Skip to main content
European Commission logo
polski polski
CORDIS - Wyniki badań wspieranych przez UE
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary
Zawartość zarchiwizowana w dniu 2024-06-18

Fluid Foods Pasteurizer and Homogenizer based on Centrifugal Hydrocavitation Reactor

Final Report Summary - FCHR (Fluid Foods Pasteurizer and Homogenizer based on Centrifugal Hydrocavitation Reactor)

Executive Summary:
The EU food and drink industry is the largest manufacturing sector in the EU, with €965 billion turnover (about 15% of total manufacturing turnover) in 2008, about 310.000 companies and 4.4 million direct employees. The SMEs, which account for 99% of the companies involved in the sector, represent 48% of the turnover and 63% of total employees.
As an example, the EU dairy industry, representing around 15% of the turnover of food and drinks industry in Europe, even remaining the world’s number one cow milk producer with 142 mln tonnes, far ahead of US (80 mln tonnes) and India (38 mln tonnes), is facing increasing competition from emerging countries (China, India, US and South-America) and loosing export share: New Zealand is now the biggest exporter.
The most relevant challenges the sector is facing in the next future are:
1. Improve competitiveness of the processors, which is increasingly decisive for the pricing.
2. Uptake innovations linked to the health and well-being of consumers.
Process steps in which innovation could answer to both challenges are:
1. alternative treatments to pasteurization, which are becoming increasingly important due to a) consumer demand for new methods of food processing that have limited impact on the nutritional content and quality of food; b) need to improve the energy efficiency of the processes. Presently this process step is realized through conventional high temperature processes, which are both high energy consuming and can somehow alter the quality of the product. Hence effort is ongoing to explore new ways of introducing energy into the system.
2. Alternative more efficient processes for homogenizing, emulsifying, dispersing the fluid, which are commonly realised through rotating blades, mixers, or pumping devices with homogenizing valves that creates a narrow passage through which the product is forced to flow out. This process stage in the treatment of food provides improved product stability, shelf life, digestion, and taste.
Alternative “non-thermal” solutions for realising pasteurization and homogenization are being studied, such as pulsed electric field (PEF) and sonication through ultrasounds (US) or combination of these solutions.
These approaches are indeed more efficient in energy terms with respect to common thermal treatment, but have a main drawback linked to the difficult scalability.
For this reason the industrial application of these technologies is still difficult and commercial solutions are still only applicable at lab or very small production scale, while no industrial solution is presently available.
The FCHR proposal is a clear step forward with respect to the available S&T solutions in the reference industry. The Centrifugal Hydrocavitator Reactor (CHR) is a centrifugal reactor based on hydrodynamic cavitation and constitutes the core of the proposed pasteurization and homogenization plant. The Hydrocavitator has been devised to improve process efficiencies of industrial mixing, emulsifying, through uniform heating and cavitation phenomena: conditions of 2000 bar and 6000 K can be achieved for very short time (1 – 10 μsec) in micro sizes of 10 nm – 10 μm.
The heart of the FCHR plant for combined pasteurization and homogenization will be constituted by the FCHR in which the fluid food passes.
The proposed Food Hydrodynamic Cavitation Reactors present the following advantages with respect to commonly applied homogenizers and pasteurizers:
• Product Quality: lower temperature need to achieve the same pasteurization grade, thanks to the cavitation energy, and no thermal gradient. This reduces the degradation of the nutrients and keeps the flavor of the food; for instance in milk: white color is enhanced, quality problems such as “sweet curdling” due to enzymatic action and oxidation are minimised, shelf-life is extended under normal refrigerated conditions. This technology may also be beneficial in the processing of high protein or high viscosity nutritional products (e.g. infant nutritional products) due to the issues commonly faced with viscosity, fouling, burn-on and back pressure using traditional heat exchangers.
• Product Digestibility: in the case of milk the action of cavitation in the disruption of the structure of proteins has been reported, granting higher digestibility. Possible capability of raising the reaction of lactose hydrolysis for the production of lactose-free milk will be evaluated.
• Process Efficiency: If we consider a target temperature of 60°C instead of 72°C to achieve the same grade of bacterial inactivation, the energy saving would of the order of the use of FCHR allows to reduce energy consumption of more than 15% with respect to common solutions. Moreover, since the proposed equipment is smaller than the common solution ones, the power losses due to thermal exchange (convection and radiation) with the environment can be reduced. The specific amount of power loss reduction depends on the size of the reactor and for a mid-size equipment, they can be estimated in 70%.
• Plant Investment: the FCHR will combine 2 process steps (homogenization and pasteurizing) in one unit operation, allowing reduction of investment costs of about 50%.
All these features of the FCHR reactor will allow the dairy SMEs in Europe to increase their competitiveness and a valuable economic advantage will be gained with respect to traditional methods for pasteurization and homogenization. The potential market for the FCHR is therefore outstanding, due to the huge range of possible applications in the food sector and to the relevant number of involved SMEs in EU. Only considering the application to the dairy and canning sectors, with a market penetration reaching 5% of the plants installed after 5 years from the commercialisation of the FCHR plants (about 115 plants/year of medium size), it has been estimated that the profits for the SMEs participants would get to about 6 M€/y.
Project Context and Objectives:
Recent data on the Food & Beverage processing sector’s analysis show that the global market for food and beverage processing equipment reached a value of about 227.36 billion $ in 2012 and is expected to grow at a compounded annual growth rate (CAGR) of 2.7% from 2012 to 2019, the beverage processing equipment contributing 18% of the revenue. The overview of the equipment market indicates that:
1. The market is highly fragmented; larger companies are acquiring small and regional market participants to increase market share and visibility. Hence, the market is expected to consolidate in the future;
2. Ability to integrate software solutions into equipment is an important factor to consider in installation of individual pieces of equipment;
3. Environment-friendly and light-weight packaging are becoming increasingly important for companies to differentiate their products;
Companies are expected to shift the production base to developing countries in Asia-Pacific because of enormous growth potential and cheap labor. Furthermore, directives and legislation concerning food safety and hygiene, as well as environment-friendly systems in America and Europe are pushing food processors to upgrade machinery.
Intense competition in the market has resulted in a constant flow of technological innovations with respect to food safety, product quality and energy efficiency. Specifically for the dairy industry, optimized sensory attributes and diverse flavors are of prime importance: completely new product lines installed is expected to generate revenue in the food processing equipment sector.
Manufacturers dealing with processing, sterilization, preservation and packaging machineries within the meat and dairy domain are addressing the need for enhanced food safety and assuring equipment reliability and efficiency. There is a continued trend of modernization, upgrade and innovation taking place across most manufacturing industries, including the reference one of Food and Beverage.
The lifestyle needs of consumers, a key theme identified in this study, are expected to drive the growth of the whole processing support market. According to interviews made with consumers, in fact, safety is the most important aspect of processed foods. As a result, legislations are being passed on to food processors to equip themselves with safer and hygienic systems.
The market continues to witness major technology innovations, and market players are increasingly looking towards improving automation and environment-friendly solution. At a global level, over 700 big companies compete in the food and beverage processing support market, their competition being based mainly on the products’ technological capabilities, price, quality, automation features, flexibility and customer service/support. The remaining part of market share is represented by small and medium enterprises, for which several major challenges still need to be faced, starting from technological development, demand for advanced solutions and energy-saving processes, potentially supporting their daily activity and improve their competitiveness.

The FCHR project intends to increase the competitiveness of fluid foods producers, most of which are SMEs, thanks to the introduction of process intensification and improved energy efficiency treatments while keeping the integrity of food nutritional and flavour attributes.

The FCHR project proposes the implementation of an integrated pasteurizer and homogenizer for fluid foods based on an alternative approach induced only by mechanical means: hydrodynamic cavitation, which consists in the generation of huge amounts of energy in the form of shock waves, due to the turbulence produced in a fluid by pressure fluctuations.
The project was promoted by the company Wixta Industries who has patented an innovative configuration of Centrifugal Hydrocavitation Reactor (CHR). The reactor has been devised for heating and vaporization of fluids and has been first tested in laboratory with the support of the University of Rome Tor Vergata (UTV), allowing to confirm that the reactor is able to continuously increase the temperature of a liquid flow of 300 l/h of 25°C with efficiency of 84%.
Although the design and the control of cavitation effect were not optimised in the prototypes implemented so far and a large margin of improvement is expected, results confirm that the CHR technology:
1. can be effectively applied for liquid heating (or vaporisation) through mechanical device.
2. can act as a highly efficient homogenizer/emulsifier
Starting from the promising results produced in these applications, the objective of the project is to specialise the CHR concept to the needs of the food sector, producing a reactor which can act both as pasteurizer and homogenizer in a single process step, with an outstanding advantage for the food manufacturers in terms of energy efficiency and quality of the product. Even though the CHR designed by Wixta has so far not been tested for this application, it is worth noticing that the use of controlled hydrodynamic cavitation for fluid food pasteurization or sterilization has been already studied in some scientific works.

Project Objectives
1. To perform pasteurization and homogenization in a single process performed with a purely mechanical process, which is therefore highly scalable, due to the absence of electric field or ultrasound emitters.
2. To substitute thermal pasteurization with a process working at lower temperature, while delivering a safe product that preserves the sensory characteristics and freshness.
3. To reduce milk processing cost, thanks to improvement in energy efficiency in the manufacturing steps (pasteurization and homogenization).

The FCHR technology is applicable potentially to all fluid food in which pasteurisation and homogenization is needed: all products in the dairy industry, emulsions of flavorings, fruit nectars, vegetables puree, egg yolks, sauces and tomato sauces, formulations for early childhood, etc. The potential application to these products, and to intermediates used for their production, represents a huge potential market for the technology.
Project Results:
The activities carried out in the project (M1 – M26 ) were mostly devoted to the achievement of the following strategic goals, from a technical point of view:

FIRST REPORTING PERIOD
1. The internal discussions with the partners about what restrictions and constraints that should be considered for the FCHR pasteurizing system, and the establishment of the end-user and system requirements for both the lab-scale and the final prototype, including external requirements (Directives, national regulations, etc.) and desirable features that the beneficiaries considered could make more attractive for the future commercial use of the processing machine;
2. The identification of the technical specifications – components (sensors, valves, pumps, electronics, etc.) and of a general layout for both the lab-scale and the final prototype;
3. A study on the existing numerical models available for simulating the fluid-dynamic behavior and assess controlled cavitation in the reactor, as well as the realization of a fluid-dynamic model and of a parametric, 3D model of the reactor itself;
4. The design and implementation of a lab-scale prototype for microbiological and chemical/physical tests, able to assess pasteurization and homogenization in the treated food,
5. the development of a test protocol aimed at verifying the effect of the FCHR technology on raw milk and juices and the realization of preliminary tests at lab scale to evaluate the feasibility of the technology and the optimal conditions/parameters for achieving the desired conditions
6. The establishment of initial discussions among the partners aimed at identifying targeted dissemination and exploitation actions that the Consortium will put into practice for an optimal diffusion and promotion of the project, together with the collection of preliminary ideas for the preparation of a specific Plan for Use and Dissemination of Foreground document, at Month 9, indicating the approach that will be adopted after the end of project for the product commercialization of the pasteurizer/homogenizer.
7. The ideation and creation of a preliminary dissemination channel, i.e. the website, with a dedicated logo, and a first press release on the launch of the project.

SECOND REPORTING PERIOD
1. A parametric study having the purpose of optimizing the new pasteurization/homogenization process and the structural/fluid-dynamic behavior inside the reactor; this was possible by the use of a 3D model of the reactor itself, refined on the basis of constant feedbacks from the results of lab tests on the FCHR prototype; the output of the study was represented by a complete set of guidelines for the geometrical optimization of the rotor and stator enhancing the cavitation effects on the passing fluids;
2. The design & implementation of the final FCHR prototype; detailed engineering construction drawings were delivered, along with materials / components selection, a dedicated control system supported by a user interface and a new type of motor allowing for higher rotation speeds favoring the cavitation phenomena inside the reactor;
3. Intense tests sessions with the final prototype aiming at verifying the microbiological safety of the processed foods and their quality in terms of nutritional properties; two main food matrixes were selected for the lab tests, according to the interest of the participant SMEs (milk and apple juice). Scale-up proposals for the final reactor have been also delivered, in order to reach industrially interesting production capacities and to be ready to develop a FCHR plant for the food processing market (post-project phase, see details in this report);
4. A technology evaluation activity having the purpose of qualifying the treatment process in terms of economic competitiveness, energy efficiency, food quality, regulations compliance, production and operation costs for the final industrial process against commercial benchmark;
5. A detailed set of dissemination actions and the definition of an exploitation plan agreed with all the partner SMEs; this work included communication related to the project, an analysis of the FCHR competitors on the market, IP protection issues and training/know-how transfer activities.

