1. INTRODUCTION:
Since the move from hand cutting and individual weighing of dough pieces to mechanical dividing methods, the configuration and ultimate efficiency of yeast raised production lines has been driven primarily by the then available divider technology.  Thus each advance in divider technology has had a significant impact on the entire production process.

In the following paper I will identify the key functions required of all dividers, review the development history of the more popular dividers and touch on what may be the next step in advancing the technology.

2. ADVENT OF MECHANICAL DIVIDING:
During the last decade of the 19-century the first mechanical dividers went into commercial production in this country.  The successors of these units, which I refer to as ram and shear dividers, are still in wide spread use today.

3. KEY MECHANICAL DIVIDING CONSIDERATIONS:
With the advent of this technology dough pieces were no longer individually weighed as part of the dividing process but instead a specific and consistent volume of dough was repeatedly delivered.  Thus the specific weight of the dough becomes a primary consideration and the entrapment of gases a key concern.  Therefore successful dividers have to deal with degassing capability along with the rheological impact of the particular technology.

Fortunately for the divider designers, almost all yeast raised doughs have a similar mixer density running approximately 69 lbs./foot cubed, with mixer density being defined as the specific gravity of the dough mass at the moment it is ejected from the mixer.

This reference density can be almost duplicated throughout the first 15 to 20 minutes of the life of the dough by applying appropriate mechanical efforts to break the gas being formed out of the gas cells and give the resulting bubbles being formed a way out of the dough mass.

Understand that the dough is continuously expanding on its way to a specific weight of around 35 lbs./foot cubed, but that through appropriate mechanical action can be brought back to mixer density without altering the rheology of the product.  Thus the degassing properties of dividers has always been a key element in their success and, throughout the last 110 years of mechanical dividing has become more and more important as dough gassing rates have increased, and weight and measure laws and best practices have become more exacting.

4. KEY ELEMENTS OF MECHANICAL DIVIDING:
There are five basic elements in successfully dividing dough using volume as the controlling item to establish weight.  They are as follows:

a. Degas
b. Isolate and pressurize
c. Meter (establish a volumetric measurement)
d. Sever
e. Remove

While the mechanical action of the earliest ram and shears and those dividers that followed up to the mid-sixties all had a degassing effect at some level, it was typical, and still is in many cases, to have to continuously hand weigh divided pieces and adjust the volume being delivered say every minute or so.  With the advent of dough pumps, mainly used on soft roll lines, the delivered dough was typically degassed and divider adjusting was reduced to either a few times for each dough, or no adjustment at all.

All dividers generally proceed by isolating a quantity of dough and then forcing the dough, under pressure into a volume measuring system.  Of interest is the fact that no matter which dividing method is used a given dough has a specific pressure threshold.  When this pressure is reached the best achievable scaling accuracy for the system in use is achieved.  Going to higher pressures will not lead to further improvements in divider accuracy.  While the threshold pressure does vary with dough, I have seen it as low as 15 PSI on pizza dough and as high as 60 PSI on specialty dough, it generally averages about 40 PSI.  Because dough held under pressure will age with faster, itís beneficial to minimize the pressure used and imperative to minimize the time the dough is held under pressure.  Thus establishing the threshold pressure of a given dough can be helpful in maintaining best possible scaling accuracy and product quality.

Metering is accomplished in either of 2 ways.  Until the first continuous mixer devices arrived on the scene, a cylinder and adjustable stroke piston were the primary method.  Continuous mix units substituted a variable speed linear delivery metering pump pumping through a fixed time cycle cut-off.  Various means were employed to sever the measured dough from the upstream mass generally involving gravity, and with the exception of continuous mix, a take away belt.

5. MORE POPULAR DIVIDING SYSTEMS:
As a tribute to its designers, machines that are functional duplicates of the first 1890ís ram and shear units are still in wide spread use.  The most significant change, besides the piston head degassing rods added in 1927 are that the new units have substituted hydraulic drives for the cam and lever systems used initially.  Thus the pressure referred to earlier is controlled directly by the pressure on a hydraulic cylinder instead of by precompressed springs.  While some attempts have been made to control scaling adjustment via a downstream check weigher itís still common to have the units directly monitored by an operator to compensate for the lack of an efficient degassing system.

