A new original algorithm, CONDOR ("COnstrained, Non-linear, Direct, parallel Optimisation using trust Region method for high-computing load function"), was developed. The goal of this algorithm is to find the minimum of an objective function in the least number of function evaluations. It is assumed that the dominant computing cost of the optimisation process is the time needed to evaluate the objective function (one evaluation can range from 2 minutes to 2 days). The algorithm will try to minimize the number of evaluations, at the cost of a huge amount of routine work. The derivatives of the objective function need not to be known. The only information needed about the objective function is a simple method (written in Fortran, C++) or a program (a Unix, Windows, Solaris executable) which can evaluate the objective function at a given point. The algorithm has been specially developed to be very robust against noise inside the evaluation objective function. This situation is very general, the algorithm can thus be applied on a vast number of situations. CONDOR outperforms many commercial, high-end optimiser and is maybe the fastest optimiser in its category (fastest in terms of number of function evaluations). When several CPU's are used, the performances of CONDOR are unmatched. The experimental results open wide possibilities in the field of noisy and high-computing-load objective functions optimisation like, for instance, industrial shape optimisation based on CFD (computation fluid dynamic) codes (The Method project uses this kind of objective function) or PDE (partial differential equations) solvers.
Radial and mixed compressors and turbines are frequently used in refrigeration cycles, refining systems, aircraft auxiliary systems, turbo-shaft engines, small gas turbines, turbochargers and other systems where light-weight compact compression or decompression is required. Prandtl demonstrated that placing a thin wire around a sphere just in front of the circumference can artificially make the laminar boundary layer turbulent and can move it downstream of the separation region. This method for delaying the separation Comoti considers that it can be used also in the blades passages of the impeller. Hopping to delay the separation of the boundary layer in the impeller’s blade passages and so, to increase the rotor efficiency Comoti used this method as a point of reference for its calculations. Comoti put a rib on the shroud, in front of the place where the boundary layer displacement occurs. To calculate the 3D flow through the passage between two blades Comoti made a program in which it introduced at the beginning the geometrical data of impeller: the geometry of impeller, the gasodynamics data at the inlet and also the rotational speed. A quasi 3D viscous code to solve the simplified Navier–Stokes equations was developed. The viscous effect is simulated by viscous body force method. This code is demonstrated to predict the internal flow field of a close impeller. The Lax–Wendroff scheme is used to obtain the discretized equations. For higher efficiency, the discretized equations are solved using accelerating techniques, such as local - time stepping and multi-grid methods. For convenience of simulating the quasi 3D flows in rotating turbo machinery blade passages, a relative cylindrical coordinate system (z, fi, r) is adopted.
Comoti made 3D models in UNIGRAFICS NX (Version 19), detailed stress analysis and manufacturing drawing for 6 centrifugal impellers with high efficiency and complex geometry. The soft for machining was performed and tested with Cam Unigrafucs. Comoti made the technology for the manufacturing of 6 close impellers. All impellers were designed using an aircraft technology: covers disks were welded on impellers in VIG technology. Comoti chose and bought the rough parts for impellers manufacturing. The impellers were milled on modernised Forest Line CNC 5 axes milling machine. The impellers were balanced, dimensional checked and sent to Nuovo Pignone for testing on the special centrifugal compressors test rig.
An optimised geometry for a centrifugal compressor impeller has been produced as a result of the optimisation algorithm developed inside the METHOD project. The optimisation process is based on an optimisation algorithm which drives a CFD computational code. This code produces information to select the best and the shortest path to the optimised geometry.