Advances in the design process for guiding mechanisms in movable car body parts
Fundamental changes in component design
- Univ.-Prof. Dr.-Ing. Peter Gust, Bergische Universität Wuppertal
- Dipl.-Ing. Volker Dürhagen, Kirchhoff GmbH & Co KG, Halver
- Dipl.-Ing. Eliseo Milonia, info-key GmbH & Co. KG, Wuppertal
- Dipl.-Ing. Ahmet Gülcan, info-key GmbH & Co. KG, Wuppertal
In automobile construction, four-bar hinges are well-established solutions for the task of guided opening of rear and front hatches. Requirements for a front hatch are, among others, a sufficient vertical and lateral rigidity, while occupying a minimal installation space at the same time. Forces resulting from the airflow towards the windshield are trying to lift the front hood, i.e. they act vertically upwards. The elasticity of the system will in the worst case lead to oscillations of the hood corners at high speed that can negatively impact the driving experience.
To prevent this, the four-bar mechanism must be sufficiently rigid. This is one of the main requirements for the hinge system. Generally, a designer has to satisfy several opposing demands during the design process. To satisfy these demands while the available design time is reduced ever further, new strategies for the design of four-bar hinges have been developed in a joint research project.
This lecture will present these new methods, which also encompass synthesis and computer-based optimization, by means of examples. In a computer-based design strategy, classical methods like the two- and three-plane design are used to determine valid hinge variants. The found variants are then rated with respect to how well they satisfy the given catalogue of requirements and the best proposal is delivered to the designer.
This way, the design process was shortened significantly while solution quality was increased at the same time.
Why do we need "new" design processes for four-bar linkages?
The automobile industry is confronted with ever new and at the same time long-known challenges. While in the late 80s the time frame for the development of a vehicle was still 50 months, it was reduced to 30 months in 2003 - a reduction of 40% . And while development times decreased, the efforts involved in vehicle development increased at the same time. Power windows and automatic seat adjusters are already standard features in many cars. It is expected that by 2015 even powered hatch and hood opening will be widely offered in standard and luxury vehicles, at least optionally . Thus, systems with increasing complexity have to be developed in a decreasing time frame.
From these demands resulted a philosophy to switch from the more complex four-bar hinges to the simpler one-bar hinges, especially for rear hatches. The design, the auto-opening (the automatic opening of the hatch after just operating the handle) and the implementation of a drive system usually are easier for a one-bar hinge than for a four-bar hinge.
A more recent statement by industry insider Helmut Becker proclaims: "The german automobile industry is solely concerned with absurd increases in performance and high-tech gadgets. In the country of carmakers the creation of affordable and economical models is considered to be beneath notice" . The exclusive trend towards more comfort thus is not sufficient. At the same time a trend arose towards cars that are more affordable regarding purchase price and gas consumption. In this context, the four-bar hinge as the smaller, lighter option, gains renewed significance. On the other hand, any current design process for a hinge system must take the increased requirements into account. The development has to be progress rapidly, but it also has to fulfill the customers' demands by sufficient safety margins. Therefore, conventional design methods are becoming insuffcient and current processes have to be extended, optimized and automated.
In the following, the typical requirements for hinge systems and the conventional design process will be described. Subsequently, a new design method for four-bar hinges is presented. This extension has been integrated into a software, developed expressly for this purpose. The software and its application in the design process will be briefly demonstrated after that. A conclusion and an outlook will close the lecture.
Requirements for automobile hinge systems
Since the various manufacturers have differing requirements, we will give here only an overview of the essential requirements in an abbreviated form (all of the numerical values here represent orders of magnitude and have been deliberately manipulated so that they do not reveal any proprietary information). Particularly, it is not possible to present a complete specification sheet. Thus, only those requirements are discussed that can directly be affected by the presented design method. The following requirements exist for hinge systems:
Kinematics; Fulfillment of the core task of guidance, i.e. the motion of the hatch from the closed into the open position. During the motion, no dripping water may enter the trunk. Furthermore, the trajectory of the head of a user bending down towards the trunk may not collide with the edge of the hatch.
