It allows a statement to be made about the proportion of rejects and rework that can be expected when operating this machine.
Machine capability study
In order to determine the machine capability, a random sample of workpieces (at least 50 pieces) determined beforehand according to statistical criteria is produced under specified conditions and then the critical values for the later function of the part are measured. The 5-M influences of man, material, measurement method, the environment and machine must not change or only change slightly.
- Human: The same person must operate the machine during the examination.
- Material: The same material must be used (e.g. aluminum pipes from one manufacturer from one batch).
- Measurement method: measurements are taken with the same measuring device over the entire duration of the examination.
- Machine temperature: The temperature of the machine should not fluctuate, i. H. the machine's systems are at operating temperature.
- Manufacturing method: The same manufacturing method (process) is used, e.g. B. turning or grinding.
Then the statistical distribution applicable to the sample is determined . For teaching purposes, a normal distribution is often assumed, but this is seldom found in production. Now you determine the position and spread of the measured quantity. From this one can determine the machine capability as a numerical value. In modern companies, a machine capability of 1.33 (corresponds to 4 σ standard deviation with normal distribution) or 1.67 (corresponds to 5 σ) is often assumed. The smaller the value, the worse the machine capability. Stricter standards are set for machine capability than for process capability, since there are other influencing variables (e.g. deviations in the material between different batches).
Two values are normally determined: C mK gives a statement about how the mean value of the sample lies within the given tolerance, C m shows how great the variance of the size is. In the ideal case, C m = C mK , but since there is usually always a discrepancy between the mean value and the nominal value, in reality C mK <C m .
Problems and application
The term machine capability has led to problems in practice, because too many influencing variables (measuring equipment, environment, employees, ...) have an effect on machine capability, so that only the process capability with the key figures C p and C pK and the stability (control) of the Process should be used. Since the machine capability is much easier and faster to determine compared to the process capability, it is often used for the initial assessment of a system / machine (e.g. for a machine pre-acceptance).
- Calculation of machine capability with qs-STAT: for machine acceptance and process qualification
- Free Excel template for calculating machine capability