DESIGN OF FISH FEEDING MECHANISM FOR RECIRCULATION AQUACULTURE SYSTEM (RAS)

(1)

DESIGN OF FISH FEEDING MECHANISM FOR RECIRCULATION AQUACULTURE

SYSTEM (RAS)

NITHIN SIVAKUMAR

Master of Science Thesis TRITA-ITM-EX 2019:523 KTH Industrial Engineering and Management

Machine Design SE-100 44 STOCKHOLM

(2)

(3)

Examensarbete TRITA-ITM-EX 2019:523

Utformning av fiskmatningsmekanism för återcirkulation av vattenbrukssystem

Nithin Sivakumar

Godkänt

2020-09-22

Examinator

Ulf Sellgren

Handledare

Stefan Björklund

Uppdragsgivare

{Namn}

Kontaktperson

{Namn}

Sammanfattning

Fiskodling eller vattenbruk är en växande livsmedelsproducerande industri för att odla fisk i konstgjorda tankar. Ett stort antal fiskar uppföds i tankar i 100-150 dagar och producerar mat för befolkningen. Med ökande befolkning och minskande världsfångstfiske uppfattas vattenbruk som en potentiell teknik för att möta den ständigt växande efterfrågan på livsmedel utan att skada vattenlevande livsmedelskedjan. Emellertid hämmar problem som tas upp i nuvarande system som mänsklig intervention och oanvänt foderråvara produktions hastigheten. Denna avhandling ger metoder och riktlinjer för att lösa utmaningarna och förenkla de mekaniska designaspekterna av lagring, transport och kontroll av utmatning processen, vilket i slutändan gynnar småskaliga producenter och entreprenörer. Lagring och distribution av fiskfoder längs vätskeflödet innebär att man överväger tvärvetenskapliga tekniska beräkningar. Den grundläggande kunskapen om fluidmekanik, diskreta element egenskaper och mekanisk design har visat sig vara en effektiv lösning för sådana utmaningar. Resultaten av detta arbete ger information om utformningen av fiskmatning mekanismen som innehåller vätskeflöde partikeldosering och utvärdering av komponenter som finns i systemet. Examensarbetet tillhandahåller lösningar som fungerar som utgångspunkt för lågkostnads konstruktion, validering och automatisering av komponenter i en utfodringsmekanism för vattenbruksindustrin.

Nyckelord: Vattenbruk, Hopper, Diskreta element, Mekanisk design.

(4)

(5)

Master of Science Thesis TRITA-ITM-EX 2019:523

Design of fish feeding mechanism for Recirculation Aquaculture System (RAS)

Nithin Sivakumar

Approved

2020-09-22

Examiner

Ulf Sellgren

Supervisor

Stefan Björklund

Commissioner

{Name}

Contact person

{Name}

Abstract

Fish farming or Aquaculture is a growing food-producing industry to culture the fish in artificially constructed tanks. A large number of fishes are reared in tanks for 100-150 days and produce food for the population. With increasing population and declining worldwide capture fishery, Aquaculture is perceived as a potential technique to meet the ever-expanding food demand without tarnishing the aquatic food chain. However, problems which are addressed in current systems like human intervention and unused feed material inhibit the production rate. This thesis provides methods and guidelines to resolve the challenges and simplify the mechanical design aspects of storage, transportation, and control of the dispensing process, which ultimately benefits the small- scale producers and entrepreneurs. Storing and distributing the fish feed along the fluid stream involves in considering multi-disciplinary engineering calculations. The fundamental knowledge of fluid mechanics, discrete element properties and mechanical design have been found to be an effective solution for such challenges. The results of this work provide information on designing the fish feeding mechanism incorporating fluid stream particle dispensing and evaluation of components present in the system. The thesis work provides solutions that serve as a starting point for low-cost design, validation and automation of components in a feeding mechanism for aquaculture industries.

Keywords: Aquaculture, Hopper, Discrete elements, Mechanical design.

(6)

(7)

FOREWORD

Firstly, I would like to thank my parents for their endless support from India and sacrifices over the years to help get me where I am. I am eternally grateful to Sherjeel Ton, CEO at Cross- Disciplinary Engineering, for all the support, practical help, and patience throughout this project.

Sincere thanks to Cross-Disciplinary Engineering for supporting me in this project. I want to thank Associate Professor, Stefan Björklund, at KTH Royal Institute of Technology for all his frequent and vital advice, ideas, and encouragement. I would also like to thank Håkan Karlsson, Technical Manager at MAFA for his tireless assistance on both theoretical and practical matters.

