SERVOVALVE CALIBRATION
SYSTEM PROJECT

Team ME 5.3

Home

Members

Background

Research

Software Development

Hardware Interface Development

Picture & Endnote

Project Sponsor

 

Liaison Engineer
Norm Englund, PE

Team Members
John Cotter
Harun
Robert Phillips
John Ulmen

Faculty Advisor
Dr. Frank J. Shih

Sponsored by
Seattle University &
The Boeing Company

 

Servovalve Operation

A servovalve is a flow control device that mechanically controls hydraulic flow over a large range of flow rates and fluid pressures. A servovalves is controlled by a small electrical input current signal which causes the torque motor to rotate a rigid flapper toward one of the nozzles. The resulting pressure differential at the two ends of the spool moves the spool to open or close ports. There is a positional feedback device called “feedback flex” that returns the flapper to the center position and balances the forces on the spool. The system reaches steady-state operation when the port openings are adjusted such that the fluid flow through the servovalve matches the flow rate specification by the electrical input current signal. To summarize, a servovalve compares the hydraulic flow rate to the electrical input current signal, and automatically adjusts the servovalve port openings until the desired flow rate is achieved. Servovalves respond very quickly to input current signal, and can have a useable frequency range of up to 150 Hz.

Servovalve Tests

The servovalve tests are intended to check for defect and evaluate servovalve performance. The types of servovalve test are briefly described. The four general categories of tests include: (1) Flow Gain: Flow rate versus input current, (2) Pressure Gain: Load pressure versus input current, expressed in psi/mA, (3) Null Leakage: Valve internal leakage flow rate versus input current and (4) Dynamic Response: Sinusoidal input current of varying frequency versus the displacement of a low-mass low-friction actuator. The solenoid valves, shown in Figure 7, can either be closed or open to change the testing condition. The servovalve tests that were specified for the project are the following.

Blocked Port Pressure
For this test, the solenoid valves to control ports C1 and C2 are closed. After opening the solenoid valve in the pressure return line R, the solenoid valve in the pressure supply line P is opened. The port abbreviations C1, C2, R, and P refer to the schematic from Figure 7. Input current control signal is then swept between plus and minus rated current values, stopping at zero input current. For the Pressure Gain plot, load pressure versus the input current is measured, recorded, and plotted. Typical Pressure Gain plot produced from the blocked port pressure test is shown in Figure 10.

Figure 10. Typical servovalve Pressure Gain plot

The blocked port pressure test simulates an actuator applying force on a very stiff specimen or performing a high pressure test on a hydraulic tube or fitting. Since there is no external device attached to the servovalve, any anomalous behavior in the hysteresis plot curve is an indication that there is something wrong with the servovalve. The blocked port pressure test provides the clearest indication of a servovalve malfunction, while the other tests evaluate its performance.

Blocked Port Pressure at Null
At null condition, ports C1 and C2 are blocked but a small flow is still possible from the pressure port P to the return port R through the flapper nozzles. The blocked port pressure at null test can measure the degree of wear in the valve. An ideal valve has half the supply pressure on the output ports at null. In practice this pressure is usually a little higher than the half pressure point.

Open Port Flow
For this test, the solenoid valves to control ports C1 and C2 are fully open. Input current control signal is then swept between plus and minus rated current values, and the flow meters FM1 and FM2 measures the flow rate in the hydraulic lines. The flow rate versus the input current plot can be used to evaluate servovalve performance such as Flow Gain, linearity, symmetry, hysteresis, and saturation. The open port flow test simulates an actuator applying a force on a very flexible and compliant specimen or position control with minimal load and friction. Again, since there is no external device physically connected to the servovalve in the test, any anomalous behavior is an indication of servovalve malfunction.

Null Leakage
At null condition, there is a small flow from the pressure port P to the return port R through the flapper nozzles. The internal leakage flow can be measured using the flow meter FM3 (shown in Figure 7). The null leakage test simulates a double-ended actuator holding zero force. This test can indicate the degree of wear in the servovalve.

Frequency Response
The frequency response test examines the dynamics characteristics of the servovalve. The servovalve is a dynamic device that is frequently used to the full extent of its bandwidth. Used in conjunction with a feedback signal (such as position or force), the servovalve can be used to create a closed-loop servo, controlling the position or force of an actuator. For servovalves to be used effectively in closed-loop mechanical systems, their frequency response must be thoroughly characterized throughout their usable bandwidth. A servovalve that is stable with a 3,000 psi supply pressure may not be stable when the pressure is raised to 5,000 psi. The frequency response test is run with no load (all control ports solenoid valves opened) and a sinusoidal current input signal is swept through the servovalve with varying frequencies. A low-mass, low-friction actuator is used in FRA (shown in Figure 7) attached to a LVDT to measures the displacement of the actuator. The resulting plots are the Bode magnitude and phase shift plots. These two graphs plot output/input amplitude versus input frequency and output/input phase shift versus input frequency.

It is often not necessary to perform every one of these servovalve tests. If the performance of a servovalve deviates from expected behavior in one of these tests, the servovalve will likely require service from the manufacturer.

Hydraulic Power Supply Operation

The United Filco hydraulic power supply is rated at 3,000 psi and 35 gpm. The fluid velocity coming out of the hydraulic power supply is about 25 fps. ME 5.3 located the necessary hoses and fittings to connect the hydraulic power supply to the test bench. Hoses with a 3,000 psi working pressure and 0.75 inch inside diameter were used to connect the hydraulic power supply to the test bench. The component testing group primarily uses a petroleum-based hydraulic fluid. Many components in commercial aircrafts use Skydrol phosphate-ester based hydraulic fluid, which meets FAA specifications for function and safety. Skydrol hydraulic fluid poses a greater health hazard and is substantially more expensive than petroleum-based fluids. ME 5.3 used a petroleum-based Mobil DTE 25 hydraulic fluid for the servovalve calibration project.

