The Standard Says a Drawing Is Read by Assuming the Tolerance Is Exactly Right

This page gives no advice on tolerancing, drawing indication or measurement. It did not come from a forum question; it comes from listing the standards this site has already cited and looking for the one it had barely touched. The browser tooling we normally use remains unavailable.

Which document, and how much of it we could read

The standard is ISO 8015, Geometrical product specifications, fundamentals, concepts, principles and rules, second edition dated 1 June 2011, prepared by ISO/TC 213 and replacing a first edition from 1985. This is the first time this site has used it, and the first time it has used anything from that committee.

The free preview runs to nine pages and stops part way through clause 5. The contents page lists thirteen fundamental principles. Only the first three have any text in the preview. The other ten exist here as names only, and this page does not say what any of them requires. We also did not read the clauses on indication rules, the annex, or any of the standards this one references. We did not check whether the document is current.

One line from the introduction is worth having before anything else. Until this edition, it says, these principles “were implied, but not formulated explicitly”. Twenty six years passed between the two editions, by our arithmetic, and the rules everyone was already working to were written down at the end of it rather than the beginning.

Three things you are asked to assume

Clause 4 is headed fundamental assumptions for the reading of specifications on drawings, and its opening says these assumptions are “the basis for the overall rules of the GPS system”. Then come the three, each phrased as “it is assumed for interpretation that”.

First, that the functional limits come from an exhaustive investigation, by experiment or theory or both, so that they are known with no uncertainty. Second, that the tolerance limits are identical to those functional limits. Third, that the workpiece functions 100 % inside them and 0 % outside.

None of the three is true of any real part. Nobody knows a functional limit with no uncertainty, tolerance limits are set with margin and manufacturability in mind rather than being copies of function, and a part a micrometre over a limit does not stop working. The standard is not claiming otherwise. It says these are assumptions, and it says what they are for: interpretation. They are the fixed points that let a drawing mean one thing rather than several.

That is a more honest piece of drafting than it first looks. A document that needs a simplification in order to work can leave it unstated and let everyone absorb it by habit, or it can put it in a numbered clause where you can see it and know its scope. This one numbered it.

A standardized truth

The introduction defines its own vocabulary before it uses it. A concept is an abstract idea. A rule is a standardized procedure for action. And a principle is “a standardized truth” based on concepts, upon which rules are based.

Put that phrase next to clause 4 and the two sit oddly together. The principles are called standardized truths, and the assumptions they rest on are things the document does not claim are true. We are not saying this is a flaw. Standardized is doing work in that phrase: what is being standardized is the reading, not the world. But it is a useful reminder that in a specification, agreed and true are different words, and the document itself keeps them apart.

It applies whether or not you invoke it

The first of the three readable principles is about invocation, and it is broader than most people assume. Once a portion of the system is invoked in product documentation, the entire system is invoked, unless the documentation indicates otherwise. Using one geometrical symbol on one drawing brings in the rest.

And you do not have to say so. The clause notes that a marking such as “Tolerancing ISO 8015” may optionally appear near the title block for information, but is not required to invoke the system. A note gives examples of what comes along: the reference temperature from one standard, the decision rules from another. We did not read either of those, so nothing here describes what they say.

And yet nothing unwritten can be enforced

Two subclauses later, the definitive drawing principle says something that sounds like the opposite. The drawing is definitive, all specifications are to be indicated on it, and “consequently, requirements not specified on the drawing cannot be enforced”.

Then, in the same subclause, the standard says its own principles and rules apply to all product specifications where the system is invoked, “even though it is not explicitly referenced in the drawing”.

Read together, our reading is that the two sentences are about different things. What cannot be enforced unwritten is a requirement: a limit, a characteristic, a demand on the part. What applies unwritten is the interpretation: how the marks that are on the drawing are to be read. The standard does not spell out that distinction, and we are stating it as our reading rather than as a quotation.

If the reading is right, it is the practical sentence to take away from the whole document. You cannot be held to something you did not write down. You are held to a particular reading of everything you did write down, whether or not you named the system that supplies it.

The same subclause adds one more thing worth knowing: a drawing may carry specifications for several stages of completion, and where it does, each indication has to say which stage it belongs to, unless it is the final stage. The final stage is the silent default.

Four words that mark a default

The terms clause carries a note that is worth more than its size. Where defaults are defined, it says, they are usually recognizable by the introductory wording “unless otherwise specified”.

This site has quoted that construction, and its cousins about agreement between customer and supplier, in a long series of unrelated standards: in the delivery condition of bolts and nuts, in the diameter tolerance of a taper pin, in the ordering information for welded studs. Each of those pages read the phrase as the hinge it plainly is. Here is a fundamental standard saying that is exactly what it is: the marker that tells you a default has been set and can be moved. Those pages are linked rather than repeated.

Thirteen principles, three of them readable

For completeness, the contents page names all thirteen: invocation, GPS standard hierarchy, definitive drawing, feature, independency, decimal, default, reference condition, rigid workpiece, duality, functional control, general specification, and responsibility.

