Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]


[Subject Prev][Subject Next][Thread Prev][Thread Next]
Title: RE: ASTM A36

>Paul Ransom wrote:
>Designers: At the very least, include
>a requirement for random mechanical
>testing in your specs. It is not
>adequate to rely on the
>"mill certs."

I'm all for quality and quality assurance, but I think this is too much for routine materials and mill certs can be accepted as evidence of conformance (note comments on mill certs below). By routine materials, I mean the standard base grades for the various structural products (summarized periodically in Modern Steel Construction and listed in Part 2 of the AISC Steel Construction Manual) and common alternative grades. The exception I recognize for this is the case of a party to a contract engaging in deceptive behavior. If that is a concern, the nature of the problem is not the mill cert, but rather the deceptive behavior. Otherwise, you'll only be adding cost for the owner unnecessarily if you call for random testing of all structural products.

Conversely, if you have special-order materials with a special characteristic (Chris Wright's T-1 plow blade story  comes to mind) that is the basis of the function and performance of the structure, special testing might be appropriate. Such testing should be clearly defined in the project specification with references to the appropriate ASTM specifications and acceptance criteria identified. What I mean by this is, I fear the result of a project specification or drawing note that attempts to do well with a general statement like "Contractor to perform random material testing ..." but actually defines nothing about what is required, how the testing is to be performed, or what test result is acceptable.

There are so many variations inherent in all that we do. Loads are an educated guess to two decimal places. Analysis is based upon rampant simplification. And so on, including material strengths. In steel, the normal variations on the strength side are small and accounted for in phi factors and omega factors. There are so many other places I'd be looking at variations before I worried about steel strength, if I thought variations were a problem. Even in concrete where we mix a batch of rocks together, our past history shows us that material variations are addressed properly in our design standards. Note that comment is NOT intended as a slam on my friends in the concrete industry -- we steel folks use plenty of concrete in steel buildings.

I think there might be a temptation implied to view a mill cert as something that it isn't. A mill cert is a summary of the production testing that a mill does on a heat of steel. It's important to note that production testing is not lab testing. As a result, I'd not take the values from a mill cert and use them for design. A mill cert only indicates that the product to which it corresponds meets the same production-testing criteria established in a given ASTM standard. That ASTM standard then establishes the design values that can be used with the AISC Specification for that product.