US3153846A - Method of making heat sensors - Google Patents

Description

Oct. 27, 1964 J |NDBERG 3,153,846

METHOD OF MAKING HEAT SENSORS Filed June 21, 1962 HYDROGEN 2Q ARGON 32. V 30 g 3! VACUUM FURNACE VACUUM 2s PUMP 25 f 27 7 I2 I :11 Q- 5 2/3 I7 20 INVENTOR.

JOHN E. LlNDBERG VACUUM FURNACE BY 0W.WM,L M

ATTORNEY United States Patent 3,153,846 METHOD OF MAG HEAT SENSORS Iohn E. Lindberg, 3206 Spring 11111 Read, Lafayette, Calif. Filed. .Iune 21, 1962, Ser. No. 204,125 '7 Claims. (131. 29-400) This application relates to a method of making heat sensors, especially heat sensors of the type disclosed and claimed in my application Serial No. 102,622, filed April 10, 1961, now abandoned.

Such a heat sensor comprises an enclosure, preferably a narrow-diameter metal tube of constant cross-sectional area and of any desired length, within which is a metallic hydride, there being provision for passage for the hydrogen gas that is emitted from the hydride upon the sensors being subjected to certain temperatures. The emission and taking up of the hydrogen act to vary the pressure inside the enclosure. The enclosure is gas-tight and its only opening is connected to a device called a responder, which itself defines a closed chamber connected only to the sensor enclosure. An alteration of the internal pressure within the enclosure therefore affects the responder, which then acts on an electrical circuit or other device to actuate a signal.

The use of various metallic hydrides is disclosed not only in my application Serial No. 102,622, but also in application Serial No. 815,406, filed May 25, 1959, now US. Patent 3,122,728, issued February 25, 1964, and in its various divisions and continuations, including application Serial No. 66,221, filed October 31, 1960, now abandoned.

The present invention deals with special problems that arise when using palladium hydride. Palladium behaves peculiarly, one might say uniquely, with hydrogen, and is particularly difficult to work with at temperatures above about 300 C. When treated in the usual manner, palladium hydride does not give consistent results. Some observers, in fact, have said that it tends to absorb hydrogen when heated above 300 C. instead of giving it off.

An object of this invention is to provide a method for making a completely hermetically sealed heat-detection transducer, completely free from environmental errors caused by such things as pressure and altitude changes, moisture condensation, and so on.

A difiicult problem is to provide a method for making palladium hydride sensors which Will assure that the sensor will give accurate results, and will give the same results each time, over and over. Another object of the invention is to solve this problem.

Other objects and advantages of the invention will become apparent from the following description of a preferred example.

In the drawings:

FIG. 1 is an enlarged view in elevation and in section of a simplified form of heat-detection device embodying the principles of this invention, the sensor thereof being broken in the middle in order to conserve space.

FIG. 2 is a greatly enlarged view in elevation and in section of the sensor in its fully ingassed state.

FIG. 3 is a view similar to FIG. 2 showing the same sensor in its fully outgassed state, or as it is before the metal is ingassed with hydrogen.

FIG. 4 is a diagrammatic view of a set-up that may be used in performing some of the early steps of the method of this invention.

FIG. 5 is a view similar to FIG. 4 of a set-up used in performing some of the later steps.

The invention deals with the use of palladium hydride as a transducing agent for the release or emission of large volumes of hydrogen when elevated to a temperature sought to be detected. When this palladium hydride is enclosed in a constant-volume container or enclosure and subjected to temperature changes, the resultant alteration of pressure within this container or enclosure is employed to actuate the responder.

The solubility of hydrogen in palladium, when treated by this invention, apparently varies as the square root of the pressure, and it decreases with increase in temperature. This solution is commonly termed a hydride, though it is not a stoichiometric compound.

