CA1126570A - Process of producing a spray-dried agglomerated product - Google Patents

Abstract

ABSTRACT
A final spray-dried product having the appearance of an agglomerated product is produced by utilizing a two-fluid nozzle for mixing an inert gas with an extract, solution or sus-pension of a food product whereby on subsequent atomization and partial degasification in a conventional spray dryer the drop-lets are fused together in an oscillating pattern within a region of spray turbulence.

Description

~ ~ Q ~ ~ 7 ~

This invention relates to a process for producing a spray-dried agglomerated product. Typical of the products which may be produced by the process of the instant invention include dry soluble coffee powders, bulked dextrins and the like.
Spray drying and agglomerating are well-known unit processes that have been employed in the food processing industry for some time. Generally, however, spray drying and agglomerating are separate processes which are carried out in stepwise fashion on coffee extracts, liquid milks and the like.
It is apparent that such separate stepwise processes leave much to be desired in terms of plant economies, utilization of equip-ment and throughput times. It would be desirous-j therefore, if a combined spray drying and agglomerating process were able to be devised so as to increase the production capabilities of a food processing plant and at the same time to reduce the amount of equipment and throughput times that might otherwise be re-quired, The concept of a combined spray-drying/agglomerating process has heretofore been described in several U. S. patents, namely, U.S. Patent No. 3,514,300 to Mishkin et al. and U. S.
Patent No. 3,151,984 to Peebles et al. In the Mishkin et al.
process, recycled fines are employed in the production of an àgglomerated soluble coffee powder, In the process described by Peebles et al., milk concentrate with added lactose crystals is introduced into a spray dryer and with proper control of drying conditions is discharged as an aggregated material having a moisture content of 10 to 20%, A second drying operation is then employed in the Peebles et al, process to further reduce the moisture content. In both the Mishkin et al, and Peebles et al.
processes, however, it appears that two separate drying steps are

- 2 - ~

~65t~

required: in Mishkin et al. a drying operation for the production of coffee fines is necessary while in Peebles et al.
not only are lactose crys-tals an essential component for intro-duction into the spray dryer but, as mentioned, to obtain the final milk powder product a second drying operation is required to reduce the moisture content of the aggregated material to the final moisture content of the milk powder.
The concept of foaming coffee extracts and other ex-tracts and suspensions of food products to control final product color, density and particle size has also been described hereto-fore. Thus, U.S. Patent No. 2,788,276 to Reich et al. teaches a process for spray-drying a foamed material such as coffee. How-ever, the objective referred to in the patent is that of produc-ing a product with discrete spherical structures which are not clumped, aggregated or otherwise agglomerated.
We have now discovered that a final spray-dried pro-duct having the appearance of an agglomerated product may be produced by employing a two-fluid nozzle for mixing an inert gas with an extract, solution or suspension of a food product and varying the differential pressures of the extract, solution or suspension whereby droplets are subsequently fused together in an oscillating pattern within a region of spray turbulence in a conventional spray dryer. We have found that coffee extract as well as dextrin solutions and other extracts, solutions or sus-pensions of various food products can be spray dried and agglomerated simultaneously in one processing step. The process is capable of producing agglomerated products having unique physical appearance as well as a range of colour attributes which may be desired.

