Fine Tuning Steering System Design using MSC Adams
This blog is contributed by engineers at PIX Moving, Siddharth Suhas Pawar and Robert Xu.
A steering system’s function is to provide the driver the ability to control the vehicle’s lateral dynamics. In the case of an automobile, the primary components of a steering system include:
● Steering wheel
● Steering column
● Steering Gear
Together, these components transmit the movement of the steering wheel to the front and sometimes the rear wheels. The steering system plays a very important role in determining the feel of a car. Moreover, the tire wear is largely dependent on how well it is tuned. Let’s now discuss the main sub-assemblies in brief:
The steering wheel — It’s the part that is used by the driver to provide steering input. The steering wheel also houses components like the horn pad and airbag system and other control switches.
The steering column — The steering column assembly holds the shaft that connects the steering wheel to the steering gear. The connection most likely involves the use of one or several universal joints. Most steering shaft are made of two sections — one hollow and the other solid. These halves are held together by soft shear pins that give away in the event of a collision, allowing the solid half to collapse into the hollow half. This minimizes the chances of injury to the driver. Such shafts are known as collapsible or telescoping shafts and form a vital safety feature along with the airbag system.
The steering gear — There are two standard steering gear designs: the rack and pinion arrangement and the recirculating ball mechanism. Because of it’s reduced steering force, many trucks, buses and SUVs use the recirculating ball mechanism. However, almost all passenger cars use a rack and pinion arrangement as it is cheaper. Steering gears are generally coupled with electric or hydraulic power steering systems (EPS, HPS) that drastically reduce a driver’s steering effort.
In this post, we will see how different kinematic software tools can be used to create and optimize a steering system design.
Steering Kinematics
Ackermann Geometry — If we assume a car to turn slowly, then ideally, all the tires have to be oriented tangentially to concentric arcs. The arcs’ center is the instantaneous center of the car. The Ackermann or A-condition stipulates that, in case of an ideal turn with no body roll or suspension effects, the inner and the outer steered wheels must turn as per the relationship –
where,
δo,A Outer Wheel Angle, in a pure ackermann turn
δi,A Inner Wheel Angle, in a pure ackermann turn
j distance of the steering axes on the road
l wheelbase
Now, the Ackermann angle is the difference between the steering wheel angles inside and outside in a true Ackerman turn.
One good yardstick to measure the A-condition is the Ackermann error. It’s deviation of the actual from the ideal Ackermann angle
where,
Δδf Ackermann Error
δo Steering Wheel Angle outside (Actual)
Δδa Toe-Out turns according to Ackerman
Δδ Artificial toe-out on turns
The Ackerman condition (A-condition) is important for passenger cars. A well-tuned Ackerman allows good turning characteristic, and prolongs tire life. During the design phase, engineers attempt to fine-tune the steering gear geometry to reduce the Ackerman error as much as possible. (This is not true for sports cars, where anti-ackerman design is preferable) Modern cars, though, do not use pure Ackerman steering, partly because it is difficult to achieve in practice and partly because it ignores important dynamic and compliant effects.
There are many other effects that must be considered during the design phase, like bump steer effect. However, for this post, we will now see how we can fine tune a steering system for good Ackerman.
A good and proven way to achieve good Ackerman effect is through simulating an initial design on a computer and analyzing the Ackermann Error. This is followed by iterative design changes and re-measurements of A-error till the designer is sure that Ackerman error has been minimized to the best possible extent. In short — fine tuning the system.
Simulation Tools
There are many software that can do this, most commonly used ones being MSC Adams and Lotus. MSC Adams is preferred by many companies and FSAE teams. This software is not very user friendly though, and the learning curve is initially quite steep. This post may be helpful in getting up to speed with using such tools.
Because the steering system linkage and its motion is three-dimensional, performing a 2D analysis is not precise. However, it is a good starting point. Working Model 2D is a great software tool for handling such 2D kinematic problems. Let’s have a quick look at what we can achieve through this tool.
