A chain really is only as strong as its weakest link.
THE MOLECULAR HOMOGENIZER
THE TRILONGATIONAL MIXING SCREW
Part I: First Principles of Mixing
Keith Luker, Randcastle Extrusion Systems, Inc.
2/11/2025
FOREWARD
Single screw extrusion mixing has progressed incrementally. All the single screw mixers do something a little bit different. One is said to be dispersive and another, distributive. Tests of mixing have relied, primarily, on the Dow Chemical method of comparison.[1]
While it can be argued that there is overlap between the mixing of single screws and twins, twins have dominated the market for extrusion compounding. Corotating twins have elongation at their intersection for mixing. Counterrotating twins act as two roll mills at their intersection. Both have advantages.
Studies comparing single screw to twins are rare and difficult. There are countless things to mix and the measurement of mixing quality is difficult, expensive, and dependent on use.
This author knows of no papers about compounding virgin polymers. This is not surprising. It seems obvious to ask, “What would be the purpose of mixing a resin company’s virgin polymer?” And, “If you compound it, you’ll add a heat/mechanical history and degrade it. Why bother?”
The purpose of mixing a virgin polymer is to make it better by improving its properties.
No one believes that resin manufacturers make perfect polymer. The chains are not all the same; the space between the chains is not uniform; the inclusions (gels, debris, etc.) are not perfectly distributed; chain alignment is inconsistent. Local diffusion is problematic. Intuitively, we know that vastly better mixing will make some polymer properties better. We just haven’t had the ability to mix at the tiniest scale of molecules. Now we can. This becomes obvious when we consider bubbles and water vapor.[2]
We test tensile bars by pulling them apart to measure the point where they fail. The failure point changes because the polymer flaws change. We make many tensile bars to get a flaw average. Suppose that we factually knew that extremely small inclusions were the cause of tensile bar failure. We’d conclude that the tensile bar failure occurs at the greatest agglomeration of inclusions. Much better distributive mixing of the inclusions would make the tensile bars stronger. We’d say the polymer is stronger.
Sometimes what tensile bars tests actually measure is the biggest flaw during elongation. Other polymer tests (impact, gas permeation, flex, light transmission, etc.) also measure polymer flaws. Many people have understood that vastly better mixing would decrease flaws and properties would improve.
Such mixing will likely produce surprises. The best mixer is the one that can do what no other mixer does. A mixer that improves virgin polymers on standardized tests and also generates, new, surprising results is the best mixer—The Molecular Homogenizer.
ABSTRACT: A new level of mixing is created by a novel, patent pending, screw. The fine mixing significantly alters polymer properties, improving the physical characteristics of virgin polymers processed through the screw. In hygroscopic polymer, the water vapor absorption is delayed. Undried hygroscopic polymers display no bubbles as the mixing keeps the gas from changing state. Rheological properties show incredible improvement.
Structurally, the screw is a Continuous Spirally Fluted Elongational Mixer (CSFEM[3]). This screw is designed from first principles for straightforward operation. Flowing melted polymer experiences multiple sequences of shear, elongation, and trilongation.
The tested screw’s predicted mixing actions in the 1007 (100 trillion) imply significant possibilities. Such an increase implies mixing to the small molecule level (such as 3 atom water vapor). Because the sequence is orderly, the mixer imposes organization and structure to the messy, tangled chains produced by a polymer reactor.
Trilongational mixing (material flowing in 3 directions) changes the alignment of long polymer chains; distributes small molecules (monomer, water vapor.), additives, particulate, actives, etc. Trilongational mixing surpasses elongation’s two direction flow and is better than shear’s one directional mixing flow. This results in the improvements we see at the macroscopic level. The mixing may enhance, indirectly, diffusion.
Mixing is now thought of as molecular homogenization. The mixer is the Molecular Homogenizer.
Applications:
• Polymer Producers: We expect a Molecular Homogenizer placed at the end of a polymer reactor (before pelletizing) to improve physical properties and lessen the need for drying.
• Extruders and Molders:
• PET Crystallization Unnecessary: Improved solids conveying in the MH allows the lower melting temperature amorphous phase to flow unimpeded.
• Improved Degassing: Inversive mixing (a new kind of mixing where the inside becomes the outside) combined with many extensive free surfaces creates excellent degassing even with only an atmospheric vent. This may allow bypassing, or at least reducing, the energy required for drying. Vacuum may be unnecessary.
• Reduced Drying Costs: For some hygroscopic material, drying or venting may be unnecessary.
• Direct Extrusion: The Molecular Homogenizer is a stable, high-pressure pump/mixer without the need for a gear pump or twin compounding.
• Toll Compounding: Save the cost and heat history with the Molecular Homogenizer.
• Improved Physicals: Often, better physicals lead to downsizing with consequent savings.
• Reclaim: Higher quality and improved viscosity often lead to increased output.
• Higher Output: PP, PFA & rigid PVC powder are known to process at higher rates through the SFEM.
• R&D: A new realm of mixedness means opportunity for new products.
• Twin Screw Users: All the benefits above may be enjoyed by adding the Molecular Homogenizer to the end of a twin screw extruder.
INTRODUCTION: The purpose of this paper is to show how the use of first principles for extrusion mixer design creates a new realm of mixing. Then, we show how this new mixing realm creates opportunities for greener processes from energy savings and increased profits.
The last significant invention in extrusion mixing was the 1958 twin screw kneading disc (Erdmenger patent 3,122,356). In the 1980’s, it was discovered that twins generated stretching (two directional elongation) at the disc intersection—an improvement over shear (one directional flow). Randcastle’s patent pending invention, the Molecular Homogenizer (MH), promotes a more powerful dispersive mixing force, trilongation, where all the flow is stretched in three directions rather than elongation which is a two directional flow.
There are two mechanical forces: compression and elongation—push and pull. In extrusion melt mixing, one force, compression, is negative and counters mixing. That is, when compression is combined with shear or elongation, mixing is significantly retarded. Different arrangements of shear and elongation through various screw designs are known. Historically, single screw designers concentrated their efforts on shear with compression. This pushed material through narrow gaps (i.e. Madock, Egan, Double Wave fand barrier screws) but, “…shear is an extremely inefficient way to mix.”[4] However, by reorienting or interrupting shear, order of magnitude improvements become possible.[5] Further, when a mixing flow is oriented in one direction and then another, “The orientation imparted by the first section is destroyed, and each “stage” of the device behaves as a separate unit.”[6]
Rauwendaal formulated criteria for mixing[7] and suggested a mixer include elongational stresses to break down agglomerates and droplets; exposure to high shear stresses for only a short time; that all fluid elements experience the same stress level multiple times. Rauwendaal created the CRD mixer in 1998 and tested it with a HDPE/PS mixture. The tested result was a “core” (representing about half of the volume) said to be about 4 microns and an outer region said to be about 20 microns due to coalescence.[8] In 2007, a similar material processed in the Randcastle AFEM[9] showed 1 micron domains and no coalescence.[10]
In 2005, the forerunner of the Molecular Homogenizer (the SFEM) created significant elongational mixing[11] and in 2011 Dow Chemical showed mixing results that were eight times better than their twin with immiscible polymer blends.[12] [13] Meanwhile corotating twins, naturally elongating material at their intersections, maintained their dominance for the more challenging mixing applications.
In 2024, Randcastle filed patents in Europe and five other countries for a novel mixing mechanism which is the subject of this paper. It uses first principle arrangements of the interrupted shear,[14] interrupted elongation and trilongation, a new mixing flow called inversive mixing and multiple extensive surfaces for efficient degassing.
