Nonlinear GCMs

Nonlinear GCMs For certain properties, model accuracy can be improved by including a quadratic term  

---

Using the nonlinear model substituting 0.915 p.gcm (the DFR value from the model 1 day after application) into Equation (6) yields a dose of 1.05 mgkg day" and an MOL from Equation (5) of 47, below the required value of 100 for margin of exposure. Performing the same calculation on the day 2 data gives a dose of 0.344 mg kg day and an MOE of 145, which is above the level of 100 required to establish a safe re-entry level. Therefore, a 2-day period is adequate to ensure worker safety. The observed values of 0.936 and 0.234 qg cm for days 1 and 2 match closely the values predicted by the model. 

GCMs can also be nonlinear for instance, Joback recommended the following for the critical parameters Tc, Pc and Vc ... 

GCM functions may be linear or nonlinear. Often, GCM schemes include extra terms for diverse structural factors. The differences among GCMs with respect to the form of their group-property function (eq. 1.6.1) are considered in the following. 

The method of Lydersen [28] is a GCM of this type to estimate the critical temperature, Tc. Other approaches to non-linear GCMs include the model of Lai et al. [29] for the boiling point, Tby and the ABC approach [30] to estimate a variety of thermodynamic properties. Further, artificial neural networks have been used to construct nonlinear models for the estimation of the normal boiling point of haloalkanes [31] and the boiling point, critical point, and acentric factor of diverse fluids [32]. 

---