GOST 30415-96
GOST 30415−96 Steel. Non-destructive testing of mechanical properties and microstructure of steel products by magnetic method (Change No. 1)
GOST 30415−96
Group B09
INTERSTATE STANDARD
STEEL
Non-destructive testing of mechanical properties and microstructure of steel products by magnetic method
Steel. Nondestructive testing of mechanical properties and microstructure of steel products by magnetic method
ISS 77.040.20
77.080.20
AXTU 0909
Date of introduction 1998−01−01
Preface
1 DEVELOPED by the Interstate technical Committee for standardization MTK 145 «monitoring Methods of steel products"
INTRODUCED by Gosstandart of Russia
2 ADOPTED by the Interstate Council for standardization, Metrology and certification (Protocol No. 10 of October 4, 1996)
The adoption voted:
| The name of the state | The name of the national authority for standardization |
| The Republic Of Azerbaijan | Azgosstandart |
| The Republic Of Armenia | Armastajad |
| The Republic Of Belarus | Gosstandart Of Belarus |
| The Republic Of Kazakhstan | Gosstandart Of The Republic Of Kazakhstan |
| The Republic Of Moldova | Moldovastandart |
| Russian Federation | Gosstandart Of Russia |
| The Republic Of Tajikistan | Tajik state center for standardization, Metrology and certification |
| Turkmenistan | Turkmengeologiya |
| Ukraine |
Gosstandart Of Ukraine |
The change in N 1 taken by the Interstate Council for standardization, Metrology and certification (Protocol No. 24 dated 5 December 2003)
For the adoption of the changes voted by national standardization bodies of the following States: AM, BY, KZ, MD, RU, TJ, TM, UZ, UA [codes alpha-2 at MK (ISO 3166) 004−97]
3 Resolution of the State Committee of the Russian Federation for standardization, Metrology and certification, dated 27 February 1997 No. 71 interstate standard GOST 30415−96 introduced directly as state standard of the Russian Federation from January 1, 1998
4 INTRODUCED FOR THE FIRST TIME
5 EDITION (February 2005) with amendment No. 1 adopted in March 2004 (I & C 6−2004)
1 Scope
This standard covers rolled, sheet, strip, structural shapes, sheets of non-magnetic coatings, tubes, layered sheets and strips of carbon, alloy and electrical steels, products made from the above metal, and installs the nondestructive magnetic method of testing of mechanical and technological properties, microstructure and resource characteristics.
The standard can be extended to other types of steel products in coordination with the consumer.
Non-destructive magnetic method of control is applied along with the test methods prescribed in the standards, by definition:
yield stress of the physical, conditional, temporary resistance, relative elongation after rupture, the relative narrowing of the cross section after the rupture of the churchyard 1497, GOST 10006;
relative uniform elongation according to GOST 1497;
coefficient of plastic anisotropy, indicators strain hardening and non-uniform plastic deformation according to GOST 11701;
true the tear resistance according to GOST 10006;
hardness according to GOST 2999, GOST 9012, GOST 9013, GOST 22975, GOST 23273;
the grain size according to GOST 5639;
poloschatosti and structurally-free cementite according to GOST 5640;
the sensitivity of steel to mechanical aging according to GOST 7268;
impact bending according to GOST 9454;
share viscous component in the fracture according to GOST 10006;
the depth of the holes according to GOST 10510;
the number of inflections according to GOST 13813;
flattening according to GOST 8695;
the angle of bend or estimate the limit of plasticity in bending according to GOST 14019;
depth bezoperatsionnogo layer according to GOST 1763;
relative deformation at a Deposit with GOST 8817;
bending according to GOST 3728.
