Improving Automotive Closures through Simulation

Closure efforts are often one of our first impressions of a vehicle, but that is just the start of the complex and seemingly conflicting requirements for a vehicle system that has steadily increased in content while continuing to drop weight. A range of technologies are frequently employed to address the challenge, including linear, non-linear implicit and explicit FEA, multi-body dynamics (MBD) and mechatronics, computational fluid dynamics and optimization.

auto_closures_051111

Global platforms now dominate the underpinnings of new nameplates, the Chrysler/Fiat C-Evo, Ford C1 MCA and GM Gamma to name a few, and each is spawning multiple variants, sometimes specific to niche or local markets. While it is possible to carry over the underbody and many other parts from, for example, a 5-door hatchback to 2 door coupe, the closures are always likely to be unique to each model. Time efficiencies achieved in closure development are significant to launch timing especially when program goals may drive the application of less traditional materials such as lightweight metals or composites. A multi-disciplined approach is required to satisfy design, safety, NVH, durability, manufacturing and weight targets.  The creation and use of custom automations can simplify the application of CAE and make it a seamless part of the design process.

Side Doors

An industry leader, Toyota Auto Body, evaluates and improves the behavior of doors by using CAE ahead of physical product testing, specifically Altair MotionView and MotionSolve. These tools allow the modeling of even the non-linear displacement-load characteristics, present in the lock, stops and seals, essential to simulate the loads that will be observed during slam load cases. The simulation results show good strong correlation with test data and have allowed an MBD driven process for developing door slam characteristics to be established.

The experience at Toyota is typical of an industry trend that is seeing OEMs and suppliers moving from a finite element based approach to a modern MBD method. FEA methods generally require longer run times and a greater modeling effort than MBD closure motion simulations. An MBD model can be built with less design data, allowing low-fidelity studies to be performed quickly and early in the design cycle to guide concept direction. As design data becomes fixed higher-fidelity MBD models can be developed as part of an ongoing system optimization.

In general side doors are becoming slimmer. This is primarily to afford more cabin space and, through lower weight, achieve improved fuel efficiency.  The application of optimization can result in lighter metallic carriers and can also be employed to inspire the design of monolithic plastic carriers that integrate multiple features.  The Equivalent Static Loads (ESL) method is a way of representing a transient dynamic event with a static load history. It is a technique that can be applied to the design and optimization of closure systems in parallel to MBD event simulation from concept through to detailed design. Flexible and rigid bodies can be included in the optimization and all optimization disciplines are supported (topology, free size, topography, size, shape and free shape).

There are alternatives to using ESL, but each has significant disadvantages:

  • An optimization can be wrapped around an MBD solution – This implementation can only offer a limited number of design variables and constraints, in addition to being computationally expensive. This alternative is also not well suited to topology, topography or free size problems
  • The peak load approach – This has been historically more common, but offers up many questions in its application; Have appropriate boundary conditions been selected for each flexible body? Is the “rule of thumb” factor used to convert dynamic loads to static loads robust for all cases? Is the modified component structural design feasible within the constraints of the closure?

The animations below show the topography optimization of an inner door panel for a door slam event.

Topography optimization of inner door panel

Topography optimization of inner door panel.

Through the use of ESL Altair OptiStruct was able to demonstrate a 19% reduction in the panel displacement. The model shown accounts for the deformation introduced by the inertial force, Coriolis force and centrifugal force in addition to the externally applied forces and joint forces. Using the ESL method ensures that designs are optimized for updated loads through an automated process that also removes any possible ambiguity about the boundary conditions.

Drop Glass

Recently Altair worked with Faurecia on the optimization of an electric window mechanism. These systems require mechanical design, controls software and hardware, plus system calibration. With a focus on the window regulator the first task was to find a regulator cable location that satisfied the requirements for normal forces with the different combinations of friction coefficients possible within the mechanism.  The MBD model was co-simulated with an algorithm development tool, in Faurecia’s case Simulink. The MBD model was built in MotionView, run with MotionSolve, which has multi-tool co-simulation capability, and the evaluation of cable positions with the different combinations of friction coefficients in the system was driven by HyperStudy.

The Faurecia team achieved an optimal design that met all of the requirements.

Dany Desrus , Leader Product Line EE & Mechatronic , Faurecia Interior Systems stated “Co-simulation of window regulator with MotionSolve and Matlab/Simulink gives a full mechatronic model of the system, allowing simultaneous design of  both mechanism and actuator. HyperStudy allows direct parameterization of data and easy optimization of system design”.

In addition to providing a solution for this program the design process is now automated for the future.

Sliding Doors and Sun Roofs

Although the heyday of the minivan may be behind us, sliding doors are still an important part of the automotive landscape especially in light commercial applications. This type of closure brings its own modeling challenges, but there is a proven process for that can be applied using MotionView and MotionSolve that finds the correct balance of solution robustness and model complexity.

Sliding door & roller and guide rail

Sliding Door & Roller and Guide Rail.

