Message Boards Message Boards

0
|
7077 Views
|
5 Replies
|
2 Total Likes
View groups...
Share
Share this post:
GROUPS:

New User - Creating mechanical mechanism model to investigate forces?

Posted 9 years ago

Hello,

I am a new user working to learn System Modeler. I've been modeling simple mechanisms to see if I can obtain outputs that match my expectations.

I'm now stuck on a pretty basic issue. I created a model of a rectangular body with mass that swings on a pivot. A prismatic joint connected to the rectangular mass represents a hydraulic cylinder. A pictorial depiction of the simple mechanism is shown below: enter image description here

The rectangular mass's rotation is now driven by a combiTimeTable with an Accelerate Rotational Source. For now, let's say that the resulting motion profiles (acceleration, velocity, and position) of the rectangular mass are exactly what we want. With this simple model, I'm looking to determine the hydraulic cylinder force profile (over the simulation time span) that would accelerate and decelerate the rectangular mass through the motion prescribed by the combiTimeTable. Right now, when I plot the prismatic joint's force (f), it's value is zero over the entire model time span. Understanding that System Modeler is acausal, I was hoping that I could find the force through a prismatic joint by defining the motion of a driven, connected part. If this can be done, I'm making a mistake somewhere along the line.

If anyone can lend me hand with this, the help will be greatly appreciated. I've attached the model and the combiTimeTable file to this post. Thanks.

Dave

Attachments:
POSTED BY: Dave F
5 Replies
Posted 9 years ago

Johan, I finally had a moment to break away from the chaos as work to play with System Modeler. Yes, I see now how you were driving the model in your earlier post.

You and your co-worked were a huge help. Thank you! The last model you posted perfectly nails what I was trying to accomplish. I would never have found the inverse block constraint on my own.

Once again, you got me past a roadblock. Thank you!

-dave

POSTED BY: Dave F
Posted 9 years ago

Dave, In my previous answer I changed the model to only follow a linear position change, so it's not the same profile and not the same model as yours. Hopefully it makes sense when you download the model.

A co-worker suggested using an inverse block constraint. This constraint can be used to find the force needed to get the wanted acceleration.

Model with inverse block constraint

Running this model results in the following:

inverseModeling Prismatic SystemModeler

More material about learning to use the inverse block constraint can be found in this example. That example can be downloaded and played with. Also attaching the model I the pasted the diagram from.

Attachments:
POSTED BY: Johan Rhodin
Posted 9 years ago

Hello Johan,

Thank you, once again, for helping me. I am traveling this weekend and won't be back in front of my desktop machine with System Modeler until Monday. I will play with this and look at your model then.

I did think about the solution you suggested, but was hoping to find a way to get the force on the prismatic joint without having to re-run the model. I'm wondering if this approach will be cumbersome in the future when I'm looking at more complex mechanisms, want to run many "what if" scenarios, or use Mathematica with System Modeler to optimize some aspect of the system.

I tried to use the prismatic flanges with a parallel translational force input (hoping that I could actually use it as an output), but got an error message. I wonder if there's another object that I might use in parallel with prismatic flanges to get the force directly from a model "driven" at the resolute. I'll experiment more Monday with this.

I'll look at your model when I have access to System Modeler, but right now I'm wondering about the all positive values of force in the Revolute Prismatic plot. The force in the prismatic should be compressive (negative) as the mass is accelerated and then tensile (positive) as the mass decelerates. Later, when friction is added to the model, the forces will be modified. With the present frictionless model, I'm expecting to see a sign change in the prismatic force since it starts off pushing on the mass to accelerate it and then finishes by pulling on the mass to slow it down.

Again many thanks for helping me get my arms around modeling within the tool. Your generosity is very much appreciated.

-Dave

POSTED BY: Dave F
Posted 9 years ago

Dave,

The reason you don't see any force in the prismatic joint is because there's nothing attached to the driving flanges. The documentation string for the force in the prismatic joint hints at this: "Actuation force in direction of joint axis". So it's only the actuation force that will be determined there. Maybe the easiest way to understand this is to look at this example: SpringMassSystem.

POSTED BY: Johan Rhodin
Posted 9 years ago

To expand on my answer, if you want to find the force necessary to drive the motion from the prismatic while still replicating the behavior

  1. Find the position/speed/acceleration necessary in the prismatic joint. This can be done by "driving" with the revolute.
  2. "Drive" the simulation from the prismatic (with for instance the translational position), while removing the driving force in the revolute.
  3. Read the force necessary for the motion from the prismatic joints variable f.

Attaching a model where I've inserted a simple linear position. enter image description here

Attachments:
POSTED BY: Johan Rhodin
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Reply Preview
Attachments
Remove
or Discard

Group Abstract Group Abstract