Mech testing of spun grey cast iron pipe

[Simon55]

Due to some recent failures of a spun grey cast iron water mains (due to graphitic corrosion from both internal and external surfaces) I would now like to determine the mechanical properties of the pipe material. This information would then be used to assess the critical flaw size and potential remnant life.
Could someone please tell me the best means for determining the tensile strength? As i understand it, tensile testing would require a very large sample population to yield meaningful results (due to the pitting etc). Can a correlation be made from the results of 4 point bend testing? How else can it be done? any comments or advice are appreciated.

 

[stanweld]

Tensile testing is not really a problem. However, is tensile strength your real concern? Centrifugal casting will produce highly elongated grains and graphite flakes in the through thickness direction of the pipe wall, which appears to be the cause of your problem.

 

[Simon55]

Perhaps this needs some clarification...
Centrifugal casting produces a variable microstructure through the wall thickness. A very fine structure of irregular graphite nodules in a ferrite matrix is produced at the outer surface, an interdendritic fine flake structure through the mid section, to a coarse flake, randomly oriented structure at the inner surface - perhaps with some rosette characteristics. Thus you are partially right - the microstructure is part of the problem. The mechanical properties of the pipes therefore vary with the microstructure at different wall thicknesses.
Once again the problem - how can I determine the tensile strength of such a variable material? Especially if it is quite pitted and there has already been some loss of section from both the internal and external surfaces.

A straight forward tensile test would produce a huge scatter in results and require a potentially very large sample population to get meaningful results. I was hoping to avoid this by using some more robust test method.... Once again, any input (thoughts, comments, advice) is appreciated.
cheers
Simon

 

[LenStott]

Simon, I don't think you will get any meaningfu;l results as any material that has undergone the influence of corrosion will give extremely variable results.