The technological development of prestressing steels was accompanied by research to establish parameters and test methods for susceptibility to H-SCC. Early prestressed concrete structures mostly used hot rolled steel rods with a yield stress around 600 N/mm2 and a tensile strength of 900 N/mm2. This steel quality, St 600/900, does not differ much from St 835/1030 used today, except that the yield strength is increased by straining to 835 N/mm2. This type of prestressing steel shows high resistance to the formation and growth of hydrogen-induced cracks under corrosive loads.
When high strength quenched and tempered and cold drawn prestressing steels with a tensile strength up to 1800 N/mm2 48-52 were introduced, damage caused by cracking was reported for the first time. The original case was of delayed fractures that could be attributed to SCC, which occurred within a few days after prestressing. This and the many other cases of damage in structures made of aluminous cement3,53,54 led to investigations that would provide more information about the mechanism of hydrogen-induced SCC and help metallurgical development and design of prestressing steel.
Cases of damage had been reported prior to this experience, for example at the 1955 Federation International de la Precontrainte Congress.55 Engineers observed wires broken due to corrosion on the reel in three projects (see e.g. ref. 3). As a result, quenched and tempered steel was effectively no longer used in the Netherlands after about 1952.
The initial tests used well-known test solutions that had been proven for unalloyed steels used for other purposes. This was not only because of a lack of information about the mechanism of hydrogen-induced stress corrosion, but also because the composition of the electrolytes that would occur during construction was unknown. Thus highly concentrated solutions of calcium and ammonium nitrate, developed for testing boiler steels for caustic embrittlement,56 were used. With better knowledge of the mechanism of crack formation in prestressing steels, and especially the influence of hydrogen generated by corrosion, test solutions with high hydrogen activity were introduced, such as saturated aqueous H2S solution, well known from tests of oil pipelines, or NH4SCN solutions. Specimen design was limited to bend specimens because they were easier and more cost-effective to produce than tensile specimens.
At the beginning of 1974, a new prestressing steel, type St 1080/1320 (St 110/ 135) (smooth and ribbed), was introduced for prestressing steel rods in prestressed concrete structures. However, within the first year of use, damage occurred in a number of structures where this steel was used, and the number of failures increased to 60 within a short period of time.3 This was unexpected since it had been assumed from the results of notched-bar tensile and fracture mechanics tests that this steel had much better brittle fracture properties than the steel St 835/1030 (St 85/105) previously in use.
This experience showed that the behaviour demonstrated in brittle fracture tests in the laboratory was not always transferable to actual conditions in a way which is valid for all prestressing steel qualities. It was found that, under construction conditions, St 110/135 is essentially more susceptible to delayed cracking than St 85/105. The two steel qualities differ mainly in their chemical composition and their microstructure, rather than in their manufacturing processes. As a result of these failures, it was decided to develop a test method that would enable definitive statements to be made about the suitability of prestressing steels under construction conditions. DlBt (formerly IfBt) and EU initiated an extensive cooperative research programme (using the research facilities of FMPA Stuttgart, BAM Berlin, MPI DuÈsseldorf, Krupp AG Rheinhausen) which resulted in the development of a test method45 that would be capable of distinguishing between susceptible and non-susceptible prestressing steel qualities and therefore identifying those that could be used safely under construction conditions. Details of this test method and a parallel development, the FIP test, are described in the following sections.
