Enhanced, Seamless Ultrasound Introduction Unit for Thermoplastic Melt Treatment

Abstract

Ultrasonic power stimulation of polymers has been employed to modify their properties for various industrial applications. It is used to disperse particles in polymers or to impregnate continuous fiber-reinforced filaments. These processes operate under extreme conditions, involving temperatures exceeding 400 °C and pressures reaching up to 60 bars. Traditional sound introduction systems rely on moving horns within cavities, which pose significant challenges in maintaining clean environments. The moving horn makes it inconceivable to seal such sound introduction systems. To address these limitations, an innovative tool that incorporates only sealed components capable of introducing sound energy without employing cavities was developed. This technology focuses on optimizing eigenfrequencies to efficiently transfer motion into the polymer channel while minimizing disturbing vibrations and sealing issues. The simulation results reveal the mode of operation between the fourth and fifth resonance of the tool. The measurements of the tool validate this theory, indicating a range of operation between 18.5 kHz and 19.5 kHz. With an amplitude of 15 μm, resulting in a minimum power introduction of 224.03 W/cm^² in PLA, the average flow rate gain is 13.6%. This new design reduces the risk of blockages and damage in the processed goods and minimizes the force required to process the polymer.

1. Introduction

Polymeric treatment has been implemented in various industrial processes, i.e., applications that involve dispersing particles or impregnating fibers in a polymeric melt [1,2]. However, these polymers are favored to be used in performance applications. Either particles or fibers can change the material properties dramatically. The base polymers need to be modified in order to utilize the benefits of the composite between the polymers and additives [3,4]. This process is mainly used to improve their properties [5,6,7]. Ultrasound technology plays a crucial role in enhancing the effectiveness of this process [8,9]. Research has demonstrated that ultrasound treatment enhances particle dispersion in melts and fiber impregnation [10,11]. One specific use case is the production of continuously reinforced filaments for 3D printing [12]. Furthermore, by introducing shear forces and energy into the polymer, its viscosity is reduced, easing the flow through narrow channels typically found in processing tools [7,13]. However, current implementations of these systems introduce secondary issues that vary in disruptive impact depending on the material system used. One notable problem associated with direct contact horns is the creation of edges and undercuts in tools, potentially damaging semi-finished products, facilitating blockages, or increasing forces required for material transfer through the system. The mechanical setup of existing ultrasonic sound introduction units consists of a horn in direct contact with the melt, transmitting energy through direct impact during operation at frequencies of up to 40 kHz and amplitudes reaching 30 μm. The relative movement of the horn compared with the surrounding components poses significant sealing challenges [14]. Additionally, the ultrasonic treatment reduces the viscosity within the designated cavities and enhances the secondary flow (see Figure 1). Various approaches have been developed to address these issues:

• Attempting to achieve tight seals, resulting in relatively low-amplitude movements [15].
• Creating a cavity in part of the horn and clamping it in its first wave node, resulting in the necessity of a dead space filled with the polymer. However, without the implementation of a venting system, the derogated polymer flows back into the primary melt stream [16].

From an economic perspective, these sealing solutions increase costs without adding value to the final product. Therefore, a totally sealed tool would address these challenges. Such a tool could potentially be achieved by either sealing the horn or by modifying the mechanism of sound introduction, which is challenging to achieve at high frequencies. But still, theoretical concepts for such approaches remain largely unimplemented due to practical challenges in achieving tools that operate in eigenforms at applied frequencies and amplitudes while also maintaining the ability to insert sound energy and seal against the rest of the production line without transmitting vibrations [17].

To realize a seamless mixing element, the polymer should stay encapsulated while receiving sound energy from the ultrasonic transducer. To achieve this, a modal design to arrange an encapsulated ultrasonic sound introduction unit is presented (see Figure 2).

The new design ensures ultrasonic power transfer to the polymer through a counter-phase swinging mass that compresses the channel. Solid joint arrangements allow the necessary deformation without creating cavities. The absence of cavities prevents the polymer from generating secondary flows, which enables the tool to prevent edges or undercuts, reduce potential damage to semi-finished products and the risk of blockages, and minimize the forces required for material transfer through the system.

