METHOD OF MANUFACTURING PARTS HAVING A DIMENSIONAL REQUIREMENT AKIN TO A TOLERANCE STACK-UP

Abstract

The invention pertains to a method of manufacturing parts, where each part must satisfy a global dimensional requirement on a specific dimension of said part akin to a tolerance stack up comprising at least two tolerances, each tolerance corresponding to a characteristic dimension of the part, where: the global dimensional requirement is fixed so that the specific dimension of the parts has a mean 0 and a standard deviation oG lying in a global acceptance range constructed on the basis of one or more statistical criteria; a reference dimension is chosen from among the characteristic dimensions, and a mean and a standard deviation oref of said reference dimension are calculated on the basis of a sample of collected parts; a range of acceptance for the other characteristic dimensions of the tolerance stack up constructed with the same statistical criterion or criteria as the global acceptance range is calculated on the basis of the global acceptance range and on the basis of the mean and the standard deviation oref calculated for the reference dimension; the manufacture of parts is steered without verification of the reference dimension by using the acceptance range calculated for the other characteristic dimensions of the tolerance stack up by fitting the adjustment parameters of a manufacturing device so that each of the other characteristic dimensions of the tolerance stack up lie in the corresponding calculated acceptance range.

Claims

1. Method of manufacturing parts produced with a manufacturing device, where each part must satisfy an overall dimensional requirement on a specific dimension of said part, said overall dimensional requirement being akin to a tolerance stack-up comprising at least two sides, each side corresponding to a characteristic dimension of the part, in which: the overall dimensional requirement is fixed so that the specific dimension of the parts has an average .sub.G and a standard deviation .sub.G in an overall area of acceptance constructed from one or more statistical criteria; a reference dimension is selected from among the characteristic dimensions comprising the tolerance stack-up, and an average .sub.ref and a standard deviation .sub.ref of said reference dimension are calculated from a sample of parts taken from among the parts produced with the manufacturing device; an area of acceptance for the other characteristic dimensions of the tolerance stack-up is calculated from the overall area of acceptance and from the average .sub.ref and the standard deviation .sub.ref calculated for the reference dimension, the area of acceptance being constructed with the statistical criteria or the same statistical criteria as the overall area of acceptance; the manufacturing of parts without verification of the reference dimension is controlled by using the area of acceptance calculated for the other characteristic dimensions of the tolerance stack-up by adjusting the regulating parameters of the manufacturing device so that each of the other characteristic dimensions of the tolerance stack-up is included in the corresponding calculated area of acceptance.

2. The method as claimed in claim 1, in which the manufacturing of parts is controlled by conducting the following successive steps for each sample of parts taken: measurement for each part of the sample of the other characteristic dimensions with the exception of the reference dimension; calculation for each of the other measured characteristic dimensions of an average .sub.i and a standard deviation .sub.i; adjustment of the regulating parameters of the manufacturing device so that each of the other characteristic dimensions of the tolerance stack-up has an average .sub.i and a standard deviation .sub.i in the corresponding calculated area of acceptance.

3. The method as claimed in claim 1, in which the area of acceptance of the other characteristic dimensions of the tolerance stack-up is calculated from the limit values of average .sub.GL and standard deviation .sub.GL forming the overall area of acceptance for the specific dimension and from the average .sub.ref and of the standard deviation .sub.ref calculated for the reference dimension.

4. The method as claimed in claim 3, in which the area of acceptance of the other characteristic dimensions of the tolerance stack-up is calculated by taking .sub.i=.sub.GL.sub.ref and .sub.i={square root over (.sup.2.sub.GL.sup.2.sub.ref)}, where .sub.i and .sub.i are respectively the average and the standard deviation of the corresponding characteristic dimension.

5. The method as claimed in claim 4, in which the limits of the area of acceptance of the other characteristic dimensions of the tolerance stack-up are calculated by taking the strictest values limits for the statistical criterion or the statistical criteria used.

6. The method as claimed in claim 1, in which the different fields of acceptance are illustrated on a control graphic (;) having as abscissa the average and as ordinate the standard deviation of the characteristic dimension and are constructed according to the limits of the statistical criterion or criteria used for the overall dimensional requirement.

7. The method as claimed in claim 6, in which the area of acceptance of the other characteristic dimensions of the tolerance stack-up is constructed graphically according to the following successive steps: applying to the overall area of acceptance in the control graphic (;) a transform according to the average .sub.ref and the standard deviation .sub.ref calculated for the reference dimension; then extrapolating the overall field of application transformed for constructing the field of application of the other characteristic dimensions of the tolerance stack-up by taking the strictest values limits for the statistical criterion or the statistical criteria used.