From the point of view of the Management activities, the following actions have been performed:
1. The preparation of the First and Second Periodic Report and of the related financial statements;
2. The preparation of an agreed version of the Consortium Agreement, submitted in its definitive, signed version;
3. The creation of a Final Report showing all the developments carried out within the project framework;
4. The submission of all the planned Deliverables for the period, and the supervision on their technical contents, on the base of what was originally planned as from the DoW;
5. The planning and organization of general Project Meetings with all the partners, and the verification of the technical developments obtained by the RTDs in accordance to what is foreseen from the funding scheme, with an internal review procedure by the SMEs partners;
6. The management and coordination of internal technical meetings at WP level in order to monitor the status of the activities and to validate the results achieved;
7. The application of a contingency plan aiming at reaching the project goals and at recovering delays induced in the activities by unexpected technical difficulties with the prototype.
8. The preparation of official documents for the Amendment request for extending the project duration;

The project activities can be thought as structured in 3 main phases:
Phase 1: Process definition and lab research - identification of requirements and specifications for the FCHR system; simulation of the hydrodynamic behavior inside the reactor and preliminary microbiological/ chemical/physical tests for the control of the fluid foods treated with a preliminary prototype of FCHR, under different operative conditions (Activities performed in Period 1);
Phase 2: Process design and engineering / Industrial implementation and validation - design optimisation through numerical simulations; mechanical design of optimised reactor, structural evaluations and material selection; implementation of a specific control system for the control of cavitation; tests carried out to evaluate functionality and to verify the microbiological safety & quality; energetic and competitiveness assessment; definition of a pasteurizing/homogenizing process design properly applicable at industrial scale in order to meet the specifications and regulation of the food production (Activities performed in Period 2);
Phase 3: Industrial Perspectives - exploitation, dissemination and training, definition of the exploitation strategy and business plan, benchmarking the technology with respect to competitors in different application scenarios and delineating a business plan for the most promising. Technology transfer and training; dissemination of results in events (Activities performed in Period 2);

The main results of each WP can be summarized as follows:
WP1 - Definition of expected requirements for the FCHR pasteurizer/homogenizer from the end-user side (in terms of target product, selection of food matrixes for the tests, regulatory constraints, technical features, users’ needs, etc.).
WP2 - Development of numerical methodologies for simulating the fluid-dynamics inside the reactor, evaluation of the effects of the variation of the parameters affecting cavitation and their dependency.
WP3 - Manufacturing of the lab-scale prototype and implementation of preliminary tests at lab scale to evaluate the dependency of the working parameters on the microbiological and physical results desirable for the FCHR.
WP4 - Optimisation of the reactor design with the support of numerical simulations, along with structural evaluations and materials selection for the specific application to foods.
WP5 - Testing on milk and juices for the verification of the microbiological safety of products and for an evaluation of their nutritional properties. In this WP the overall treatment process was also defined and its applicability to an industrial scale was analysed.
WP6 - The results of lab tests on the prototype led to an analysis of the FCHR performance from both a technical and an economical point of view. Scale-up considerations and the selected parameters for operating the prototype were included in this analysis.
WP7 - According to the SMEs’ specific interests and businesses, a comprehensive study of the market has been performed with the final aim of defining a strategy for the future use of the FCHR technology. IP protection issues, competitors’ analysis, exploitation plan and dissemination actions were included in the official project documents as agreed by the Consortium.
WP8 - Management activities were successfully carried out to support the project development and prosecution; these will be further detailed in a specific section of this report.

We can state that:
a) The fluid-dynamic model of the reactor was properly carried out by UTV and the results on assessing controlled cavitation inside the reactor were detailed in D2.2 and D2.3--> MS1 achieved in P1;
b) The lab-scale solution for the FCHR pasteurizer/homogenizer was delivered to Cork and preliminary microbiological and physical tests were carried out on the prototype --> MS2 achieved in P1;
c) The executive, engineering CAD drawings of the final prototype were delivered by LABOR --> MS3 achieved in P2;
d) Three proposals of FCHR industrial plant layouts were delivered to the SMEs, taking into account different approaches for the scale-up of the reactor to reach the desired production capacities, with an analysis of the possible benefits of each approach --> MS4 achieved in P2;
e) An intense period of tests has been carried out at UCC for the evaluation of hydrocavitation effects on milk and juices pasteurization/homogenization and a detailed technology assessment has been delivered, taking into account technical and economical aspects of the FCHR new process --> MS5 achieved in P2;
The Consortium Management has been carried out by WIXTA during the whole project, and support on the management tasks in the moderation of technical discussions and the preparation of dedicated progress meetings came from the Technical Manager UTV and from the other performers.
In summary, we can state that:

o Most of the strategic and operative goals of the project have been achieved;
o The technological choices and strategic decisions for the optimal execution of the project’s activities and research have been always presented and agreed to the SMEs within the Consortium; their involvement in the development of the project work has been precious and always encouraged;
o Delays were assessed in the course of the project that needed attention from the Management Board and the application of a contingency plan; the prompt reaction of the partners in managing the critical aspects of the research led to the fulfillment of the Consortium obligations towards the REA and to the satisfaction of the beneficiary SMEs for what concerns the expected results of the project, results that will undergo an engineering and scale-up activity in a post-project phase with the final aim of introducing the FCHR technology in the target market (details in the Final PUDK).

MAIN RESULTS
1) Definition of Specifications
The machinery sector represents a fundamental part of the engineering industry and is one of the industrial mainstays of the Community economy. The Machinery Directive 2006/42/EC provides the regulatory basis for the harmonization of the essential health and safety requirements for machinery at European Union level.
Essentially performing a dual function, the Directive not only promotes the free movement of machinery within the single market, but also guarantees a high level of protection to EU workers and citizens. Being a "New Legal Framework" Directive, it promotes harmonization through a combination of mandatory health and safety requirements and voluntary harmonized standards. A brief overview of the EU Directives regulating the design and manufacturing of industrial machinery has been reported in D1.1 – FCHR Requirements, which concentrates on those directives that refer to food processing machinery, as in the case of the FCHR system, in particular:
- Directive 98/37/EC (The Machinery directive 98/37/EC of 22nd June 1998, valid until 2009) and
- Directive 2006/42/EC (Machinery directive of 17th May 2006 of the European Parliament and of the Council of 17th May 2006 on machinery, amending Directive 95/16/EC. This aims at the free market circulation on machinery and at the protection of workers and consumers using such machinery. It is intended at defining essential health and safety requirements of general application, supplemented by a number of more specific requirements for certain categories of machinery.
Guidelines for the implementation of the Machinery Directive have been taken into consideration in this Deliverable, as well as a description of how the Consortium members in FCHR project are implementing the Directives in their countries.
The FCHR Consortium, aware of the potentiality of the new technology underlying the project, also intends to carefully evaluate the EHEDG guidelines, that mostly integrate and include Regulations of the European Commission on materials and hygiene rules for the design of machinery intended for food processing or production, in order to eventually establish the compliance to such regulations as a system requirement.
The European Hygienic Engineering & Design Group (EHEDG) is a consortium of equipment manufacturers, food industries, research institutes as well as public health authorities and was founded in 1989 with the aim to promote hygiene during the processing and packing of food products.
The principal goal of EHEDG is the promotion of safe food by improving hygienic engineering and design in all aspects of food manufacture. EHEDG actively supports European legislation, which requires that handling, preparation processing and packaging of food is done hygienically using hygienic machinery and in hygienic premises.
The EHEDG guidelines include the following:
- Regulation (EC) no. 1935/2004 on materials and articles intended to come into contact with food, and repealing Directives 80/590/EEC and 89/109/EEC,
- Regulation no. 2023/2206 on good manufacturing practice for materials and articles intended to come into contact with food,
- Regulation no. 852/2004/EC on hygiene of foodstuff

Directives on juices production
The composition of fruit juices, concentrated fruit juices, dehydrated fruit juices and fruit nectars, their reserved names, their manufacture and labelling characteristics are subject to specific Community rules under Council Directive 2001/112/EC, which was amended in August 2009 by the Commission Directive 2009/106/EC in order to be adapted to technical progress taking account of developments in relevant international standards, in particular the Codex Standard for fruit juices and nectars (Codex Stan 247-2005) which provides information on the quality criteria for fruit juices and nectars, and indicates a set of further criteria assessing the products’ authenticity and composition, and the AIJN Code of Practice for evaluation of quality and authenticity of juices.
The FCHR Consortium analysed and reported the main indications of the Fruit Juice Directive (Council Directive 2001/112/EC of 20th December 2001), which has been considered of relevance for the project.

Market Overview and Traditional Pasteurization/Homogenization processes
The food and drink industry is the largest manufacturing sector in terms of turnover and employment in the EU. It is also the 2nd leading manufacturing sector in terms of value added and number of companies in the EU. The share of the food and drink industry in manufacturing in terms of employment is stable and its share in terms of turnover and value added has registered slight variations between 2004 and 2008. In 2009, the turnover registered a more important decrease due to the combined effects of the decline in factory-gate prices and reduced exports. Figures for the Food & Drink Industry in Europe report the following data (referring to 2010):

4.1 million employees (leading employer in the EU manufacturing sector),
The Food & Drink market is a fragmented one, comprising about 274.000 companies, the 99.1% of which are Small and Medium Enterprises (SMEs),
It purchases and processes the 70% of the European agricultural production, contributing to a trade balance of around 9.8 billion € and exporting 65.3 billion € in food and drink products to third countries.
The main sectors of interest in the EU Food and Drink market for the purpose of the FCHR project will be the Dairy Industry market and the Fruit Juice market. A short overview of the structure and trends of these markets were reported in Deliverable D1.1 with specific focus on the Consortium countries, Italy, Norway, Ireland and Malta, as well as forecasts for the next years.
A general overview of the traditional treatments for milk and fluid foods pasteurization and homogenization was provided. Even if innovative treatments are currently investigated and tested for improving fluid foods processing – such as PEF, ultrasonication, combination of these two methods, high pressure processes (HPP), UV light technologies etc. – traditional pasteurization remains the benchmark against which the FCHR mechanical treatment performances will be assessed in the course of the project.

The system/user requirements task in FCHR is intended to identify, collect and organize the requirements coming from the SMEs’ needs and interests, from an operational point of view. The goal is obviously that of assessing a complete, consistent and clear set of requirements through which the different components of the system can be outlined, and the related specifications gathered for a proper system implementation. The techniques used for the identification of the FCHR pasteurizer/homogenizer requirements have been the following:
- Plenary meeting (kick-off meeting in Rome) for performing a brainstorming session,
- Further discussions among the partners and exchange of emails/information to be included in the official documents concerning normative, external constraints and further necessities identified by the SMEs,
- Use of a supporting model for requirements collection and sharing of the chosen (IEEE) shell structure with the explanation of how these should be filled in,
- Iterative revision of the document for updates and improvements.
The template used for the user requirements files was also used in this task so as to provide the partners with a unique tool for gathering information. The methodology applied for the collection of such requirements proved efficient and intuitive, and the immediateness of the requirements setup process, as it was suggested, made the involvement of all the project partner easy, allowing for fruitful exchange of information and requests among the beneficiary SMEs and the performer. This led to a shared vision of what the FCHR pasteurizer and homogenizer will have to perform, and of what it will have to be like, in order to be competitive and bring innovation in the addressed markets.

2) Technical Specifications and Layout
Following the establishment of the lab scale and final prototype requirements for the FCHR system, the technical specifications identification phase was started. This task was mostly related to the definition, in accordance to what is required by the beneficiary SMEs and by the industrial environment, of the best layout for the system.
In order to allow for the design and implementation steps, the features of the FCHR cavitation reactor were outlined in this task, with a distinction between the lab-scale prototype that will be shipped to Cork for preliminary tests on raw milk and juices by UCC partners, and the final prototype. The technical specifications have been collected so as to respond and fulfil the already established requirements, using the same template for convenience. This in fact allowed to recall the ID of the requirement linked to each single specification. Furthermore, the layout of the overall process and instrumentation scheme, together with the choices made for the components and the control system were drafted.
It must be considered that, the requirements encoding activity carried out in Task 1.1 was performed already specifying in most of the cases measurable criteria related to the required performance. The translation of the requirements into technical specifications has then been realized with the following steps:
• checking the technical feasibility of the requirement,
• defining the technical parameter related to the requirement (if not done in the definition of the requirements)
• confirming or modifying the quantitative criteria for each parameter, which constitutes the technical specification, which will then be used to assess whether the system meets the established requirements.
Like in the case of the requirements, the specifications on the small-scale cavitator were collected in tables for convenience, this allowing to directly link the technical choices indicated in the table with the need/request made by the SMEs.

The design of the reactor is implemented using a single rotor configuration. It is constituted by a rotor in high speed rotation inside a stator. The inlet of the liquid is realized in the upper part of the reactor, while the outlets are located in the radial part of the stator. The liquid is accelerated by the centrifugal speed inside the rotor and collected in the external radial part of the stator.
The rotor is realized in two symmetrical parts in which the channels were the liquid passes are located.
The following characteristics are implemented according to the specifications outlined in the related task:
• According to LABSPEC1 the reactor is realized in AISI316.
• According to LABSPEC2 the diameter of the rotor is 10 cm.
• According to LABSPEC3 the reactor is realized in a vertical configuration with the rotor/stator in the upper part and the motor in the lower part, connected by a flange specifically implemented to grant the mechanical stability and avoid vibrations due to the high speed of rotation. An hydraulic seal is located on the shaft.
• According to LABSPEC4 a discharge opening is designed in the lower part of the stator to allow the discharge through a tap.

According to LABSPEC5 the outlet is realized from the radial part of the reactor. However, to avoid interference with the circumferential seal that were produced by longitudinal outlets, only 2 radial exits were designed. It was considered that, due to the presence of a rest area where the liquid is collected in the external part of the stator, the difference between radial and longitudinal exits would not be relevant.
According to LABSPEC9 and LABSPEC10 the motor must be high speed (up to 10.000 RPM) and the speed must be controllable through an inverter.
For this reason, the motor selected is a specific high speed asynchronous motor, specifically designed for high speed applications (granted up to 20.000 RPM) thanks to the high precision mechanical design, the use of special ball bearings. The lubrication is air/oil dosed through and electronic system.