Until this arrangement, it was not understood that one of the two mechanical forces—compression—was so detrimental to extrusion mixing. Stated another way, if there were no compression mixed in with shear and elongation, they would mix vastly better. Fundamentally, this is one of the reasons the Molecular Homogenizer produces such surprising results.
Fundamentally, the task of extrusion mixing is to arrange elongational and shear forces in three dimensions without degrading these forces. This paper shows the mixing forces arranged in sequential directions that mirror the three dimensions. Achieving a quantum leap in performance requires new principles:
· Inverse Mixing: A new flow where the outside flow becomes the inside flow.
· Multiplicative Mixing: The mixing mechanism in each mixing element, being identical, is multiplicative with a large multiplier of 100 (instead of two in a static mixer). However, it is without the error multiplication of static mixers.
· Trilongation: Stretching in three directions simultaneously.
· Axial Mixing: To mix at the molecular level, axial mixing—at an extremely small scale—is required.
· Equal Treatment: All the material passes through the same mixing sequence.
· Maximize Mixing Forces: Remove the forces that degrade shear and elongation.
This arrangement creates exponentially improved mixedness such that surprising results start to become mundane. For example, processing through the unvented Molecular Homogenizer screw is known to:
· Improve Physical Properties: I increase tensile-at-yield greater than the virgin polymer.
· Increase Viscosity: Restored undried hygroscopic regrind to the viscosity of the unprocessed, undried virgin polymer.
· Slow Water Absorption: Processing slowed water absorption from the typical 4 to 6 hours when exposed to atmosphere to over 72 hours.
· Sequestered Water Vapor: Moisture analyzer measurements showed the moisture content of an unprocessed material entering the Molecular Homogenizer at 0.328%; it then reported that the Molecular Homogenizer processed polymer had 0.171% moisture.
· Vent: Vent even nylon with just a single atmospheric vent.
While mixedness is the primary purpose of this paper, the Molecular Homogenizer has other important properties. Two of the most important are:
o Improved Solids Conveying: More efficient solids conveying is important when mixing two materials where melts at a substantially lower temperature than this other. This occurs, for example, when combining amorphous and crystalline PET. In conventional solids conveying, the lower melting material prevents the solids from moving forward. The Molecular Homogenizer resolves this problem. In other words, amorphous PET does not need to be recrystallized.
· Venting With Extensive Free Surfaces: Extensive free surfaces are particularly useful for breaking bubbles and therefore degassing. There are 35 extensive free surfaces in the tested screw.
Capitol Expense: This Molecular Homogenizer is single screw extruder that generates stable, high pressure combined with unrivaled mixing at far less expense than a twin. Patents are pending in five countries and Europe.
FIRST PRINCIPLES OF FLOW:
Introduction: There are only two mechanical forces: Compression and elongation. Of these, compression is a negative mixing force and elongation is positive.
Elongation can be broken down into shear, elongation, and trilongation (aka three dimensional elongation). Shear mixing is simpler to understand and model and dominated single screw efforts through the 1960’s.
The invention of the kneading disc in 1958 marked a significant advance in dough mixing and then was incorporated into the twin screw extruder. This coincided with our understanding of mixing and the benefits of elongation over shear. In the 1970’s Gogos, Tadmor, and others understood that elongation produced far better dispersion and distribution than shear. The corotating twin elongates the flow at the intersection of the screws (and singles screws have no similar intersection); twins with kneaders dominated the extrusion compounding market for all the more demanding compounding applications.
Today, almost all corotating twin screw presentations display kneading discs because they amplify elongation. Without kneading discs, twins would not work as well. Nevertheless, twins with kneading discs are far from perfect. In the 1980’s, numerical methods and computational fluid dynamics began to model elongation such as in the schematic below:
o Bypass Flow In Kneaders: The kneading block elongation is shown below by color. Red is the most intense elongation, yellow less so and green insignificant. So, a serious twin drawback is that the vast majority of the flow bypasses the region where elongation occurs:
o Uneven Intersection Mixing: Twins improve their mixing by creating many intersections. But:
o Some Material Is Never Elongated: As each intersection mixes only a minor portion of the total flow, some material always bypasses the intersection and is not elongationally mixed. This allows otherwise dispersible material to remain in the flow.
o Some Material Is Over Mixed: Some material flow through the elongating region more than once and so is mixed more than the rest. Overmixing results in degradation.
o Various Degrees Of Mixedness: Taken together the above describes the corotating extruder as a machine that must mix unevenly.
Various incremental improvements are known to aide twin screw mixing. For example:
o Reverse Flights: In order to improve mixing, reverse flighted elements are often placed downstream of the kneading blocks. This forces the flow to remain in the kneading blocks under higher pressure. Fluid pressure is a form of compression pushing equally and opposite in all directions. Pressure pushes against the twin’s extensive mechanism which reduces elongation significantly restricting elongation (Fig. 8-9).
o Various Kneading Disc Widths: In order to aide dispersion, wider kneading blocks are used. To aide in distribution, narrower kneading discs are used. Since neither mixes all the flow evenly, some is better dispersed and some not and likewise, distribution. This describes a nonuniform mixer.
In 2007, Randcastle patent 6,932,431 B1 was awarded. It was shown by Dow Chemical (Fig. 23), to be dominantly elongational and 8 times better than Dow’s twin. In 2024, Randcastle filed for a new patent in Europe, the US, Canada and other countries. The name of the new mixer is the Molecular Homogenizer as it benefits from new mixing principles that increase mixing performance exponentially. These new principles include:
1. Optimized Shear, Elongation And Trilongation.[15]
2. Dynamic Multiplicity.
3. Equal Mixing Of All Flow.
4. Inversive Mixing.
5. Extensive Free Surfaces.
6. Micro-level Axial Mixing
Matter has three dimensions, X, Y, Z. The task of extrusion mixing is to create a mechanism that mixes in all three dimensions. It’s easiest to describe the screw and barrel in terms of dimensions while we describe the flow in terms of directions. We will use the standard convention of a stationary screw with a barrel rotating around it[16] with three dimensions as below:
Fig. 1
It is important to understand that the when the barrel rotates around the stationary screw (in our minds, not in reality as the screw actually turns). X is best thought of as the axis of the screw. The Z dimension is perpendicular to X and Y.
First Principles Of Single Screw Mixing:
1. Optimize shear, elongation and trilongation:
Avoid Compression/Resistance: Compression agglomerates; agglomeration is the opposite of mixing. The fluid form of compression is pressure.
Maximize Mixing Force: Pressure substantially reduces mixing force. It doesn’t matter whether the shear, elongation or trilongation in single or a twin. Operate at zero pressure everywhere possible.
2. Dynamic Multiplicity: Static mixers are well known. Static mixers are multiplicative often having a multiplier such as 2. This screw has dynamic multiplicity of 100 per mixer and seven mixing elements. Whereas static mixer generate layers in the 256 to 4,097 range, 100 trillion mixing actions are calculated in the seven mixer tested.
3. Treat All Material Equally: It’s counterproductive to mix one part while you give the rest a pass. That’s what happens in corotating twins—some of the melt is, some elongated and some just goes by.
4. Inversive Mixing: Make the outer flow become the inner flow. In other words, turn the flow inside out every mixer so that the material on the screw changes places with the material on the barrel.
5. Large And Many Extensive Free Surfaces: The outer surface of an extensional flow promotes degassing since the extension can break bubbles and release gas in a vent. The larges such surfaces for the same mass, the better.
6. Mix Axially: No matter how uniformly you drop different materials into a screw, they are not mixed at the molecular level. Small scale axial mixing is required. d
MECHANICAL FORCES
There are only two mechanical forces that we can exert on an object or a flow. These are compression and elongation.
