2d vs 3D models

[livewire]

Hi,

I would be very interested to have any feedback regarding the generation of 3d components from 2D, as I am concerned about the introductions of errors.

By way of example, if I model a 3D cylinder using a 2D clipped surface that is extruded, the external dimensions are not maintained properly (as per object info), whiist if I use the VW 3D cylinder the external dimensions seem to be maintained.

This would seem to be a a result of the error introduced by the representation of 2D circle as a set of segments within the extruded 3D cyclinder? Can this be changed or improved?

I seem to get the same effect when I subtract solids. If I subtract a 3D threaded shaft from a 3D cyclinder I loose the defintion on the outer edge of the cylinder and the resulting thread appears to be composed of a number of visible segments in wireframe and final quality Renderworks.

My goal is to export the resulting data as an IGES file for use with a CNC machine and am concerned that the aberrations will form part of the machined piece.

Any feedback much appreciated.

Thanks in advance,

The issue seems to be associated with solid subtraction. If you take two cylinders created by an extruded circles of different radius and subtract solids, then the resultant tube has an external dimension that is smaller than that of the original larger circle.

I am fairly new to VW, so I may have made a blindingly obvious mistake!

Thanks,

[Gytis]

I can't duplicate what you describe, i.e. extruded circles, solid subtract from larger, dim change.

Are you perhaps being stymied by display preferences? Go to VWPrefs>Edit>3D conversion res> set to Very High. This will increase the number of display segments, thus making a smoother arc.

As I'm sure you know, CNCs are limited by a similar "resolution" issue, insofar as the positions along an arc are always interpolated from x and y, and thus never really truly an arc.

[Guest]

By way of example, if I model a 3D cylinder using a 2D clipped surface that is extruded, the external dimensions are not maintained properly (as per object info), whiist if I use the VW 3D cylinder the external dimensions seem to be maintained.

Can you be very specific as to the settings (sizes, extrusion height, etc) and steps you are taking?

This is needed for an attempt to duplicate and either clear confusion, or enter the issue as a quality control ticket.

[Guest]

I deleted the duplicate post you entered over at the Machine Design Forum.

Please refrain from posting the same issue to multiple forums. Typically, someone will see the first one and answer the post.

[livewire]

Apologies, I was not sure if the two forums would be treated in isolation.

Thanks,

[propstuff at MJLA]

I tried this and got different results.

Circle 600?, concentric circle 300?

Extrude 600 circle to 1000, 300 circle to 1200

Subtract inner extrude from outer

Resulting cylinder reads in the OIP as 599.07924 instead of 600.

Same procedure with 1000? and 500? circles

Result reads as exactly 1000

N.

[livewire]

Thank you, your help is appreciated

I have just tried a simple experiment in which I have created two cylinders from extruded circles, one of 125mm dia (100mm extrusion) and one of 120mm (100mm extrusion), aligned them centrally and then moved the smaller one up 4mm. Subtracting the smaller from the larger results in a solid with a width of 124.5mm (as shown by the object info and also as measured with the tape tool) as opposed to 125mm? I may be missing somthing but I cannot find a rational for the dimension to be reduced by 0.5mm?

[livewire]

Having experimented some more I have found that if I repeat the process multiple times I get different results 124.83, 124.74... and sometimes 125mm (what I would expect). I thought it may be related to the Document Settings/Units but I cannot find the correlation. I have been using the Machine Design (Metric) template.

I assume that the internal mathematics is always done at the highest precision and that unit presentation and significant digits is only a display issue?

[propstuff at MJLA]

I think you may be seeing a bug or a flaw in the 3D engine.

The numbers shown in the OIP should be the correct size. :-(

N.

[livewire]

Thanks for the input. I will be seeing the mechanical engineering company today so I will do some comparisons of the IGES files to see if the actual 3D model is correct or incorrect and feed back my findings...

[livewire]

More info:

I gave an IGES file containing two hollow cylinders created in an identical fashion in VW (but showing external dimensions for the solid at 125mm and 124.7mm), to the engineer who imported them into his machining package. The CAD package used to define the milling path sees the cylinders as both having dimensions of 125mm (or very close), that is to say the size they should be. This is good news, but I still feel a little uncomfortable and would appreciate some feedback from the guys at Nemetschek. I have the VW file if they would like it.

Thanks in advance,

[Guest]

Sorry for the delay - I've been traveling.

Send the information (situations that fail) to bugsubmit@nemetschek.net.

This is certainly a quality issue.