Rigidity to stabilize against lifting forces; Due to aerodynamic forces during travel, uplifting forces act on the corners of the hood (in front of the windshield). The hinge kinematic has to properly counteract these forces.
Rigidity to prevent lateral movement; In the open position, the hinge has to possess sufficient rigidity to preclude lateral movement.
Operation; During opening and closing, no excessive friction forces may occur in the hinge itself, so that usually plain bearings (bushings) are used. The hinge must be able to withstand at least around 6,000 duty cycles for front hoods and around around 30,000 duty cycles for rear hatches without developing additional play. During operation, the hatch should open automatically, starting at a predefined angle between 14° and 25°, usually by way of a gas spring. During closing, the manual forces for operation below 14° should be small enough that the hatch just falls shut and latches automatically.
Installation space; In addition to all the requirements concerning rigidity and kinematics, the hinge also has to fit into the prescribed installation space. Usually the installation space is not defined as an independent entity given by "height x width x length". Rather, the specific installation space that may be used by the hinge is defined by all adjacent components, and it is provided as a CAD model. Thus, the designer has to construct "his" hinge directly into this installation space. In the case of collisions with neighboring components, an iterative coordination process with the customer begins, to see if additional space in that place (or somewhere else) can be granted for the hinge.
Those are the requirements that are essential for the kinematics. There are already, as mentioned above, a large number of additional requirements, relating e.g. to corrosion properties or to specific requirements for locking the hinge in the open position.
Another essential requirement for front hoods is the protection of pedestrians, which will be explained briefly. There is a more general requirement, prescribing that in the event of a collision with a pedestrian, the hood of a car has to be sufficiently deformable to prevent lethal injuries in as many cases as possible (measured as the HIC value – the Head Injury Criterion). Since most vehicles with powerful engines no longer possess sufficient room below the hood to allow for the necessary deformation, the hinge has to lift the hood upwards if a collision of a pedestrian with the bumper is detected. Figure 1 shows such a hinge that can raise a hood, using mechanical energy stored in a torsion spring. The torsion spring is mounted on an additional lever that pushes the hinge upwards once the action is triggered. The whole process takes no more than 12 ms from start to finish and is completely reversible.
Figure 1: Prototype of a hinge with pedestrian protection using a triggered torsion spring as mechanical actuator manufactured by Kirchhoff GmbH & Co. KG
It is easily apparent that the addition of such a pedestrian protection feature will greatly affect the overall design of the hinge kinematics.
The development process
Figure 2 shows a general model for product development, based on VDI standard 2221. This has been extended by two additional layers here. Layer 1, the process layer, has been extended by the important modules of quality design and manufacturing process design. Into these modules, three important milestones that usually are employed in vehicle development have been inserted as examples. SOP here stands for „start of production“, the start of mass production, a date that can "never" be moved, at least not by a supplier.
Figure 2: Layered model for systematic development and design
In layer 2, the layer for methods and tools, various methods and tools are shown - grouped by the corresponding process step - which are employed during product development. The third layer is the competency layer. It shows that to employ certain specific methods or tools (and this is especially true for complex software systems), corresponding specific know-how has to be acquired first.
The hinge development exemple presented here focuses on the two fundamental processes of design and specification, and the main emphasis is placed on the design of the hinge. Due to the requirement to fit into a specified installation space, in practice this is generally an iterative process alternating between the two steps.
The "conventional" hinge development process
Figure 3 depicts the conventional process for hinge development. A sequence of process steps of this type generally emerges if there is no opportunity to fit the process specifically to a certain product. In step 1, the installation space is imported into a CAD tool and it is assessed how much room is available for the hinge. In the next step, it is checked if a pre-existing hinge (or parts of one) might be re-used for the new hinge. Even if only certain parts can be re-used, there will be synergy effects, like e.g. corresponding savings in tool costs. Due to the specific installation space situations, though, most of the time no components can be re-used for the new hinge. Still, as long as the sheet metal thickness remains the same, at least the hinge pins and bushings can often be kept as common parts.