Compression: When compression is applied equally and opposite in all three dimensions, it is called pressure. When we apply compression or pressure to the cube, sphere or bubble, mixing does not occur. The force is applied in all three dimensions inwardly. If the cube or sphere were made of loosely held particles, then compression or pressure will push these particles together causing agglomeration. When pressure is applied to the bubble, it shrinks, becoming denser and this too is agglomeration. Agglomeration is the opposite of mixing and is caused by compression and pressure. You can think of pressure as an anti-mixing force.
Fig. 2
Elongation: The other mechanical force is elongation. Here we are concerned with melted polymer. Melted polymers are laminar rather than turbulent and mix in predictable ways. We focus on, “…presenting the principles that relate fluid flow to mixing.”[17] [18]
Fig. 6 (Fig. 4 + Fig.5) Fig. 5 Fig. 4 Fig. 3
Above, Fig. 3-4 have fixed gray boundaries. The frictional boundaries act as drag slowing the flow[19] represented by arrow length. This is often thought of as the flow through a tube or slot. The pressure pushes (acts compressively) material forward. Fig. 3-4 flow lines vary depending on material.[20] The flow is driven forward (pushed) by pressure (compression).
Fig. 4 has a moving boundary that pulls (an elongational force) material forward in parallel lines—one layer dragging (pulling) its neighbor along. Larger Y letters denote the degree of flow as does the length of the arrows. Since Fig. 1 specifies Y as the direction of barrel rotation, simple shear is labeled Y. It will be useful, later, to imagine that the moving boundary is pumping the material forward.
Effect of Pressure On Shear Mixing
Simple shear, in the sense of geometric lines, is one directional (linear). The shear mixing does not involve the X or Z direction. Pressure flow, Fig. 3B and 4A represent the typical range of polymer flows. When pressure flow is combined with Fig. 5 it pushes on the shear flow. This disrupts and retards the simple shear mixing as shown in Fig. 6. This changes “shear” into a remnant of what is desired. It becomes increasingly like the poor mixing of pressure flow (usually only thought of as mixing during purging).
Shear reduction is not exclusively one way. Restriction to flow reduces the little transport that shear creates. An extrusion die creates resistance. The greater the resistance, the greater the degradation of the shear force.
This point needs further emphasis. The conceptual shear flow in Fig. 5 occurs only when you don’t push on the shear flow. While Fig. 5 is technically a pump, it is an extremely weak pump. This is readily overcome by upstream extruder pressure (such as in Fig. 3B or Fig. 4A). Often in the thousands of pounds per square inch, such pressure reduces the effective shear and so mixing is less effective.
Preserving some shear such as by using low pressure ahead of the barrier, is known. For example, Barr’s much greater barrier clearances in the VBET (compared to the Double Wave Screw #US6227692) reduces the pressure required to push material over the barrier. This preserves some useful shear for mixing, lessens the temperature rise and makes the shear more effective.
Or, consider a UC mixer (Maddock) mixer. Essentially, it is a narrow slot where material is pushed through. The narrow slot is intended to create mixing shear. The input pressure reduces that shear. Compare the UC mixer (Fig. 7 left, normal pressure) to a mixing screw with the same narrowed slot but where there is a lowered input pressure head of the slot (Fig. 7 right).[21] The contrast in Fig. 7 is stark:
Fig. 7: Light Shining Through 0.010 Thick Film
Effect of Pressure On Elongation
Schematically, simple elongation moves in two directions. Below, molten polymer is represented by a line of length L. If the polymer is stretched beyond its limits, it will break. The maximum elongational length occurs just before it breaks.
Fig. 8
Combining compressive forces (pressure) against extension is not fully appreciated; it’s very bad for mixing. Everyone who has pulled on a melted strand (i.e. a melt leaving a die) knows that it doesn’t take much force to stretch it out. Gravity alone will readily stretch most polymer strands as it falls toward the ground. By comparison, extruders develop very high opposing pressures. Imagine the difficulty of trying to pull a melted strand by hand while your hands are being opposed by the pressures that a screw can generate. The concept is shown graphically below where Fig. 3B reduces the elongation:
Fig.9 Effect Of Pressure On Elongation
Elongational Flow: Below, fixed boundaries are shown 90 degrees to the moving boundary. The pressure is zero (as in a starve fed screw). Heavier flow lines denote stronger extension and greater velocity. The letters denote a relative amount of each direction by size. Fig. 10A shows asymmetrical elongation in two dimensions zY where material sticks to a point and is dragged by Y (the barrel in a single screw).
Fig. 10 C Fig.10B Fig. 10A
A natural 2:1 draw ratio[22] occurs until converging into the four parallel lines (plug flow where Y is non-mixing). Flow is primarily Y and during convergence zY. Elongation is 2 directional flow and mixes better than 1 directional shear. We can think of the elongational mixing in 10A as area mixing since it involves two dimensions.
Fig. 10B shows the flow tethered to the fixed vertical boundary pulled by the moving boundary Y. The flow is strongly Z and nearing the moving boundary weakly y.
TRILONGATION
Fig. 10C shows the innermost flow labeled X. This flow is moving toward the reader. Close to the fixed boundary, X flow is slowed by the boundary resistance. Away from the boundary, the X flow drags the yZ flow downstream. The combined flow (Fig. 10B and Fig.10C) is xZy flow or three-direction flow. Just as two directional flow (10A) mixes better than one directional flow (shear), three directional flow mixes better than two directional flows. This three directional flow is called trilongation. We can think of the mixing in 10B/10C as mixing the entire volume.
We can think of the increasing mixing performance, too, in the same way we consider the relationship of line, area, and volume. That is, a line has one dimension, area has two dimensions, volume has three dimensions. For mixing it’s best to think of directions where there’s one for shear, two for elongation and three for trilongation.
Regarding combined Fig. 10B/10C, X flow is linear and so appears to poorly mix. However, the X flow diminishes over its length, Fig. 11. Eventually all the X flow is converted to yZ elongational flow. This is shown in the Fig. 15 view. Note that the velocity of the X flow slows as it diminishes.
Fig. 11
MOLECULAR HOMOGENIZER MIXING ELEMENT
Fig. 12 links Figs.10B/C, Fig. 5, and 10A into a single conception of first principles creating an orderly, non-chaotic flow.[23] The orientation of mixing flows is reoriented with each ‘stage.’ This arrangement creates exponentially improved mixedness.
Fig. 13 Fig. 12
Upper Fig. 13 is a 2D embodiment of upper Fig. 12 flow adding the structure of a mixer. Here the gray vertical fixed boundary of upper Fig. 12 becomes Pump 1. The mixer then repeats the gray boundary becoming Pump 2 of upper Fig. 13. Partially filled channels (C1, C2, C3) and flights complete the conception.
The orange arrows point to five extensional free surfaces. In the tested screw, there were 7 mixing elements in a 36/1 L/D screw for a total of 35 free surfaces. Extensional free surfaces contribute to breaking bubbles which is necessary for degassing during vented extrusion. The free surfaces can be linked to a single vent or isolated. [24] Free surfaces are unconstrained which minimizes frictional temperature increase. Note that in upper Fig. 13, there is no pressure pushing on the shear or elongation. This maximizes the mixing.
