I just had this confusion after reading Luf's remarks . I always felt that duration does matter.When a pipe is heated to higher temperature intially we have elastic strains coming into play and latter plastic strains .Ultimately bcos of temperature we have self limiting stresses induced.This condition is represented by expansion stress range term we come across in the analysis.For selflimiting stresses and consequently its associated terminal reactions in the final operating state it takes some time to attain .The final condition of the piping configuration is an equilibrium.Is time not an important factor to achive equilibrium.If yes then there is a transient state before the piping system achives equilibrium.Coming back to the turbine problem ...the heating during startup must be taking some time .This interveaning time must be enough for the piping to achieve the equilibrium i,e the final state.Now coming to friction ..it comes to play during the time the system is tending towards equilibrium and its role ends after the system reaches equilibrium.So in the final equilibrium state the forces at the terminal points are due to the stiffness of the whole piping system.Does that mean that during the transient state the terminal loads are a combination of that due to stiffness and friction , both which are varying in magnitude as the piping system tends towards equilibrium.The friction load is decreasing and the stiffness loads increasing plus at the same time selflimiting itself.Going by these thoughts of mine the loads experienced by turbine must be less during startup than during final operating state.