Hi....

Today I got iges file from my customer with so many disjoint surfaces.
I tried to join the surface to single solid body, but its not working. I also have turn on "join surface" under iges import.

Any can give me advice to get solid body from many disjoint surfaces ?
I will give you the file if you want to try.

Thanks !!!

Hi Anis,

Before you try anything else and ask everyone to have a go, can the customer supply the model in another format, you'd probably have more success with STEP.

-

Hi Anis, definitely what Danny says about getting some alternate file formats is a good bet - in fact if possible ask for the widest variety of formats that they can make, like IGES, STEP, and SAT.

Both STEP and SAT maintain the solid information so they can avoid re-joining issues.


One other thing you can try with join is to scale your objects down in size by 1/10 or 1/100 and then try to join them, that may shrink the gaps between pieces to within the join tolerance.

To quickly scale everything, select it all, then run the Transform/Scale command, type in 0 and push Enter to specify 0,0,0 as the scaling origin, and then type 0.1 (or 0.01 for 1/100) and push enter and your objects will be scaled.

- Michael

Hi Danny,

Unfortunately, my customer always use iges as std format to send the file to their vendor.
Sometimes I have question in mine, why many company including big automotive company like Toyota, Honda etc usually use IGES instead STEP or another better neutral format ?

Thanks

Hello,

Are you able to post the IGES file?

Hi OSTexo,

PM me and provide your email and I will send the file to you....
Sorry the file is confidential.

Thanks

Hi Michael...

> One other thing you can try with join is to scale your objects down in size by 1/10 or 1/100 and then try to join them,
> that may shrink the gaps between pieces to within the join tolerance.

Thanks, its much better.
I scale it down 1/100 and now the total number of surfaces = 30.
Looks like there are some bad geometry like overlapping each other.
Should edit the surface manually...

Thanks

I have sent the file to several peoples in the forum.
You can add comment or posting screen shot to this forum.
But please dont posting the 3D file.

Thanks

Hi Anis

You really need a step file, I couldn't do much with it even with messing around with tolerances in rhino. You can see the unjoined edges here over 600 of them out of 6000 or so edges overall so a load of work to recreate. There are some very small sliver surfaces in there which are not nice. You could rebuild these areas but its not fast work. you will spend a lot of time sorting it out a new file will save so much time.

Out of interest what software do they use to create this?

regards

Jamie

Hi Jamie...

Thanks for tried to solve it use rhino.
Yes, the surfaces is complex and also have "bad geometry".
I dont know exactly the software they are using to create those model, I will ask to my customer.

Looks like there is no "magic" button to solve this case.
I will examine one by one manually... :)
But at least Michael's tips to scale down is help to solve this case.

Thanks Jamie....

Jamie you are right, it is an "UGS NX IGES V5" file with > 2500 single surfaces.

Anis I have send the STEP file and I hope you can use it. :)

Hi Ralf,

How do you the file come from NX, I have opened the file use notepad but cant any phrase with "NX"...
Thanks for sent to me the STEP file, I have open it and still the same condition. I mean contains many disjoint surfaces.

I am appreciate your help, thanks.... :)

Hi Anis & Ralf - is the STEP file you are referring to recently one created by a conversion from the IGES file?

Instead of that you want to get a STEP (and possibly an SAT) file created from the original application that exported this IGES file, exported from the original native model data.

That may help keep the data all connected together as a solid more easily rather than trying to deal with the individual surfaces in the IGES file.

- Michael

Hi Michael...

I think Ralf export from Iges.
Which do you think better comparing export to iges, step or sat ?

Thanks

Hi Anis,

> Which do you think better comparing export to iges, step or sat ?

Each of these formats has some complexities in slightly different areas, so I would suggest getting as many different formats as possible.

So instead of choosing one between IGES, STEP, and SAT, instead request getting the same model data in each of those formats, including 3DM and Parasolid if they are able to do those.


But if the original data is messy and possibly contains holes and overlaps in it then of course that will not get automatically solved by using some particular file format.

If the original model data is ok then one of the formats that contains solid information (STEP or SAT) will save you some joining work.

- Michael

Ok, then how about the precision / which is more accurate ?

Hi Anis, are you talking just about numeric accuracy? They are all pretty close to equivalent in the representation of numbers.

Those formats are text file formats and I think the number of digits to use for the text representation of numbers is something that is up to the particular application generating the file.

But it is very unlikely that you will run into problems with not enough digits in the number formatting used in any of those formats.

- Michael

thanks Michael... :)

I recently ran into a model that appeared the same. It "appears to be a simply constructed solid, the many of the fillets are naked. There doesnt appear to be anything like overlapping trim boundries or fincky edges or anything. I'm thinking this might be something related to "Tolerances" set in other apps that then dont propogate unless the same tolerance is known in the receiving app.

We've discussed this before Michael, with regard to Rhino and you described your thoughts that having that variable tolerance value was a design mistake that lead to these types of situations.

I've requested the file from Anis also to have a looksee at it.

Hi Burr, well tolerances come into play mostly when various kinds of geometry is created.

Many operations cannot be calculated directly, they go through a kind of iterative refinement until the result is close enough to the desired accuracy, with the desired accuracy being the tolerance.

Various things like general surface/surface intersection, fillet surface creation, sweeping, etc... are all things that use that kind of refinement.

If you have set a tolerance that is too loose, then it means the the surfaces that you generated will kind of sag away by some significant amount away from where their ideal locations would be. That's something that happens at the time that the surface is created, it doesn't have a whole lot to do with just saving data that has already been created.

Various kinds of systems can also have an error tolerance value that is associated with edges, which kind of marks the distance of the maximum gap between the 3D edge and the surfaces that are associated with it. This kind of error value assigned to an edge is not really too crucial to be sent down by export processes because it can be calculated later.


- Michael