Residual stresses using thermal expansion

[Settler]

Hello all,

I would like to simulate the procedure of the dipping of a medical tool into liquid nitrogen and measure the amplitude of the residual stresses after that.

Does anybody know any location to find tutorials on how to define two different temperature half-spaces (liquid nitrogen and air) and how to create a "moving" object between those two half-spaces.

I want to see also the effect of the dipping speed, that's why I need the two half-spaces.

Thank you very much!

I hope to reading from you soon!

 

[Settler]

Anyone please?

 

[mvp23]

create a box or cylinder that contains your liq. i would use a box as you dont have to deal with cylindrical coordinates. i will leave it to you as how to how design this box based on your problem. create a horizontal partition to separate the liquid nitrogen region from the other region. you can impose initial temperatures on thse regions if required. if you have to have a moving object coming into the liq nitrogen, you will have to use Abaqus Explicit to perform a dynamic analysis. in doing so you will be able to specify a velocity to the moving object. basically you will have to perform a coupled thermal analysis if you would like to see the stresses induced by the thermal problem.
to assess the effect of dipping speed, just perform a parametric study by varying the velocity of the moving object. one way to do it is to change the velocity and run the job every single time. the automated process will include writing a python script for parametric studies.

 

[sticksjoerd]

You can also use the user subbroutine film.f. That way you can program a virtually moving sink (your liquid N2) to gradually move up the tool, as a function of time. You can even change the heat-transfer coefficient as a function of for example temperature, if you want to.

Unfortuately Abaqus requires a complete Intel Fortran set-up for compiling user subroutines at runtime. If this poses a problem, you can take a look at CalculiX, a free, open-source non-linear FE solver, which uses the same keywords as Abaqus (a sub-set) and also supports film.f. It compiles with gfortran, so it's easy to do on any Linux system.

 

[Settler]

Thank you very much mvp23 and stickjoerd for the answers. I would like to follow up with some questions as I'm pretty new to heat/transfer analysis and abaqus in general.

1. How can I create a box with a liquid container in it. Also I don't want the temperature of the liquid to change over time.

2. I don't know where or how to find film.f subroutine.

3. Do you know anybody that has done it before so that I can find some kind of tutorial online?


Thank you again.

 

[mvp23]

settler - i haven't had the need to model a physical fluid cavity at any point. but i looked around online and one post had suggested that if it is required to have a fluid then you would probably have to run a co-simulation of abaqus explicit and CFD. i had suggested a box ( ie a closed region) , but didn't know you were interested in modeling the fluid. also, you would have to define contact between the tool and the fluid.

 

[Settler]

I was thinking of not even modelling the fluid but have a region of constant temperature.

What do you think?