FIGS. 1 to 3 illustrate the structure of a typical sensor 10. A sensor tube 11 may be a non-porous tube of constant cross-sectional area. In applications where the tubes 11 are to be bent or curved around corners, metal is the preferred material. Suitable metals are pure iron, which is impermeable to many gases, stainless steel, nickel-A, and molybdenum, for example. In applications where bending is not required and minimum diffusion is desired, the tube 11 is preferably made from non-porous quartz, ceramic, or special glass. In any event, the inner surface of the tube 11 should not react with the materials it contacts, including hydrogen gas. A typical sensor tube 11 is preferably about 0.040" to 0.060 outside diameter with a wall thickness of preferably about 0.005 to 0.015". Such tubes 11 are preferably about two to forty feet long, although they may be longer or shorter.

FIGS. 2 and 3 show a sensor 10 having a preferred form of palladium transducing agent 12 enclosed in the sensor tube 11. Here, the transducing agent 12 is a filament of palladium wire about 0.005" to 0.030" in diameter, for example. The more nearly the filament 12, when ingassed, fills the tube 11, the less the remaining space; hence the greater the increase in pressure when the same quantity of hydrogen is emitted at any temperature. Care should be taken that the ingassed wire 12 does not fill the tube 11 so tightly that it cuts off passage of gas. A responder 15 is attached to one end 16 (FIG. 1) of the finished tube 11, and the other end 17 is sealed closed.

To make this unit, two things I have found to be vital: (1) the palladium filament 12 must not be overheated; i.e., it must not be heated above about 400 C., and (2) it must be cycled while being ingassed and outgassed in an excess of hydrogen. In order to get the very best results, the tube 11 should also receive special treatment.

In the preferred process the tube 11 is first degreased. For example, with one end of the tube 11 submerged in acetone, compressed air is applied to the surface of the acetone (which may be in a bottle) to blow the liquid acetone through the tube, flushing it out and degreasing the tube 11. Other degreasing solvents that do not react with the metals involved may be used.

One end of the tube 11 is then silver-soldered to a valve that is connected to a tank of hydrogen, and the tube 11 is flushed with hydrogen preferably at about 25 p.s.i.a. for about ten minutes. During this time, the acetone apparently also acts to carry away moisture with it, thereby helping to prevent oxidation when the temperature is later raised. The connection between the valve and the tube 11 is then cut, or the tube 11 otherwise detached.

In order to assure stable operation of the completed fire detector, the preferred process next removes all atmospheric gases and vapors, which had earlier entered the sensor tube 11. For this purpose, the sensor tube 11 only is preferably put into a vacuum furnace 20 and purged with hydrogen. Thus, FIG. 4 shows a vacuum furnace 20 with the end 17 of the sensor 10 extending out from the furnace through a vacuum-tight fitting 21. A second vacuum-tight fitting 22 is at this time plugged by a plug 22a. The end 17 of the sensor tube Ill is connected then to a tube 23, which may be connected directly to a cylinder 24 of hydrogen, or may, as shown, be connected to a manifold 25, which in turn is connected to the cylinder 24 through a valve 26. There is also a valve 27 between the tube 23 and the manifold 25. The manifold 25 is connected (1) through the valve 27 and the tube 23 to the tube 111; (2) through a tube 28 to a pressure-vacuum gauge 29; (3) through a valve 30 to a tank 31. of argon or other gas that is chemically inactive with respect to hydrogen, to palladium, and to the tube 11, (4) through the valve 26 to the tank 24 of hydrogen; (5) through valves 32 and 33 to a vacuum pump 34; and (6) through valves 32 and 35 and a tube as to the vacuum furnace 20.

With this apparatus the system may be purged at room temperature with hydrogen, as follows: The valves 30 and 32 are closed, and the valves 26 and 27 are opened to send hydrogen into the end 17 of the tube 1 at a pressure of about fifty p.s.i.g., asindicated by the gauge 29. However, the valves 33 and 35 are open; so a vacuum of about one micron of mercury is drawn on the end 16 of the sensor tube 11 by the pump 34. This operation continues for the time needed to assure the sweeping of any remaining gases and vapors which might have been trapped on the surfaces of the sensor tube 11 or the transducing agent 12.