2 Normative references
The present standard features references to the following standards:
GOST 27.002−89 reliability of the technique. Basic concepts. Terms and definitions
GOST 27.202−83 Reliability in technique. A technological system. Methods of reliability evaluation by parameters of quality of manufactured products
GOST 1497−84 (ISO 6892−84) Metals. Test methods tensile
GOST 1763−68 (ISO 3887−77)Steel. Methods for determining the depth bezoperatsionnogo layer
GOST 2999−75 Metals and alloys. Method of measurement of hardness by Vickers
GOST 3728−78 Pipe. Method of bend test
GOST 5639−82 of Steel and alloys. Methods of detection and determination of grain size
GOST 5640−68 Steel. Metallographic method for the evaluation of microstructure of sheets and strips
GOST 7268−82 Steel. The method of determining the propensity to mechanical aging test the impact strength
GOST 7564−97 hire. General rules of sample, blanks collection for mechanical and technological tests
GOST 8695−75 Pipe. Test method for flattening
GOST 8817−82 Metals. Test method for sediment
GOST 9012−59 (ISO 410−81, ISO 6506−81) Metals. Method of measuring hardness Brinell hardness
GOST 9013−59 (ISO 6508−86) Metals. Method of measuring Rockwell hardness
GOST 9454−78 Metals. Test method for impact strength at low, room and elevated temperatures
GOST 10006−80 (ISO 6892−84) of the Pipe metal. Test method tensile
GOST 10510−80 (ISO 8490−86) Metals. Test method for the extrusion of sheets and strips for Eriksen
GOST 11701−84 Metals. Methods tensile tests of thin sheets and strips
GOST 13813−68 (ISO 7799−85) Metals. Test method for bending of sheets and strips thickness less then 4 mm
GOST 14019−80 (ISO 7438−85) Metals. Methods of bend test.
GOST 15467−79 quality Control of products. Basic concepts. Terms and definitions
GOST 15895−77* Statistical methods of quality management. Terms and definitions
_______________
* On the territory of the Russian Federation there are GOST R 50779.10−2000, GOST R 50779.11−2000.
GOST 16504−81 System of state testing products. Testing and quality control. Key terms and definitions
GOST 18321−73 Statistical quality control. Methods of random selection of piece products
GOST 20736−75* Statistical acceptance control of quantitative trait. Plans control
___________________
* On the territory of the Russian Federation GOST R 50779.74−99.
GOST 22975−78 Metals and alloys. Method of measuring Rockwell hardness at low loads (for Super-Rockwell)
GOST 23273−78 Metals and alloys. Hardness measurement by the method of elastic rebound of the striker (shore a)
GOST 27772−88 rolled products for structural steel constructions. General specifications
Sections 1, 2 (Modified edition, Rev. N 1).
3 General requirements
3.1 non-destructive magnetic method of control is applied in the presence of a stable pair or multiple probabilistic relations between controlled terms of quality and magnetic properties of the steel.
All the probabilistic estimates used in this standard are applied at a confidence level not lower than 0,95.
When continuous or single-piece non-destructive magnetic method of control adopted is likely to provide norms of standards should be ensured in every batch.
3.2 Correlation between magnetic characteristics and quality indicators is determined at each facility based on the information file for each grade or groups of grades, differing mainly by carbon content.
Allowed the grouping of different grades of steel and similar profiles of hire, if calculated by the combined sample regression equation has a significant correlation coefficient.
If necessary, the control is carried out taking into account other structure-sensitive characteristics, the chemical composition of the metal, the technological parameters and conditions of metal.
(Changed edition, Rev. N 1).
3.3 Terms, basic concepts and notation — in accordance with GOST 27.002, GOST 16504, GOST 15895 GOST 15467, GOST 18321, GOST 20736.
4 controls
4.1 For non-destructive testing according to this standard apply devices that measure one or more structural-sensitive characteristics with a basic accuracy of no more than 5% in the working range of measurements.
4.2 the results of measurements of magnetic characteristics of metal magnetic method should not affect other ferromagnetic body and the electromagnetic field, the characteristics of which do not meet the requirements and conditions of operation of the devices.
5 procedure for the preparation of the control
5.1 Selection of samples for testing — according to GOST 7564.
5.2 the Number of samples subjected to NDT is to be stipulated in the normative documents for the products.
5.3 the Number of measurements of the magnetic parameter and the direction of installation of two-pole sensors on the test plots of samples shall be stipulated in the normative documents for carrying out non-destructive testing.
5.3 a Procedure for determining the limit of the product must be specified in the normative document on metalloizdelie.
(Added, Rev. N 1).
5.4 Upon non-destructive testing of quality indicators estimated on the basis of «satisfactory-unsatisfactory», set the allowable limit of the measured magnetic characteristics, guaranteeing the established norms adopted in the standard probability.
5.5 use refined the quality of steel products, isklyuchaya error destructive tests.
5.6 the lower bound of the confidence interval of correlation coefficient in absolute value must be above its critical value at the significance level of 0.05.
6 Procedure control
6.1 Mandatory definition are subject to the statistical characteristics for each information array, the composition of which is given in Appendix A.
6.2 the Value of acceptance numbers are calculated by the formulas:
where — norm of the
-th indicator of quality according to the corresponding standard features, the normalized lower;
norm
-th quality indicator according to the corresponding standard features, the normalized upper;
— residual standard deviation
on measure of quality, defined by the formulas:
where ,
— quality indicators determined by non-destructive and destructive tests;
— the sample volume;
— standard deviation
-th indicator of quality;
— correlation coefficient;
the value of student criterion for the accepted confidence probability.
