ENGINEERING TOLERANCES - INTRODUCTION

Audio: Engineering tolerances – introduction

For young mechanical design engineers, engineering tolerances are probably the most challenging topic regarding drawing creation, but they are definitely worth learning. In my previous article, Engineering drawing – dimensioning, I wrote about the main elements of dimensioning. Furthermore, I was also writing about defining the object’s shape and form using different dimensioning methods.

But in order to fully understand the dimensioning concept, we must learn about the engineering tolerances and how to correctly specify them on the engineering drawing.

In this article, we will build upon dimensioning methods and go one step further; we will dive into the engineering tolerances. First, we will learn the basics and then move to more complex topics.

What are tolerances?

The main intention of the engineering drawings is to communicate design intent in language that is clear and understandable for manufacturing personnel. In other words, the whole design process, from idea, concept sketches, CAD modeling and drawing creation is done with the intent to manufacture physical component. But the problem is that we cannot manufacture anything to its exact dimensions as defined on the drawing. It is just impossible to do it. But in order to make sure that despite these inconsistencies in the manufacturing process, we still can fit multiply different components in functional assembly, we have to allow dimensional variation. The permissible variation of a dimension is called tolerance.

The tolerancing can be divided into two groups:

Plus/Minus tolerancing.
Geometrical and positional tolerancing.

In this text, we will look into the basics of Plus/Minus tolerancing.

Reasons for size variations

Let us now investigate one simple example. Imagine a wood plank 1,5m long that we would like to cut with the saw on 1.2m. We take a measuring tape and mark where we want to cut. After we are finished with the cutting, we measure the plank, and the plank is slightly shorter than 1.2m. We are scratching our heads, completely confused. Why was this plank shorter when we measured it?

Let us analyze what could have gone wrong:

It could be that the surface that we used is not entirely straight. When we measure from the bottom side of the surface, we get a shorter plank than when we would measure it from the top side of the surface.
When we were marking where we wanted to measure, we measured it directly on 1.2m or slightly left or the right from the 1.2m mark on the measuring tape. Was our pen sharp or blunt? Did we draw the cutting line straight (parallel to the cutting path)?
Did we cut with the saw directly following the line, or did we move the saw blade slightly to the side to compensate for the saw thickness?

As you can see from the example above, for a simple plank cutting process, many variables could influence the final size of our plank. Now try to correlate this example with a much more complex manufacturing process than wood plank cutting with a hand saw.

Some of the reasons for the size variations:

Geometric accuracy of machine tools,
Deflection and vibration of a workpiece,
Tool deflection and tool wear,
Error due to clamping,
Alignment errors,
Structural deformation etc.

As we can see, there are many reasons why manufactured component dimensions could defer from the defined dimensions on the drawing.

Basic terminology

The plus/minus tolerancing relate to the linear and angular dimensions. Like everything is defined on the engineering drawings by standard, the basic terminology for the tolerances is defined by standard ISO 286-1:2010.

The tolerances are defined after the dimensional value. They can be defined with the signs “+,” “0”, “-“and “±.” The dimension of a feature on the component lies between the two limit deviations.

Let us now take a look into the graphical representation:

Nominal size – N – dimension of the feature defined on the drawing. This dimension represents the dimension of the feature with the “ideal” form (marking with the letter N is no longer in use, but for easier understanding, we will use it).
Actual size – is determined by measuring the defined dimension after it has been manufactured.
Upper limit of size (ULS) – is the larger permitted size of the two limit sizes.
Lower limit of size (LLS) – is the smaller permitted size of the two limit sizes.
Upper limit deviation – ES (for holes), es (for shafts) – is defined as upper limit minus the nominal size.
Lower limit deviation – EI (for holes, ei (for shafts) – is defined as lower limit minus the nominal size.
Fundamental tolerance (IT) – the difference between the upper limit deviation and lower limit deviation. Tolerance is an absolute value.

In the picture below, we can see an example of the toleranced part. We defined the nominal value as 75 mm with a tolerance range ± 10 mm. This means that our final product can be manufactured in the range between 65 and 85 mm.

Entry of the tolerances on the drawing

When defining the tolerances on the engineering drawing, we can enter them in a few different ways:

Limit dimension

We can specify tolerances on the drawing in the way that instead of the nominal size and the limit deviation, we enter the upper limit of size (ULS) and the lower limit of size (LLS).