Let us now consider the operation of a responder, to see how the finished productworks.

FIG. 1 shows a responder 15 comprising two circular plates 41 and 42, preferably ot non-porous metal, between which is bonded (as by brazing) a thin metal flexible disc or diaphragm 43. The plates 41 and 42 are hermetically sealed together and are in electrical contact for their full peripheries and over a substantial margin, but in the center the diaphragm 43 preferably has a spherical depression 44 called a blister, which is free to move relative to the plates 41 and 42 and constitutes the active or movable part of the diaphragm 43. Use

' of a diaphragm with a blister 44 makes possible the use 41 is formed with a recess 46 in its upper surface.

Then the valves 26 and 27 are closed to cut off the supply of hydrogen and, while maintaining in the tube 11 the pressure at about one micron (0.001 mm.) of mercury, the furnace 20 is then heated to about 2200 F., and held at that temperature about 30 minutes. This heating cleans up the tube 11 and also apparently dissolves any oxides present into the metals.

The temperature is lowered from the 2200 F. to about room temperature, and a length of palladium wire 12 is then put into the tube 11. Then the tube 10 is brazed to the responder 15, as by brazing it in a hydrogen atmosphere, using a suitable brazing alloy. Here, as at other times, care must be taken not to heat the palladium above 400 C. Next the tube 10 may be put back into the vacuum furnace, connected as shown in FIG. 5, with the end 16 coming out through the fitting 22, and heated to a little lower than 400 C.no more in any event. Then the valves 26 and 27 are opened to send hydrogen into the tube 11 at a pressure of about 50 p.s.i.g. Then the furnace is gradually cooled to room temperature, still holding the hydrogen pressure at 50 p.s.i.g. The operation is then repeated several times, heating and then cooling the palladium-always keeping it below 400 C. and always having an excess of hydrogen under pressure. This operation need not be done in the furnace 20, but can bedone in air with electrical heating or gas furnace heating.

Next, the sensor 10 ,is heated to the set point-where ever it is supposed to cause actuation of the respondere.g., 250 F. or 350 F. and, with the valve 26 closed, the valves 32 and 33 are opened and the tube 11 is evacuated until the responder 115 is exactly at the actuation point. Then the tube 11 can be sealed, if it tests all right, testing being done by heating it and seeing whether the responder 15 operates at the proper temperature. If not, the hydriding process is repeated before sealing the tube end 17.

For many uses, it is preferred to add a noble gas inside the tube 11. To do this, the furnace 20 is cooled to a desired temperature and argon is added from the cylinder 31. It may be added at, say, 150 F. or 250 F. Or a chart may be used to determine the room temperature and pressure corresponding to the actuating pressure at any desired temperature, and that amount is then added at room temperature. Once the pressure in the tube 11 is correct for the fire detector to respond to the selected temperature, the tube end 17 is sealed by some suitable means, such as by a Cuplat wirebrazed inside it.

The force necessary to deflect the blister 44 against an electrode 47 can be chosen to accommodate a wide range of values by suitable choice of mechanical parameters. Once this force is determined, the dimensions of the sensor tube 11 and the amount of transducing agent 12 may be chosen by design to provide the force necessary to obtain contact between the blister 44 and electrode 47 at a certain temperature.

in addition to mechanical design considerations, the necessary deflecting force may also be altered by precharging the anti-sensor chamber 48 with a gas under pressure or by partially evacuating it. To accomplish this, gas is forced into (or withdrawn from) the tube 49 after its attachment to the plate 42 and before it is closed by its cap 50. The required deflecting pressure against the blister 44 becomes greater as more gas is present in the chamber 48.