The layout of a single block of the plant is defined in the preliminary scheme. The layout of the final system will be better defined once the results of the lab phase will allow the definition of the working parameters (flow, temperatures, need of pre-heating), in order to define the need of series/parallel treatments to achieve the desired plant capacity and treatment (homogenization and pasteurization grade) for the fluid food. In particular:
• The full capacity of the plant will be realized by using parallel reactors, which could be implemented in parallel blocks, in order to exploit also partial capacities, or by using parallel reactors in the same line, in case also keyed on the same motor shaft.
• The eventual need to treat the fluid in more than a passage in the reactor could be realized by series reactors in the same line, keyed on the same motor shaft.
The features considered in the preliminary layout are:
• washing line with 3 way valves,
• heat recovery with an heat exchanger between the outlet and the inlet to realize a pre-heating,
• cooling on the outlet of the reactor to bring it immediately to the needed storage conditions (4°C),
• indication of temperature, pressure and flow transmitters to be logged in the control system,
• temperature control actuating on the variable speed motor,
• temperature and flow alarms as specified above (not shown in the layout)

3) Simulation of the hydrodynamic behaviour
The tasks planned under WP2 intended to develop:
i) a study on the cavitation phenomenon focusing on relevant numerical models for addressing a fluid-dynamic simulation,
ii) a fluid-dynamic model of the reactor in order to assess the cavitation phenomena inside it, and of
iii) a parametric 3D CAD representation of the reactor based on the most modern geometrical modelling techniques.
The intention of the study at point i) is to comprehend the mechanisms governing cavitating flows for choosing the best numerical approach for the fluid-dynamic simulation of the reactor. To do this, UTV has identified and reported in a document:
- a description and a discussion on different aspects in cavitating flows,
- the most relevant contributions in the scientific research about cavitation and
- a detailed review of the state-of-the-art numerical models for including cavitation in standard computational fluid-dynamic (CFD) models. Moreover, a comparison among 11 commercial and open source CFD software applications has been addressed.
The model developed in task ii) takes into account the rotation movement of the rotor with respect to the stator, the dynamic unsteady interaction between rotor and stator domains, the mass transfer between liquid and vapour phases governed by cavitation. In order to simplify the model for reducing the required computational resources, only a periodic portion of 36° has been simulated and adequate boundary conditions at the cutting zones have been imposed.
The model has been used for the simulation of the preliminary shape and layout of the reactor and was able to compute pressure field and velocity vectors inside it, together with the occurrence of cavitation zones. Different rotational speeds have been simulated.
Finally, the model has been used in the course of task iii) for both manufacturing purposes and for the numerical simulations of the functional behavior. The parameterization of the model makes the geometries prone to be changed for quick and easy modifications and shape optimization of the reactor. After the interpretation of the results coming from the preliminary simulations a list of priority and importance of these parameters on both the rotor and the stator has been discussed.
The study on the physics of cavitation showed that cavitation is a complex flow dynamics phenomenon, whose investigation needs to be addressed with specific mathematical models and dedicated numerical tools. While the study of the inception of cavitation can be studied with classical mathematical models of flow dynamics (estimating flow pressure using simplified formulas and/or numerical models), the development of cavitating flows and supercavitation requires specific numerical approaches based on two-phase flow modelling and complex three-dimensional meshes. Scientific literature presents several models for dealing with the study of the development of cavitating flows. There is not a unique methodology, but several approaches based on different assumptions and simplifications. Different methodologies mainly differ on the contribution to the flow dynamics of the evaporation and condensation rates. Twelve different approaches have been studied in depth and the corresponding equations analyzed and tested. Most approaches are based on semi-empirical models.
The overview of available software unveils that only a few of computer program can support the modelling of cavitating flows. In some cases the mathematical approaches are well documented, in some other the details of implementation are hidden to the user. Since the cavitation problem applied to a hydrodynamic reactor includes the dealing with complex geometries and complex fluid-condition well-tested and reliable numerical software is required. A good interface with computer-aided modelling environment is also required for addressing the optimization loop.
Starting from the preliminary study on the physics and modelling of cavitation, the fluid-dynamic model has been built from the geometrical computer-aided model of the device by extracting the internal cavities by Boolean geometric operations. The main purpose of the model is to describe the overall fluid-dynamic behavior in a global and significant way in order to produce enough information to compare different engineering solutions and gather design guidelines. Moreover, the approach in modeling aims to produce a model able to catch sufficient details in fluid dynamics, but enough simplified to be solved on a standard workstation in a reasonable amount of time. The model is very complex because the phenomenon to be modelled is complex too. In particular, the model takes into account the rotation movement of the rotor with respect to the stator (sliding interfaces), the dynamic unsteady interaction between rotor and stator domains, the mass transfer between liquid and vapour phases governed by cavitation. In order to simplify the model for reducing the required computational resources, only a periodic portion of 36° has been simulated and adequate boundary conditions at the cutting zones have been imposed. The cavitation has been assessed using the Shenerr-Sauer model, one of the most used models according to the review of the state-of-the-art.

The entire fluid-dynamic domain can be split into two main zones: the rotor zone and the stator zone. Both zones have been discretized using tetrahedral elements whose dimensions have been controlled in order to be smaller on the features of interest. In particular, mesh elements have been concentrated in the rotor channels and in the interface between them and stator zones in order to facilitate the convergence of the first steps of simulation which are the most computational demanding.
According to the basic specification, two simulations have been performed, changing the rotational velocity of the rotor from 3000 r.p.m. to 6000 r.p.m. Other design parameters have been assumed constant.
The results of the simulation are presented by using graphs and plots. Concerning with the preliminary results, both pressure fields and velocity vectors have been computed. These quantities are useful for a deeper understanding of the fluid-dynamic behaviour inside the device. The cavitation phenomenon has been investigated by reporting the volume fraction of vapour produced inside the reactor.
In both simulated cases (rotational speed of 3000 and 6000) r.p.m. looking at the velocity of the fluid, the fluid accelerates passing from the inlet channel to the radial channels. A second acceleration is present at the bifurcation of the Y channels. In this zone, since the cross section area of two channels is lower than the area of the corresponding radial channel, the fluid accelerates for the fulfillment of the continuity (see below for the results at 3000 r.p.m.).
Another interesting zone is the intersection between the radial channel and the transverse channels. Here, the flow is slightly perturbed by the sudden change of width, but it is not deviated in the transverse channels. According to the simulation results, the direction of the flow inside the transverse channels is not well defined and the fluid tends to stagnate.
All these results about velocity, although plotted and discussed for a specific relative position, are present during the entire rotation of the rotor with very similar behaviour.
The pressure field in the stator presents an important gradient from the axis of the rotor to the outer parts similar to that computed in the rotor. This gradient is amplified in the circumferential collecting channel in which the fluids accumulates and near the outlet orifice. Some intensifications of the pressure are present in the channels connecting the reaction chamber to the circumferential collecting channel.
Results of the simulation of the reactor rotating at 3000 rpm limited to the rotor zones. Pressure field is depicted as a volume rendering and velocity vectors are depicted as 3D arrows. Boundary outliers (arrows of huge magnitude) are due to local numerical inaccuracy at sliding interfaces.
According to the pressure results, the lowest pressure is recorded in the bottom part of the rotor. By plotting the volume fraction of the secondary phase the location of cavitation sources can be observed. Two main zones of cavitation have been pointed out, both belonging to the rotor domain. The first one is located in the bottom and deepest zone of the rotor, since it is the zone at the lowest level of pressure. The second source zone for cavitation is located in the first portion of the radial channels. In these regions the fluid is coming from the inlet channel and after entering in the radial channels accelerates producing a drop of pressure which reaches the thermodynamic conditions to produce cavitation. The vapor is then transported by the flows and implodes again becoming liquid in the radial channels. Simulations do not underline the occurrence of cavitation in the outer parts of the reactor and cavitation does not happen in the stator regions.
All the simulations have been repeated for two different values of rotational speed in order to underline the differences. It has been observed a similar behavior of the two cases but the increase of the rotational velocity has a positive effect on cavitation which occurs in wider zones. Moreover, with a higher rotational velocity, the pressure drop inside the reactor is greater.
With a higher rotational speed, in the rotor there is a lower pressure at the portion near the axis of rotation and a higher gradient throughout the structure (remember that the output pressure has been set as constant and the pressure field is relative). The same differences are noticeable in the stator.These differences in the pressure field also influence the occurrence of cavitation. For the high rotational velocity case, the cavitation occurs in the same zones of the low velocity case, but it is slightly amplified. In particular, there is an extension of the zone subjected to cavitation of 20÷30% more. In the same proportion, the volume fraction increases too. Cavitation does not happen in the stator regions.
Since the model is very complex and several simplifications have been included, the confidence of the results is limited to the description of a global behavior rather than a precise and punctual evaluation. Nevertheless, the positive outcome of the model has been extremely useful to have a detailed idea about what happen inside the machine.
From the interpretation of the simulation results we can conclude that the first source of cavitation is mainly located at the inner part of the rotor. The second source zones are located about 1 cm above the previous one, along the radial channels of the rotor. In these zones there is a combination of two effects: the low pressure as already discussed and an additional pressure drop due to the acceleration of the fluid inside the channels.
An increase of rotational velocity gives a benefit towards the production of cavitation, since the pressure drop inside the reactor is greater.
Concerning with the interpretation of the results about the fluid flow, the device has some zones whose utility is limited as the transverse channels of the rotor and the lower part of the fluid domain of the stator. These considerations have been used for defining some important strategies for the geometrical optimization of the reactor.
In order to support the manufacturing process and to prepare the fluid-dynamic domain for the simulation, a complete 3D model of the reactor has been built. The assembly has been modeled by using the modern Computer-Aided Design (CAD) parametric approach. In relation to traditional design, a feature-based approach pays more attention to form and topology than dimensional precision in the conceptual phase, which makes the methodology very suitable for preliminary design and optimization. In a feature-based approach which supports constraints, profiles are sketched to capture this “design intent”.
All the parts have been modeled using a functional strategy including explicit reference to variables subjected to optimization. Constraining equations have been included in order to take into account the geometrical and practical limitation and to simplify the subsequent morphological optimization.
The complete CAD assembly is made of 73 parts and 14 geometrical and dimensional changeable parameters have been chosen. Several constraint equations have been deduced for limiting the variation of these parameters in order to obtain a valid and reasonable shape.
The model is composed by the following main parts:
- A stator which has been divided in two halves (four pieces) for assembly and maintenance purposes (a detailed view below);
- A rotor which has been divided in two halves for manufacturing purposes (detailed view below);
- A cap with different inlet openings;
- A closing cap;
- A shaft;
- Two ball bearings;
- A coupling joint;
- A dynamic sealing;
- Four sets of screws;
Each component has been modelled in details as a separate part and then assembled in specific assembly files. Since the fluid-dynamic behaviour investigation is executed on a geometrical modeled derived from the global assembly, a geometrical verification has been also performed to check the suitability of the assembly for the other task in the research project.
The preliminary results of the fluid-dynamic model have been used for correcting and integrating the observation concerning with the geometrical parameters optimization in order to simplify and enhance their best choice.

4) Preliminary prototype and tests
The adaptation activity for the preliminary prototype involved the following actions, carried out by LABOR in collaboration with WIXTA, with the support of UTV:
- A modification of the reactor’s design, performed by UTV and WIXTA,
- The implementation of the reactor by LABOR with the support of a mechanical workshop for the manufacturing;
- The design of the test bed plant, including selection, purchase and integration of the necessary valves, sensors, pumps and of the motor to be coupled to the rotor, as well as the user interface(LABOR);
- The assembly of all the parts for the realization of the test-bed, with subsequent test and verification of proper hydraulic functioning of the whole prototype to be sent to UCC for tests on raw milk (LABOR).
The starting point for the adaptation of the reactor’s geometry was the identification of specifications, carried out in Task 1.2 of the project, mainly:
• Prototype capacity: 20 l/h
• Stability to be reached after > 30 secs
• Quantity treated per test: 1 liter
• Feeding tank and output tank: water and ice baths (10 liters) available at Cork
• Inlet pressure: useful to have a pump at inlet to test different pressure conditions, even though the cavitator presents a self-pumping behavior,
• Variables to be measured: Inlet T, Inlet P, Output T, Output P, Flow, time, RPM, electric power input,
• Tests under different conditions: RPM, inlet temperatures, inlet pressures
• Test different passages of the same milk in the plant (simulating series of cavitators)
• Computation of output power from inlet T, output T, Flow
• Material: AISI 304 or 316, suitable for food processing
The Description of Work foresaw “a preliminary lab-scale prototype (50-120 l /h) to be implemented by WIXTA on the basis of the designs already available from previous applications, and according the specifications setup in WP1”. However, the following specified features defined for the lab scale prototype, made it evident that it was not possible to perform tests on already existing prototypes. In particular:
• necessity to perform tests on raw milk having a very small reactor (down to 20 l/h),
• necessity to use Inox as material for all parts in contact with food,
• necessity to improve ease of inspection, and
• necessity to modify the sealing to avoid contact of the food with not specified material
All these requests, analyzed and evaluated by the Consortium, led to the necessity to apply for a new solution, realized by making some improvements in the reactor designs, and thus to an implementation phase that was not planned in the project activities.
Starting from the technical specifications drafted in Deliverable D2.1 and concerning the realization of the lab scale prototype, the following components were purchased by LABOR for proceeding with the implementation activity.
1. A pump PU 01,
2. The FCHR reactor,
3. The FCHR motor and inverter,
The sensors that were acquired to measure the interesting parameters of the process are the following:
1. FT01: flow meter to measure flow passing to the reactor the range 0-100 l/h,
2. TT01: temperature meter to measure temperature of the input to the reactor in the range 0-40°C,
3. TT02: temperature meter to measure temperature of the output from the reactor in the range 0-100°C,
4. PT01: pressure meter to measure pressure of the input to the reactor in the range 1-3 bar,
5. PT02: pressure meter to measure pressure of the output from the reactor in the range 1-3 bar,
6. RT01: encoder to measure the rotational speed (RPM) of the rotor in the range 0-12.000 external, not logged,
7. WT01: Watt meter to measure the electrical power consumed by the motor in the range 1-5 kW,
8. TC01: inlet temperature in the input thermostatic bath,
9. TC03: temperature of the output thermostatic bath,
10. PC01: inlet pressure by actuating on the pump,
11. RC01: RPM of the rotor by actuating on the inverter.
These sensors and their technical specifications will be shown later on in this report, when describing how the FCHR prototype was conceived and the different parts selected).
The user interface acquires and reports the data from these sensors, as well as the calculation of the following physical quantities:
1. WT02: Thermal power transferred to the liquid,
2. E01: Electrical energy used by the motor (Wh)
3. E02:Thermal energy transferred to the liquid (Wh)
4. η inst: Instantaneous efficiency calculated as WT02/WT01
5. η avg: Average efficiency c.
According to what was discussed with WIXTA in this task, the main changes to be implemented on the small-scale prototype were made to the shape and features of the rotor. In this section we will report some CAD designs of the initial rotor, meant to be mounted on the FCHR prototype, and those of the re-designed one, actually present on the prototype sent to the partners UCC.
The main features of the new reactor configuration, respect to the existing one initially planned for the project, can be summarized as follows:
- Outlet of the fluid: holes initially present have been substituted by slots on the external circumference of the stator; the optimum position for such slots was chosen so as to avoid problems with the seals realized through an o-ring. Multiple tangential and radial outputs have been created. The discharge of the fluid – used for emptying the machine when this is not running – has been provided in the lower part of the stator by means of an inclined hole, creating, in parallel, a slight slope of the stator;
- Seals: two different seals have been kept on the new design, even though, at high pressures, one of them proves sufficient; In order to provide an adequate housing on the stator some geometrical modifications have been suggested by UTV and an updated CAD model provided.
- Motor: the motor to be used was purchased by LABOR and WIXTA (maximum power 1 kW); LABOR was in charge to design and manufacture a physical support favoring the cavitator and the motor’s alignment on the prototype; this was implemented via a flange included between the motor and the stator.
The FCHR prototype can be thought as an assembly of three basic parts:
o Sensors,
o Electrical equipment, and
o Mechanical parts.
Each of these parts was detailed in Deliverable D3.1. Before shipping the prototype to Cork for testing the FCHR process on raw milk, LABOR mounted and assembled all the components and provided the system with the flange allowing for the alignment of the motor with the hydrodynamic cavitator, pre-testing the functioning of the single parts with water.