~;S~

The process has the advantage that it eliminates the necessity of incorporating standard agglomerating equipment such as agglomerating towers in a soluble coffee processing unit.
Moreover, the process allows for an increase in the capacity of - a soluble coffee processing unit by permitting the conversion of an existing agglomerating tower into a spray agglomeration tower.
The process of the present invention comprises inject-ing an inert gas such as nitrogen, carbon dioxide or the like into an extract, solution or suspension of a food material typically coffee extract or a dextrin solution utili~ing a two-fluid nozzle and the subsequent atomization and partial degasification of the extract, solution or suspension utilizing a conventional spray nozzle accompanied by a spray pattern oscillation due to unequal extract, solution or suspension and gas pressures so as to produce a spray dried/agglomerated product.
The invention will be better understood by reference to the following detailed description, examples and the accom-panying drawings.
In the drawings, Figure 1 is a flow diagram of thespray drying agglomeration process of the present invention as specifically related to the processing of a coffee extract.
Figures 2 to 5 are photomicrographs of soluble coffee powder spray agglomerates produced in accordance with the pro-cess of the present invention and conventional soluble coffee products which have been spray dried and also agglomerated.
Figures 2 and 3 illustrate spray agglomerates produced in accordance with the process of the present invention. Figure 4 illustrates a spray dried soluble coffee powder produced by a conventional commercial process. Figure 5 shows an agglomerated soluble coffee product prepared on conventional commercial equipment utilizing the spray dried soluble coffee powder illustrated in Figure 4.
The total magnification is 61~ for the photomicro-graphs shown as Figures 2 to 5. The relative particle size and shape of the product shown in Figure 2 may be contrasted with that shown in Figure 4 and the relative particle siæe and shape of the product shown in Figure 3 may be contrasted with that shown in Figure 5. Further, the unique physical form of products obtainable by the process of this invention may be noted from Figures 2 and 3.
Figures 6, 7 and 8 are photomicrographs of several dextrin products including a spray agglomerated product prepared in accordance with the process of this invention. Figure 6 illustrates a commercially available bulked dextrin product.
Figure 7 illustrates a spray dried agglomerated dextrin product prepared in accordance with the process of this invention and Figure 8 illustrates a drum dried dextrin product. The magnifi-cation in these photomicrographs is lOOX.
Referring to Figure l, coffee extract from a percola-tor set and evaporated to 25-35% concentration is fed by means of a high pressure air piston pump (supply on demand) to the liquid intake of a two-fluid nozzle where an inert gas such as nitrogen is intimately mixed with the coffee extract to produce a foamed extract. The motive force from the air piston pump is used to move the foamed extract to a conventional high pressure core type nozzle in the spray dryer. Check valves are position-ed both between the air piston pump and the two-fluid nozzle and between the inert gas supply and the two-fluid nozzle. The extract and gas pressures at the two-fluid nozzle are balanced such that when the piston pump provides the maximum extract pressure to the two-fluid nozzle, the extract pressure is greater than the air pressure and similarly when the pump pressure is at its low point, the gas pressure is equal to or greater than the liquid extract pressure to the two-fluid nozzle. This variation in gas and liquid pressures results in an oscillation of the nozzle pattern in the spray dryer as well as a variation in the amount of gas mixed with the extract. This pulsating, although at pressures above the minimum atomizing pressures, replicates the partial agglomeration which is associated with operating pressure spray nozzles at the minimum atomizing pressures where-by a small drop in nozzle pressure would result in no atomiza-tion to occur. The frequency of nozzle spray pattern pulsation can be controlled by the relative sizing of the feed pump and spray nozzle within the spray dryer. Spray particle collision and subsequent fusion can be ensured by more frequent oscilla-tion of the spray pattern. The pump pressure and consequentlythe extract pressure may vary substantially with a minimum pressure variation of 10 to 15 psig required. Although the most efficient agglomeration occurs when the gas and extract pressures are almost equal, some agglomeration will occur at lower gas to extract pressures, the degree of agglomeration being reduced with the lower gas to extract pressures. The agglomeration will occur at both low and high spray nozzle pressures as the critical parameter is the extract to gas ratios expressed in terms of liquid and gas pressures to the two-fluid 65~YO

nozæle. Thus the operating range for spray dryer nozzle pressures is determined by desired feed rate to the spray dryer.
This agglomeration process may be applied to other food products in addition to coffee extract such as dextrins. If desired, the spray dried product may be after dried to obtain products with lower moisture contents.
The following are detailed illustrative but nonlimit-ing examples of this process:

Coffee extract at 35% concentration and 98F. is passed by means of a Graco air piston pump through a Spraying Systems two-fluid pneumatic mixing nozzle (Type 40100/1253283.
The extract and nitrogen gas pressures at the mixing nozzle were 100 psig and 98 psig respectively. A Spraying Systems (Whirljet Type 1-1) nozzle was utilized in the spray dryer whereby an average foamed extract pressure of 95 psig was maintained on this nozzle. The spray dryer inlet and outlet temperatures were 430F. and 268F. respectively. The product produced had a mean particle size of 834 microns (contrasted with the particle size for a control of 283 microns), a bulk density of 13.8 gms/100 cc. and a final product moisture of 1.1%. No change in final product flavours is apparently detect-able.