Alternatively, the deflecting pressure may be effectively lowered by precharging the inside of the sensor tube 11 and the sensor Chamber 51 with gas. In this case, if the ambient pressure in the sensor chamber 51 is greater than normal, less than normal gaseous emission from the transducing agent 12 is required to deflect the blister 44 against the electrode 47. Most gases may be employed for this purposes; however, ideally the gas should not react chemically with its surrounding materials. Particularly suitable here, as well as in the tube 11, are the noble gases, such as helium, argon, neon, and xenon, especially since they do not readily diffuse through most materials. As a consequence, a precharged pressure of argon, for example, may be maintained for an indefinite length of time to retain a desired biasing of the diaphragm 43, as described.

The responder 15 may be connected to an alarm circuit which, as shown in FIG. 1, may be a simple visual indicator consisting of a lamp 52 in series with the conducting wire 53 and a source 54 of electrical current, which may be a battery, as shown, or may be a source of alternating current. A return path for the electrical circuit may be provided by grounding either one of the plates 41 or 42 and is shown as a ground wire 55.

In operation, when the sensor 10 is exposed to heat at a level high enough to cause the argon gas to expand and increase in pressure, in the sensor chamber 51, it exerts pressure upon the blister 44. This pressure tends to move the blister 44 away from the plate 41 and toward the plate 42. The pressure in the sensor chamber 51 is a function of the temperature of the sensor 10, and in general there will be a one-to-one correspondence between the temperature of the sensor 10 and the pressure within the sensor chamber 51. This pressure, if great enough, .will cause the blister 44 to make contact with the electrode 47, but no contact will be made unless the temperature of the sensor 10 is at or above a definite level.

A similar action takes place at a. higher temperature when a portion of the transducing agent is heated above its threshold temperature and emits hydrogen. Thus, in this embodiment, the argon serves for actuation at a pre determined over-all temperature of the tube 12, and the hydride causes actuation at a predetermined spot temperature.

When the sensor is exposed to heat at a level high enough to cause the blister 44 to make contact with the electrode 47, current flows from the battery 54 through the lamp 52, the conductor 53, the electrode 47, and the blister 44 to the plates 41 and 42, and returns to the battery through ground line 55. This current flow causes the lamp 52 to light and provides a visual indication that the temperature of the sensor 10 is at or above a certain level. In this sense, the device shown in FIG. 1 functions both as an all-point end as a threshold temperature indicator. When heat is removed from the sensor 10, the transducing agent 12 cools and reabsorbs its previously emitted gas, resulting in reduction of the pressure exerted upon the blister 44. The blister 44 moves away from the electrode 47, breaking the electrical circuit, and the lamp 52 goes out.

In practice, the sensor 10 is placed in the area whose temperature is to be monitored, while the responder may be located upon or behind a shielded Wall or at some easily accessible area. Thus only the sensor 10 itself need be exposed to possible heat sources, and it contains no element of the electrical circuit. In this manner, protection for the responder 15 and its associated alarm circuit may be provided.

After being treated according to this invention, the palladium hydride 12 no longer suifers the disadvantages which observers noted in the past with palladium and hydride. It gives off gas, when heated, as a linear function, and takes up the same gas when cooled. It can be heated much hotter than 400 C.up to about 2000 F.and will still function properly and repeatably. However, if it is to be used at such high temperatures as 2000 F., it does tend to have an adverse effect upon a stainless steel tube 11. The tube 11 may be molybdenum, or the filament 12 can be wrapped with a molybdenum ribbon, as disclosed in the prior applications mentioned earlier.

To those skilled in the art to which this invention relates, many changes in construction and widely differing embodiments and applications of the invention will suggest themselves without departing from the spirit and scope of the invention. The disclosures and the description herein are purely illustrative and are not intended to be in any sense limiting. For example, the responder may also be put into the vacuum furnace; or the sensor may be heated electrically in a vacuum by using the resistance of its own tube walls, without using a furnace.