2. Materials and Methods

2.1. Material

Two polymers were processed in the subsequent experiments. The first material was “Ingeo™ Biopolymer 2003D” manufactured by “NatureWorks” (Plymouth, MN, USA), which is based on polylactic acid (PLA) with a melting point of 210 °C. Throughout this research, only the base material PLA will be referenced. The second material employed was “Dahlpram^® 009” produced by “Airtech Europe Sarl” (Niederkorn, Luxembourg). This product is specifically designed for cleaning purposes within the 3D-printing sector and exhibits a wide range of processing temperatures (200–400 °C). In the following study, this material will be referred to as the cleaning material. The final tool was manufactured using Toolox44, a high-performance steel alloy characterized by its exceptional dimensional stability during processing. This choice of material enabled the efficient production of the tool itself, ensuring optimal precision and consistency throughout the production process.

2.2. Modeling

A modal design minimizes the structural rigidity to optimize sound power transmission into the polymer being mixed. Figure 3 shows a graph illustrating the dynamic stiffness over frequency of a structure near its first eigenfrequency. The lowest point of stiffness represents the resonance frequency and decreases the structure’s stiffness to the damping of the structure (a tuning fork’s functional principle) [18,19,20]. This effect allows us to maximize the transmission of sound power from the ultrasonic transducer (Hielscher, type UIP1000hDT transducer/sonotrode, Teltow, Germany) to the polymer through a closed channel.

The proposed closed seamless mixing element employs an encapsulated polymer system that allows transmitting the ultrasonic power from an ultrasonic transducer. The device includes two specially designed masses configured to exhibit the modal swinging mode at the operational frequency. These masses compress the channel that encloses the polymer, as illustrated in Figure 4. The mixing tool is engineered to operate in resonance within a predetermined frequency range (19.2–19.8 kHz) set by the ultrasonic transducer. The key design considerations include (1) mass dimensions, (2) the integration of a cooling system, (3) the decoupling of vibration, and (4) maintained thermal conductivity.

These considerations focus on the compatibility with existing production lines. The oscillating masses were designed with a 15 mm key width and cooling holes to function as a heat break. To mitigate potential interference with sensitive components due to vibrations, decoupling holes were drilled along and crosswise to the channel. However, heat transfer remains possible due to the solid joint connections to the channel. Harmonic analysis conducted using ANSYS 2022 R2 software demonstrated the expected deformation of the masses during operation. The material model was based on the material “Toolox44”, with the following properties: a density of 7.85 g/cm³, an E-modulus of 210 GPa, and Poisson’s ratio of 0.31. The expected deformations were small and allowed a linear calculation with a mesh made of 1.9 million knots and 1.3 million elements. The left-hand side of Figure 5 presents the three-dimensional representation of the simulation model, revealing the displacement behavior of the mass positioned on the opposite side of the sound introduction unit. This visualization focuses attention on the boundary conditions influencing the dynamic response of the system under investigation. The model features fixed end-mounting conditions for the polymer-filled channel and directional sound introduction perpendicular to one of the two masses. The side view on the right side of Figure 5 illustrates the effect of decoupling holes parallel to the polymer channel during compression. These holes facilitate the deformation of the channel without transferring vibrations to the heating elements. The decoupling holes perpendicular to the channel separate the sound introduction area from the production line, visible in the three-dimensional representation on the left side of Figure 5. The frequency response evaluation was facilitated by assessing the speed of the counter-mass, influenced by the force introduction. To validate the initial ANSYS model, three-dimensional printed prototypes were employed.

To validate the simulated results against experimental data before manufacturing the tool out of Toolox44, the material model utilized in the simulation was modified to match the “VeroBlue” material properties from Stratasys (Eden Prairie MN, USA), with the following characteristics: a density of 1.19 g/cm³, an E-modulus of 2.5 GPa, and Poisson’s ratio of 0.42.