Here is the initial layout of our rack and pinion steering gear (top view). The steering gear was purchased, which implied that the length of the rack, rack travel and the maximum pinion rotation is fixed. The tie-rods lengths are fixed too, but it was possible for us to decrease it’s length if needed, by machining.
There are many parameters that affect the Ackermann. These are —
● The alignment of the steering arms
● The length of the steering arms
● The alignment of the tie-rod
● The length of the tie-rod
Note that the above parameters automatically fix the rack’s position. The most important step towards minimizing Ackermann error is to ensure that the steering arms are correctly aligned. The steering arms must be tilted in the longitudinal plane of the car and their extensions must meet the intersection of the rear axle and the middle of the car. This step has the greatest effect on Ackermann steering.
Working Model 2D accepts .dxf files, a very useful feature. The dxf file as shown above is now imported into WorkingModel2D. This is followed by defining joints for the mechanism and then simulating it. The simulation data can be collected in the form of steering angles of both wheels and then plotted to check the deviation from pure Ackermann steering.
Many measurements can be taken from WorkingModel2D. In the above simulation, we measured the steering angles and the rack travel. Notice how we did not employ a proper rack and pinion meshed gear arrangement, instead we created a much simpler frictionless joint between them which does not affect the accuracy of this analysis. The rack travel was set at 50mm and a maximum inner angle of about 41 degrees is observed. Here are two plots from one of these simulations.
Both plots give an idea of the deviation of the system design in consideration, with respect to true Ackermann. The green scatter curve is plotted from the equations given before, while the blue scatter curve is derived from the simulation data. The angle difference in the second plot refers to the difference between the outer and inner steered tire angles. The Ackermann error is the vertical difference between the two curves and is maximum at the end of a turn.
Once this setup is achieved, the designer now needs to evaluate different steering gear geometries by varying the parameters that affect the Ackermann. After a few simulation trials, it becomes quite clear as to what parameters are affecting the Ackermann error the most and require tuning.
The simulations can be run till an acceptable Ackermann error is achieved (we brought it down to 3 degrees) and the final steering geometry may be used as a starting point for a 3-D simulation.
Let’s now see how we can use MSC Adams Car, a good computational tool for running kinematic and dynamics analyses.
MSC Adams Car contains a slew of vehicle subsystems that allow the user to build virtual test prototypes and analyze it quickly. The tool’s greatest strength is that it allows one to focus on studying how design changes affect vehicle performance rather than spending a lot of time on preparing virtual models. A good starting point (highly recommended) is this official manual. Let’s now see how we can use this tool to optimize a given steering system design.
For this analysis, we need a steering sub-system, a test- rig and a suspension sub-system. These sub-systems are available as standard templates.
Once this setup is selected, this standard template model needs to be modified to match the current design in consideration kinematically. This is done by modifying the location of hard points, which are located at link ends. The option is available in Adjust > Hardpoint > Table. The location of hard points is entered in terms of co-ordinates. It is therefore ideal and a good practice to place the origin at the center of steering system. This leads to symmetry making it easy to check for errors in location of hard points.
Once the right hard point locations are entered, the model is ready for analysis! For kinematics, the location of hard points matters the most. Also, the fine tuning and optimization of the steering system is done by changing the hard point location. The analysis option for our problem is in Simulate > Suspension Analysis > Steering. Depending on the steering system parameters, the upper and lower steering parameters are set. One can choose, to have rack travel or pinion rotation as input for simulation.
The simulation is now ready for review and can be viewed under the Review > Animation Controls option. More importantly, the plots for Ackerman Error can be viewed in the Review > Postprocessing Window. The simulations are shown below.
The plot below is obtained from the first simulation. The second simulation is a check for bump steer, a self-steering effect observed as a car goes through a suspension stroke. No bump steering is observed.
