To what extent is the safety factor necessary?

Since all applicable codes and standards are using inherently “Safety Factor” in the design rules and formulae, if we need to use further safety factor to design components? For instance in ASME Section VIII, Div.2, whether a new safety factor (In addition to S.F= 2.4) should be added to result of finite element analysis?

Ans.:

It depends on many parameters, but to be frank I would have to say that “Yes, we need a new safety factor”.

Level of individual’s ignorance in the following areas may make error in results. These items would determine whether you need to impose further safety factor in designing or not.

1. Wrong estimation of loads or inadequate loading data
2. Simplification in boundary conditions, restraints, and joints
3. False impression of boundary conditions
4. Simplification in material properties or wrong material properties
5. Selection of wrong analysis or lack of appropriate theories
6. Ignorance of environmental effects
7. Misunderstanding of assembly methods
8. Wrong lifecycle expectations
9. Change of service condition by operator or wrong customer usage profile
10. Improper interaction of parts

All the uncertainties make a discomfort zone for designers. All the codes and standards assume the design data are completely correct, and the safety factors are not to be used because of individual’s ignorance or because you might make mistake in some assumptions. To me, with all the experiences and evidence found, the following factors are recommended beyond the safety factor of applicable codes and standard:

• For main loads such as pressure and weight S.F.= 1.6 – 2
• For occasional loads such as seismic and wind loads S.F.= 1.3 – 1.4
• For thermal loads and secondary stresses: S.F.= 1.3 – 1.4