一种评估汽车照明的新方法

NCAP-照明:一种评估汽车照明的新方法

Keywords: Lighting development, Simulation, Temperature simulation, Stray light, visual impression 1 Introduction In the development process of headlamps not only decisions about volume or shape of outer lens and housing are necessary in an early state of the project, to an increasing degree also the internal styling and the lit and unlit impression of the headlamp should be fixed before starting any tool. Therefore the need for sufficient simulation tools is increasing. The calculation of the light pattern and the photorealistic visualization is already state of the art in the lighting development. Further simulation tools e.g. for heat dissipation of outer lenses, stray light investigations or the visual impression of the lit function have been developed. To use this in the development process it is important to have the knowledge about the limits of the simulations, and to plan also prototypes for the necessary review in reality. 2 Concept phase The very first investigations done in the headlamp development process are more or less related to the question, if the available volume is sufficient to integrate the required functions. The bases therefore are typically a few rough sections like the horizontal section shown in figure 1. For the evaluation of the photometric performance it is obvious to judge the size and shape of the reflector itself. But the surrounding has also to be considered. Important for the spread of the light distribution are the angles to the optical axis of some characteristic rays as shown in figure 1.     Figure 1: Horizontal section to investigate feasibility in concept phase Another aspect is the expected temperature of the outer lens. In this early stage the distance between the filament and the outer lens is compared with the experience from previous realized headlamps. 3  Design phase During the design phase many aspects related with lighting technology have to be evaluated. The simulation of the light pattern itself and their visualization on real or virtual scenes is already state of the art in the lighting development. Further evaluations e.g. the heat dissipation of outer lenses, stray light investigations, influence of bulbs shields or the visual impression of the lit function are necessary. 4  Outer lens temperature The temperature load of the outer lens of a headlamp is caused by the overall temperature inside the headlamp overlaid by the specific light and heat distribution caused by the optical system. Due to this approach the work on a fast simulation tool for the heat dissipation of the outer lens was started in the software which is used for the lighting design. First, basic investigations were carried out in a simplified headlamp set-up, a reflector with a halogen bulb and a dirty lens with 20% light transmission. In this set-up the temperature load caused by the filament and the bulb itself was analyzed. In parallel, special bulb models and calculation algorithm for the lighting simulation tool ASAP were created and optimized until the agreement between measurement and simulation was good. Next, complete headlamps had to be analyzed, i.e. the convection inside the headlamp has to be taken into account by the simulation. This is done with an approximation of the air temperature inside the headlamp behind the outer lens, which depends mainly on the bulb power and the distance between filament and lens. Also the prediction of the temperature caused by the ECE temperature cycle test is included. Figure 2: Outer lens temperature simulation in comparison to measurement The SAE internal heat test leads to a different temperature load of the outer lens. There are differences in the voltage, in the degree of soiling and in the outside temperature. Therefore another set of parameters was carried out to correlate simulation and measurement. Now this method allows a fast simulation of the outer lens temperature already during the design of the optical surface for “conventional” headlamp shapes. Important for the application of these simulations is the knowledge about some limits. Because only the temperature load caused by the light distribution itself is specifically calculated for this application and the overall temperature inside the headlamp is a prediction, the accuracy depends on the similarity of the headlamp shape with the lamps used for the correlation. This means for example that the temperature of the lens in a headlamp with the reflectors of low and high beam on top of each other will differ from the simulation. In front of the upper reflector the real temperature will be higher than simulated, because there is an additional heat load caused by the reflector located below. Influence of additional parts In the development process often some parts, which define the visual impression, are not completely fixed when starting with the design of the optical surfaces. These can be some additional shutters in signal lamps, inner lenses or even the bulb shields which are not finally defined. Bulb shields are in principle used to cover the direct light of the bulb and because of that an optical element. But in headlamps with clear outer lenses the design of the bulb shield is mainly driven by the visual impression combined with some mechanical aspects. Only the trimming line of the shield is defined by the photometrical needs. As an example the influence of the leg design of a bulb shield is shown in figure 3. The bulb shield with 2 legs was taken into account during the development of the reflector; the corresponding light distribution is shown in the upper simulation. As shown the light distribution has a homogeneous decrease of intensity to the sides and the over head illumination is also considered. With the bulb shield with only one but wider leg on the bottom the quality of the light distribution is reduced. There are darker areas in the side illumination of the pattern and the overhead illumination is completely lost.     Figure 3: Reflector with bulb shield with 2 legs (left) or 1 leg (right) and the influence on the light distribution: simulated light distribution with 2 legs (top), with 1 leg (bottom) This kind of simulation is often necessary to show to all parties the influence of mechanical parts on the photometrical performance of the system. Stray light investigation A wide field for simulations during the development process is the evaluation of stray light. There are two effects which cause stray light: the direct reflection of light coming from the reflector or PES module on a reflecting part (e.g. reflector side walls, metalized bezel) or a multi reflection first on the outer lens and then on reflecting parts. Both effects are shown in figure 4.

 

Figure 4: Stray light in headlamps caused by direct reflection (top) or double reflection (bottom)

The investigation of stray light is important, because it can have several negative effects, e.g. it can cause stripes in the near field of the car or it can lead to glare problems. The stray light caused by direct reflection on the bezel like shown in the upper sketch in figure 4 could also be investigated with a vertical section of a drawing. But in general the effects are more sophisticated like shown in figure 5 and only visible in a 3 D ray tracing. And even more complex are stray light investigations for swiveling low beams (see figure 6).

 

Figure 5: Ray tracing for stray light investigations

The results of the stray light simulations in figure 6 show some hot spots with more than 1000 cd. For US applications there is a limit of 125 cd in the glare zone 10-90U. Since these simulations were done with a perfect mirror as bezel, there is a good chance to reduce the glare level by introducing a structure on the bezel surface. The question is how much this measure will reduce the maximum stray light spots. For this estimation only simulations or only tests on prototypes do not give a secure result. A combination of both leads to the most trustable results.

The simulations with the reflecting bezel lead to the areas of the bezel which are involved in the reflection of stray light. With this information it is possible to discuss the zones for structures with all involved parties of design and manufacturing in an early stage of the development process.

 

Figure 6: Stray light investigation for a swiveled low beam Bi-Litronic module

Visual impression of lit function

For signal functions beside the conformance of the light values the visual impression is the most important feature. For the demonstration of the visual impression it is the best to make real parts. But also if parts are available and some changes are required, the simulation of the lit function is helpful to define corrections. Figure 7 shows the simulation of the visual impression of a turn signal reflector before and after a correction in the lower right corner.

 

Figure 7: Simulation of the lit impression of a turn signal reflector before (left) and after (right) correction

Conclusion and Outlook

The examples shown above demonstrate the necessity of simulations during the development process. And they also show that the results of simulations can not be taken automatically as true. Before a simulation tool will deliver trustable results, there is a phase after the development process itself, where the correlation between simulation results and the reality has to be worked out. This has to be done for algorithms and for light source models.

But even then simulations can not show exactly the results which will be measured or seen later with parts. One point is the agreement of the CAD data used for simulation with the shape of the real parts having some tolerances. Also the quality of the surfaces (e.g. lacquered and metalized reflectors) will never be exactly like the theoretical description. Additionally also the model of a light source can never include all possible variations of the real light sources. All this leads to the point, that for the interpretation of simulation results always the experience of the conversion from theory to practice has to be considered.

For the application of simulations also the time can be a restrictive point. This does not only mean the time for the calculation itself but also the time to prepare the data for simulation. In some cases the preparation time and the simulation needs more or less as long as to make samples. This is a field where further improvements are necessary.

Another field for further development of the tools is the simulation of surfaces with diffuse reflection. The simulation of direct reflections works very well, for surfaces with flutes or with structures new models have to be developed and verified. It is similar for the transmission through any kind of material. As long as they are clear and have a smooth surface, the agreement of the simulation results is very good. For structured or opaque materials further improvements of the simulation tools are necessary.