1.4 Integration of a Subsystem Level in the Derivation
1.4.1 Targets of Full Vehicle Development
The subjective evaluation of ride comfort is a key method in the development process of a car. However, this method cannot be used in early stages of develop- ment as prototypes are still not available. Therefore, objective and computable criteria for ride comfort are needed. Such objective criteria, often in the form of characteristic values, should be related to relevant subjective criteria. A compre- hensive collection of subjective evaluation criteria is shown in Fig.1.6.
Representative characteristic values for ride comfort should fulfil different requirements:
Fig. 1.6 Evaluation criteria for ride comfort adapted from Heiòing and Brandl (2002, p. 115)
Computability
The characteristics shall be computable out of a simulation.
Measurability
Under real testing conditions, the values need to be easily identifiable out of measured signals from the full vehicle.
Completeness
The relevant ride comfort criteria have to be described by the targets completely and comprehensively.
Assignability
The transfer from subjective to objective criteria needs to follow a distinct process.
A literature research results in a huge amount of possible descriptions for ride comfort. Many of these approaches are based on correlation of subjective and objective evaluation, either with analytical weighting methods (Cucuz 1993;
Hennecke1994; Klingner1996; ISO 2631) or with representation of the unknown correlation by neuronal Networks (Albrecht and Albers 2004; Stammen and Meywerk2007).
The usage of these values is often restricted to selected contact points and corresponding directions between driver and vehicle or to specific excitation pro- files. Additionally many approaches summarize ride comfort in one value, which does not correspond to conventional testing methods in subjective evaluation and therefore not meets the mentioned requirement of assignability.
Therefore many OEM8do not work with such approaches and use specific, not weighted values for the description of relevant and standardized excitations instead.
An example could be the description of the response of a car while passing a cleat, as shown in Fig.1.7. In this case, the hardness and the decay behavior due to the impact are of particular interest in subjective evaluation. Characteristic values can be defined by the Peak-to-Peak value (P2P), the vibration frequency (fd) as well as the decay constant (δ) resulting out of the measured acceleration at the seat rail of the driver. These values fulfil the requirements above and can be monitored along the development process of the car.
Apart from the behavior of the car when passing a cleat, further characteristic values can be defined. These differ with respect to the operational methods of each OEM, but still the classification is usually carried out in similar categories. In detail this concerns the frequency dependent transfer behavior of the body, the previously described step response or the response on stochastic roads. Similar to Fig.1.7, characteristic values can also be defined in these cases. For instance, the body response over frequency can be separated in the range of its natural frequencies and in corresponding resonances of subsystems at higher frequencies. For stochastic roads effective values, like root mean square values (RMS), can be defined. In the product development process, these characteristic values are then used as a basis for a derivation of properties.
8OEM: Original equipment manufacturer; corresponding to the common definition in the auto- mobile sector, the term OEM means manufacturers of vehicles, selling them under their own brand.
For this purpose, primarily a comparison of current vehicles in one class of different brands is conducted and characteristic values are determined. Depending on the differences between competitor’s vehicles and the OEM’s own vehicles, advantages and disadvantages are identified. Subsequently, new target values for a successor or a corresponding new vehicle class are defined. The relationship between predecessor and new target values can be depicted using a bar diagram, shown exemplary in Fig.1.8.
At this precise moment, models on subsystem level already can support the development process as competitor’s vehicles can also be measured on subsystem Fig. 1.8 Definition of targets of full vehicle development for different characteristics
Fig. 1.7 Step response of the body when passing a cleat
test rigs, for instance on a kinematics and compliance test rig, while the determi- nation of component properties is not practicable in a limited time frame. The parametrized models allow for analyzing the contribution of determined parameters on subsystem level to the behavior on full vehicle level, supporting the basis for decisions concerning the derivation process.
In the present figure, natural frequencies of the body on front and rear axle as well as Peak-to-Peak and decay values have been defined as target values. Concrete values of the predecessor are visualized by a dashed line, while target values for the new vehicle are represented by a range of possible values, allowing for a tolerance when designing parameters. Additionally common areas within these values are depicted, which depend on the conditions of the particular criterion and vehicle class.
Overall the depicted process provides a possible approach for a transparent and integrated development process based on full vehicle characteristics which can be transferred to different objective criteria of subjective evaluation for ride comfort.