Surfboolean Problem.
Anqi L
Qianqian Fang
hi Anqi
to make sure surfboolean works as expected, you must "reorient"
the surface triangles consistently (i.e., the node order must be
in counter-clock-wise order, in order words, the normal dir points
outward).
To correct surface element orientation, you need to call surfreorient() or meshcheckrepair(). Then, in the surfboolean call, you need to flip one of the surface's direction by swapping any two columns.
For example
[no1,fc1]=meshabox([0 0 0],[10 10 10],1); % a
box mesh
[no2,fc2]=meshasphere([10 10 10], 8, 0.4); % a
sphere mesh
[no1,fc1]=meshcheckrepair(no1,fc1); %
orient surface 1
[no2,fc2]=meshcheckrepair(no2,fc2);
% orient surface 2
[no3,fc3]=surfboolean(no1,fc1,'and',no2,fc2(:,[2 1
3])); % two surfs should have reverted direction
plotmesh(no3,fc3)
For simple compartment surface, the flipping in direction may not be necessary, but if you have multiple shells, this can produce different results.
hope this helps.
Qianqian
--
Regards,Anqi
You received this message because you are subscribed to the Google Groups "iso2mesh-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iso2mesh-user...@googlegroups.com.
To post to this group, send email to iso2mes...@googlegroups.com.
Visit this group at https://groups.google.com/group/iso2mesh-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/iso2mesh-users/d5275abb-d440-47ef-a2f2-7e9c71687c0e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Anqi L
在 2019年7月5日星期五 UTC+2下午6:04:37，Qianqian Fang写道：
On 7/5/19 10:50 AM, Anqi L wrote:
Hello,
I have a wired problem while using the function Surfboolean. For two intersecting mesh surface, while using the Surfboolean with option 'and', I am expecting the surface of the intersected region, but sometimes i obtain the mesh that belongs to option 'diff'. Or when i use option 'union', but I obtain the mesh for option 'first' etc.
It seems that sometimes, the option and the mesh generated is not coincide. However, this only happen occasionally.
Option 'and' is always correct (cause nothing needs to be removed from it), but I do not have the fourth column in newelem indicating which part is the element belongs to.
Thank you for helping me out! Also, thank the author for developing such nice meshing tool!
hi Anqi
to make sure surfboolean works as expected, you must "reorient" the surface triangles consistently (i.e., the node order must be in counter-clock-wise order, in order words, the normal dir points outward).
To correct surface element orientation, you need to call surfreorient() or meshcheckrepair(). Then, in the surfboolean call, you need to flip one of the surface's direction by swapping any two columns.
For example
[no1,fc1]=meshabox([0 0 0],[10 10 10],1); %a box mesh
[no2,fc2]=meshasphere([10 10 10], 8, 0.4); % a sphere mesh
[no1,fc1]=meshcheckrepair(no1,fc1); % orient surface 1
[no2,fc2]=meshcheckrepair(no2,fc2);% orient surface 2
[no3,fc3]=surfboolean(no1,fc1,'and',no2,fc2(:,[2 1 3]));% two surfs should have reverted direction
plotmesh(no3,fc3)
For simple compartment surface, the flipping in direction may not be necessary, but if you have multiple shells, this can produce different results.
hope this helps.
Qianqian
--
Regards,Anqi
You received this message because you are subscribed to the Google Groups "iso2mesh-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iso2mes...@googlegroups.com.
Anqi L
在 2019年7月5日星期五 UTC+2下午6:04:37，Qianqian Fang写道：
On 7/5/19 10:50 AM, Anqi L wrote:
Hello,
I have a wired problem while using the function Surfboolean. For two intersecting mesh surface, while using the Surfboolean with option 'and', I am expecting the surface of the intersected region, but sometimes i obtain the mesh that belongs to option 'diff'. Or when i use option 'union', but I obtain the mesh for option 'first' etc.
It seems that sometimes, the option and the mesh generated is not coincide. However, this only happen occasionally.
Option 'and' is always correct (cause nothing needs to be removed from it), but I do not have the fourth column in newelem indicating which part is the element belongs to.
Thank you for helping me out! Also, thank the author for developing such nice meshing tool!
hi Anqi
to make sure surfboolean works as expected, you must "reorient" the surface triangles consistently (i.e., the node order must be in counter-clock-wise order, in order words, the normal dir points outward).
To correct surface element orientation, you need to call surfreorient() or meshcheckrepair(). Then, in the surfboolean call, you need to flip one of the surface's direction by swapping any two columns.
For example
[no1,fc1]=meshabox([0 0 0],[10 10 10],1); %a box mesh
[no2,fc2]=meshasphere([10 10 10], 8, 0.4); % a sphere mesh
[no1,fc1]=meshcheckrepair(no1,fc1); % orient surface 1
[no2,fc2]=meshcheckrepair(no2,fc2);% orient surface 2
[no3,fc3]=surfboolean(no1,fc1,'and',no2,fc2(:,[2 1 3]));% two surfs should have reverted direction
plotmesh(no3,fc3)
For simple compartment surface, the flipping in direction may not be necessary, but if you have multiple shells, this can produce different results.
hope this helps.
Qianqian
--
Regards,Anqi
You received this message because you are subscribed to the Google Groups "iso2mesh-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iso2mes...@googlegroups.com.
Qianqian Fang
Hello Qianqian,
I have tired to correct the surface element orientation before calling function Surfboolean(), as in sample1.png two surface mesh have reverted orientation after I called meshcheckrepair() and inverted one of the mesh. But still when using surfboolean() with command "and" I obtained something in sample2. Is there any reason why it is happening?
hi Anqi
orientation/boolean does not usually work with open surfaces, but I am glad to see that cork is able to process (to some extent).
Because for an open source, there is no definite definition of "out" or "in", I am not entirely sure the expected behavior of the boolean operator.
Qianqian
To unsubscribe from this group and stop receiving emails from it, send an email to iso2mesh-user...@googlegroups.com.
To post to this group, send email to iso2mes...@googlegroups.com.
Visit this group at https://groups.google.com/group/iso2mesh-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/iso2mesh-users/8fc3e961-cd0e-4765-9b33-d354833a1125%40googlegroups.com.