The objectives on the preliminary tests are to evaluate if the reactor can be considered a good alternative technology to the traditional pasteurization and homogenization of milk. Therefore, the preliminary tests done at UCC evaluate the capacity of the reactor to emulsify and pasteurize raw milk.
Raw milk is an emulsion, i.e. a mixture of milk-fat globules, various solids and water. The fat in milk is presents as globules of non-uniform size, ranging from 0.20 to 2.0 µm. Homogenization is a technological process commonly applied in the dairy industry to stabilize the fat emulsion against gravity separation. Homogenization is achieved by disrupt the fat globules into much smaller and uniform ones (size to less than 1.0 µm) which allows them to stay evenly distributed in milk. In the experiments conducted, the particle size distribution of milk processed by the prototype was determined using a laser light-scattering instrument (Malvern MasterSizer®) and the particle size distribution was compared to raw milk and milk homogenized.
Raw milk can be considered a favourable media for micro-organisms development, like lactic acid bacteria, enterobacteria, Listeria, Staphylococcus, Pseudomonas, Bacilli, Clostridia, due to its composition and pH. Pasteurization is applied to raw milk to reduce the number of harmful micro-organisms to a level at which they do not constitute a significant health hazard. However, it has effects on the flavour, colour, texture and nutrients. Preliminary analysis on the ability of the reactor to reduce the microorganisms present in milk were performed by inoculate the milk with lactic acid bacteria strains commonly used in Cheddar cheese manufacture. Therefore, raw milk was inoculated with 0.03% (w/v) DVS R604 (Direct Vat Strains of Lactococcus lactis subsp. Cremoris and Lactococcus lactis subsp. Lactis) at 30°C for 30 min. At the end of this period, a sample of inoculated milk was taken to measure microbiological count before treatments. Moreover, a sample of milk was pasteurized at 63⁰C C for 30 min (LTST pasteurization) to check the effect of pasteurization on the milk without cavitation.
Before starting each experiment, the prototype was rinsed with warm water (50-60⁰C) for 2 min and left empty. At the end of the experiments, the equipment was rinsed with water and 10 L of caustic soda (0.5%) were run in recirculation through the reactor for 10 min. Then, the equipment was rinsed with water and 10 L Horolith V (0.5%) where run for 10 min in recirculation through the reactor. At the end, the equipment was rinsed with water and left empty. All solutions are at room temperature. The motor was set up at 2500 RPM and no pressure was applied.
The effectiveness of homogenization was discussed on the basis of comparison of fat globules particle size distribution measured using Malvern MasterSizer®. The Mastersizer measures the size of particles by measuring the intensity of light scattered as a laser beam passes through a dispersed particulate sample. Depending on their size, the particles scatter the light with a different angle. The angular scattering intensity is used by the instrument software to calculate the particle size distribution. Prior to measurement, milk samples were dispersed in deionised water by stirring in a sample dispersion unit. The parameters of fat globule size distribution were calculated using the polydisperse optical model, assuming a refractive index of fat globules of 1.46 absorbance of 0.00 and the refractive index of the aqueous medium (water) as 1.33.
Due to homogenization, milk fat globules size should change to a distribution around 1 µm. In all the graphs, the purple line indicates the particle size distribution of homogenized milk, whereas the red line shows the particle size distribution of raw milk.
As a result of this task, milk has been processed by FCHR technology using different combinations of the parameters controlled by the operator. Moreover, different physical configurations have been tried.
Preliminary results on the pasteurization effect of the FCHR in milk showed that the reactor did not result in pasteurization. In the conditions used in the experiments, it seems that the microbial growth reduction was due only to the temperature used, as opposed to sonication.
Preliminary analysis on the homogenization ability of the prototype reported that fat globule disruption was not achieved under the operating conditions used. However, in a few experiments partial homogenization was reached and it is expected that better results could be obtained with a higher motor speed. Therefore, the FCHR prototype with the current configurations and settings is not presenting ideal results. However, the key variable appears to be motor speed and steps to be taken include the substitution of the motor with a powerful one. After motor substitution, milk will be processed by the FCHR and analysed for particle size distribution and microbial inactivation. Further analysis on inactivation of surrogates of bacteria, rheological properties, and protein physical state and denaturation were carried out at UCC on the final prototype.

5) Simulation model and reactor improvements
With the purpose of optimising the shape and geometry of the reactor and enhance the effects of cavitation on milk, further simulations were carried out at UTV. The proposed and tested modifications are driven from the observations come from the experimental campaign on the prototype built according to the initial design.
- In a first part of the activity, the results of the experimental tests have been reviewed focusing on mechanical and fluid-dynamic aspects and trying to relate the specific behaviour to each geometrical feature. In particular, preliminary experimental tests showed that the initial design and configuration of the reactor did not produce evident benefits on the treatment of milk. The most likely reason of this unsuccessful functioning is that the machine is not able to produce enough cavitation to process the milk.
- In a second part, the guidelines for the modification to the device geometry have been defined. They guided the update of the corresponding computer-aided geometrical model and the computational fluid dynamic one.
According to the interpretation of results of the experimental tests, 2 strategies have been pointed out:
1. The first one concerned the increasing of the rotational speed of the machine, since the production of cavitation is influenced by the pressure drop inside the device which is somehow proportional to the rotor speed;
2. The second strategy focused on a very relevant improvement of the processing geometries inside the rotor.
The stator geometry of the device is not influenced by the modifications. This choice is driven by the consideration that the update of a single internal part as the rotor is more straightforward to be manufactured and assembled and does not require a new configuration of the layout.
- In a third part, the fluid dynamic simulations concerning the improved geometry have been repeated in order to virtually compare the improved design with the initial one. The fluid dynamic model of the optimized reactor has been developed following the same numerical strategy and methodology of the initial design and it is based on a parametric geometrical description of the features in order to easily update the entire simulative workflow.
Results of the simulation showed that, in the optimized design, the pressure changes are more uniform presenting a lower gradient from the hub to the radial extremity. This feature testifies that the entire geometry works in favor of the fluid processing and the fluid treatment is uniform throughout the device. Another positive aspect is that the mean value of the pressure is lower and it is a benefit with respect to the production of cavitation.
Furthermore, the pressure gradient in the transverse hole testifies that the fluid inside these regions is moving and it does not stagnate as it happened in the original design. Thus, also these secondary regions positively contribute to the fluid process. The new tapered geometry of the inlet holes accelerates the fluid causing an important initial pressure drop, which affects the pressure field of the entire rotor channel. The fluid in the inner part of the rotor is moving too and no stagnating zones are reported.
In conclusion, the positive outcome of the simulations confirmed the positive impact of the proposed modifications on the global functioning of the device and so they have been embodied as design specifications and mechanical drawing.

The experimental tests performed by the Research Team in Cork which are discussed in detail in the corresponding deliverable, showed that the initial design and configuration of the reactor does not produce evident benefits on the treatment of milk. Two are the possible reasons of this unsuccessful functioning: a) the machine is not able to produce enough cavitation or b) the type of produced cavitation is not suitable for the desired process.
Actually, an already published research about food science showed that hydrodynamic cavitation has positive effects on the treatment of milk and so the first option seems the most improbable.
In order to perform a comprehensive review of the reactor design we should investigate which aspect(s) has/have the most important influence on the production of cavitation. Again, experimental tests showed that the modification of some working parameters such has inlet/outlet pressure or the rotational speed of the main shaft has a small influence on the overall performances. This behavior suggests that the modifications should be implemented inside the reactor, trying to amplify the production of cavitation.
As a preliminary observation, the modeling of cavitation is not an easy task at all. The cavitation model introduces a high level of complexity and imprecision may occur. Moreover, as any other representation of a real phenomenon, the model is based on assumptions and includes simplifications. All these aspects together with the modeling strategies have been carefully investigated and all the choices that have been made have strong theoretical bases.
For all these reasons, we are confident that the results of the model are reliable, even if they may contain some imprecision and approximations. The amplitude of these unpredictable errors is lower enough to allow a useful global assessment of the behavior at a level which is sufficient for design purposes.
First of all, the model is very complex and it is quite a unique specific approach. According to the best knowledge of the research team, there are no contributions in the scientific literature dealing with such a complex scenario (3D flow, cavitation, unsteady effects, rotating domains, narrow channels, etc.). A direct comparison to similar models is therefore impossible at this moment.
On the other side, the comparison with the experimental observations is complex too. Actually, the purpose of the model is to predicts and assess the production of cavitation, but a specific (and direct) measurement of cavitation is very complex. Optical qualitative measurements seem the most widespread, but they require a transparent device and dedicated video acquisition system. Moreover, the presence of narrow annealed channels and a high speed rotor make the observations still more complex.
According to the interpretation of the results of the experimental tests, there is the need for an important improvement of the reactor. The first strategy has concerned the increasing of the rotational speed of the machine, since the production of cavitation is influenced by the pressure drop inside the device which is somehow proportional to the rotor speed.
On the other side, experimental tests showed that for the current layout configuration the maximum speed provided by the motor is 3000 r.p.m. Recirculation has a positive but very limited effect.
All these observations suggest a deep improvement inside the rotor, not only in the layout configuration or in the working parameters.
Coming back to the observations discussed in the deliverable D2.2 we underlined that the geometries of the rotor play a more important role with respect to those of the stator.
Moreover an optimization of the rotor geometries keeping constant those of the stator has the advantage to keep the original layout of the assembly. In this case, it is possible to substitute the two rotor halves in the device without designing again the complete layout and the assembly features.
In order to keep the original stator geometries the changes of the rotor have to be compatible with them and with the rotating shaft connections. It means that the new version of rotor must be assembled on the rotating shaft using the same keyway and locking nut and has to freely rotate (with adequate clearances) inside the stator. These requirements introduce some constraints and limitations in the redesign of the geometries, but several improvements are still possible.
The optimization is performed following the subsequent strategy:
- Try to improve the filling of the rotor --> more homogenous fluid dynamic field
- Try to reduce the volume and the zones of stagnating flow --> more regions become active and take part in the fluid processing;
- Try to extend the zones where the cavitation occurs --> the reactor increase the fluid treatment
- Try to increase the amplitude of cavitation --> the reactor increase the fluid treatment
- Try to reduce the clearance between rotor and stator --> an increased shear effect has a benefit especially in homogenization process.

The modifications are implemented keeping the following constraints:
- The rotor halves have to be congruent with the previous prototype (compatibility with the stator geometries, same hub connections, same boundary dimensions)
- The general layout of the rotor has to be the same in order to preserve the initial design intent suggested and required by WIXTA.
- The manufacturing of the rotor halves has to be possible using standard technologies.
Five important modifications of the rotor geometries have been implemented. They are summarized in the following list together with the motivations and the expected improvements:
1. The number of axial inlet holes has been increased. The holes have now a reduced diameter with a tapered region in order to accelerate the fluid. This modification is expected to produce a more uniform filling of the reaction chamber and an initial acceleration of the fluid with a consequent pressure drop which aims to fuel cavitation.
2. The position of the axial inlet holes is changed. They are located farther from the rotation axis of the rotor. This modification has two expected benefits. First of all, the number of the holes can be increased since the surface is wider in the areas farther from the axis. Moreover, the fluid enters in the reaction chamber in a region where the centrifugal effects are higher (the centrifugal acceleration is proportional to the square of the distance from the rotation axis).
3. The inlet chamber has a reduced volume. The simulation on the initial configuration showed that the inlet chamber has a large volume of stagnating flow. A reduction of the available space should decrease this unwanted behaviour.
4. The transverse holes have a different location. They have been placed nearer to the rotation axis in order to exploit their contribution on the accelerations of the fluid inside the chamber.
5. The transverse holes have a different inclination. The initial transverse holes were parallel to the rotation axis. In the revised design they have an inclination of 48°. This modification has been suggested by the results of the fluid dynamic model which unveil that the fluid is stagnating in that region. A positive inclination causes the fluid to be subjected to an acceleration gradient and contributes to its flow.
6. The lateral clearance between the rotor and the stator has been reduced by increasing the axial thickness of the rotor. This modification has two expected benefits. First of all, the smaller fluid thickness between fixed and moving parts increases the local velocity gradient with an increase of the pressure gradient too. Moreover, the reduced thickness is expected to amplify shear effects between rotor and stator and these effects are expected to have a mechanical contribution to the homogenization process.
The fluid dynamic model of the optimized reactor has been developed following the same strategy and methodology of the initial design, which are herein summarized:
- Starting from the 3D Computer-Aided model of the rotor and stator geometries, the fluid domain is built by volume subtraction. A comparison between the fluid dynamic domain of the initial design and that of the improved one is reported.
- A single portion of 45° degrees has been considered in order to use the cyclical symmetry periodic boundaries.
- Moving wall condition (ideal walls, no slip) has been imposed to the rotor internal surfaces
- Stationary wall condition (ideal walls, no slip) has been imposed to the stator internal surfaces
Fluid-dynamic simulations showed that the proposed modifications seem to produce important benefits to the production of cavitation. The presence of zones with cavitating flow is extended and the amplitude of the phenomenon increased. Finally, the improved design of the rotor is compatible to the same manufacturing processes of the previous one and the estimated cost for production is almost the same.

