hexagon logo
Register Sign in
  • State Not Answered
  • Replies 6 replies
  • Subscribers 7 subscribers
  • Views 160 views
  • Users 0 members are here
Options
Related

How to simulate the Slalom maneuver in A/Car

viktor.kulagin
Hello
I want to simulate the Slalom maneuver (which means driving between the cones located with equal interval).
First, I tried to set up the Sine-Sweep Steer according to Slalom characteristics. I mean, to find correct values of steer frequency and steering angle to ensure model riding with proper lateral displacement (which more than a half of car width, obviously). But the path of the model (when you check it in a postprocessor) doesn't look like as desired to be (in a picture attached). There is an explanation for this issue, I know
So, Sine-Swept Steer maneuver doesn't meet the requirements.
I remember that one day I found a picture of the FVA staying on a road with cones (for slalom maneuver, I suppose). So, there must be a tool to build this road. But I didn't find anything looks like a cone in Obstacles menu in Road Builder.
If you know how to simulate the slalom maneuver, please, tell me about it.
Thank you in advance.
Attached Files (1)
  • 0
    Hi Viktor,
     
    1. Sine-Sweep event should have served the purpose. But, from the picture you posted it seems like your vehicle is drifting. May I know whether the model was built by you or is the Model from   ADAMS Shared database? Use the shared database model, and simulate the sine-swept steer event and compare the results.
    2. Another solution is to build a road using road builder with the desired path as center line to the road. (*.rdf). Then you can use course event option for simulations.
    3. The third solution is defining path using event builder option in the car.
     
    Regarding Cone in obstacles in road builder, I have not seen such an option in road builder. May be the cones were CAD files imported for animation purpose.
     
    Hope this helps.
     
    Regards
     
  • 0
    1. Well, yes, I forgot to mention that this model is built by ourselves, not from acar_shared database. And also, I didn't mention which conditions I had used for this maneuver (velocity 30 km/h, steering angle 180 deg, initial and maximum frequency 1.042 Hz). But, anyway, this lateral displacement is okay, because in this maneuver steering angle is controlled, not the path or lateral displacement. There is the same issue as, for example, in Single Lane Change maneuver, where you can describe a lane change as 1-step sine signal for steering wheel, and after this signal ends the model doesn't ride in direction parallel to the start. And this is not an issue of a model, but an issue for vehicles at all. The picture attached shows lateral displacement of Testarossa during side-swept steer maneuver (velocity 100 km/h, angle 15 deg, frequency 1 Hz), and the same thing happens there.
    2. Hmm, that's a good idea. Don't know why I didn't think about it.
    3. Yes, it's another good solution, but not so easy to do.
     
    Regarding Cones: probably you are right, but I'd like to have an opinion from MSC guys if they can confirm your suggestion.
     
    Thank you for your help a lot.
    Attached Files (1)
  • 0
    It all depends if you want an open or closed loop driving event.
    For closed loop, use the File Driven Events. Look at the file:
    mdids://acar_shared/driver_controls.tbl/iso_lane_change_dcf.xml
    as an example.
    This file will call
    <acar_shared>/driver_data.tbl/iso_lane_change.dcd
    for the driver path.
    But using this method involve the dynamics of the driver.
     
    If you use a open loop, the method to use is still File Driven Events.
    But instead of a closed loop steering, you would want it to be Open, with Control Type = function.
    You could use a function of the style
    180d*sin(2*pi*time/2)*step(time,1.5,0,2.5,1)*step(time,7.5,1,8.5,0)
    and tune the start and end times of the step functions to make it come out straight.
     
  • 0
    I think you'll find if you measure yourself driving a slalom, the first cycle is not a sine wave but rather a cosine step. You can explore these ideas with an Excel spreadsheet and a geometric vehicle model before wasting a lot of time solving inside ADAMS. When you figure out the geometric input you can put it into ADAMS.
     
    Alternatively, and particularly if your manoeuvre is in the non-linear region, you will need closed-loop control as Jesper says.
     
    Plan C is to ignore the compass heading error in the model and only analyse the "developed" part of the slalom.
     
    Only you can decide which of these solutions suit you because only you know what is the problem to which this model is the solution.
  • 0
    Thank you, Jesper and Damian, I've already started working with the path description in Excel and File Driven Events, and it really works.
  • 0
    Viktor, it's not a breach of protocol if, once you have learned something, you answer your own question on here. That way other people can learn from your learning!
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement