Varying temperature input

[strogy]

I am trying to create a scrip (using apdl), which would give results of element elongation change due to temperature change.

Assume simple truss, with multiple elements which have different load as a "body" temperature depending on height - 2to3 levels - 2to3 different temperatures.

The idea is: use input file (with temperatures over 2-3 year period - each day has different temp); assign it to script file and get output in form of table listing time increments (1,2,3 ...) and elongation (due to temperature change).

Hope you get the idea, if not ... i can explain it in details

Hope you guys now a script that could be used for this!

Thanks

 

[flash3780]

A truss? It sounds like a good application for a beam model. You could run a static model through many load steps with varying temperatures.

You might look into the BF command, which prescribes a temperature to the selected nodes. Also, the *DIM command gives you the option of creating a table that varies temperature with position. Something like:

*DIM,temptable,table,3,,,y
temptable(1,0)=  0,10,20  ! Y-Coordinate, Elevation (meters)
temptable(1,1)= 22,20,18  ! Structure Temperature (Celcuis)
bf,,temp,%temptable%

Clearly, you'll be creating a lot of load steps, so maybe referencing an external file with a *DO loop might make sense. Look into the *TREAD function to pull data into ANSYS from an outside file.

I'm not clear what you're using this model to figure out: Perhaps you could elaborate on your goals?

If you're dealing with variations in outdoor temperatures, you may also consider that the high-temperature and low temperature can vary significantly in a day. Good luck.

//signed//
Christopher K. Hubley
Mechanical Engineer
Sunpower Incorporated
Athens, Ohio
--

http://engineeringliberty.wordpress.com

 

[strogy]

Thanks for advices!

Yes, temperature is varying daily, taking this in consideration would make the task even more difficult. I've assumed different temperatures at different heights.

E.g. for a simple truss: bottom chord elements have +7; diagonals and studs +11; top chord +14.

So temperature is changing daily; but it is not represented as a function of something (and also yearly temperature readings are not exactly sinusoidal); so output/results are varying. And at one point - damaged is introduced in one of element. Looking at the results (using various algorithms) it is possible to detect it - the time when damage occurred.

So this is the plan.

During the weekend, I'll have a look at *DO and *TREAD functions. That is what I need - to have another external file (with temperature readings - approx. [(3x365)x3] matrix.

Thanks!