Linear or Ordinate Dimensions

[macduff]

All,
What's preferred in today’s industry? Linear or Ordinate dimensions. I prefer Linear and guessing Machinist would like to have that way as well. Am I correct in this assumption?

I know there's no right or wrong answer, I'm just trying to get a feel what machinist would prefer.

Thanks,


Macduff
Colin Fitzpatrick
Sr. Mechanical Designer
macduff's SW page
Inhouse System
Pentium(4)2.80GHz
Ram 1.00 GB
SW2005 Office SP 3.1
Windows 2000 SP4.0
NIVIDA Quadro4 750 XGL

 

[ajack1]

Most machinists I have talked to prefer ordinate, especially boxed, a list of X and Y dimensions with a description, far easier to work to rather than matching up linear dimensions.

 

[MechCT]

Ordinate is fine in some situations... But don't forget that sometimes your drawing will have a standard tolerance for dimensions over a certain length and you can really screw up your part by dimensioning that way if you're not careful.

I think ordinate dimensioning is fine if you have a simple part that doens't require a high degree of accuracy applied to it's tolerances. Or if you have a drawing that either doesn't have any general tolerances or says something like, "ALL LINEAR DIMENSIONS ARE ±0.5, UOS."

 

[chancey]

The biggest problem with ordinate is that you get tolerance stack up. Our toolmakers like ordinate so that is what we give them but when we have to keep feature to feature relations tight we will use a linear dimension. So you end up with a drawing mostly in ordinate with some linear dims.

 

[ajack1]

Why do ordinate dimensions stack up if you come from a common datum, surely it is linear dimensions that stack up, if you do not come from a common datum?

 

[ewh]

You measure from an edge to a hole. One tolerance. You do the same for another hole. Another tolerance. These holes may have an important relationship to each other, so if you dimensioned directly between the holes, you eliminate one of those toleances.

 

[ewh]

The exception to the above is if you used basic dimensions.

 

[ajack1]

Ewh that is true if you have only two holes but if you have say four, dimensioning from one to the next stacks up to a greater tolerance does it not?

If they matter surely putting a positional tolerance or MMC solves that.

 

[ewh]

I would use basic dimensions and a positional tolerance, but if not, I would dimension from one hole to the others. You are right that a tolerance stackup would occur with more than two holes in my example.

 

[TheTick]

I prefer toolmakers that can add and subtract as needed and read prints across the full scope of acceptable print standards.

I make things easier to read for machinists when I can, but design comes first.

 

[ajack1]

I would have thought that making drawings as easy as possible to read was a very important part of good draughting, regardless of whether toolmakers can add or subtract.

 

[MadMango]

Drawings need to be clear and unambiguous, and fully define the design intent of the item depicted. I hate when I see prints on the shop floor with math scribbled all over it, disregarding the manner in which it was dimensioned and toleranced.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[ewh]

Making drawings that are easy to understand IS a very important part of drafting. While making it easier for the fabricator to make the part is important, it is more important to define the part so that it functions as intended.

 

[ctopher]

How a part is dim depends on the part design. You have to look at it how you think it will be made. Any type of dim will work if it's easily interpreted by the machinist and inspection. Also important that the dwg is clear to read even when it is reduced in size.
Having some knowledge of the machine shop and inspection practices helps.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[drawoh]

I prefer ordinate dimensioning on anything complicated, primarily because it is more readable. I have seen some truly hideous linear dimensioned drawings.

Tolerance stack up is irrelevant for me because all my holes are located by positional tolerances. As noted above, if you have more than two holes, you have tolerance stack up, ordinate dimensioning or no ordinate dimensioning.

Sometimes, on simpler parts, linear dimensioning is clearer, in which case I use it.

JHG

 

[ringman]

IMO, ordinate dimensioning lends itsself to panels and such that tend to have holes that are not necessarily related. To use ordinate dimensioning for machined parts having critical patterns or sets of patterns is inviting problems.

 

[wgchere]

I disagree, Ringman. When you use ordinate basic dimensions the GD&T holds all pattern relationships for you.

Tolerance stackup--there isn't any. Stackup is when one feature is dimensioned off another feature, thus inheriting the other feature's tolerances and any tolerances that feature inherited, or in an assembly situation where parts are adding their tolerances to the whole assembly's tolerances. If all the ordinate dimensions (cartesian) have the same tolerance then ANY two features have the same tolerance between them. True, the tolerance of each feature is added to the tolerance between them, but this isn't stackup—you'll get the same thing with baseline dimensioning but take up more space. If this is a problem, adjust the tolerance so it adds up to what you want. Or use GD&T.

 

[ringman]

After some thought, I am back to where thought I was. I still do not believe there is a case that would justify the use of ordinate dimensions having a positional tolerance applied to the lot of them such as is shown in the standard (FIG.1-50), which defines the method for ordinate dimensioning, (but not necessarily in conjunction with positional tolerances).

If there is such a case, might someone be so kind as to briefly describe for the benefit of all of us.

 

[wgchere]

Ringman,

Fig. 1-50 has no positional tolerance, just cartesian tolerances (± applied to the dimension). By the by, I have never used the method in Fig 1-50, I usually use Fig. 1-49 and callout holes with a leader. and usually all my ordinates are basic and my hole coding is applied to patterns, not sizes, 6X Ø.138 THRU, HOLES "A" [GD&T FRAME], 5X Ø.138 THRU, HOLES "B" [GD&T FRAME], and so on.

I see ordinate dimensioning as just another of many tools for accomplishing the task of clearly transmitting the information to manufacturing and inspection.

wgchere