Lower Fig. 13 is a 3D representation of two elements of the Molecular Homogenizer mixing element. The 36/1 L/D of the tested screw had seven mixing elements. The mixing action in each is the same. This multiplies the result in the same way as the result is multiplied in a static mixer. Unlike a static mixer, errors are not copied and repeated.[25]
For clarity, upper Fig. 13 (the conceptual cross section) and lower Figure 13 are shown below to the same scale so that the conceptual parts line up vertically with the 3D representation below:
Fig. 14
INVERSIVE MIXING
Better Degassing: By “inversive mixing” we mean that an inner flow becomes an outer flow; or, an outer flow becomes an inner flow. Such a change has significant benefits. In a conventional vented extruder, there is a rolling flow under the vent. The purpose of the vent is to extract gases. While the surface of the rolling material is exposed for degassing, the center of the viscous flow is not. Single screw degassing is therefore very limited. Degassing improves greatly with inverse mixing since the central flow becomes the surface flow.
B
Below Fig. 15 shows a cylindrical unwrapped projection[26] of two elements of the mixer. The flows are projected onto the channels of a mixing element
Fig. 15.
The arrows show the Y flow direction on the P1 and P2 pumps. These are the areas of simple shear flow (Fig. 5).
The first pump, P1, pulls the material away from C1 where yZx flow in C1 (Fig. 10B/10C) is converted to simple shear over P1. This reorients the C1 flow. The P1 pumps at a constant rate draining the C1 channel.
The constant flow from P1 stretches over C2 (see upper Fig. 14) before the flow forms a second XYZ flow upstream of P2. Because the flow from P1 is constant, the XYZ shape becomes cylindrical. X mixing is enhanced. XYZ flow is reoriented into simple shear.
P2 then pulls the material from C2 and delivers a constant flow to C3. Initially stretching occurs over C3 and is again converted into XYZ flow. Since the flow from P2 is constant, the shape of the flow in C3 changes to a conical flow.
Fig. 16
In the view below, cross sections A-A and B-B are shown.
Fig. 17
Above left, P1 pulls material from C1. Because the X flow sticks to the fixed surface and the barrel, the outermost flow, yZ, is drained away. X flows downstream until it is entirely converted to yZ flow. Note that the X flow is on the inside and the yZ flow in the cross section is on the outside.
Fig. 18
Once the P1 gap is fixed, the thickness after stretching will be half of P1. P2 must be less than that or it will have no effect. When the P2 clearance is slightly less than half the clearance, then the flow length in C2 doubles. Since the X flow is linear, by doubling the X flow length, the axial mixing is increased two-fold. This is important because, no matter how uniform the solids feed of the mixture is made,[27] it is not uniform at the molecular level. Improving the axial mixing becomes a key element in mixing to the molecular level. The increase in X length is not limited to a multiplier of 2.
Note that the green arrows point to the outer flow in C2 that becomes the inner flow in C3. This is called inverse mixing. The x or axial flow mixing (although converted to Zy) is weak. Inverse mixing can be thought of as the inside becoming the outside or the top becoming the bottom. In terms of axial mixing—weakly mixed in 9B-9C—X is moved from the innermost flow to the outermost flow i.e. next to the barrel.
Since the objective of mixing is to arrange mixing flow in all three dimensions, axial mixing is greatly improved. Since the feed is inherently, definitionally, poorly mixed (on a microscopic scale), this is extremely important.
From an engineering point of view, the objective may be to balance the mixedness of the three dimensions using the X, Y, Z directions[28] and inverse mixing and axial mixing contribute greatly.
Note that the mixed material flowing from the first C3 element flows into the next mixing element into C1. The mixing method of the first element, like a static mixer, is then multiplied in the second and subsequent mixers. Every element inverts the flow.
C3 FLOW
As seen previously, the flow in C1 and C2 are trilongational and experiences inverse flow where the inside becomes the outside. This is not the same as the flow in C3. As the material leaves P2, it is stretched over the first part of C3 and thins to half the P2 gap as shown in Fig. 18. Unlike the flow in C2 and C3 (where there are adjacent upstream pumps P1 and P2), this material flows against a flighted channel where it rolls somewhat like the flow in a conventional vent. P2 pumps the same amount across its length and here the P2 flow becomes the inner flow in C3. The flow from P2 adds uniform layers C3 flow.
AXIAL MIXING
It is quite common to mix 2% color concentrate in extruders. In the 1 inch extruder that made the LDPE films below in a conventional screw, each black color concentrate pellet forms a distinct parabolic line. The axial mixing of the Molecular Homogenizer shows greatly improved results in axial mixing.
To better quantify the issue, consider previous Fig. 7 below. It too was extruded on a 1 inch extruder with much less color concentrate into a 10 mil film. A light shines through both samples:
UC Mixer Molecular Homogenizer
Fig. 7
The notoriously poor mixing of small extruders is readily visible in the UC mixer. There around 100 pellets per L/D in a 1 inch extruder. [29] At 0.5% yellow color concentrate and 0.5% red color concentrate, there would be about 1 yellow pellet every two L/D’s and one red pellet every two L/D’s. Axial mixing is necessary so that the yellow and red combine. Since the average colors are far apart along the screw axis, this picture demonstrates the Molecular Homogenizer’s axial mixing. Essentially, the different color pellets must “catch up” to each other for the mixing to take place.
There are two mechanisms that account for the axial mixing. The first is material transformation. For simplicity, imagine that one element of the spiralling Molecular Homogenizer is straightened so that the channels are parallel to the screw’s axis as in Fig. 19:
Imagine that a purple color concentrate pellet flows into the C1 channel completely surrounded by clear polymer pushing it along. The pellet is a cylinder 0.125 inch long and 0.125 inch diameter. Since the surrounding material is clear, it cannot be seen in the graphic. As the pellet moves forward, the 0.04 inch P1 pump clearance first removes the outer clear material and then start to pull in the color pellet, 19A, pulling it in the Y direction.
In 19B, the barrel continues to pull in the pellet and starts to transform a triangular plain. Eventually the pellet is completely converted into a triangularly plain, 19C. Since the P1 clearance is 0.04 inches, the pellet will completely drain over P1 in 0.125/0.04 = ~4. The original pellet length is four times longer or 0.5 inch at the base of the triangle.
Since P2 is less than half the P1 clearance, the length of the purple pellet will double again to eight times it’s original length or 1 inch. This mixing repeatedly takes place in either of the 7 mixers tested. Additionally, the pellet will stretch over C2 starting from the P1 gap of 0.04 to 0.02 inches. Leaving P2, it will stretch again to 0.01 inches.
The second contribution to axial mixing flow is the core flow during trilongation. The X flow, Fig. 17, moves axially faster than the stretching flow.
The combination of faster X flow and transformational flow creates axial mixing flow. It is not only yellow and red pellets that experience axial flow mixing. The principle applies to the molecular level so that, while a feeder may input a poor mix (a pellet of “A” next to a pellet of “B”), axial mixing improves the input significantly.
A chain really is only as strong as its weakest link.
Fix those weakest links on the microscopic scale,
the chain gets stronger.
THE MOLECULAR HOMOGENIZER
A SINGLE SCREW DYNAMIC MULTIPLICATIVE MIXER
Part II: Mixing Results, Applications, Conclusions
Now that you’ve gone through the first principles of mixing, we’ll show you some of what happens when we put material through the Molecular Homogenizer.
Most of the following work is about virgin polymers processed through the Molecular Homogenizer. There’s unvented, undried hygroscopic color concentrates pellet processed through the MH. The strand displayed no bubbles despite the hygroscopic carrier and the adsorption of the pigment. Intriguingly, those pellets were then processed, undried, in a conventional screw making film without bubbles. The unprocessed, undried color concentrate pellets processed in the conventional screw made a net.
