Simulate Reality - Deliver Certainty Through the Virtual Weld

Article information

* Director R&D, Simufact Engineering GmbH, Hamburg 21079, Germany
** CTO & Managing Director, Simufact Engineering GmbH, Hamburg 21079, Germany
*** Technical Manager, MSC Software, Seongnam, 13591, Korea
Corresponding author : inhyuck.hwang@mscsoftware.com
Received 2016 May 2; Accepted 2016 June 16.

Abstract

Welding is an absolutely essential component of industries such as the automotive industry, the construction industry and even the aviation industry. Although it is a widespread technology it is still characterized by lots of uncertainties. This still requires well experienced and highly skilled workforce to design and perform safe welding processes. The early knowledge of distortion and residual stresses is almost an issue which is influenced mainly by the welding parameters and the fixture design. But more and more engineers want to know as well final properties of the assembled components. With the beginning of the computer age in the 1970s and 1980s last century, the numerical prediction of manufacturing processes using FEM was gradually getting better and has established itself in the industry since the 1990s as a standard tool. Unlike in metal casting and forming industry, however, the everyday use of FEM-simulation tools for welding processes eked out a shadowy existence for a long time. This paper will give a short classification of welding simulation types and a structured overview on the technical questions. Selected case studies and the benefits achieved through simulations with the software Simufact.welding are discussed. Finally an outlook on future developments will be given.

1. Introduction -About welding

Commercial application of welding processes in industrial production requires a high degree of planning reliability. The proper development of a welding schedule, which is necessary for the definition of welding sequences, intervals, and programming of welding robots, is important for the quality of a welding process. Additionally, one should be able to choose the right material, the applied welding process, as well as the correct application of clamping tools.

Welding is defined as a “permanent connection of components through application of heat and/or pressure”. The components are connected either by melting or by heating, and by applying additional forces (pressure). There is no other joining process which allows such resilient and dense connection with minimal space requirements. Furthermore, welding is the method of choice when it comes to the joining of assemblies with high levels of complexity. There are over 100 different welding processes depending on the specifics of heat input and pressure application.

Although joining is the main application field of welding processes, it is also used in deposition processes in order to create durable and hard surfaces by means of cladding.

With Additive Manufacturing a new kind welding is established to create components from powder material.

2. Current trends

2.1 Why the virtual weld?

There have been two main demands from industry to researchers during the past few years. First, in order to move further towards energy efficiency in the automotive industry, the importance of lightweight construction is increasing. The development of new high-strength materials is a serious challenge for joining technology. Second, decreasing energy consumption during the production itself is desirable. The amount of filler material is reduced and lap joints are replaced with butt joints in order to reduce the overall weight of the product. Welding processes with low energy input, numerical welding simulation, and application of virtual welding trainers contribute to the overall goal to make joining technology more energy- and material-efficient.

Due to a lack of experienced welding specialists, available professionals are forced to complete more demanding work in less time. Simultaneously, “easy” tasks are given to less experienced workers. This situation leads to an increased amount of mistakes due to insufficient abilities of staff members and overloading of experienced specialists. Labor costs, as well as overall costs for development and trial tests, are expected to increase.

Simufact.welding simulation software offers the possibility to capture the institutional knowledge of welding processes. They allow virtual try-outs that help to investigate process parameters and their influence on the results of an applied welding process, as well as support in finding and documenting convenient process parameters.

The focus of this modern simulation package is not only an appropriate calculation of the relevant effects but also to give engineers fast modelling tools and fast access to results, which is of vital interest for industrial users and a big challenge with respect to the complexity of welding applications.

2.2 Disciplines of welding simulation

According to Radaj1) welding simulations have to be subdivided into three several disciplines:

  • 1) Structural simulation

  • 2) Material simulation

  • 3) Process simulation

Process simulation deals with the complex physics of material melting & solidification, formation of weld pool geometry and related effects. These kind of information are needed as an input for structural and material simulation. Since it is very complex and time-consuming it has been used nowadays more for academic purposes with specialized simulation software. Alternatively to the expandable simulation of above described effects one can use simplified analytical formulas and statistics in order to describe an “equivalent heat source”.

Typically the main goal of structural calculations made by Simufact.welding is a prediction of welding distortions. Due to the implementation of material models, it is also able to calculate metallurgical effects like material phase fractions, material conditions, and resulting local material properties, as well as further effects like transformation induced plasticity and transformation strains. This will increase the predictability and accuracy of the simulation and becomes more and more essential for multi-stage processes

With its new version Simufact.welding 5.0 demonstrates continuous migration of all three disciplines into one platform and one tool - the first application unifies all three disciplines into one complete model approach is Resistance Spot Welding - please refer to chapter 3.2 too.

2.3 Typical pains

Welding distortions often are not completely avoidable. Nevertheless, they represent a challenge only if requirements of the welded structure are not fulfilled. Following picture denominates selected fields of application in which Simufact.welding is used to control and minimize welding distortions and other known pain points. Other applications like energy sector, healthcare or heavy vehicles can utilize the software capabilities as well.

It can be stated that among all other potential problems the early knowledge of distortion contributes most to cost savings which can be achieved with virtual preoptimizations.

