i have a simple GRP pipe anchored one side
lenght 1000mm, no fluid, no pressure

coef. of thermal expansion (x1,000,000)(mm/mm/°C)=18
(inserted in special execution parameters window)

axial_dilatation=alpha*L*delta_T

where alpha= 18E-6mm/mm/°C

ambient temperature -27°C

we have 3 temperatures to test:
T1=30°C
T2=50°C
T3=70°C

i thought the alpha would always "18":

axial_dil(T1)=alpha*1000*30=0.000018*1000*30=0.54mm
axial_dil(T2)=alpha*1000*50=0.000018*1000*50=0.9mm
axial_dil(T3)=alpha*1000*70=0.000018*1000*70=1.26mm

(or instead of temp 30,50,70 have i to use 30+27, 50+27, 70+27?)


BUT if i run the simulation i obtain an axial dilatation

axial_dil(T1)=0.872mm
axial_dil(T2)=1.178mm
axial_dil(T3)=1.484mm

infact in "special execution paramenter" if we flag the "print alphas and pipe properties" we obtain, starting the run:

alpha1=0.000872
alpha2=0.001178
alpha3=0.001484


Question1:
if thermal dilatation is "18" (really 18E-6mm/mm/°C), why the thermal coef. seem to change at various temperatures, obtaining different lenght expansion at different temperature?

Question2:the alpha1, alpha2, alpha3 are the coef. of thermal expansion at T1,T2,T3?
It seem no because if i calculate by hand:
axial_dil(T1)=0.000872*1000*30=26.16 >> 0.872 : ????

question3:
How does caesar calculate these alphas starting from "18" given in input?


The problem i want to calculate axial expansion by hand to confirm the caesar calculus. and i need to know the axial thermal expansion coeff. at varous temperatures.

If you manually specified coefficient of thermal expansion, it's specified as a percent (i.e. mm/mm and incorporates temperature) and remains constant to what you specify. CAESAR calculates it in the same manner.

However, for many materials, coefficient of thermal expansion as percent per unit temperature is not constant along the entire range of temperature.

There is a symbol on the input window that looks like this ">>". Clicking this will open up a dialogue that gives you additional information from the piping input window. Change your temperatures, record the values in Excel, and back it out into units of % / unit temperature.

I ran a model similar to yours and I can see the proper thermal strain. It appears to me that you must have some additional axial restraint on your cantilever that prevents free axial growth.
A few points on entering thermal strain values...
The Special Execution Parameters collects the "FRP Coef. of Thermal Expansion" in (micro-length/length/degrees C). When I see the word "coefficient" used with thermal strain, I think (length/length/degree C). If the word coefficient is not used, I see it as pure strain (length/length).
The enhanced Temperature/Pressure input window (opened, as Michael states, by clicking on the chevron, ">>") shows "Thermal Expansion" next to the Temperature field. No coefficient here, this is the strain associated with that temperature and based on the growth from the specified ambient temperature. FRP pipe elements show 0.0000 for these thermal strains.
Finally, for User Defined Materials, where there is no database providing strains for temperatures, you can specify this strain in place of the temperature since very small values here (default > 0.05), are taken as strain rather than temperature.

thank you

(i forgot to tell i use caesar 2016).

If in Special Execution Parameters you flag "print alphas and pipe properties" and you run the error checker, you will obtain an "alpha" value for every temperature.

Doing it for a METAL system, the obtained alphas are the same of the enhanced Temperature/Pressure input window.

BUT

doing it for a GRP system Inside The enhanced you obtain the "alphas" but meanwhile inside Temperature/Pressure input window "FRP pipe elements show 0.0000 for these thermal strains.".

So the question still is:
-you have a GRP system.
-you insert a manual value in special execution parameter (eg. FRP coef ot thermal Expansion mm/mm/°C = 18)
-the "alphas" obtained by error checker (after having flagged the "print alphas" in special execution paramenters windows"): where do they come from (if the temp/pressure input windows contain all ZERO for these thermal strains)?

please see image at
http://it.tinypic.com/view.php?pic=efs2sx&s=9


more:
If i want to know the thermal expansion (mm/mm) at every temp for a GRP pipe, have i to take the "alphas" obtained from the error checker?