8. The method as claimed in claim 6, in which the control graphic (;) is displayed on a control monitor.

9. The method as claimed in claim 1, in which the overall area of acceptance is constructed from one or more statistical indicators selected from among: a first capability index Cpk defined by the formula: $\mathrm{Cpk}=\frac{\mathrm{Min}\left(\mathrm{TS}-;-\mathrm{TI}\right)}{3.Math.}$ a second capability index Cp defined by the formula: $\mathrm{Cp}=\frac{\left(\mathrm{TS}-\mathrm{TI}\right)/2}{3.Math.}$ a centring coefficient Cc defined by the formula: $\mathrm{Cc}=\frac{}{\left(\mathrm{TS}-\mathrm{TI}\right)/2}$ where: is an average and a standard deviation of a measured characteristic dimension on several parts; TS is an upper tolerance of the measured characteristic dimension; TI is a lower tolerance of the measured characteristic dimension.

10. The method as claimed in claim 1, in which the selected reference dimension corresponds to the characteristic dimension having the least controllable manufacturing parameters from among the different characteristic dimensions comprising the tolerance stack-up.

11. The method as claimed in claim 1, in which the tolerance stack-up comprising two sides corresponding respectively to a first characteristic dimension and a second characteristic dimension of the part, in which: the reference dimension corresponds to the first characteristic dimension of the tolerance stack-up; the area of acceptance of the second characteristic dimension of the tolerance stack-up is calculated from the overall area of acceptance and from the average .sub.ref and the standard deviation .sub.ref calculated for the first characteristic dimension, the area of acceptance being constructed with the statistical criteria or the same statistical criteria as the overall area of acceptance; the manufacturing of parts is controlled without verification of the first characteristic dimension by using the area of acceptance calculated for the second characteristic dimension of the tolerance stack-up by adjusting the regulating parameters of the manufacturing device so that the second characteristic dimension of the tolerance stack-up is included in the corresponding calculated area of acceptance.

Description

DESCRIPTION OF FIGURES

[0057] Other characteristics and advantages of the invention will emerge from the following description, which is purely illustrative and non-limiting and must be viewed in conjunction with the attached diagrams, in which:

[0058] FIG. 1 diagrammatically illustrates the restrictions due to a dimensional requirement depending on several particular characteristic dimensions, in the particular example of the deposit of a thermal barrier layer in a passage section of a mechanical part;

[0059] FIG. 2 is a graphic of density of probability illustrating the variability of the thermal barrier layer;

[0060] FIG. 3 is a graphic in (;.sup.2) illustrating the evolution of the passage section prior to and after the deposit of the barrier layer;

[0061] FIG. 4 is a graphic in (;.sup.2) illustrating determination of the field of acceptability of the passage section prior to deposit of the barrier layer as a function of the statistical values of the barrier layer and of the field of acceptability of the passage section prior to deposit of the barrier layer;

[0062] FIG. 5 is a graphic in (;) illustrating determination of the field of acceptability of the passage section prior to deposit of the barrier layer as a function of the statistical values of the barrier layer and of the field of acceptability of the passage section prior to deposit of the barrier layer;

[0063] FIG. 6 is a graphic in (;) illustrating the final extrapolation step for determining the field of acceptability of the passage section prior to deposit of the barrier layer.

[0064] FIG. 7 is a diagram illustrating a production chain integrating control and regulating of production with sampling of parts.

DETAILED DESCRIPTION OF THE INVENTION

[0065] The general principle of the invention comprises assimilating a dimensional requirement relating to a dimensional characteristic specific to a tolerance stack-up where each side corresponds to a characteristic dimension of the part.

[0066] From among the different characteristic dimensions forming the tolerance stacku, one is identified whereof machining is the least manageable and/or whereof the rectification is the most difficult. This characteristic dimension is selected and serves as reference for monitoring production of parts, and is called reference dimension.

[0067] Machining that is barely or not manageable is machining having a very random result despite identical input parameters and identical manufacturing conditions.

[0068] The other characteristic dimensions of the tolerance stack-up serve as adjustment variables for satisfying the dimensional requirement required by taking into account the values of the reference dimension.

[0069] The final assembly sees its conformity sanctioned by statistical tolerances, comprising, for example, a centring coefficient Cc and a capability index Cpk. In this way, it is advantageous to take into account the impact of the non-controlled side, called reference dimension, on the final assembly not only in terms of decentring of the average but also in terms of dispersion engendered on the registered values.