You’ll find bubbles prominently mentioned regarding undried hygroscopic resins in an unvented extruder with the Molecular Homogenizer. It’s the norm that we don’t see visible bubbles in a very wide range of materials. Everyone wants to know where the water vapor is going. Hint: It’s not going out the hopper.
You know, now, that venting is improved with the multiple, extensional free surfaces and inverse mixing. There are two examples that follow—pretty good ones. We successfully vented with just an atmospheric vent while doing reactive chemistry; the other is with a nylon blend (nylon absorbs a terrific amount of moisture) in a production run.
There are surprises, too, in the rheology work. You’ll wind up asking yourself, “How can that happen?" You might wind up asking that a lot.
There are surprising things that we’ve observed that are not in the presentation yet. For example, a material was processed on a control screw. The rate was limited by its hardness entering the strand pelletizer. When we ran the Molecular Homogenizer, it hardened faster so we could increase the rate by 40%.
In another case, two reputable companies tested the molecular weight of their own materials. The molecular weight didn’t change.
Frankly, some of these things seem impossible. But, I make it a habit not to argue with reality or with the numbers from the experiments that our partners are doing on their materials.
Finally, thanks to the skill of a colleague, Tom Cunningham, we have a picture of dispersion actually happening just where you’d expect it—in the elongational flow.
If you’re interested in working with, please give us a call. This is virgin territory. There’s lot of room to explore.
Keith Luker, President
Randcastle Extrusion Systems, Inc.
PICTURE OF MIXING RESULTS
The predecessor of the Molecular Homogenizer model CSFEM was the SFEM. Significant work was done on the SFEM and this gives us an indication of previous performance.
Fig. 15 is a photograph of an SFEM cooling experiment.[30] [31] The extruder processed PP and blue color concentrate. While all the conceptually drawings above are starve fed, Figure 15 was necessarily flood fed.[32] While this created some pressure, note that C1 is always open downstream. This keeps the pressure pushing on P1 to a minimum. The extruder conditions were adjusted so that melt was just completed in C1.
Fig. 21 Fig. 20
C1 is marked is marked 1 and 2 to identify blue “lines” of about the same length. Label three identifies a line about half as long. In total there are 2.5 lines. When we look at the number of lines in C2, the lines are blurred and so not readily countable. While we cannot see any more distinct lines (lines are composed of pigment and this breaks up into individual particles[33]), it’s not unreasonable to suggest that there are at least 25 lines or 10 times better mixed than in C1. It might be better to say that, qualitatively, C2 looks 10 times better mixed than in C1. C3 displays no distinct lines but, since the mixing mechanism is the same as for the transition from C1 to C2, it too is 10 times better than in C2. A rough estimate is that this single mixing element made the input 100 times better than it started.
In the Melt Homogenizer, the C3 channel becomes the C1 channel of the next mixer. Each of the following mixers multiple the previous—by 100. For a Melt Homogenizer with three mixers, that would be 1003 or it would be a million times better than it started. Polymer molecules, while very long for molecules, are still astonishingly small from our macroscopic prospective. Some of the Molecular Homogenizer screw have
At some point (at the moment unknown), there are diminishing returns. “Satisfactory” mixing depends on the application. For many mixing applications, a million times better will be “sufficient” and Melt Homogenizers with only 3 mixers become very attractive for, say, 24/1 L/D extruders, injection molders and blow molders.
Color Mixing Distribution and Dispersion Shown Taking Place
Color concentrates, often less expensive ones, have pigment agglomerations. Besides making the surfaces rough, they must be dispersed in order to be effective in some applications such as 3D printer filament. There, they can jam in the printing extruder, the hot end and the nozzle (tip).
Randcastle technology lets us show you pigment distribution (rearrangement of particles) and pigment dispersion (breaking of pigment agglomerates) below. Here, we’ve frozen the action onto the first of the series of Molecular Homogenizer mixing elements using 2% blue concentrate in an enlarged view of Fig. 21:
Most dispersive mixing takes place with the red oval. As the agglomerated pigment flows from the terribly mixed input, we see it being torn apart in the gross and medium dispersive regions of elongational force. The remaining leftovers, continually overcome by the stretching force, show the fine dispersive mixing. Once dispersion is accomplished, we see the fine pigment particles are then greatly distributed. Importantly, this graphic shows the mixing of only the first of the series of mixing elements of the Molecular Homogenizer. Each element repeats the mixing and exponentially reaches down to the molecular level.
IMMISCIBLE POLYMER BLENDS:
Significant work was performed by Dow Chemical who quantified their work below:[34] [35]
Fig. 24
The left axis of Fig. 24 shows the domain size. The average of the twin is about 2.5 microns and the domain size for the SFEM is about 1.25 microns. The dispersion of the domains is 8 times better than Dow’s twin (the volume of a sphere is 1/8 when the diameter is reduced to half). The twin average for distribution on the horizontal axis is about 3.5 while the average for the SFEM is about 0.9 or around 4 times better than the twin. This is our predecessor screw.
MIXING RESULTS: MOLECULAR HOMOGENIZER (CSFEM).
Rheology:
A study of PLA viscosity was conducted.[36] Randcastle pelletized the material through the Molecular Homogenizer and the rheology was performed by Raj Krisnaswamy of CJ Biomaterials. Below, the Molecular Homogenizer that processed the PLA is shown with processing conditions.
An empty Molecular Homogenizer screw uses 5.9 amps to rotate in the barrel. The Molecular Homogenizer used only 6.4 amps of the possible 13.5 amps[37] to process the material. This implies that the mixing uses only a small amount of energy and this minimizes thermal and mechanical degradation.
Fig. 25
Fig. 16 shows multiple viscosity curves using a parallel plate rheometer. Below, the dried, unprocessed virgin polymer, PLA9 acts as the control. Note that all but two of the curves, PLA 2 and PLA 3, have comparable viscosity. This implies that, like the low amperage of Fig. 15, the mixing did not degrade the polymer in those trials.
Fig. 26
In Fig. 26, PLA 7 is virgin, dried, PLA processed through the Molecular homogenizer. Compared to the unprocessed control, PLA9, the curves overlap confirming that the mixing did not degrade the polymer viscosity.
Fig. 26
PLA 2 viscosity is lower. This is not due to thermal or mechanical degradation. PLA2 was exposed to atmosphere and had absorbed water vapor. Fig. 28 below shows one of the unexpected results of this new quality of mixing. PLA5 is the undried PLA2 reprocessed in the Molecular Homogenizer. The viscosity increased to nearly the same level as the virgin, unprocessed material, PLA9. Since it was processed once to become PLA2, it is regrind. Some processors do not have driers and must process their PLA regrind without drying. This can create several problems such as physical property loss, change to product uniformity, change in die flow, product finish and, of course, create bubbles.
The increase in viscosity is very surprising. The expected result of reprocessing polymer is a lower viscosity. The implication is this new quality of mixing acts differently. Somehow, two mixing histories restored the viscosity despite the water vapor.
Fig. 28
PLA 3 is the same as PLA 2 except that it was processed with an atmospheric vent and some water vapor likely escaped. It is interesting that, when PLA3 was reprocessed into PLA 6 (the blue triangle) that the beginning part of the viscosity curve is higher than the virgin material. It is happily surprising that more mixing of undried PLA created a viscosity greater than the unprocessed virgin in any part of the curve as the virgin unprocessed material had no water vapor.[38]
Molecular Weight, Output, Moisture Absorption: Another PLA study was performed by Randcastle with NatureWorks on their 2003D material. In the chart below, the results are show for unprocessed virgin material, a general purpose control screw and the Molecular Homogenizer. Lot LF1928B121 was used for all the tests. The line consisted of a non-vented Randcastle 1 inch 36/1 Taskmaster, single strand die, 8 foot water trough and strand pelletizer. The same process temperatures were used for all tests. The first experiment determined the maximum output of the control screw. At any higher output, the single strand was too soft to pelletize in the 8 foot water trough. The output was 68.3 g/m. The screw was changed to the Molecular Homogenizer where the output was also 68.3 g/m although starve fed to maximize the mixing.
