Pos Tol applied to pattern (how to?) 6

[fcsuper]

I have a X,Y grid pattern of holes laid out flat plate. Datum A is on the facing surface. Datum B runs through the part in the X direction. Datum C runs through the center of the part in the Y direction. (See the attached drawing.)

I wish to apply positional tolerance to the holes of the pattern.

What's needed, what's basic, what's reference?

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[KENAT]

The equally spaced issue comes up frequently on here (you must have seen it) and while not forbidden in ASME Y14.5M-1994 I'd strongly suggest using 49 X .500 BASIC instead.

Arguably the 1.750 Dimensions are not required and if given should be either basic or reference (depending on if you really think you need them).

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[PaulJackson]

Lawyers only ask questions in court that they already know the answer to. fcsuper, you indeed are quite experienced in this stuff... why don't you tell us what you think is the answer to your question?

Paul

 

[ringster]

Personally, I'd need to know more about the function and assembly before making any judgements as to correctness or what is needed vs ref dimensions.

"Equally spaced" is entirely redundant and should not be used. I would likely question the use of the mid planes as datum features for this application, without further insight. Tol on hole location vs hole tol? Seems tight.

Kenat,

You might want to rethink the 49X. I believe 7X in a vert and horiz direction to be more appropriate.

 

[ctopher]

It is complicated to dim off of the theoretical CL.
The tol needs to be added for overall length of part, CL to 1st hole, each space between holes, then position of each hole.
To add these up and then equal 3.500 could be confusing for the machinist and inspector.
I agree that the equally spaced is redundent. I would use only the 7X. The 1.75 dim should be basic.
Which dimensions are needed is hard to tell without knowing the design.
Some examples:

http://engineering.dartmouth.edu/mshop/pdf/introdr.pdf

Here is a good site to find docs:

http://www.scribd.com/doc/9299264/Geometric-Dimension-Ing-and-Tolerancing

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 08; CATIA V5
ctopher's home (updated Aug 5, 2008)
ctopher's blog
SolidWorks Legion

 

[fcsuper]

I'm looking for input because a scenario similar to this one is coming up for me, and discussion regarding this has been ...well... anyway, I am looking for some input.

I cannot get too much into function. The holes line up with an array that hovers over one whole row at a time. I will say that we are concerned with how the holes are located to each other and don't much care about the edges, per se. There are two alignment pins that locate the plate (I didn't model those for the example). One pin is the upper left, one in the lower right. The part is designed to be symmetric about the C datum for handling.

I don't see much issue with using the centers of the part as the datums, as this is not difficult for machinists to handle, though I may have to take additional action to allow the part to be more readily inspected.

Futher input is welcome.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[powerhound]

You shouldn't have a dimension from the center of the part to the outer row of holes. When you locate the hole pattern symmetrically within the datum structure it is implied centered. The 1.75 can be reference at most.

If the hole pattern itself is more critical than where the pattern lies in the plate then consider composite position. Be sure to consider position at MMC if applicable. The only other thing I would change goes along with KENAT and that is making your .500 dimensions basic. Using equally spaced can become a bear but if your "equally spaced" is applied to basic dimensions, then most of the problems associated with it go away. I'm still not a fan of it though.

Powerhound, GDTP T-0419
Production Manager
Inventor 2009
Mastercam X3
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[PaulJackson]

The primary needs a form control.
The secondary needs a size constraint and an orientation to the primary.
The tertiary needs a size constraint and an orientation to the primary and secondary.
The hole, countersink, and angle each need a tolerance.

The descriptions of the basics work ok as they are but could be improved by making them more universally interpretable by losing the English language and boxing the value. There are a number of ways to do that… try one and bounce it off this group and count the yeas and nays.

I don’t see any problem with the opposing surfaces of the length and width being the secondary and tertiary datum features and referencing the basics from the center plane… if that mirrored function… but since you revealed that two alignment pins actually do the job of a pattern secondary (two round pins in two round holes) or a secondary and a tertiary (two round pins in a hole and a slot)… I would put the dimensional reference line at the center of the 2X pattern or the center of the 1X secondary… tolerance the pattern for position to A (which controls both orientation and spread) or the orientation of the secondary to A then the position of the tertiary to A|B. Then I would tolerance the edge profile “all around” from A|B or A|B|C which ever reflects function. Naturally the basic dimensions to the holes - countersinks would originate from that functional reference as well.

Paul

 

[fcsuper]

I'll post an updated image tomorrow with more detail.

I'm also been asked to completely disassociate the edges of the part from the pattern, without the use of locating pins or any other reference. The suggestion was to use "witness holes" to control the location of the datum. Am I completely off base by seeing this as trying to create a datum by using the very features created to mark the datum (chicken-and-the-egg self referencing).

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[PaulJackson]

I horsed around with your picture... this is using the locating pins as a secondary 2X pattern.

Paul

 

[PaulJackson]

or... if there were no locating pins.

Paul

 

[btrueblood]

"The primary needs a form control.
The secondary needs a size constraint and an orientation to the primary.
The tertiary needs a size constraint and an orientation to the primary and secondary.
The hole, countersink, and angle each need a tolerance."

Why? Maybe the "straight line rule" is good enough. By the above logic, the holes should also have a circularity tolerance, and perhaps cylindricity as well...unless the block tolerances (unknown) are "good enough". More importantly, none of that really addresses fc's question.

