US2700537A - Humidity changer for air-conditioning - Google Patents

Description

Jan. 25, 1955 N. A. PENNINGTON 2,700,537

HUMIDITY CHANGER FOR AIR CONDITIONING .Filed June 29, 19571 5 sheets-sheet 1 j zal l 'III Il, /IIlI/IIIIILYIIII/ lI/Il/l ATTORNEY Jan. .25, 1955 N. A. PENNINGTON 2,700,537

HUMIDITY CHANGER FOR AIR CONDITIONING Filed June 29, 1951 3 Sheets-Sheet 2 NEAL A. PENNmaroN,

INVENToR;

ATTOR NE Y.

N. A. PENNINGTON 2,700,537 HUMIDITY'CHANGER FOR AIR CONDITIONING 3 Sheets-Sheet 3 wir a1/ M14/MATURE Jan. 25, 1955 Filed June 29, 1951 /vE/vro/P,

ATTORNEY nite Patent 2,700,537 HUMIDITY CHANGER FOR AIR-CONDITIONHG Neal A. Pennington, Tucson, Ariz., assignor of one-fifth to Robert H. Henley, Tiptonville, Tenu., and onefourth to Roger Sherman Hoar, South Milwaukee, Wis.

Application June 29, 1951, Serial No. 234,301 21 Claims. (Cl. 261-83) My invention relates to new and useful improvements in a humidity-changer for air-conditioning, and more particularly to a combined humidifier-dehumidifier, primarily for use in a universal (i. e., both summer and winter) air-conditioning apparatus.

This present application is a continuation, as to all common subject-matter, of my copending application for improvements in universal air-conditioner, Serial No. 765,554, iled August 1, 1947, now abandoned without prejudice to this present application and one other continuation-in-part, Serial No. 234,800, tiled July 2, 1951, and embraces the elected subject-matter of that parent case, plus certain subsequent improvements. This elected subject-matter is that part of my complete apparatus which effects moisture transfer.

Outdoor air, when heated in winter for use indoors, thereby almost always acquires too low a relative humidity, and hence requires to be humidiiied. Outdoor air, when cooled in summer for use indoors, thereby almost always acquires too high a relative humidity, and hence requires to be dehumidied.

Accordingly it is the principal object of my present invention to devise a s` ple mechanism, adapted to perform selectively, at will or under automatic control, these two functions of humidifying and dehumidifying the incoming air.

It is a further object to devise this mechanism so that it performs these functions by moisture-exchange between two streams of air (one incoming and one outgoing), transferring the moisture in either direction at will or under automatic control, without any rerouting of the two main air-streams.

In addition to the objects above stated, I have worked out a number of novel and useful details, which will be readily evident as the description progresses.

My invention consists in the novel parts and in the combination and arrangement thereof, which are defined in the appended claims, and of which one embodiment is exemplified in the accompanying drawings, which are hereinafter particularly described and explained.

Throughout the description the same reference nurnber is applied to the same member or tosimilar members.

Figure 1 is a longitudinal vertical central section of my apparatus.

Figure 2 is a transverse vertical section of my apparatus taken along the line 2 2 of Figure l.

Figure 3 is a horizontal section of the rotary moisturetransferrer, and adjacent parts, of my apparatus, taken along the line 3-3 of Figure 2.

Figure 4 is an enlargement of a portion of Figure 3, to illustrate my means for preventing the leakage of air past my rotary moisture-transferrer in either air-passage, and for preventing the leakage of air from one air-passage to the other at the periphery of said transferrer.

Figure 5 is a diagrammatic detailed showing of the gear-shift for changing the rotation speed of my heatexchanger and my moisture-transferrer, this gear-shift being merely indicated in Figure l.

Figure 6 is a chart, which shows the psychrometric interaction of the air-streams as they pass through my rotary moisture-transferrer, and of the particles of the latter as it rotates across the air-streams, all under' humid summer conditions.

Figure 7 is a psychrometric chart, which shows the same under dry summer conditions.

Figure 8 is a psychrometric chart, which shows the same under Winter conditions.

Referring now to Figure 1, we see that 11 is the main container of my invention, in which 12 is an air-inlet from outdoors. Centrifugal fan 13 impels this air into passage 14, thence through filter-pad 20, and thence through the upper portion of rotating wheel-like moisturetransferrer 15.

This moisture-transferrer 15 is built Very similar to the aluminum wool pad of my Patent No. 2,464,766, but employs novel means to prevent the leakage of air from one air-passage to the other, and to prevent the leakage of air past the periphery of the moisture-transferrer. ghes'e novel means will be described and explained later erem.

The rim 16, ribs 17, and hub 18 of my moisturetransferrer are of substantially the same width in an axial direction. Each of the sectors between successive ribs is fully and rigidly but loosely stuffed with some airpervious liquid-absorbing non-heat-conducting packing 19, which may be excelsior or other similar filamentous material, or corrugated asbestos paper or the like with the corrugations extending axially.

The packing should be impregnated with some appropriate solution of an vhygroscopic liquid or salt.

