Efficient Stamping Simulations

By on February 7, 2019 in DESIGN/MODELING/SIMULATION

Focusing on production planning, process engineering, tool design, manufacturing.

Lightweighting and safety needs have rapidly expanded applications of press hardening and newer-generation cold-forming steels, as well as warm and hot formable aluminum grades to Body-in-White components in recent years. Product design, manufacturing engineering and production technologies have delivered tremendous innovations toward efficient production of parts with optimally developed and tailored properties. Digital engineering and validation tools have also kept pace with strong innovations of their own to actively support these recent developments.

Pre-eminent demands placed on digital engineering tools are to facilitate early and reliable decisions on critical aspects of part design and manufacturing process, and to enable detailed and accurate engineering of dies and processes guaranteed to produce acceptable parts.

These demands translate into concrete needs:
1. Enable design assessments starting very early during design; eliminate/ minimize design-triggered downstream production issues
2. Dovetail design process seamlessly into development and maturation of the manufacturing process
3. Account for all necessary thermomechanical sheet, die and process conditions in simulation
4. Provide reliable, objective and detailed feedback on all quality metrics predicted on physical parts
5. Provide detailed diagnostic and issue resolution tools
6. Ensure quick turnaround on simulation for objective decision-making based on all necessary and exhaustive what-if studies

The AutoForm product portfolio supports the automotive sheet-metal forming process chain from concept, styling and product design all the way to part production. The focus lies, however, in production planning, process engineering, tool design and manufacturing.

Computational speed and accuracy of digital validation tools are crucial to achieve the following:

• Efficient engineering workflow with maximum flexibility in the early phase and high accuracy in the final phase
• Proper representation of the physical process with precise tool definition
• Accurate results for feasibility and springback to optimize the forming process and process parameters
• Process capability and robustness to strive for zero-defect tool and part production

This article shows how to achieve the above benefits over the engineering workflow for a typical stamped panel. High-quality targets can consistently be achieved when following this step-by-step workflow.

1. Systematic development and validation of full-stamping process

Rome was not built in a day. Neither can a stamping process be finalized within a week. To define all process parameters, and design all tool surfaces, multiple studies are needed for making and validating crucial decisions on dies and process. Maturity of die faces and process grows over these iterative simulation studies to a final process digitally proven to produce acceptable parts in a repeatable fashion.

Figure 1 shows a generic timeline of milestones over product development and the engineering functions that are executed toward these milestones. Objective awareness of and persistent balance between cost, manufacturable quality and achievable timing is crucial to realizing intended part function.

Part design, manufacturability assessments and manufacturing planning need to go hand in hand from early on.

Part design, manufacturability assessments and manufacturing planning need to go hand in hand from early on. It is a crucial that simulation tools support product development through simultaneous and progressive validation—from early designs all the way to start of production.

Manufacturability assessments on early part designs need to reliably identify the most critical geometry issues and afford flexibility for very rapid what-if studies on alternatives still open at that early stage, including material grade and gauge. This phase may be called conceptual stampingprocess feasibility when only the process basics are analyzed.

We work with conceptual tool surfaces that can be modified easily. Drawbeads are represented using sophisticated, compute-efficient line-bead models. Tool surface design, simulation and forming evaluation are executed within an integrated simulation environment.

In the next phase, which may be called full-stamping process feasibility, full-process simulations of the intended process layout, or die operational lineup, are executed. All production stages, as well as crucial intermediate details such as part positioning and relaxation between stations, pad closing, progressive trimming, are evaluated. Drawbead action continues to be represented by line-bead models, but bead and binder set is executed with geometrical beads.

The main outcome of this phase is a confirmation of the viability of intended operational layout and/or validation of the same after needed changes. Very importantly, there should be an understanding of springback mode and magnitude. This is necessary to either confirm this can eventually be compensated or that early product or process countermeasures may be necessary to mitigate and manage the same. Also in this phase, systematic process-improvement methods are leveraged to arrive at a process that can produce acceptable parts. An assessment of robustness is used to validate its capability or repeatability.

The feasibility phases are wrapped up with these important outcomes:
• Process parameters and conceptual tool surfaces are established for a viable process
• Springback is confirmed to be compensatable
• The process is validated to be “capable” or sufficiently robust

The conceptual tool surfaces become the basis for the nominal computer-aided design (CAD) tool design, which can be built “first time right” in CAD without any costly modification loop. The task of creating nominal CAD surface based on the conceptual surfaces is called process design CAD–nominal.

By simply replacing the conceptual surfaces with the CAD surfaces in the last simulation of the feasibility phase, the nominal CAD tool design may be readily validated. This phase is called full-stamping process validation-nominal.

Following this validation, springback and its progression over different stations is carefully studied, and an appropriate compensation strategy— involving decisions on which stations/their tools need to be compensated, how and by how much—is proposed. The objective is to develop and validate a process for producing dimensionally compliant panels in a repeatable manner. The main output of this phase is the simulation-validated compensation vector fields for the compensated tools.

The compensation vector fields are imported into CAD, and the relevant tools may be rebuilt as native CAD surfaces. This phase is called process design CAD-milling. The resulting CAD design not only includes compensation, but also geometrical beads, fillet clearance, tool relief, scaling (draw station) and press compensation.

In the last phase, a confirmation simulation based on the CAD-milling data is performed. This phase is called full-stamping process validation for milling. Following simulation, CAD tool surfaces may require minor adjustments. This validation/re-validation then needs to be complemented by a full robustness assessment to confirm process repeatability.

And last, but not least, milling surface data released following all this extended development and validation needs to be used as is to machine the surfaces on the physical tools—without any undocumented and unvalidated changes. (See Figure 2.)