Coffee extract at 40% concentration and 70F. is passed by means of a Graco* air piston pump through a Spraying Systems^
two-fluid pneumatic mixing nozzle (Type 100150/189251), The extract and nitrogen gas pressures at the two-fluid mixing nozzle were 180 to 300 psig and 200 psig respectively. An Trademark ' .. .

.

6S'7(~

average nozzle pressure of 250 psig was maintained on a Spraying Systems (Whirljet Type 2-2) nozzle in the spray dryer. The spray dryer inlet temperature was 360F. and the outlet tempera-ture was 260F. The bulk product mean particle size was 640 microns, the density was 19.1 gms/100 cc. and the final product moisture was 3.2%.

A dextrin (Morex 1918) at 60% concentration and 190F. is pumped by means o~ a Graco air piston pump through a Spraying Systems* two-fluid mixing nozzle (Type 100150/189351).
The liquid and nitrogen gas pressures at the two-fluid nozzle were 700 psig and 300 psig respectively. A Manton-Gaulin positive displacement pump was then used to produce a higher nozzle pressure of 1100 psig at a Spraying Systems nozzle (Whirljet Type 1-1) in the spray dryer. The inlet spray dryer temperature was 560F. and the outlet temperature was 270F. A
feed rate of 300 to 350 lbs./solids per hour was obtained. The - bulk product density was 4.6 gms/100 cc.
While no particular theory is advanced for the results achieved by the spray agglomeration process of this invention it appears every pressure nozzle employed for spray drying has a minimum threshold pressure which is required for proper atomiza-tion to occur. By operating in this narrow band of threshold pre~sures, large spray droplets can be formed but control in operations is difficult and varying spray patterns are obtained.
By utilizing two-fluid nozzle pneumatic mixing as a means of aerating the feed material to a conventional spray nozzle above the threshold pressure:

Trademark 13.;~;5 ~C) .

(I) No extremely wide fluctuations in spray pattern will occur.
(II) Fine control on ratio of extract to inert gas which controls final product density can be obtained.

With the spray agglomeration process of this invention it is thus possible to produce a spray dried agglomerated product, i.e., soluble coffee, and dextrin having a new physical form, i.e., large fused spherical buds, employing high inert gas to extract, solution or suspension ratios.

,. 1126St70 SUPPLEMENTARY DISCLOSURE

It is known in the art of spray drying that, for a particular spray nozzle system, optimum atomization occurs when the liquid flow exiting from the nozzle is of sufficient magni-tude to cause the nozzle to deliver the spray over a wide area in the form of small droplets (fine spray). Conventionally, high pressure positive displacement pumps are employed to force the liquid through the spray nozzle at the desired, substantially uniform, flow rate to form a spray pattern which conforms to the structural dimensions of the drying tower and which has drop size uniformity which conforms to the drying capability of the tower.
As is well known, reciprocating positive displace-ment pumps develop a fluctuating pressure and fluctuating flow rate of discharge liquid. The fluctuating flow rate of liquid through the spray nozzle creates fluctuations in the spray pattern (conical angle of discharge) and also non-uniformity of drop size in the spray. For the most part, in conventional spray drying practice, the magnitude of these fluctuations is minimal, and creates a change in conical spray angle of about 3-4. Conven-tionally, operations are conducted with systems which developsubstantially uniform flow rates. In many instances, a gear-type positive displacement or multi-piston pump is employed to assure the desired unfiromity of flow rate at the high pressures.
Also, in many installations a dome-type accumulator is located between the putnp discharye and the spray nozzle to even out the flow rate to the nozzle.
For some liquids, namely cottage cheese whey, fluct-uations in flow rate through the spray nozzle (commonly called "slugging) when using a high pressure reciprocating pump has been minimized by the injection of gas into the whey in the system between the pump and the spray nozzle (cf. Hanrahan, U.S. 3,222,193)-. .,. 10 a - :llZ65~0 For a particular spray drying system, the flow rate through the spray nozzle should be of a value for the nozzle to deliver a spray of desired atomization without endangering the operation by 05 creating too wide a spray angle and thus wetting the side walls of the drying tower. Too low a flow rate will create little or no break-up of the liquid into drops or will develop drops of such large size they will be incompletely dried in the tower.

