|
[Sponsors] |
August 16, 2011, 19:57 |
Importing Rhinoceros 3d model to SolidWorks
|
#1 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
The title of my this topic, may be different but here is the story:
I spent almost three days in trying to import my 3d model from Rhinoceros to SolidWorks Flow simulation for wind pressure analysis of my curved roof. First I experimented with the export extensions. I tried to export my model from Rhino to: .3ds, .obj .vrml .step .dwg .dxf .sat ... and all other available extensions. I even tried the trial plugin "3ds to SolidWorks" Nothing, no succes. Some of above mentioned extensions can be opened in the SolidWorks, some not. But even those that can be opened, can not be used in Flow Simulation, which is my goal (importing my 3d model from Rhino to SolidWorks for wind pressure analysis). Finally after today I saw that .igs export from Rhino gives best results. With it, I succeded to import my 3d Rhino model into Solid Works. Again a new headacke begins: As I read somewhere on this forum, Flow Simulation requires all entities to be solid, not surfaces. Because of that, I had to use the "Extrude" and "Thicken" commands to convert my surfaces to solids. But there comes the headacke: probably because those are not SolidWorks surfaces, but imported ones, I simply can not Extrude or Thicken them. Especially the two way curved ones. I am getting messages: "The surfaces could not be extruded/thickened" and similar. After couple of hours I succeeded in thickening all of them. It has something to do with the order by which I thicken them, and by the values (in milimeters). Sounds strange? Yes it does to me also. When I finally made it to convert all surfaces to solids, then I Created a new Flow simulation project, set general settings, initial mesh, surface goals, went to Solve->Run, and was finally excited that I will get my results, a new dissapointment emerged: Quote:
Again a new problem emerges - the pressure results in the Surface plots, can not be seen unless you click on those furfaces. But then the blue selection color is appearing as close as you zoom in the project. Take a look: Did anybody else have any similar experiences with export of Rhino to SolidWorks? Thank you P.S. I saw a tv show on National Geography about the designer from Maserati. He created the 3d model of new Maserati in 3d Studio Max, and then the Engineers did the air pressure analysis in SolidWorks Flow Simulation. So I guess they had to to export the 3d model from 3d Studio Max to SolidWorks. 3d studio max is not a NURBS modelling application, in contrary to Rhinoceros, which is. So if exporting from 3d studio max to SolidWorks is possible, then it has to be much easier to export from Rhino to SolidWorks, right? |
||
August 19, 2011, 07:25 |
|
#2 |
New Member
Vertex Wrangler
Join Date: Apr 2011
Location: Belgium
Posts: 25
Rep Power: 15 |
If you had to spend a few hours to get your model to thicken, it probably will not be a logical and clean solid. You can get lucky with SolidWorks and push it enough to give you a result, but it won't fly when processing further down the line inside of Flow Simulation.
Exporting from Rhino to SolidWorks should be pretty easy, it depends on the original Rhino model. If you got these two on the same PC, you can bring in .3dm files (native Rhino) without needing to go through .step or .iges |
|
August 19, 2011, 10:45 |
|
#3 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
Thank you for the reply Kevin.
I am waiting for somebody to reply for a couple of days. Thank you Rhino can not create real solids, just surfaces connected in a way that they look like solids (for example cube created in Rhino can not be solid, just 6 plain surfaces connected to each other, with an empty space in the middle of the cube). There for every 3d model imported from Rhino to SolidWorks, is consisted of surfaces. As I heard that SolidWorks Flow Simulation can work only with solids, this means I need to convert all my Rhino surfaces into solids. About importing .3dm files into SolidWorks: Believe me I already tried using both original .3dm and export .igs extensions. And .igs proved to be better - simply after opening my .3dm file, converting surfaces to solids (mostly by "Thicken" command), Setting up a geometry of my Flow simulation, setting up mesh, Goals and when I try to solve it (Flow Simulation->Solve->Run) I am getting the message: "There is no fluid volume in the project. Please check geometry or boundary conditions." This is not the case with .igs extensions, as I can do the Flow Simulation with no problems, and solve it. But it seems that the results might not be accurate? I guess that was what you tried to tell me - what did you mean by this: Quote:
|
||
August 20, 2011, 06:10 |
|
#4 |
New Member
Vertex Wrangler
Join Date: Apr 2011
Location: Belgium
Posts: 25
Rep Power: 15 |
Take this example, it's a valid solid in SolidWorks, but it's insane.
