yade-users team mailing list archive

[Fu zuoguang <question295521@xxxxxxxxxxxxxxxxxxxxx>]

Question #295521 on Yade changed:
https://answers.launchpad.net/yade/+question/295521

    Status: Answered => Open

Fu zuoguang is still having a problem:
Dear Jan，

Thanks for your suggestion and e.g. I am afraid that I had a wrong
understanding of contact normal analysis and plotting.

As my original standpoint, all the contact normal data obtained by
(i.geom.normal) should be uniformly located in the circle edge (Fig.1)
in Cartesian system.

Fig.1 dir：http://tinypic.com/view.php?pic=66gfmf&s=9

They can be used to represent the unit vector with the start point O. 
So, in my test codes uploaded yesterday, I used (++), (-+), (--) and (+-) to define respectively 4 quadrants and considered that the number of data in each group may almost the same in isotropic condition. But I was wrong, they are not the same at all. As reported by many tested cases in Yade, the number of data in 3rd quadrant is the largest and that in 4th quadrant is larger than the others. I think it is caused by Yade calculation system, and another DEM platform may have the same case.

Your opinion is very useful as the solution. You skillfully transformed
all the vectors which are originally in 1st and 2nd quadrant,
respectively to 3 and 4 quadrant BY (angle+pi). As shown in Fig.1,
vector 1 is to be equivalent to 3 and 2 to 4. Thus, all the thetas are
just in (0, pi), which can perfectly meet your point as

‘plotting normal is meaningful only for values (0,pi)’.

And your dealing with 
'quadrant_1_num+quadrant_3_num= quadrant_2_num+quadrant_4_num' is reasonable.

I hope that I had no wrong in getting your points. But I had not seen
all the details above in previous papers and simultaneously I am afraid
that this method may get fabric tensor calculation into trouble.

Did you have some related papers which have recorded these details? If
necessary, please tell me that.

cheers
zuoguang

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.