[0070] In this way, according to the proposed monitoring method an average .sub.ref and a standard deviation .sub.ref of the selected reference dimension are calculated from a sample of parts taken in the manufacturing flow, where .sub.ref is the average of the decentring measured on the sample concerning the reference dimension, and .sub.ref is the standard deviation measured on the sample.

[0071] The overall dimensional requirement is fixed so that the specific dimension of the parts has an average .sub.G and a standard deviation .sub.G contained in an overall area of acceptance constructed from one or more statistical criteria. The overall area of acceptance can, for example, be constructed from a centring coefficient Cc and a capability index Cpk where the aim is for the values to be contained within certain limits.

[0072] Given the overall area of acceptance of the specific dimension for a part to be acceptable in terms of production, and having calculated the average .sub.ref and the standard deviation .sub.ref for the reference dimension, an area of acceptance can be calculated for the other characteristic dimensions of the tolerance stack-up, this area of acceptance preferably being constructed with the statistical criteria or the same statistical criteria as the overall area of acceptance.

[0073] In this way, controlling the characteristic dimension or the other characteristic dimensions to be inside the area of acceptance now calculated ensures that the specific dimension also stays in its overall area of acceptance, and therefore the dimensional requirement for the part is respected, therefore validating the analysed part.

[0074] Consequently, the manufacturing flow of parts is controlled without need to verify the reference dimension during monitoring, by using only the area of acceptance calculated for the other characteristic dimensions of the tolerance stack-up.

[0075] Such a method is particularly advantageous since this ensures respect of the overall dimensional requirement, which can be complex, without the need to be too strict for what relates to the characteristic dimensions since the method takes into account not only the average .sub.ref of the reference dimension but also its standard deviation .sub.ref.

[0076] As is preferred, the area of acceptance of the other characteristic dimensions of the tolerance stack-up is calculated from the limit values of average .sub.GL and standard deviation .sub.GL forming the overall area of acceptance for the specific dimension and from the average .sub.ref and of the standard deviation .sub.ref calculated for the reference dimension.

[0077] According to a particular embodiment, the area of acceptance of the other characteristic dimensions of the tolerance stack-up is calculated by taking .sub.i=.sub.GL.sub.ref and .sub.i={square root over (.sup.2.sub.GL.sup.2.sub.ref)}, where .sub.i and .sub.i are respectively the average and the standard deviation of the corresponding characteristic dimension.

[0078] According to this calculation, the limits of the area of acceptance of the characteristic dimensions of the tolerance stack-up can be defined by taking the strictest values limits for the statistical criterion or the statistical criteria used.

[0079] The calculated field of acceptance is preferably valid for all the other characteristic dimensions of the tolerance stack-up.

[0080] In the embodiment where the tolerance stack-up comprises only two sides, one of the two characteristic dimensions of the tolerance stack-up is taken as reference dimension, then the area of acceptance of the other characteristic dimension of the tolerance stack-up is calculated from the overall area of acceptance and from the average .sub.ref and the standard deviation .sub.ref calculated for the reference dimension, this area of acceptance preferably being constructed with the statistical criteria or the same statistical criteria as the overall area of acceptance. The manufacturing flow of parts is then controlled without verification of the reference dimension by using the area of acceptance calculated for the second characteristic dimension of the tolerance stack-up.

[0081] The present invention is described hereinbelow in reference to a particular but non-limiting example of the production of mobile blades for engines in the field of aeronautics, or any similar part, whereof a passage section must be covered by a thermal barrier layer.

[0082] Such a thermal barrier layer considerably influences the passage section of these bladings. The sections of passage of many bladings form the object of a statistical definition, imposed by the final use of the produced, corresponding to the overall dimensional requirement.

[0083] The method of deposit of thermal barrier, although very repeatable in general, is difficult to control, and even more to rectify. For this reason, when non-conformities relative to statistical criteria imposed on the passage section are observed, acting on the deposit of thermal barrier is generally not possible. The sole possibility is to act on the value of the passage section of the part after machining and prior to deposit of thermal barrier, which is possible by machining the foot of the part differently.

[0084] For the passage section prior to deposit of thermal barrier there is a value recorded in the plan, with a nominal and tolerances, but this nominal and these tolerances assume standard deposit of thermal barrier, which should finally result in a compliant passage section, that is, being in the overall field of acceptability of the specific dimension.

