Design and Performance Analysis of a Roll Damping Function for an Electromechanical Active Roll Control System Meindert Solkesz, Department of Precision and Microsystems Engineering Welcome, i’m very happy that you all could come to my presentation about my graduation project. My name is Meindert Solkesz, and today I will present a part of the work I did the past year. The title of my graduation project is ‘Design and Performance analysis of a roll damping function for an electromechanical active roll control system. First I will give some background information in which I explain what an eARC system actually is. Then I will continue with the objective of this thesis, and after that i will dive into the design part of the roll damping function. Followed by some results of the tests I carried out. But first, i introduce you the ford motor company in Cologne. After all this is where I spend most of my time.

The Ford Motor Company a brief introduction Ford Werke - Merkenich 500.000 m2 of design centres, test tracks, equipment Global Advanced Vehicle Dynamics Active suspensions Active safety I worked at GAVD where i worked on my graduation project Also projects concerning active safety such as Stability control/ traffic sign recognition / lane keeping assistance After i introduced the company, i would like to introduce the test vehicle that was available

SUV demonstrator vehicle Range Rover Sport -specs 4.2L supercharged V8 (390Hp) Hydraulic Active Roll Control system Airsprings front & rear 2700 kg Here she is A normal passenger car weighs about 1500kg...

Terminology Degrees of Freedom Translational directions This vehicle has six degrees of freedom... Roll most important, traditionally roll is reduced by a stabiliser bar Translational directions x-direction: longitudinal y-direction: lateral z-direction: heave Rotational directions θx: roll θy: pitch θz: yaw

Terminology suspension linkages dampers springs stabiliser bar Here you see a front suspension of a conventional vehicle Connection between wheels and body Therefore responsible for the coupling between road inputs and body Stabiliser bar in conventional vehicle reduces roll in conventional vehicle and it works like this

Stabiliser bar even input uneven input However how large this angle will be give a certain cornering speed is dependent on the stiffness of the stabiliser bar. This effect of the stiffness on the driving behaviour of the vehicle is explained next

Stabiliser bar Weak vs. stiff weak stiff Cornering (handling) Compromis between handling and comfort The best would be to change the characteristics of the stabiliser bar according to the driving situation. Straight driving (comfort)

Active Roll Control (ARC) Suspension layout Actuator it is possible to manipulate the roll angle Compromise improved Hydraulic actuator mounted orginally in the vehicle as mentioned at the beginning of this presentation, however there are some advantages and disadventages:

Active Roll Control Hydraulic vs. Electromechanical system Hydraulic pros Lots of power Relatively simple Knowledge available cons Continuously running oil pump Slow Electromechanical pros Energy efficient Fast cons Complex Proven technology Slow: therefore good in slow manoeuvres such as cornering, not good to filter out bumps in road for good comfort Trent of electrification in the automotive industry, and demands for less fuel consumption Expectation is that electromechanical is better, therefore implemented in demonstrator vehicle

SUV demonstrator vehicle Range Rover Sport –modifications for eARC Electromechanical actuators at front & rear Sensors Programmable control hardware Data acquisition hardware LCD-screen It turned out software was not compatible with eARC because of lack of damping

Software modification motivation Lack of damping in the implemented actuators Demand for a roll damping function in the existing control software Potential of energy regeneration This is where my graduation project started

Objective “Design a roll damping function for an electromechanical active roll control system (eARC) and analyse its performance” Steps taken: Modify veDYNA computer model Validate computer model with SUV demonstrator vehicle Modify the eARC controller with a roll damping function Analyse the open-loop for stability Analyse the closed-loop performance Implement the controller in the real vehicle Performance test of the roll damping function with regard to power consumption, comfort and handling

Validation Kinematics & Compliance test rig Quasi static measurements

Validation Parameters that qualified for validation: Tyre stiffness Vehicle body mass, COG, inertia and roll centre Vertical stiffness suspension Roll stiffness A lot those parameters not only relevant for computer model, but also for the eARC controller That’s all about the validation and computer model I will now continue with the development of the roll damping function

Roll damping function Control loop Turned out that the system became unstable, unexpected when increasing damping Analysed response of vehicle body as a function of the roll rate road input

Simulations Performance low level controller no stabiliser bar demanded tq = 0

Roll damping function Control loop

Roll damping function Open-loop

Open-loop frequency response Bode plot Difficult to analyse stability with gain and phase margin

Open-loop frequency response Nyquist plot Magnitude Phase Therefore nyquist

Nyquist stability criterion Most common phase & gain margin, easier however the vector margin which combines both stability creteria

Open-loop frequency response Nyquist stability criterion

Phase lead filter

Notch filter

Modified roll damping function

Open loop with filter Optimal stability Now that the stability is ok, let’s have a look what the closed-loop performance with regard to noise rejection looks like

Performance 4 poster rig measurements

Closed loop response Optimal stability demanded tq = 0 activated roll damping

Closed loop response Optimal performance demanded tq = 0 activated roll damping Amplification for higher frequencies less relevant, because human body is tolerates it better

Stability optimal performance Nyquist stability criterion Stability ok, now let’s investigate the performance in real life with regard to power consumption, comfort and handling as mentioned at beginning

Closed loop response Optimal performance 1.8 passive activated roll damping 1.8

1.8 Hz Passive stabiliser bars Active with roll damping

Power consumption 4 poster test Roll damping function activated No net energy regeneration Not as expected

Power consumption 4 poster test

Comfort

Handling Slalom manoeuvre Without rolldamping With rolldamping Now the results of the tests lead to the following conclusions

Summary Computer model modified and validated Stability analysis performed Stability improved with the use of notch and phase lead filters Functioning of the roll damping function judged with regard to power consumption, comfort and handling

Conclusions Stability of the roll damping feedback loop is improved with a combination of a notch and lead filter. Filters tuned for optimal stability do not provide optimal performance The roll damping does not regenerate energy. The roll damping function has a positive / a negative / no influence on the comfort. The roll damping function improves the handling.

questions

Validation Vertical stiffness – vertical position

Validation Vertical stiffness - avarage Rebound stop Rebound spring Air bellow Spring aid

Validation Vertical stiffness – 31 sec bounce cycle

Validation Vertical stiffness - 300 sec bounce cycle

Validation Roll stiffness

Validation Roll stiffness

Active Suspension principles Fully active Semi active energy added no energy added Example: damper with adjustable damping coefficients Fully active energy added Example: electromechanic/hydraulic actuators at 4 corners

Vehicle model veDYNA

Controller high-level

Controller low-level

Roll damping function Frequency response

Damping ratio Paragraaf subkop

Paragraafkop Paragraaf subkop

Rollrate signal method Vertical acceleration sensors Lateral acceleration sensors Rollrate sensor

Rollrate signal estimation Vertical acceleration sensors

Roll damping function Control loop

Open loop measurement Paragraaf subkop

Power consumption Simulation -double –step-steer manoeuvre Average power consumption: 184 Watt 68 % decrease of powerconsumption

Power consumption Simulation -Bumpy road Average power consumption: 31 Watt 50 % increase of powerconsumption

Validation Kinematics & Compliance test rig

Closed loop response Optimal performance passive filtered 6

6Hz Paragraaf subkop Passive stabiliser bars Active with roll damping

Closed loop response Optimal performance passive filtered 7.6

7.6Hz Paragraaf subkop Passive stabiliser bars Active with roll damping

Open loop filtered Bode plot