A second experiment was then performed to determine the maximum output of the Molecular Homogenizer screw. The output was increased by 40% before the pellets were too soft to pelletize. Pellets were returned to NatureWorks for analysis:
PLA Research Courtesy NatureWorks
Additional Testing Planned For Impact and DSC
Fig. 29
Summary of Results[39]:
· Molecular Weight:
o Control Screw: The control screw showed a typical 5% loss in Mw compared to the unprocessed virgin material.
o Molecular Homogenizer: The MH showed no Mw change.
§ Most people would expect that increased mixing would reduce the molecular weight at 68.3g/m output.
§ Most people would expect that increased mixing would reduce the molecular weight even more at the higher output of 96g/m at 34 degrees higher temperature.
o The increased mixing does not seem to degrade the PLA.
· Moisture Absorption:
o Control Screw: It is not surprising that the reduced Mw hampered water absorption.
o Molecular Homogenizer: A 70% increase in moisture absorptions is very surprising. Since crystallinity was not detected, the change occurred in the amorphous structure. Since trilongation creates XYZ mixing, it creates an amorphous structure with greater free volume and so greater moisture absorbance.
New Specific Gravity Research Is Underway:
Measurements are currently underway to measure the effect of Molecular Homogenizer processing on density. While preliminary in nature, a few materials have been measured and the specific gravity was reduced by about 1.5%.
This makes sense given the PLA results as measured by NatureWorks. When the density decreases, the interstitial space or free volume increases leaving more room for the 30% greater gas volume for water vapor.
This also makes sense for the physical property improvements that are observed.
At this moment, we think that this is a natural result of trilongational flow. That is, trilongational flow aligns some of the polymer chains in the transverse direction reducing the specific gravity. If so for most polymers, then each pound of plastic produced would have 1.5% more volume. The potential saving in energy and reduction in pounds produced, is significant.
Amorphous PET Feeds In Molecular Homogenizer
It is usually recommended that PET be crystalline rather than amorphous during extrusion. Typically, the crystalline polymer proceeds through a heated dryer warming these pellets. When these heated pellets contact the amorphous PET, it can become tacky and is no longer free flowing.
However, the Molecular Homogenizer screw behavior is different. The MH has processed undried crystalline material without visible bubbles. So, the undried crystalline PET feeds are room temperature with 25% amorphous PET. Even though amorphous PET absorbs more water vapor than it’s crystalline phase, both were extruded without bubbles.
Bubble Free Undried, Unvented, Hygroscopic Polymers
Undried hygroscopic polymers are well-known to produce visible bubbles in unvented extruders[40]. When hygroscopic polymers are only partially dried, visible bubbles may not be apparent. Instead, the water vapor can migrate to the inside surface of the die. The water vapor exits the die on the surface of the molten extrudate leaving telltale ovals as the shallow surface bubbles are pulled from the die.
Another source of bubbles occurs when a polymer is cooled quickly as in a water trough. The outside of the strand freezes quickly while the insides stays warm. Eventually, the inner material shrinks creating a shrinkage void or a vacuum bubble.
The Molecular Homogenizer processed undried hygroscopic polymers that leaves the die and falls by gravity. This gives any water vapor additional time to expand and form a bubble. The slow air cooling prevents vacuum bubbles. The cooled strand’s surface is checked for telltale ovals.
Looking for bubbles in hygroscopic undried polymer is so well-known and so easy to observe that, when undried polymers don’t show bubbles (in a non-vented screw) it’s startling and unpredicted.
All of the following materials were processed through the Molecular Homogenizer showing no sign of bubbles during the gravity fall or on the surface: PMMA, PET, PET with 25% amorphous reclaim, PEEK, PVA, PVA with 3% reactive agent, PLA, SAN color concentrate (black and white), PC, coffee chaff in LDPE.[41]
While all of these examples are remarkable, some of these undried experiments deserve special comment:
· Amorphous PET With 25% Reclaim: Undried PET is very well-known to absorb water vapor and form bubbles. Undried PET reclaim is amorphous and known to absorb even more water. No bubbles were seen. This has serious implications as PET has two phases. Bottles and regrind are the amorphous phase where the melting point is less than the drying temperature. So, in order to dry the material, it must first be crystallized or the screws do not feed. Crystallization, which consumes energy and handling may be avoided.
· Reactive Extrusion PVA: Undried PVA produces a bubbled extrudate. A reactive chemical was added to dried PVA and created a bubbled extrudate in a conventional screw—implying that the bubbles were not water vapor. Yet, in an undried PVA with 3% reactive agent, no bubbles were seen in the Molecular Homogenizer extrudate.
· Coffee Chaff: Coffee chaff is the thin papery skin that comes off the coffee bean. When heated with LDPE, the chaff breaks down, releasing gases such as carbon dioxide (CO2), water vapor (H2O), carbon monoxide (CO), and various volatile organic compounds (VOCs). The Molecular Homogenizer produced no bubbles in the extrudate. (The comparison pictures, footnote 33, are dramatic.[42])
· Black And White SAN Color Concentrate: SAN is hygroscopic. Both carbon black and titanium dioxide are absorb surface moisture. The undried materials were both pelletized. No bubbles were visible in the strand cut pellets. The pellets were then processed in a conventional screw in film line, Fig. 30 and 31. The unprocessed color concentrate pellets were processed on the same film line in Fig. 32 and 33.
Fig. 33 Fig. 32 Fig. 31 Fig.30
In these examples we can begin to think about processing without drying—as none of these examples was dried. The cost of drying all the hygroscopic polymers is immense. It may not be necessary.
Where Do The Bubbles Go? This is a very natural question. Water vapor is not a bubble. There are no bubbles in undried hygroscopic polymer pellets. [43] There is only water vapor--an invisible gas. We only see bubbles (Fig. 31, 33, 37 bubbles breaking into holes) because conventional single screw extruders are terrific, compressive, agglomerating machines. Conventional screws agglomerate the individual H2O molecules in the hygroscopic polymers. When enough molecules are pushed together, they become bubbles in the melt.
The process is similar to cloud formation. Water vapor, the gas in the atmosphere, is invisible. We only see clouds when the gas condenses into visible droplets as they change state from gas to water. In screw processing, the objective of mixing is to keep the H2O molecules separated so that it remains a gas and cannot change state into a droplet that forms a bubble.
Simply put, the Molecular Homogenizer’s mixing is so good that it prevents the water vapor gas (and the other gases in the interstitial space between the polymer molecules) from getting together. The question, “Where do the bubbles go?” is so natural because, prior to the Molecular Homogenizer, we hadn’t imagined such mixing were possible.
