expand the tools for property prediction

24
Summary

• There is a need to use computational tools to
  elucidate the interaction between physical
  processes
• A flexible EWeldPredictor computational tool that
  integrates welding engineering, finite element
  analyses and supercomputing technology was
  introduced in welding engineering education
• Potential of this tool in teaching was
  demonstrated and the feed back from students are
  presented
• Future directions to expand the application scope
  and merging with experimental laboratory work are
  stressed