Jump to content

Datum System with Free Surface


---
 Share

Recommended Posts

Hi,

an arbitrary shape can bind up to six degrees of freedom. It's not easy to analyze a shape and to derive safely "only" the intended "main" degrees of freedom. Furthermore, we are dealing with measured data, therefore even a "plane" is not a perfect plane. Which degrees should be bound? From real world applications we derived the information how surfaces were usually interpreted and therefore we "defined" that all degrees of freedoms were bound if the surface is specified as primary datum.

The ISO norm is also not helpful in this case how to derive degrees of freedom and which directions are specified (therefore you have to define a coordinate system where the concrete directions are taken from).   

Regards,

Bernd

Link to comment
Share on other sites

Please sign in to view this quote.

Hello Bernd, 

Thanks for the reply. I would like to discuss more on this topic. Would be possible to have a 1on1 session.

regards

TK

Link to comment
Share on other sites

And to add the last argument, we will bind all degrees of freedom because that is the default according to ISO 5459 (see table A.1):

image.thumb.png.7e4aa755a50ad7553767f16437986c2f.png

According to ISO it is the default case to bind a plane, a line and a point. This will result in binding all remaining degrees of freedom. Of course you could define on your technical callout only to bind one situation feature but this is not supported from the GOM software. 

Reason for this (as Bernd Alpers has mentioned already) it is that is hard to impossible to estimate a line, plane or point from a complex surface.

Therefore we choose a different approach to compute the actual datum system. The complete algorithm is described in the techguide: https://techguide.gom.com/en/gom-software-2022/article/cmd_gdat_create_datum_system.html#mark-3. Please have a deeper look. In short instead of estimate the situation features we compute only a transformation from nominal to actual and apply it to a coordinate system. Unfortunately this will only work for the default case. Therefore GOM software do not allow to choose a single situation feature and therefore it will bind all degrees of freedoms. 

Hope this helps

Christoph Schult

Link to comment
Share on other sites

  • 4 weeks later...

Thankyou both for the explanation.

How would one then use a prismatic feature as a Datum feature. As per ISO 5459. It can still have 1 Translation along the line of a plane free.

I have then the folowing example. The highlited surfaces are datums. The inner walls are Datum A and have no draft angle. If i am not wrong the Prismatic principle could be used here.The bottom flat surface is Datum B. The A should lock 5 degrees of freedom and the 6th would be locked Datum B. Can you guide me how i can build this in GOM. 

 

image.thumb.png.54ba740b0486b9302df3815b94810796.png

image.png.ab6782733a2bc9e5e2ce37624426718c.png

image.png.03c4710024af772447ccbfe80134162d.png

Link to comment
Share on other sites

This invariance class is not supported from the GOM software.

But you would have the same problem to construct a line and plane from the surface as discussed in the surface problem. In my point of view there are 3 quality solutions for this problem is the GOM software (I am aware of the fact that all of them change the original the original check slightly; but together with your customer/constructor you have to make the decision which changes is suitable for you.

1) You make one surface on booth, the ground plate and the walls and use only a primary datum

2) You ignore the information in cylinders in the wall. And make a common datum from two pairs of parallel planes to create a datum with the same invariance class.

3) You change the order of the datums, i.e., the ground plate is primary datum. If you only want to use to stop the translation you can construct a point in both actual and nominal and use this point a primary datum instead of a plane. Nevertheless, in this case you are using the wall as surface but as secondary datum.

As I mentioned, that are only workarounds and deliver slightly other results than the original definition. But in my point of view this is the nearest approximation that is possible with the current version of the GOM software.

 

Hope this helps.

Christoph Schult

Link to comment
Share on other sites

  • 1 month later...

Another workflow would be to perform a local best fit on the  free form surface that restrains all six degrees of freedom then use the CAD coordinate for your analysis.

Link to comment
Share on other sites

 Share

×
×
  • Create New...