INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Are you an
Engineering professional?
Join Eng-Tips Forums!
• Talk With Other Members
• Be Notified Of Responses
• Keyword Search
Favorite Forums
• Automated Signatures
• Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

#### Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

# Secondary datum feature must perpendicular with primary datum feature?5

## Secondary datum feature must perpendicular with primary datum feature?

Yes.

(OP)

Season

### RE: Secondary datum feature must perpendicular with primary datum feature?

Datums are always 90 degrees to one another. Datum features don't have to be. The difference in the definition of those words should clarify why the answer to your first question is "yes."
Looking at the ASME Y14.5 standard, Figure 4-7 is the most obvious example of non-perpendicular datum features.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

### RE: Secondary datum feature must perpendicular with primary datum feature?

Datums are not always perpendicular; at least there isn't a statement to that effect. One can't say a datum point is perpendicular.

The planes making up a DRF are defined to be, though they don't always get named. For example, there are 3 unnamed planes in a DRF for a spherical datum feature.

It looks like in the '2009 version there was a semantics fight over datums, theoretical datums, datum features, datum feature simulators (theoretical) datum feature simulators (physical). There's five great questions for the GDTP test. I get the desire for precision in language, but I don't get how they could be confused in context. Was there an inspector somewhere rejecting parts because a datum feature wasn't perfect, when he thought it had to be since someone called it a datum?

In the meantime, no one questioned "A datum reference frame is three mutually perpendicular intersecting datum planes" as if unbounded, mutually perpendicular planes were not guaranteed to intersect. Maybe geometric planes aren't infinite in Y14.5. If so, what happens if they are too small?

Here's a puzzle. Where is the DRF for skewed cylindrical features that are the basis for datum features when they don't share a mutual perpendicular?

### RE: Secondary datum feature must perpendicular with primary datum feature?

Dave, I think you know what I meant. It's true that any two random datums derived from a print don't have to be perpendicular, but datums that comprise a DRF are perpendicular, according to Y14.5. That was the intent of the OP's question.

Not sure what the "skewed cylindrical features" in your last question look like. Could you attach a sketch? You refer to features (plural), so from what I am picturing you'd have to home in on one of the "features" as the primary, and the other as the secondary. The result would be perpendicular things.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

### RE: Secondary datum feature must perpendicular with primary datum feature?

Grab a rod. Now grab a second rod. Hold them so they aren't parallel and not perpendicular to each other. That's skewed.

It was an unqualified blanket statement about datums, not DRFs, regardless of OP context. In particular they were contrasted with datum features, which are not DRFs.

### RE: Secondary datum feature must perpendicular with primary datum feature?

Then, if you wish to create a DRF from the two rods, you'd have to choose a "datum feature" to be primary (perhaps one of the rods?). The you'd have to choose a "datum feature" to be the secondary. Despite the convoluted shape of the actual part, wouldn't the secondary datum need to be perpendicular to the primary datum? (Recall that it's about degrees of freedom...)

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

### RE: Secondary datum feature must perpendicular with primary datum feature?

Each rod controls 3 degrees of freedom, so the two skew rods control 6. Both can have identical priority and the pair can be the primary and only datum reference required.

I'd say that one could create any arbitrary DRF as long as the two rods are located and oriented to it for the purposes of locating other features.

To drag this back to the apparent OP problem, I think it is less valuable to worry about the DRF which is theoretical and think more about what one would use to gage the part and whether a realizable gage can be made. Even if a CMM will be used at the end, if a part will engage other real parts, then it should be possible to create a gage to validate it and by going through that thought process it's easy to see if the choices are suitable.

### RE: Secondary datum feature must perpendicular with primary datum feature?

Very interesting discussion about DRF construction, datums, etc.

I would, however, like to notice one thing regarding the original drawing. In theory with the angled secondary cylindrical datum feature B there is no need for tertiary datum feature reference in the profile callout. A primary and B secondary fully constrain the detail relative to the datum feature simulator. Would you agree?

