Question about ISO_13920 2

[SteelAndSteel]

It is possible to make a welding drawing of the item in the picture without adding any measurements, (if you have drawings on details 1 to 4), because the items is aligning with the edges of item 1. So what I wonder is who I shall interpret the general tolerance concerning distance in this case, because the distance is 0 from the edges.

 

[MintJulep]

Your drawing likely has everything at nominal dimensions and form,and thus every edge lines up perfectly to your design ideal.

However real parts will have deviation from nominal, and therefore not all the edges will line up perfectly.

You need to add dimensions to define what to do with real parts when they don't line up perfectly.

 

[SteelAndSteel]

But i think that should be covered by ISO_13920 i am just not sure if i interpret it rigth, in the table for linear dimensions it only cover the distance from 2 to 30, not from 0 to 2? But of course i get angular dimension, straightness, flatness and parallelism from ISO 13920.

 

[CheckerHater]

Nobody is holding gun to your head forcing you to limit yourself to ISO 13920
It all starts with the function. You have to figure out which tolerances will make your part work. Then you compare them with ISO 13920 tolerances. If ISO tolerances are acceptable, you specify them on the drawing. Your part may require tighter tolerances, and then you’ll have to specify them on the drawing in addition to ISO tolerances or instead of them.
It is possible that part could be produced cheaper but will function well with some tolerances being larger than ISO 13920. In this case larger tolerances may be specified on the drawing as well.
No, 0 inch/mm, 0 degrees, 90 degrees and so on are usually implied and not specified on the drawing.
Nevertheless it doesn’t mean your drawing shouldn’t have any dimensions whatsoever.
Your dimensions specify how your part looks and works after being assembled.
So, position of center of your round element may be important. It affects assembly with other parts.
Some companies require specifying envelope dimensions, especially for large heavy parts – it affects packaging and transportation.
At the end of the day you are the boss – you know how your part works and it is your goal to produce the drawing the way that nobody will come back to you to ask questions.
Good luck!

 

[SteelAndSteel]

Well, i base my design on earlier drawings that someone else has done and those drawings had no dimensions and aparently that has worked, to change that i need to make an argument and that is why i asked about how i shall interpret ISO 13920 when there are surfaces who are flush with no Linear dimension.

But of course the way you describe it is the correct way to do it, but sometimes you are not the king of the circumstances.

 

[3DDave]

You don't need an argument to change, you need careless fabricators who don't care about repeat business.

As long as the variation from ideal that is produced by your fabricator is smaller than the application has tolerance for, there will be no motive to change. It is only when items that meet all the drawing requirements fail to fit or function that there is a need to change the drawings.

 

[MintJulep]

So here's the reality.

Specifies general tolerances for linear and angular dimensions and for shape and position of welded structures in four classes, these being based on customary workshop accuracy.

In other words: ISO 13920 codifies exactly what happened before ISO 13920 existed.

Basically "Line 'er up as best you can and weld the sucker"

So in effect specifying "Tolerances per ISO 13920" is functionally the same as specifying no tolerances at all.

It just makes you feel good because "I called out a standard."

 

[CheckerHater]

MintJulep, you forgot to quote Clause 7:

"Non conformities
A decision on the acceptance of components not complying with this standard may be made on the basis of the suitability for their intended purpose."

 

[KENAT]

Oh no we're not in this territory again are we - see earlier threads about ISO 2768:-(.

Posting guidelines faq731-376

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

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[SteelAndSteel]

Ok, i have attached 2 pictures one is the drawing and the other one is who the item ended up after manufacturing. Can you say based on Clause 7 that the manufacturer has done an error here?
i hope this isn't a stupid question.

 

[SteelAndSteel]

The drawing

 

[pmarc]

This isn't stupid question at all.

The picture with actual geometry shows incompleteness of the weldment drawing pretty nicely. The fact that the drawing shows two plates aligned to the left does not automatically mean they will always be welded this way in reality, thus the drawing should clearly specify the maximum amount of misalignment allowed. This value, which is a locational tolerance, can't be found in ISO 13920 (according to the second paragraph in clause 4.3).

