Progress 02/15/19 to 02/14/24
Outputs
Target Audience:Ice cream freezers have traditionally been "black boxes" where what goes on inside is not very well understood. Thus, design of these freezers remains primarily trial and error. The primary target audience for this work are manufacturers of ice cream processing equipment, who desire a better understanding of freezer operations. Ice cream manufacturers and research scientists will also benefit from the knowledge developed. Changes/Problems:Numerous problems have arisen in the early stages of this project: • We experienced a significant delay in getting the newly donated ice cream freezer from Tetra Pak into the lab and then getting it commissioned. There was a significant learning curve for training with the students involved. • On the modeling side, significant difficulties arose in finding suitable personnel. An international search did not uncover an adequately trained post-doc, so the decision was made to utilize a second graduate student on the modelling side. • To further impede the experimental plan, the pandemic shut the lab down for an extended time, meaning that no experiments were conducted for several months. Even then, progress was slow during the early years because of covid limitations. These delays resulted in the lack of any progress on Objective 3, attempting to optimize freezer operations. We will continue to work with Tetra Pak to develop these ideas beyond the scope of the current project. What opportunities for training and professional development has the project provided?One MS student in Food Science, Katherine Helbig, started her training working on this project. She was trained on operating the new freezer donated by Tetra Pak Hoyer, developed methods for making sorbet with the desired characteristics, conducted experiments, and collected data for subsequent analysis. A PhD student then picked up the work to complete the ice cream studies. Several undergraduates were recruited to help with both data collection and analysis. One MS student in Biological Systems Engineering developed the CFD model. How have the results been disseminated to communities of interest?In addition to publications and presentations, this work has been shared at the UW-Madison Frozen Dessert Center technical symposium each year since its inception. Attendees of this symposium are ice cream manufactures, ingredient suppliers and equipment manufacturers. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Objective 1. Effects of dasher design on SSF operation and product attributes. Several phases of research were covered under this Objective. First, the effects of 5 different dasher designs and various operating parameters on residence time distribution and ice development in sorbet were conducted. Second, the start-up dynamics of the freezer were studied. Third, the properties of ice creams made at different conditions were evaluated. Five different dasher designs were studied: (1) solid dasher, 0.55 m2 of surface area and 64.3% volume displacement, (2) multi-dasher and solid beater, 0.74 m2 and 54.1% displacement, (3) standard dasher and solid beater, 0.70 m2 and 51.2% displacement, (4) multi dasher and open beater, 0.66 m2 and 28.9% displacement, and (5) standard dasher and wing beater, 0.62 m2 and 262.2% displacement. The mean residence time of fluid within the SSF running sorbet varied from a low of 100 s for the solid dasher to a high of 191 s for the standard dasher and wing beater. In general, mean residence time increased with the volume displacement of the dasher design. The spread of residence times also increased with increasing volume displacement. Surprisingly, for unaerated sorbet, there were no significant differences in mean ice crystal size or distribution of ice crystal sizes as a function of the dasher design, despite the large differences in residence time distributions. Apparently, the freezer is somehow compensating for different flow conditions to produce similar ice crystal sizes. The start-up sequence of making ice cream was investigated with the standard dasher and wing beater since this set-up had the longest mean residence time. Upon system start-up, draw temperature and product viscosity (torque on motor) reached the steady-state values within the first 5 minutes of operation. Development of steady-state overrun, however, took 10 minutes to finalize, with an overshoot of about 15% in overrun when targeting a final value of 75%. Ice crystal size and distribution stabilized after about 5 minutes of operation, similar to the build-up of ice mass as seen from the viscosity (torque) reading. Air cell sizes decreased during the first measurement at 3 min to stabilize after 10 min of operation, following the pattern of the overrun development. The extent of fat destabilization (partial coalescence) increased over the first 5 minutes, peaking at nearly 80% before decreasing to about 70% and stabilizing there after 10 minutes of operation. Operating parameters and dasher design on residence time distributions and ice cream structure: Dasher RPM: Dasher RPMs of 200 and 300 were evaluated across the 5 different dasher designs. Mean residence time distributions in ice cream varied similarly