A proposed solenoid-actuated valve would open for a short time to pass a very small quantity of fluid. The valve could be used to meter a specified total quantity of the fluid through repeated actuations.
In the operation of a typical solenoid-actuated, momentary-opening valve of older design, the solenoid is first actuated to magnetically pull a poppet away from a valve seat to effect opening. Shortly thereafter, the current in the solenoid is turned off, allowing a spring to halt the motion of the poppet away from the seat and then push the poppet back into the seat to effect closure. In the proposed valve, there would be no return stroke; instead, to effect momentary opening, the poppet would be toggled between two "closed" positions. One advantage offered by this approach would be that without a return stroke, a shorter opening time could be achieved. Another advantage would be protective redundancy; if a leak were detected with the poppet in one of the two "closed" positions, one could simply switch the valve to the other "closed" position.
In the proposed valve (see figure), a poppet would be spring-loaded and latched into either a rearward or a forward "closed" position. The spring-loading and latching action would be effected by a Belleville spring, which would be stable in either "closed" position and would snap ("oil-can") between the two stable positions when the poppet was moved through the middle position. Two armatures (magnetically permeable plungers) would be mounted on the poppet, separated by an inner yoke. There would be two solenoids; a forward-closing and a rearward-closing coil.
The armatures would be positioned, relative to the solenoids, to configure the magnetic-flux paths to obtain the desired magnetic forces. The configuration would be such that upon application of electric current to the forward-or rearward-closing coil, the resulting magnetic force would pull the poppet toward either the forward or the rearward "closed" position, respectively. When neither coil was ene rgized, the force of the Belleville spring would hold the poppet in whichever "closed" position was reached most recently.
This work was done by George Yankura of Caltech for NASA's Jet Propulsion Laboratory.