In contrast to the teachings and efforts of the prior art, the method of spray drying of the instant invention purposely utilizes flow rate fluctuations of considerable magnitude of coffee extract liquid through the spray tower spray nozzle to successfully develop a physically unique soluble coffee agglomerate product. The flow rate fluctuations of coffee extract through the spray nozzle are controlled to uniformly cycle (sine wave) between set maximum and minimum flow rate values.
Concurrently, flow rates of inert gas, such as N2, through the spray nozzle are controlled to uniformly (sine wave) cycle between set maximum and minimum values. The liquid coffee extract and gas flow rate cycles are controlled to be directly out of phase (180 sine wave curve out of phase).
The liquid coffee extract maximum flow rate through the spray noæzle causes the nozzle to develop - a spray pattern of a maximum area of coverage; i.e., maximum conical angle of spray of fine droplets just short of wetting the side walls of the drying tower.
The liquid coffee extract minimum flow rate causes the spray nozzle to develop a spray having a signficantly smaller conical spray angle and to , a .

~Z65~7~

develop drops of a much larger size than those developed at the maximum flow rate. The large drops are slower to dry than ~he small drops from the wide spray angle.
05 The concurrent cyclic inert gas flow (directly out of phase with the liquid flow rate through the respective spray nozzle) is at the highest flow rate when the largest coffee extract drops are formed and sets up (along with the tower drying air) a turbulence which, accompanied by the cyclic collapse and expansion of the spray angle (spray pattern) causes the relatively large number of fine particles from the rapidly dried small droplets to impact the fewer, large, partially dried, particles produced from the slower lS drying large drops ~o form agglomerates of coffee particles. The agglomerates are unique in that they are essentially comprised of several relatively large, substantially spherical, particles of dried soluble coffee partially embedded, or otherwise adhered to a large pa~ticle, or particles, of partially dried soluble coffee (cf. Figures 2 and

3). These agglomerates differ in appearance from conventional soluble coffee agglomerates which are (usually) formed from uniformly sized wetted small particles (cf. Figures 7 and 8).
The critical aspect of the invention and that which leads to the production of the desired unique agglornerates is the capability of the system to rapidly cycle the fluctuating flow of the liquid coffee extract and the inert gas stream (opposed cycle) through controlled maximum and minimum flow rates. A two-fluid venturi flow regulator (flowrator) is successfully used for this critical phase of the method of the invention.
C

,) ,l.Z~SqO

As the concept depends for its ultimate success on having large drops of slowly drying liquid present to impact with the smaller dried or partially dried particles, a conventional afterdrier may be employed 05 to finish dry the agglomerates to a desired uniform dryness.
The process has the advantage that it eliminates the necessity of incorporating standard agglomerating e4uipment such as agglolnerating towers in a soluble coffee processing unit. Moreover, the process reduces the detrimental heating effects on the flavor quality of the soluble coffee as compared with the stepwise process of spray drying followed by agglomerating and allows for an increase in the capacity of a soluble coffee processing unit by permitting the conversion of an existing agglomerating tower into a spray agglomeration tower.

; Brief Description of the Drawings A more complete understanding of the method of the invention may be obtained by reference to the following description and claims, taken together with the following drawings in which:
- Fig. 9 is a schematic sketch of the arrangement of the principal equipment components 2S employed, including the coffee extract high pressure reciprocating pump, the inert gas supply source, the two-fluid venturi flowrator and the drying tower spray nozzle, and illustrates the flow path of each of the two fluids.
Fig. 10 is a view, in partial section, of the two~fluid venturi flowrator.
Fig. 11 is a sectional view of the spray ~, nozzle and shows the developed spray angle and ~ pattern of the exiting coffee extract at maximum : 35 flow rate.

. . , . ..... , . . , ., . .. . .. , , ~, .

~' ¢' ,' ' ..

65~7~

Fig. 12 is a partial sectional view of the same spray nozzle as shown in ~ig. 3 and shows the developed spray angle and pattern of the exiting coffee extract at minimum flow rate.