to the sorbet results noted above and were not significantly influenced by variations in RPM. Mean ice crystal and air cell sizes were not affected by dasher design but were slightly smaller at higher RPM. The extent of fat destabilization was seen to vary with dasher design, with higher displacement dashers giving high fat destabilization, nominally because of the higher shear rates within the barrel. Dasher RPM also influenced extent of fat destabilization, with higher RPM causing significantly greater fat globule clustering because of the increased shear rates within the barrel. Overrun The effects of overrun (50, 75 and 100%) on residence time distribution and microstructure development for each of the dasher designs was investigated. Here, in addition to the effects of dasher design seen before, overrun was also seen to have a significant effect on mean residence time and overall distribution of residence times. An increase in overrun caused a significant increase in mean residence time across all the dasher designs. With higher overrun, there was less actual mix within the barrel. Increasing overrun did not affect mean ice crystal size although there was a slight decrease in mean air cell size as overrun increased. Increasing overrun caused a significant decrease in extent of fat destabilization, as the shear rates within the barrel increased with higher air content. Throughput rate Flow rates of 200, 300 and 400 L/h were studied for each of the 5 different dasher designs. As expected, increased throughput rate caused a significant decrease in mean residence time and overall distribution of residence times. However, there was only a small effect of throughput rate on decreasing ice crystal size and no effect on air cell size. The main effect of increased throughput rate was to decrease the extent of fat destabilization. Apparently, the time spent within the barrel played a primary role in causing partial coalescence of fat globules. Objective 2. Design and operation of SSF using CFD simulations. In this study, the fluid flow and heat-and-mass transfer that occur inside an SSF were visualized using a CFD simulation model. The validity of the simulation outcomes was determined by means of two experiments: an axial temperature profile analysis of the inside of an SSF and an analysis of the residence-time distribution of a sorbet mixture as it was transported through the SSF. The axial velocity dominated the flow initially as the fluid flowed out of the inlet pipe into the larger cylinder. Thereafter, this flow was mainly dominated by its rotational movement, which achieved a maximum tangential velocity of 1.81 m s-1, while the mean axial velocity was 0.0053 m s-1. The maximum velocity occurred along the wall of the cylinder, which was consistent with the actual tangential velocity of the blades. The mixing patterns within the flow were characterized by the spatial distribution and density of velocity vectors. Specifically, regions where velocity vectors were closely packed indicated more intense mixing activity. This dense arrangement of velocity vectors signifies areas of high fluid momentum exchange, illustrating how the flow's mixing behavior is intricately influenced by the local velocity field. Centrifugal forces drove the mixing of the product's ingredients as the mix flowed along the surfaces of the dasher, the wing beater, and the cylinder. As it flowed along the cylinder wall that had been scraped, the product was depicted by the recirculating motion of the velocity vectors that occurred between the blade and the cylinder wall. The magnitudes of the vectors were positively correlated with the temperature, and this result was consistent with the nature of shear-thinning non-Newtonian fluids. Further examination of the data by means of shear rate profiles established that the highest shear rates occurred as a result of the blade's scraping. Although the complex phenomena that occur inside a scraped surface freezer will be influenced by many different parameters (dasher speed, mass flowrate, geometry, etc.), the present study produced CFD outcomes that successfully characterized the complex mixing process. In addition, the CFD simulations were able to generate a temperature profile as well as a product residence time that were comparable to those achieved by the physical experimentation, thus validating the accuracy of the CFD model. With the developed model, the complex flow patterns that occur inside an SSF were easily identifiable by means of visual contours that would otherwise have been physically impossible to discern. Objective 3. Develop "optimal or enhanced" SSF designs for frozen dessert manufacture Unfortunately, because of the delays and impediments to this project during covid years, we were unable to make any progress on this aspect. Discussions are ongoing with Tetra Pak on how our results influence their design decisions.
Publications
- Type:
Journal Articles
Status:
Awaiting Publication
Year Published:
2024
Citation:
Panditharatne, D., Sipple, L., Jung, S., Hartel, RW., Choi, C.Y., Application of computational fluid dynamics to improve scraped surface freezer operation for producing frozen desserts, J. Amer. Soc. Agric. Biol. Eng.