6) Design and Implementation of the Final Prototype
According to the results of the simulation and those of the preliminary tests, the main changes that have been implemented on the small-scale prototype concern the shape and features of the rotor. Besides that, some other modifications have been implemented already in the preliminary prototype, in order to improve the global functionality, operations and reliability of the device. The modifications have been driven by practical considerations and by the results of the experimentations.
The improved features of the new reactor configuration, with respect to the ones available at the beginning of the project, can be summarized as follows:
- Outlet of the fluid: holes initially present in the stator have been substituted by slots on the external circumference of the stator for facilitating the flowing of the fluid, especially in presence of high-density fluids. Their position has been chosen to be compatible with the external o-ring seal. Multiple tangential and radial outputs have been manufactured. The drainage of the fluid, for emptying the machine when this is not running, has been provided in the lower part of the stator by means of an inclined hole, creating, in parallel, a slight slope of the stator;
- Seals: a more efficient seal system, based on a pre-loaded dynamic barrier has been implemented in the new design. In order to provide an adequate housing on the stator, some geometrical modifications have been manufactured. The improved sealing system has revealed to be more reliable and its functioning is not altered by wear.
- Cage: The cage of stator in which both sealing system and bearings are located has been simplified and manufactured in a single part. This modification takes advantage of the simplified mounting requirement of the sealing system and its compact dimensions.
- Outlet channels: in the improved design, the device the outlet channels of the fluid have been doubled in order to favor the mixing from two opposite device zones.

To deal with the need of increasing the rotational speed of the FCHR, it was decided to substitute the motor which was selected and used in the preliminary prototype. First an upgraded version of the initial motor was shipped to Cork and mounted on the preliminary prototype, passing from 1.1 kW to 4 kW electrical power. However, tests carried out on the prototype showed no significant improvement on the speed achievable (2800 RPM vs 2500 RPM).
For this reason, it was agreed to select a new motor normally used in mandrels, in order to achieve higher rotation speed and allow maximum speed of 6000 rpm (even 9000 RPM). This was decided after preliminary tests in order to check if the cavitation effect is improved and consequently the pasteurization and homogenization phenomenon. Due to the increase of the power of the motor, it was also necessary to substitute the inverter and the electrical cables used in the previous design. As the power of the motor in the final prototype was increased to meet the technical requirements, all the electrical equipment had to be also improved.
The most important component of this upgrade is the frequency converter, which has a power of 15Kw. Its size and its fitting into the electrical cabinet didn’t influence the layout inside of the cabinet, as the original housing was chosen in the first place to be larger than necessary and the global inside layout of the components didn’t encounter any obstruction.
The FCHR prototype can be thought as an assembly of 3 parts:
1. Mechanical and hydraulic part: FCHR reactor, valves, pipes
2. Electrical part: Motor, inverter and electrical appliances
3. Control system: Sensors, actuators and user interface
The FCHR prototype is located in the dairy processing hall in the School of Food and Nutritional Science at University College of Cork (UCC) in Ireland.
Preliminary tests to optimise the working parameters on the final prototype were carried out during May 2014, to evaluate if the reactor can be considered a good alternative technology to the traditional pasteurization and homogenization of milk. Tests were carried out using the first version of the rotor design; results are included in D3.2 – Final results of lab tests on the small scale prototype. The testing of the prototype under optimised conditions will be included in D5.1 –Results of field tests. On the other hand, some preliminary tests on the prototype were carried out at UCC to evaluate the basic performance of the prototype with water.

No further or new programming was needed to this new device, but only its power is changed compared with the previous version. The energy cables were also changed and made available in suitable size to face the greater electrical consumption. All connections were checked and tested for signaling and functionality within the new layout. As mentioned before, the only active part is the motor command (speed regulating) realized by a frequency converter that accepts analogue proportional voltage signal ranging from 0 to 10 volt. Even for this section, a special electronic board is needed and basically it is different from the others because of its characteristic to generate signal rather than acquire. The electronic unit is composed of housing unit and specialized acquisition board. The connection to target application computer is a standard USB port.
The entire software was conceived with National Instruments LabVIEW development system that drives its own hardware system. The software is developed on a Microsoft Windows platform and the screenshot is conceived to follow the physical layout of the prototype. The operator simple needs to start the software that immediately is connected to sensor and give the whole status of the system. User can choose the motor speed and start the data recording. Some input data are not provided by the sensors but must be provided by the user (example: fluid density) but they are computed as well as the others values even in the graph widget, in the log files also and in calculated values boxes. Before starting the test (start/stop button), the user can choose the test’s physical conditions: Density, RPM for the hydrocavitator, logging frequency.

7) FIELD TESTING AND SCALE UP
Several experiments were carried out with the purpose of testing the prototype capability in reaching pasteurization and homogenization results on milk and juices. The following set of experimental analyses were carried out in the course of Task 5.1 that led to the following results:

Milk processed at 4000 and 6000 rpm (1st rotor head). The effect of different FCHR speeds on the chemical and physical stability of milk was evaluated. Raw milk and the prototype were pre-heated to 50⁰C before tests at 4000 and 6000 rpm.
Milk was collected after one passage through the reactor, after discarding about 500 ml of milk. Between tests, the prototype was cooled down with water. Temperature, pressure and flow rate were recorded by the software overtime. Analysis on the particle size distribution of milk samples shown that milk treated at 4000 or 6000 rpm did not give differences in particle size distributions.
Analysis on the milk creaming rate showed that samples treated by FCHR-F had a slower creaming rate compared to raw milk and pasteurized milk. Between the two speeds used on the FCHR-F, 6000 rpm gave less cream separation compare to milk processed at 4000 rpm. Homogenized milk did not show any cream separation during 7 days.
Analysis on the rheological properties of rennet-induced gels showed that homogenized milk generated stronger gel and the gel was produced in a shorter time compared to the other samples. Pasteurized milk produced weak gel and the gelation time was longer compared to the other samples. Milk treated by FCHF did not differ from raw milk in terms of gelation time and gel strength.

Milk processed 5 times at 6000 rpm (1st rotor head). Raw milk and the FCHR-F prototype were pre-heated to 50⁰C before tests at 6000 rpm. The outlet pump was on during the experiments. Water at 50⁰C was passed through the FCHR-F then the inlet tube was transfer to the milk pre-heated to 50⁰C and a milk sample was collected after one passage through the reactor. The first 500 ml of milk were discarded to exclude the possibility of collecting milk sample diluted with water. The prototype was cooled down with water and the temperature of milk and reactor were brought to 50⁰C. Milk was passed a second time through the reactor using the procedure described for the first passage and a sample of milk was collected. Further five passages were done with the same method and a sample of milk was collected at the fifth passage. Analysis on the particle size distribution of milk samples shown that increasing the number of passages of milk through the FCHR-F generates a better disruption of fat globules.
Analysis on the milk creaming rate showed that samples treated by FCHR-F had a slower creaming rate compared to raw and pasteurized milk. Moreover, at day 7, the layer of cream was less thick in the samples treated by FCHR compared to the other milk samples. The increased number of passages was effective on reducing the cream thickness and the creaming rate. Homogenized milk did not show any cream separation during 7 days.
Analysis on the rheological properties of rennet-induced gels reported that homogenized milk generated a stronger gel in a shorter time compared to the other samples. Pasteurized milk produced a weak gel and the gelation time was longer compared to the other samples. In this experiment, milk treated by FCHR-F after 5 passages had a gelation time similar to homogenized milk. However, the gel strength was not comparable to the homogenized milk. The other milk samples did not differ from raw milk in terms of gelation time and gel strength.

Milk processed 10 times at 6000 rpm (1st rotor head). The effect of the number of passages through the FCHR-F on the particle size distribution of raw milk was studied in this experiment.
The method tested in the previous experiment was used to process the milk up to 10 times through the FCHR-F. Milk was collected after 1, 2, 5, 8 and 10 passages through the FCHR-F. Between samples collection, the reactor and the milk were cooled down and the temperature was led to 50ºC. The reactor was on during the all experiment. No significant differences in particle size distributions were observed between 5, 8 and 10 passages. Therefore, the next experiment was done using maximum five passages through the FCHR-F. Moreover, the particle size distribution of the FCHR-F samples was closer to the milk passed only once through the conventional homogenizer.

Effect of FCHR treatment for 5 passages on bacteria inactivation (1st rotor head). To study the contribution of the FCHR-F to milk pasteurization, raw milk was inoculated with 0.03% (w/v) DVS R604 (Direct Vat Strains of Lactococcus lactis subsp. Cremoris and Lactococcus lactis subsp. Lactis) and incubated at 30°C for 30 min. A sample of milk was taken to measure microbiological count before treatments. Lactic acid bacteria grew in milk to about 8 log CFU/ml milk in 30 min at 30°C. After pasteurization at 63°C for 30 min they are reduced to 2 log CFU/ml milk.

Milk processed 5 times at 6000 rpm (2nd rotor head). The FCHR-F rotor head was replaced with a new rotor head in July 2014 at UCC. Milk was analysed after treatment by the FCHR-F equipped with the new rotor (FCHR-F-N). Analysis on the particle size distribution of milk samples shown that increasing the number of passages of milk through the FCHR-F-N generates a better disruption of fat globules. 5 passages through the FCHR-F-N seems to be necessary to have a homogenization effect comparable to the one obtained by processing the milk by conventional homogenizer one time.
However, processing the milk by FCHR-F-N did not give better results in terms of milk homogenization compare to the results obtained using the first rotor head under the same operative conditions (6000 rpm of speed and pump on).
Analysis on the milk creaming rate and accelerate stability showed that samples treated by FCHR-F-N had a slower creaming rate compared to raw and pasteurized milk. The increased number of passages was effective on reducing the cream thickness and the creaming rate. However, the creaming rate was slightly faster in the samples treated by FCHR-F-N compared to the samples treated by FCHR-F. Homogenized milk samples did not show any cream separation during 7 days.

Effect of 5 passages through the FCHR at 6000 rpm on bacteria inactivation (2nd rotor head). The test performed with the 1st rotor head was repeated with this second one. Lactic acid bacteria grew in milk to about 8 log CFU/ml milk in 30 min at 30°C. After pasteurization at 63°C for 30 min bacteria were reduce to 2 log CFU/ml milk. Homogenization slightly increased the bacteria, 3 log CFU/ml milk were count.
In the tests using the FCHR-F-N, the micro growth decreased by increasing number of passages of the milk through the reactor. Five passages reduced the bacteria to 5 log CFU/ml milk and the same reduction was found in milk processed five times through FCHR-F.

Milk processing by pre-heating at 40°C and then passing through the FCHR until 70°C are reached. The objective of this study is to evaluate the ability of the FCHR-F-N to inactivate lactic acid bacteria in milk pre-heated to 40⁰C and processed through the FCHR-F-N at 6000 rpm until the temperature of the milk in outlet reached 70ºC.
Raw milk was inoculated with 0.03% (w/v) DVS R604 (Direct Vat Strains of Lactococcus lactis subsp. Cremoris and Lactococcus lactis subsp. Lactis) at 30°C for 30 min.
Milk was pre-heated at 40° C and passed through the reactor more times until a temperature of the milk in outlet reached about 70ºC. A total of 7 passages were performed and a sample of milk was collected after each passage. Analysis on the particle size distribution of milk samples shown that processing the milk starting from a temperature of 40ºC and passing the milk until a temperature of about 70ºC was reached did not generate a better disruption of fat globules. There were no evident differences in particle size distribution between samples passed two times up to seven times throughout the FCHR-F-N.
Lactic acid bacteria grew in milk to about 8 log CFU/ml milk in 30 min at 30°C.
After pasteurization at 63°C for 30 min number of lactic acid bacteria was reduced to 2 log CFU/ml milk. Homogenization slightly increased the number of lactic acid bacteria to 3 log CFU/ml milk.
Milk processed by FCHR-F-N shown a slight reduction in bacteria growth when milk was passed from one to four times through the reactor. After five, six and seven passages the number of lactic acid bacteria decreased to 5, 4 and 3.5 log CFU/ml milk.

Processing of juice at 50°C for 5 times at 6000 rpm. One batch of raw apple juice was received from a local apple juice producer. The juice was processed the day of the delivery.
The method used for milk was tested also on apple juice. Water at 50ºC was passed through the FCHR-F then the inlet tube was transfer to the juice pre-heated to 50ºC and a juice sample was collected after one passage through the reactor. The prototype was cooled down with water at 50ºC and the temperature of juice was brought to 50ºC. Juice was passed a second time through the reactor using the procedure described for the first passage and a sample of juice was collected. Further five passages were done with the same method and a sample of juice was collected at the fifth passage.During the treatment of juice by FCHR-F, an increase of about 6ºC in temperature was measured
No differences in pH were recorded. pH of all the samples was 3.27.
There were no differences in particle size presented in raw and homogenized samples.
Analysis on the particle size distribution of juice samples shown the presence of some particles with a bigger size in samples processed by FCHR-F.
The samples were tested for stability during storage and the results showed that all the samples presented a sediment. There were no differences in the sedimentation rate and thickness of the sediment between samples.
Results on the total bacteria count showed a reduction in the number of bacteria naturally presented in raw apple juice.