### RE: Secondary datum feature must perpendicular with primary datum feature?

pmarc,
Probabbly the same concept can be applied to fig 4-53 (2009) where tertiary datum feature C is not needed?
In other words, if the profile callout is to A (primary)and B (secondary)only, and the Ø12mm hole has the same DRF, the meaning of the drawing won't change?
Am I unconditionally correct or you have to add some caveats for the two schemes to be equivalent?

Thank you

### RE: Secondary datum feature must perpendicular with primary datum feature?

It probably would not suffice as a practical installation, which would mean verifying the part would be problematic. In an idealization, like a typical CAD assembly, the software would come to one conclusion, but for an actual part the secondary feature is too small and too short to adequately provide a reasonable solution. One could, with any little variation, turn the part several degrees to either side of a nominal orientation and the nominal for that variation may not align any other features.

### RE: Secondary datum feature must perpendicular with primary datum feature?

#### Quote (3DDave)

It probably would not suffice as a practical installation, which would mean verifying the part would be problematic. In an idealization, like a typical CAD assembly, the software would come to one conclusion, but for an actual part the secondary feature is too small and too short to adequately provide a reasonable solution. One could, with any little variation, turn the part several degrees to either side of a nominal orientation and the nominal for that variation may not align any other features.

3DDave,
And this is where the tool that you had such bad opinions about could come into play - customized datum reference frame ;-]

### RE: Secondary datum feature must perpendicular with primary datum feature?

The thread is focusing on theory and mfg and QA. But... is this an actual part for mfg or is it a case-study? If an production part, do the datums and the sequence communicate he function and fit up the part?

Certified Sr. GD&T Professional

### RE: Secondary datum feature must perpendicular with primary datum feature?

greenimi,

I would say that your question about fig. 4-53 touches different topic.

In this (Season's) example two datum features are enough, at least theoretically, to fully constrain the part. If the cylinder B was longer and bigger in diameter, there would be no question about it in reality too.

In Fig. 4-53 the part will not be fully immobilized using A and B datum targets only. That is why tertiary datum hole C has been defined. I agree, it is not needed for angular relationship of the cutout and the hole (simultaneous requirement would do the job), but I do not agree that in this particular case there would be no geometrical difference between having and not having C in the profile callout.

I think I know what your next question is going to be

### RE: Secondary datum feature must perpendicular with primary datum feature?

I would say that datums don't have to be perpendicular to each other. I agree with 3DDave that Y14.5 uses some of the different terms rather loosely. We have Datum Features, Datum Feature Simulators, Datums, and a Datum Reference Frame. Unfortunately Y14.5 uses the term "datum plane" and "datum axis" to refer to different entities in different contexts and this creates confusion.

I believe that Fig. 4-6 in Y14.5-2009 shows everything correctly. 4-6 (c) shows the datums, and two of them are datum planes. The other datum is a datum axis. The datums happen to be mutually perpendicular in this example, but in general they don't have to be. 4-6 (d) shows the datum reference frame, which consists of 3 planes that are by definition mutually perpendicular. I would have preferred that these planes be described as "planes of the datum reference frame".

Fig. 4-7 muddies the waters, partly because of the fact that many of the entities are coincident. The geometry labeled "First Datum Plane" is in fact a planar Datum Feature Simulator, Datum Plane A, and a plane of the datum reference frame, which are all coincident. So is the geometry labeled Second Datum Plane. The geometry labeled "Datum feature simulator of datum feature C" is a planar Datum Feature Simulator and Datum Plane C, which are coincident. The geometry labeled "Third datum plane" is a plane of the datum reference frame.

By the way, where I've used the term "Datum Feature Simulator" I meant the Theoretical Datum Feature Simulator. The descriptions don't account for physical datum feature simulators or simulated datums. That would just complicate things ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

### RE: Secondary datum feature must perpendicular with primary datum feature?

Regarding the degree of freedom ocnstraint in the OP example, I agree with pmarc that A and B fully constrain all 6 DOF's. Datum feature C can't constrain anything and should not be referenced.

This is different from the situation in Fig. 4-53, where datum feature C constrains a degree of freedom that has no effect on the circular runout tolerance in this case.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

### RE: Secondary datum feature must perpendicular with primary datum feature?

I agree with pmarc's DOF analysis as well - Datum C is not required.