Trying to mesh this would be suicide: http://img.photobucket.com/albums/v2...filletcide.jpg Thickening can make small edges, knife edges, all kinds of nastiness. If you're not careful. I think maybe you didn't need thicken. Make a backup copy of your part. Then delete all the thickens, and try to thicken again, now see if you get a "make solid from enclosed volume" checkbox in the thicken feature. "Solid" in any modern solid modeler is a bit of a misnomer, a cube is SolidWorks is also just 6 surfaces glued together and considered closed and solid. But it isn't really. Want proof? Sketch a rectangle, extrude it to get a cube and then go to Features > Face > Delete Face (make sure the radio button in the feature is set to only delete) and pick a side. Now take a peek inside! |
|
August 20, 2011, 06:44 |
|
#5 |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
thank you for reply Kevin
I did not know this about solids in solidworks, thanks for the information. About the "make solid from enclosed volume" checkbox, I did not manage to find it. I do not have that option in my SolidWorks 2011, or maybe I am looking at the wrong place. Here is the screenshot of my Thicken command: |
|
August 20, 2011, 16:51 |
|
#6 |
New Member
Vertex Wrangler
Join Date: Apr 2011
Location: Belgium
Posts: 25
Rep Power: 15 |
No, it only appears when it finds a closed volume. Your geometry does not apply to that. However, you may have more luck with the following: Surfaces > Surface Knit and select all of the faces that are joined together (if it's a sheet or sorts I imagine all of them) and then do a thicken.
|
|
August 20, 2011, 19:34 |
|
#7 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
Thank you for the reply Kevin.
But the Insert->Surfaces->Knit did not worked on my .igs model. It did worked on my .3dm model, but I can not do the Flow Simulation on the .3dm model (I am getting a message: "There is no fluid volume in the project. Please check geometry or boundary conditions" when I try to run the Solver). It does not matter, thank you for the advice. What I am now more interested in, was your quote: Quote:
|
||
August 21, 2011, 06:30 |
|
#8 |
New Member
Vertex Wrangler
Join Date: Apr 2011
Location: Belgium
Posts: 25
Rep Power: 15 |
Refer to the screenshot I linked to earlier. See those little fillets? No way that is going to mesh in simulation!
|
|
August 21, 2011, 15:47 |
|
#9 |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
I see.
Even though I will not get an error message when I try to set up my initial mesh? And even though that fillet is not created in Rhino, but in Solidworks? |
|
August 29, 2011, 06:31 |
|
#10 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
Hi Kevin and George,
SW is not the best surface modeller I too have sometimes problems doing the simpliest things that are very easy for me in Catia V5 althought SW is very similar to Catia, it is simply not as powerful, hence the price difference It is also hard to tell what the issue is in your model and someone would have to have a look on them if it's really geometry issues or user errors in the project settings. I cannot tell how good you know your way in SW FlowSimulation. Ideally try SW support if they can help you find out what's wrong and why it doesn't work. For the imported model form 3D Studio Max, I never did try to import anything from 3D SM but often the steps in between are not shown in TV shows. I know 3D SM has polygon models and it's a pain to use that in SW as you get facetted models which are a horror for a CAD system to work with if you don't reconstruct your model from that which I believe was done to get a working model for SWFS. I know of Voxdale who laser scanned their race car and then had to get 3D surfaces from the point cloud and a 3D solid model at the end to run it on FloEFD for Creo (former FloEFD Pro). And about the filetted model, have you tried to run it in SWFS or is that not from you? I have seen many creazy models with filets and they all worked in FloEFD and SWFS. You might have these problems with standard CFD mesher but shouldn't have them with FloEFD or SWFS. Regards, Boris |
|
August 29, 2011, 07:55 |
|
#11 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
Thank you for the reply Boris
The model you see was exported from Rhino, and Rhino is a NURBS application, like SolidWorks, not Mesh application like 3d Studio Max. So from that point, I guess there would be no problems. Yes, I did tried to run it in Soli Works Flow Simulation. After hours and hours of testings: I discovered two strange things: First of all the analysis can not be done with the Rhino .3dm file. As when everything is set, domain, fluid, mesh, goals, and when you click on Run-> Solve the following message appears: Quote:
Another problem emerges then: Some surfaces can not be Thickened. For some strange reason, the Thickening of the some surfaces, depend on the order by which you thicken them. I discovered that you always need to Thicken the problematic surface first. Then the others. Boris are you advising me to use CATIA for this kind of stuff (wind impact on curved surfaces)? |
||
August 29, 2011, 08:13 |
|
#12 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
Hi George,