Processing of juice at 50°C at 3000 and 6000 rpm. In this experiment, the batch of apple juice used in the first experiment was processed after 7 days of storage at 4ºC. This time should allow the growth of the natural microflora present in the juice and the FCHR-F could be tested on the capacity to inactivate the existing microorganisms.
Water at 50ºC was passed through the FCHR-F then the inlet tube was transfer to the juice pre-heated to 50ºC and a juice sample was collected after one passage through the reactor.
Two different FCHR speeds were tested, 3000 and 6000 rpm. A slight increase in pH was recorded in these samples compare to the juice analysed in experiment one.
There were no differences in particle size presented in raw, homogenized samples and samples processed by FCHR-F at 3000 or 6000 rpm.
The average of the particles present in pasteurized juice was bigger compared to the other samples.
Tests on the simulation of the stability of the samples during storage showed that all the samples presented a sediment.
Polyphenol oxidase activity did not differ between homogenized samples and samples processed by FCHR. Pasteurized juice had a lower residual activity compared to other samples.
Results on the total bacteria count and yeast count showed a reduction in the number of bacteria and yeast naturally presented in raw apple juice when juice was pasteurized or pre-heated to 50ºC and processed by FCHR-F at 3000 or 6000 rpm.

Main conclusion on milk processed by FCHR-F:
During the treatment of milk by FCHR-F, an increase of 5-10ºC in temperature is measured. Results from the experiments reported show that milk pre-heated to 50ºC and passed five times through the reactor presented chemical and physical characteristics closer to the milk conventionally homogenized at 180 bar and passed one time through the homogenizer.
Moreover, milk processed by FCHR-F reduced lactic acid bacteria population inoculated into raw milk, especially after five passages, however results were no comparable to traditional pasteurization. The second rotor head did not give improvement in milk homogenization and bacteria inactivation compared to the first rotor head.
Main conclusion on juice processed by FCHR-F:
During the treatment of juice by FCHR-F, an increase of 5-10ºC in temperature is measured. Results on juice processed by FCHR-F shown that there are no benefits in processing the juice multiple times through the FCHR-F in terms of the physical characteristics analysed and microbial inactivation capacity. Pre-heating juice before processing the product throughout the FCHR-F has an effect on bacteria and yeast inactivation. When juice was pre-heated to 50⁰C and processed by FCHR-F bacteria and yeast were completely inactivate and the inactivation was maintained during storage. Moreover, the physical characteristics of the samples were not altered by the FCHR-F process. FCHR-F technology seems to give more promising results on juice than milk treatment.

o Scale-up Evaluation
The tests carried out on the prototype were necessary to define an approach to develop a solid scale-up for the FCHR hydrocavitator process. As it was highlighted in the previous sections, for both the Pasteurization and Homogenization processes, 5 passages through the hydrocavitator head are needed to reach a good level of disinfection and homogenization of the treated milk.
Specifically, the FCHR prototype CF2 configuration behaves better than CF1 configuration for the Pasteurization process so this will be considered as the key geometry for the scale-up. The working conditions for both pasteurization and homogenization are found to be the following:
- Milk flow rate: 250 l/h
- Engine speed: 6000 RPM
- Temperature: <60°C
- Hydrocavitator passages: 5 passages
These parameters characterize the hydrocavitator reactor in its completeness. Therefore, its size and shape cannot simply be changed or adapted to higher flow rates to meet an industrial workload in terms of processed milk per day. This means that the scale-up cannot coincide with a “bigger hydrocavitator head” where x-y-z dimensions are resized to increase the head capacity. The geometry of the rotor is the limiting factor where a small change can influence the whole process making it inefficient.
So, a modular approach will be drafted in the following, and this will imply one or more FCHR hydrocavitator heads, in configuration CF2, working in series or in parallel to cover the needs of small, medium and large pilot plants where the hydrocavitator head becomes a black box that cannot be altered. Two different pilot plant sizes are proposed and studied to meet the needs of different milk producers according their working day capacity: medium and large plant size. The size of the plant is expressed as liters of milk processed per hour:
- Medium Plant size: 1000 l/h of milk
- Large Plant size: 6000 l/h of milk
The test proved that 5 passages through the cavitator head are necessary for achieving Pasteurization and Homogenization of milk, so, in order to maintain a continuous milk flow rate, 5 hydrocavitator heads must work in series.
Below we report the industrial proposition of the concept applied during the laboratory tests: five passages through the head, cooling down of milk between each passage. Of course, this is not an applicable and valid solution at an industrial scale as it does not save costs nor optimize the items:
- In this scheme each hydrocavitator head is matched to a single motor, whereas it should be more appropriate to match more heads to a single motor reducing the control component;
- The cooling of milk between each passage was a need to study and characterize the reactor at laboratory level and discriminate the contribution to pasteurization coming from a thermal effect with respect to that coming from the cavitation effect.
The increase in temperature as ‘side effect’ is indeed an aid towards the pasteurization goal combining both the phenomena (cavitation and temperature) when the prototype is running. Stated that each cycle increases the temperature of milk of about 8°C, it is possible to fix the final temperature of the product and pre-heat the milk at the beginning of the process by means of a heat exchanger.
The P&I of the FCHR 250 l/h module where these improvements and optimizations have been applied.

The P&I scheme for a FCHR based plant of the capacity of 1000 l/h of milk consists of four 250 l/h FCHR modules assembled together. The proposed scale-up does include a heat recovery by means of a heat exchanger between the cold inlet milk (4°C) and the processed milk (60°C) after its passage through five hydrocavitator heads.

Results for the pasteurization of apple juice by means of the FCHR 250 l/h module are reported. The working parameters are:
- Apple juice flow rate: 250 l/h
- Engine speed: 3000 & 6000 RPM
- Pre-heating temperature: 50°C
- Hydrocavitator passages: 1 passage

The pasteurization tests conclusions on apple juice are:
a) No benefit is assessed in processing the juice multiple times through the FCHR in terms of the physical characteristics measured and microbial inactivation capacity: one passage is enough to inactivate bacteria and yeasts;
b) No difference in terms of bacteria and yeast inactivation between the juices processed at 3000 or 6000 rpm meaning a less energy consumption;
c) Temperature of the juice affects the bacteria and yeast inactivation (pre-heating at lower temperatures does not give the same results).


8) TECHNICAL ECOMOMICAL EVALUATIONS
With the final aim of verifying the effectiveness of the FCHR process and of comparing its performance with the traditional pasteurization and homogenization methods, we are here providing an overview of the system’s maturity with respect to its target applications.
Market Segments: Dairy Products, Food & Beverage Industry, Food Processing Machinery
Target Applications/Customers: Raw milk pasteurization and homogenization processes (customers: dairy companies), Pasteurization of fluid foods like juices, sauces, infant formulations, eggs yolks, etc. (customers: food and beverage producers), Production plants implementation and selling (customers: food processors and food machinery producers)
Maturity: FCHR prototype not fully ready to serve the target market(s), needing further engineering and validation tests work; Price: final plant expected price is less than the setup target, derived from commercial benchmark; this is far more evident in the case of apple juice FCHR treatment plants, for which initial investment costs prove significantly reduced if compared to the commercial ones. At this stage there is evidence of the advantages that the FCHR solution might bring but no proof of the technology being mature enough to compete in the market with traditional processing techniques; more promising results have been achieved for apple juice processing (details later on this document). Optimization needed for the whole process of pasteurization and homogenization; results of bacterial inactivation need to be demonstrated on real pathogens to prove in line with the regulation requirements for food; Further tests need to be performed to assess the efficiency of the FCHR process on foods other than milk and apple juice (tests on more dense fluids required, planned in a post-project phase).

Starting from the final results of the project against the planned objectives, we can state what follows:
1. The prototype developed during the 26 months of the project had a different performance in treating milk with respect to apple juice; the operative conditions for both processes were reported in D5.2 and show how the hydrocavitation reactor is capable to achieve pasteurization and homogenization results for both foods with different working conditions, which prove more efficient and cost-effective in the case of apple juice; on milk, however, the obtained pasteurization grade still cannot be considered as comparable with the traditional, thermal treatment;
2. From a technological point of view, the FCHR solution presents the following advantages and limitations:

MILK
o The processed milk needs to be pre-heated to a temperature of about 50°C and to be passed 5 times through the reactor to present chemical/physical characteristics which are similar to those of conventionally homogenized milk;
o An increase of about 5-10°C per passage was observed;
o Pasteurization conditions, although achieved, could not induce a final treatment which is comparable with the traditional, thermal one;
o Processing milk for a certain number of passages inside the reactor is found to be necessary for the achievement of pasteurization conditions; lab tests showed that 5 passages are required to obtain such goal in the case of milk;
o The combined effects of hydrocavitation and temperature increase in the liquid allows to achieve the above mentioned results; however, no quantification of their specific contributions could be possible;

APPLE JUICE
o An increase of about 5-10°C in temperature was observed during apple juice treatment;
o It appears clear from the laboratory tests that increasing the number of passages through the reactor does not imply any beneficial effect, so that 1 passage only was considered as sufficient to achieve a satisfactory pasteurization and homogenization grade;
o Pre-heating up to 50° C was necessary before starting processing the raw juice by means of the FCHR prototype;
o Pasteurization and stability results, as well as stability during storage and unchanged physical characteristics of the treated juice confirm that the FCHR processing technique is promising on this food, even more that if it is applied to raw milk. The result that is to stress here is that when treating apple juice with the FCHR pasteurization occurred after one passage only through the hydrocavitator, at a temperature of about 60° C (increase in T after one passage about 8°C), which is lower than the traditional pasteurization temperature (72 °C). This allows at the same time to achieve the objective of reducing the energy consumption associated to the process and to keep the quality of the treated food unaltered. Energy consumption and efficiency of the FCHR process will be detailed in a dedicated paragraph in this document.

The most demanding process in terms of energy consumption is the pasteurization process which requires 58.8 Wh/l of milk while the homogenization process is a very low energy-demanding process with only 6.5 Wh/l.
If we consider both Pasteurization and Homogenization processes, we have to uniform the two-energy contributions due to the different nature of the energy source: steam for pasteurization and electricity for homogenization. The first contribution is expressed as thermal kWh while the second one as electrical kWh. The correlation between the two units of measurement is:
Thermal equivalent kWh = Electrical kWh / 0.46
So the homogenization process energy consumption contribution can be rewritten as 6.5 / 0.46 = 14.1 thermal equivalent W/h meaning that the traditional Pasteurization-Homogenization process requires 58.8 + 14.1 = 72.9 thermal equivalent Wh/l of energy.
The FCHR technology aims to be a valid alternative for these processes. From the P&I we know that a 250l/h module works with five rotor heads in series, each of them consuming 3,5 kW for a total of 17,5 kW, which means 17,5 / 250 = 70 electrical Wh/l or else 152.17 thermal equivalent Wh/l of energy.
We can conclude that the FCHR technology is not mature enough to compete with the traditional Pasteurization-Homogenization technology, so further investigation and development of the FCHR prototype within optimization of the hydro-cavitation rotor is absolutely required in order to compete with the existing state of the art technologies in terms of energy use and efficiency.
On the other hand if we talk about Pasteurization of apple juice the conclusions are quite different. We can assume to apply the same Pasteurization energy data for milk also to apple juice. Indeed traditional Pasteurization process is a thermal treatment, that is, subdued to the ability of the liquid media to receive heat.
The parameter which expresses this characteristic is the specific heat, Cp, which for milk is about 3.9 kJ/Kg°C. The average specific heat of fruit juices -ranging between the 12 and 15% of solids - is about 3.8-3.9 kJ/Kg °C therefore comparable with the Cp of milk.

Traditional Pasteurization requires 58.8 thermal Wh/l of apple juice of energy while the FCHR technology is able to process 250l/h by means of a single module working with one hydrocavitator rotor head consuming 3.5 kW, meaning 3.5/250 = 14 electrical Wh/l or else 30.4 thermal equivalent Wh/l of energy.
Finally, data must be normalized taking in account the heat necessary to pre-heat the juice in order to achieve a satisfying Pasteurization grade.
The P&I of the scale-up for both milk and apple juice, see D5.2 Scale up evaluation, presents a heat exchanger at the end of the line in order to recover the heat of the liquid food exiting the process, using it to pre-heat the media entering the process.
While for milk the T of outlet is enough to guarantee a proper heat exchange with the inlet of milk, for the apple juice some extra energy must be spent to provide a DT of 19°C to the inlet stream of apple juice. Indeed the outlet stream is not hot enough to pre heat the liquid to the working temperature for the FCHR process.
This energy contribution affects the efficiency of the process by reducing the total amount of litres that can be processed for each kWh from the calculated value of 30.4 Wh/l to the final value of 46.8 thermal equivalent Wh/l***, anyway providing an overall satisfactory result for apple juice if compared to the traditional value of 58.8 Wh/l improving the efficiency of the process of about the 20% in terms of energy spent.
As evident from the estimates above, we can conclude that, whilst the FCHR treatment applied to milk does not provide an overall improvement in the energy efficiency of the process, as it basically consumes almost twice the amount of energy which would be necessary to thermally treat it, the new technology applied to apple juice performs the pasteurization process with a significantly reduced energy demand.
This, together with the reduction in the treatment temperature (about 60°C for apple juice) shows how the FCHR device can be effectively considered competitive in performing pasteurization of apple juice, with respect to currently available energetic and economic benchmarks.

Milk Treatment Costs Analysis
Starting from the data available from the RTD performers regarding the costs sustained during the prototyping phase as for motors, rotor heads and related equipment provided on the FCHR system, we can proceed with an analysis of the economic viability and advantages of the technology. An evaluation of the costs for one FCHR module and for the final medium-size plant (meaning a production capacity of about 1.000 l/h) is reported in this section.
As reported in Deliverable D5.2 on the scale up of the FCHR system, the prototype implemented in the project lifetime was characterized in terms of implementation costs. Due to its prototypal nature, of course, the real sustained costs were relatively high if compared to the desired massive production of FCHR modules for sales. We must say, however, that due to the hydrocavitation reactor configuration and its optimized shape for the achievement of satisfactory pasteurization/ homogenization conditions, there is no high margin for the reduction of the scaled-up plant, as increasing the production capacity does not mean increasing size or dimensions of the reactor itself.
It is, then, necessary to couple several reactors together to achieve the desired quantity of processed food, and additionally, manufacturing costs for the reactors cannot be reduced with the increase in the reactors number.