[0085] When deviations to the specification are observed during deposit of the thermal barrier (or in terms of deposit of thermal barrier the specification has changed without the final need having been modified), respect of this nominal and these tolerances prior to deposit of thermal barrier engenders nonconformities on the passage section after thermal barrier. Therefore which new nominal and which new tolerances must be applied to the passage section of the part prior to deposit of thermal barrier must be known so that the individual and statistical criteria relating to the conformity of production are respected.

[0086] This problem is illustrated in FIG. 1 illustrating the variability 1 of the passage section prior to thermal barrier 10, the variability 2 of the thermal barrier 20, and the variability 3 of the passage section after thermal barrier 30. It is evident in this FIG. 1 that the passage section is smaller due to deposit of the thermal barrier, which is logical, but above all it is more dispersed due to the variability of the thermal barrier.

[0087] This particular example of deposit of a barrier layer in a passage section can resemble a tolerance stack-up having two links, one of the links (the thickness of thermal barrier deposited) being non-controllable, and the other of the links (the passage section prior to deposit of thermal barrier) contributes to serving as adjustment variable to finally ensure conformity of the assembly.

[0088] The final assembly will see its conformity sanctioned by statistical tolerances, consisting classically of a Cpk and a Cc. This is the reason for which it is important to take into account the impact of the side not controlled on the final assembly not only in terms of decentring of the average but also dispersion produced on the recorded values.

[0089] The assembly of a value of passage section prior to thermal barrier having as its characteristics an average .sub.SdP prior to BT and a standard deviation .sub.SdP prior to BT, with deposit of thermal barrier causing obstruction to the passage section having as its characteristics an average .sub.BT<0 (since deposit of a barrier layer reduces the passage section) and a standard deviation to form a final passage section having as its characteristics an average .sub.SdP final and a standard deviation .sub.SdP final having to respond to some criteria translates into the following relationships:

.sub.SdP final=.sub.SdP prior to BT+.sub.BT

.sup.2.sub.SdP final=.sup.2.sub.SdP prior to BT+.sup.2.sub.BT

[0090] A graphic tool called Plan Variance-decentring can be used, in which the average of the decentring is carried as abscissa and as ordinates the variance (that is, the square of the standard deviation ). In this graphic, each population is consequently illustrated by a point.

[0091] The above relationships translate into the fact that the effect of deposit of thermal barrier is to translate the population of sections of passage of the part in the variance-decentring plane of a vector {right arrow over (v)} of coordinates (.sub.BT; .sup.2.sub.BT).

[0092] If variability of the thermal barrier 20 such as illustrated in FIG. 2 is considered, characterised by an average .sub.BT and a standard deviation .sub.BT, translation in the variance-decentring plane can be represented as on the graphic of FIG. 3 where point 4 corresponds to the section of the passage prior to thermal barrier, while point 5 corresponds to the section of the passage after thermal barrier.

[0093] The technical solution proposed consists of representing in the variance-decentring diagram the area of acceptance containing all the populations of sections of passage after deposit of thermal barrier admissible in light of statistical criteria listed in the manufacturing plan.

[0094] Typically, in the event where the statistical criteria consist of a Cpk and a Cc less than 1, this area of acceptance D1 resembles a tent hoisted by walls, as is illustrated in FIG. 4. This area of acceptance D1 is then offset by a vector {right arrow over (v)} to produce the area of acceptance D2 in which the population of sections of passage should be located prior to deposit of thermal barrier so as to respond, after deposit of thermal barrier (embodied by a translation of vector {right arrow over (v)} in this graphic) to the passage section after deposit of thermal barrier criteria included in the plan.

[0095] As seen in graphic of FIG. 4, the area of acceptance D2 for the sections of passage prior to deposit of thermal barrier constructed in this way tolerates larger sections of passage (as the effect of the thermal barrier is to obstruct the passage section), but less dispersed (as the dispersion produced by the deposit of thermal barrier is added to that initially present in the population prior to deposit of thermal barrier).

[0096] For easier graphic reading of fields of acceptance, it is preferable to use diagrams of standard deviation-decentring type, also called diagrams (;), in which is carried as abscissa the average of the decentring and as ordinates the standard deviation .

[0097] In such a type of graphic the representation of statistical criteria and their respective limits of acceptance defining the different fields of acceptance is simpler, and therefore far easier to execute.

[0098] For example, in this type of diagram (;), the iso-Cc are illustrated by vertical straight lines and the iso-Cpks by oblique straight lines cutting the axis of abscissa at the level of tolerances.

