Manufacturing costs are influenced by a number of factors: material, finish, whether or not the welds need to be certified to name a few. When we look at drawings to manufacture a part, we often don’t know anything other than what is on the drawing so we cost and manufacture that part based on what the drawings show. But, when we have an opportunity to work with our clients more closely, we can better align engineering with purchasing and manufacturing to reduce production costs.

Impractical tolerances is one factor that can significantly increase manufacturing costs. Some design software defaults to specific tolerances that may not be necessary. These default templates are used without thought to the functional tolerance requirements.

In manufacturing, the larger, more forgiving tolerances which means less scrap, better yeild and less waste. Tight tolerances could require much more expensive production methods, machine tools, inspection devices and a significantly greater amount of total processing time. These costs can increase exponentially when making multiples of the same part.

For example, if a part does not have to connect to another part on one end to function, then manufacturing it to 1/16″ instead of 1/32″ means there’s less time on the machine as well as potentially reduces labour costs.

Obviously, fit is exteremly important. Loose tolerances will result in gaps and can lead to equipment failure. As well, sometimes consistency is needed from part-to-part which will dictate a method of fabrication.

Alternatively, seldom would we make a part with no tolerances; unless it’s say a prototype or will have additional fabrication in the manufacturing process.

How do you find the appropriate balance?

Job shops that do small batch production of OEM parts is a great place to start. They can be agile and respond to needs as you are looking to identify greater manufacturing efficiencies. Before you commit to a tight tolerance consider:

  1. How does the part function in its end use?
  2. What tolerances are important for the environment in which the part will be used?
  3. Does the part interact with any other part or parts? If not, you might not need such tight tolerance. If so, you need to be aware of the potential for tolerance accumulation.
  4. If the same part has different attributes requiring tolerances – such as both a diameter and a radius – which is the more critical dimension?
  5. Is there a very tight tolerance taht may be countered by some interaction within the part and therefore, not be practical?
  6. Are you willing to test a slightly out-of-spec part to see if a looser tolerance will work for your application?

Do I really need ±0.001 vs. ±0.01 (or ±0.1)?

There are times when tight tolerances are absolutely mandatory – when the difference between a product that works and one that fails means dimensional tolerances expressed in decimal places that go much further than you initially expected. In these scenarios, small variations in each part’s dimensions can multiply, especially with complex assembilies, resulting in tolerance buildup and unacceptable variations in the intended design.

The good news is that there are instances where looser tolerance may be perfectly acceptable. By carefully considering the end use, you can distinguish between critical and non-critical tolerances, and minimizing your cost of quality.

The cost of default tolerances

There are standards for different industries in place and handbooks that provide default tolerances. In fact, most drawings will name a default tolerance under the heading “Unless Otherwise Specified”. Accepting default tolerances without understanding whether or not they’r etruly necessary, or simply not noticing them at all, can needlessly drive up costs. Here are some common examples:

  • An extra decimal of tolerance is specified not for a technical reason but because it is “the way it’s always been done”
  • In a project wiht a number of components requiring a tolerance of ±0.001because of the way they must fit together but a separate part automatically defaults but it may not be needed
  • In an effort to keep mechanical drawing labels consistent to the same decimal point, a zero is accidentally misplaced, labelling a tolerance ±0.001 instead of ±0.010
  • Two parts that failed to fit together correctly are re-specified at tighter tolerances – but the original failure was due to the parts being cut at two different loose tolerances.

There is always going to be a gap between design, purchasing and manufacturing but by asking the right questions during the process will help uncover greater efficiencies.

Contact us