I claim:

1. A method of making a heat sensor, comprising the steps of:

inserting a palladium wire into a hollow tube,

cyclically heating said tube to a temperature above 250 F. and no higher than 400 C. and then cooling it to a substantially lower temperature while sending hydrogen at high pressure into said tube and maintaining the hydrogen at high pressure in said tube to cause said palladium wire to ingas some of said hydrogen,

then withdrawing excess hydrogen from said tube; and

sealing oif the tube interior from the atmosphere.

2. The method of claim 1 wherein before sealing off the interior of said tube from the atmosphere said passage is filled with gas that is chemically inactive with re spect to hydrogen, to said palladium, and said tube.

3. A method of making a heat sensor, comprising the steps of:

purging an open end tube with hydrogen,

6 evacuating the purged tube, inserting a palladium wire into the purged tube, heating said tube to a temperature above 250 F. and

no higher than 400 C., forcing hydrogen at high pressure into said tube, lowering the temperature of said tube while maintaining the hydrogen under pressure therein, to cause said palladium wire to ingas some of said hydrogen,

repeatedly heating and cooling said tube to cyclically ingas and outgas hydrogen several times into and out from said palladium,

then withdrawing excess hydrogen from said tube, and

sealing off the tube interior from the atmosphere.

4. The method of claim 3 wherein before sealing 01f the interior of said tube from the atmosphere said passage is filled with a noble gas.

5. A method of making a heat sensor, comprising the steps of:

degreasing a tube of metal that does not dissolve hy drogen therein, by flushing with solvent, evacuating the tube to a vacuum of about one micron of mercury,

then purging the tube by alternately (a) flushing with hydrogen, to drive out substantially all other gases from said tube and (b) evacuating said tube, raising the temperature to about 2200 F. while hold ing a vacuum of about one micron of mercury, cooling said tube to about room temperature, inserting a clean palladium filament into said tube, heating said tube to about 400 C., and

sending hydrogen under pressure into said tube,

reducing the temperature of said tube to about room temperature to cause said filament to ingas hydrogen, cooling said tube,

repeatedly and cyclically heating and cooling said tube,

withdrawing excess hydrogen from said tube at a desired actuation temperature, and

then sealing off the interior of said tube from the at mosphere.

6. The method of claim 5 wherein before sealing oil the interior of said tube from the atmosphere said pas sage is filled with noble gas.

7. A method of making a heat sensor, comprising the steps of:

degreasing a tube of metal that does not dissolve hy drogen therein,

then purging the tube by alternately (a) flushing with hydrogen, to drive out substantially all other gases from said tube and (b) evacuating said tube, raising the temperature to about 2200 F. while libld ing a vacuum of about one micron of mercury, cooling said tube to about room temperature, inserting a clean palladium filament into said tube,

heating said tube to about 400 C.,

sending hydrogen under pressure into said tube,

reducing the temperature of said tube to about room temperature to cause said filament to ingas hydrogen, cooling said tube,

re-heating and re-cooling said tube several times,

heating said tube to a desired actuation temperature,

Withdrawing excess hydrogen from said tube, and

then sealing off the interior of said tube from the at mosphere.

Kollsman Apr. 11, 1939 Polye Nov. 20, 1956

Claims (1)

1. A METHOD OF MAKING A HEAT SENSOR, COMRPISING THE STEPS OF: INSERTING A PALLADIUM WIRE INTO A HOLLOW TUBE, CYCLICALLY HEATING SAID TUBE TO A TEMPERATURE ABOVE 250*F. AND NO HIGHER THAN 400*C. AND THEN COOLING IT TO A SUBSTANTIALLY LOWER TEMPERATURE WHILE SENDING HYDROGEN AT HIGH PRESSURE INTO SAID TUBE AND MAINTAINING THE HYDROGEN AT HIGH PRESSURE IN SAID TUBE TO CAUSE SAID PALLADIUM WIRE TO INGAS SOME OF SAID HYDROGEN, THEN WITHDRAWING EXCESS HYDROGEN FROM SAID TUBE; AND SEALING OFF THE TUBE INTERIOR FROM THE ATMOSPHER.

Images