Figure 3: Conventional development process for four-bar hinges
In the next step, the joint positions are placed in the CAD system (in Germany this usually is the software CATIA by Dassault Systems), according to the experience of the designer. In the CAD system it is usually also possible to do a basic kinematic analysis, i.e. the hatch geometry is rotated around the placed jpint positions, and the designer assesses if the requirements concerning the guided motion are fulfilled. After suitable joint positions have been found by an iterative process, the actual three-dimensional geometry of the hinge system parts can be designed in detail. After deriving mechanical drawings for all parts, a first prototype is constructed. In the conventional process, prototypes are constructed using customary methods of sheet metal working. A very common method is the milling of auxiliary bending tools, which are then used to fashion pilot parts for the top, the bottom, and the two connecting rods of the hinge from rough parts that have been laser-cut from sheet metal (cf. figure 4a). This has the advantage that a limited series can also be manufactured using these same bending tools. Until the start of mass production, often up to 200 sets of such near-series prototypes may be produced. An alternative for initial prototypes is 3D milling from solid material. This has the disadvantage, though, that each set requires the same production effort.
Audi TT front hood hinge (formed from sheet metal, which is typical for vehicle hinges)
VW Phaeton rear hatch (components fashioned from forged aluminum)
With these near-series prototypes, the vertical and lateral rigidity properties of the hinge can then be measured and compared to the manufacturer-specific requirements.
After some initial installation tests at the customer, there usually follows an iterative process yielding the final design that is then used to fashion the tools for mass production off of it.
A fundamental disadvantage of this process is the purely manual determination of the joint positions. Even just slight geometrical variations applied to the joint positions may already result in a completely different kinematic behaviour of the hinge. As an example, a project by Kirchhoff GmbH & Co. KG for a powered hinge (i.e. an automatic rear hatch) can be cited. Here, a displacement of the upper joint of the long rod of 1mm horizontally and 1.5mm vertically yielded a reduction of the necessary drive torque of 50%. This is only one example from many illustrating the possible influences on hinge kinematics. It is thus very difficult to assess all the effects that a change in geometry may produce, purely based on experience. This makes it harder to assess and optimize solutions, resulting in a fairly complex process that may easily take up to three weeks for a new hinge kinematic.
The new hinge development process
In the context of vehicle development, as described above, three weeks for the design of a new hinge kinematic is entirely too long. Thus, a revision of the process was imperative. Following that assessment, the company info-key GmbH & Co. KG developed a new software tool, ASOM (Analysis, Synthesis and Optimization of Multi-bar systems), which by now is also available as a commercial version.
The development of a new software tool alone was not sufficient, though, to achieve not only a significant reduction in development time, but also the desired improvements in quality of results. According to the layered model for development (figure 2), the new tool had to be integrated into the existing process and corresponding expertise in the use of the new process had to be gained.
Figure 5 shows the process after it was extended by the new contents. After checking if pre-existing hinges may be re-used, additionally two two-dimensional zones (one on the car body, one on the hatch) are defined as closed polygons to describe the available installation space. These zones, together with similar descriptions of e.g. the water channel and the hatch corners, are then passed to the software ASOM.
On the basis of the given restrictions, ASOM performs a multi-bar synthesis, where the resulting multi-bar system can afterwards also be modified manually - under consideration of all restrictions and synthesis parameters. The trajectories or traces of selected points can be displayed, to verify the fulfillment of exclusion conditions and quality criteria. Energy storage elements like gas springs can also be added or synthesized and their properties can be modified by the user. Diagrams can be used to display forces, among other quantities, which can thus be used as quality criteria (see below). All the stages of the process can be performed in real-time.
In the following step, the designer has the task to define quality criteria for an efficiency analysis and weight them against each other. For a hinge design these are e.g. the position of the instantaneous center of rotation to improve the rigidity against uplift, or the height of the hinge that directly affects the rigidity against lateral movement. Another very simple criterion is e.g. the sum of all link lengths which can be used to gauge compactness and weight of the hinge.