For an impregnating liquid, I prefer glycerine, or one of the polyethylene glycols (such as triethylene, or polyethylene 200 or 300). Glycerine I consider the best; the drawback of high viscosity, which has prevented the use of this absorbent in liquid circuits, is no drawback to it as an impregnant of such a rotary moisture-transferrer as hereinafter described.

For an impregnating hygroscopic salt solution, I prefer some such combination as that described and claimed in my copending application for improvements in hygroscopic composition of matter, Serial No. 231,445, filed June 13, 1951, to which application reference may be made for further details.

When in the claims I refer to a salt, I to include a combination of salts.

The use of a rotary moisture-transferrer packed with an inert air-pervious carrier having a rigid space-structure, and impregnated with a liquid absorbent combines the simplicity of reactivation heretofore possessed only by solid adsorbents, with the lack of resistance to airflow heretofore possessed only by liquid absorbents.

It also possesses the advantages of complete counteriiow and one-to-one correspondence, applied to moisture transfer, that the aluminum-wool-packed wheel of my Patent No. 2,464,766 applies to the transfer of sensible heat. Thus, in my rotary moisture-transferrer, the sensiintend thereby clear into the packing before the tion passes into the other air-stream and begins to reverse the process. This causes the packing to act as though it were composed of many discrete minute specks. discussion of Figures 6, 7, and 8, which follows, I shall refer to an almost infinitesimally small-angled almost innitesmally thin (in the direction of air-flow) sector of such specks, as a particle, and shall refer to the plane in which it rotates as a level of my wheel.

Two further characteristics of my rotating moistureexchanger are: its ability to transfer moisture in from either air-stream to the other at will by merely changing I its R. P. M. (this will be explained later herein, in connection with my speed-changer); and its ability to utilize desirable highly-hygroscopic liquids or salt solutions, whose high viscosity or corrosiveness renders them unt for use in a spray or flowed over a lixed pad.

Moisture-transferrer 15 rotates clockwise in Figure 2.

It has two alternative speeds of rotation, as will be explained later herein.

That sectoral portion of the incoming air rst treated -by this moisture-transferrer, passes therefrom into bypass 23, and thence into the outgoing air-stream, for

purposes which will be explained later herein. If damper 24 be closed and damper 25 be opened, this air, instead of being thus discarded, will pass with the rest of the L incoming stream. v If there be no damper 25, and damper 24 be closed, all the incoming air will by-pass the bypass.

In any event, the main stream of incoming air continues on in passage 26, until it encounters heat-transferrer 27, rotating at a speed of about 30 R. P. M.

This heat-transferrer 27 is preferably of the sort of the aluminum wood pad of my Patent No. 2,464,766, already alluded to.

Instead of packing the heat-transferrer with metal wool, it would be permissible to pack it with a foraminous carrier impregnated with some non-hygroscopic liquid, such as mineral-oil of the sort employed in my copending application, Serial No. 505,924, led October 12, 1943, now Patent No. 2,536,081, issued January 2, 1951.

Let us now compare moisture-transferrer and heattransferrer 27. Except for the nature of the packing, these two transferrers could be identical. However, l have shown them slightly different, so as to illustrate two alternative means for preventing the leakage of air past them in each air-stream, or from stream to stream.

My heat-transferrer 27 rotates in a casing 28, spanned by bridges 29, with sectoral openings between each face of the casing and the corresponding bridge, all as shown and described in my Patent No. 2,464,766.

But my moisture-transferrer 15, although having very similar bridges 30, has a single shroud 31, which projects inwardly from the main container 11, nearly touching the center line of the periphery of rim 16 of the moisture-transferrer. The rim carries two annular pieces of felt 32, one being on each side of, and touching or nearly touching the sides of, the inwardly projecting edge of the shroud. In the horizontal section portrayed by Figure 4, we see that, in addition to partition 33, which separates the two air-passages, I have between the two bridges 30, in extension of the partition, two small plates 34 attached to the two bridges 30, the shroud 31, and the wall 11 of the container, and almost touching the rim 16 of the rotating moisture-transferrer and the two felts 32.

These expedients could equally well be substituted in connection with the heat-transferrer. They serve to prevent appreciable leakage of air past the moisture-transferrer, or from passage to passage.

In both types of transferrer, it is desirable to hold the packing in place in some convenient manner, as for example by metal screening secured to each face of the transferrer in any convenient manner, as for example (see Figure 2) by being out into sectors 35 inserted between successive ribs 17, and in turn held in place by wires 36, strung in spider-web formation through transverse holes 37' through the ribs close to the edges thereof.

The incoming air, after having been cooled by heattransferrer 27, is further cooled by passing through evaporative pad 38, which I call my secondary pad. Water from tank 39 is sucked through pipe 40 by electric pump 41, by which this water is impelled through feed pipe 42 to the top of pad 38, whence it trickles down through this pad, so much of the water as is not evaporated therein, being returned to the tank through pipe 43.

The tank is supplied with water from any convenient source through supply pipe 46 and ball-cock 47.

The air then enters the room or other enclosure through louvres 44.

Exhaust air leaves the room or other enclosure through louvres 45, and is cooled by passing through evaporative pad 48, which I call my primary pad, and which is supplied with water from tank 39 by pump 49 in exactly the same manner as secondary pad 3S, already described.

