The error in stamped parts produced from the time a new die is put into production until its first sharpening is called the initial error; the error in stamped parts produced after the die has been sharpened approximately 20 times until it fails and is scrapped is called the conventional error; and the maximum allowable error of the last batch of qualified stamped parts produced before the die fails and is scrapped is called the limit error.
On-site, the basis for determining the sharpening life of a die is the burr height of the punched and blanked parts. Since any formed part involves blanking operations (blank blanking or punching), this is especially true for compound dies. Therefore, it is crucial to check and measure the burr height of the stamped parts against the die and compare it with the enterprise standard or JB4129-85 "Burr Height of Stamped Parts".
The initial error of a stamping die is usually the smallest error in the stamped part throughout the die's entire lifespan. Its magnitude depends primarily on the manufacturing precision and quality of the die, as well as the dimensions of the stamped part, the material thickness, and the size and uniformity of the clearance.
The manufacturing precision and quality of the die, in turn, depend on the die-making process. For medium-carbon steel composite stamping dies with a material thickness t ≤ 1 mm, both experimental results and production practice have shown that the burr height of dies manufactured by wire electrical discharge machining (EDM) is 25%–30% higher than that of dies manufactured using forming grinding or NC/CNC continuous trajectory coordinate grinding (i.e., precision grinding). This is because the latter not only has higher machining precision but also a surface roughness Ra value that is an order of magnitude smaller than the former, reaching 0.025 μm. Therefore, factors such as the manufacturing precision and quality of the die determine the initial stamping precision of the die and also contribute to the initial error of the stamped part.
The standard error of a stamped part is the actual error it possesses from the first sharpening of the die to the final sharpening and the production of the last acceptable stamped part. As the number of sharpening cycles increases, the dimensional increment caused by natural wear of the cutting edge gradually increases, and the error of the stamped part also increases. When the error exceeds the limit deviation, the stamped part is unacceptable, and the die fails and is scrapped. The dimensions of the upper hole and inner shape of the stamped part gradually decrease due to die wear; the blanking dimensions of the outer shape gradually increase due to die wear. Therefore, the upper hole and inner shape of the stamped part are marked with a unidirectional positive deviation tolerance and the die is made according to a size close to or almost equal to the maximum limit dimension. Similarly, the blanking dimensions of the stamped part are marked with a unidirectional negative deviation tolerance and the die is made according to a size close to or almost equal to the minimum limit dimension. This expands the standard error range of the stamped part, increases the number of sharpening cycles the die can undergo, and improves the die life.
The limit error of a stamped part is the actual maximum permissible dimensional error of the stamped part with limit deviations. These stamped parts are usually the last batch of acceptable stamped parts produced before the die fails and is scrapped.
A comprehensive analysis of the fluctuations, trends, and patterns of errors in various types of stamping dies throughout the die's lifespan reveals that: the dominant component of stamping part errors remains constant; the error increment due to natural wear of the cutting edge or cavity gradually increases with the number of strikes made after die grinding; and some error increments are irregular and unpredictable. Therefore, the errors in various types of stamping dies are composed of several parts, including constant errors, gradually increasing errors, systematic errors, and random errors.
