Caito wrote:Hi guys, big revival.. but related.
Today I modelled the quarter car, similar to this one:
The basic difference is that I also modelled a damper between the tire and the road.
Usually people don't model tyre damping at this stage, because, as DaveW just said, it is quite small compared to the proper damper, meaning there is no significant difference in the response of the whole system. From the values I usually see, the damping coefficient from the tyre is between 1000 to 10000 times less than the damper, while the stiffness of the tyre is about 10 times larger than the suspension. But as you already put it, leave it there.
Caito wrote:I would like to know a little bit about the limitations of the models. Leaving the fact that is just a quarter of a car. What I find disturbing is that the spring labeled Kt is attached to the ground and to the unsprung mass. The problem with this is that you can never have your car leave the ground.
Moreover, If you get a disturbance with a BIG(and negative) dxr/dt the Kt spring tries to instanstly stretch a lot, as that would never happen it pulls the sprung mass violently. So the model should be subject to a road that does not change very abruptly.
What bothers me is that when the road "goes down" the spring pulls the car towards the road which, in reality, it's not that way. What really happens is that the sprung mass is pushing the wheel to the ground, the wheel is not being pulled by the road.
To sum up, with positive dxr/dt the system looks pretty real to me. It fails to represent the reality with a negative dxr/dt.
In reality I don't know if this represents a problem to the model or not. Anyway, do you think of any other similar problem? Does anybody know which could be the solution to the problem presented?
Bye bye!
Caito.-
You are quite right at your observations, and these problems are because you are considering an ideal spring, that can expand on and on, and that it is 'fixed' at the ground.
If you consider that your spring can expand up to a certain limit (a bit more realistic), than you need to check if the displacement of the tyre is greater than this value. If it is, your car will be airbone. Then it depends on what sort of simulations you want to perform. If you want to continue the simulation after the car is in the air, from this point you have to consider a free body flying due to its (at this point) initial vertical velocity and let gravity bring it back. For landing, you need to check when the distance between the tyre and ground equals the length of the spring.
You could do this for the compression as well. Imagine that you have a big disturbance up, the tyre (and suspension) as you modelled will compress up to unreallistic values. I first noticed this on one simulation when the amplitude of the oscillations of the tyre was larger than the height of the car... You need to again stipulate one limit, and check what happens when you reach this limit.
The problem of the spring pulling the car back (that bothered me a lot as well) is similar, although there is one thing. The spring as you modelled (and as we all do) is massless, so has nor inertia. So yes, if it is attached to the road, it will move intantly at any disturbance. You need to model it as a 'unidirectional' spring. If the road presses the tyre up, the tyre will react. But if the road pulls the tyre down, it doesn't do anything. It is not glued to the ground. It is the opposite problem of a cable: if you try to stretch it, it reacts, if you try to compress it, it doesn't. The problem is that the 'compressed' and 'extended' situations are related the unstretched length of the spring, not the equilibrium position as in your diagram. So intead of using, say, xt = 0 to define your transition, you need to use (xt - xr) and compare to the unstretched length. And if you couple this to the extension limit of the spring, if the road goes down too much, you are airbone for a few moments.
Although these modifications can be made, most simulations I see for simpler models like this don't take these issues into consideration. You will see that you can model quite a good range of parameters until you get to extreme cases of loosing contact with the road. For passenger cars at least.
I would suggest that you go now to half car models, where you can see pitch and roll (depending on which half of the car you get), before you go to bicyle model, which involves steering. You could also include some basic suspension geometry.
