This article explores the relationship between the damper and its mounting and outlines the objective techniques that would aid product development and benchmarking accuracy.
Damper hysteresis has been identified as a potential measure of the efficiency of a given damper architecture at all velocities and accelerations. Hysteresis within the damper is a well studied characteristic and in isolation its influence is understood. In motorsport applications, very low levels of mounting compliance are required and as a result ‘pure’ suspension damping can be achieved in a relatively straightforward manner (assuming correct valve architecture).
Road cars are very different because of the required compromise between chassis performance and NVH issues. With customers becoming more critical and wheel/tyre sizes increasing, the chassis engineer is faced with a difficult task. Damper mountings and strut top mounts are critical to achieving the right blend of control and isolation. However, the complexity of the combination of both mount hysteresis and damper hysteresis presents a major obstacle to the development of potential improvements.
It is well understood that damper performance is a function of acceleration. However, current methods for assessing damper performance do not fully take this into account and only display force / velocity or force / displacement. As a consequence, damper performance cannot be fully assessed. In order to gain a better understanding of damper performance, it is necessary to develop new test methodologies and new ways of presenting and analysing the measured data. Figure 1 and figure 2 highlight the difference in the force, velocity, acceleration envelope covered by the current and proposed methods.
Figure 1
Recent work by Lotus Engineering and Kingston University has looked at improving what is currently available and focussed on two key areas. Firstly, the need to move away from standard sine waves and redefine the waveform used for a damper test. Secondly, to generate new methods of post processing and analysing the data.
There are several benefits of adopting the approach above. By moving away from the standard sine wave profiles and adopting waveforms that are based on a mathematical model of a road profile, it is possible to fully exploit the force / velocity / acceleration envelope. Figure 1 clearly highlights the shortcomings of the current sine wave input. The large areas of missing data clearly illustrate that this technique has its limitations. The new damper excitation profile lasts only 30 seconds and fully populates the operational envelope as shown in figure 2.
Figure 2
All the experimental work was carried out on a Roehrig EMA2K damper dynamometer, with calibrated Misco accelerometers connected to the data capture card. The EMA2K uses linear motors to drive the damper piston, and is therefore not subject to the constraints imposed by a scotch yoke mechanism which produces the accepted sine wave input. A similar effect could also be achieved with a hydraulic dynamometer. The excitation profile imported was correlated to the actual drive profile produced by the dynamometer actuator to ensure the requested profile was achieved. It is clearly understood that damping rates are affected by fluid temperature and to ensure confidence in the results, the damper temperature was monitored by an infra red temperature sensor. It was maintained at 25oC ± 2oC through out all the testing.
A twin tube Kayaba damper was used during the initial study. A full range of shims, springs, discs, base valves and pistons are allowed for the adjustment of low, medium and high speed damping, in both compression and rebound. Other damper types have also been used within the later stages of this research to determine the robustness of the new testing methodologies. An adapter was fabricated to allow the top mount to be attached to the damper during testing. This top mount was also highly tuneable by varying the hardness and geometry in both the upper and lower elastomers.
From the early research at Lotus, it was recognised that one of the key criteria for any new testing/analysis methodology would have to be clear and concise data presentation. Therefore, considerable efforts were expended on deriving suitable plot types. MATLAB routines were written to process the raw data from the damper dynamometer into a more manageable and more informative format.
Figure 3
The first process was to produce a 3D plot of force, velocity and acceleration as shown in figure 3. A level of surface roughness (‘noise’) can be seen in this type of plot. This ‘noise’ is the hysteresis present in the damper or damper / top mount combination. This plot gives a visual indication as to where in the operational spectrum the hysteresis is predominant. There are numerous absolute and statistical values that can be extracted from these curves to fully define the damper’s performance. However, it was considered that these did not fully illustrate many of the differences that were thought to be present (from a damper perspective). Therefore, a number of other analytical approaches were required.
Viewing this 3D plot in the force, velocity plane is similar to the standard 2D force, velocity plot currently available. However, more details are available due to the modified test profile fully populating the force, velocity, acceleration envelope. Unlike standard force vs velocity plots, this 3D plot exposes the location of the hysteresis in relation to the acceleration experienced. When the 3D plot is viewed in the force, acceleration plane, shown in figure 4, the hysteresis can again be seen as ‘noise’ along the lines of constant velocity.
Figure 4
One method to quantify the hysteresis looked at evaluating the total force difference at each velocity point. This gave rise to a curve quantifying the amount of force hysteresis as shown in figure 5.
Figure 5
Extending this line of enquiry, the next approach looked at quantifying the magnitude of the hysteresis as a function of acceleration. This data was then processed to give a density plot in the acceleration / velocity plane, indicating where the hysteresis occurred. To reduce the data set, it was decided to set a threshold below which the hysteresis was not displayed. Allowing the user to quickly highlight the areas of damper cycle most effected, see figure 6. Again a number of absolute and statistical values can also be presented to aid with evaluating damper performance and pinpoint where the hysteresis predominates.
Figure 6
The approach illustrated above has provided greatly enhanced performance data. In particular, it has clearly defined the relationship between mount stiffness (static and dynamic) and the operating efficiency of the damper during highly transient inputs. From the quadrant diagram (Figure 6), it has been possible to measure the influence of differing top mount configurations on the damper valve ‘opening’ characteristics. Indeed, it has been shown that, across a very large percentage of those damper configurations and types tested, the key characteristic is dominant across the opening stages of both the rebound and compression valves (Quadrant 1 and 3).
This is of particular interest to the ride engineer who is seeking to derive the optimum arrangement for secondary ride. Typically the secondary ride issues—compliant yet well damped ride over highly complex road surface changes—can be affected by the correct choice of intake (recoup) plate thickness (and / or control spring). This was clearly shown within the quadrant plot, but not within typical peak force / velocity data or displacement loop diagrams. Furthermore, the influence of damper / mount characteristics, whether by tolerance variation, or by design, can be integrated within the tuning process and its effect minimised.
It was successfully shown during development of this methodology, that a softer top mount (lower static stiffness) could be used, if an appropriate mechanical arrangement was selected that maintained or enhanced the damper hysteresis characteristics. Therefore, it will be possible to benchmark potential top mount configurations during the early design stage, select those that produce minimal increases in damper hysteresis and hence significantly improve ride optimisation activity during product development.
Keywords: Damper Testing, Vehicle Ride Tuning, Damper hysteresis, chassis performance, Damper performance