Some time during the early 30ís pressure dividers that included a dough filled tank under air pressure, a three cylinder piston arrangement to divide the dough, a built in inverted cone rounder and product table provided the first step in the automation of bun dividing.  Even though they provided a batch operation, having to be shut down frequently to reload, and holding the dough under pressure for an extended period of time caused significant quality problems, this type unit was the main stay of bun production right up to the early fifties, and flour tortilla lines more recently.

In 1953 the still popular K Head, a single rotating drum machine with 4 cylinders forming 8 delivery ports, when combined with 2 sided pistons, came into use and revolutionized the soft roll side of our industry setting in place the productive structure to support the rapid growth of the fast food industry that started becoming a factor in the fifties.  While the K Head does not appear to follow the pressure rule expressed earlier, in fact the combination of the knife, wedge, exit end hopper skirt, preloaded cylinder and the timing of the piston rise all combine to emulate the pressure required of more conventional machines.

In the mid-sixties the simplest dividing system burst on the scene and almost took the industry by storm.  The divider consisted of a linear metering pump which delivered dough on a continuous basis through a shaping chamber and opposed cut-off gates dropping the dough piece directly into a waiting pan.  The divider was simple, inexpensive and fast.  However the different nature of the bread, because of the upstream process, found acceptance mainly in the South East with the rest of the country basically rejecting the product because it did not match their perception of the type of bread they preferred.  Unfortunately at the time no one was able to device a way to couple the continuous mix system to a depositor that would directly load bun pans since the product characteristics would have in all probability proven very successful in the production of soft roll products.

Taking a lead from the dividing simplicity of the continuous mix systems, in the early eightyís the first of what is now term extrusion or rotary dividers went into commercial use.  The approach was to combine a pressure controlled, vacuum loaded version of the industry standard twin screw dough pump, with a linear metering pump and a simple cut-off.  This approach made it possible to feed the new system with conventional dough and process the resulting dough balls through a conventionally configured rounder, intermediate proofer and sheeter, moulder, panner.

The result proved to be very effective partially due to the very efficient degassing capability eliminating, in most cases, the need for a full time operator.  The unit also provided a more consistent and finer grain structure and, in general, a more uniform product.  It also allowed greater flexibility in dough piece weights, and almost unlimited throughput speeds and piece per minute rates and could be configured to provide 2 or 3 independent lanes of product to feed multiple sheeters.  In fact, based on the capability of these units the high end throughput rates of the fastest lines in the world increased from the 14,000 lbs./hour to the 20,000 lbs./hour range for bread lines and from a high end of 6,000 lbs./hour to in excess of 12,000 lbs./hours for single line soft roll lines.  Because of its efficiency, extrusion dividing is becoming the norm for soft rolls and probably would have for bread except that customer expectations for certain specialty products includes an open grain structure and a subtle difference in cell shape.  Where as rotary dividers answered a bakers normal prayer for an even tight grain structure and no tunnels, no matter how hard we tried, we were unable to trick the rotary divider into opening up the grain structure enough to satisfy the desired characteristics of certain specialty bread without introducing other unacceptable negatives.  Thus the many combination white bread and specialty lines were forces to remain with the older ram and shear technology to protect their specialty bread franchise.

In the late 80ís an off chute of the extrusion divider that was based on the Earl Glassí patent was imported from Australia and went into commercial production.  This unit substituted opposed rollers feeding into a closed chamber for the twin screws of the first extrusion divider and a single drum, something like a single cylinder K Head drum, positioned at the bottom of the closed chamber to accomplish the metering function.  By varying the speed of the two drums the pressure created as the dough was pulled through the 3/8 gap between the drums would force feed the dough into the piston open to the bottom of the closed chamber while the piston on the opposite side discharges its dough piece onto a take away belt.  Speeds to 210 pieces per minute could be maintained and the degassing effect of the pinch roller system was very efficient.  Best of all, once formulation issues had been addressed the product characteristics from bread produced on this unit were similar to bread produced on the old ram and shear units providing a universal divider capability.  However the penetration into the market has apparently been limited by a variety of issues.