In place of each of these two pads, I could use the pads and sprinkler of the copending application of McKinney and self, for improvements in Evaporative Air- Cooler, Serial No. 231,394, led June 13, 1951, now Patent No. 2,681,217,y issued June 15, 1954.

Adiabatically cooled by primary pad 48, the outgoing air in turn cools heat-transferrer 27, being itself warmed in the. process, and then passes through radiation-shield 50 into chamber 51, where it is joined by the by-passed. portion of incoming air entering the chamber through-*bypass 23. l

In this chamber there is an air-heater, which might be an electrically-heatedv grid of German-silver, o-r any equivalent air-heating means; but, as shown, is a furnace. (preferably gas) 52, which heats. the air by means, of tins 53 on its ue 54. This furnace, and its ue. and fins, are represented here merely conventionally.

The furnace or alternative form of air-heater, could be located in an adjoining chamber, with heat-exchange means (such as heat-transferrer 27) transferring heat from that chamber to chamber 51.

The outgoing air, greatly heated by the furnace, then passes through radiation-shield 55.

Chamber 51 should be lined with some highly heatreflecting material, backed by heat-insulating material.

Radiation-shield 50 may be of any convenient construction which will shield heat-transferrer 27 from the direct rays emanating from air-heater 52, while permitting the free passage of air. In other words, it should be substantially impervious to heat-radiations, and yet pervious to air. I prefer a stationary pad of much the same sort of aluminum-wool as employed as a filler for heat-transferrer 27. Radiation-shield 55 is similar, and similarly protects moisture-transferrer 15 from the rays of airheater 52. Y

The outgoing air, having been raised in temperature by passing through heat-transferrer 27 and through radiation- screens 50 and 55, and by the addition of by-passed incoming air, and by air-heater 52, then passes through moisture-transferrer 15, where it dries and heats the hygroscopic packing thereof.

Thence it is sucked through centrifugal fan 56 into passage 57, whence it passes outdoors through exit opening 58. In this opening is butterfly valve 59.

From passage 57, in the opposite direction there extends a draft-passage 60, which connects with furnace 52 in such manner as to furnish draft-air thereto. The amount of this draft can be regulated by butterflyvalve 59.

By thus utilizing the exhaust air as draft-air for the furnace, I not only avoid the expense of an additional draft fan, and obtain preheated draft-air, but also utilize the principle that somewhat damp air is better than dry air for draft purposes. I have found that this simple expedient of utilizing my hot exhaust air for draft purposes results in a fuel-saving of about 15%.

Motor 61 drives shaft 62 through pulley 63, belt 64, and pulley 65. Fans 13 and 56 are keyed to, and driven by, this shaft 62.

This shaft 62, through gear-reduction 66, drives shaft 67 at a very slow speed (about 3 R. P. M., or less), and drives shaft 68 at a relatively-faster speed (about 25 to 30 R. P. M.). On these speeds, see later herein. Shafts 67 and 68 enter speed-changer 69, the details of which are shown enlarged and somewhat symbolically in Figure 5.

In that figure, slow shaft 67 terminates in male clutchmember 70. Aligned with shaft 67, ythere is a driven shaft 71, on which and keyed thereto there slides gear 72 and female clutch member 73 integral with this gear.

Fast shaft 68 terminates in female clutch-member 74 and gear 75 integral therewith. Aligned with shaft 68, there is a driven shaft 76, on which and keyed thereto there slides gear 77 and' male clutch-member 78 integral with thisV gear. A fixed dog 79 engages and locks this gear when in its upper position.

Driven shaft 71 drives sleeve 80 which is keyed to rotary moisture-transferrer 15. Driven shaft 76 drives shaft 81 which isl keyedy to rotary heat-transferrer 27.

The operation and object of this whole speed-change system, just described, isl as` follows.

Gear 72 and its female clutch member 73, and gear 77 and its male clutch member 78, are raised or lowered simultaneously. When they areboth in their lowered position gears 72 and 75 are disengaged, and both clutches are set. The slow speed of' shaft 67 is transmitted, through shaft 71 and sleeve 80 to moisture-transferrer 15; and the fast speed of shaft 68 isv transmitted, through shaft 76 and shaft 81, to heat-transferrer 27.

When they are both in their raised position, the situation shown in Figure 5, both clutches are released, gear 72 engages gear 75, and gear 77 engages dog 79. The fast speed of shaft 68 is transmitted through gears 75 and 72 toy shaft 71, and thence through sleeve 80 to moisturetransferrer 15; and' the engagement of gear 77 with dog 7 9 locksv shaft 76 and thence shaft 81 and heat-transferrer 27 against rotation. In place of gear 77 and dog 79, any convenient form of brake could be used.

The optimum rotation-speeds differ somewhat for various materials, but can` easily be experimentally determined for each. The optimum fast speed for both exchangers is of the order of 25 to 30 R. P. M., but somewhat more would be permissible.

The optimum slow speed for a fully-impregnated moisture-exchanger is of the order of 3 R. P. M. or less. For example, for excelsior fully impregnated with triethylene glycol, it is 2 to 3 R. P. M. or excelsior fully impregnated with the best hygroscopic salts, itis 1/3 to 1/2 R. P. M For asbestocel fully impregnated with the best hygroscopic salts, it is of the order of l/s R. P. M. On the eiect of underimpregnation, see later herein.