DETAILED DESCRIPTION OF T~IE D_AWINGS
Referring to Flg. 9, liquid coffee extract is fed by means of a high pressure, single cylinder air driven piston pump (1) (supply on demand) to the liquid intake (12) of a two-fluid flowrator (2). An inert gas supply source (3) under high pressure is fed to the gas intake (14) of the two-fluid flowrator (2). The gas pressure is adjusted to remain substantially constant at a set value by adjustment of valve (4) and pressure gage (5). The flow rate of coffee extract is partially controlly (see below) by adjustment of the air pressure (6) of the air driving the pump and, optionally, the coffee ' extract flow control valve (7). The Elow ra-te .C~

variation is indicated by the change in pressure at pressure gage (8) in the conduit to the two-fluid flowrator. The motive force from the air piston pump and the inert gas supply pressure is used to 05 move the mixture of extract and inert gas through a conventional high pressure core-type spray nozzle (9) in the drying tower (10). The combined coffee extract and gas pressure at the intake of the spray noz~le is mc~sured by th~ prcssure g~c (11) The difference in pressure readings at gages (8) and (11) represents the liquid coffee extract pressure differential (~PL) across the two-fluid flowrator (2) and the inert gas differential pressure (~PG~
across the two-fluid flowrator is calculated from the difference in pressure readings at gages (11) and (5~. -The two-fluid flowrator (Fig. 1~ is of the venturi-type and is (or is equivalent to) an assembled unit manufactured by Spraying Systems Co., Wheaton, Illinois as shown and described on pages 47 et seq of Industrial Catalog 27 (1978). As shown in Fig. lo, the coffee extract enters at port (12) and passes through a constricting passageway, "extract nozzle"
(13). The inert gas (N2) enters at port (14) and passes through a constricting passageway, "inert gas nozzle" (15). Both "passageways" are of the type to constrict flow such that the velocity (at constant pressure) of both fluids increases at the points (orifices) of outlet (16) and the venturi effect of . each influences the flow rate of the other.
Thus, the two-fluid flowrator not only controls the flow rate of both fluids (coffee extract and inert gas) but also affects the flow rate of one fluid with respect to the flow rate of the other.
Thus, even though the pressure (5) of the inert gas ,~ , ., .. , . . . . . ,, . . ~

F~
' ~ i ~
- 1~2i570 to the flowrator intake remains constant, a change in coffeé extract fluid pressure (8) (increase) at the flowrator intake port not only causes a controlled increase in extract flow but also provides a smaller 05 than expected (absent the flowrator) decrease in inert gas flow due to the partly compensating vent`uri effect. Similarly, a decrease in flow rate of extract through the flowrator creates a slight increase in inert gas flow even though the upstream gas pressure remains substantially constant and may be at a lower intake pressure than tha~ of the extract.
As will be understood by those knowledgeable in the art of fluid flow, it is the pressure differential (~PL) across the two-fluid flowrator coffee extract passage as measured by gage readings (8)-(11) which influences the coffee extract flow rate and similarly, it is the pressure differential (~PG) across the two-fluid flowrator inert gas passage as measured by gage readings (5)-(11) which influences .the flow rate of inert gas. In addition, the venturi effect in the two-fluid flowrator developed by each fluid further influences (partly compensates the reduction in flow when the ~P is reduced) the flow rate of the other fluid. The flowrator is designed such that the venturi effect of the coffee extract flow influences the inert gas flow to a greater degree than the gas flow venturi effect has on the extract flow rate.
As stated previously, the combined coffee extract and inert gas pressure at the intake (11) to the spray nozzle drives the mixture through the nozzle and atomizes the mixture in the tower. By holding the inert gas pressure constan~ (substantially) at (5), the single cylinder pump (6) fluctuating ~16-.. .. . ... . .. . .. .