- Type:
Theses/Dissertations
Status:
Accepted
Year Published:
2024
Citation:
Gallagher (Sipple), Lauren, Effect Of Dasher Design on Residence Time and Microstructure of Frozen Desserts Produced with a Continuous Scraped Surface Freezer, PhD Dissertation, Food Science
- Type:
Theses/Dissertations
Status:
Accepted
Year Published:
2023
Citation:
Panditharatne, Dimuth, Application of computational fluid dynamics to improve scraped surface freezer operation for producing frozen desserts, MS Biological Systems Eng.
|
Progress 02/15/22 to 02/14/23
Outputs
Target Audience:The results of this work will benefit all ice cream scientists around the world as well as ice cream manufacturers. Of particular interest, manufacturers of ice cream equipment will benefit from insight into design choices for the working parts of ice cream freezers. Changes/Problems:As noted previously, the pandemic has caused this project to be delayed considerably, hence the need for a second extension. Although we have been productive this past year on both fronts, we still need additional time to complete the work as planned. What opportunities for training and professional development has the project provided?To date, we have had two MS students complete their degrees working on this project, one in Food Science and the other in Biological Systems Engineering. One PhD student remains working on this project in Food Science, with the aim to complete Objective 2 work. Several undergrad students have been helping with this project, mostly as assistants when the freezer is being ran and for their work on analyzing structure distributions (air, ice). As noted, they are learning the practical aspects of food manufacturing, including process engineering and safety/sanitation. How have the results been disseminated to communities of interest?This year, we have presented aspects of this work at both ASABE and CoFE meetings, as well as at our local Frozen Dessert Center conference. We plan to present recent results at the upcoming ICEF meeting in Nantes, France. What do you plan to do during the next reporting period to accomplish the goals?Objective 1: We will further explore (i) key design elements including the dasher, the blades, the rotor and cylindrical wall and (ii) key operational conditions such as rotational speed and mixing patterns and corresponding heat and mass transfer. We will summarize the computational outcomes and present the outcomes in an international conference in July 2023 and publish the final results in a refereed journal. Objective 2: The factorial design of ice cream making with the different dasher designs will begin shortly and continue through approximately the end of the year. Objective 3: In the next few months, we will have a meeting with all involved parties, particularly the freezer manufacturer, to discuss the work planned for this objective. Covid has substantially impeded progress on this project and this objective, as stated, is probably a little beyond what we can accomplish now, even with the additional extension. The group will decide how best to meet the project objectives in the most reasonable way.
Impacts
What was accomplished under these goals?
Objective 1: Fluid flow and the heat and mass transfer that occurs inside a scraped surface freezer was studied by using a CFD simulation that was validated in two experiments, an axial temperature profile analysis of inside the SSF and a residence time distribution study of sorbet as it was transported through the SSF. The complex phenomena that occur inside a scraped surface freezer is influenced by many different parameters that include dasher speed, mass flowrate, geometry, etc. However, by carefully simplifying and choosing the correct input boundaries, the CFD model created in this study was able to successfully characterize freezing and mixing. Consequently, the simulation predictions compared favorably with the experimental measurements for the residence time distribution, although the predicted axial temperature variation did not perfectly match the experimental results. We feel this is more the error in the measurements than the model predictions. Moreover, by means of the post-processing results it was possible to visualize and thus analyze the phenomena occurring inside the SSF. Objective 2: Residence time distributions (RTD) have been measured in sorbet for each of the 5 different dasher designs supplied by the freezer manufacture. In general, the lower the volume displacement of the dasher/beater system, the longer and more spread was the RTD of dye passing through the freezer. That is, the dasher design with the greatest solid displacement of fluid within the freezer (lowest liquid hold-up) had the shortest average residence time and the narrowest distribution of residence times. In preparation for conducting a factorial experimental design with ice cream made with different dasher designs, we have been conducting steady-state experiments to determine how long the freezer has to operate prior to stabilization of the structures. This has proven to be quite an interesting set of trials, which will undoubtedly provide value to both equipment manufacturers and ice cream makers since the start-up process is one that creates a substantial amount of waste. Preliminary results show that both ice and air cell distributions stabilize within the about 10 minutes of when a frozen products begins to exit the freezer. After that, for the next 60 minutes or so, ice and air structures were quite consistent. The most interesting developments, with results still being analyzed, involve the earliest stages of freezing, where the foam apparently transitions from protein stabilized immediately upon product appearing at the draw to fat globule stabilized within just a few minutes. We are still analyzing the fat globule changes.