Operation costs details
The analysis we are performing on the FCHR reactor’s production for commercial exploitation must be completed with the costs associated to the general management and operation of the plant. This includes salaries of the personnel involved in the process supervision and control, costs for maintenance of the plant and for the purchase of necessary consumables / equipment for substituting parts that might be damaged for the prolonged use.
An estimation made by the technical partners of the project setup the lifetime of one module to about 10 years of continued operation, which is comparable to currently available pasteurizers for industrial use. The commercial pasteurization equipment does not operate necessarily for the whole day, so that for the FCHR plant we have also estimated a duration of the processing operations of about 8 hours per day.
An overview of these costs is reported in D6.1; these are summed to the initial investment costs to obtain a realistic estimate of the overall implementation, installation and management/operation expenditure for a medium-size processing plant. Costs per year are meant to take into account electricity consumption above all, as the reactor is moved by an electric motor.

Referring to section 3.3 it has been estimated that it is possible to process about 15 l/kWh of milk meaning 0.0667 kWh of energy spent for single pasteurized-homogenized litre of milk. Having an average cost of 0.15 €/kWh it is possible to calculate the benchmark cost expressed as euro per litre of the value of 0.01 €/l (cost therefore related to traditional Pasteurization-Homogenization processes). From an economical point of view, we can therefore state what follows:
- the FCHR pasteurization and homogenization plant investment costs range between 80.000€ - 384.000€ for milk treatment (for a 1.000 l/h and a 6.000 l/h plant capacity, respectively);
- It is evident that, respect to commercial benchmarks, the hydrocavitation-based technology used by the FCHR plant for fluid foods processing proves more or less comparable as for initial investment costs in the case of milk. The main costs for the plant implementation are in fact associated to the rotor heads manufacturing and to the necessity to couple several reactors for scaling up the system to the desired – industrially appealing – production capacity. This raises the initial cost of the plant up to a commercial one, proving comparable to the cost of traditional pasteurizers and homogenizers used in the Food and Beverage industry for the treatment of the same volumes of fluid foods.
- From the estimates made in this document, the final cost per litre of milk processed by the FCHR plant proves comparable to the one of a commercially available pasteurization plant on the market:
0.010 € / litre (commercial benchmark) vs 0.028 € /l (FCHR treatment)
Apple Juice Treatment Costs Analysis
In the case of apple juice pasteurization, the estimation of the plant costs is different, as the process via the FCHR hydrocavitation is somehow simplified. Applying the same considerations made for the milk economic analysis, setting the lifetime of the plant to about 10 years of operation not operating necessarily for the whole day.
As evident from the comparison, the FCHR treatment results very competitive in terms of costs savings per litre of juice processed on a yearly basis:

0.030 €/ litre (established from benchmark) vs 0.010 €/litre (FCHR treatment)
This result proves in line with our expectations; in fact, as already shown in paragraph 3.3 ‘Analysis of the energy consumption for the FCHR process’ – the FCHR process is improving the efficiency in terms of processed litres of apple juice for unit of energy spent in the process. From an economical point of view, we can state what follows:
o the FCHR pasteurization and homogenization plant investment costs range between 25.000€ - 120.000€ for apple juice treatment (for a 1.000 l/h and a 6.000 l/h plant capacity, respectively);
It is evident that, respect to commercial benchmarks, the hydrocavitation-based technology used by the FCHR plant for fluid foods processing brings a significant advantage in the case of apple juice, by reducing the initial expenditure for the plant purchase of a relevant amount. The main costs for the plant implementation are in fact associated to the rotor heads manufacturing and to the necessity to couple several reactors for scaling up the system to the desired – industrially appealing – production capacity. This aspect is extremely reduced when considering the treatment applied to apple juice, as a lower number of passages within the reactor proved to be necessary to perform the desired processes.

9) Dissemination, exploitation and training
The activities planned under WP7 have the main scope of supporting the technical development of the FCHR technology by communicating and interacting with an external audience, so as to raise awareness on the commercial potential of the hydrocavitation technology and to establish contacts with end-users, interested groups or stakeholders. The objective of this WP is also that of facilitating the uptake of results and the commercialization phase.

The dissemination and communication plan conceived in the course of the FCHR project was not a static tool, but rather a flexible and living document which was updated regularly by the Consortium members, following the developments of the new pasteurization and homogenization technology and the changing or evolving interests of the beneficiaries of the technology itself, the SMEs.
A table – revised and updated in the Second Period of the project - shows the plan, and reports the type of activity agreed, the responsible partner(s) and the timing at which communication was carried out, with a small description of what the partners delivered, and the type of audience addressed by the message.

Dissemination in FCHR project was meant as a continuative activity having the final aim of promoting the project developments and of keeping the potential end-users of the technology updated regarding the technology. The main actions carried out under this task were:
D7.1 press release – an initial message from the SMEs, launching the project and describing its technological objectives was done in Italian, English and Norwegian languages in Period 1, and then diffused through few portals. An updated version of the press release was created at M9 to support the diffusion of the news regarding the delivery of the FCHR preliminary prototype. Communicates and articles were sent to online magazines / websites to further promote the project work and progresses.
D7.2 - Website - Put on line in November 2012 with general and, obviously, non confidential information about the project. Available at: www.fchrtechnology.com. Together with the creation of the graphics for the website, the Project Coordinator also implemented the logo. The website was constantly updated in the course of the Second Period and until the end of the project; the objective is that of keeping the pages online for at least 2 years after the end of the research project to support the exchange of information with interested companies even out of the project framework.
FCHR brochure – with the purpose of supporting the dissemination activities and the future presence of the Consortium at events of relevance in the food processing sector, WIXTA implemented a preliminary project brochure to be spread via email/mailings/newsletters and, in the course of the Second Period of the project, created a poster to be possibly used at national/international events and fairs in the food sector.
Exhibitions and participation to events in the food processing sector have been proposed, and will be evaluated until the official end of the project. A list of European fairs and conferences where dissemination could be performed was reported in the Final PUDK deliverable.
The task ultimately generated D7.5 (video clip showing final project results), that took advantage of the SMEs’ You Tube accounts to be diffused and made available to the public. This was realized by WIXTA, and showed the main outcomes of the research funded by the REA in the project as well as the expected benefits of the hydrocavitation technologies applied to fluid foods.

Hydrodynamic cavitation technology has come of age from initial skepticism in the early to mid ‘90s to a well accepted fact that it can bring about intensification of processes such as emulsification, micro and nano-suspensions, cell disruption, water disinfection through pathogen destruction etc., through the physical and mechanical effects of the cavitation phenomena. Hydrodynamic cavitation is a better choice as compared to the sonochemical reactors due to the fact that high velocity returning fluid jet can be used also for bringing out spatial uniformity due to the fluid mixing. The energy efficiencies of hydrodynamically cavitating reactors for their physicomechanical cavitational effects are found to be substantially higher. Newer designs of hydrodynamic cavitating chambers are being patented quite regularly, showing the renewed interest of the research community.
A dedicated section of the Final PUDK has been created to report examples of hydrodynamic cavitation reactors developed by companies at a global level, so as to provide information on potential competitors of the FCHR technology. The step forward in the analysis of FCHR competitors is that of comparing the features of these reactors with the ones of the FCHR pasteurization and homogenization system. Here a list of what might be considered the set of direct competitors:
i) HYDRO DYNAMIC INC. – Shockwave Power Reactor
ii) SPX – APV cavitator
iii) QUANTUM VORTEX – KAP 1500 (associated with KAVITUS technology)
iv) DYNAMIC WATER TECHNOLOGIES – VRTX
v) CAVITATION TECHNOLOGIES INC. – Nano-Reactor TM
It must be stressed here that most of these existing hydrocavitation reactors on the market are still not directly linked to the Food and Beverage sector, even if some of their producing companies are periodically sending out communicates and articles referred to this type of application.
The only products that can be actually put in correlation with the FCHR system are (i), (ii) and (v) as they might represent direct competitors of our product, which is specifically developed for the food application. The table for comparison is structured so as to include the name of the producing companies and of their products, and, line by line, descriptions and data referred to the main analysis criteria. This list of analysis criteria includes both criteria which represent the strength of the FCHR innovation but also criteria representing general requirements of the market.
We must stress here that the table comparing FCHR with these technologies has been created by taking into consideration the final FCHR scaled-up system, meaning the cavitation reactor that will be adapted to an industrial scale to represent a product ready to compete with others in the market.
As will be detailed later on in this document, in fact, the main output of the project is a prototype and additional work will be required to make it a real final product for the market. The detailed post-project plan to reach this goal has been included in the exploitation strategy of this document.
A SWOT analysis for the FCHR plant has been also included in the PUDK Deliverable, reporting the envisaged strengths/weaknesses of the system, as well as the opportunities/threats of the external environment in which the FCHR solution will be introduced (the food processing market).

The exploitation strategy task involved activities aimed at planning and defining, in agreement with the other beneficiary SMEs in the Consortium, a preliminary exploitation plan for the project results. This was done by means of an intense email exchange in the final months of the project, by proposing discussion points in occasion of the Consortium meetings, with the support of information and materials exchange on the possible role of each partner in the exploitation of the FCHR machine.
All this material, as well as an analysis of the progress of the work in the project, led to the identification of an exploitation route that will be detailed in the Results section of this WP.
The key achievements of this task were:
o The definition of a commonly shared Exploitation Plan & Strategy, based on the lab results on the prototype that represented the ground for technical and economical considerations and that supported the analysis of the FCHR commercial potential;
o The advancement in the knowledge of the current patents/ publications status, and the identification of the main pieces of foreground representing the project IPs to be possibly protected;
o The evaluation of the future steps necessary to bring the FCHR technology in the market and the realization of a plan for the post-project activity leading to the FCHR commercialization, as well as the definition of a parallel marketing plan supporting the introduction in the market;
o The analysis of current technologies potentially hindering or mining the entry of the FCHR processing technology in the market, supported by an analysis of the direct competitors of the FCHR;
o Finally, training and know-how transfer have been taken into account; a small manual has been created by the RTDs for the SMEs to support the operation of the machine, and know-how transfer activities have been performed in occasion of the project meetings, when the prototype and its functionalities were shown. A complete training session, supported by presentations and materials, will be organized as part of the Final Meeting of the project.
The task produced the following deliverables:
- D7.4 Final plan for using and disseminating knowledge - The Plan for Use and Dissemination of knowledge is a document that summarizes the marketing action needed to commercially exploit the result of the project, according to the interest of the beneficiary SMEs, and includes details of dissemination strategy outside the consortium.
The outlined exploitation strategy for the project results is meant to have the following phases and roles:
• WIXTA, owner of the CHR patent, will maintain the ownership on the FCHR and the possibility to patent eventual upgrades emerging from the project. WIXTA will have the right to use the simulation model and the optimized design of the reactor without any constraint in any sector apart from food.
• For the application to the food sector of the FCHR, FENCO will have the exclusive license, upon agreement with WIXTA on the sector of application and/or territorial basis. WIXTA will have the right to license to third parties in other sectors/territories following this agreement.
• WIXTA will keep the FCHR after the end of the project, allowing the two end-users of the project to receive the system for 6 months each for tests on their production lines;
• ELEC will be exclusive manufacturer of the control system for the FCHR plant, upon definition of fair and reasonable purchase agreement at market conditions.
• EPLE will have the possibility to get the FCHR equipment from FENCO at special conditions, to directly exploit it in production facilities.
• Equally, GLENILEN will have the possibility to receive from FENCO, at special price, the FCHR system to be used in their daily milk processing activities.
• Both EPLE and GLENILEN will have the possibility to receive the system for free from WIXTA for tests in their production lines for an agreed period of time (6 months), incurring only the shipping expenses.

In summary, we can state that the main output of our research project is represented by the FCHR prototype that is now properly working and capable of heating and producing cavitation effects in the treated fluids. In this sense, even if ambitious, the scientific and technological objectives of the research are to be considered as achieved.
However, due to its prototypal nature, the reactor will need further optimization, engineering and tests to become a product ready to be introduced in the market. Before that, specific scale-up activities will have to be performed on the reactor to obtain an industrial-scale plant having processing capacities of about 6.000 l/h, which is the capacity actually processed by conventional pasteurizers/homogenizers.
According to what stated above, the SMEs – supported by their partner RTDs – have decided to draft a POST-PROJECT ACTIVITY PLAN, which describes, at a very high level, the work to be done after the end of the project to scale the FCHR reactor up to the desired production capacity, meaning the steps that will be necessary for the design and implementation of the FCHR industrial plant.

A possible way to run this phase is to find other funds for the follow-up of the project. The SMEs are exploring the possibilities in each country but in case it turns out not to be possible, a financial plan will be set up and agreed upon. At the moment, internal discussions are taking place among the partners; a direct involvement of the SMEs in the post-project phase is considered as possible and desirable, WIXTA and FENCO leading the activities.
A marketing and communication plan will be necessarily developed in parallel with the technical activities to promote the product and accompany the entry of the FCHR system in the market. Dedicated campaigns and personnel effort will be devoted to the definition of the communication strategy for the product and of the actions to support sales. The plan will be put into practice from the 1st testing phase on until the end of the engineering and scale-up work, in continuous contact with the technical staff providing information on the plant features, advantages respect to state-of-the-art systems and overall performance in treating milk and juices. These details will be of crucial importance to focus the message of communication and to be able to optimally structure a set of marketing actions supporting the delivery of the FCHR technology in the reference market.

