A.6.3.2 Characteristic values for families of tests
(1) A series of tests carried out on a number of otherwise similar structures, portions of structures, members, sheets or other structural components, in which one or more parameters is varied, may be treated as a single family of tests, provided that they all have the same failure mode. The parameters that are varied may include cross-sectional dimensions;, spans, thicknesses and material strengths.
(2) The characteristic resistances of the members of a family may be determined on the basis of a suitable design expression that relates the test results to all the relevant parameters. This design expression may either be based on the appropriate equations of structural mechanics, or determined on an empirical basis.
(3) The design expression should be modified to predict the mean measured resistance as accurately as practicable, by adjusting the coefficients to optimize the correlation.
NOTE: Information on this process is given in annex Z** of ENV 1993-1-1.
(4) In order to calculate the standard deviation s each test result should first be normalized by dividing it by the corresponding value predicted by the design expression. If the design expression has been modified as specified in (3), the mean value of the normalized test results will be unity. The number of tests n should be taken as equal to the total number of tests in the family.
(5) For a family of at least four tests, the characteristic resistance Rk should then be obtained from expression (A. 11) by taking /?m as equal to the value predicted by the design expression, and using the value of k from table A.2 corresponding to a value of n equal to the total number of tests in the family.
This annex is in preparation.
A.6.3.3 Characteristic values based on a small number of tests
(1) If only one test is carried out, then the characteristic resistance Rk corresponding to this test should be obtained from the adjusted test result /?adj using:
in which r\k should be taken as follows, depending on the failure mode:
(2) For a family of two or three tests, provided that each adjusted test result /?adj, is within ± 10% of the mean value Rm of the adjusted test results, the characteristic resistance Rk should be obtained using:
(3) The characteristic values of stiffness properties (such as flexural or rotational stiffness) may be taken as the mean value of at least two tests, provided that each test result is within ± 10% of the mean value.
A.6.4 Design values
(1) The design value of a resistance Rd should be derived from the corresponding characteristic value Rk determined by testing, using:
7M is the partial factor for resistance;
7sys is a partial factor for differences in behaviour under test conditions and service conditions.
(2) For a family of at least four tests, the value of 7M may be determined using statistical methods. NOTE: Information on an appropriate method is given in annex Z#) of ENV 1993-1-1.
(3) Alternatively 7M may be taken as equal to the appropriate value of 7M for design by calculation given in Section 2 or Section 8 of this Part 1.3.
(4) The appropriate value for 7sys should be agreed between the client, the designer, the testing organisation and the competent authority.
(5) For sheeting and for other well defined statndard testing procedures (including A.3.2.1 stub column tests, A.3.3 tension tests and A.3.4 bending tests;) 7sys may be taken as equal to 1,0.
(6) For other types of tests in which possible instability phenomena, or modes of behaviour, of structures or structural components might not be covered sufficiently by the tests, the value of 7sys should be assessed taking into account the actual testing conditioas, in order to achieve the necessary reliability.
(1) The provisions given in Section 7 should be satisfied.
This annex is in preparation.
BSI — British Standards Institution
BSI is the independent national body responsible for preparing
British Standards. It presents the UK view on standards in Europe and at the international level. It is incorporated by Royal Charter.
Was this article helpful?