the pic doesn't show clearly, you may click it to get a better view in a seperated window
Wow, that is pretty complex.
Can you get the file into an iges, step or a parasolid format?
If so just open it into a SW session and then you will have youe surfaces
to work with.
Thanks, Scott for your reply!
I read a little bit about iges format, I can find matlab functions that can make iges format and NURBS. But, I think the iges file generated in matlab is based on mesh (as shown in the pic, the matlab discretize the cylinders/shperes to pieces of planes).
I am thinking if I can just "draw" those cylinders and shperes in SW directly. Ideally I can import a batch file of commends, each of which creates a single cylinder or sphere as a vector diagram, instead of a discretization form (meshs or points). If this idea works, I can convert my data file into a batch file in matlab.
So, I may need to know
1, the commend syntax of SW drawing a cylinder and a shphere
2, the procedure of importing the batch file of commends
3, whether or not SW can idenfiy the intersections of all objects, in other word, whether or not SW can "remove" the parts inside objects and leave the surfaces there.
Then I could mesh it and caculate the surface area in SW.
Theoretically I could slove the intersections of two cylinders intersecting at arbitary angle and sitting at arbitary position, and also the intersections of a cylinder and a sphere, then mesh/discretize it for the surface area. But it may take lots of time to program this in matlab. Since this is a CAD thing, I guess SW could be able to do it, but I am not familar with any CAD software at all.
Sounds like you are trying to get into the deeper depths of Solidworks then
I have knowledge for. Sound like more of an API thing which is more of the programing
aspect of it.
As far as Mathcad, I am just learning that program in my new job and really new at it.
But I have seen threads on this forum speaking to compatibility between Mathcad and Solidworks.
I will research and try and find some for you.
As far as the batch, that is similar to macros which Solidworks runs.
I am doing my best to try and help you, but this may be a bit out of my league.
couldnt you just estimate it fairly well based on the number of edges x the area of 1 cylinder?
sounds a little over serious to be so exacting...
Neil, you are right, I have used this approximation for the preliminary simulation tests. When the cylinder radius is small, the overlap is not a big issue, but when the radius is bigger, the approximation doesn't hold. Besides, I am doing this for a technical paper and my program will be released in the future, so the approaches I used are supposed to be mathmatically convincing.
as an approximation how about just making one node with half long cylinders and multiple by no. of nodes.
heres a quick doodle of that type of thing that allows you to use the measure tool to find the surface area.
2 configs to use one with sphere at node and another with just cylinders.
seems like the difference between assuming only cylinders vs spheres is likely less than 1%
use the dimensions of sketch4,7,8 to suit your sizes
Neil, this looks great! But I think it may take me a while to understand it, I have to start with the GUI of SW first...
I tried Rhinoceros which I am a bit familar with, but that is evaluation version.
what I am doing is to draw cylinders as solids around lines (direction and length can be determined by 2 pts), and draw spheres as solids based its center and radius, and those info can be provided my fibers.dat. Then union all solids as one single solid, but I got difficult with unioning the 3rd cylinder (as shown below left, the one cylinder is not connected)
I also opened your file in Rhinoceros, it is very good that those are surfaces (pardon me, I almost know nothing about CAD).
But why not take cylinders intersection into account? I suppose it should be a few extra clicks in SW. And why take half of the cylinder?
When the fibers are thin, the approximation is good, but when fibers radius is bigger, all the overlaps have to be taken into account.
I think the difficulties to me are to draw surfaces (as you have shown, which I will learn), and to write a script to read my data file.
attached file is corresponding to the thiner fibers
fibers_50^3_r=1.dat.dat.zip 6.8 KB
I am sorry I dont have alot of time to spend on this.
The idea was to give you an idealised situation with one typical node to evaluate for nett area changes.
You would assume the whole volume can be modelled from considering the variation of one node x the number of nodes.
Attached is another 'improved' SW file and an IGS for Rhino. The sponge texture is just for fun
I included a face fillet between the cylinders and the sphere - perhaps this more closely resembles the material in reality(?)
The construction is a mess now but it gives you something to play with.
To find the surface area first switch to the 'one node' configuration.
Use the measure tool to select one face area and RMB and select tangency.
The displayed area is 2x the outside area = inside +outside.
Change the dimensions of your geometry and run again.
Possibly this can be automated by macro and excel table.
I have read integration of MathCAD to SW and Matlab to SW. You are right, this is more of an API thing, I have tried drawing a cylinder and a sphere and then unioning them, and I could get the surface area as one of the mass properties. I think I can start with macros.
If I am not mistaken, in your pictures the faces of interest are having their own color.
Why don't you select them by color, using either the Display Manager or Power Select?
For more details - CLICK HERE
After all the faces are selected use the Measure command in the Tools Menu.
Can you post the file here?
I think I should learn a little bit more about SW basics before I can go that far
my data file is attched in my post above, but not in any SW format, just some points coordinates