(8)

(9)

NOMENCLATURE

Notations

Symbol Description

𝑑_𝑠 𝑆𝑐𝑟𝑒𝑤 𝑎𝑢𝑔𝑒𝑟 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 (𝑚𝑚)

𝑑_𝑐 𝑆ℎ𝑎𝑓𝑡 𝑜𝑢𝑡𝑒𝑟 𝑑𝑖𝑎𝑚𝑡𝑒𝑟 (𝑚𝑚)

𝑟_𝑝 𝑅𝑎𝑑𝑖𝑢𝑠 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑝ℎ𝑒𝑟𝑖𝑐𝑎𝑙 𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒 (𝑚𝑚) 𝐶𝐹₁ 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑝𝑖𝑡𝑐ℎ 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟

𝐶𝐹₂ 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑓𝑙𝑖𝑔ℎ𝑡 𝑚𝑜𝑑𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 𝐶𝐹₃ 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑚𝑖𝑥𝑖𝑛𝑔 𝑝𝑎𝑑𝑑𝑙𝑒 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 𝜌_𝑏 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝑎𝑠 𝑐𝑜𝑛𝑣𝑒𝑦𝑒𝑑 (𝑘𝑔/𝑚³)

𝐹₀ 𝑂𝑣𝑒𝑟𝑙𝑜𝑎𝑑 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝐿 𝐿𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑐𝑟𝑒𝑤 𝑎𝑢𝑔𝑒𝑟 (𝑚)

𝑒 𝐷𝑟𝑖𝑣𝑒 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦

𝐹_𝑚 𝑀𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝐹𝑎𝑐𝑡𝑜𝑟

𝐹_𝑓 𝐹𝑙𝑖𝑔ℎ𝑡 𝑚𝑜𝑑𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝐹_𝑝 𝑃𝑎𝑑𝑑𝑙𝑒 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝐹_𝑑 𝐶𝑜𝑛𝑣𝑒𝑦𝑜𝑟 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝐹_𝑏 𝐻𝑎𝑛𝑔𝑒𝑟 𝑏𝑒𝑎𝑟𝑖𝑛𝑔 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝑁 𝑆𝑝𝑒𝑒𝑑 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑐𝑟𝑒𝑤 𝑐𝑜𝑛𝑣𝑒𝑦𝑜𝑟

𝐻𝑃_{𝑡𝑜𝑡𝑎𝑙} 𝑇𝑜𝑡𝑎𝑙 𝐻𝑜𝑟𝑠𝑒𝑝𝑜𝑤𝑒𝑟 (kW)

𝐻𝑃_𝑓 Frictional Horsepower (kW)

𝐻𝑃_𝑚 Material Horsepower (kW)

𝑃₂ Pressure of water at the Pipe junction (Pa) 𝑃₃ Pressure of water at the tank surface (Pa) 𝑉₂ Velocity of water at the Pipe junction (m/s) 𝑉₃ 𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑦 𝑜𝑓 𝑤𝑎𝑡𝑒𝑟 𝑎𝑡 𝑡ℎ𝑒 𝑡𝑎𝑛𝑘 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 (m/s)

𝜌 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 𝑜𝑓 𝑤𝑎𝑡𝑒𝑟 (𝑘𝑔/𝑚³)

𝑔 𝐴𝑐𝑐𝑒𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑑𝑢𝑒 𝑡𝑜 𝑔𝑟𝑎𝑣𝑖𝑡𝑦 (9.8 𝑚/𝑠²) 𝑉_{ℎ𝑜𝑝𝑝𝑒𝑟} 𝑉𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 ℎ𝑜𝑝𝑝𝑒𝑟 (𝑚³)

𝑉_{𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒} 𝑉𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑝ℎ𝑒𝑟𝑖𝑐𝑎𝑙 𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒 (𝑚³) 𝑀_𝑝 Mass of single-feed particle (grams)

𝑁_𝑝ℎ Total number of particles that can be accommodated inside the hopper 𝑀_𝑝ℎ Total mass of the particles accommodated inside the hopper (kg)

(10)

𝑁_𝑝𝑝 Number of particles that can be accommodated between one pitch of the screw conveyor

𝑀_𝑝𝑝 Total mass of the particles accommodated between one pitch of the the screw conveyor

𝑊_𝑝𝑒 Total weight of pellets in conveying tube (kg) 𝐿_𝐴𝑔 Length of flexible screw auger (m)

𝑊_𝐴𝑔 weight per unit length of flexible screw auger (kg/m)

𝑃𝑑 𝑃𝑎𝑐𝑘𝑖𝑛𝑔 𝑑𝑒𝑛𝑠𝑖𝑡𝑦

(11)

Abbreviations

AISI American Iron and Steel Institute

ASTM American Society for Testing and Materials

CAD Computer Aided Design

CEN European Committee for Standardization

DN Diameter Nominal

FCR Food Conversion Rate

PVC Poly Vinyl Chloride

SS316 Stainless steel AISI 316

RAS Recirculation Aquaculture System

SMS Sveriges Mekanförbunds Standardcentral

(12)

1 INTRODUCTION

This report focuses on the design and evaluation of storage, transportation and dispensing components present in the fish feed mechanism for Recirculation Aquaculture System (RAS).