f~

t~
~lZ6570 pressure at (8) is changed by the restrictive orifices of the flowrator to a substantially co~stant pressure at (11) even though the flow rate of coffee extract fluctuate B .
05 In summary, the spray nozzle flow rate requirements conform to the drying tower structural dimensions and drying capability. The flowrator is sized to provide the desired maximum and minimum coffee exLract u~d lner~ flow raLe~ ~o conforlll wi~h the spray nozzle atomization requirements (conical spray angles and drop sizes). The coffee extract pump is oversized and would deliver vastly larger amounts of extract to the spray nozzle if the flowrator were not in the system. The flowrator reduces the overall flow rate of both coffee extract and inert gas. The - flowrator, however, causes wide fluctuations in the overall reduced flow rate of coffee extract when the extract pressure differential across its intake and the intake to the spray nozzle are slightly altered.
The venturi effect of the flowrator causes less : magnitude of fluctuation of the inert gas flow (the venturi effect partly compensates for the slight increase in overall pressure at the entrance to the spray nozzle when the extract is at its maximum flow). The maximum flow of extract, accompanied by the minimum flow of inert gas (and vice versa), maintain a substantially constant pressure at the intake to the spray nozzle -- it is the change in flow rate of the liquid coffee extract which has the ,~ 30 far greater effect on the angle of spray delivery and drop size.
Thus, for a flowrator of the types described above, the following changes (fluctuations) in liquid coffee extract and inert gas (N2) flow rates to a spray nozzle located in a drying tower are obtainable.

.. , . .................... . . . .. ., .. , .. .. .. _ _, . . _ _ '" ~C

~lZ6570 ;

TABLE I
Pressure differential (~P) across the two-fluid flowrator 05 Extract Inert Gas Inert ~as Extrac~ Flow Rate Z Increase Flow Rate % Decrease - (lb/in') (lb/in~) (lb/hr) S.CF/min.
31 20 35 --- 2.8 ----31 30 79 126 2.5 10.4 31 40 140 77 2.2 12.0 31 60 250 150 2.0 9.1 _. ..
--- 3.1 ----66 164 2.9 6.5 100 52 2.6 10.3 220 120 2.4 ~.6 --- 2.4 ----105 91 2.1 12.5 185 76 2.0 4.7 300 62 1.9 5.0 From Table I it can be observed that, at a constant inert gas pressure differential (~PG) of 31 lb/in2, a drastic increase (150%) in flow rate of coffee extract from the flowrator is obtained if the single piston pump on its pressure stroke increases the pressure of the liquid intake to the flowrator to increase the coffee extract differential pressure (~PL) across the flowrator from 40 lb/in2 to 60 "
lb/in~ while, concurrently, the inert gas flow is limited to a 9% reduction by the venturi effect of the flowrator.
Without the installation of the venturi-type flowrator in the system and for the same inert gas pressure differential and change in coffee extract pressure differential, the flow rate of the coffee extract is increased only about 20% and the inert gas flow (at 31 lb/in2) is reduced to nil.

. . . , _,, _ ., .

~ C~
.
, l~Z~i570 Consequently, the venturi-type flowrator provides for large variations in coffee extract flow to the spray nozzle with concomitant small variations in inert gas flow even though the coffee extract liquid 05 pressure to the intake of the flowrator may be greater than that of the inert gas pressure.
Employing similar flowrators but of different sizes will, of course, permit larger (or smaller) flow values for the two fluids but will have similar affects on the flow rate fluctuations.
Thus, although the flow of coffee extract will fluctuate throughout a wide range and cause the spray n~zzle to alternate the spray pattern from maximum to minimum spray angles, the inert gas flow is always present to create the desired turbulence within the spray patterns.
At the maximum coffee e~tract flow rate through the spray nozzle, the spray angle of droplet discharge is at its maximum (cf. Fig. 3) and the droplet size is minim~M. Conversely, the minimum flow of extract through the spray nozzle yields drops of the largest size and the smallest spray angle pattern, ~Fig. 4).
Also, the inert gas flow rate is at its peak when the drops are of the largest size (when the highest degree of turbulence is desired) and at its lowest rate when the lighter, smallest drops are being sprayed.
The controlled alternating expansion and collapse of the spray pattern, within the required limits 30 dictatèd by the drying tower dimensions and atomization energy needs, plus the ever present, but varying inert gas flow rate, create the turbulent conditions for the production of the agglomerated coffee product.
The more rapidly drying smaller coffee particles impact the slower drying wetter larger particles to __ .