2. Phase-tailored computational settings for optimal accuracy and simulation throughput

As mentioned, process maturity gradually increases over the development of the stamping process. This maturity implies more detailed representation of die and process conditions in simulation. It is important to complement such detail with commensurate computational refinement for accurate outcomes.

On the flip side, high levels of computational refinement applied during early phases, lacking sufficient precision in the representation of materials, die and process, will lead to undeserved precision in simulation outcome. This precision comes at a high computational cost and should not be mistaken for accuracy.

During conceptual stampingprocess feasibility, the focus is on the drawing operation only. Splits and wrinkles are the targets that need to be addressed and eliminated. The bending enhanced membrane (BEM) element formulation is specially designed for this task and is part of concept evaluation (CE) simulation control settings. Using the BEM is the computationally efficient way to evaluate and eliminate splits and wrinkles.

Once splits and wrinkles are eliminated, the full-cycle stamping process feasibility phase starts. Draw-in is evaluated and optimized for stretch, skid lines and to reduce material consumption. The entire process is simulated, including secondary/line dies. Besides optimizing developed trim lines, and ensuring acceptable formability outcomes, springback is an important focus—its progression over the process, its magnitude and mode.

This review is important for countermeasure versus compensation decisions. During this phase, development of tool surfaces is accomplished inside the simulation environment. This affords a crucial opportunity for efficient and systematic improvement and finalization of these surfaces, based on simulation outcomes. This phase is carried out with concept evaluation plus (CE+) simulation control settings.

In the stamping-process validation phase, tool surfaces engineered in the simulation environment are replaced by copies of these surfaces rebuilt in CAD. The elastic plastic shell (EPS) element formulation, which is part of the highly refined final validation simulation control settings, is applied. In this phase, the full range of results are re-evaluated with the benefit of the high accuracy, and precision, accompanying the final validation (FV) settings—splits, wrinkles, draw-in, skid lines, springback and surface defects. Compensation—morphing, scaling, drawshell—is applied, optimized and validated.

Acceptable and finalized nominal is evaluated for robustness/repeatability. Process details—tonnages, gaps, binder and pad bearing, etc.—as well as tool surfaces are finally released for machining on the basis of this finedetailed and extended validation. (See Figure 3.)

3. An application showcase

Application of systematic process improvement, over the stamping optimization phase, was outlined start to finish in an earlier article on the press hardening of a B-pillar (“Digital Validation & Beyond,” Adithya Ramamurthy, AutoForm Engineering USA, Inc.; Lightweighting World, Sept&Oct 2018).

Automotive closure panel designs have increased in complexity to accommodate geometry elements and assembled components that specifically facilitate lightweighting and safety goals. In acknowledgment of this fact, the present showcase of simulation efficiency is based on a complex door inner panel.

As process maturity grows, the simulation represents real or intended die, process and material conditions very closely. With the use of appropriately refined simulation controls, simulation outcomes provide a very accurate representation of physical panel characteristics. Refined simulation controls, also called final validation, however, increase computational time. In this study, evolution of simulation time and predicted thickness has been evaluated over the progression of engineering phases. While gradually tightening the simulation controls, the final required accuracy can be achieved efficiently.

The simulation time for each engineering phase is shown in Figure 4. Performing a simulation of the draw operation in the conceptual stamping process feasibility phase takes 11 minutes, whereas the stamping-process validation simulated with geometrical bead takes about three-and-a-half hours. The other engineering phases are between these two extremes.

The graph also shows how many simulations can be performed over different engineering phases compared to one simulation in the stamping-process validation phase when geometrical beads are used. In the conceptual stamping-process feasibility phase, 18 simulations of the drawing operation can be performed. From an engineering and process optimization point of view, this means 18 modification and optimization iterations can be made in the same time that it takes for one stamping-process validation simulation with geometrical beads. In more general terms . . .

• During early phases, more simulation iterations are necessary to finalize details of die and process conditions. Simulation controls applied during these phases—CE and CE+— provide very short turnaround to enable this.

• As the process matures, fewer simulation iterations are necessary. Accuracy and reliability of simulation outcomes are crucial. This is enabled through refined simulation controls, FV.

During early phases, more simulation iterations are necessary to finalize details of die and process conditions.

It is also important that as the maturity of simulations progresses, simulation outcomes “converge” to reliable and accurate predictions. Figure 5 provides an overview of thinning distribution outcomes from the different simulation phases. It is very clear that the distribution or pattern of thinning is very similar between the different phases, although magnitudes increase in the later phases.

Part, tool surface and process conditions evolve and mature over different phases of stamping-process engineering.

Thinning values at six critical points are tracked over the different simulation phases in Figure 6 and can be seen to converge well over this progression. The spread of thinning values is narrow—less than three percentage points—over the different simulation phases, except at points 2 and 6. Point 2 is a location with relatively large changes in deformation modes—bending, unbending, plane strain deformation and biaxial tension—over the forming process. Under these conditions, the EPS (used during the final validation phase) performs better than the BEM (used during the conceptual evaluation phase).

Difference in thinning observed at point 6 is due to the difference in tool shape in the vicinity of this location— finalized CAD tool surfaces carry a more crisply defined step on the addendum than in the very early version of the tool used during concept evaluation. This element of tool geometry, evidently, was refined to final shape over the later simulation phases.

4. Conclusions

Part, tool surface and process conditions evolve and mature over different phases of stamping-process engineering. Associated changes are significant early on, driven not the least by changes to part design. Phase-calibrated digital standards and simulation controls are essential to the overall efficiency of the stamping-engineering process, as well as to the quality and cost of engineering outcomes.

Authored by Kidambi Kannan