The apparent increase in the ultimate tensile strength at a pit is at variance with what would be expected from both a simple analysis and from the numerical model。 A possible explanation was put forward by Castel, Francois, and Airliguie。7 If a bar does not have a completely uniform cross section and material composition throughout its length, then it would clearly be expected to fracture at the point where the material is weakest。 If the position of the pit does not coincide with the location where the steel is weakest, an apparent increase in strength (where this is based on the minimum cross-sectional area) will be measured。

Corrosion initiates where micro-differences in environment or materials’ composition allow micro-electropotential differ-ences to develop。 Steel strength is also influenced by minor differences in alloy composition。 If the differences that result in one section being of stronger composition than the next also caused corrosion to develop preferentially at the same location,

Fig。 6—Decreasing rate of penetration at pit with increasing corrosion。

(a)

(b)

Fig。 7—Test bars from Series UB: (a) lightly corroded (Test 1) and heavily corroded (Test 2) test bars; and (b) variation in cross section and diameter along heavily corroded bar。

then it is apparent that bar strength would be reduced by a lesser amount than the bar section。 Given that pure iron generally corrodes less but is weaker than a mild steel alloy, one can spec-ulate that corrosion pits are more likely to develop where the metallic structure is composed of a stronger alloy。

Figure 8(c) shows the influence of corrosion on ductility。 Two sets of data are represented。 Open squares represent strain at fracture and are plotted against the left-hand axis。 Filled triangles represent elongation at fracture (elongation measured over the fracture on a gauge length of five times the bar diameter), and are plotted against the right-hand axis。 Both measures show a reduction in ductility with increasing corrosion。 The reduction in ductility is evidently more marked than the reduction in strength shown in Fig。 8(b)。

Comparison with results from numerical model: HW series tests

The numerical model was also applied to the test bars even though the variation in cross section along test bars was not measured in full detail。 To simplify input data, it was assumed that the mean loss of section occurred at all incremental lengths, except for one increment where the area loss corresponded to the area of the largest measured pit。 Analysis showed that results were not sensitive to the presence of smaller pits。 For a maximum section loss of 8%, corresponding to the highest section loss plotted in Fig。 8(b) and (c), no significant loss of yield strength was measured, and the average strain at maximum load decreased from 0。20 to 0。15 (Fig。 8(d))。 The numerical model results lie within the range of the experimental scatter shown in Fig。 8(b) and (c)。

Residual mechanical characteristics of corroded bars: UB series

Plots of stress (based on original cross section) against elongation measured for the two bars from the UB series shown in Fig。 7(a) are presented in Fig。 9(a)。 The bar in Test 1 had suffered only a 2% loss of section, while the Test 2 bar suffered a maximum section loss of around 70%。 The reduction in both strength and ductility is clearly evident。

Plots obtained from the numerical model are also presented in Fig。 9(a)。 In this case, the bar cross section was taken from the incrementally-measured variations throughout the gauge length。 The plot shows that the model provides a reasonable representation of behavior, with estimated reductions in both strength and ductility corresponding well to measured values。 In Fig。 9(b), the variations in strain along the length of the bar calculated by the model are also shown, and can be related to the variation in the cross section presented in Fig。 7(b)。 The very high strains calculated at the most severely attacked sections are evident, with fracture deemed to occur when the local peak reaches a limiting value。 Overall elongation, however, is related to the integral of strain along the bar, that is, to the area under the plot。 It can be seen that the contribution of the highly localized peak strain does not provide a correspondingly large contribution to overall elongation。 锈蚀钢筋的力学性能英文文献和中文翻译(6):http://www.youerw.com/fanyi/lunwen_135205.html