Steel Properties in a Technical Inspection

Why steel properties are refined

In the design of new structures, the design properties of steel are assigned from standards that set values on the basis of statistical processing of sample tests at steelworks. In service and reconstruction, at the technical inspection the properties of the steel must be refined for each structure or group of structures. As mass inspections show, this approach reveals considerable reserves of load capacity in structures.

How much each property matters depends on the working conditions. For example, at low temperatures the tendency of steel to brittle fracture is especially important, and for crane girders the fatigue strength characteristics are. The properties can be established by direct methods (testing samples and specimens taken from the structures, which is more accurate but laborious) and indirect ones (non-destructive testing, less accurate but simpler).

The kind and number of tests depend on an analysis of the existing documentation: KM drawings (steel grades, electrodes, metal products), KMD drawings (deviations from the design), steel certificates, work records and logs, and earlier inspection reports. Tests are mandatory when there is not enough information about the material of the structures, and when damage linked to the quality of the metal (lamination, cracks, brittle fracture) is found.

In studying and testing the metal the following are determined:

• chemical composition: carbon, silicon, manganese, sulphur and phosphorus (for all steels), chromium, nickel and copper (for low-alloy steels) and, where needed, nitrogen;
• yield strength, tensile strength and elongation in tension;
• impact toughness at −20 °C for carbon steel and −40 °C for low-alloy steel (after mechanical ageing where needed);
• the distribution of sulphide inclusions by the Baumann print method.

For less critical structures working at low stresses and not at very low temperatures, some of the tests may be omitted.

Number of samples and specimens, and their sizes

The number of samples and specimens taken from the metal of structures should be the minimum that gives sufficiently reliable results, and likewise the volume of metal cut out: more and larger specimens raise the labour and weaken the structure. To identify the material, the metal of the structures is divided into lots: a lot consists of elements of one type of rolled product (by number, thickness and steel grade) that belong to the same structures. The number of samples from each lot must be at least the number set by the norms.

For chemical analysis, swarf obtained by drilling particular places on the structures is taken, working the whole cross-section of the rolled product or a symmetric half of it. Before a sample is taken, the surface is cleaned of grease, dirt, paint and corrosion products down to bright metal; the sample is taken without lubricant.

For tension tests, blanks are cut out of the structures from which cylindrical specimens 3 mm in diameter or more, or flat specimens 0.5 mm thick or more, are made; short specimens are preferable in service conditions. The basic cylindrical specimen is 10 mm in diameter; for sheet steel and shaped sections up to 25 mm thick, flat specimens are used with the rolled surface layers kept. Allowances are needed when cutting so that the specimen is protected from work hardening and heating; for flame cutting, at least 15 mm from the cut line to the edge of the finished specimen for rolled products up to 60 mm thick and 20 mm above 60 mm. For impact toughness, standard specimens with a U or V notch are used, depending on how critical the structures are.

Where samples are taken

Sampling places are chosen with the stress level of the element in mind: the properties of the steel must be established at their most unfavourable limits, so the properties of the metal of the most heavily loaded elements are of most interest. The elements with the highest stress level are identified first. In service, blanks are usually cut out with a flame and the places weakened by the cut are strengthened by arc welding, so the sampling places must be accessible.

The texture of the rolled elements has a significant effect on the results, so the rolling direction should be established and specimens taken and made: from sections and shaped products along the rolling direction; for plate elements whose rolling direction is not known, along the main force flow.

Analysing the results and assigning design properties

Chemical analysis gives the percentage of the elements in the steel. First, attention is paid to the harmful impurities, sulphur and phosphorus, which make steel red-short and cold-short: depending on operating conditions, a raised content may rule out further use of the structures. Particular attention is paid to the degree of deoxidation of the steel, judged roughly from the silicon content. To refine it (and when deoxidisers without silicon are used) Baumann prints are used: a sheet of photographic paper soaked in a 5% solution of sulphuric acid is laid on a prepared section; after 5–10 minutes the macrostructure appears, showing the sulphide inclusions. On the print of rimmed steel many dark bands (sulphide rows) show, of semi-killed steel single sparse rows, and of killed steel there are none or only sparse point inclusions.

Weldability is also assessed from the chemical composition; it is a composite characteristic including joint strength, resistance to hot and cold cracks and cold-shortness. In welded structures in service, the condition of the welds is an indirect assessment: if no cracks caused by welding itself are found, no further assessment is usually needed. Weldability is usually assessed from the carbon equivalent: steel is considered to weld well when it does not exceed 0.45%; ordinary low-carbon steel welds well with carbon up to 0.22%, sulphur up to 0.055%, phosphorus up to 0.050% and silicon up to 0.22%.

The main strength characteristics are established from the results of mechanical tests. Since the load capacity of existing structures is determined by the limit state method, the characteristic and design resistance of the steel must be assigned from the test results. If 10 or more specimens have been tested for each lot, the characteristic resistance can be determined by statistical methods: the value of yield strength with a reliability of at least 95% is taken (the distribution of yield strength is close to normal); with a limited sample it is taken into account that the statistical characteristics may differ from the parameters of the whole population, and the confidence probability is set according to how critical the structures are. In all cases the characteristic resistance is not taken higher than the minimum yield strength obtained.

If the number of tests is limited and the results cannot be used directly, the steel grade or its analogue under the norms of the period of construction is established from the test material and the design properties are assigned with experience in mind. The material reliability factors for existing structures are set by the applicable design norms for steel structures.

Rapid analysis and indirect methods

Indirect non-destructive methods (hardness testers, ultrasound and so on) are less accurate but fast and do not weaken the structure; they are used for a rapid assessment of strength and for choosing the places where direct tests are needed. Final conclusions on load capacity are based on direct tests of samples and specimens.