Introduction
2-D Drawings Based on 3-D Models
A. Additional Issues on Sections of Original
Paper:
A1: Specifying Basic Dimensions
A2: User Specified Basic Dimensions
A3: Use of a note to specify basic dimensions
A4: Use of a general note to tolerance basic dimensions
A5: Use of CAD system specified basic dimension for size
of a feature of size with only the tolerance shown on the drawing
A6: Dimensioning of 90° angles.
A7: Use of CAD system specified basic dimensions for
datum target location, size and orientation
B.
Issues when using toleranced dimensions on a solid model.
B1: Use of toleranced dimensions with no nominal specified
B2: Use of max, and min dimensions
B3: Use of a toleranced dimension (which is not a feature
of size) with no datum sequence specified
B4: Specifying both a basic and a toleranced dimension
to a part feature
B5: Use of overall dimensions when CAD system specified
basic dimensions are present
B6: Use of the word “true” when dimensioning a feature
not in true projection
B7: Rules for spacing of dimension lines
B8: Extension lines
C. Fundamental Rules
C1: Each surface of the model is at a basic relationship
to every other surface on the model.
C2: CAD system specified basic dimensions applying at
assembly levels.
C3: Implied toleranced 90° angles and CAD system
specified basic 90° angles.
D. View dependent tolerances
D1: Use of the “between” symbol
D4: Use of the “All around” symbol
D5: Use of “Unilateral” tolerance zones with profile
D6: Use of “Unequal bi-lateral” tolerance zones with
profile
D8: Dimensioning to a true profile
D2: Use of the “Straightness” symbol, applied to a surface.
D3: Use of the “Profile of a line” symbol
D4: Use of the “All around” symbol
D5: Use of “Unilateral” tolerance zones with profile
D6: Use of “Unequal bi-lateral” tolerance zones with
profile
D7: Bi-directional positional tolerancing rectangular
coordinate method
D9: Limited length or area indication
D10: Use of a chain line to specify a partial datum
E. Datum specification issues
E1: Datum feature symbol location
E2: Datum feature symbols on extension lines
E4: Dashed leader lines
E3: Specifying equalizing datum targets.
E5: Datum symbol termination
F. Display issues
F1: The display of cutting planes for sections on a non-orthographic
views
F2: Display of a section cut from a non-orthographic
view .
F3: Direction for reading notes
F4: Application of Dimensions
F5: Terminating leader lines
F6: Text not in plane of screen
Conclusions
Related Topics
By: Lea Irwin
Date: 28-Oct-1998
Introduction
In his paper, Alex does a good job of identifying many of the issues encountered
with the direct application of the ASME standards Y14.3, Y14.4 and Y14.5
on digital drawings. However, I submit that the subject is not really
one of the application of these standards to digital drawings, but rather
to 3-dimensional models. The term “drawing” has a very specific meaning,
one of a series of 2-dimensional views of a product with a specific set
of rules as to its construction and interpretation. I believe a distinction
needs to be made between this 2-dimensional presentation and the 3-dimensional
presentation in order to allow the reader to avoid confusion. Therefore,
I consider the 3-D model as an entity distinct from a Drawing.
Much of my focus deals with the presentation of the symbols and their
interpretation based on that presentation. I do not believe the dimensioning
and tolerancing concepts defined in the standards are any different in
3-D than they are in 2-D. The question is, how do we communicate the dimensioning
and tolerancing information necessary to ensure the manufacture of the
desired product?
The Y14 standards are written assuming a 2-dimensional medium. If that
medium is paper or computer, it is still 2-dimensional. We have devised
a language based on that 2-dimensional representation of a part. On a
drawing, we cannot represent the actual part in 3 dimensions, so we provide
a series of 2-dimensional views of the part which are sufficient for the
reader to infer the 3-dimensional object which is to be made. With the
use of 3-dimensional modelers, we need to devise a means to communicate
the same dimensioning and tolerancing design intent, but we now have a
3-dimensional representation of the model with which to work.
3-Dimensional Presentation of Annotation (e.g. dimensions and geometric
tolerancing symbols)
I believe that many of the issues which arise from the application of
the drawing standards to 3-dimensional models come from the literal application
of the annotation using the same techniques applied on drawings. Although
the following statement is an oversimplification, it exemplifies my primary
point. In 2-D, we had to connect the annotation to a 2-dimensional representation
of a 3-dimensional object. We have well specified rules which a reader
employs to infer the 3-dimensional portion of the part to which the annotation
pertains. For example, if a GD&T callout is applied to a profile of
a surface in a view, the tolerance applies to the surface which goes “into”
the paper, the one perpendicular to plane of the view. In 3-D, we can
connect the annotation to the actual 3-dimensional object. The reader
no longer needs to infer the affected geometry, it can be communicated
directly. We need to address the real underlying issue of indicting the
controlled geometry in 3-D. The ambiguity which results from utilizing
a 2-D presentation technique in 3-D is really a symptom, not a problem
in and of itself.