The object of the two speeds for the moisture-transferrer will be given later herein when I discuss the psychrometry of my apparatus.

T1 and T2 are two thermostats (but might as well represent one double-stage thermostat), and H is a humidistat. These could be placed at any strategic location in the room or other enclosure which is being air-conditioned, or equivalently in the outgoing air-passage just inside the louvres 45. These three stats control the turning on and off of pumps 41 and 49, and valve V which supplies fuel to furnace 52, all in a manner described in the second continuation-in-part of parent application Serial No. 765,554, which second continuation-in-part is identified earlier herein as Serial No. 234,800, tiled July 2, 1951. Said parent case has now been abandoned without prejudice to the two continuations in part.

Let us now turn to Figure 6, which shows the psychrometric circuits of the air in the humid summer use of my machine; and also a temperature-moisture plot of two typical particles of my moisture-transferrer, one on the outdoorward face and the other on the indoorward face of the moisture-transferrer.

From the point of view of the sensible-heat and heatcapacity of a particle as defined earlier herein, I deal with the totality of the carrier, the sorbent, and the absorbed water. From the point of View of the vapor-pressure of the particle, I deal with merely the absorbed water, inasmuch as the vapor-pressures of the carrier and of the sorbent are zero, or at least negligible.

Psychrometric charts are well-known devices for plotting the thermodynamic characteristics of the atmosphere. But the use to which I put these charts in connection with my present invention is also a discovery of my own. Inasmuch as there is, to my knowledge, no available literature on this use, I shall now explain it.

The chart which I use has nearly vertical lines representing dry-bulb temperature, horizontal lines representing dew-point and also vapor-pressure and moisture content, northwest southeast lines representing wet-bulb and constant enthalpy, and curving representing relative humidity.

For plotting the thermodynamic characteristics of particles of my moisture-exchanger, I employ the dry-bulb lines to represent sensible temperature, and the dew-point lines to represent vapor-pressure. Thus at any point on the chart, a particle of my moisture-exchanger plotted thereat, and atmosphere plotted thereat, will be in both temperature equilibrium and vapor-pressure equilibrium with each other: i. e., in complete equilibrium.

But we must note and always bear in mind one very important difference between plotting the characteristics of the atmosphere and plotting the characteristics of my particles: namely that although the dew-point lines are lines of constant moisture-content as respects the atmosphere, they do not perform this function with respect to my particles. For, in this latter connection the lines of constant moisture-content are practically coincident with the relative-humidity lines near the bottom of the chart, veering more and more to the left of those lines (although still very slightly) as we approach the top of the chart.

The wet-bulb lines have no apparent significance, as applied to my particles.

A plotted line-of-change of characteristics of one of my particles, indicates: gain of moisture if it veers to the left of one of my almost relative-humidity lines, loss of moisture if it veers to the right thereof, loss of sensible heat if it veers to the left of one of the dry-bulb lines, gain of sensible heat if it veers to the right thereof.

This ends the explanation of how the psychrometric characteristics of one of my particles can be plotted on a chart which was designed merely to plot the characteristics of atmosphere.

And now to discuss the psychrometric interaction between my particles and air.

Between two plotted points, which respectively represouthwest northeast lines .i

' characteristics represented by a.

`creasingly less than that of the air.

sent a particle and air in contact with it, moisture flow from the point higher up on the chart to the point lower; and sensible-heat iiows from the rightmost point to the leftmost.

But, as to heat-iiow, this does not tell the whole story, for change in sensible heat takes place in three interrelated ways, namely: (l) The moisture-exchange, if condensation, tends to warm up the particle, by conversion of latent heat into sensible heat; if evaporation, it tends t0 cool the particle, by conversion of sensible heat into latent heat. (2) The sensible-heat of the transferred moisture is averaged into the sensible-heat of the transferree, but has no direct etl'ect on the temperature of the transferrer. (3) Between the air and the particle, each with its temperature modified by the two foregoing steps, there is a ow of sensible-heat by direct contact.

Finally there is an important empirical fact as to my moisture-transferrer, namely that (due to the heavy temperature-dierential) the sensible-heat exchange eifected thereby between the atmosphere and a particle is very much more rapid than the moisture-exchange, and hence predominates during the early part of the passing of the particle across either air-stream.

It is particularly to be noted that inasmuch as each moisture-exchange particle moves transversely across each air-stream, no single point on the plot of either corresponds to any single point of the plot of the other. Each point on the plot of either air-stream represents the average of the conditions of that stream at a given level of the moisture-exchanger, and hence corresponds to a whole half of the plot of a given particle, which half plot represents the changes in characteristics of the particle as it crosses the airstream at that particular level.

With these Preliminaries out of the way, let us now consider circuit abcda, which constitutes the plot of the change in characteristics of each particle of the outermost level of the moisture-exchanger as that particle crosses the two air-streams at that level, and interchanges sensible heat and moisture therewith.

The incoming outdoor air enters this level with characteristics A.

As one of the particles on the outdoorward face of the moisture-exchanger leaves the outgoing air-stream and enters the incoming air-stream, it has thermodynamic The reason for this will appear later. The atmosphere which it encounters has characteristics represented by A, and it will continue to encounter air of characteristics A all the way across the incoming stream.

