Hi,
  there are different functions like fcube, ftanh, tanh etc that can be used to model an amplifier. Can somebody tell me if there is any specific one that is preferred over the other functions to model an amp or can i use any one of them and just change the gain to suit the requirements?

thanks.

If you just want a generic limiting amplifier function you could use

L*tanh(g*x/L)

L is the maximum output level
g is the gain

If you want to adjust the sharpness of the limiter (sharper limiting is more linear) you could use the following adjustable model

g*x/[1+(g*abs(x)/L)^s)]^(1/s)

s is the sharpness factor

You could also use analytical functions for real amplifiers such as bipolar and cmos differential pairs (found in most circuits text books).

If you need an accurate behavioral model then you will need to spend time figuring out which behaviors you need and cobble together good models for those characteristics.