Analyzing Thickness

[DaSalo]

Is there a quick way to locating the thickest point on a body that has complex curvature on all faces? For example, when I create an interference body using the simple interference analysis tool I commonly end up with something that has no planar faces. I would like to know the exact point of maximum interference and how much interference that is. This would equate to the thickest point on the interference body but I'm not sure how to exactly locate that point. Perhaps there is a better way to do this type of analysis?

I suppose I am looking for a tool that will allow me to select a non-planar face as the reference face, another non-planar face as the analysis face, and then show me the point of maximum separation between those two faces measured normal to the reference face at that exact point.

Thanks for any help.

 

[JohnRBaker]

When asking about specific capabilities or functionality, PLEASE indicate which version of NX that you're currently using.

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

http://www.siemens.com/plm

http://www.plmworld.org/museum/

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

 

[JohnRBaker]

Depending on what version of NX you're running, if it's one of the more recent one and you have the Studio Design module you may have a function which will measure the wall thickness of any entire part and display the values as a color-coded fringe plot. Check to see if you can access...

Analysis -> Molded Part Validation...

...and if this dialog opens, select the Analysis Type 'Thickness' and if there's only one body in the part file it will be selected automatically so all you have to do is select the 'Calculate Thickness' button (it the icon on the left that looks like a 'calculator').

If you don't own the correct modules, then there is very little that can be done to find model thicknesses, at least not without a lot of drudge work.

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

http://www.siemens.com/plm

http://www.plmworld.org/museum/

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

 

[cowski]

I suppose I am looking for a tool that will allow me to select a non-planar face as the reference face, another non-planar face as the analysis face, and then show me the point of maximum separation between those two faces measured normal to the reference face at that exact point.

The tool that comes closest to this description is Deviation analysis; use the face to face and maximum distance options. I'm fairly certain it is part of the core modeling application and should not need an additional license, it has been in every version of UG/NX that I have used. That said, this is most likely the 'drudge work' that John alluded to; faces to be analyzed will need to be selected individually.

 

[DaSalo]

Sorry, we're using NX versions 2-7.5. For this particular model I am working in NX 6. I'll check on the Studio Design module but I highly doubt that we have it. Deviation analysis sounds like a very good possibility. Never used that tool before so sounds like a good time to get familiar with it.

Thanks for the suggestions.

 

[JohnRBaker]

Unfortunately, while Deviation -> Checking... can be used, after a fashion, to measure the minimum and maximum distances between two faces, it will NOT show you the location of these conditions, however if you play with it a bit, Deviation -> Gauge... might be able to get you something usable.

And YES, this is the "drudge work" I was commenting about, particularly when you compare it to how the 'Mold Part Validation' tools work, which are virtually automatic and requires little or no interactive setup to use.

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

http://www.siemens.com/plm

http://www.plmworld.org/museum/

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

 

[man2007]

"Unfortunately, while Deviation -> Checking... can be used, after a fashion, to measure the minimum and maximum distances between two faces, it will NOT show you the location of these conditions"

@John,
Deviation -> Checking... will definitely show the Maximum or Minimum Deviation point (location) with * on the surfaces, but make sure that you have selected "Maximum Distance" or "Minimum Distance" from the "Deviation Option" dropdown list.

 

[DaSalo]

Deviation checking was a useful tool for what I needed to know. As man2007 mentioned, it does graphically show the point of max or min deviation(depending on which you ask for) with a "*". It also gives you the x,y,z coordinates of this point in the info window so you can easily create a datum point at that location by manually entering the numbers.

 

[Toost]

You are correct in that it shows where it has measured, but it will not search for the worst case points, only show which of the measured points are outside of the "allowed tolerance" .- It will drop points in a rectangular pattern and measure to corresponding points. The trick is then to play with the tolerance value to isolate the point or a few points of interest.

Another tip is the Analysis - measure distance if one knows the measuring direction, projected distance - maximum clearance.

 

[DaSalo]

@Toost

It does report the number of points that are outside the specified tolerance but it also singles out the point of maximum deviation. That is what I needed. I just set the tolerance at .001. The interference body I was looking at was thinner then that everywhere so it just marked the point of max deviation and told me nothing was out of tolerance. It might not have dropped a point on the exact worst case position but I'm very confident it gave me a result that was within .0001 of the actual worst case deviation. For what I'm doing that is good enough.

Maximum clearance is very useful in a lot of situations but because of the geometry that I was working with I just couldn't get what I needed using that. I did try.

Thanks for the tips.