[MBDyn-users] Deformable joint

Pierangelo Masarati masarati at aero.polimi.it
Thu May 7 11:10:23 CEST 2009


Doug,

thanks for the contribution.  It's under review.

Cheers, p.

Doug Baldwin wrote:
> I've attached a working copy of a reformulated vehj3 "deformable joint", and
> have a question regarding this code.  What I've done so far is to convert
> the vehj2 "deformable displacement joint" from using a 3D Constitutive to
> using a 6D Constitutive.  I'd appreciate an independent look at this code
> before I add in the vehj "deformable hinge" code.  The test simulations I've
> run comparing this new vehj3 with the original vehj2 "deformable
> displacement joint" appear to produce similar results.
> 
> My question is as follows.  Most of the code was indeed cut and paste.
> However, there are just a few lines that I'm not certain of, which I've
> marked with CHECKME.  These lines seem reasonable to me, and the simulation
> results seem to be good.  But still, I'd like to know that these are good
> formulations.
> 
> The attached zip file includes joint.cc, vehj3.cc, and vehj.h, and was
> compiled successfully with mbdyn v1.3.6 .
> 
> An example of the lines I'm referring to are as follows:
> 
> 	tilde_kPrime = Vec6(tilde_dPrime, RotManip::VecRot(R1hT*R2h));
> 	ConstitutiveLaw6DOwner::Update(Zero6, tilde_kPrime);
> 
> Doug
> 
>>> What can the user community
>>> (and I) do to help complete the implementation of the "deformable
> joint"?
>> It's basically a matter of cut'n'paste to merge the deformable
>> displacement joint and the deformable hinge residual and Jacobian matrix
>> contributions into those of the deformable joint.  This has been done
>> for the elastic case (dependence on linear and angular strains) but not
>> yet for the viscoelastic case (dependence also on strain rates).  The
>> reason is that the simpler elements underwent significant refactoring
>> when their "invariant" form was formulated, so the deformable joint was
>> deferred.
>>
>> Cheers, p.
> 
> 




More information about the Mbdyn-users mailing list