Venting Undried Nylon; Venting During Reactive Chemistry; Venting Possibilities
None of the previously discussed hygroscopic polymers produced bubbles. There is a limit to how much water vapor and other gases the Molecular Homogenizer can handle before bubbles are created. Then, venting becomes necessary. In Fig. 12, the orange arrows identified extensive free surfaces. When bubbles form on extensive surfaces, they tend to break as they are stretched, and this allows the gas to escape. The more extensive surfaces (35 in the tested screw), the better the degassing. The better the mixing, distributive and inversive, the greater the chance of bringing bubbles to the surface for degassing.[44]
Here are two examples where the amount of gas in the material created bubbles in the unvented screw and where an atmospheric vent was sufficient to make extrusions without bubbles:
· Nylon: Nylon is very hygroscopic polymer absorbing moisture in the 2% to 9% range. It is expensive to dry polymers. Below, a production nylon blend was processed into film for a week. Drying was necessary. Conditions were well established. The screw was changed to the Molecular Homogenizer and undried material processed. Bubbles were still created, left. A single atmospheric vent was opened, exposing the extensive free surfaces, and the film on the right was produced. It was interesting that:
o The output rate was the same (rpm was increased and extruder starve fed).
o The temperatures were kept the same as the dried material.
o No die adjustments were necessary to produce the same tolerance film—implying the viscosity was the same.
o Pressure was the same as the dried material—implying the viscosity was the same.
Fig. 37 FIG.36 Fig. 35
o Undried PVA With More Than 3% Reactive Agent: PVA with 3% reactive agent was one of the materials where the strand did not show bubbles. However, the experiment also mixed 6, 9 and 12% reactive agent. At 6%, bubbles were created. A single atmospheric vent was opened. No bubbles were seen. The experiment proceeded and no bubbles were seen at 9 or 12%.
o Venting Possibilities: The cost of drying hygroscopic polymers is huge. It may not be necessary even for nylon. Since the Molecular Homogenizer has new and many extensive free surfaces and polymer multiple polymer inversions, since most materials absorb far less moisture than nylon, it may be that PET, for example, will not need drying.
Slowed Water Absorption
The Molecular Homogenizer mixer slows the rate of moisture absorption. This is particularly useful for regrind where it takes more than a couple of hours to grind and reprocess the material. For example, PMMA regrind will absorb moisture in 4 to 6 hours creating problems (bubbles, lower viscosity, property degradation, and roughened surfaces).
The SCFEM processed dried PMMA through a water trough and strand pelletizer on a rainy, summer Friday afternoon in New Jersey. The bag of pelletized material was left exposed to the atmosphere over the weekend and it continued to rain. On Monday afternoon, 72 hours later, the pellets were extruded (undried) in a conventional screw. There were no bubbles. This demonstrates that the time for processing regrind can be substantially extended for product improvement.
After 30 days, the pellets were again extruded in the conventional screw and bubbles were readily apparent.
Moisture Release:
Undried polycarbonate powder was sealed in a paint can with nitrogen to preserve the moisture content. The undried polycarbonate powder was also processed through the Molecular Homogenizer through a water trough and strand pelletizer (no bubbles were seen) before being sealed in a paint can with nitrogen so that the two could be transported. These materials were taken to the moisture analyzer at American Leistritz. Left, the PC moisture level of the powder is 0.328%. Right, the moisture level indicates 0.171%.
Fig. 39 Fig. 38
A moisture analyzer calculates the water percentage by heating the polymer and measuring the weight loss over time. This result is consistent with the previous experiment on slowed water absorption for better processing of regrind. Here, the rate of moisture release, at elevated temperatures, is slowed.
Improved Physical Properties
A Molecular Homogenizer (MH) and a general-purpose control screw processed virgin dried PETG. The pellets were then molded in a family mold and compared at Pennsylvania College of Technology: [45]
• Elongation At Yield:
• The MH screw was 17.9% better than the general-purpose screw.
• The MH screw was 6.7% better than the virgin material.
• Izod Impact Strength:
• The MH screw was 22.4% better than the general-purpose screw.
• The MH screw was -18% worse than the virgin pellets.
The polymer “standard” is that a virgin polymer’s physical properties are degraded with processing. This study shows the virgin polymer’s tensile at yield was improved by Molecular Homogenizer processing. Further, there was substantial improvement over the conventional screw both for tensile and impact.
APPLICATIONS
• Polymer Producers: The Molecular Homogenizer has demonstrated post-reactor improvement in physical properties and rheology—despite an additional processing histories. Since the reactor operation already includes a pelletizing processing history, we expect the same or better results from a Molecular Homogenizer placed at the end of a polymer reactor for pelletizing. This will create polymer with better properties and lessened or no need for drying.
• Extruders and Molders:
• PET Processors:
• No Need To Crystallize: The Molecular Homogenizer’s solids conveying allows amorphous PET with crystalline PET despite its normal clogging behavior.
• Energy Savings: With the increased efficiency of the Molecular Homogenizer’s venting mechanisms, consideration of energy savings is considerable.
• Toll Compounding/Direct Extrusion: The Molecular Homogenizer mixer is a stable, high-pressure pump without the need for a gear pump. Besides improvement to virgin polymers, many applications will not need a toll compounding step.
• Pharmaceuticals: Actives (drugs) are particularly sensitive to mixing for their uniform biological distribution over time. Direct extrusion or the addition of the Molecular Homogenizer to an existing twin can enhance the delivery.
• Reduced Drying Costs: Nylon and other hygroscopic polymers may not need drying.
• Reclaim:
• Slowed Water Absorption Rates: Increased time for processing without need for drying.
• Improved Viscosity: Die flow becomes more consistent for less reclaim.
• Quality Improvement: The Molecular Homogenizer does not show surface roughening, bubbles, or physical property degradation.
• Downsizing: Improved physicals often lead to downsizing with consequent savings.
• Improved Degassing: Inversion and the many elongating free surfaces can remove the need to dry. Atmospheric venting may be sufficient instead of vacuum venting.
• Better Mixing of Particulate: While this paper has concentrated its efforts on virgin materials, mixing of particulate from the SFEM is known.[46]
• Higher output: PP, PFA & rigid PVC powder are known to process at higher rates through the SFEM screw.[47]
· R&D: This new realm of mixedness means opportunities for new products.
· Twin Screw Users: Enjoy enhanced mixing of actives and other small molecules by adding the Molecular Homogenizer extruder to your twin screw.
MOLECULAR HOMOGENIZER CONCLUSIONS
• Mixing Mechanism:
• Zero Pressure: Zero pressure operation optimizes the mixing of shear, elongation and trilongation.
• First Principles: Based on first principles, the combination of one-directional shear, two-directional elongation, three directional trilongation forms an enhanced mixing element.
• Inversion: Mixing inversion creates multiple enhancements to these first principles by turning the flow inside out and bottom to top for heating or cooling at the barrel and for much improved gas extraction.
• Uniform Treatment: All the flow is subject to the shearing and elongation in each element. This is not true of twins.
• Extensive Free Surfaces: These many surfaces make for greatly enhanced degassing.
• Axial Mixing: Great improvement in axial mixing allows creates superior mixing.
• New Opportunities: This new realm of mixing opens many prospects including:
• Energy Savings: The amount of energy and labor spent drying materials is enormous. By reducing this cost, we can help move to green while increasing profits.
• Physicals: Improved physicals can lead to downsizing and increased profits.
• Potentially: Reduction in specific gravity –the notion of using 1.5% less oil and gas to make the same volume of polymer—is a very green idea.
• Simplicity: The mixing mechanism is fixed. It is “plug-and-play.” It doesn’t require the purchase of the many complicated parts (i.e. kneading discs, reversing flights, etc. or their time consuming exchange per material for improved properties).
• Cost: Singles screw extruders have lower capital costs than twins.
THANKS:
This paper would not have been possible without help from many people. I would like to express my thanks for their contributions:
• Tom Cunningham, Bamberger AMCO.
• George Venturini, Randcastle.
• Raj Krisnaswamy, CJ Biomaterials.
• Charlie Martin, and Brian Haight of American Leistritz.
• Kirk Cantor, PhD., Professor, Plastics and Polymer Engineering, Pennsylvania College Of Technology.