Publications
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2022
Citation:
Gallagher, L, Rankin, SA, Hartel, RW, Panditharatne, D, Choi, C, Alvarez, G, Feasibility of CFD application to improve scraped surface freezer operation and design for frozen desserts, Conference on Food Eng, Raleigh NC (2022)
- Type:
Theses/Dissertations
Status:
Accepted
Year Published:
2022
Citation:
Panditharatne, ND, Application of CFD to improve scraped-surface freezer operation for producing healthier frozen desserts, MS Biol. Syst. Eng, University of Wisconsin-Madison
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2022
Citation:
Panditharatne ND, L Gallagher, R Hartel, CY Choi, Design and Performance Assessment of Scraped Surface Freezer Designs for Producing Healthier Frozen Desserts using CFD, ASABE Annual International Meeting, Houston, TX.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2022
Citation:
LR Gallagher, K Helbig, S Rankin, G Alvarez, and RW Hartel. 2022. The effect of dasher design on residence time and microstructure of frozen desserts produced with a continuous scraped surface freezer. Poster presented at Frozen Dessert Center Conference, Madison, WI.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2022
Citation:
Panditharatne ND, L Gallagher, R Hartel, CY Choi (2022) Design and Performance Assessment of Scraped Surface Freezer Designs for Producing Healthier Frozen Desserts using CFD, Poster presented at Frozen Dessert Center Conference, Madison WI.
|
Progress 02/15/21 to 02/14/22
Outputs
Target Audience:The results of this work will benefit all ice cream scientists around the world as well as ice cream manufacturers. Of particular interest, manufacturers of ice cream equipment will benefit from insight into design choices for the working parts of ice cream freezers. Changes/Problems: As noted last year, this project has been off to a rough start. There were complications in finding a suitable post-doctoral researcher for the modeling side and we ultimately decided to switch to a second graduate student for that work. Of major concern was the pandemic, which began shortly after this project started. The six month shut down and slow return to business caused significant delays in the experimental side of things. What opportunities for training and professional development has the project provided? After graduating one MS student in Food Science last year, there are currently two graduate students working on this project. One is a PhD student in Food Science working on freezer operations and the other a MS student Biological Systems Engineering, who is conducting the modeling studies. In addition, there are several undergrad students in Food Science who actively assist the PhD student in freezer operations. They are learning practical aspects of food manufacturing, including safety/sanitation and process engineering. How have the results been disseminated to communities of interest?We had a poster presentation at the 2021 ASABE meeting and have another accepted for presentation at the 2022 ASABE meeting this summer. This conference is for engineers. In 2021, we also had poster presentations at our annual Frozen Dessert Center Technical Symposium in Madison, WI. Attendees include technical representatives from various ice cream manufacturers, ingredient suppliers and equipment suppliers. What do you plan to do during the next reporting period to accomplish the goals? We are in the process of beginning the ice cream studies where we residence time distributions, microstructure (ice, air, gat globules) development, and temperature profile for different dasher designs and operating conditions. In addition, the computer model is being refined so that results from the model can be compared to experimental measures of temperature profile and residence times.
Impacts
What was accomplished under these goals?
After a rough start to the project (due to equipment issues and the pandemic), this past year has seen significantly more progress on Objectives 1 and 2. The Tetra Pak C700 series freezer was successfully installed and commissioned. Five different dasher configurations were made available by Tetra Pak to study the effects of dasher design on frozen desserts. A preliminary study on freezing of sorbet was completed. The different freezer configurations were studied at different operating conditions with ice crystal size distributions measured and compared. Interestingly, only small differences in mean ice crystal size were found, nominally because the freezer itself automatically sets conditions to ensure good product. While this was a surprise, to some extent, this confirmed that the algorithms used by the freezer manufacturer are indeed robust. The results showed a trend toward slightly larger ice crystal sizes for the dasher designs that had longer average residence times. Since then, we have initiated residence time distribution studies on sorbet, with the assistance of Tetra Pak Hoyer, who have provided assistance in developing a dye injection strategy. These experiments have recently been completed to compare different dasher designs and the data are being analyzed for possible publication. Preliminary results show that, as expected, more open dasher designs with internal mixing show a longer average residence time and have a greater spread of residence times. With the help of Tetra Pak, we have also measured temperature profile along the axial dimension of the freezer, data that is needed for verification of the computational model. Work on developing the CFD model has progressed slowly, in part because of the inconsistency in personnel. With a new grad student in place, things have progressed to the point where we have begun generating model data for comparison with experimental results. Preliminary results suggest that the temperature profile for the dasher design that was modeled follows closely with the experimental value, suggesting that the model is sufficient. Continued modeling for particle flow profile will generate residence time distributions for comparison with the measured values.