These and more criteria are compared in an efficiency analysis, and then the designer receives a suggestion for a hinge that performs "best" based on the criteria he defined according to his experience. The result of the analysis thus depends directly on the experience of the designer. In this context, ASOM also allows, by way of an independent and easy to use rule-set module, for the integration of multifarious criteria. With this module, simple, but also very complex, conditions can be defined freely and can be considered additionally during synthesis. These condition-sets then embody the experience of a designer, in a way, which makes it possible to incorporate the core know-how of a designer in a very individual manner. Without the designer's experience with the specific requirements his customer has for a hinge, the process cannot be applied successfully. This experience is captured by the rule-set module and is thus made accessible for use in future projects or even for use by future employees of the developing company.
Figure 5: The new, extended development process
Due to the automation of the process, due to the fact that the designer is now able to confine himself to dealing only with valid hinge variants, and especially due to the fact that with a proper rule-set, essential quality criteria can be considered and refined in an ever more automatic way, it was possible to speed up the process of hinge development considerably. Due to the computer-aided evaluation of the quality criteria, the process is made much more transparent and comprehensible. This also can represent a crucial advantage in discussions with customers.
Altogether, the process of the development of an initial functioning hinge kinematic could be shortened to about five days. This represents a major reduction.
To further optimize the overall process, research into prototype creation using a generative rapid prototyping (RP) method was carried out. In the RP method, the components usually made from sheet metal were instead generated by selective laser sintering (SLS) and infiltrated with metal in a subsequent step. In that way, the components gain almost the same properties as construction steel (like e.g. ST37) and they can already be used for test-installs in a prototype vehicle. An advantage of the method is, that components can often be generated almost completely automatically and over the time of a weekend. The effort, in comparison with conventional methods like milling and bending, is reduced by a factor of 2 - another essential reduction of development time. On the other hand, here as well, as with any new software, additional time is needed to build up the necessary expertise, by gaining experience from multiple test cases. Without enough experience in the new method, the results should not be presented to customers.
All in all, the development time was reduced as planned by the described method, and also the results of the development were made more comprehensible.
Conclusion and Outlook
The project was a success for all participants, and the extension of the methodology has proven reliable, as described.
Still, it could make sense to extend the process by yet more modules. Especially the detailed design of the 3D geometry of the hinge has to be extended with the application of the finite element method (FEM). This is already used today for virtual testing, i.e. for the computational verification of the mechanical properties. So far, though, a full integration into the development process with a corresponding modification of the interface has not been attempted. Also, a combination of multi-body simulation and FEM would be useful. In this way, even the elastic deformation of the hinge can be considered as an influence on the hinge kinematics. Currently, computations to test this integration are carried out with the software system Recurdyn (Supplier and co-operating partner: FunctionBay GmbH, Infanteriestr. 19, Haus 3, 80797 München), which may soon be ready for presentation.
Analogously the use of topology optimization will be useful. On that aspect, studies are made with the system CATOPO (Supplier and co-operating partner: CES Eckard GmbH, Hessbrühlstr. 61, 70565 Stuttgart), a software that is especially well-suited for the use in a design process. Thus, an integration into the described process is expected to be possible without many problems.
Ultimately, complex development processes always have a potential for further refinement. In the scope of this lecture, only some high points of the work that was actually carried out could be presented.
- Harvard Business Review on car manufacturing, date unfortunately unknown.
- Market and Technology Survey - Automatic drives for doors and hatches, Arthur D. Little, Munich 2004
- Pedestrian protection - Kühn, Matthias; Fröming, Robert; Schindler, Volker - textbook, Springer 2007
VDI report no. 2050
Technology of motion, 2008
Linkages, curved guides and controlled drives in machine, device and vehicle manufacturing - Conference, Fulda, September 23 and 24, 2008, VDI Verlag GmbH, Düsseldorf 2008, page 241ff.