Another significant difference between interpreting annotation on a drawing
and annotation on a 3-D model is that the 3-D model is free to “move”.
On a drawing, the representation of the part is static. In a 3-D modeler,
the part model may be manipulated, i.e. moved and rotated. This is a definite
benefit as it allows the viewer to study the model in a more natural manner.
He/she can manipulate the model in space to gain a better comprehension
of its structure. On a drawing, we have to resort to multiple views, sometimes
detail or auxiliary views, in order to communicate the 3-dimensional characteristics
of the part. It is this ability to move the part from a static viewing
direction which causes the literal application of 2-dimensional annotation
techniques to become problematical. For example, if GD&T is placed
such that it points to an edge of a surface in a particular viewing orientation,
when the 3-D model of the part is rotated, the meaning of the annotation
becomes ambiguous as it now points to an edge shared by one or more surfaces.
We should devise a means for applying annotation to 3-D models which best
allows for their interpretation in 3-D and not require the viewer to use
static views of the model to interpret the annotation. Although this sounds
a bit philosophical, I'll offer it for the reader's consideration. We
have actually been limiting ourselves by working in 2-D, albeit out of
necessity. We should be able to express our annotation in a more natural
fashion, i.e. 3-D, rather than continuing to limit the 3-D world by insisting
on the use of more restrictive 2-D annotation techniques.
2-D Drawings Based on 3-D Models
The author recognizes the concerns with regard to the generation of drawings
and hard copy output from 3-dimensional models. While many companies are
moving toward annotating their 3-D models instead of relying on production
of 2-D drawings, there is still a need to be able to produce paper based
output. However, this need not necessarily imply that the techniques used
to annotate a 3-D model be those used on 2-D drawings. If a drawing is
produced, then the rules from the Y14 standards still apply.
A. Additional Issues on Sections
of Original Paper:
The following comments pertain to the six subsections of Alex's white
paper. The following are intended as further consideration of the issues
and are not meant to serve as solutions.
Issues when using measured distances of the model as CAD system specified
basic dimensions of the part
We need to keep in mind that the idea is to communicate the dimensions.
This is not intended as an affront to the reader of this paper, who certainly
understands that. However, it is easy to forget this point and become
embroiled in discussions of “where the dimension comes from”.
A1: Specifying Basic Dimensions
There may be cases where it is advantageous to explicitly show a basic
dimension, in order to properly communicate the design intent of the part,
even if the value of the dimension is obtained from the CAD model.
If dimensions displayed on a 3D model are obtained from the model, should
they even be allowed to have different values? Do we really need to present
dimension values other than those from the model?
A2: User Specified Basic Dimensions
The question posed by Alex is a fair one but I submit this falls into
the category addressed by my introductory comment on dimensions. The need
is to be able to communicate the dimension value. It is intended as a
basic dimension for the purpose of manufacturing or inspection. We need
not confuse the intent with the mechanism used to obtain the dimension
value, i.e. interrogation of the CAD model.
A3: Use of a note to specify basic dimensions
A4: Use of a general note to tolerance basic dimensions
I would like to add that issues A3 and A4 are very much related
to the ability to analytically process dimensions specified in general
notes. We need to be able to enable electronic processing of the digital
data models and general notes are problematical.
A5: Use of CAD system specified basic dimension for size
of a feature of size with only the tolerance shown on the drawing
I refer back to my introductory comment on dimensions. The CAD system
can certainly provide values from the model. If the values are not explicitly
displayed, is there really any point of confusion? Is it simply a matter
that any explicitly displayed dimension on a CAD model should take precedence,
in terms of interpretation, over dimensions interrogated from the model?
I am assuming the value will be the same but an explicitly displayed dimension
may be toleranced whereas an interrogated dimension probably would not
be.
A6: Dimensioning of 90° angles.
No additional comments.
A7: Use of CAD system specified basic dimensions for
datum target location, size and orientation
No additional comments.
B.
Issues when using toleranced dimensions on a solid model.
B1: Use of toleranced dimensions with no nominal specified
The value used to model the feature is easily obtained from the CAD model.
The question is what value should be used to manufacture the part. Is
this issue any different on a 3-D model than it is on a drawing?
B2: Use of max, and min dimensions
Is this issue any different on a 3-D model than it is on a drawing?
B3: Use of a toleranced dimension (which is not a feature
of size) with no datum sequence specified
Is this issue any different on a 3-D model than it is on a drawing?
B4: Specifying both a basic and a toleranced dimension
to a part feature
I refer back to my introductory comment on dimensions. The CAD system
can certainly provide values from the model. If the values are not explicitly
displayed, is there really any point of confusion? Is it simply a matter
that any explicitly displayed dimension on a CAD model should take precedence,
in terms of interpretation, over dimensions interrogated from the model?
I am assuming the value will be the same but an explicitly displayed dimension
may be toleranced whereas an interrogated dimension probably would not
be.
B5: Use of overall dimensions when CAD system specified
basic dimensions are present
No additional comments.