2.3. Experimental Setup

Figure 6 illustrates the impedance test stand utilized for determining the dynamic properties of a VeroBlue material-based model (printed with Stratasys, Objet260 Connex3). In this configuration, a shaker (Bruel & Kjaer, Type 4809 Vibration Exciter/Shaker, Nærum, Denmark) excites an impedance sensor (Bruel & Kjaer, Type 8001 Impedance Head). This sensor is designed to measure both force and velocity alongside the sound introduction unit. Simultaneously, a vibrometer (Polytec GmbH, Type PDV-100 vibrometer, Waldbronn, Germany) is employed to record the velocity of the counter-mass that is on the opposite side. The dataset is recorded with an FFT analyzer (Onosokki, type CF-7200, Yokohama, Japan), while the signal generator (Yokogawa, type FG300) controls the shaker amplifier (Kepco, type “Bipolar Operational Power Supply/Amplifier”, Samseong-dong, Seoul). And the force and velocity signal from the impedance sensor is amplified by the signal conditioner (Bruel & Kjær, type “Nexus Conditioning Amplifier”). The identification of the eigenfrequencies can be performed by plotting the velocity over frequency. Peaks in velocity occur in modes of interest. The velocity of the counter-mass, resulting from the force introduction, can be evaluated as a frequency response. From the simulation validated by 3D-printed models, four crucial frequencies arise in the sonotrode excitation range (see Figure 7). The third eigenfrequency (from the left) is a bending mode of the sound introduction unit, corresponding to a bending mode, potentially causing damage to the heating elements. The fourth eigenfrequency represents the optimal operational mode. The following minimum has similar deformation properties. However, the compression of the channel is impaired because the counter-mass has a minimal displacement. The fifth eigenfrequency induces oscillation along the channel and should, therefore, not be stimulated.

Thus, the ideal operational frequency falls between the fourth eigenfrequency and the subsequent minimum of the counter-mass speed. The design and manufacturing of the sound introduction unit from the material Toolox44 for integration into the manufacturing process were carried out based on the measurement results of the 3D-printed models.

3. Results and Discussions

Figure 8 shows the final results from the measurement of the dynamic properties of the sound introduction unit made of Toolox44. These measurements were subsequently utilized within the impedance test stand described in Section 2.2. The left graph plots the force detected by the impedance sensor, while the right graph illustrates the resulting speed of the counter-mass. The excitation signal was restricted to the frequency range of 15 kHz to 25 kHz, resulting in the absence of the first two resonances depicted in Figure 7. Within this frequency range, however, the three resonances from the simulation were also observable in the measurement. Discrepancies between the simulation and measurement outcomes can be attributed to the simplified simulation without damping. The critical frequencies are distinctly visible in Figure 8 in the right graph at 17.2 kHz, 18.5 kHz, 19.5 kHz, and 19.8 kHz. Hence, the operational frequency ranges between 18.5 kHz and 19.5 kHz. Installed in the production line (see Figure 9), the tool was driven by the sonotrode without the polymer, which found its frequency of operation at 19.1 kHz, which is within the functional frequency range. This frequency is visually indicated in Figure 8. The peak observed in the graphical representation on the right-hand side of Figure 9 corresponds to the operational frequency, quantified as the velocity of the counter-mass measured with a vibrometer (Polytec GmbH, Type PDV-100 vibrometer). During the experimental trials, the sonotrode control system reported a conversion efficiency of approximately 100% of the total power without any indication of power loss. It is important to consider that this seemingly perfect efficiency may be influenced by the measurement methodology employed at the amplifier stage. Further investigation into the measurement techniques and data acquisition processes would be warranted to fully understand and validate these findings. Subsequent thermal treatment of the cleaning material (250 °C) and polylactic acid (PLA) (220 °C), facilitated by heating elements deployed in the designated boreholes, enables the production of initial semi-finished products. The air-blown cooling holes successfully prevent the temperature transfer to the sonotrode during operation. The ultrasound excitation of polymers within the sound introduction unit is achieved using an amplitude of 50% of the maximum displacement of 30 µm. This results in a power consumption of approximately 50 W for the cleaning material and 94 W for PLA (polylactic acid). Figure 10 presents two graphical representations illustrating variations in displacement from 20% to 50% (blue curves) and corresponding total and net power outputs (orange and black curves), while the operating frequency is around 19,500 Hz.

