Pilot stage of servovalves and high response valves is mostly of open center type such as flapper nozzle or jet pipe valves. This paper introduces a pilot valve of closed center type and examines its dynamic performance. The valve mainly consists of a main spool, and a through the spool shaft as a pilot stage. The pilot orifices are formed by the pilot shaft angular input displacement, and closed by the main spool linear movement, thus the spool follows the pilot shaft input without need to extra feedback loops. Pilot stage of closed center type consumes very small amount of on demand pilot oil during the main spool movement only.
The simulation shows outstanding step and frequency response. This new valve enjoys high stability with strong immunity against disturbances, flow forces, and negative damping. The high pilot pressure and flow gains enable replacing three or more stages valves with valves of two stages only.

Die stetige Weiterentwicklung von Computer- und Simulationskapazitäten ermöglicht die Abbildung von immer komplexeren hydraulischen Systemen bei gleichzeitiger Erhöhung der Simulationsgüte. In vielen Anwendungsfällen zeigt sich, dass die erhöhten Kapazitäten nicht zur Reduzierung der Rechenzeit, sondern zur Erhöhung der Modellkomplexität und Vorhersagegenauigkeit verwendet werden. Der effiziente Einsatz der gegebenen Ressourcen bleibt somit in den meisten Fällen eine der wichtigsten Randbedingungen. Aus dem Bereich von physikalischen Experimenten sind verschiedene klassische statistische Methoden bekannt (z.B. statistische Versuchsplanung), die häufig auch im Bereich von Computerexperimenten eingesetzt werden. Diese Methoden können jedoch die speziellen Eigenschaften von Computersimulationen, wie zum Beispiel keine Messstreuung in Simulationen, nicht ausnutzen, wodurch die vorhandenen Ressourcen und simulationstechnischen Möglichkeiten bei Weitem nicht ausgeschöpft werden. Über die letzten Jahre hinweg wurden spezialisierte Methoden im Bereich Versuchsplanung, Meta-Modellierung, Analyse und Optimierung entwickelt, welche bei einem gleichzeitigen Einsatz maximale Informationen und beste Ergebnisse erzielen.
Die hier geplante Präsentation zeigt ein statistisch sinnvolles Gesamtkonzept von Versuchsplanung über Modellbildung und Analyse bis hin zur Systemoptimierung zum Einsatz bei Computerexperimenten.

This paper discusses the influence of the internal geometry on the behavior of spool directional valves. This study is based on both the flow modeling on control orifices and internal channels and test carried out with an overspeed sensor and a distributing valve applied to hydroelectric turbine control. On hydroelectric power plants, several hydraulic components of the speed governor are particularly designed for this application, operating under unusual pressure and flow rate values when compared to industrial and mobile fields. This fact demands low-scale manufacturing of the components where these must meet the requirements of standards such as IEC 60308 and ISO 10770 series. These standards establish steady-state and dynamic characteristics that must be achieved under specific operate conditions, characterizing the hydraulic component. Aiming to support the analysis and design of directional on/off and continuous control valves, a model based on the principles of fluid mechanics have being developed which allows the study of the influence of internal geometry on the behavior of both flow rates and pressures. Features like control orifice shapes, radial clearances, and internal channel sizes are considered in this analysis. The simulations are carried out by Matlab/Simulink where the model is applied to the analysis of a directional valve operating as a turbine overspeed sensor and a 2,500 L/min distributing valve. Therefore, it was possible to analyze the model robustness on different size of valves. The theoretical results are compared with the experimental data obtained in laboratory considering the specifications from international standards mentioned above. Some characteristics such hysteresis, dead zone, internal leakage, stabilization test time are evaluated under different internal geometry parameters that leads guidelines to be taken during performance tests. With the validated model, different directional valve configurations applicable to discrete control (on/off) or continuous control (such as proportional valves) can be evaluated. The knowledge about these important phenomena that influence valve characteristic curves such load pressure, internal leakage, and control flow rate depending on spool displacement and supply pressure is systematized. These results can help both professionals involved on operation and maintenance by the identification of possible valve failing causes and design engineers on the definition of dimensional tolerances.

In recent years, a trend towards speed-variable pump drives has become apparent. By using an axial piston pump with variable displacement, motor speed and volume flow can be decoupled providing a degree of freedom for controlling hydraulic processes.
This degree of freedom has been used in prior studies on a holistic analysis of static losses in electro-hydraulic drive systems to increase energy efficiency. By developing an operating-point-oriented motor speed feed forward control, energy savings of over 20% for static and quasi-static process cycles were proven in comparison to pure speed-variable or pure displacement-variable control strategies.
At the same time, however, a crucial shortcoming of the static loss model had to be observed in dynamic cycles. Fast acceleration sequences lead to increased copper losses in the electric drive, which can dominate the estimated energy savings when following the energy optimal operation points. As a result, the energy-optimized two-variable control led to increased energy consumption in dynamic injection molding processes, when it was compared to traditional process control with a variable-displacement-pump at constant speed.
In this work, the aforementioned problem will be countered with a model predictive control approach. For that purpose, a time-discrete dynamic loss model of all drive components was formulated. In contrast to the conventional operating-point-oriented optimization a new cycle-oriented optimization can be introduced, minimizing the energy consumption for any hydraulic cycle. In the model predictive approach a known, recurring control task is transformed into an equivalent mathematical optimization problem for the entire cycle. By means of formulated constraints, the required process dynamics, such as considering the physical limits of the drive components, can also be ensured.
In simulations with the new, validated dynamic loss models of all subcomponents energy savings of up to 30% were achieved in comparison to the known static approach to efficiency optimization. In particular, the functional verification of the model predictive concept for highly dynamic processes was proved, in which the static optimization became inefficient.