Like shown in Fig. 3 distortion dependents on the degree of fixity significantly. Although the physical basics are well known and can be easily understood there is nevertheless a great uncertainty because every single process is unique with respect to its geometries, boundary conditions, fixture designs and their implications to the resulting distortion.

Fig. 3

General interaction between deformation, stresses and degree of fixity according to2)

Fig. 1

Different areas of welding simulation (according to1)) and their coupling

Fig. 2

Typical pain points of welding in selected industry branches

Only simulation is able to analyze these complex interactions that lead to unwanted distortions.

3. Case studies

3.1 Laser-welding of a car door(Courtesy: AUDI AG Ingolstadt)..

Compared to conventional laser-welding processes the so-called remote-laser-welding enables shorter cycle times and higher flexibility. The engineer has more options for the weld sequence and weld position planning. But along with shorter cycle-times and changes in the weld sequence the process will respond differently with respect to distortions3).

Above pictures illustrate the complexity of such a welding process and underline the need of virtual process optimization with the help of a dedicated simulation tool. Simufact.welding was used first to understand the existing process design and to compare measured distortions against predicted ones.

After passing the initial calibration step and analyzing the main influences on the distortions occurred, several set-up modifications and different weld sequences should be tested to find a set-up with minimized distortions.

With a sub-model, using real component shapes, the heat source parameters and weld-pool dimensions were calibrated – refer to Figs. 6. This step is essential to put in the right energy density and energy distribution for the subsequent structural analysis. For this kind of analysis Simufact.welding offers a special functionality called “thermal calibration”. This initial simulation uses real shapes and weld sequences but it is not computing any mechanical response to the thermal input. Since this pure thermal simulation is very fast one can secure the most important input parameters for the fully coupled mechanical-thermal solution within shortest time. Customer has reported less than 4 hours effort to get sufficient input data4).

Fig. 6

Fast thermal calibration of weld-pool using submodels4)

Fig. 4

Remote laser welding cell at AUDI AG4)

Fig. 5

Fixture design for remote laser welding of a AUDI car door

Calibration step is followed by the fully coupled transient thermal-mechanical simulation which provides appropriate distortions for the given fixture and weld sequence design. Fig. 7 illustrates how close simulation predicts reality with well calibrated input data.

Fig. 7

Measured vs. calculated distortion after remote laser welding of a car door4)

Once the engineer could validate the simulation model versus real measures he can start to find correlations between initial boundary conditions and distortion evolution. This enables to create dedicated process and design modifications and find desired optimum.

Time to solution was reported to be less than 2 weeks which is sufficiently quick to achieve remarkable time and cost savings within the development cycle4).

3.2 Resistance Spot Welding (RSW) of a car wheelhouse

Resistance Spot Welding (RSW) is the latest development in the powerful Simufact.welding software. It enables a fully automatic execution of the RSW-process, typically consists of 6 sub-stages.

The existing product and process design was suffering from extra-clamping and unwanted high distortions. Additional costs were spent due to a higher number of machined parts and spare cylinders for the fixtures, more sensors than planned and a reduction of throughput caused by increased cycle times.

With the new, revolutionary modelling approach of Simufact.welding 5.0 it took less than 2 hours to create the complete model, 1 hour for calibration, 4 hours for execution and another 1 hour for analysis and review.

The core of the advanced automated process and modelling set-up is besides of a data-base with typical weld-gun geometries a fully automated kinematic description of the process. The engineer has just to define the location of the spot weld, the right direction (upper and lower torch), torch type (X-, C-spot type) and the right order of the multiple spot welds as shown in Fig. 9. After assigning the appropriate properties of the weld guns (current vs. time characteristic) one can execute the simulation. Approaching of torches, clamping, welding and release are done fully automatically.

Fig. 9

Software GUI - model setup - definition of spot weld location and orientation

Fig. 8

Fixture design of a car wheelhouse, prepared for multiple resistance spot weld

After execution typical results like peak temperatures and distortions (as shown in Fig. 10) but also phase fraction, residual stresses and nugget geometries can be analyzed.

Fig. 10

Typical result values of a RSP simulation

Doing so will lead within 2 … 3 days to a new optimized set-up with the following main improvements:

  • 1) Reduction of distortions by 70%

  • 2) Reduction of fixture weight by 25%

  • 3) Improved production throughput (not specified)

With those savings reported from the customer one can imagine how fast a return on invest can be achieved.

3.3 Downstream study - manufacturing of a crash box (real case demonstrator)

Passenger safety is one of the most critical criteria in process-, part- and design approvals. Every safety-critical component has to pass several enhanced testing cycles including virtual crash tests.

Without going too muchin detail regarding the welding analysis (laser welding of two roll-formed half-pipe bodies) this example shall illustrate the importance of deploying virtual process chains.

It’s a fact that most crash simulations do not consider the manufacturing history of the parts and components within the assembly. They are using averaged and locally homogenized material properties as inputs for their calculations.

Similar statements can be given for welding simulations. Usually modelling is using CAD geometries assuming material properties from a database.