If the values of the results of nondestructive testing beyond a limited acceptance numbers, the party being tested arbitration methods
.
6.3 Level of quality score in the party complies with the requirements of normative documents, if for each characteristic the following conditions are true:
Controlled products that meet the above conditions, the test is not exposed, and the test results are tabulated the calculated values of the quality indicators.
6.4 Products that do not meet 6.3 shall be tested according to GOST 1497, GOST 1763, GOST 2999, GOST 3728, GOST 5639, GOST 5640, GOST 7268, GOST 8695, GOST 8817, GOST 9012, GOST 9013, GOST 9454, GOST 10006, GOST 10510, GOST 13813, GOST 14019, GOST 22975, GOST 23273.
6.5 To assess sovpadaet of the results of determination of quality indicators non-destructive and destructive methods of the manufacturer shall be subjected to parallel testing by these methods at least 10% of controlled quantities of metal during the whole monitoring period of hire.
6.5 and Finished products of steel products are subjected to non-destructive testing the manufacturer and the consumer before the commissioning and during the operation process controlled by the frequency set forth in normative document on metalloizdelie.
(Added, Rev. N 1).
6.6 Pipes and wires made from blanks supplied with the assessment of quality indicators, are subjected to parallel testing of these methods in the amount necessary for the formation of a representative sample over the period of monitoring.
7 Processing of results
7.1 To ensure the unity of the method and for sopostavimy results of non-destructive magnetic control of mechanical properties of rolled products and pipes, it is recommended to adhere to a formalized decision-making procedure when constructing mathematical models of mechanical properties. The description of technology of the automated construction of mathematical models is given in Appendix B.
7.2 carrying out of calculations of pair and multiple correlations and the construction of regression equations by the recovery method of correlation of data dependences inconsistent measurements, i.e., measurements obtained on the samples taken separately, but belonging to a given population, according to the method specified in Appendix V.
7.3 Evaluation of sovpadaet of the results of determination of quality indicators non-destructive and destructive methods is carried out using control charts, analytical or graphical methods.
It is acceptable to combine the control map results along with the process control of the mechanical properties group of the thickness of rolled and steel grades.
7.4 the Number of deviations outside the control range, shall not exceed 5% during the whole monitoring period. If unsatisfactory test results, the control parties shall be conducted in accordance with the requirements of state standards and technical conditions for steel.
7.5 Evaluation of quality indicators is satisfactory if the shift of the distribution with respect to the Central line does not exceed ±0.5. With a larger displacement of the center of the distribution of deviations is carried out correction of the regression equations; a conclusion on the need of the specified adjustment shall be made on the basis of processing a sample of at least 50 parties.
7.6 test report enter the number of the normative document according to which products are delivered, grade of steel, thickness, size of controlled products, heat number and batch, the values of the magnetic characteristics and quality indicators.
7.6 and the test of metal point: a normative document according to which they are made; the conditions and modes of operation; a means of measurement; the measured values of the magnetic characteristics; the parameter values of mechanical and technological properties, microstructure and resource characteristics, calculated on the basis of physically grounded connection with magnetic characteristics for each grade of steel and manufacturing of products indicating the source, which presents the dependence used.
(Added, Rev. N 1).
7.7 In the test report for products monitored under this standard shows the mechanical properties in the units established product standards.
7.8 In the case of continuous or piece of non-destructive control in technological production flow in the test report specifies the quality level of the party, provided with the regulatory documents on products, adopted in the standard confidence level.