• Morgan Bartholomew, Pennsylvania College Of Technology.
• Darryl Nazareth, Easton Technologies.
• Andrea Auchter, NatureWorks.
[1] Link to, https://www.randcastletechnology.com/s/Comparison-Of-Flow-Striations-Of-Various-SSE-Mixers-To-The-Recirculator-And-Elongator-Apdf.pdf
[2][2][2] P. 27, “Where do bubbles go?”
[3] Previous structures are found in the technical papers section link at https://www.randcastletechnology.com/home/papers AFEM is Axially Fluted Elongation Mixer; SFEM: Spirally Fluted Elongational Mixer.
[4] Mixing In Polymer Processing, Edited By Chris Rauwendaal, (Marcel Decker, 1991), 13
[5] Ibid., 14.
[6] Ibid 13.
[7] Polymer Extrusion, Chris Rauwendaal, 4th edition, (2001), 507.
[8] C. Rauwendaal, A. Rios, T. Osswald, P. Gramann, B. Davis, T. Osswald, P. Gramann, and B. Davis, “EXPERIMENTAL STUDY OF A NEW DISPERSIVE MIXER,” Antec 1998.
[9] Ibid., footnote 1
[10] K. Luker, “SUMMARY RESULTS OF A NOVEL SINGLE SCREW COMPOUNDER LUKER PAPER REFERENCE,” Fig. 10, Antec 2007.
[11] Randcastle US Patent 6,932,431 B1.
[12] Link to, “3 Dimensional Mixing An Orderly Mixer For Single Screw Compounding,” https://www.randcastletechnology.com/s/2015-Plastic-Technolgy-Conference-Finished-1.pdf, 7.
[13] Link to, “FACILE TPO DISPERSION USING EXTENSIONAL MIXING,” https://www.randcastletechnology.com/s/ANTEC11.pdf
[14] Mixing In Polymer Processing, Ibid., 14
[15] Shear flow is in one direction. Elongational flow is two directional. “Trilonational” flow is three directional.
[16] Polymer Extrusion, Ibid., 337.
[17] Ibid, 15.
[18] Standard extrusion lines are made right to left and the drawing numeration and flow reflects that convention.
[19] This type of mixing is most often seen during purging and is usually not considered a useful mixing mechanism.
[20] Polymer material flow shapes vary due to viscosity, non-newtonian, shear thinning, and other behavior.
[21] Link to “NEW HIGH OUTPUT VINYL COMPOUNDING SCREW,” Fig. 17, https://www.randcastletechnology.com/s/vinyl1.PDF Best Paper Award, 2007 Vinyl Conference.
[22] Link to: https://www.randcastletechnology.com/s/Quality-High-Output-Single-Screw-Vented-Production.pdf p. 21, Plastics Technology Conference 2023.
[23] Orderly or patterned flow allows regular arrangement of molecules that is superior to chaotic flow: https://www.randcastletechnology.com/s/The-Molecular-Homogenizer-Sept-21-2021-Plastics-Technology-Conference.pdf , Plastics Technology Conference 2021.
[24] https://www.randcastletechnology.com/s/Quality-High-Output-Single-Screw-Vented-Prowduction.pdf , Plastics Technology Conference 2023.
[25] Link to, GREAT MIXING AND SUSTAINABILITY-ELONGATION, 2021, Plastics Technology Conference, 2021, p.9, shows the error per element is perpetuated rather than erased in each element.
[26] See Normal Cylindrical projection, https://en.wikipedia.org/wiki/Map_projection for a detailed explanation.
[27] In order for single screws to compete with twin screw mixing, it’s common to grind and mix the ground feedstock to the extruder. When compressed in a convention single screw, the mixture is as good as the ground mix. Ultimately, this procedure fails as the ground materials cannot be made to the molecular level. Before that, the bulk density may decrease so much that the solids conveying section fails or the production rate becomes uneconomical.
[28] It may be the objective of the polymer producer to have imbalanced mixing to enhance some particular property improved over others by adjusting the mixer design (such as by altering length, width, height of the mixer components.
[29] This is more readily pictured, p, 4, https://www.randcastletechnology.com/s/HME-Book-Initial-Submission-2011-F-with-Molecular-Homogenizer-addendum.pdf. On a large extruder, the input mixture is much better since every flight could contain both colors.
[30] Link to, https://www.randcastletechnology.com/s/Antec-09-INVESTIGATION-INTO-A-HIGH-OUTPUT-POLYPROPYLENE-SCREW.pdf
[31] A similar result is shown for RPVC, link to https://www.randcastletechnology.com/s/PVC-FLOW-STREAMS-IDENTIFY-ELONGATIONAL-FLOWS.pdf
[32] Attempts were made to extract a starve fed screw but the stress of extraction made the remains crumble.
[33] The actual dispersion (breakup) of pigment agglomerations is shown in detail at, https://www.randcastletechnology.com/s/Great-Mixing-And-Sustainability.pdf p. 18-19.
[34] Ibid. Footnote 12.
[35] This chart and additional pictures were presented at Antec for the paper. The twin screw comparison was made after the paper’s submission and incorporated into link https://www.randcastletechnology.com/s/2015-Plastic-Technolgy-Conference-Finished-1.pdf
[36] Link to, https://www.randcastletechnology.com/s/PLA-Processed-Through-Dynamic-Multiplicative-Mixer-002.pdf is the presentation given at Polymer Technology.
[37] Often, we process novel polymers using conventional screws where the objective is the greatest output possible. Maximum amperage is often the limiting variable, and the 13.5 amps is regularly exceeded.
[38] These results are thanks to Raj Krisnaswamy, CJ Biomaterials.
[39] The summary is Randcastle’s rather than NatureWorks.
[40] It has been suggested that, since there are no bubbles, that the Molecular Homogenizer is venting water vapor out the hopper and “back-venting.” Our experience with conventional screws is that this does not happen. Instead, telltale bubbles appear first in the extrudate; if you continue to process this way, the water vapor is squeezed from the pellets during compression and condenses on the water-cooled feed section. Eventually, this forms a puddle in the feed and the solids conveying stops. When the hopper is removed, a water/pellet mix is seen. If the water didn’t condense, then the steam would leave the hopper and make the pellets wet. Single screw hopper venting is not a known degassing mechanism for water vapor.
[41] Link to additional pictures, https://www.randcastletechnology.com/s/The-Effects-of-Molecular-Homogenization.pdf p. 15-20.
[42] https://www.randcastletechnology.com/s/The-Effects-of-Molecular-Homogenization.pdf p. 13-14
[43] To be clear, pellets can have a hole, usually centered along the strand axis of a cylindrical pellet. Or, pellets can have what are called vacuum bubbles. Both occur when the outside of the pellet is rapidly cooled which freezes the outside before the inside. The stress creates a bubble. In rare cases, air can be entrained in the feedstock that can then be extruded. None of these “bubbles” is the result of water vapor in the pellet.
[44] A conventional degassing two-stage single screw is extremely limited by comparison as it mixes so poorly, has no extensive surfaces, and no inversive mixing.
[45] Link to, Capstone Research and Testing for the Molecular Homogenizing Screw, Morgan Bartholomew, Pennsylvania College of Technology, 5/14/24.
[46] Link to https://www.randcastletechnology.com/s/2015-Plastic-Technolgy-Conference-Finished-1.pdf
[47] Ibid footnote 28,29. The information about the PFA increase is unpublished. Normally, PFA is only processed at rpm up to around 25 rpm. Processed on an RCP-0625, a straight line output with rpm was produced at up to 92 rpm.