Sensors and Control Hardware

The test bench is equipped with four types of measuring devices. There are three Cox internal turbine flow meters (FM1~3), two Sensotec A-5 series pressure transducers (PT1~2), a linearly variable differential transducer (LVDT) for frequency response test, and a Ohio Semitronics current transducer that measures the control signal to double check the input current signal.

Flow meters FM1 and FM2 measures the flow rate in the + Flow and the – Flow line. The internal turbine flow meter (Cox Model ANC 16), shown in Figure 11, outputs a digital pulse train at 10 VDC. The frequency of that pulse train signal corresponds to the amount of fluid flow through the body of the flow meter. It is shown that the flow rate ranging from 0 to 60 gpm corresponds to pulse train signal frequency of 0 to 1400 Hz. A separate and different flow meter FM3 (Cox Model ANC 8-4) is used to measure the leakage flow rate.


Figure 11. Cox internal turbine flow meter


Figure 12. Sensotec pressure transducer

Pressure transducers PT1 and PT2, shown in Figure 12, are linear analog voltage devices with output voltage proportional to pressure. The Sensotec A-5 Gage series pressure transducer, shown in Figure 10, is rated at 5000 psi and outputs 0-5 VDC. LVDT and the current transducer are also both linear analog voltage devices with output voltage proportional to displacement and current, respectively. The current transducer will be discussed in more complete detail in sections on hardware interface development.

For controlling the servovalve tests, five normally closed Parker solenoid valves (SOL1~5), shown in Figure 13, manage the flow of hydraulic fluid through the test bench. By selectively open and close solenoid valves, the user can configure the test bench for any of the required servovalve tests. For subsequent data acquisition, ME 5.3 obtained pin-out diagrams for the pressure transducers and the flow meters.


Figure 13. Parker solenoid valves

National Instruments DAQ Card

Boeing required ME 5.3 to use a National Instruments (NI) DAQ card for the servovalve calibration project. ME 5.3 narrowed its search to two models, an older E-Series and a new M-Series model DAQ card. After researching the advances in NI DAQ cards and consulting with our liaison engineer, a NI-PCI 6221 M-Series DAQ card was chosen for the project. The M-Series cards support all of the functions needed for the project and cost less than the E-Series cards. Since the M-Series cards are relatively new, ME 5.3 reviewed the technical documentation extensively. The M-Series DAQ card supports C++, LabView, Visual Basic, and Microsoft .Net programming languages. The specifications of the NI-PCI 6221 are tabulated.

Two 16-bit analog outputs (833 kS/s)
NI-MCal calibration technology
NIST-traceable calibration certificate
NI-DAQmx Measurement Services to simplify configuration &measurements
Correlated DIO (8 clocked lines, 1 MHz)
24 digital I/O; 32-bit counters; digital triggering

ME 5.3 also selected NI BNC 2090 DAQ breakout panel, shown in Figure 12, since it has the same pin capabilities and supports all the functions of the NI-PCI 6221 M-Series DAQ card. A breakout panel is easier to use than hooking up wire leads directly to the DAQ card. The BNC-2090 is a shielded, rack-mount adapter with signal-labeled BNC connectors, spring terminal blocks, and component locations for passive signal conditioning. The BNC-2090 consists of 22 BNC connectors (18 analog, 2 digital, and 2 user-defined), a 68-pin male connector, and 28-spring terminals to provide easy connection. The panel also has signal conditioning capabilities (simple passive, low pass or high pass filter, voltage attenuator circuits).


Figure 12. BNC 2090 DAQ breakout panel

Software

In the Request for Proposal, Boeing requested a Windows-based software for the control and data acquisition portion of the Servovalve Calibration Project. C++ was selected as the programming language since it is fully supported by the M-series DAQ card and is a common programming language. During the research phase of software development, National Instruments provided extensive documentation and example programs to demonstrate the capabilities and codes recognized by the DAQ card. In addition, other C++ programming books were referenced in order to gain sufficient knowledge and skills to explore the full potential of the DAQ card.

Install and Test DAQ and Breakout Panel

The installation of the NI DAQ card was straightforward, due primarily to NI’s extensive support for their DAQ card and breakout panel. NI measurement devices and application software NI-DAQ 7.x was installed on the computer. The software includes two NI-DAQ drivers: NI-DAQmx and traditional NI-DAQ. These drivers have an Application Programming Interface (API) which includes a library of functions, classes, and properties for creating applications. NI-DAQmx is used with Windows 2000/NT/XP, LabVIEW, Measurement Studio 7.x, or Measurement Studio .Net.

Since a computer was used for the project, ME 5.3 selected a Peripheral Component Interconnect (PCI) card. PCI is a local bus standard developed by Intel Corporation which connects directly to the microprocessor. This is not to be confused with PCI-X (PCI extended) which is an enhanced version of PCI and was designed jointly by IBM, HP and Compaq to increase performance of high bandwidth devices. The NI-PCI-6221 card is installed into a PCI slot in the computer.

The BNC-2090 breakout panel comes ready to use. The break out panel was installed on the equipment rack under the computer and then connected to the card using a SHC68-68-EPM cable. The SHC68-68-EPM cable is an extended performance shielded cable designed specially for the M-Series DAQ card that allows connection to standard 68-pin accessories. The application software automatically installed the software Measurement & Automation Explorer (MAX) to confirm, test, and configure the NI-DAQ card settings. MAX also has its own software-based test panel for troubleshooting the DAQ card.

Next Page - Software Development

 
| Home | Members | Background | Research | Software | Hardware | Picture & Endnote | Project Sponsor |
For problems or questions regarding this web contact cotterj@seattleu.edu