FC, you say the hole-to-hole pattern is what's important, and the outer form of the part could presumably be toleranced by +/- dimensions if you wanted to. My understanding of the best (least confusing) way to control the hole pattern is to call out one or preferrably two holes as new datums (say D and E), given position tolerance by whatever means you want (GDT or +/-), at whatever tolerance is required, to your existing datums A-B-C. The two new datum holes, perhaps one at each corner of the pattern, are then used to control position for the remainder of the pattern.

Callout would look something like:

64x 0.150 THRU (etc. for c'sk)
TP/0.5 dia/A/B/C
TP/0.001 dia/A/D/E

 

[btrueblood]

Ok, Paul, I like your 2nd picture a bit better, and explains why you would use form control on the outer boundaries.

I seem to be parroting Kenat in my first reply, and apologize, he said it better elsewhere in this forum:

"Were they a complete engineering drawing fully complying to the standards, no. However, when looking at illustrations on this site (as opposed to actual drawings/extracts of drawings) I wish we could treat them like figures in 14.5 and invoke paragraph 1.1.4. It even states "EXAMPLES ARE INCOMPLETE BY INTENT" in the top left."

 

[fcsuper]

Paul,

Yes, my picture was incomplete by intent. Some of the details added to my pics in your edits already exist on the actual drawing. I don't think I need to profile the edges to get the assumed perpendicularity or to control their size.

Regarding the first example, I misspoke before (a little bit). Although there are corner pin holes (as you drew), only one is engaged at a time (and it doesn't matter which one that is). The part is located with two pins, the second being in a slot across from the locating hole. So, I don't know if I could meaningfully reference the pin holes without making one arbitrarily important over the other. Your example would create two different datum B's, if I'm reading it correctly.

I guess the question on both your examples is whether the pattern itself should be basic? Also, your method of calling out the pattern is more intuitive to my needs, however Y14.5 says it this way "7X .500 (=3.5)" Even if I turn this basic, it concerns me that the .500 is the driving dim when I'm more interested in controlling the overall pattern of 3.5. Thoughts?

Blue,

"The two new datum holes, perhaps one at each corner of the pattern, are then used to control position for the remainder of the pattern."

The difficulty I'm having with this (and I've considered this) is that I cannot think of any way to logically fixture the part to the holes I would declare as the new datum while maintaining the symmetry of the pattern.

Both Blue and Paul,

In Paul's example two, what is the purpose of declaring the hole itself as datum B? I'm having trouble understanding how example two provides a complete callout.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[PaulJackson]

fcsuper,

In both mark-ups that I offered datum feature B is not one of the 2X locating pin holes at a time... or one of the 64X features. It is the entire pattern simultaneously either 2X or 64X.

paul

 

[ewh]

... it concerns me that the .500 is the driving dim when I'm more interested in controlling the overall pattern of 3.5.

As long as the dimensions are basic, it makes no difference, as they would add up to 3.5 basic.

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[fcsuper]

Paul,

I guess where I'm confused is in how datum B is being used in each of the examples.



Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

 

[KENAT]

So sticking to the initial question/information and not going into the location by dowel pins or whatever.

I was wrong, it's 56 X .500, 8 rows of 7 'gaps'. 7 X is just plain wrong to me because there's more than one row/column, unless you want to add the lines through the center of all the holes, which seems messy.

As to the centering 1.75 like Powerhound I believe the 'implied centered' is in effect so I'd probably leave it off. If you do put it though I'd prefer as reference but, I don't think making it basic is explicitly 'wrong' but is probably unnecessary and could perhaps be argued double dimensioning but I’m not sure that holds water. This is what I meant in my initial brief post.

Your concern over the total size of pattern 3.5 being more important than the individual .5 pitches doesn’t seem too relevant if you are using position since there is no stacking of tolerances.

Given that you effectively ‘swap’ your primary functional datum from one hole to another during operation it does mean using the hole & slot as secondary and tertiary datum’s gets a bit more complicated. Maybe making one of the holes in the pattern, or even the entire pattern the secondary datum makes sense and then using the common ‘slot’ as a clocking feature and relating the other 2 pin location holes back to this. This is getting out of my comfort zone though.


KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[btrueblood]

"what is the purpose of declaring the hole itself as datum B?"

As Paul said, the datum is being defined as all of the holes together, i.e. the pattern. By defining the pattern as datum B, the profile tolerance then has something to work from.

"The difficulty I'm having with this (and I've considered this) is that I cannot think of any way to logically fixture the part to the holes I would declare as the new datum while maintaining the symmetry of the pattern."

Not clear what you mean. It sounded like you want the holes in a tightly controlled pattern (tighter than the rest of the part features), which makes any two holes a perfect place to drop the part over a couple of fixturing pins. Maintaining the symmetry of the pattern - pick a pair of holes that straddle the centerline if you want, or from corner to corner...the symmetry of the pattern is defined by the [7x 0.5], and the "lines drawn perpendicular are to be interpreted as perpendicular"...the pattern is symmetric because you drew it that way, not because of the location of the datum you chose to control it.

 

[fcsuper]

"As Paul said, the datum is being defined as all of the holes together, i.e. the pattern. By defining the pattern as datum B, the profile tolerance then has something to work from."

Ah. Ok. Maybe I mistated something. First, I'm not going to use the profile. Second, the relationship of each of the instances within the pattern to each other is my main concern. In that effort, I arbitarily choose the part center of the datum to give these a common frame of reference. Is there a better way?

I was going to put a new example, but I'm not ready it, I think, until I get some clarity on this. Sorry if I'm being dense. Everyone's assistance so far has been very much appreciated. (Patterns suck)

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group