The particle, being much hotter than the atmosphere, will tend to cool olf rapidly. If there were no moistureexchange, the plot of the characteristics of the particle would lie downward along one of the almost relativehumidity lines.

But as the particle has a higher vapor-pressure than the air, it will give up moisture to the air. This moistureexchange will cause the plot of the particle to Veer to the right. But the cooling due to evaporation will speed of the plot toward cooler vertical lines. Thus the plot will reach b, a point of vapor-pressure equal to that at A, very quickly, after having traversed only a small sector of the incoming air-stream.

The lesser temperature of the air from that of the still continues to draw the due to the vapor-pressure of the particle being now inexchange becomes less, as the plot of the particle apstics tend from A toward a. Each successive narrow will have its character.

7 sector of the air will have its characteristics tend from A toward lower and lower points on the abc plot, until those of the final sectors of the air will tend from A toward c.

The plot of the characteristics of the particles in the next level of the pad will be similar to abc, but will lic a little to the right thereof on the chart, until finally the characteristics of the particles in the indoorward face of the pad, will be represented by abc. v

The average characteristics of the sectors of the incoming air-stream which rst encounter these successive particles at separate levels, will be represented by AB. By discarding, through by-pass 25,`the incoming air of the rst air-sectors, the retained air will average in its passing through the pad, a change in characteristics from A to B.

Let us pause a moment to summarize the reasons for this discarding. From the foregoing pyschrometric description, we have seen that, as each particle (i. e., narrow sector) of each level of the moisture-transferrer enters the incoming air-stream, the particle is so hot that it continues to give up moisture and sensible heat to the incoming air, whereas what We want it to do is to extract moisture from that air and not to heat that air any more than can be avoided. But by the time the particle has crossed about a fifth of the stream, it has cooled sufliciently so that it has given up all. or practically all of its sensible heat, and has begun to extract moisture from the air. Hence, by discarding through the by-pass', this one-fifth of the air-stream, the effect on the remaining four-fifths of the stream is just what we want, namely dehumidication with only such heating as is due to the heat of condensation.

And an important by-product is that we conserve the sensible heat of the by-passed air, for use over again in the outgoing. stream,

Turning now to the outgoing air7 let us assume that it enters the moisture exchanger with characteristics H. Then by the same reasoning as detailed hereinbefore, the plot of the characteristics of particles `lill be cd'a in the indoorward level of the moisture-exchanger, and cda in the outdoorward level; the plot of the average characteristics of the first sectors of air of the outgoing stream will be Hl; and the plot of the average of most of the air-sectors will be Hl. Only, as all the outgoing air gets discarded anyway, we dont need to by-pass the air of the early sectors.

I have thus described the psychrometry of merely that part of my apparatus comprised in my moisture eX- changer. The complete psycnrometry of my apparatus will be givenin the second continuation-iu-part already mentioned.

Let us now turn to Figure 7, which shows the psychrometric circuits of the air in the dry summer use of my machine with the furnace shut oli; and also a temperature-moisture plot of two typical particles of my moisturetransferrer, one being on the indoorward face of the moisture-transferrer and the other on the outdoorward face of the moisture-transferrer.

Comparing the relative ch-aracteristics of the incoming air (A) and the outgoing air (H) in Figure 6, with the corresponding relative characteristics in Figure 7, we see that in Figure 6 the moisture-differential and the temperature-differential opposed each other, whereas in Figure 7 both differentials tend to transfer moisture in the same direction, namely from the outgoing stream to the incoming stream. Thus in the humid air case (Figure 6) 'e had to use slow rotation, in order to give the temperature-differential the preponderance and thus transfer moisture from the incoming air. Whereas in the dry air case. (Figure 7) either the fast or the slow rotation would transfer moisture in the direction which we desire, namely to the incoming air.

Referring back to Figure 6, we see that the characteristics of a particle in the outdoorward face of my moisture-transferrer pass through three phases as the particle crosses the incoming air of characteristics A. From a to b, inasmuch as is it hotter and has a higher vaporpressure than the air, it gives up both sensible heat and moisture to the air. Some of its original sensible heat is, of course, absorbed by evaporation. From b to c, inasmuch as the particle is still hotter than the air but now has alower vapor-pressure than the air, it still gives up initial sensible heat (and in addition the heat created by 8 condensation) to the air, but new extracts moisture from the air.

In Figure 7, the plot of the characteristics of a corresponding particle is represented by the loop mpnqm. Note that the characteristics traverse this loop counterclockwise.- Each particle of the 'outdoorward level of my moisture-transferrer, in the dry summer case now under consideration, enters the incoming air-stream with characteristics in, and encounters (all the way across this stream) air with characteristics A. inasmuch as the particlehas a much higher vapor-pressure than the air, although only slightly cooler than the air, there will be an initial burst of moisture discharge from the particle to the air, with resulting evaporative cooling or the particle' and hence of the air. Hence the loss of vmoisture and sensible heat by the particle, in this initial stage, from m to p. From then on the moisture-loss is less rapid, and the particle gains sensible heat from p to ny when it leaves the incoming stream.