We could also have the case when one of the limit deviations equals zero. When the upper limit of size is equal to the nominal size, then ULS=N, ES/es = 0.

When the lower limit of size is equal to the nominal size, then LLS=N, EI/ei = 0.

Unequal bilateral tolerancing

The unequal bilateral tolerances specify a nominal value and the limit deviations.

In the case when the ES/es ＞ 0 and EI/ei ＜ 0:

In the case when the ES/es ＞ 0 and EI/ei ＞ 0:

In the case when the ES/es ＜ 0 and EI/ei ＜ 0:

Equal bilateral tolerancing

The equal bilateral tolerances specify a nominal value and the limit deviations when the ES/es = EI/ei.

Unilateral tolerancing

The unilateral tolerances specify a nominal value and the limit deviation in one direction only.

In the case when the ES/es ＞ 0 and EI/ei = 0:

In the case when the ES/es = 0 and EI/ei ＜ 0:

General tolerances

In order to simplify and improve the clarity of the drawing and speed up the drafting process, we are not defining every single tolerance on the drawing. We can refer to the tolerancing standard relevant to the dimensions that do not have limit deviations called out. We are usually referring to these as general tolerances. They can be overridden by defining local ones to the dimension.

One of the standards we can use is ISO 2768-1:1989. ISO 2768-1:1989 defines general linear and angular tolerances for the manufacturing methods involving cutting and forming (for other manufacturing methods, they are defined with different standards).

ISO 2768-1:1989 contains four different classes: fine (f), medium (m), coarse (c), and very coarse (v). We can write general notes on the drawing or refer to the standard in the title block.

For form and position, general tolerances were defined by ISO 2768-2:1989. This standard was withdrawn, but most engineers are still using this standard. This standard defines three classes: H, K, and L.

In the drawing title block, for that reason, you can often see remarks: ISO 2768 – mK (ISO 2768 – class medium, and ISO 2768-2 class K)

As stated above, ISO 2768-2:1989 was withdrawn and it was replaced with ISO 22081:2021.

Let us look now at the drawing example below. We have a small plate with two holes and a chamfer on every edge. As you can see, we did not define any tolerances on dimensions. But that does not mean that we don’t have them defined on the drawing. We can see in our title block that we are referring to the ISO 2768 – mK standard. That means that all of these dimension tolerances are defined by ISO 2768 medium-class. If we would look at our table defined by ISO 2768-1:1989 we can see that the length of the plate has a tolerance of ±0,3 mm and that our chamfers have ± 0,5 mm. That way, we can check every dimension on the drawing in which range is defined.

The reason why we need to define general tolerancing standard on the drawings is because of fact that the drawing is a legally binding document (check Introduction to the Engineering drawings). Imagine that you do not define the general tolerancing standard on the drawing, and your supplier is working with his standard of tolerances. His standard is approximately around ±35 mm. And then, you receive a component not with a length of 100 mm but with 135 mm. And you cannot get a refund or rework for the component because your supplier has a legal document to prove that you did not specify the tolerance range. If you have the general tolerancing standard statement on the drawing, the story is drastically changing in your favor.

Functional and non-functional dimensions

In Engineering drawing – dimensioning, I was writing about the Functional and non-functional dimensions. This concept is so crucial for tolerances that I will repeat it here.

When we are dimensioning an object, naturally, some dimensions will be more important than others. Some dimensions will be critical to the correct functioning of the component, and these are called functional dimensions. Other dimensions will not be critical to correct functioning, and these are called non-functional dimensions. Functional dimensions are obviously the more important of the two and therefore will be more important when making decisions about the dimension value.

The non-functional dimensions are primarily defined as general tolerances. The functional dimensions require more attention. We must make sure that we define proper deviation limits to ensure the proper function of the part, and for that, we must define proper tolerances.

Scope of the applying tolerances

In this text, we discussed the plus/minus tolerancing. But in order to get a complete understanding of the tolerancing and to properly define the feature’s form, size, orientation, and location on an object, it is necessary to acquire knowledge of:

One of the most important considerations when applying tolerances is fit. The fit refers to the mating of two mechanical components. In other words, two different components are in some kind of relationship with each other, and using fits, we are making sure that this relationship is defined correctly. Fits are a big topic in mechanical design engineering; you can find it here Engineering tolerances – Fits.

