gsphelp  States and errors

<< Click to Display Table of Contents >>

Navigation:  GSP models >

gsphelp  States and errors

Previous pageReturn to chapter overviewNext page

For solving the GSP non-linear differential equations (NDEs) GSP defines the operating point by states (or 'free states') in a state vector. Using the appropriate aero-thermodynamics equations, maps and other relations, all engine parameters can be directly derived from the states. As such, the states represent the unknowns in the NDE set to solve for. The NDEs are depending on the state vector and each NDEs has a error variable representing the deviation from a valid solution. The GSP solver iterates towards the solution where all errors (i.e. the error vector) are zero (within the user specified tolerance). Note that the NDEs can not be considered equations that can be represented by a series of mathematical expressions (functions of state variables) but rather represents the outcome of several algorithms using thermodynamics, table/map look-ups, internal iterations etc. This is what makes the non-linearity so significant.

For a simple turbojet model (tjet.mdl) for example, there are 4 states and 4 errors. For more complex models such as turbofan models with several schedulers, the number of states and equations may easily rise up to 20 or more.

 

Although most states and errors are set up automatically by GSP, the user can have control over states and errors using component model options. The current maximum amount of model state variables is set to 50. Please contact NLR when your model exceeds this maximum to discuss options.