I am not familiar with that particular modelling issue, but I guess you should investigate what precision is experimentally feasible, with regard to the expected/desired precision for your model, according to the final purpose of the modelling effort.  It happens sometimes that modellers are working with a degree of precision that is not experimentally achievable, or has not yet been achieved experimentally.  "Exact" can only have a meaning with regard to a defined level of accuracy for a determined application scope, in my opinion.  

Hi Mr. Didier, thank you for your answer.

These parameters are generally material dependent, u have to check the type of material that u r using, for example, for Reinf Cement concrete , under compression, stiffness recovery is about (10-20%) as compared to under tension, stiffness recovery is taken as 70-80%. The difference is basically due to two different mechanism under two loading condition RCC fails by crushing in compression, whereas, RCC fails by Crack formation under tension. Stiffness Recovery parameter basically takes into account the micro effect of crack opening and Closing. therefore under compression stiffness recovery is at lower amount  (Crushing), whereas under tension, this phenomenon is higher (70-80%). You can also use EXPERIMENTAL RESULTS TO CHECK the EFFECT OF THE values, then U can CHoose the BEst VALUE for FURTHER EVALUATION.  BEST of LUCK

Thank you Sanjeev. 

So, it is based on the assumption. When the experimental investigation has been carried out, the numerical analysis follows the trend of experimental result. Hence, the value of both compression and tension recovery are based on the assumption at certain range?

Regards

It can be understood this way.

Abaqus Analysis User's Guide/ 23.6.3 Concrete damaged plasticity /

Figure 23.6.3–7 Uniaxial load cycle (tension-compression-tension) assuming default values for the stiffness recovery factors:wc=1 and wt=0