[MBDyn-users] Moving pendulum
Andrea Zanoni
a.zanoni.mbdyn at gmail.com
Thu Nov 2 10:34:16 CET 2017
Hi Andrea,
can I ask you what is the purpose of your simulation? I guess you want
to solve an inverse dynamics problem, since you are trying to prescribe
the motion of the pendulum.
If that's the case, be aware that MBDyn is fully able to solve inverse
dynamics problems, but the documentation about this particular problem
type is still lacking... So feel free to ask here everything you might
need. One thing though is certain: if you want to solve a problem of
this kind, you have to integrate the acceleration yourself to get the
angular velocity and the angular position: the complete kinematics of
the pendulum must be prescribed.
If you just want to impose the angular acceleration of the pendulum and
let MBDyn integrate the path for you, use an "angular acceleration" joint.
Also, I'd like to point out that
- revolute hinge joints support friction (see section 8.12.38 of the
input manual);
- a nonlinear damping couple can be applied using the deformable axial
joint (8.12.11), to which a nonlinear viscous constitutive law can be
attached (section 2.9).
For very "strange" constitutive laws, you can compile MBDyn with GiNaC
support (www.ginac.de) and define your own symbolic constitutive law.
Again if you want support into how to do that, just ask here.
Cheers
Andrea
Il 31/10/2017 08:45, Louis Gagnon ha scritto:
> Hi Andrea and welcome to the list,
>
> On 30/10/2017 23:16, Andrea Palazzi wrote:
>> Hi,
>>
>> I just realized that my previous message was in html, so here it is
>> again...
> This was not an issue, both your messages arrived.
>>
>> I'm new to mbdyn, i'm trying to use it to model what basically is a
>> pendulum moving on the xz plane.
>> The pendulum moves along a defined path with a variable acceleration;
>> i also need to model the pin friction (can be constant) and a damping
>> device (damping nonlinear to rotational speed).
> You want to impose an acceleration or want to let gravity do its job
> and nothing else?
>>
>> I'm studying the documentation an so far I've modeled the simple
>> pendulum based on the tutorial file (a body and a reference), but
>> this doesn't seems the right path to get the pendulum moving.
> Have a look at the joint elements section, you have many options
>>
>> Instead a total pin joint with two scalar functions for x(t) and z(t)
>> seems more appropriate: is that right or there are a better way to
>> simulate this?
> For a simple problem like this any method that works for you is fine,
> I would not worry about "better" ways
>>
>> Is there a way to specify the acceleration (or even better the jerk)
>> and let mbdyn convert it to a position?
> Not sure what you mean
>>
>>
>> Any hint on how to model the friction?
> If it's constant then just apply a constant couple at the position of
> the joint
>>
>> Bye
>> Andrea
> Good luck with your models
> -Louis
>> _______________________________________________
>> MBDyn-users mailing list
>> MBDyn-users at mbdyn.org
>> https://mail.mbdyn.org/cgi-bin/mailman/listinfo/mbdyn-users
>>
>
--
---
Andrea Zanoni
Dipartimento di Scienze e Tecnologie Aerospaziali
Politecnico di Milano
Ph. (+39) 02 2399 8035
More information about the MBDyn-users
mailing list