6. TWIN PISTON BREAK THROUGH:
In mid 1997 a beefed up version of a typical 8 pocket extrusion soft roll make-up line went into production in a Southwest frozen dough manufacturing facility.  The unit was purchased to feed a 12,000 lb./hour freezer belt with a variety of soft and hard roll products including a pretzel stick.  At the time of installation the plant was running 3 European style lines feeding another 12,000 lb./hour freezer belt and needed more production.  A second freezer belt, primarily used for bread and sweet goods had available time but there was only room to put in one new make-up system and the plant needed to be able to maximize the use of the second freezer belt at its 12,000 lb./hour rated capacity.  Thus the extrusion type system which could out produce all 3 of the old style systems was the logical choice.

While the line was successful in its original form, significant formulation changes were required to achieve the product desired and, in general product shelf life was limited to around 12 weeks vs. the 16 weeks desired.

To hopefully over come these problems without decreasing throughput we borrowed a twin piston type continuous feeder typically used in the meat and cheese industry and substituted it for the screw portion of the original system.  Thus the vane extrusion manifold was now fed by the new twin piston meat pump.  The manifold, rotary cut-off and balance of the make-up system was not altered.

On start up we immediately noticed that the delta dough temperature between the infeed hopper and the cut-off dough piece dropped from the 6° to 8° F seen earlier to 2/10 of 1° F.  The old style European machines typically had a delta T of 3/10° F.  With this indicator the plant switched to the formulation and mixing parameters used on their conventional machines.  The result was the new line essentially duplicated the product characteristics and shelf life of the old style lines while continuing to maintain better than 3 times the old style dividers.

Note that this was achieved with a frozen type dough, which included dough temperatures in the low 60ís, absorption rates sometimes below 50% and elevated gluten levels.

The twin piston pump supplied, which has now been in production for in excess of 2 years, uses twin sleeve and vacuum headed pistons to alternately load and deliver dough to the downstream manifold on a continuous basis.  With one sleeve extended, thus trapping a load of dough inside it and the associated piston moving forward and delivering a controlled rated of dough to the downstream manifold, the second sleeve pulls back, then the piston pulls back sucking a new load of dough into the created cavity.  Once filled the sleeve moves forward sealing off the captured dough and, as the active piston completes its stroke the second piston takes over the feeding task while the first piston goes through its reloading cycle.  The outfeed of the 2 pistons is coordinated by an articulated downstream valve.  The result is product delivery at either a constant volumetric rate or alternatively as what ever rate is necessary to maintain a continuous downstream pressure.

Following the success with the frozen dough operation a second twin piston pump was configured to accept either a single or double extrusion cut-off system and then tested on a number of different bread lines.  The initial tests were on a line that ran very specialized products including rye, pumpernickel, and high particle breads.  In every case the product characteristics matched those delivered by the old ram and shear confirming the lack of any negative rheological impact of the system.  However whereas scaling on the frozen dough system proved to be more than adequate, on the initial bread tests it varied in the range of +/- 10%.  Tests were then run with a vacuumized hopper and a stuffing auger, and with and without vacuum pistons.  At this point the unit appears to be capable of holding in the +/- 2% range.  While this might equal many of the ram and shear units still running our target is the +/- _% typical of extrusion dividers and we still have a way to go.  The good new is of course the ability of this new technology to handle high throughputs, up to 30,000 lbs./hour, high multiple stream piece rates and do so without negatively effecting product rheology.  The bad news is we are not quite there yet, but we are trying.

For those of you either listening to or reading this, we have not tried the unit out on other systems, such as string lines and extrusion lines.  This may be an answer to the challenges of various other type systems where a constant feed is required and the product is relatively sensitive.

There are 2 other technologies that we may see in the future that I wanted to mention in passing.  One is stress free lines, which are basically low pressure extrusion lines that divide dough on the belt.  This type of line has been making inroads in other type products and may prove effective on bread production.  The second is the possibility of measuring dough mass through one of the various non-intrusive mass flow meters currently in use.  I actually ran extensive tests using this technology in the early 80ís and determined that while it did work it was not a practical application at that time ­ dough is just to "non-Newtonian".  Itís possible that at some point the mass flow technology will become usable in dough systems.

Thanks for your time and attention, I will be happy to take any questions.