On entering the outgoing stream with characteristics n, the particle lirst encounters air with characteristics even to the left of I. Thel greater temperature diierential, due to this fact, offsets the heat of condensation, with the result that there is less bulge at q than at p. Apart from this the plot from n to m. is the converse of the plot from m to n.

Loop m'p'nqm' portrays the change of characteristics of a particle of the indoorward level of the moistureexchanger.

In Figure 7, as in Figure 6, AB represents change in characteristics of the average of the by-passed initial sectors of the incoming air. HI' represents the change in characteristics of the corresponding initial sectors of the outgoing air, which (for the same reasons as in Figure 6') is not by-passed. AB represents the change in characteristics of the average of the main portion of the incoming air. HI represents ditto for the outgoing air.

A comparison of lines AB and AB shows that those sectors of the incoming air which the particles first traverse, emerges from the moisture-transferer not so much cooled, but slightly more humidified, than the rest of the incoming air. In fact, in practice, due to slight differences in air speed, this temperature difference may be slightly more, and this moisture diferenc'e slightly less, than as shown. So no harm results from leaving the by-pass open. And it would be extremely inconvenient to have to keep opening and closing the bypass, as the furnace goes on and off, due to humidity changes, during summer operation of my machine. Accordingly the psychrometry of Figure 7 is based on the assumption that the by-pass is open.

And although fast rotation would probably produce more efficient humidication of the incoming air, the same considerations of simplicity of controls lead me to keep the moisture-transferer set for slow rotation throughout the summer.

Before turning to Figure 8, the following preliminary discussion is' in order. lf the speed of rotation of my moisture-transferrer be increased, the amount of temperature fluctuation and vapor-pressure liuctuation of each particle thereof `becomes less and less. Thus the loops abcda, a'b'cda, rnnm, and mnm of Figures 6 and 7, shrink practically to points at about 25 to 30 R. P. M.

The ability of my moisture-transferer to transfer moisture from an air-stream ot' lesser vapor-pressure to an air-stream of greater vapor-pressure is due to iiuctuating the vapor-pressure of each particle (this being in turn accomplished by liuctuating its temperature), so that that vapor-pressure will alternately be higher than that of the high-vapor-pressure air-stream, and lower than that of the loW-vapor-pressure air-stream. It follows that, if these uctuations be reduced by speeding up the rotation, this effect will nally cease, and moisture will ow with, rather than against, the vapor-pressure differential between the two streams. Thus there lies somewhere between my optimum low-speed rotation and my optimum high-speed rotation a critical R. P. M., below which a propertemperature-differential can cause moisture to pass from air-stream to air-stream against the vapor-pressure dilferential, and above which critical R. P. M. moisture must perforce pass from air-stream to air-stream with the vapor-pressure differential. p

Above this critical R. P. M., as rotation is further sped-up, the transfer' of both sensible-heat (in the direction of the temperature differential) and moisture (in the direction of the vapor-pressure differential) becomes more and more e'icient, inasmuch as the increase in the number of rotations per minute more than offsets the decrease in the amount transferred per rotation.

And although the amount of moisture-transfer and sensible-heat transfer per rotation decreases as speed increases, this is more than offset by the increase in the number of rotations per minute.

At about or 30 R. P. M., the transfer of both sensible heat (in the direction of the temperature differential) and moisture (in the direction of the vapor-pressure differential) becomes almost as though there were an intimate direct contact between the two streams, without the intervention of the moisture-transferrer, and yet without either stream mingling with or interfering with the ow of the other.

Faster rotation than this would cause appreciable entrainment of air from one streams to the other, and would also interfere with the air-flow.

Figure 8 is a psychrometric plotting of how, in winter, with the moisture-transferrer rotating at fast speed, in one alternative setting of my machine, the incoming air increases its temperature and moisture content from A to B, by counter-flow exchange with the outgoing air from H to I. L and L represent the practically constant characteristics of two particles, of the outdoorward and indoorward levels respectively of the heatexchanger, at about 25 to 30 R. P. M.

In winter, the by-pass is closed, the furnace is of course on, and one or both of the evaporative pads are operating, depending on the amount of humidification necessary.

And now to briefly summarize the operations of my rotary moisture-transferrer.

It always transfers sensible heat from the air-stream of higher temperature to the air-stream of lower temperature.

As for its transfer of moisture. At optimum low rotation-speed for any given combination of carrier and impregnant, the loops (which plot the characteristics of the particles) are long, and moisture-transfer occurs from the air-stream of lower temperature to the air-stream of higher temperature, regardless of the relative vaporpressures of the two streams. But at high rotation-speeds the loops contract to mere points, and moisture-transfer occurs from the air-stream of greater vapor-pressure to the air-stream of lesser vapor-pressure, regardless of the relative temperatures of the two streams.

Condensation of moisture on the particle from either air-stream, adds sensible-heat to that stream, by increasing the temperature of the particle, and thence increasing the temperature of the air. Evaporation of moisture into either air-stream, subtracts sensible heat from that stream.

The net effect of the above principles results in the three operations discussed in connection with Figures 6, 7, and 8, whereby:

(l) On a warm humid day, changer rotating slowly, the by-pass open, and the furnace on, the incoming air will be dehumidified and somewhat warmed. Other means, described but not claimed herein, are employed to then cool the incoming air.