In present case study from a German OEM the complete manufacturing sequence of a crash box was investigated. Starting with sheet rolling, followed by a roll-forming operation the cold metal forming stages were simulated by means of Simufact.forming, the leading software approach in metal forming.

Work hardened half-pipe bodies with residual strains and stresses from cold roll-forming (Fig. 13) were taken and a 2-stage welding simulation was performed. Stage 1 contains welding of the two half-pipes, stage 2 represents the welding of the previously welded half-pipes against a ground-plate.

Fig. 13

Typical roll-formed shape of a half-pipe for crash boxes - here: equivalent strain distribution

Fig. 11

Achieved reduction of total distortion through welding simulation

Fig. 12

Car frame with crash boxes

From stage to stage the local distribution of the mechanical properties from the previous stage was used as an input for the subsequent stage.

Finally a crash simulation was performed. For this purpose an impact energy was applied to the manufactured crash box and the energy consumption, represented through the impact height was calculated. This final stage was carried out with two approaches: on one hand the “classical” way of using just CAD data with averaged material properties was performed. On the other hand the entire manufacturing history of each components was considered. Fig. 15 illustrates in an excellent manner the importance of it: remarkable differences occur between the two cases.

Fig. 15

Comparison of calculated energy consumption and impact height

Fig. 14 Welding stages: stage 1 (top) and stage 2 (bottom)

Fig. 14

Welding stages: stage 1 (top) and stage 2 (bottom)

The crash box which owns its forming and welding history consumes the impact energy faster (blue curve) and provides a larger impact height compared to the crash box using a CAD model with initial material properties from a data base (red curve). The results of this investigation were validated against test models.

It can be demonstrated that the consistent consideration of manufacturing chains in the virtual development will lead to more accurate results and will contribute to more safety in component and process design.

4. Conclusion

For welding simulation, Simufact Engineering offers a comprehensive simulation software suite: Simufact. welding.

Simufact.welding offers the possibility to calculate welding stresses, distortions, and the evolution of material properties from a graphical user interface. This means:

  • 1) To investigate possible problems up front

  • 1) To investigate possible problems up front2) More insight into welding processes

  • 3) More methodical optimization of processes

  • 4) Verification of quality of welding seams

  • 5) Investigate process chains that appear during manufacturing

Selected case studies with different welding applications demonstrate the economic benefits of faster welding process design as they are:

  • High efficiency of the development process due to a reduced number of expensive failed hardware trials

  • Reduced expenses of manufacturing of prototypes

  • Reduction of machining and straightening costs

  • Reduction of development times which leads to shorter time-to-market

  • Decrease of material and energy consumption for experimental investigations

  • Reduction of manpower needed for experiments

  • When bidding on a project, efficient feasibility studies lead to winning offers

For more information please refer to www.simufact.com.

Acknowledgement

The authors would like to thank first of all our valuable customers, especially the AUDI AG in Germany for providing us real case studies and their valuable feedback. Furthermore the authors would like to acknowledge the contributions of Dr. Ingo Neubauer, Julian Litzkow (Simufact Engineering) and Fernando Okigami (TekniCAE, Brazil).

References

1. Radaj D. Eigenspannungen und Verzug beim Schweißen –Rechenund Messverfahren. DVS-Verlag Düsseldorf 2002;(in Germany).
2. Porzner H. Piossibilities of numerical Simulation for Evaluation and Optimisation of Welded Designs, Mathematical modelling of weld phenomena 5, Proc. of the 5th International Seminar, Numerical Analysis of Weldability. In : Cerjak H, ed. London: IOM Communications. 7382001. p. 701–724.
3. Thater R, Wiethop P, Rethmeier M. Welding Simulation in Car Body Construction. Laser Technik Journal, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 12(2)2015;:33–37. (in Germany).
4. Philipp Wiethop (Audi AG), Raphael Thater (Fraunhofer IPK), Schweißverzugssimulation im Karosseriebau. Proceedings 16. Round Table Simulating Manufacturing, Marburg, Simufact Engineering GmbH 2015;(in Germany).

Article information Continued

Fig. 1

Different areas of welding simulation (according to1)) and their coupling

Fig. 2

Typical pain points of welding in selected industry branches

Fig. 3

General interaction between deformation, stresses and degree of fixity according to2)

Fig. 4

Remote laser welding cell at AUDI AG4)

Fig. 5

Fixture design for remote laser welding of a AUDI car door

Fig. 6

Fast thermal calibration of weld-pool using submodels4)

Fig. 7

Measured vs. calculated distortion after remote laser welding of a car door4)

Fig. 8

Fixture design of a car wheelhouse, prepared for multiple resistance spot weld

Fig. 9

Software GUI - model setup - definition of spot weld location and orientation

Fig. 10

Typical result values of a RSP simulation

Fig. 11

Achieved reduction of total distortion through welding simulation

Fig. 12

Car frame with crash boxes

Fig. 13

Typical roll-formed shape of a half-pipe for crash boxes - here: equivalent strain distribution

Fig. 14

Welding stages: stage 1 (top) and stage 2 (bottom)

Fig. 15

Comparison of calculated energy consumption and impact height