Smart questions
Smart answers
Smart people
Join Eng-Tips Forums

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

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

Join Eng-Tips
*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.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

joejack7 (Industrial) (OP)
26 May 05 16:38
I need some justification on this one...

Our design group has released a drawing with a hole pattern on a sheet of honeycomb core.  The requirement of the holes is TP of .250 LMC relative to Datum 'A' only.  Datum 'A' is the planar surface of the honeycomb.  My question is...

In controlling the position of these holes is it possible to control to only 1 datum feature?  Since this is concerning POSITION with basic dimensions, doesn't there need to be more controlling features, ie.  datum 'B' and/or 'C'?

Joe Jackson
Goodrich Corporation
Helpful Member!  ajack1 (Automotive)
26 May 05 16:48
A planar surface is not enough as it ensures the part is flat but does not control rotation.
joejack7 (Industrial) (OP)
26 May 05 16:56

I thought so, but when addressed the design group claimed that the pattern itself is "floating", but...

doesn't that contradict the purpose of positional requirements?

Also, 2 of these locations are for an insert with a drilled hole.  They did the same for these 2 holes, TP relative to only datum 'A', then labeled these 2 holes as Datum 'B' and 'C' respectively.  They then use these datum features (holes) to control surface profile of the exterior features!  

My point is...if the hole pattern itself is not contained as you stated, then how are these 2 features gonna control any other features of size?

BTW...  Is there any justification of this principal in the ASME 14.5, I am not finding any direct explanation of this?

Joe Jackson
ewh (Aerospace)
26 May 05 17:09
I could understand using datums B & C to control the other holes, but I am not sure what you mean by "exterior features".  It could mean that the perpendicularity of B & C are controlled relative to A, then the edges are cut relative to B & C.  This does seem backwards, but perhaps in the fabrication process it is simpler to trim the piece after drilling, in which case this would work.
Helpful Member!  Heckler (Mechanical)
26 May 05 17:13

If you want to control the position of the feature pattern as well as the features within themselves then use a composite true positional callout.  In ASME Y14.5M it spells out all the typical uses of TP.  I suggest getting the spec and validate this foryourself.  Is this honeycomb core being machined in its solid state or stretched out?  If you're designing a panel the features for potted inserts are usually completed after the panel has its face sheets assembled.
joejack7 (Industrial) (OP)
26 May 05 17:32
More detail...

The honeycomb is being fabricated as a sandwich with inserts for drilling.  The drilling of the 2 "datum Holes" are for post fixturing on a fixture plate for CNC routing.  The drawing is for the build of the sandwich with 2 holes drilled after cure for location purposes.

The problem is that the design is trying to control the location of the plugs and the 2 holes in relation to the plug centers with the TP callout relative to only 1 planar datum 'A'.

My problem is that I am trying to convince or justify, if you will, to the designers that this is a problem in manufacturing.  I need justification from the ASME standard or likewise stating that it is theoretically impossible to control the position of the plugs and holes relative to 1 planar datum!  Any help with that is greatly appreciated.  Thanks.

Joe Jackson  
Heckler (Mechanical)
27 May 05 1:25
After thinking about what you're doing it sounds logical.  The two holes will become datum holes for the rest of the features on the honeycomb panel, correct?  So basically, these two holes are not critical in the X & Y axis (which is why the TP .250 LMC) just the z axis which is datum A all this with respect to the machine milling the holes.

The rest of the features will be referenced from these tooling holes?  Unless I'm missing something this sounds okay to me.
RedPen (Mechanical)
27 May 05 17:41
I agree with Heckler and ewh, it is somewhat unusual but not incorrect.  It may be incomplete, which seems to be the problem you have with the single datum reference.

If I understand your example correctly, Datum A is one planer surface of the panel.  The two holes (Datums B & C) are drilled into this surface.  The existing callout controls perpendicularity to the surface (A) but doesn't control their location relative to the plugs for instance.

If the other dimensions and tolerances are such that the holes are located "close enough" to the center of the plugs to work and the holes are perpendicular to Datum A, then all subsequent features can be relatively precisely located using all three Datums A, B & C.

Sr. Designer/Lead Checker
drawoh (Mechanical)
27 May 05 18:29

   The ASME standard explicitly calls out a procedure where you have two positional tolerances for the hole pattern.  One pattern is from datums A, B and C, and allows a TP tolerance of something like 1.5mm dia.  The other is from datum A, and allows 0.25mm dia.  The indication is that you are not concerned about the location of your pattern, but you are concerned about the location of the holes with respect to each other.  

   Your pattern cannot just float.  There has to be a tolerance for how far it can float.

ringman (Mechanical)
27 May 05 19:14
After reading all the inputs sofar and thinking on it,  it appears that the drawing is attempting to combine detail and assembly information on a single drawing.

It would seem that the holes can be related using LMC with relation to the OD of the insert.  That would of course require a datum/s relationship for that purpose. A separate detail would best serve that purpose.

At the assembly level, the  two holes could then be used to establish a datum reference frame for the location of other holes or the perimeter of the honeycomb panel.
tunalover (Mechanical)
28 May 05 12:16
Form-critical printed circuit boards or ceramic substrates for hybrid circuits are sometimes dimensioned such that the outline of the part is controlled relative to a single datum hole.  The dimensions from the hole to the boundary are basic and the boundary is controlled with a profile tolerance.  The primary (perpendicular) datum would be 'A' for example.  The secondary datum is assigned to the centerline of the hole, call it 'B'.  In this case only TWO datums are required to adequately define the boundary and the datum hole.  In this example, the two mutually perpendicular datum planes passing through the hole's axis are free to rotate and a third "clocking" datum is not required.  A second hole can be located relative to the first also using datums A and B in this way where a third clocking datum is not required.