The steering system can now be optimized by fine-tuning hard point locations to minimize the Ackerman error. Compared to 2D analysis, this analysis is much more precise. Once a good-enough design is obtained, it is possible to export the final steering geometry (3D) as a step file. This can be done using the File > Export option. This exported geometry can be used as a guide to place the steering gearbox in the car and also to design the steering arms.
We used this method to design the steering system for PIX 4.0.
The final turning obtained is quite good! No slip was observed and the required turning radius was also achieved. We had done about 40 simulation trials in 2D to get a good understanding of the steering kinematics. When we turned to 3D simulation with MSC Adams, we came to an optimized result in about 6 design iterations.
Let us now discuss another kinematics problem.
We wanted to analyze the kinematics of a trailer that was attached to PIX 6.0, our latest four-wheel steered electric and autonomous vehicle. PIX 6.0 is designed to carry trailers that each serve a unique purpose — one trailer could be used for ferrying people, another for logistics and yet another for a self- driving coffee bar or a hotel. The trailers can be designed as per the needs of the market. PIX 6.0 is designed to have 3 steering modes –
● Two-wheel steering
● Four Wheel Steer
● Crab Steering
The engineering questions that naturally arose were —
● Where do we attach the trailer to the vehicle?
● Should a hinge joint or a fixed joint be used for the attachment?
so as to get the best turning radius and least tire wear (contact forces)
To solve this problem, we tried 18 different cases. The cases varied by the location of the joint between the vehicle and the trailer and the type of joint. This time, MSC Adams View software was used for this analysis, not Adams Car.
The most accurate way to perform the simulation is to take into account the suspension and steering system kinematics. However, this is time-consuming and not necessary. The turning performance can be predicted with sufficient accuracy by using universal and hinge joints between the tire and the chassis. It is also possible to mimic the Ackermann turning fully using this setup.
Here are some simulation runs using this setup: (videos embedded)
Such simulations help us predict the vehicle’s behavior under different steering modes. The first simulation depicts heavy turning loads on the wheels and the wheels seem to be collapsing. In reality however, the wheels will slip on the road under the high loading. MSC Adams can measure these contact forces as well. Using this simulation and the force results, we ruled out using a fixed joint in the four-wheel steering mode.
It is also possible to roughly compare the turning radius for the different cases. To do this, we traced the projection of the center of mass point of the trailer on the ground as the vehicle takes a turn. Again, let’s plot it! Plotting is one of the best ways to visualize and compare data.
Here’s one such plot
The green curve depicts the scenario of 4-wheel steering (crab-mode) with a fixed joint. The red curve depicts a four-wheel steering with a hinge joint at front and the blue curve depicts a two-wheel steering with a hinge joint at front (Figure 11). This plot shows that both crab-mode (with a fixed joint) and a four-wheel steering (with a hinge joint) will work equally well.
In order to arrive at a conclusion, many such plots were made based on joint type and joint location.
This data led us to choose a hinged joint instead of a fixed joint, because a hinge joint allowed the best turning radius and well as low contact forces when compared with a fixed joint. The data also informed us the where to place the hinge joint.
This way, virtual simulations can be very helpful and time-saving and allow engineers to take decisions quicker. The other way is to test physically, which may not be practical. An accurate enough virtual simulation can be a great guide to engineers to know what cases to rule out quickly. MSC Adams Car is a powerful tool to perform vehicle dynamics simulation. This tool can be quite a pain in the beginning, but once we get used to it, we can really appreciate the detailed nature of different types of analysis that it provides.
References:
1. Harrer, Manfred, and Peter Pfeffer, eds. Steering handbook. Cham: Springer International Publishing, 2017.
2. Hanzaki, A. Rahmani, P. V. M. Rao, and S. K. Saha. “Kinematic and sensitivity analysis and optimization of planar rack-and-pinion steering linkages.” Mechanism and Machine theory 44.1 (2009): 42–56.
3. Getting Started Using ADAMS/CAR. MSC Software Corporation
4. Getting Started Using ADAMS/VIEW. MSC Software Corporation
