Learning Forum

[m.caragiuli]

To better explain the question consider a rubber band between two pins each one belonging to a body. Look at the attached picture

rubber band equivalent force.png

[Rameez_ul_Haq]

,if you know the forces acting on the ends of the rubber band, then using a force directly at the locations where pins are connected onto the two bodies will be okay. If the heights of the pin is considerable according to your thinking (I would had compared the height of the pin with the size of the bodies), then using a remote force is going to be true. I would try using the force instead of remote force as much as possible since using remote force increases the constraint equations within the analysis, making it take longer time to solve than if only force is being used.
But keep in mind that if you are using static structural module, you need to apply boundary conditions somewhere within the model to avoid rigid body motion. Thats up to your judgement where to apply these boundary conditions.
You can calculate the force acting on the rubber band if you know how much the band has stretched in reality when attached between these two bodies, then refer to the stress-strain or force-displacement graph, find the force from the graph where the displacement is half of what you have noted in reality.

[m.caragiuli]

Hi first of all thanks for your answer.
Actually, the geometric model does not include the pins. They should be modelled and unified to the bodies, that's why I thoght to apply a remote force since it allows to specify the location for the scoped region. In my case, the location is a face of the body (where theoretically should be the pin) and the point given by the coordinates of the pins.
Regarding the boundary condition, by looking at the image, the top face of the upper body was constrained by a fixed support. Since the two bodies are in contact that constrain should be enough to avoid rigid body motion.
My doubt was also about the type of contact. The two bodies could slide on each other, but not separate. So since the action of the rubberband is to avoid separation maybe it could worth replacing the contact by a frictionless one, that instead allows separation, otherwise the effects of the rubberband could not be visible. Don't you think?
Thanks!

[Rameez_ul_Haq]

,remote forces are usually used when two bodies are welded or if you are cutting a body from somewhere and then putting a part of it into your model (there might be several reasons for doing this, for example the part which you are not including is not experiencing too much stress according to your judgement even before conducting the analysis, this part might not be the critical region of study for stresses/strains/displacements, etc).
Assume you are just attaching the rubber band at the pins and pins are bolted or bonded onto the bodies. However, the connection between the pin and rubber band is not a weld neither is a continous body which you can cut. What I am trying to say here is that if the rubber band for example is pulled perpendicularly (w.r.t to its length direction) then the bodies won't be experiencing any internal forces or stresses, however if you try to model this case as a remote force, the two bodies in Ansys will. If the band was also welded onto the pins and then you try to move it up, then and only then the two bodies will also experience the internal forces and stresses. (Well, I am here assuming that the pins are straight without any hook at the top). This is when you can use a remote force in Ansys. Moreover, even if you weld the bands onto the pins, and then try to pull it, the band is a hyperelastic material, meaning that it will deform alot before it starts transferring any forces/moments onto the bodies. If you try to model this case inside Ansys using remote forces, it will not be correct since the remote force doesn't know how much the band has already deflected before it starts transferring forces and moment onto the bodies. These two are separate things and will result in different answers i.e. you will observe a difference in results in reality and in Ansys.

[Rameez_ul_Haq]

,if I was in your place, I would had used the bottom face of the beneath body as the fixed support. Because by common sense the bottom body must be lying on something like table. And also, you are right that the constrain by contact should be enough to avoid the rigid body motion of the other body where a support is not applied (given that you are not applying a frictionless contact since that can still result in a rigid body motion of this body).
Regarding the contact, I would go for bonded since I already know that the two bodies are coming into contact with each other, and not going out of contact. If the latter was also a possibility, then I would had chosen frictional contact between the faces in contact. But you should keep this in mind that my choice for bonded is because it is a linear contact (I hope you are already aware of what linear and non-linear contact is and how it can affect the time taken for the analysis to complete) while frictional contact is non-linear. Plus, the area of transfer of forces between these bodies will remain constant in a bonded contact which is not actually the case in reality. For a frictional contact, the area of transfer of forces can change over time. This won't affect the stress results some distance away from the contact (making use of St. Venant's principal here) but if you are interested in the stresses locally near the contact between these two bodies, then the results will be different in reality and in the case if you choose bonded. Frictional contact can give you acceptable stress results near this contact which might be comparable to the reality.
I won't choose no separation since the bodies are still slide over each other and hence the upper body can perform a rigid body motion (similar reason to why I won't be using a frictionless contact).
Please click on 'Yes, this answered my question' if you found the answers useful.

[m.caragiuli]

thanks again. The contact point is clear. But there is an issue regarding the force.
Let me be clearer about the simulation. Maybe I didn't say that I will not include the rubber band in the geometry. The geometric model includes only the two bodies. There are no pins, no rubber band. Thus, the idea is to approximate the behavior of the rubber band (that is inserted between the two pins) by applying a couple of equal and opposite internal forces (remote forces) acting on the faces where should be (but are not present) the pins. IIf I use a force load ( and not remote force) I cannot define the point to scope the force (that is the global coordinate of the pin) since the force can be scoped to a geometry and a named selection, but how can I scope it to the point corresponding to the x, y, z coordinates of the pin? If I insert a remote point it would require a remote force, thus is there a way to scope the force to a point that is not included in the model?

[Rameez_ul_Haq]

,I don't undestand what do you mean by, "the idea is to approximate the behavior of the rubber band". There is something whose behavior you want to approximate, but you are not including that in your model. Does it make sense near you? You are only including two bodies in your model, so the solver is only going to show you results for those two bodies. You can retrieve deformations, stresses, energy etc, anything that you want, but only for those two bodies. Anything excluded from the model cannot be predicted.
But still if you want to proceed with what you are doing, then I think I have already answered you. If, in the reality, the height of the pin is quite big, then you can go for a remote force. If not, then just directly use the force on a local small area on the faces of the bodies (those small local areas where the pins are supposed to be bolted/welded). And yes, you cannot change the location of force if you just choose force instead remote force.

[m.caragiuli]

I know the magnitude of the force associated with the band at its maximum length, that is corresponding to the distance between the two bodies. So the model depicted in the picture represents the ultimate configuration in which the ban is at its max stretching, thus the force is maximum. So, according to my opinion, there is no need for modelling the band since I want to evaluate the transfer of the load to the bodies to detect the stress. If I am wrong please let me understand why.
I appreciate your response and your courtesy.

[Rameez_ul_Haq]

,well thank you for your comment. If you are interested in detecting the stresses on the two bodies only, then what you are saying is absolutely correct and makes sense.
Please click on ''Yes'' if these replies have answered your queries :)

[peteroznewman]

The attachment from your original post is inserted here for ANSYS staff to see, who are not permitted to open attachments.
A rubber band is like a spring, when stretched to come distance it generates a force F.
When you want to represent the effect of the force F, each body sees the force F. It is wrong to divide that by 2 just because you are cutting out the band and replacing it with a force.
I suggest you represent the rubber band with a spring element.

[m.caragiuli]

Hi Peteroznewman can you make it clearer the reason why the force should be F and not half per side?
Thanks

[peteroznewman]

Say you have a 100 cm long rubber band. Anchor one end to a pin and put a force sensor on the free end.
Stretch the free end by 50 cm (strain = 0.5) and find the force is F.
Cut the rubber band down to 10 cm and stretch it by 5 cm (strain = 0.5), the force is still F.
Cut the rubber band down to 1 cm and stretch it by 0.5 cm (strain = 0.5), the force is still F.
If you replace the pin-anchored end with a force sensor, it will also measure F.