The CAD for these closures is often complex but will often only provide a template for two important contact types, the roller to guide and the door/roof to seal. The actual track of the roller in the guide is usually a very narrow path within the guide. Managing the size of this contact, by drawing on prior experience, will greatly reduce the solve time required by the contact algorithms. In the closed position the contact against the seal will be represented as discrete forces around the closure. These forces are calculated from a force/deflection curve generated from test or simulation data.

Tailgates and Liftgates

FIAT applied a stochastic HyperStudy experiment to a MotionView BIW model with a flex body representation of a liftgate to design a closure robust to leaks over even the bumpiest of roads. The study was required to consider all production tolerances and deliver a result that was 3 sigma (~99.73%) effective. The following images detail the model and quality improvement process.

Quality process vehicle

Model and Quality Improvement Process.

Quality process flowchart

The positions and directions of door hinges, air strut attachments and lock were all varied within manufacturing tolerance. The stiffness of the air struts and the seals were varied as design parameters. This gave a total of 29 experimental variables.

Baseline design

Baseline design. 

The baseline design showed no issues (force excursions to zero), although there was potential for a leak over rough roads:

Force over time

Force vs Time Over Normal Road Surface.

Force vs Time Over Rough Road

Force vs Time Over Rough Road Surface.

The study revealed that through a reduction of specific manufacturing tolerances a product that achieved between 4 and 7 Sigma was achievable with no major design changes. The two distributions are plotted below.

Plot of Reduced vs Large Tolerances

Plot of Reduced vs Large Tolerances.

Hoods and Trunk Lids

Leaving aside the modeling of closure dynamics and focusing on the creation of mass-efficient structural components, consider the design of a hood Inner panel.  A typical design requirement is to increase the stiffness of the inner panel of an automotive hood or trunk. Variables in the design are the bead reinforcements and the adhesive layouts to the outer panels, with an optimization objective of minimizing weight. The images below show an industry solution for a hood achieved under the constraints of a maximum adhesive volume and a minimum material constraint for the part.

Hood support structure from OptiStruct

Hood Support Structure Designed Using OptiStruct.

This result was achieved using OptiStruct through a combination of topology and topography optimization. The topography optimization was used on the panel to find the optimal bead patterns and the topology optimization was used to identify the optimal layout and attachment points for the adhesive.

Future Closures

There is no doubt that there will be an increasing diversity of closures in the future; there are now many cars with rear hinged “coach doors” and demi-doors for side closures, trunks and lift gates have seen innovation with vehicles like the Skoda Superb and the Hyundai Curb concept shown at the NAIAS in Detroit this year.  Combining these design advances with multiple sourcing for global programs, new material adoption and higher performance expectations, it becomes clear that the efficient application of simulation to multiple aspects of closure design throughout the vehicle is essential to meeting program timing.

In this short article there hasn’t been the opportunity to explore all simulation applications in closure design. Other technologies include model morphing which can reduce the time for developing initial CAE models, eliminating the need to create concept CAD for the re-use and study of architectures, systems and parts. Manufacturing simulation was not discussed and can be performed as either 1-step forming assessments for early part feasibility or incremental stamping analysis to validate the manufacturing process. RADIOSS non-linear implicit analysis offers an alternative to older solutions for slam, seal deflection and other closure analyses.

It has been demonstrated many times that once a proven simulation methodology has been established there is further time and cost saving that can be realized through process automation. Automation increases quality and repeatability, plus a single standard method allows data comparisons across programs. Clearly the most important advantage is freeing engineers from manual, repetitive, non-value added tasks, and liberating their creativity to develop the next generation of automotive closures.

► Watch an on-demand webinar on Closure Simulation from the Automotive Webinar Series

Review a Door Slam Simulation for Durability Case Study

Review a Case Study on Closure Simulation

Learn about  the Altair HyperWorks Products referred to in this article

Learn about  Altair HyperMesh
Learn about  Altair HyperView
Learn about  Altair OptiStruct
Learn about  Altair HyperStudy

Tony Norton

Tony Norton

Executive Vice President - ProductDesign at Altair ProductDesign
Tony leads the Americas based Altair ProductDesign teams in the delivery of early concept (industrial design, design exploration, testing & prototyping) and advanced simulation driven design (cutting-edge modeling, optimization, methods development & automation) to our customers. Before joining Altair UK in 1996, he worked at both Ford Motor Company and GEC-Marconi Avionics. He moved to Michigan in 1999 to join Altair US, and holds a Bachelors degree from The University of Hertfordshire in England.
Tony Norton
Tony Norton

About Tony Norton

Tony leads the Americas based Altair ProductDesign teams in the delivery of early concept (industrial design, design exploration, testing & prototyping) and advanced simulation driven design (cutting-edge modeling, optimization, methods development & automation) to our customers. Before joining Altair UK in 1996, he worked at both Ford Motor Company and GEC-Marconi Avionics. He moved to Michigan in 1999 to join Altair US, and holds a Bachelors degree from The University of Hertfordshire in England.