TWI Frequently asked questions
by Christoph Wiesner
There are two published methods for doing this. The first one was published in the UK pressure vessel code PD 5500. This allows the estimation of Charpy energy levels in the range of 20J to 47J using 1.5J/°C.
This criterion to convert Charpy requirements for energies in the range 20-47J has existed in some form since the adoption of a low temperature appendix to the BSI pressure vessel code. It existed in its current form in BS 1515 Appendix C (1972). In earlier versions 1.35J/°C and 1ft lb/°C were adopted. The criterion is not referenced but appears to be based on the transitional behaviour of typical ferritic steels such as BS 1501-161 made at the time of the early standards and also reflects ranges of Charpy properties quoted in the relevant steel specifications.
Modern steels typically have lower transition temperatures and steeper transition curves than material manufactured in the 1960s and 1970s. This can mean that inappropriate estimates of lower energy requirements could be made using the criterion based on a given higher Charpy energy (e.g. a 27J requirement based on a 47J measurement). Safe estimates will, however, always result when converting up from lower energies.
The relationship was also originally included in the atmospheric storage tank code BS 2654 when prepared by the then relevant committee CPE/12 in 1965, on the basis of a suggestion from BP (Harry Cotton) considering 1960s steel properties. The slope of the ft lb vs °C Charpy transition curve was about 43 deg, corresponding to 1 ft lb/°C in the range between 30 and 40 ft lb.
The second method is based on a Charpy reference curve, as shown in Fig.1. The temperature axis is indexed to the temperature, T 27J, at which 27J is obtained. This allows this temperature T 27J to be determined from Charpy results obtained at other temperatures (within the range shown). This relationship was proposed by Burdekin (1981) and has been used for deriving Charpy requirements in BS 5400 Pt 3.
The curve has been derived for older type structural and pressure vessel steels and its use for weldments and modern steel exhibiting steep Charpy transition curves has not been validated extensively. The first method with its limited applicability should be used under these circumstances.
Fig.1 Form of Charpy transition curve assumed for interpolation as given at present in BS 5400: Part 3 |
Examples for application of method 2 are given below:
Recent work by TWI (Pisarski et al 2001) has shown that it is often non-conservative to use the Charpy interpolation technique as described above. It is advised that expert advice is sought before safety-critical applications of the above methods are made.
| Measured Charpy energy, J | T-T 27J, °C from Fig.1 | Test temperature, T, for measured Charpy °C | T 27J, °C |
| 41 | +10 | -20 | -30 |
| 81 | +30 | -60 | -90 |
| 10 | -20 | +40 | +60 |
References
Burdekin F M: 'Materials aspects of BS 5400 Part 6'. In: 'The design of steel bridges.' Proceedings of the International Conference, University College, Cardiff. March 1980. 55-72. (K C Rockey and H R Evans Eds), London, Granada Publishing, 1981.
PD5500:2000: 'Specification for unfired fusion welded pressure vessels'. London, British Standards Institution, 2000.
H G Pisarski, B Hayes, J Olbricht, P Lichter and C S Wiesner: 'Validation of idealised Charpy impact energy transition curve shape', Proceedings, Charpy Centenary Conference (CCC 2001), Poitiers, France, 2-5 Oct.2001. F-92024 Nanterre Cedex, France; Societe Francaise de Metallurgie et de Materiaux (SF2M); 2001. Vol.1.