Annex a (mandatory). The composition of the characteristics subject to mandatory definition with non-destructive magnetic method of control of mechanical properties
APPENDIX A
(required)
Table A. 1
| Marking |
Definition | Regulatory document |
Tasks |
| The sample matrix of observations |
GOST 15895 | Data collection | |
( |
Indicators |
GOST 15895 | The representativeness of the sample |
The sample size |
GOST 20736 | The adequacy of observations | |
| The average value |
GOST 27.202 | Evaluation of the main statistical characteristics | |
| The standard deviation |
GOST 27.202 | Evaluation of the main statistical characteristics | |
|
The confidence interval of the average value |
GOST 27.202 | A definition of the limits of change |
| Statistics student’s t test to test the hypothesis about equality of averages |
GOST 27.202 | Test of data homogeneity and stability technologies. The pooling of samples | |
| Statistics Fisher test hypotheses about equality of variances |
GOST 27.202 | Test of data homogeneity and stability technologies. The pooling of samples | |
| The correlation coefficient to assess linear relationships between the indicators |
GOST 27.202 | Assessment of the level of linear correlation. Verification of the hypothesis according | |
| Statistics student’s t test to check the significance of correlation coefficient |
GOST 27.202 | To test the hypothesis of significance of correlation | |
| The residual standard standard deviation error of the regression |
GOST 15895 | The establishment of confidence limits of the regression equation | |
| Multiple correlation coefficient between the target and the set of impact indicators (characteristics determined, if necessary, multi-factor control) |
GOST 27.202 | Assessment of the level of multiple linear (linearized) dependence | |
| Acceptance number quality indicator, regulated from below |
GOST 27772 | Certification of products | |
| Acceptance number quality indicator, top regulated |
GOST 27772 | Certification of products |
Annex B (reference). The description of technology of the automated construction of mathematical models using computers
APPENDIX B
(reference)
B. 1 Preparation of media and control data input
In the process of preparing the initial information in machine-readable form by the technical control consists in verifying each number on a wrong symbol.
Error of training data are identified through a printout of information and analysis of calculated tables of the main statistical characteristics of indicators of mechanical properties, chemical composition of the steel, the magnetic properties and other parameters.
After adjusting count data statistical characteristics and proceed to the formation of the working array and data analysis using selective methods of mathematical statistics.
B. 2 organization of the working array. Analysis of test results
Of the plurality of parameters comprising initial information, the suspended form a group of factors (work array) containing all the influencing variables and the indicator of mechanical properties.
Values of indicators of quality, do not carry information in the context of the problem being solved and the corresponding values of the independent influencing variables from the sample are removed. In this case, the statistical characteristics recalculated.
The exclusion of outliers is done based on qualitative and quantitative analyses of the sample.
When a large number of observations used «three Sigma rule», the observation is excluded if its deviation
exceeds 3
, where
— mean square value of the quality indicator.
According to a more accurate evaluation criterion of the anomalous values is considered an ordered sample of observations
where is the number of observations in each indicator.
To evaluate the affiliation and
this together and make a decision about excluding or leaving
The results compare with the table value of the criterion of Smirnov on the computation of critical values at a probability
, which are determined by:
for a given size and level of significance
of 0.05.
If 
This criterion is used for small samples with a volume of 50.
B. 3 characterization Study of the distribution and bringing to normal
The target (indicator of quality) of the formed groups of factors examined for normality of distribution.
Verification of normality of distribution of indicators is carried out according to the criteria: the Pearson sample size in excess of 200; Kolmogorov for the sample size of 100 or greater and von Mises-Smirnov, for sample size greater than 50.
In case of absence of normality of distribution of proceeds from the baseline to the other variable
through functional transformation of the data.
In the case of normality of distribution of the target, or satisfying the normality calculate its statistical characteristics have a known distribution for these characteristics it is possible to set confidence bounds changes, and then the evaluation of future models are justified with probabilistic and statistical point of view, which allows you to move to the next stage of modelling for this scheme.
If the transition to normality is not implemented, then this entails the unreliability of the statistical estimates of the future model.
B. 4 Estimation of volume measurements
If the sample size for the target parameter is not less than calculated according to the following formulas, that is, the transition to the next stage of statistical processing of data, otherwise the information is collected for the replenishment of the sample and the modeling process is carried out for the augmented sample in accordance with the scheme.
Let is the mean value of observations in a simple random sample and probability
where is selected limit value error;
— some small probability;
— the General average.
As an approximation of the minimum volume of sample is set to
where is the abscissa value for a normal curve which cuts off at the «tails» area
.
B. 5 Analysis of paired dependencies
The presence of a linear correlation between indicators and
is determined by a comparison of the correlation coefficient
and correlation ratio
.
If the difference 
If the difference is 
and
.
With the aim of identifying species based on a curvilinear constructed correlation field and the empirical regression line, establish connection and performance
, select the analytical formula
All of the chosen constraints must reflect the qualitative dependence of mechanical properties on the influencing indicators.
B. 6 building a model
As a statistical method of establishing a relationship between a dependent variable and a set of influencing parameters (
) used step-by-step method of constructing a multiple regression allows you to include or exclude explanatory variables
in order of importance.
Parameter estimation is performed for linear and linearized models of the form:
where is the baseline set (
) or values derived from (
) by algebraic transformations;
,
are the coefficients of regression, estimation of model parameters.