The preview gives text for the first three only. Some of those names are suggestive and some are not, and we are not going to guess at any of them from the name. What the list does show is the shape of the thing: a document whose job is to state, once, the assumptions the rest of the system runs on.

One smaller observation while we are in the front matter. Two of the normative references carry a footnote reading “to be published”. This is the third time this site has found a standard pointing at a document that did not exist yet, on three unrelated documents from three different decades; the earlier pages are linked rather than restated, and we are flagging the repetition rather than presenting it as a discovery.

What this settles and what it does not

• Reading a drawing assumes the functional limits are known with no uncertainty, that the tolerance limits are identical to them, and that the part works 100 % inside and 0 % outside
• The standard calls these assumptions and says they are the basis for the overall rules of the system
• Invoking any part of the system invokes all of it, and no marking near the title block is required to do so
• Requirements not specified on the drawing cannot be enforced, while the interpretation rules apply without being referenced there
• That the two sentences are about requirements and interpretation respectively is our reading, not something the standard spells out
• Where a drawing covers several stages of completion, each indication must name its stage unless it is the final one
• The wording unless otherwise specified is the recognizable marker of a default
• The introduction calls a principle a standardized truth, and defines concept and rule alongside it
• Twenty six years separate the two editions, by our arithmetic, and the introduction says the principles had been implied but not formulated explicitly until this one
• Ten of the thirteen principles are names only in the preview, so this page says nothing about what they require
• We read none of the referenced standards, none of the indication clauses and no annex, and did not check whether the document is current
• No advice on tolerancing, drawing indication or measurement is given

The habit worth taking is to ask, of any specification you are handed, what it is assuming in order to be readable at all. Most documents do not tell you. This one does, in three sentences, near the front, and the assumptions turn out to be the ones a buyer and a supplier most often disagree about later.

A later page found a general standard from the same committee doing the same housekeeping, removing rules that had never been stated at all.

This page covers step 6, the documentation. The whole order is substrate, thread, head, drive, finish, documentation, and why doing it out of order is rework rather than a tweak is in specifying a screw.

Common questions

What does the standard assume about tolerance limits?

That they are identical to the functional limits, and that the functional limits themselves are known with no uncertainty. Both are stated as assumptions made for the purpose of interpretation.

Does it really say a part works 100 % inside and 0 % outside?

It says it is assumed for interpretation that the workpiece functions 100 % within the tolerance limits and 0 % outside them. It presents this as an assumption underpinning the rules, not as a claim about how parts behave.

Do I have to write the standard number on my drawing for it to apply?

No. The invocation principle says that once a portion of the system is invoked in product documentation the entire system is invoked, and that a marking near the title block is optional and not required to invoke it.

Can a requirement that is not on the drawing be enforced?

The definitive drawing principle says requirements not specified on the drawing cannot be enforced. The same subclause also says the standard applies even though it is not explicitly referenced in the drawing. Reading the first as being about requirements and the second as being about interpretation rules is our reading, not something the standard spells out.

What if a drawing covers more than one stage of manufacture?

It says each indication has to state which stage it refers to, unless it is the final stage.

Why does the phrase unless otherwise specified matter?

A note in the terms clause says defaults are usually recognizable by that introductory wording. It is the marker that a value has been set by a standard and can be changed by the parties.

What are the thirteen fundamental principles?

The contents page names them: invocation, GPS standard hierarchy, definitive drawing, feature, independency, decimal, default, reference condition, rigid workpiece, duality, functional control, general specification and responsibility. The preview gives text for only the first three, so this page does not say what the other ten require.

Is this standard current?

This page does not say. We did not check, because the catalogue page refuses requests from this site.

Is this tolerancing guidance?

No. No advice on tolerancing, drawing indication or measurement is given here.

References

• ISO 8015:2011, Geometrical product specifications, fundamentals, concepts, principles and rules, second edition 1 June 2011, prepared by ISO/TC 213. Free preview of nine pages, covering the front matter, clauses 1 to 4 and subclauses 5.1 to 5.3

This is a digital document rather than a scan, so text extraction is reliable on it; every quotation here was still checked word by word against the file. The free preview stops part way through clause 5, so of the thirteen fundamental principles the contents page names, only three have any text available to us. The other ten appear on this page as names and nothing else, and we do not guess at their content from their titles. We did not read the clauses on indication rules, the annex, the bibliography, or any of the standards this one references, so none of their contents are described. We did not check whether the document is current or superseded, because the catalogue page refuses requests from this site. This is the first time this site has used this standard, and the first time it has used anything from that technical committee. The following are our own readings and arithmetic, not statements in the standard: that twenty six years separate the two editions; and that the sentence about unenforceable unwritten requirements and the sentence about the standard applying unreferenced are about requirements and interpretation respectively. The footnote marking two normative references as still to be published is the third instance of that pattern this site has recorded, which is flagged here as a repetition rather than offered as new, with the earlier page linked. Earlier pages that turned on the phrase unless otherwise specified are linked rather than restated. This article did not come from a forum question; it comes from listing the standards this site had already cited and picking the one with almost no coverage. No advice on tolerancing, drawing indication or measurement is given, and no brand is named. The browser tooling this site normally uses remains unavailable.