That is why I don't think Clause 7 should be used to say that manufacturer made an error or not. Regardless of whether we like the wording (and presence) of the clause or not, it applies when actual geometries of welded parts do not comply with the content of the standard. However, since the standard defines no general tolerances for location, there is nothing to comply with with regard to this characteristic.

So did manufacturer make an error? No, he did not. He delivered part satisfying all requirements defined on the weldement drawing (assuming that the measured distance was 10+/-0.5, not 10+/-1). It is not his fault that the requirements were poorly defined.

 

[CheckerHater]

pmarc, could you elaborate on ISO 13920 NOT specifying locational tolerances?

 

[pmarc]

What would you like me to elaborate? The second paragraph in clause 4.3 makes it clear, doesn't it?

 

[CheckerHater]

The second paragraph in clause 4.3 talks about "tolerances of form and position, e.g. coaxiality and symmetry"
I am talking about OP drawing showing dimension 10±1. It's linear dimension (it isn't angular, is it?) and linear dimensions are covered in Para. 4.1 and Table 1.
How come ISO 13920 "defines no general tolerances for location" then?

 

[pmarc]

In my world location of items as shown in OP's drawing is controlled by position or profile tolerance, and not by geometrically ambiguous linear dimensions (not to mention any kind of general tolerances). Position and profile are locational tolerances, and because there is nothing about position and/or profile in ISO 13920, that is why I said: "the standard defines no general tolerances for location". I agree, the statement was too general. There are situations where locational relationship between two or more items can be controlled by simple linear dimension(s) - it is just that this case is not amongst them.

As for 10±1 dimension, this dimension is not on the drawing posted 22 Mar 14 12:30, so I do not understand why you want to apply general tolerance from Table 1 to it.

 

[CheckerHater]

If I understood this correctly:
On OP's drawing vertical part is shown aligned with the edge of horizontal part.
The fabricator produces part where vertical part is 10 mm away from the edge.
OP believes it should be 0 (within some tolerance)
What all of it has to do with position and profile?

 

[pmarc]

OP believes it should be 0 (within some tolerance)

So what does "within some tolerance" exactly mean? Where is this tolerance defined?

What all of it has to do with position and profile?

If two surfaces are shown coplanar, like on OP's drawing, is it enough to just show them like that and do nothing more, or should there be an additional tolerance specifying how much off the coplanarity these two surfaces are allowed to be? If I want to use geometric tolerances for that (because I believe that linear dimensions are ambiguous, and because I know that Table 1 in ISO 13920 does not give me general tolerance for linear dimensions less than 2, what should my first choice be? Since I work to ISO, I am choosing between position and profile (they basically mean the same according to that standard). Thus position and profile in my previous reply.

 

[SteelAndSteel]

Pmarc, so it is 10+/-0.5 because for parallelism?

Another question

In clause 4.3 it say , the following table 3 apply to both overall dimensions of a weldment, a welding assembly, or a welded structure, and also for sections for which dimensions are indicated.

i am a bit unsure who i shall interpret the term overall dimension

In a dictionary it says that overall means:
from one extreme limit of a thing to the other: the overall length of the bridge.

 

[pmarc]

Yes, 10±0.5 is because of parallelism.
If you want to meet the parallelism tolerance specified in Table 3, the absolute difference between measured distances shouldn't be greater than 1 (although one can rightly say this isn't 100% true).

As for "overall" dimension, yes, this is how I would interpret it. Something like total length, total height, etc. Is this what the authors of the standard had in mind? I have absolutely no idea.

Regardless of that, I think you already noticed how many problems hide behind use of standards that deal with the animal called "general tolerances". It actually does not matter whether we talk about welded (ISO 13920) or machined (ISO 2768) parts - these standards simply leave so much space for different interpretations that it is practically impossible to avoid ambiguity, even for such a simple part like yours.

One of the fundamental rules in dimensioning and tolerancing states that dimensions and tolerances shall not be subject to more than one interpretation. Thus, if one really cares about following this rule, she/he should stay away from general-tolerances standards as far as possible.