(2) On a warm dry day, with the moisture-exchanger rotating slowly, the by-pass open, and the furnace off, the incoming air will be humidiled and somewhat cooled. The other means, above mentioned, are then employed to further cool the incoming air.

(3) On a cold day, with the moisture-exchanger rotating rapidly, the by-pass closed, and the furnace on, the incoming air will be humidified and considerably warmed.

And now finally to describe a further improvement of my rotary moisture-exchanger.

The amount of my improved impregnant, or of any other hygroscopic salt or combination of salts, or of one of the glycols, or of any other impregnant, to be used to impregnate excelsior, corrugated asbestos paper, or other air-permeable carrier, is very important.

It would naturally be supposed that the more impregnant the better, up to the maximum which the carrier will hold. It is true that the more impregnant the carrier will hold, the more moisture it can extract from and give up to the air. But it turns out not to be true that we ought to employ the maximum non-water with the moisture-excomponents of the impregnant which the carrier will hold; for I have found that this introduces an undesirable phenomenon, which l shall call weeping: namely, that When my rotary moisture-transferrer stands still in humid atmosphere, a considerable portion of the moisture absorbed by the impregnant will drip away, carrying some of the impregnant with it. Weeping reduces the efficiency of moisture-transfer thereafter, unless and until the lost impregnant is replaced (whereupon the cycle is repeated), or unless a compensating change (to be discussed hereinafter) is made in the apparatus. Weeping is also messy, and subjects the apparatus to corrosion.

However, I have found that weeping can be effectively eliminated, in all atmospheres normally encountered, by reducing the weight of the non-water components of the impregnant carried by a given weight of carrier, to a fraction of the possible maximum which the carrier could carry. This fraction can be determined in either of two ways, and the resulting underimpregnation can be fully compensated for, in either of two ways, all of which will now be described.

Starting with a given carrier and an impregnant of chemical composition definite except as to amount of water-dilution, it is a simple matter to determine by cutand-try the amount of dilution which will cause the carrier to be fully impregnated, which term I define as meaning that the carrier is impregnated by soaking it with the maximum quantity of the impregnant in question which the carrier is capable of holding when the nonwater components of the impregnant are in an aqueous solution of such strength as to be in equilibrium with an atmosphere of 35% relative humidity.

The carrier is then underirnpregnated, according to one or the other ofthe following two formulas:

(l) To a degree of underimpregnation of approximately one-third the amount of non-water components required to fully pregnate; the quoted term meaning the quotient obtained by dividing the actual quantity of the non-water components per pound of `carrier by the quantity per pound of carrier required to fully impregnate the carrier.

(2) To a degree of underimpregnation such that the impregnant solution will be in equilibrium (at equal temperatures) with an atmosphere of relative humidity, or somewhat less.

The two alternative ways of compensating for either of the above methods of underimpregnation are as follows: (l) approximately double the thickness of the wheel; or (2) increase the speed of the wheel, from the optimum slow speed of rotation to a new speed approximately inversely proportionally to the degree of underimpregnation. There is no need to change the optimum fast speed of rotation, which fact renders the second method of compensation preferable to the first method.

Thus far we have considered as a humidity-changer the two air-passages, the fans, and my moisture-transferrer l5, merely per se, or in combination with an air-heater 52 to reactivate the hygroscopic packing in the moisturetransferrer, or in further combination with my by-pass 23. But if we still further add my heat-transferrer 27, and specify slow rotation for my moisture-transferrer, the resulting subcombination of my complete universal airconditioner, can still be regarded as merely a humidity changer. This subcombination without the addition of further features, has many uses, such as in chemical processes or storage rooms, for which dryness rather than coolness is the predomnatng consideration. Or it could be employed as a dehumidifying unit for air-conditioning, the cooling being supplied in some other manner than in my own universal air-conditioner.

Let us, then, consider this subcombination by itself. See Figure l. The air in the upper passage passes to the left through the slowly rotating moisture-transferrer 1S, after which the non-dehumidified or least-dehumidif'ied portion of the air is carried off through by-pass 23. This by-passed air carries all or practically all of the sensible-- heat which has been transferred into the upper air-stream from the lower air-stream by the moisture-transferrer. The non-by-passed dehumidiied air, carrying the heat of condensation resulting from its dehumidification then has practically all of this heat extracted from it by my very efcient heat-transferrer 27, and then passes on for whatever use may have been planned for it.

Meanwhile the air in the lower passage, from whatever source derived', passes in countercurrent toward the' right. First, it extracts from heat-transferrer 27 all the heat which that heat-transferrer extracted froml the upper air. Then it is mingled with the hot air from bypass 23. Then it is further heated by air-heater 52;. Then its heat, acquired in all these ways isutilized to dry the hygroscopic packing of moisture-exchanger 15, and thus reactivate the moisture exchanger.

The chief feature of all this is that we have here an air-drier, in which nearly all the sensible-heat put into my moisture-transferrer is recaptured by means of my bypass, and nearly all the heat of condensation developed by dehumidifying the upper air-stream is recaptured by means of my heat-transferrer, and all this recaptured heat is reused (by preheating the reactivation air) to helpI reactivate my moisture-transferrer, thns greatly reducing the load on my air-heater.