Publications
- Type:
Theses/Dissertations
Status:
Published
Year Published:
2021
Citation:
Helbig, K. (2021). Effects of scraped surface freezer dasher design and operating parameters on ice crystal size in sorbet, MS Thesis, University of Wisconsin-Madison.
|
Progress 02/15/20 to 02/14/21
Outputs
Target Audience:All ice cream manufacturers and research scientists Changes/Problems:After a difficult year and a half to start the project (see previous Project Report), it seems we are making better progress on both Objectives 1 and 2. However, we plan to request a no-cost extension and re-budget categories to allow us to pay staff to complete the project. What opportunities for training and professional development has the project provided?One MS student in Food Science, Katherine Helbig, completed the degree requirements and defended her thesis successfully based on this project. A PhD student in Food Science has begun work on the second phase of Objective 1. In BSE, a new grad student started working on Objective 2, modeling flow properties in the SSHE. He is already familiar with CFD, but is learning the food side of processing. Several undergrads have been involved with freezer operation, so are learning practical aspects of food manufacturing, from safety/sanitation to process engineering. How have the results been disseminated to communities of interest?• Poster presentation on CFD modeling at ASABE 2021 annual meeting (virtual) • Poster presentations on sorbet work and preliminary freezer trials on ice cream at Frozen Dessert Center Technical Symposium, October 2020 (virtual). What do you plan to do during the next reporting period to accomplish the goals?• Complete residence time distribution study for different dasher designs. • Continue experimental design on different dasher designs to determine effects on product structural attributes. • Evaluate CFD flow models for different dasher designs.
Impacts
What was accomplished under these goals?
After a rough start to the project (due to equipment issues and the pandemic), this past year has seen significantly more progress on Objectives 1 and 2. Objective 1: A preliminary study on freezing of sorbet was completed. The four different freezer configurations were studied at different operating conditions and ice crystal size distributions measured and compared. Interestingly, only small differences in ice crystal size distribution were found, nominally because the freezer itself automatically sets conditions to ensure good product. While this was a surprise, to some extent, this confirmed that the algorithms used by the freezer manufacturer are indeed robust. Still on Objective 1, we have initiated residence time distribution studies, with the assistance of Tetra Pak Hoyer. These are currently underway to compare different freezer designs. This year will see the start of freezing ice cream with the different conditions to generate data for the CFD modeling. Objective 2: Work on developing the CFD model has progressed slowly, in part because of the inconsistency in personnel. With a new BSE grad student in place, things are now progressing more readily. He has begun generating model data for comparison against our French colleague's results. We will then move ahead to modeling the different freezer designs for comparison against the experimental results.
Publications
|
Progress 02/15/19 to 02/14/20
Outputs
Target Audience:All ice cream manufacturers and research scientists Changes/Problems:Numerous problems have arisen in the early stages of this project: • We experienced a significant delay in getting the newly donated ice cream freezer from Tetra Pak into the lab and then getting it commissioned. There was a significant learning curve for training with the students involved. The freezer is now working well, although upgrades for dye injection to measure residence time distribution and measuring temperature profiles along the barrel are still dragging on. • On the modeling side, significant difficulties arose in finding suitable personnel. An international search did not uncover an adequately trained post-doc, so the decision was made to utilize a second graduate student. That student has now been identified and has begun the modeling work. • To further impede the experimental plan, the pandemic shut the lab down for an extended time, meaning that no experiments were conducted for several months. What opportunities for training and professional development has the project provided?One MS student, Katherine Helbig, started her training working on this project. She was trained on operating the new freezer donated by Tetra Pak Hoyer, developed methods for making sorbet with the desired characteristics, conducted experiments, and collected data for subsequent analysis. An undergraduate student working with Chris Choi was trained on CFD methods in anticipation of subsequent graduate work. However, this student did not develop as hoped and was replaced. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?In the next reporting period, we wil work on: • Complete sorbet study, where the effects of different dasher designs on ice crystal size distributions are studied. • Complete CFD training and initiate model development utilizing data from the sorbet trial.
Impacts
What was accomplished under these goals?
The first year of this project began with a number of delays and then ran into lab closure due to the pandemic. Under Objective 1, the first step was to conduct experiments on sorbet to confirm the results of a previous study with our French collaborators. By the end of the first year, the final formulation for that sorbet was confirmed along with the necessary physical properties of freezing point depression and viscosity. This was needed to feed into the CFD modeling. Under objective 2, the student was trained on using the CFD software for modeling flow within the SSF.
Publications
|
|