Certified Sr. GD&T Professional

### RE: Secondary datum feature must perpendicular with primary datum feature?

I objected to the addition of customized reference frames as a part of the standard because it's horrific and fails to describe the physical embodiment of the mating part function.

"All" you need to do to inspect the part is to create a plane through A and then an axis through B and then a plane that is through B that is parallel to A. Measurements should come from the intersection of the B-Axis with the A-Plane orthogonal to the A-Plane and B-Plane and the un-named plane that passes through the intersection and is perpendicular to the A-Plane and B-Plane.

However, finding a repeatable method for finding the B-Axis is going to be tough; small differences in measurement will result in wide swings in the derived orientations.

### RE: Secondary datum feature must perpendicular with primary datum feature?

(OP)
3DDave

Am I right on the sketch?

Season

### RE: Secondary datum feature must perpendicular with primary datum feature?

B-Axis would be horizontal in that picture and in the middle of the datum B feature.

### RE: Secondary datum feature must perpendicular with primary datum feature?

(OP)
AS we discussed earlier, all datum planes is mutually perpendicular, if B-axis is horizontal, then it will aganist "mutually perpendicular"，or something I missed?

### RE: Secondary datum feature must perpendicular with primary datum feature?

I agree with 3DDave. You might want to consider datum targets to refine the portion of datum feature B that is used to determine the axis.

Certified Sr. GD&T Professional

### RE: Secondary datum feature must perpendicular with primary datum feature?

SeasonLee,

This is the type of situation where Y14.5's datum reference frame theory starts to show its limitations. It's built on the concept of building a Cartesian coordinate system on simple geometry with orthogonal relationships, where the degree of freedom constraint is clear. When the datum feature geometry is more complex, things get ambiguous.

In your example, A gives you a datum plane. The datum feature simulator for B would be a cylinder, that is at the non-orthogonal basic angle relative to datum plane A. This secondary simulator would constrain the remaining 3 degrees of freedom - two translations and one rotation. Datum B would be an axis down the middle of the cylindrical simulator, which would be at the same basic angle relative to A.

The datum reference frame, which is three perpendicular planes, would be established in the set of simulators (the plane and the cylinder). It would be logical to put the DRF origin at the intersection of Datum Plane A and Datum Axis B, and orient one of the DRF planes to Datum Axis B. But there is no rule stating this - the DRF origin and axis directions could be specified anywhere the designer wishes (by basic dimensions and angles).

The key thing, that Y14.5 doesn't make completely clear, is that the datums don't have to be mutually perpendicular but the planes of the datum reference frame do.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

### RE: Secondary datum feature must perpendicular with primary datum feature?

axym: Well said.

Certified Sr. GD&T Professional

### RE: Secondary datum feature must perpendicular with primary datum feature?

Good discussion -- I stepped away for a day and lots of other comments have rolled in. Evan expressed my initial thinking in the best way. Datum features don't have to be perpendicular to one another, and the derived datums from each don't have to be perpendicular. But the planes of the DRF do need to be. (That's why I referred to Fig. 4-7.)

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

### RE: Secondary datum feature must perpendicular with primary datum feature?

SeasonLee,

I agree with 3DDave that Datum feature B will probably not provide effective rotational constraint, despite being capable in a theoretical sense. To remedy this, you might consider referencing A as primary and B-C as secondary. Based on your original attachment, perhaps this is more functionally meaningful and closer to what you had in mind at the start. I would argue that this method adds some ambiguity, but that would likely be outweighed by the the practical benefits of improved rotational constraint.

On the more general subject of datum reference frames, planes, and so forth:

I've never seen much point in worrying about three mutually perpendicular planes of a datum reference frame. All that matters is the relationship between the theoretical part geometry and the actual part geometry. A Cartesian coordinate system is not required to define that relationship.

pylfrm

#### Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

#### Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Close Box

# Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

• Talk To Other Members
• Notification Of Responses To Questions
• Favorite Forums One Click Access
• Keyword Search Of All Posts, And More...

Register now while it's still free!