~ . .

iS70 create agglomerates which are comprised of small spherical particles fixed to a larger particle (or particles) of partially dried soluble cofee. The agglomerate is then dried to a desired degree as it 05 falls to the base of the tower.
Figures 2 and 3 illustrate spray agglomerates produced in accordance with the process of the present invention. Figure 4 illustrates a spray dried soluble coffee powder produced by a conventional commercial process. Figure 5 shows an agglomerated soluble coffee product prepared on conventional commercial equipment utilizing the spray dried soluble coffee powder illustrated in Figure 4.
The total magnification is 61X for the photomicrographs shown as Figures 2 to 5. The relative particle size and shape of the particles shown in Figure 2 may be contrasted with those shown in Figure 4 and the relative particle size and shape of the product shown in Figure 3 may be contrasted with that shown in Figure 5. Further, the unique physical form of the particles and the agglomerates obtainable by the process of this invention may be noted from Figures 2 and 3, Coffee extract at 34% concentration and at 98F
was passed by means of a air piston pump (Graco Model No. 206842) manufactured by Gray Company, Inc., Minneapolis, Minnesota) through a Spraying Systems Co. two fluid venturi-type flowrator device (set-up No. 23B. cf. page 50 of Industrial Catalog No. 27, 1978). The valve on the flowrator was adjusted for full flow. Nitrogen was concurrently passed through the flowrator.

'~
~, ! I ~ i llZ6570 The extract and nitrogen gas pressures at the intake of the two-fluid flowrator were 160-170 psig and 120 psig, respectively. The discharge from~the two-fluid flowrator was then passed through a t 05 conventional spray nozzle (Whirljet Type 1-1 Spraying Systems, Co., Wheaton, Illinois) in the spray dryer where the nozzle pressure was maintained at 126 psig (substantially constant). The Graco~ air piston pump utilized to deliver the extract pressure to the 10 two-fluid venturi-type flowrator was also utilized as the motive force to provide the required hydraulic Ir pressure to the spray nozzle.
The degree and frequency to which the extract ' pressure varied was a function of the relative size 15 of the feed pump to both the two-fluid flowrator and i;
the spray nozzle and, also, the feed rate required at the spray nozzle. In this particular example with an average feed rate of 142 lbs/hr of extract the spray nozzle pattern oscillated once every 10 20 seconds. At the low end of the piston stroke (suction~
the extract pressure was about 10 lb/in2 lower than at the high end of the piston stroke resulting in a significantly reduced (30%) liquid feed rate at this point. The following.data were obtained:
TABLE II
Extract Extract Flow N2 Flow Particle Pressure Rate Rate (Drop) At Flow- ThruN Pressure Thru Spray Nozzle Size - 30 'rator SprayA~ Flowrator Spray Discharge Distri- `
Intak~ _Nozzle_ Intak~ Nozzle Spray Angle bution ~lbs/in ) (lbs/hr.) (lbs/in ) SCF/min (0) (~) 160 110 120 2.4 40-45 300 -1000 Avg.- 700 35 170 170 120 2.1 60-70 150 - 500 , Avg.-400 _ _ .. .. ,, .,, , , . ,,, , , . _ . .. . . ... _ . _ ., _ . _ _ . _ _ 11 ~ 6~ 7 ~

Atomization ranged from considered normal at the highest extract flow rate to marginal at the lowest extract flow rate.
The spray dryer inlet air temperature was 05 maintained at 415~F; the agglomerated product had a bulk density of 12.3 grns/lO0 cc. The color of the soluble coffee agglomerates was 33.0 photo units (as measured by a Lumetron Photovoltmeter; (cf. U.S.
Patent No.3,~2/~3v) and a u~ean particle size of 735 ~ with a standard deviation of 300 ~. ~or this particular example, the product moisture content out of the spray dryer was 5-5.5%. A conventional after-dryer could ~e employed to reduce the moisture to any desired lower value. The example product contained a relatively small amount (10-25% by wgt.) of satellite particles (fines) but these particles were larger than those satellites produced in conventional spray drying. A microscopic inspection of the agglomerates revealed small hollow spherical particles adhered to one or more larger particles, also, in most cases to have hollow centers. Fig. 7 and 8 are photomicrographs of the product of this example. As with typical spray drying of soluble coffee, higher extract concentrations would produce denser particles (and agglomerates) with the particular system described in this example.

Soluble coffee agglomerates were produced accordin~ to the method of the invention with processing conditions similar to those in Example 4 but employing a different pump, two fluid venturi-type flow regulator, spray nozzle, and a larger spray drying tower.