no, I'm not advising you to use Catia. This was just as an example of the different capabilities of the CAD systems. As I said, I cannot tell how well .3dm files work when importing in SW. As for the error message you received, it is hard to tell if the geometry is ok. I would also guess it is best to user a better CAD suited exchange format like STEP or IGES rather than .3dm. Although it is a NURBS model the import and export to a CAD system should be done by the best supported interfaces for both tools. For example (my personal opinion) the STEP imoport into Catia is really bad, I have the feeling it works better in SW or Pro/E. But even if you found the best suitable exchange format you might have to tweak the geometry a little to work properly, fill gaps that were created or bad surface joints etc. It can be connected in which order you thicken surfaces. As Kevin said, it is often good to morge all surfaces and then thicken them but in general thickening can also cause problems if the surfaces have for example a radius or that would vanish or intersect or things like that. I could imagine the bulky shape can cause a porblem when thickening, but it is really hard to tell from the images. |
|
August 29, 2011, 08:28 |
|
#13 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
Thank you for the reply Boris,
Do not worry I was not insulted, I just asked - if CATIA is a better application for this kind of job, then I would definitively do it in CATIA. .step did not worked in SW, as I am having problems with Thickening the surfaces. I aslo tried .3ds, .obj .vrml .step .dwg .dxf .sat ... and all other available extensions. And only the .igs gave me final result, none of the others. Can you clear up this a little bit please: Quote:
But what do you mean by radius that will vanish? |
||
August 29, 2011, 09:41 |
|
#14 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
I didn't take it as insulting, just wanted to say that it was not my intention to say you shall work with Catia.
For the radius that will vanis, imagine a "J" like line, a vertical line at the end is a partial cicle with another line pointing towards the vertical but not intersecting. The distance of the tip of the one line to the vertical one is 4mm and the radius of the circle is 3mm. Now extrude this to a surface and make an offset to this surface to the inner part between the two lines. First you will meet an intersection starting with 2mm but who cares, then at 3mm and larger offset you will undercut the circles radius and thefore this wont work. These problems appear in many freeform surfaces if not considered from the beginning. This is what I meant. |
|
August 29, 2011, 09:48 |
|
#15 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
I didn't take it as insulting. I don't want to tell you what programm you have to use and this was just as an example.
For the radius that will vanis, imagine a "J" like curve, you have a vertical line ending in a partial circle (for example 310°) and then again a line tangantial pointing towards the vertical line but not intersecting. The end of the one line is 2mm away from the vertical line and the circle has 3mm radius. Now extrude it to a surface and make an offset. When you make an offset 2mm and larger the surfaces of the two lines start intersecting, this is not a problem but as soon as you reach 3mm you undercut the circles radius and what will happen to the circle now? This is what is not working. I cannot tell if this is the case in your model but if not considered free form surfaces can cause such problems when creating an offset. |
|
August 29, 2011, 14:02 |
|
#16 | |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
Quote:
Any screenshots? If you think you already answered me a lot, I would understand. Thank you. |
||
August 30, 2011, 05:27 |
|
#17 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
Below a little example of a line. It works to make an offset to the outer region but to the inner region you can see it starts to intersect but the offset is still smaller than the radius so it is possible.
And here a surface where you can see it works for 2mm but above 3mm you'll get an error that the radius is undercut and not present anymore, so what shall it do now. This is not a fully defined status. The two red surfaces would work, but the circular surface does not. I hope this helps. |
|
August 30, 2011, 05:29 |
|
#18 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
Didn't work, lets try again:
|
|
August 30, 2011, 05:51 |
|
#19 |
Member
Join Date: Oct 2010
Posts: 63
Rep Power: 16 |
It does.
You are a true value of this sub-forum. Nobody gives replies in here, like you do. I owe you a beer! |
|
August 30, 2011, 05:56 |
|
#20 |
Disabled
Join Date: Jul 2009
Posts: 616
Rep Power: 24 |
Thank you, happy to help.
Boris |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
multiphaseInterFoam for RAS turbulence model | chiven | OpenFOAM Bugs | 8 | December 6, 2017 03:08 |
Reynolds Stress model in CFX vs Fluent | Tim | CFX | 1 | October 7, 2009 07:19 |
help for different between les model (subgrid-scale model) | liuyuxuan | FLUENT | 1 | October 2, 2009 16:25 |
2 stage axial turbine model convergence issues | sherifkadry | CFX | 2 | September 7, 2009 21:51 |
importing model from staad pro to ansys | ras | ANSYS | 0 | August 14, 2009 16:54 |