Merge faces into one FACE? Is it possible? 3

[opetrenko1]

I'm trying to measure wetted area distribution of a sheet body consisting of hundreds of faces.
for this:
1. sheet body is trimmed by a plane (say, x=0).
2. area of trimmed body is measured.
3. plane is moved, say +1" (x=1)
4. GOTO 2.

currently this is done tediously by hand, and not automated with "goal analysis" due to the following
PROBLEM: face measurement feature is not updating, as some of the faces of measured body appear (or disappear if trimming out the other side)

So, hence the
QUESTION: is it possible to make face measurement feature update properly?
OR: combine faces into one FACE. (ICEM, i know, is capable of this)
PS. Geometry is too complex to slice is along and accross to put surface mesh on it...

THANKS!

 

[opetrenko1]

P.S.S. Sorry. I have access to NX6.0 and NX8.0

 

[jerry1423]

Insert (pull-down) -> Trim -> Join face

 

[opetrenko1]

Thank you.

But it doesn't work. (I use "Convert to B-Surface")
for almost every face (90%-95%)comprising the sheet body I get is
"face is not parametrical-rectangular"

I see this menu item is abcent in NX8.0.
May be they came up with something better?

 

[opetrenko1]

In NX8.0 I guess the analog would be
Insert -> Trim -> Delete Edge.

But this tool doesn't even give the error. It simply doesn't select any internal edges.

 

[jerry1423]

can you show a picture ?

 

[Cid1979]

Can you attach the file so others can work on it with there ideas?

 

[opetrenko1]

Thank you, all.
I'll prepare an example file.

 

[JohnRBaker]

Are you simply measuring the surface area, or are you trying to optimize it to some value?

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.

 

[opetrenko1]

Thank you everybody looking into this topic.

Here I prepared a simplified example (NX6.0) of a nozzle with a centerbody. I put a couple patches on the outer surface. In reality the whole surfaces is made of patches. It would also have surfaces of pylons connecting outer and inner surfaces, fillets and other details.

http://www.cap-ny153.org/Nacelle Diagram.jpg

The example goal is to measure wetted area distribution of the outer surface of the nozzle with centerbody.

For this a plane is trimming the body. Then area measurement is made and repeated after each surface shift. Along with other measurements of flow area and others it can provide good estimate of losses in the nozzle. And could be used as part of design optimization, sensitivity analysis and other.

using goal analysis in the embedded sheet would significantly speed up the analysis. Unfortunately, area measurement fails (does not update ) once a face in the set of measured faces disappears as it get trimmed away (why not to measure it as 0in² and move on?) And it definitely does not start including new faces as I move plane forward. (why not allow measuring surface a BODY, not only separate faces?)

you can see on the chart the failure point is position of the plane when purple patch is trimmed away completely

So I thought the solution would be is to merge somehow all the faces into one. But UG, compared to ICEM, has very limited capabilities of that (Insert -> trim -> join) as I described in the post yesterday.

THANKS!

P.S. if you disable trim body feature, enable trim feature and move trimming plane around, you'll see that the portion trimmed away is changing from left to right. A problem somewhat similar to what I was describing in

http://www.eng-tips.com/viewthread.cfm?qid=285907

but couldn't create a example for.

 

[opetrenko1]

error: "after each surface shift"
should be "after each plane shift" sorry

 

[JohnRBaker]

I'm sorry, but if you think your image showed us a "simplified example (NX6.0) of a nozzle with a centerbody" you better recheck your images

You imply that you have to select faces one at a time. If you're using the...

Analysis -> Measure Face...

...function, have you looked at the 'Selection Intent' options when you selected the faces? After all, there are several options including 'Body Faces'.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.

 

[cowski]

I think the simplified example is the .prt file attached.

I tried out different selection intent options, but no matter what was used the measure face feature became unassociative and gave incorrect results when one or more of the faces were cut off completely by the datum plane.

http://www.nxjournaling.com

 

[JohnRBaker]

Sorry, I missed that

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.

 

[cowski]

The .jpg is a cool diagram, but I had the same initial reaction!

http://www.nxjournaling.com

 

[opetrenko1]

cowski,
yes. It seems selection intent is part of the GUI and is not preserved in the feature.

trim sheet feature is even more puzzling: when a surface defining flow area (a cone trimmed by the multi-patched duct) is moved along the duct (to measure area distribution along the nozzle) the timmed\retained regions swap several times. For that I end up using trim body, with significant overhead of trimming all patches to sew into one sheet body.

this example though gave me an idea: it seems the trim sheet feature stores which face of the body was selected for trim sheet feature. May be cutting out a tiny piece in the middle of the flow area sheet, patching it, and selecting a patch would make trim sheet feature robust. I won't blame you, if you call it way distorted.

the .jpg intent was to illustrate one of the applications.

Thanks

 

[Toost]

To reply to the initial question, no it is not possible to merge all faces into one single face, the needed math would be enormous.
It might be possible to merge into a "polygon face" which what is what NX CAE is using for FEM meshing. Those faces have the ability to delete edges etc.
Maybe, ( i dont know) you can measure the area there.

 

[opetrenko1]

Thank you, Toost

NX8 crashes when exporting real geometry into the polygon file, but survives the example. However, how do I import the polygon file? Don't see options in Open or Import
Thanks

 

[JohnRBaker]

For the record, I have opened a PR against the Measure Face feature failing when the number of faces changed since if this had worked as expected, the rest of this discussion would have been moot.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

UG/NX Museum:

http://www.plmworld.org/p/cm/ld/fid=209

To an Engineer, the glass is twice as big as it needs to be.

 

[opetrenko1]

Thank you John