Method for determining tolerance intervals for dimensioning a part

1. A method for determining tolerance intervals for a set of dimensions of a part, said tolerance intervals being defined so that one part for which the dimensions are included in said tolerance intervals observes a set of constraints with a determined failure probability p, the method comprising:

generating a separation function for each constraint to be observed, each function being able to indicate whether a given dimensioning of a part meets or not the corresponding constraint, and

iteratively building a set of tolerance intervals increasingly large by testing, at each iteration, by the separation functions, whether the dimensionings included in said tolerance intervals observe the constraints with a failure probability below the probability p,

wherein the step for building the set of tolerance intervals comprises an iterative application of the following steps:

a) globally enlarging current tolerance intervals relatively to preceding tolerance intervals so that there exist more dimensionings included in the current tolerance intervals than in the preceding tolerance intervals,

b) randomly sampling according to a multivariate Gaussian law centered on rated values of dimensions of the part, a plurality of dimensionings included in the set of current tolerance intervals,

c) testing, by the separation functions, whether each dimensioning is acceptable relatively to the constraints,

d) testing whether a percentage of unacceptable dimensionings relatively to the constraints is less than the failure probability p, and

e) in a case of a positive test in step d), repeat steps a) to d).

2. The method according to claim 1 , wherein during the first iteration of the step for building the tolerance intervals, each tolerance interval is reduced to a determined rated value of the corresponding dimension of the part.

3. The method according to claim 1 , further comprising, in a case of a negative test in step d), a step e′) for reiterating steps a) to d) by taking up as preceding tolerance intervals last tolerance intervals for which the percentage of unacceptable dimensionings relatively to the constraints is less than the failure probability p.

4. The method according to claim 1 , wherein the step a) is applied by perturbing ends of the preceding tolerance intervals symmetrically according to a Gaussian law.

5. The method according to claim 4 , wherein the step a) further comprises at an end of the perturbing, computing a total volume of a hypercube, sides of the hypercube are formed by sizes of the tolerance intervals, and comparing the volume of the hypercube relatively to the hypercube formed by last tolerance intervals for which the percentage of unacceptable dimensionings relatively to the constraints is less than the failure probability p.

6. The method according to claim 5 , wherein if the obtained volume of the hypercube is less than the one formed by the last intervals for which the percentage of unacceptable dimensionings relatively to the constraints is less than the failure probability p, step a) is reiterated by taking up again as preceding tolerance intervals the last tolerance intervals for which the percentage of unacceptable dimensionings relatively to the constraints is less than the failure probability p.

7. The method according to claim 6 , wherein the step for building the tolerance intervals is stopped after a determined number of unsuccessful iterations defined by:

the volume of the current hypercube is less than the volume of the last hypercube corresponding to the tolerance intervals for which the number of unacceptable dimensionings relatively to the constraints is less than the failure probability p, or

the result of step d) is negative.

8. The method according to claim 1 , wherein each constraint is modeled by a function, and the step for generating separation functions comprises the application of the following steps:

forming a Monte-Carlo population of dimensionings of the part,

initializing, for each constraint, a corresponding Kriging separator, and applying, iteratively, the following steps:

determining, from among the dimensionings of the Monte-Carlo population, the dimensioning for which the estimation accuracy of a constraint by the corresponding Kriging separator is minimum, and

computing, for the determined dimensioning, a value of a whole of the constraint functions, and updating the Kriging separators accordingly.

9. The method according to claim 1 , wherein the part is a turbine engine blade.

10. A non-transitory computer readable medium storing a program comprising code instructions that, when executed by a processor, cause the processor to execute the method according to claim 1 .

11. A method for validating a manufactured turbine engine part, comprising, for each manufactured part:

measuring the whole of the dimensions of the part,

verifying that each dimension is included in a corresponding tolerance interval, wherein the tolerance intervals are determined by applying the method according to claim 1 , and

if each dimension of the part is included in the corresponding tolerance interval, validating the part, and

if at least one dimension is not included in the corresponding tolerance interval, applying an additional processing operation on the part in order to determine whether the part is valid or not.