I hope this helps!


ringman (Mechanical)
28 May 05 14:42

The last sentence of your explanation is confusing.  Can you reviiew to see if error in typo or something.

Also I am a little lost on your explanation of not requiring a datum for orientation on the circuit board.

tunalover (Mechanical)
28 May 05 17:21
A picture would really do the trick here but I'll try to explain without it.  The second hole would have a feature control frame (fcf) that would invoke a datum reference frame formed by 'A' (a datum plane) as the primary datum and 'B' (two datum planes aligned with the axis of the first hole) as the secondary datum.  This datum reference frame leaves one unconstrained degree of freedom but, unless clocking is desired, it is not needed.  Sufficiency is the key;  if it's not necessary to invoke a tertiary datum then it should not be invoked.


ringman (Mechanical)
28 May 05 18:48
I guess I am having trouble in the concept of a circuit board or ceramic substrate that would not require 'clocking'. Datum B would result in an axis through which the mutually perpendicular planes MAY be passed, but, without orientation are useless.  

Is there something that I am missing on the geometry of the hypothetical circuit board in the discussion?
tunalover (Mechanical)
28 May 05 20:24
Without clocking, the inspector can choose any angular orientation of his choosing.  It's simply a matter of how he wishes to establish his coordinate system.  The tolerances don't care what coordinate system is used or how it is oriented!  Normally, the dimensioning method will infer the coordinate system (often with one datum running lengthwise of the board and the other at a right angle to it).


ringman (Mechanical)
28 May 05 21:30
Perhap I have erred in thinking that Y14.5 was applied on this drawing.

I don't believe that Y14.5 allows inference of a feature as a datum.
tunalover (Mechanical)
29 May 05 9:54
Indeed it does.  Y14.5 does not require that all degrees of freedom of a datum reference frame be constrained in a feature control frame!


ERE (Structural)
29 May 05 22:30
A feature control frame does not have to have three datum references. However, if a feature is used as a datum, it is to be identified with a datum reference symbol and cannot be implied. Implied datum features went away with the 1982 standard. Datum reference symbols identify physical features and are not applied
to center lines, center planes, or axes. (See Y14.5M-1996 section 4.3.2)

RE the circuit board problem...  if only the two datums are established, you are saying that the rotational orientation of the part is not important. (See Y14.5M-1996 section 4.4.2) If the dimensioning scheme allows the inspector to infer the datums, why not just identify the datums and remove any ambiguity?


tunalover (Mechanical)
30 May 05 10:07
Who said that a feature can have an implied datum?


ringman (Mechanical)
30 May 05 11:59

Perhaps this is the statement that is the cause for misunderstanding.  

"Normally, the dimensioning method will infer the coordinate system (often with one datum running lengthwise of the board and the other at a right angle to it)."

I believe that myself and others have interpreted it to be accepting of implied datums.

ringman (Mechanical)
30 May 05 12:24

Im not sure whether you got the justification you were looking for or not.  After reading through the responses and trying to digest, I think not so.

I am of the opinion that for you to apply LMC to ANY hole, a datum, other than a plane would be required. (Or in addition to). This will effectively control the minimum edge distance. Do you agree and has that portion of your situation been resolved?

tunalover (Mechanical)
30 May 05 12:58
Ringman and others-
I should have chosen my wording more carefully.  I said:

"Normally, the dimensioning method will infer the coordinate system (often with one datum running lengthwise of the board and the other at a right angle to it)."

What I should have said:

"Normally, the dimensions will infer the coordinate system (often with one of the datum reference planes running lengthwise of the board and the other at a right angle to it)."

The datum assigned to the hole axis has two mutually perpendicular datum reference planes passing through the axis.  The best choice of the angular orientation of these two planes is inferred from the dimensions between the two holes.

drawoh (Mechanical)
30 May 05 13:27

   You are describing something I have done several times, mostly on sheet metal.  The extreme case is that I bent all four edges to work as gussets.  It is difficult to locate holes from an edge, and easy to locate the holes from each other, since they are located on the same flat surface.

   Datum_A is the bottom surface.  Datum_B is a hole, and Datum_C is a second hole.  The flanged edges are located by sloppy tolerances or a sloppy GD&T profile.  

   I can now register the part precisely, and inspect it.  

   The GD&T standard shows simple parts with circular shapes and hole patterns inside, where there are only two datums, the base and the OD.  This makes sense to me, as long as there is only one hole pattern, and there are no rotation issues.

   Once the geometry requires you to control rotation, you need three datums.

tunalover (Mechanical)
30 May 05 14:11
This practice is good for when one or more holes are created BEFORE the part outline.  In PCB manufacturing, the router often uses one or two holes as a reference when forming the outline.  This is amenable to "step and repeat" processes where many circuit boards are created from a large sheet of material.

You said "This makes sense to me, as long as there is only one hole pattern, and there are no rotation issues."

I've had many hole patterns on the same part dimensioned this way using 'A' as the primary and 'B' as the secondary.  There is no limit to how many hole patterns can be dimensioned this way in one part.


Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Back To Forum

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:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close