10) Protecting IPRs
The main innovations in the project will be protected by means of patenting. In case relevant achievements in the food sector will allow for the filing of a further patent, the initial patent owner - WIXTA - will be in charge of evaluating the application and of submitting it to the Exploitation Manager (Mr. Ennio Ghillani from FENCO) and to the Consortium. An agreed strategy will be therefore outlined for the specific patent application.
Not all the results expected from the project represent patentable knowledge. At the moment we write, in fact, the main piece of foreground generated within the project framework that might be protected by patenting is represented by the CAD design of the optimized reactor and the layout of the pasteurizing/ homogenizing plant based on the effects of controlled cavitation as developed in the project lifetime.
Both the simulation work carried out and the control system of the FCHR are innovations with less patentable characteristics respect to the reactor design. However, even if not patentable, the other results of the project might represent a useful innovation for the SMEs, as they might increase the differentiation of the commercial offer of the owners or the know-how that they will be able to use in their businesses.
With the purpose of supporting the application for a patent, we have realized a patents / publications search, showing that no comparable system has already been developed for the food application, or is commercially available on the market yet.

As part of the exploitation activities, we must mention a relevant progress in the direction of the definition of the future use of the FCHR prototype. As already detailed in this section, at the end of the project the hydrocavitation prototype developed by the RTDs will be owned by WIXTA, which will be able to patent the new configuration and geometry of the reactor. The physical system, tested in Cork by the University College partners, could be available for the SMEs to test it in their facilities, integrating hydrocavitation as a novel technology in their production lines.
However, two main aspects related to the future use of the prototype emerge:
1. During internal discussions held in occasion of the Second Period project meetings, the partner SMEs have agreed to wait for the final processing plant implementation, scaled up to at least 1.000 l/h treatment capacity, following the path outlined in this Deliverable and reported in Section 3.4.2 (Route towards the FCHR commercialization).
This was decided because of the intention of the beneficiaries to receive a final, optimized system, ready to be installed in their facilities as a complete and stand-alone plant, to be possibly used in parallel with the conventional productions, at least at the beginning of its functioning. The FCHR prototype, as it was realized in the project, in fact, would need a dedicated integration activity in the companies’ production lines, investments for changes to these lines and adaptations that might be necessary to connect the reactor to existing systems, and this was considered as source of delays potentially affecting the companies’ businesses.
2. At the same time, the University College of Cork found that the field testing phase performed in the lifetime of our funded EU project raised great interest in the Food, Science and Environmental Sciences Faculty personnel, for which further developments and studies on the prototype and on cavitation effects inside the reactor would be highly encouraged. To this end, a prosecution of the activities on microbiological/chemical and physical tests on milk and other fluid foods – among which juices – would be very interesting, as they might produce advanced technological results going beyond the expectations pursued in the FCHR project. Students and academic staff might be interested in taking part to the field testing activities, to produce additional results and publish scientific papers supporting the FCHR technology development and use.

As a consequence of these considerations, a specific agreement has been created at the end of the project between WIXTA, owner of the final FCHR prototype, and UCC.
This will detail the terms and conditions by which WIXTA will grant UCC the permission to use the FCHR equipment, for a period of 1 year after the end of the project. The results of the tests representing the purpose of the Agreement will be owned by WIXTA, and publications will be regulated by a specific clause not to infringe confidentiality or to hinder the company’s commercial interests.
Potential Impact:
The 4 tangible results expected from the project are here listed and their output described in terms of commercial potential and impact.

1 – Numerical simulation model of the hydrodynamic behaviour
As a result of the simulation work carried out by UTV, the following concrete outputs were delivered:
- Consolidated set of guidelines for the geometrical optimization of the rotor;
- Consolidated set of guidelines for the geometrical optimization of the stator;
- Simulation graphs showing cavitation zones and relative intensity.
The simulation reports include the description of the modifications suggested in order to improve the efficiency of the reactor as they were suggested in the Second Period of the project. Fluid-dynamic simulations showed that the proposed modifications seem to produce important benefits to the production of cavitation. The presence of zones with cavitating flow is extended and the amplitude of the phenomenon increased. Result no. 1 will be owned by WIXTA (100% ownership), that will be able to use the simulation model for the design of the reactor in the food applications or for any other use of the hydrocavitator. In this sense, no direct commercialization is foreseen for this specific result. Its use will be – in any case - necessary for the future implementation of the prototype for food/other applications of the hydrocavitation technology.

2 – CAD design of the FCHR reactor and definition of the plant layout for the pasteurizer / homogenizer
As a result of the collaboration of UTV and LABOR for the design of the 2 FCHR prototypes, the following concrete outputs were delivered to the SMEs:
- The definitive, executive CAD design of the prototype as it was shipped to our partners at UCC for the tests;
- The list of the components and settings for the prototype operation;
- A small operation manual as a support to the users for the start of the tests on the prototype and for the solution of any potentially upcoming problem;
- 3 proposals of P&I (process and instrumentation scheme) for the final FCHR plant, meant as a final product to be implemented after the end of the project, validated by the Consortium’s industrial partners; the P&I schemes propose a pasteurization and homogenization plant scaled-up to the desired processing capacity;
At this stage, the final prototype has been delivered and the layout of the scaled-up plant has been provided; however, starting with a prototype, further optimization and engineering work will be needed to access the food processing markets and to implement a finite product.
Possible obstacles to the commercialization might be represented by initial skepticism of the producers towards a novel technology, the time for the engineering activity and for the validation of the technology at an industrial scale that may bring to a decrease in the stakeholders’ interest and finally the economic difficulty of several companies in investing in innovation. All these aspects have been evaluated in the SWOT analysis and a plan has been provided for facing these challenges.

3 – Control system for the control of cavitation inside the FCHR
As a result of the work done on the prototype by LABOR and supported by the simulation models of UTV and by the consolidated knowledge of WIXTA on cavitation effects, the know-how on the prototype’s motor was transferred to the beneficiary ELEC, whose main business lays in the design and implementation of electric motors. The foreground received as result of the FCHR project is thus meant to further enlarge the commercial offer of the company as it is specifically applied to hydrocavitation reactors.
The indications on how to drive the motor controlling cavitation inside the reactor were transferred to ELEC, as well as the datasheet of the selected motor of the final prototype, showing all the technical features and the operations that need to be done to properly induce cavitation phenomena in the treated fluid. The control system representing Result no. 3, developed specifically for the FCHR prototypes, does not represent a result having properties of a stand-alone product. This will in fact be necessarily coupled with the reactor to be sold and commercialized.
ELEC, as unique company operating in the field of motors and controllers’ development will benefit from the production of dedicated control systems for the reactors that will be introduced in the market by WIXTA.

4 – Results of tests carried out on the prototype with different food matrixes
The main output of the laboratory tests carried out in the course of the project on the small-scale prototype and on the final FCHR prototype was the microbiological assessment of cavitation applied to foods. Intense testing sessions were planned so as to validate the hydrocavitation technology for potentially improving pasteurization and homogenization effects on milk and juices. The results of tests, described in a complete an accurate report, were delivered to the beneficiary SMEs, showing the features of the treated food matrixes in terms of bacterial inactivation, fats reduction, safety, product quality, etc.
This result was mostly included in Deliverable D4.2 which reports laboratory data and graphs showing the effect of cavitation on milk/juices under the selected working conditions. A direct commercialization of this result is not expected and not possible, as it mostly concerns with the outputs of the lab tests showing the effectiveness of FCHR in treating milk and other fluid foods, achieving pasteurization and homogenization grades which should be equal to the conventional ones. Publications and articles reporting the results of these tests will be possible, on the contrary, and might be used for marketing and dissemination purposes in the next future, once the project has been concluded.

Target market for FCHR is pasteurizing and homogenizers equipments in the replacement and new plants in fluid food manufacturing: dairy industry, canning industry, condiments and sauces. The dairy sector recorded a particularly weak performance with regards to exports, losing on average 20% in export value in 2009 compared to 2008. However, it continues to be the fourth most important export product of the EU food and drink sector and the most innovative among the food sectors in Europe. For these reasons, this sector can be considered the primary reference market for the FCHR technology. However, also the canning industry is a relevant market where the results could be applied.
The objective of FCHR Consortium is to reach after 5 years of market development of the technology a market share of 5%. The market penetration would be based on an initial lower number of plants, going in parallel with the marketing phase, reaching the target number at year 5. It should be considered that the number of plants expected to be sold by FENCO, exclusive manufacturer on territorial/sector basis, is estimated to be 33% of the total number indicated. The remaining plants would be manufactured by other licensees of the technology in other sectors/markets.
The mechanical treatment introduced in FCHR will bring the following impact in the dairy market operation and business:
1) Scalability: the system will be easily scalable with parallel reactors with flow rates of several thousand liters per hour.
2) Energy Savings: it has been estimated that the FCHR process, coupling the two actually separate processes of pasteurization and homogenization of fluid foods, and mostly of milk, will allow for an energy saving of about 20% with respect to the actual expenditure for a mill processor; this is an important advantage of the FCHR technology, if we consider that pasteurization is the most energy-demanding process, representing (for milk, as an example) about 33% of heat and the 20% of the global electrical consumption for the overall process;
3) Pasteurization grade: Equivalent to that of traditional pasteurization;
4) Digestibility and quality of food: the new process will grant higher digestibility especially in the case of milk; in fact, lower temperatures are needed to achieve the same pasteurization grade, thanks to the cavitation energy, with respect to thermal gradient. This reduces the degradation of the nutrients and keeps the flavor of the food;
5) Energy Efficiency: thermal efficiency of heating > 90%, same as a heat exchanger;
All these features of the FCHR reactor will allow the dairy SMEs in Europe to increase their competitiveness and a valuable economic advantage will be gained with respect to traditional methods for pasteurization and homogenization. The potential market for the FCHR is therefore outstanding, due to the huge range of possible applications in the food sector and to the relevant number of involved SMEs in EU. Only considering the application to the dairy and canning sectors, with a market penetration reaching 5% of the plants installed after 5 years from the commercialization of the FCHR plants (about 115 plants/year of medium size), it has been estimated that the profits for the SMEs participants will get to about 6 M€/y.
An hypothesis of the commercial benefits that each of the partner SMEs will gain from the use of the FCHR results is reported here, as well as the strategic commercial objectives of each company.

WIXTA
Develop a detailed market analysis to access the food processors’ market, starting from the Italian one where the initial exploitation of the FCHR will be carried out;
Enter the market of milk processing by selling the FCHR reactor (3.000 l/h);
Target the milk and juices sector by scaling up the FCHR device to an industrial level
Optimize the geometry of the reactor and develop new configurations for the extension of the process to different food matrixes;
Investigation of new applications / markets where to exploit the new technology;

FENCO
Differentiate the company’s commercial offer by developing the new FCHR plant for milk and juices processing;
Starting the selling activity in the Italian market;
Evaluate the possibility to exploit the company’s established network to access other European markets;
Provide support to WIXTA in the optimization of the plant layout and in the validation phase;

ELECTRICARS
Optimization of the control system for the plant and production of the control for the sold plants for which the company will be exclusive producer;
Development of a marketing strategy together with WIXTA and FENCO to access the Maltese sector;
Investigation of new applications / markets where to exploit the knowledge gained with respect to motors and controllers;

EPLEBLOMSTEN
Support to the other partners in the definition of an exploitation strategy targeting the Norwegian markets and in particular the juices production one;
Support for the post-project activity aiming at verifying the effectiveness of the FCHR plant on the processing of different food matrixes;

GLENILEN FARM
Support to the other partners in the definition of an exploitation strategy targeting the Irish market and in particular the dairy products one, including cheese and yoghurts;
Support for the post-project activity aiming at verifying the effectiveness of the FCHR plant on the processing of different food matrixes;


Expected Benefits per participant

WIXTA
As exclusive producer of the FCHR reactor for FENCO and other licensees, the company will keep the system developed within the project framework and will target the Italian market by selling an estimated number of 210 reactors in the 3 years after the market entry of the technology, thus receiving revenues for about 2.36M€ (considering the price of a single reactor to be around 40K€).

FENCO
As exclusive manufacturer of the FCHR plants, once the post-project activity has been concluded and has delivered a complete, industrial-scale system, FENCO will sell about 23 plants in the Italian market after 3 years from the product commercialisation, gaining revenues for about 850k€ (considering the price of a single plant to be around 130.000€).

ELECTRICARS
The company’s role in the commercialisation of the FCHR reactors will be that of exclusive supplier of the control system for the sold plants; then, following the estimates performed for WIXTA, considering the 210 reactors sold after the start of the commercialization phase, ELEC will be gaining around 210k€ (considering the price of a single control system around 2.500€).

EPLEBLOMSTEN
As end-user, the company will receive the FCHR system for application in their production lines at a special condition from FENCO. Benefits for the company will be associated to energy savings (about 20%) respect to the current processes and about a 3% increase in revenues for the selling of their products (enhanced product quality, brand marketing, etc.).

GLENILEN FARM
As end-user, the company will receive the FCHR system for application in their production lines at a special condition from FENCO. Benefits for the company will be associated to energy savings (about 20%) respect to the current processes and about a 3% increase in revenues for the selling of their products (enhanced product quality, brand marketing, etc.).
List of Websites:
The FCHR website represents the main interface of the project towards the external, used by the Consortium to promote the research funded by the REA and to diffuse the concept of the hydrocavitation technology applied to foods, also showing the results that the project is achieving.
Deliverable D7.2 – Website, submitted at the very beginning of the project, had the main purpose of summarizing the structure of the FCHR webpage, whose domain, www.fchrtechnology.com was registered in October 2012 and put online at the beginning of November 2012.

The implementation of a public website has been planned in the DoW with the purpose of:
1. To raise awareness on the features of the new pasteurizing and homogenizing system into the target market (dairy and juices ones), and
2. To show the results of the project and to foster the application of new fluid foods processing technologies.

The contents of the webpage have been limited to all the non-confidential information that can be spread regarding the new approach of cavitation applied to foods, so a series of iterative revisions of the texts was made by the beneficiaries during the project and especially in occasion of the updates of the pages, in order not to infringe confidentiality.
In the course of the Second Period of the project a new version of the website was implemented and released by the Project Coordinator WIXTA, with the purpose of getting the information on the project more attractive to the public, and to provide additional details and materials on the FCHR technology.
The domain of the website was kept unchanged: www.fchrtechnology.com.

All the relevant contact details are reported in the dedicated page on the website (Wixta, Coordinating Company, contact details of the CEO, Mr. Cristian Isopo).