COP v3.0:structural-testing; publishing-results

17.5 Publishing Results 

Test results must not be published as load-span data. Before they can be offered as a design guide they must be checked and interpreted by an engineer and have suitable factors applied to the test result figures. It is highly desirable that this person has witnessed at least some of the tests being conducted. As fastener pull-out is not included in testing, published load-span tables should encourage the designer to check for pull out values of specified fasteners when fastening into timber or ply less than 30 mm thick, into composite materials, or into steel sections less than 1.2 mm thick.

17.5.1 Allowance for Variation 

17.5.1A Factors to Allow for Variability of Structural Units

No of units testedCoefficient of Variation (COV)
 5%10%15%20%
11.21.461.792.21
21.171.381.641.96
31.151.331.651.83
41.151.31.51.74
51.131.281.461.67
101.11.211.341.49
201.061.131.211.29
1001.011.011.011.01

Coefficient of Variation (COV) is derived from Standard Dev / Av for a group of tests.  The COV to be applied to a series of tests is the average overall spans measured during a test schedule conducted over a continuous period using identical material.

For serviceability loads, a coefficient of 5% may be assumed unless there is evidence that a higher figure is warranted. For strength loads, a coefficient of 10% may be assumed unless there is evidence that a higher figure is warranted. For clip fastened products the applied depreciation factor may not be less than 30%.

The number of units tested relates to the number of valid results derived from a given fastener configuration and purlin spacing.

For values between those listed in the table, interpolation may be used.

 

17.5.2 Interpreting UDL Load Results 

When using data from testing for producing graphs, tables, or other design aids, the following conditions apply:

  1. Data must not be extrapolated except where a minimum of four span-combinations within the limitations of (b) below can be shown to provide a statistically reliable load span graph. In such cases, an extrapolation of a further ±10% at either end may be calculated.
  2. Interpolation of data between different spans of a specific type of test is acceptable only in the following circumstances:
    1. Where the data is taken from tests for a single type, size, and profile of cladding and type and spacing of fastener.
    2. Where at least three different spans or support spacings have been tested for the same type, size, and profile of cladding and fastener details, and that in all three tests the failure mode was the same.
    3. Where test loads were derived from the same test criteria.

17.5.3 Interpreting Point Load Results 

Point load requirement for trafficable roofs is defined in NZS 1170 as 1.1 kN, which approximates a 100 kg person on one foot. (Note: many workers carrying tools or equipment exceed this limit.) To allow for statistical variation this is factored by 20% and a test load of 1.32 kN is applied to give a pass/fail.

Type A roof (unrestricted access roofs) is required to withstand a 1.32 kN test load on the rib without deformation or excessive deflection, (serviceability load) and a 2.41 kN strength load.

Type B roof (Restricted access roofs) is required to withstand a 1.32 kN test load in the pan or over two ribs without deformation or excessive deflection, (serviceability load) and a 2.41 kN strength load.

Type C Roof (non-trafficable) is required to withstand a 0.6 kN test load to the pan or over two ribs, and a 1.1 kN strength load. See 3.6.2 Roof Traffic.