In order to define the feature’s form, orientation, and location on an object, you need to familiarize yourself with the Geometrical Product Specification (ASME standard name: Geometric dimensioning and tolerancing – GD&T). Using geometrical and positional tolerances, we can define the relationships between different features on an object, and we can establish what features are used to specify the origin of measurement.

After we created a drawing and defined all the dimensions and tolerances, somehow, we had to check if we specified them correctly. We need a numerical answer to the questions: Will the shaft fit into the bearing? Can we fit two plates in the groove? What is the worst-case biggest length of the plate etc.? We can answer these questions by performing a tolerance analysis and tolerance stackup ( part 1 and part 2).

All the topics listed above are done before the component is manufactured. In order to round up all this knowledge, the last puzzle is to understand what happens with the component after it is manufactured. Different component inspection methods are used to determine if the component was correctly manufactured.

Factors to take into account when choosing tolerances

When defining tolerances for a part, mechanical designers take into account a variety of factors, including:

Functionality: The part must be able to perform its intended function within the specified tolerances. For example, if the part is a bearing, it must be able to rotate freely within the specified range.
Manufacturing Process: The tolerances must be achievable with the manufacturing process being used. For example, a tolerance of 0.025 mm may be achievable with CNC machining but not with casting.
Cost: Tighter tolerances may require more precise and expensive manufacturing processes, which can increase the cost of the part. The designer must balance the need for precision with the cost of the part.
Assembly: The part must be able to fit and function properly with other parts in the assembly. The designer must consider the assembly tolerances and ensure that all parts are compatible.
Material: The material properties also play a role in determining tolerances. For example, a part made of a brittle material may require tighter tolerances than one that is more ductile.
Industry Standards: The designer must consider industry standards and regulations that may apply to the part, such as tolerance standards set by the American Society of Mechanical Engineers (ASME) or the International Organization for Standardization (ISO).
Experience: The experience and expertise of the designer also play a role in determining tolerances. An experienced designer will better understand what tolerances are achievable and what is necessary for the part to function properly.

In general, the designer will consider all of these factors and make a trade-off between the cost, functionality, and manufacturability when choosing the tolerances for a part. The designer will also consult with the manufacturing team to ensure that the tolerances are achievable and that the part can be manufactured efficiently.

How to define tolerances on the part?

When defining engineering tolerances, no hard rule is written on how to do it. Everybody has a different approach, different skills, and experience levels. Here is the approach that I take:

Define functional and non-functional dimensions. Define the areas of interest for the part in a question and “connect” those areas with the other parts influencing this area. Usually, I would print the drawings and do the tolerancing for interconnected parts simultaneously.
Analyze the materials of different interconnected parts. This also includes analyzing the environmental requirements for the parts (for example, environment temperature).
Analyze the manufacturing process of the part.
Define the correct standard for general tolerances based on the chosen manufacturing process (for example, ISO 2768-1:1989). Usually, non-functional dimensions would be defined with general tolerances.
Analyze general tolerances and loosen up ones that can be loosened up.
Analyze the functional dimensions. Define the tolerances and define the fits where they are required.
If necessary, define geometrical and positional tolerances (Geometrical Product Specification or Geometric dimensioning and tolerancing – GD&T)
Perform the tolerance analysis and stackup.

Closing words

It is impossible to create a component with the exact dimensions as defined on the engineering drawing. There are many reasons for that, and to ensure that we can fit multiple components together despite the manufacturing error, deviations are allowed. These deviations we call tolerances.

Specifying correct tolerances to ensure that components fit together can be complicated. In order to properly define a feature’s form, size, orientation, and location on an object, it is necessary to acquire extensive knowledge about the tolerances. These could be discouraging for young mechanical design engineers, but these concepts can be mastered with the proper guidance, persistence, and lots of learning and practice. I would suggest you study tolerances vigorously; a proper understanding of them will make your professional life much easier.

Now you have a great overview of the basics of engineering tolerancing. However, I suggest you go through the text once more and identify areas you think need more understanding and clarity. Then, once you have identified those areas, start building up your knowledge in those areas.

To make it easier for you to find related posts, check the “Further reading” chapter below. Do you have any questions or need something to be clarified better? Leave a comment below, and I will give my best to adjust the post accordingly.

Do you like what you read? Then, share it with your friends, colleagues, and on your social media channels. And if you want to become a part of the Newtonians, make sure to subscribe to our newsletter!