|
|
Hi,
I'm posting this primarily to ask the advice of Jerome Berger, the author or
the Cloth Simulation prog, although I would welcome anyone elses input.
I have posted an anim ( ClothSimTest.mpg ) in p.b.a which illustrates the
problems I am having with Cloth Sim.
Firstly, here is my code:
//Start code
global_settings {
step 0.2
viscosity 0.1
}
cloth {
x_size 20
y_size 25
strenght 1
dist 2
hnoise 0.01
vnoise 0.01
elongation 0.0001
shear 2
damping 0.001
intersections true
position { 40.000000*x,10*z,40*y }
read_file "zammer.raw"
attach { 1,0,<40.000000,5,10> }
attach { 1,1,<40.000000,45,10> }
attach { 1,0.5,<40.000000,25,10> }
}
ground {
-30, 0.5
}
cylinder { <40.000000,0,-31>,<40.000000,0,10>,1,0.25}
cylinder { <40.000000,50,-31>,<40.000000,50,10>,1,0.25}
sphere { <20,25,-31>, 20, 0 }
//End code
I am initially running the simulation with -s1000, then with -s50 per frame.
I am using Pov to write a new .cl file each frame, which updates the vectors
I wish to animate. This method seems to work.
However, I the problems I have come across are:
1/ The cloth always seems to resemble a thin rubber membrane. What settings
should I use to model a heavy curtain?
2/ The cloth seems to 'vascilate' (correct word?).. to flutter far too
much.
3/ The main problem (especially in my anim) is the fact that, rather than
slide over the sphere, the cloth adheres to the surface of the sphere, and
then stretches. It seems to make no difference whether I set the friction
variable to 0 or 1.
4/ When I used translate in an anim (placing it last in the cloth
statement), the cloth moved about from frame to frame quite unpredictably.
Is translate supported within animation? I'm having problems also with the
way the position option works.. but this is more to do with my not really
understanding the syntax required.
Basically, I am unclear as to what the various options do. Maybe I am
expecting the prog to work in a way in which it was not designed for?
Any help would be greatly appreciated.
Thanks,
--
Andy Cocker
---------------------------------------------------------------
listen to my music at:
www.mp3.com/lunarland
---------------------------------------------------------------
'I spilled spot remover on my dog. He's gone now. '
'I went to a restaurant that serves "breakfast at any time."
So I ordered french toast during the Renaissance. '
- Steven Wright.
---------------------------------------------------------------
Post a reply to this message
|
|
|
|
Andy Cocker wrote:
> However, I the problems I have come across are:
>
> 1/ The cloth always seems to resemble a thin rubber membrane. What
> settings should I use to model a heavy curtain?
>
I'm not sure, since I haven't made enough tests to find parameters that
would do it. Using a shear of 2 seems a bit too much, especially since
you've kept the default value of 0.05 for the stiffness (shear has a
side effect on the apparent stiffness unfortunately)
> 2/ The cloth seems to 'vascilate' (correct word?).. to flutter far
> too much.
>
I'm not sure that I see what you mean in the animation you posted, but
maybe raising viscosity (or damping) will fix that.
> 3/ The main problem (especially in my anim) is the fact that,
> rather than slide over the sphere, the cloth adheres to the surface
> of the sphere, and then stretches. It seems to make no difference
> whether I set the friction variable to 0 or 1.
Sounds like a bug. Could you mail me the complete code (including pov
script and ini) so that I can run tests please? (mail is better than
news since I can access it every day whereas I can only read news on
fridays for the time being)
>
> 4/ When I used translate in an anim (placing it last in the cloth
statement),
> the cloth moved about from frame to frame quite unpredictably. Is
> translate supported within animation? I'm having problems also with
> the way the position option works.. but this is more to do with my
> not really understanding the syntax required.
>
This is normal: when you run the program for the first time, the cloth is
positionned in its default position (or in the position you specified
with the "position" statement), then translated and rotated according to
your specs. Then the simulation is run and the resulting geometry is
saved to the raw file in its current position (including the effect of
the translations and rotations). Therefore when you re-read the raw
file, the cloth has already been translated and if you add a new
tranlate statement, it is applied again.
In other words, the easier way to use translate with animation is to only
use it in the first frame since it will be taken into account in the raw
file anyway.
Jerome
--
* Abandon the search for truth, * mailto:ber### [at] inamecom
* Settle for a good fantasy. * http://www.enst.fr/~jberger
*********************************
Post a reply to this message
|
|
|
|
>> 3/ The main problem (especially in my anim) is the fact that, rather
>> than slide over the sphere, the cloth adheres to the surface of
>> the sphere, and then stretches. It seems to make no difference whether
>> I set the friction variable to 0 or 1.
> Sounds like a bug. Could you mail me the complete code (including pov
> script and ini) so that I can run tests please? (mail is better
> than news since I can access it every day whereas I can only read
> news on fridays for the time being)
>
It's more or less fixed in the version that's currently on my hard drive.
The "or less" part is the reason why I haven't released it yet, but it
should be available next week.
Jerome
--
* Abandon the search for truth, * mailto:ber### [at] inamecom
* Settle for a good fantasy. * http://www.enst.fr/~jberger
*********************************
Post a reply to this message
|
|