Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Creeplife-Larson Miller Parameter PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

10.

Creep Life Prediction: Creep tests take a long time to perform making the
generation of design data expensive and the lead time between developing a
new alloy and its exploitation excessive. The fact that there is an Arrhenius
relation between creep rate and temperature has led to a number of time-
temperature parameters to be developed which enable extrapolation and
prediction of creep rates or creep rupture times to longer times than have been
measured. They also enable rating comparisons to be made between different
materials. It is important that no structural changes occur in the region of
extrapolation, but since these would occur at shorter times for higher
temperatures it is safer to predict below the temperature for which data is known
than above. One parameter used is the Larson-Miller Parameter.

æQö
This is derived by taking natural logs of the Arrhenius equation: ε = Aexp− ç ÷
è kT ø
(Note that k is being used here instead of R so that Q is quoted in joules per
atom. Also, if logs to the base 10 are used, the Larson-Miller Constant values
given below are multiplied by log10e = 0.43429)
• Q • Q
ln ε = ln A − , ∴ ln A − ln ε = . If we assume that the creep strain to rupture
kT kT
ε r is a constant over the temperature range of interest, and the major part of the
creep strain is steady state creep, then the average creep rate over the life to
• ε
rupture, t r , of the specimen is: ε = r .
tr
æε ö Q
∴ ln A − lnçç r ÷÷ = ln A − ln ε r + ln t r =
è tr ø kT

∴ T (C1 + ln t r ) = = P ...(11), where C1 = ln A − ln t r is the Larson-miller


Q
k
Constant and P is the Larson-Miller Parameter for a particular stress, since
Q = (Q0 − vσ ) . Plotting experimental data of
1
ln t r versus gives a straight line with slope
T
1
P and intercept at = 0 of - C1 . The values
T
of C1 range from 35 to 60, but is typically 46,
and Figure 26 shows that it is independent of
stress.

Figure 26: Stress rupture data plotted as


ln(rupture time) versus reciprocal of absolute
temperature.
Figure 27: Shows a plot of stress
versus Larson-Miller
Parameter for the alloy
Astroloy.

It is clear that for a given stress,


the Larson-miller Parameter for
the material can be obtained and the time to rupture predicted for any
temperature. Alternatively, given a lifetime required at a specified temperature,
the L-M Parameter can be calculated and the maximum allowed stress looked up
on a graph like Figure 26. E.g., For a 100,000 hour life at 870oC (1143 K),
P = 1143(ln105 + 46) = 65.7.103. This corresponds to a stress of 85 MNm-2.

In practice, alloys used in applications like turbine blades experience a wide


range of stresses and temperatures for varying times during start-up, take-off,
steady flight, landing and shut-down. Creep strains accumulate during these
stages and design could assume a 'worst case' scenario, but this would lead to a
weight penalty in the aero-engine and less payload. Alternatively, an approach
similar to Miner's Law in fatigue can be adopted, and this is called the 'Life
fraction rule'. This states that rupture occurs when the sum of all the fractions of
the rupture life at different stress/temperature combinations becomes equal to
unity. I.e.
æt t t ö t 1
å çç t 1 + t 2 + t 3 .....÷÷ = å i t i = ò t dt =1 ....(12), where ti is the time spent at a
è R1 R2 R3 ø Ri R

particular stress/ temperature combination where the time to rupture is t Ri . The


data in Figure 9 can be used to carry out such a calculation.

You might also like