Learning Forum

[ericpflor]

Hi,

I'm trying to simulate creep strain on a concrete column along with construction sequency. The part of construction sequency is OK by using birth and death controls. My problem now is to model creep behaviour once the creep equations on ANSYS do not consider the age of concrete when load is applied (I call it "t0"), but only the time after a load is applied ("t", in other words, time when creep strain is analysed).

My creep equation is a function of f(t0,t). Example: If you apply a load when concrete is 7 days old, the creep function will act different from a load applied when concrete is 14 days old.

Ansys creep equations (there are 15 types available on the database, the one I'm using is 'time hardening', function of (stress,t,T) - I'm not using temperature) are not a function of t0.

The image attached explains my situation:

 

Day 0

concrete is poured to column C1

 

Day 7

Load P1 is applied to C1 leading to elastic strain on C1.

 

Day 14

Load P2 is applied leading to elastic strain on C1 and C2.

C1 is 14 days old. Creep strain on C1 is equivalent to f(P1,t0=7,t=14)

 

Day 21

Load P3 is applied leading to elastic strain on C1, C2 and C3.

C1 is 21 days old. Creep strain on C1 is equivalent to f(P1,t0=7,t=21) plus f(P2,t0=14,t=21)

C2 is 14 days old. Creep strain on C2 is equivalent to f(P2,t0=7,t=14)

 

My point is: Is there a way to model these different creep functions according to different "t0"? Or: Could ANSYS "read" the element age according to the time of its birth?

Thanks in advance!

[Wenlong]

Hi,

Thanks for sharing your problem. Very interesting yet complicated.

My opinion is you can add temperature-dependent creep parameters, and use temperature only as an indicator. It means ignoring all the thermal expansion and other temp-dependent material properties. 

By applying a temperature boundary condition, you can control the creep coefficients by temperature, in other words, by time, because temperature changes with time in your input amplitude. 

Hope this helps.

Bests,

Wenlong

[ericpflor]

Thanks, Wenlong!

I'll try that immediately.