1~
;

llZ6570 Coffee extract at 34h concentration and at 97F
was delivered by a Graco~ Model No. 206421 air piston pump through a Spraying System Co. two fluid venturi-type flow regulator (flowrator) set-up 05 No. 42 (cf. page 48 of Industrial Catalog No. 27, 1978) to a Whirljet Type 5-6 spray nozzle. Nitrogen gas was concurrently forced through the flowrator and spray nozzle.
The pressure at the intake to the spray nozzle was maintained at 135 lb/in2 (substantially constant).
For this run, with an average feed rate of 663 lbs/hr, the air driven single cylinder pump created a spray nozzle pattern oscillation once every 3 seconds. At the low end of the piston stroke (suction) the extract pressure was approximately 78 lbs/in~
lower than at the high (top) of the piston stroke.
The following data were obtained:

TABLE III
Extract 20 Extract Flow N Flow Particle Pressure Xate R2te ~Drop) At Flow- Thru N Pressure Thru Spray Nozzle Size rator Spray A~ Flowrator Spray Discharge Distri-Intak~ N~zzle Intak~ Nozzle Sprav Angle bution 25 (lbs/in') (lbs/hr.) (lbs/in') SCF/min (0) t~) 156 241 160 3.0 40-45 300-1000 Avg. 700 2341085 160 2.4 60-70 120-200 Avg. 400 . Again, atomization ranged from considered normal at the highest extract flow rate to marginal at the lowest extract flow rate.
Spray dryer conditions in the larger tower were maintained to dry the agglomerated product to about 5% moisture. The physical properties of the soluble coffee agglomerated product were similar to those of the product obtained in Example 4.

, _ .. . . .... ... .

!

Claims (11)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. In a process for producing a spray dried food product by spray drying an extract, solution or suspension of a food product selected from the group consisting of coffee, dextrin and combinations thereof into which an inert gas has been admitted to foam the extract, solution or suspension prior to spray drying, the improvement of passing the extract, solution or suspension through a two-fluid nozzle at a pressure which varies from higher than to lower than the pressure of the inert gas and subsequently atomizing and partially degasifying the extract, solution or suspension through a nozzle in an oscillating pattern in a spray drying chamber and recovering a spray dried food product having the appearance of an agglomerated food product.

2. A process as in Claim 1 in which the inert gas is nitrogen.

CLAIMS SUPPORTED BY THE SUPPLEMENTARY DISCLOSURE

3. In a process for producing a spray dried agglomerated soluble coffee product wherein an aqueous soluble coffee extract is caused to cycle between selected process maximum and minimum flow rates through a drying tower spray nozzle, the improvement which consists of passing the extract concomitantly with an inert gas through a two-fluid venturi-type flow regulator to regulate the cyclic flow rates of both fluids from their respective supply sources to the spray nozzle.

4. The process of Claim 3 wherein the inert gas is nitrogen.

5. The process of Claim 3 wherein the aqueous extract contains about 35% by weight of soluble coffee solids.

6. The process of Claim 3 wherein the cyclic period of flow rate fluctuation for both fluids is substantially uniform and has a period ranging from about 3 seconds to about 10 seconds.

7. The process of Claim 3 wherein the flow rate of the coffee extract through the flow regulator is caused to uniformly cycle from maximum flow to minimum flow and return to maximum, the inert gas flow is caused to concurrently and uniformly inversely cycle from minimum to maximum to minimum flow.

8. The process of Claim 3 wherein the magnitude of change in flow rate of the aqueous coffee extract through the flow regulator and spray nozzle is greater than the magnitude in change of the concomitant flow rate of the inert gas through the two devices.

9. The process of Claim 3 wherein the two fluid flow regulator regulates with restricting flow orifices the flow rates of both the aqueous coffee extract and the inert gas which, in turn, cause venturi effects to partially compensate the flow restriction effects.

10. The process of Claim 3 wherein the maximum flow rate of extract is that flow rate which exits from the spray nozzle in the form of a spray having a conical spray angle ranging from about 60° to about 70°.

11. The process of Claim 3 wherein the minimum flow rate of extract is that flow rate which exits from the spray nozzle in the form of a spray having a conical spray angle which ranges from about 40° to about 45°.