The change in frequency may be attributed to temperature changes, but these remain within the operational range. Analyzing the trend of achieved power highlights the disappearing stiffness of the sound introduction unit in consideration of the achieved displacement. This relationship indicates inherent power losses within the sonotrode. The sonotrode exhibits significant variations in power consumption when processing different materials, specifically between the cleaning material and the PLA. Notably, doubling the amplitude from 20% to 40% requires a power increase of approximately 27% (from 34 W to 43 W) for the cleaning material and 57% (from 42 W to 66 W) for the PLA. This disparity in energy consumption suggests that the excess power is likely converted into acoustic energy directed toward the PLA. Considering the deformed area within the processing channel, which measures 0.1964 cm² (refer to Figure 11 on the right), the calculated power densities at 50% of the maximum amplitude are:

• 254.58 W/cm² for the cleaning material at 50 W;
• 478.62 W/cm² for PLA at 94 W.

The difference in power consumption between these materials yields a sound power transmission of 44 W, resulting in a power density of 224.03 W/cm².

This power transfer generates additional thermal and mechanical stresses on the polymer, leading to reduced viscosity. The decrease in viscosity manifests as increased flow rates while maintaining temperatures and a constant pressure of 20 bars at the entrance of the sound introduction unit.

Table 1. The amplitude-dependent flow rate measurements for PLA. The comparative measurements with and without ultrasonic treatment demonstrate the flow rate increase caused by the sound power introduction.

Measurement	0% Ampl. [g/min]	50% Ampl. [g/min]	Gain [%]
1	6.20	7.00	12.9
2	6.30	7.20	14.3
3	6.06	6.88	13.5

presents the amplitude-dependent flow rate measurements for the PLA, starting with comparative measurements without ultrasonic treatment as the initial state. After each comparative measurement follows a measurement with ultrasonic treatment. Across three measurements, the average flow rate gain is 13.6%, indicating the presence and efficacy of the ultrasonic treatment.

4. Conclusions

The purpose of this sound introduction unit was to introduce the sound power produced by the sonotrode into the channel filled with the polymer while keeping the channel closed and decoupling the vibration from the production line. This was achieved by creating a counter-phase motion in opposing masses, which compressed the channel and decoupled holes drilled along and crosswise to the channel. A design guideline was drafted, and the challenges between the simulated design, the modal analysis of the tool, and the behavior in the functional system were given. It was outlined that there was an offset between the designed and final frequencies of the operating mode. However, the measured total power and net power overlayed onto each other, caused by the modal design of the ultrasonic sound introduction unit. This may have interfered with the measurement methodology employed in the amplifier stage. Further investigation into the measurement techniques and data acquisition processes would be warranted to fully understand and validate these findings. However, this study demonstrates significant differences in power consumption and acoustic energy conversion during the ultrasonic processing of various materials. Specifically, the sonotrode exhibited substantial variations in power usage when processing the cleaning material versus polylactic acid (PLA). These findings suggest that the excess power is likely converted into acoustic energy directed toward the PLA, resulting in increased thermal and mechanical stresses on the polymer. The calculated power densities at 50% of the maximum amplitude revealed substantial differences between the materials, with the PLA exhibiting higher values (478.62 W/cm² at 94 W) compared with the cleaning material (254.58 W/cm² at 50 W). This disparity led to sound power transmission directly to the PLA of 44 W, generating a power density of at least 224.03 W/cm². These observations are supported by the experimental measurements showing average flow rate gains of 13.6% across three measurements after implementing the ultrasonic treatment while keeping the temperatures (PLA: 220 °C; cleaning material: 250 °C) and the pressure at the entrance of the sound introduction unit constant (20 bars). The conducted research demonstrated the functional ability impressively. A tool was designed to decouple the movement of the tool itself from the areas of the heating element and sensor integration, as well as the sealing to the rest of the machine.