The Excavator Hydraulic Jack Forces are Changing Direction in Simscape Multibody

I am trying to model a mechanism of excavators in the Simscape multibody.

I am using prismatic joint to simulate the functionality of hydraulic jacks with the following settings:

I motion the hydraulic jacks (using PS Ramp bock) and track the sensed "actuator force" parameter via a scope block.

So technically, I am doing an Inverse dynamic analysis to calculate the needed force on the hydraulic jacks and because during my simulation, all of the jacks (modeled with Prismatic joints) are getting opened, I am expecting that all the actuator forces be positive; however, the sign of force is changing over the simulation (!) which I believe it is wrong as I set to the motion of prismatic joints in one direction and the actuator force should be along this direction too. So I was wondering, why this is happening?!

[Yifeng Tang]

It's hard to tell without seeing the rest of the model. The way you actuating the joint and sensoring the force for inverse dynamic analysis looks OK to me.

One wild guess: is the gravity on and do all components have the correct mass and inertia? I'm wondering whether it's possible that at some point, the mechanism will move due to gravity and the joint force needs to slow it down trying to maintain the input position. Also, if the mechanism is decelerating, the inertia from all the masses may require a negtive force.

[Medalan]

Hi @Yifeng Tang

Thank you for the comment and also my apologies for the delay.

Actually, I think You somehow mentioned the right reason. I checked the model and I noticed one thing. In "inverse dynamics," the inputs are "q," "dq," and "ddq," and the derivation of "dq" and "ddq" is automatically performed in the "PS Ramp" block. So, in my case, the motion is constant velocity. In this scenario, the prismatic joint should sometimes apply a negative force to prevent increased speed.

I think this is the main reason I had negative force; However, it also shows that this strategy is just suitable for static analysis. For an accurate simulation, I have to use the hydraulic blocks and directly feed the prismatic joint with force, instead of motion.

[Yifeng Tang]

I'll summarize this conversation in the "Answer". Feel free to comment. Happy modeling.

Answer 1

[Yifeng Tang]

The way the joints are actuated and the joint forces are sensed for inverse dynamic analysis looks OK.

As a position signal is used as the actuation signal, it's 1st & 2nd order derivatives (velocity and acceleration) are also implied. In this particular case, the position changes linearly, so the velocity remains constant and the acceleration is zero. This requires that all forces sum to zero at the joint. As the mechanism moves, the sum of the gravity force may drive the mechanism to move on its own and the hydraulic force may be negative to enforce the zero acceleration we provided as actuation signal.