[0099] In this way, an overall area of acceptance D.sub.G, that is, the area of acceptance of the passage section after deposit of thermal barrier, defined by a centring coefficient Cc and a capability index Cpk, has the appearance of a house as is illustrated in FIG. 5.

[0100] But here, the effect of the deposit of the thermal barrier on a population illustrated by a point in this diagram is not similar to simple translation of the overall area of acceptance.

[0101] Therefore the area of acceptance D.sub.i.sub._.sub.inter the values of the passage section is o f calculated prior to deposit of the thermal barrier by taking:

.sub.SdP prior to BT=.sub.SdP final.sub.BT

.sub.SdP prior to BT={square root over (.sup.2.sub.SdP final.sup.2.sub.BT)}

[0102] The variance-decentring for constructing this field is therefore passed through the diagram, and it is converted in the diagram - to quantify the limits thereof in terms of tolerances, Cc and Cpk, as illustrated in FIG. 5.

[0103] The new area of acceptance D.sub.i.sub._.sub.inter calculated in this way in the plane - no longer has the same form as the initial field since it is not a simple shift of the overall area of acceptance.

[0104] For example, when the area of acceptance after deposit of thermal barrier is constructed by imposition of a Cpk and a Cc, the area of acceptance after deposit of thermal barrier D.sub.G in the plane - has the form of a house, whereas the area of acceptance prior to deposit of thermal barrier D.sub.i.sub._.sub.inter constructed by the method described hereinabove resembles a crossed vault hoisted by two walls, as seen in FIG. 5.

[0105] If the aim is to delimit the area of acceptance prior to deposit of thermal barrier by statistical criteria similar to those employed for the passage section after deposit of thermal barrier, that is, a centring coefficient Cc and a capability index Cpk, it suffices to list a house in the crossed vault and calculate the preferred area of acceptance D.sub.i.sub._.sub.final (see FIG. 6). Determination of this area of acceptance D.sub.i.sub._.sub.final can be done by calculation or by extrapolation graphic.

[0106] The method proposed is particularly advantageous as it takes into account the impact of the defective process, that is, of the dimension whereof the manufacturing is minimally regulated and/or minimally rectifiable, not only in terms of average of the decentring but also in terms of dispersion.

[0107] Initial naive reasoning results in simply degreasing the nominal prior to thermal barrier of the average value of the deviation observed in the deposit of thermal barrier.

[0108] In the example above of the passage section with deposit of a thermal barrier layer, the method proposed takes into account the fact that the variability observed on the deposited thickness of thermal barrier has repercussions on the passage section. This prevents cases where good production relative to statistical requirements in the framework offset by the average deviation observed on the deposit of thermal barrier was sanctioned as poor after deposit of thermal barrier, because its dispersion resulted in its being deemed unacceptable in terms of the statistical requirements after deposit of thermal barrier.

[0109] The method proposed can be executed in a manufacturing chain of parts, which can be fully or partially automated, where checks during production regulate the manufacturing flow, that is, adjust the manufacturing conditions to ensure that the manufactured parts continue to respond to the required criteria of qualities.

[0110] FIG. 7 gives an example of such a manufacturing chain in which a machining device, such as, for example, a 5-axle machine, is used to make parts according to a specific instruction. The specific instruction can, for example, relate to an overall dimensional requirement. In place of the machining device a manufacturing device not limited to the machining of parts could of course be used.

[0111] In this automated production chain parts are taken when leaving the machining device to form a sample and are transmitted to a measurement device, which measures one or more characteristic dimensions of each part of the sample taken. Such a measuring device can, for example, be a three-dimensional measuring device having sensors, which automatically measure the preferred characteristic dimensions of each of the parts.

[0112] The measuring data originating from the measurement device are then transmitted to a calculation device, which processes them to calculate one or more statistical indicators representative of one of the characteristic dimensions of the parts.

[0113] The calculated value of the statistical indicator is then compared to a reference instruction on the characteristic dimension to regulate the manufacturing flow. More precisely, the results of this comparison optionally adjust the input parameters of the machining device.

[0114] If a deviation is noticed, implying an error, for example if the value of the statistical indicator on the characteristic dimension is outside an acceptable range defined by the reference instruction, corrective measurements are determined by a corrector to adjust the input parameters of the machining device. The aim of the modifications of the input parameters of the machining device is to correct the deviation noticed so that the value of the statistical indicator on the characteristic dimension is again in an acceptable range.