B6: Use of the word “true” when dimensioning a feature
not in true projection
The use of the word “true” is one of the techniques devised to overcome
the fact that our presentation medium was a 2-D representation of a 3-D
part. In 3-D, all dimensions on the model should reflect the actual measured
size. The value of the dimension does not depend on the direction from
which the model is being viewed.
B7: Rules for spacing of dimension lines
I concur with Alex's point. The rules for dimension spacing are an artifact
of a paper presentation and have their origins in clarity of an image
printed on paper.
B8: Extension lines
This is another case where we need a mechanism to indicate the 3-D geometry
being dimensioned. Extension lines seem to be one of those tools which
are 2-D oriented but are not immediately inapplicable in 3-D. These issues
do need to be investigated.
C. Fundamental Rules
C1: Each surface of the model is at a basic relationship
to every other surface on the model.
Are we getting too restricted by the literal interpretation that, because
a CAD model can provide a measured value (i.e. a basic dimension) for
any surface on the model, that each dimension should be considered as
having been explicitly specified?
C2: CAD system specified basic dimensions applying at
assembly levels.
Is it sufficient to control the visibility of the part dimensions in the
context of an assembly?
C3: Implied toleranced 90° angles and CAD system
specified basic 90° angles.
No additional comments.
D. View dependent tolerances
D1: Use of the “between” symbol
D4: Use of the “All around” symbol
D5: Use of “Unilateral” tolerance zones with profile
D6: Use of “Unequal bi-lateral” tolerance zones with
profile
D8: Dimensioning to a true profile
If we do not restrict ourselves to the use of 2-D techniques for indicating
the controlled geometry, many of these issues are easily solved. The issue
is that we need to directly indicate the 3-D geometry now that it is available.
Should we propagate a 2-D mechanism into 3-D?
D2: Use of the “Straightness” symbol, applied to a surface.
D3: Use of the “Profile of a line” symbol
D4: Use of the “All around” symbol
D5: Use of “Unilateral” tolerance zones with profile
D6: Use of “Unequal bi-lateral” tolerance zones with
profile
D7: Bi-directional positional tolerancing rectangular
coordinate method
D9: Limited length or area indication
D10: Use of a chain line to specify a partial datum
These issues are good examples of a dependency on the 2-D nature of the
drawing presentation. An equivalent mechanism should be derived for 3-D
representation which communicates the design intent while being as robust
as possible in a 3-D environment where the part can be dynamically manipulated.
E. Datum specification
issues
E1: Datum feature symbol location
E2: Datum feature symbols on extension lines
E4: Dashed leader lines
If we do not restrict ourselves to the use of 2-D techniques for indicating
the controlled geometry, many of these issues are easily solved. The issue
is that we need to directly indicate the 3-D geometry now that it is available.
Should we propagate a 2-D mechanism into 3-D?
E3: Specifying equalizing datum targets.
No additional comments.
E5: Datum symbol termination
No additional comments.
F. Display issues
F1: The display of cutting planes for sections on a
non-orthographic views
Just a small addition to Alex's points. We need to be careful as we devise
presentation techniques for 3-D that they do not add unnecessary clutter
to the model. As all annotation may now be presented on the model at the
same time, the annotation techniques should not add clutter or obscure
the geometry. For example, planes will be more obtrusive than lines.
F2: Display of a section cut from a non-orthographic
view .
No additional comments.
F3: Direction for reading notes
No additional comments.
F4: Application of Dimensions
I concur with Alex's point. The rules for dimension spacing are an artifact
of a paper presentation and have their origins in clarity of an image
printed on paper with ink.
F5: Terminating leader lines
No additional comments.
F6: Text not in plane of screen
Another example of a dependency on the 2-D nature of the drawing presentation.
An equivalent mechanism should be derived for 3-D representation which
communicates the design intent while being as robust as possible in a
3-D environment where the part can be dynamically manipulated.
Conclusions
I concur with Alex's assessment that the display issues are a central
question in this discussion. I would counter however, that the use of
the phrase “human interpretable” to mean “interpret using literally the
same techniques and rules applied to a 2-D presentation” is too restrictive
and somewhat misleading. It seems that humans can interpret what is seen
on the 3-D display even if the model can be dynamically manipulated.
Related Topics
It is no longer sufficient to have the annotation “just look right” to
a person. We need to explicitly identify and capture in the CAD data model,
the portions of the part to which the annotation pertains. This is essential
to enable the integration of computer based design and manufacturing packages.
However, as this is along the lines of a solution, I have not included
it in the previous material.
Another subject is that of dynamically querying the 3-D model to obtain
information about the controlled geometry. I feel this is an extremely
powerful tool and one which should be leveraged to its fullest extent
in 3-D modelers. However, as this is something of a solution instead of
an issue, I have purposefully avoided the topic. The technology is rapidly
becoming available which makes this viable, even in applications outside
of the CAD system. Also, because this tool cannot be leveraged in a paper
presentation of a part model, it is not sufficient in and of itself. It
needs to be combined with adequate visual interpretation of the annotation.
Issues when using ASME standards for solid models L. Irwin