Having now described the various features and components of my invention, I wish it to be understood that my invention isnot to be limited to the specific details herein described.

I claim:

l. A rotatable' cylindrical transferrer for use in an air-conditioning unit, to transfer, between two air-streams passing through said unit, a thermodynamic characteristic of air, said transferrer comprising: a wheel-like casing, having spokes, a hub, and a rim, all imperforate, and all of substantially the same width in an axial direc tion, said spokes dividing the" casing into sectors; a packing of inert air-permeable liquid-absorbing material impregnated with a non-volatile liquid, said packing completely filling each sector, and being packed into each sector with such compactness as to remain freely airpermeable and yet be so self-sustaining as to be substantially immovable with respect toV the casing during the rotation of the casing even in a vertical plane; and means secured to the casing at each face thereof to retain the packing therein.

2. A. transferrer according to claim l, in which the inert material is fllamentous.

3. A trans-ferrer according to claim 2, in which the inert filamentous material is excelsior.

4. A transferrer according to claim l, in which the inert material is corrugated asbestos paper, the corrugations of which run axially of the cylinder.

5. A transferrer according toclaim 1, in which the impregnant is non-hygroscopic.

6. A transferrer according to claim 5, in which the nonhygroscopic impregnant is mineral oil.

7. A transferrer according to claim l, in which the impregnant is hygroscopic.

8`. A transerrer according to claim 7, in which the impregnant is an non-volatile hygroscopic liquid of the group consisting of glycerine and the ethylene glycols.

9. A transferrer according to claim 8, in which the impregnant is triethylene glycol.

10. A transferrer according to claim l, in which the impregnant comprises an hygroscopic salt.

l1. A rotary wheel-shaped moisture-transferrer divided into sectors, each sector being substantially lillecl with a composition of matter, which composition consists of an air-permeable carrier impregnated with a watersolution of an hygroscopic salt which has a degree of underimpregnation of approximately one third.

12. A moisture-transferrer according to claim 11, characterized by the fact that its axial thickness is approximately twice that appropriate to the same carrier fully impregnated with the same impregnant.

13. A rotary wheel-shaped moisture-transferrer, divided into sectors, each sector being substantially filled with a composition of matter, which composition consists of an air-permeable carrier impregnated with a watersolution of an h-ygroscopic substance the quantity of whose non-water components is` less than the minimum quantity of such components which, in water-solution in equilibrium with an atmosphere at 90% relative humidity, would produce weeping from the carrier.

14. A moisture-transferrer according to claim 13, characterized by the fact that its axial thickness is materially less than the product of the reciprocal of the degree ot underimpregnation by the' axial thickness appropriate to the same carrier fully impregnated with the same impregnant.

15. A moisture-transferrer according to claim 1l, combined with means for rotating it at approximately three times the angular velocity optimum for the same carrier fully impregnated with the same irnpregnant for moisture transfer from the cooler to the warmer of two air-streams across which the moisture-transferrer is rotated.,

16. A moisture-transferrer according to claim 13, corn-- bined with means for rotating it at an angular velocity which is approximately the product of the reciprocal of the degree of underimpregnation by the angular velocity optimum for the same carrier fully impregnated with the same impregnant for moisture transfer from the cooler to the warmer of two air-streams across which the moisture transferrer is rotated.

17. A moisture-transferrer according to claim 13, combined with means for rotating it at an angular velocity materially faster than the angular velocity optimum for the same carrier fully impregnated with the same impregnant for moisture transfer from the cooler to the warmer of two air-streams across which the moisturetransferrer is rotated.-

18. A rotatable cylindrical transferrer for use in an air-conditioning unit, to transfer, between two air-streams passing through said unit, a thermodynamic characteristic of air, said transferrer comprising: a wheel-like casing, having spokes, a hub, and a rim, all imperforate, and all of substantially the same width in an axial direetion, said spokes dividing the casing into sectors; a packing of inert air-permeable liquid-absorbing material capable of being impregnated with a non-volatile liquid, said packing completely filling each sector, and being packed into each Sector with such compactness as to remain freely air-permeable and yet be so self-sustaining as to be substantially immovable with respect to the casing during the rotation of the casing even in a vertical plane; and means secured to the casing at each face thereof to retain the packing therein.

19. A transferrer according to claim 18, in which the inert material is lamentous.

20. A transferrer according to claim 19, in which the inert lilamentous material is excelsior.

21. A transferrer according to claim 1S, in which the inert material is corrugated asbestos paper, the corrugations of which run axially of the cylinder.

References Cited in the tile of this patent UNITED STATES PATENTS 1,206,977 Batter Dec. 5, 1916 1,481,221 Nuss Jan. 15, 1924 1,541,147 Ikeda et al. lune 9, 1925 1,583,238 Saidder May 4, 1926 1,912,784 Miller et al. June 6, 1933 y2,223,227 Robic Nov. 26, 1940 2,227,773 Warren Jan. 7, 1941 2,302,807 Shoeld Nov. 24, 1942 2,503,523 Stuart Apr. l1, 1950

Images