Criterion step-by-step build regressions based on the reduction in residual sum of squares of the equations (B. 6), it is introduced in the regression, the variable most influencing the decrease in this step, excludes the least affected.
The procedure of constructing the model continues until until you exhaust all various 
. Step-by-step construction suggests the movement in areas promising from the point of view of reducing the residual sum of squares. The final choice of model is determined by statistical reliability in General, and the statistical reliability of each of the resulting estimates
of the model parameters.
At each step of building a regression model to calculate its characteristics:
where the sum of the squared deviations of the dependent variable from its mean;
is the number of observations for each variable.
— the number of variables in the regression equation at this step;
— regression coefficient;
— standard errors of the regression coefficients, computed as the matrix elements of the inverse correlation.
Estimation of the parameters of the regression model according to the method of least squares is chosen at each step such that the value characterizing the measure of the spread of the experimental data against the predicted model values were minimal.
When evaluating the quality of the model values of indicators of reliability of the regression coefficients is compared with a limiting value of statistics t-test
(
adopted level of significance,
is the number of degrees of freedom), and the value of
the reliability coefficient of the multiple correlation coefficient is compared with the tabular value of Fisher statistics
— the adopted significance level
If 
-th regression coefficient is deemed to be reliable. If
__________________
* The text matches the original. — Note the manufacturer’s database.
First, you need to obtain a model with reliable estimates of regression coefficients and multiple correlation coefficient, the minimum approximation error and standard error of the model estimates.
Stop should be on the models built, which has reliable estimates of the regression coefficients
, robust estimates of
the multiple correlation coefficient, the lowest standard error of estimation of the model, a sufficiently high coefficient of multiple correlation
as a measure of determinism of the relationship of the target variable
with the independent variables
, and also has a variable composition
, is acceptable in the context of the problem being solved.
Annex b (informative). Methods of recovery and regression of dependencies from empirical data
THE APP
(reference)
To recover quantitative agreement between the values of the mechanical properties of rolled and measured physical parameters in the case when samples are inconsistent and have different number of dimensions, it is proposed the method of finding the coefficients of the control equation based on the recovery of correlations. The basis of application of the method of recovery is the stability properties engendered by this technology, and the normal law of the joint distribution of measurement values
When you restore the dependencies of different formulation of problems are reduced to a mathematical scheme of minimization of average risk on empirical data.
It is believed that the indicators and
associated regression dependence if each value
of the index
corresponds to the number
obtained by random testing over the indicator
according to the conditional probability density
is defined in accordance with the law
.
Full knowledge of regression requires the restoration of the conditional density 
called regression.
Task restore the function of the conditional expectation in this case is formulated as the task of reconstruction regression is one of the main problems in applied statistics.
Statement of the problem consists in the following.
When tested randomly and independently appear in the measurement values . In this environment, the Converter operates
puts in the number
obtained as a result of the implementation of random testing, according to the law
Properties of the environment and the law
Requires a pair of random independent samples in General have different volume
,
,
to recover the regression, that is, in the class of functions 
.
Here ,
the volumes of independent samples on the indicators
,
and
the designation of a class of regression functions with different parameter values
belonging to
a range of values.
The task of rebuilding the regression is reduced to the problem of minimization of the functional
on the set 
functions in a situation where the joint probability density
It can be shown that if the regression 
. If the regression does not belong
The proximity of functions is understood in terms of a metric (quadratic measure):
Write a formula (B. 4) in the General form
where — average risk;
in the problem of minimization of average risk when restoring a regression from empirical data ,
…
The minimum value of (B. 6) is achieved with a confidence probability , called the reliability of the restoration.
Practical solution to the problem, ensuring minimization of the average risk of recovery of a regression with a given reliability on samples of the beds in the volume, is to construct equations of the selected 
of the joint distribution of values of indicators
,
.
where is a vector of parameters including the reference value of the joint distribution of measurements
,
including average value
,
average quadratic deviation
,
and the paired correlation coefficient
. Is a solution of the equation relative to
the sample reference values. In particular,
,
is determined by the equation of the ellipsoid
with stretching, the corresponding assigned confidence probability and sample size.
Setting statistical hypotheses on limit values 
, which allows to determine the gauge factor
and offset
recovery regression
between mechanical characteristics and the measured physical parameter.
As a rule, equations (B. 7) are relatively non-linear , therefore it is advisable to use one of the approximate methods of finding solutions at iteration
step
where is a unit vector in the direction of the gradient;
— the step value.
The electronic text of the document
prepared by JSC «Code» and checked by:
the official publication of the
M.: IPK Publishing house of standards, 2005
