Control plan – Design

The design menu makes it possible to create control plans by posting their efficiency curves, as well as the average quality level after testing and the average number of parts tested.

Courbe

Select the controls that you want to post at the bottom of the window.

Designing a standardized control plan

To design a standardized plan all you need to do is choose three parameters:

  • The sampling level (S1 to S4 for special controls, and I, II or III)
  • Acceptable quality level (NQA [AQL]) in the standardized AQL list
  • The type of sampling for revitalization (normal, reinforced, reduced)

All the plans selected will be calculated, the efficiency curves drawn, the P95 (a non-quality that has a 95% chance of being accepted) and the P10 (a non-quality that has a 10% chance of being accepted) will appear

Normalisé

The acceptance probability can be known for a given non-quality rate:

in the example below we input 4% as the default, Ellistat then calculated the acceptance probability as 37.48%.

Normalisé2

The non-quality rate can be known for a given acceptance probability:

in the example below a 50% acceptance probability was input, Ellistat then calculated the assoicated non-quality level as 3.32% of non-compliant parts.

PA

Designing a customized control plan by imputting the sample sizes and the acceptance rates

In addition to standardized plans, Ellistat enables you to design any type of plan.  The white boxes are input cells.  When you input a value into a white box, the control plan is reset to adapt to the input.

In the example above we created a single control plan with a sample size of 40 parts with acceptance of 1 non-compliant part per 40.

We also created a double control plan with two samplings of 25 parts, an acceptance if no non-compliant parts were controlled in the first sampling or if 2 non-compliant parts or less were controlled over a total of 50 parts.

Création du plan

For the measurement control plan we choose the size and a factor k for one of the two methods (Sigma or s) and Ellistat calculates the equivalent control plan using the other method.

Designing a custumized control plan for imput at P95 and P10 rates.

Another way of creating a control plan consists of inputting the two design points of an efficiency curve (P95 and P10). Ellistat then calculates all the plans following this efficiency curve.

In the example below we want to calculate a plan that yields a P95 of 1% and a P10 of 6%.  After having input the two values you click on the “generate” button to obtain all of the corresponding plans.  Notice the almost perfect superposition of the efficiency curves.

P10