1.1 Background

Recirculation Aquaculture System is an emerging technology in the field of Aquaculture systems to meet the demand for seafood in the market. The ever-expanding seafood demand is depleting the resources in the ocean. Recirculating aquaculture systems offer sustainable solutions for meeting such requirements by using closed-loop water treatment and circulation [1]. One of the essential subsystems in a RAS is the grow-out tank. Thousands of fishes are raised in these grow- out tanks for 100-150 days depending on the species. A process diagram indicating the working of a RAS system is shown in Figure 1.

Figure 1 Recirculation Aquaculture System

3 Bio-Filtration (ammonia removal) Feeding

mechanism

Fish wastes and uneaten feed

Mechanical filter 2

Solids Removal 1

Grow-out Tank

Bio- Filter 4

Dissolved gas Control (Oxygenation

)

Aeration Unit Recirculated

water “in”

Recirculated water “out”

(15)

1.2 Purpose

Meeting the food demand for the culture of the fishes in the aquaculture system is a critical process as they incorporate a regular feeding schedule. There are several working models for dispensing the fish feed on the water surface in culture tanks [2], but there is no existing design which incorporates feed dispense with the fluid stream. In the fluid stream dispensing method, the feed particles are not allowed to settle down due to water current, which significantly reduces the amount of uneaten fish feed. In addition to this, surface dispensing results in bloating of fish and eventually reduces its life. The thesis work gives information on designing of a feeding mechanism for Recirculation Aquaculture System (RAS) and evaluation of the storage, transportation and dispensing components present in the fish feeding mechanism. The feeding mechanism incorporates fluid stream fish feed dispensing and control of feed rate (kg/hr) to maximize the feed conversion rate (FCR) or to minimize the amount of uneaten feed.

1.3 Delimitations

The thesis work was carried out with assumptions in places where there is a lack of experimental data. The assumptions made are closely associated with the real-time situation. The delimitations that exist in this work are presented below

 The Hopper deformation was simulated by assuming the Bulk material as a liquid which exerts pressure equally in all the directions. The behaviour of the discrete elements was not taken into consideration in the static structural simulation.

 The forces exerted by the discrete element particles on the components of the driving shaft and driven shaft sub-assembly are assumed with a conservative factor of safety.

 The sensor signals used for simulating the closed-loop control model developed in Simulink were generated manually to check the feasibility of the control algorithm. No experiments were conducted to extract the data for the simulation.

1.4 Research questions

The various research questions that were addressed and answered in this thesis are,

 Is it possible to facilitate fluid stream dispensing of feed particles ?

 How far can material intensive design be elimated in the feeding mechanism ?

 To what extent can the human intervention be reduced in the feeding process?

(16)

2 FRAME OF REFERENCE

2.1 Recirculation Aquaculture System (RAS)

In a recirculation aquaculture system, the feeding mechanism plays a significant role in deciding the overall efficiency, i.e. the Feed Conversion Rate (FCR). The feeding device in this thesis consists of four specific parts, namely the feed pellets, hopper to store the feed pellets, screw auger for feed transportation and a dispensing approach.

2.1.1 Relevant theory

The basic definitions and the technical terms used throughout the report are consolidated and elaborated in this subsection.

Bulk density of a powder contained in a known volume is the ratio of the Mass of the Powder to the Volumetric capacity of the container, including the inter-particle void volume. Bulk density is expressed in kg/m³. Bulk density depends on the spatial arrangement of the particles in all the layers of the powder bed. A powder contained in a known volume can have a range of Bulk density depending on its preparation, treatment, and storage.

Packing density (Pd) is a parameter that defines how efficiently you arrange particles in a given volume with minimum void space between the particles in 3D dimensional space. For perfectly ordered and closely packed spheres, the packing density is found to be 0.74 according to [3] and decrease till 0.54 for random packing and different encapsulated volumes [4] (see Figure 2 [5]).

Untapped Bulk Density is defined as the Bulk density of the powder as it is stored. Even the slightest tap of the container rearranges the spatial arrangement of the powder layers to fill in the interparticle voids.

Figure 2 Random packing (left) and perfect packing (right) of dice in a cylinder

(17)

The Powder's Effective Angle of Internal Friction is the measure of friction within the adjacent particles when its sheared internally. The effective angle of internal friction is stated in degrees.

The Hopper Wall Angle (𝜽) in Figure 3 represents the angle in degrees of the conical hopper wall measured from a line drawn vertically along the axis of the hopper.

The Powder's Wall Friction Angle expressed in degrees is the measure of the sliding friction at the powder-wall interface. In this study, the powder is the fish feed pellets.

Feed Conversion Rate (FCR) also called the Aquaculture production efficiency, is the ratio of kg of feed dispensed in the culture tank to the kg of fish produced.

𝐹𝐶𝑅 =𝑘𝑔 𝑜𝑓 𝑓𝑒𝑒𝑑 𝑑𝑖𝑠𝑝𝑒𝑛𝑠𝑒𝑑 𝑘𝑔 𝑜𝑓 𝑓𝑖𝑠ℎ 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑 2.1.2 Feed pellets

The feed pellets shown in Figure 4 are to be stored and transported to the culture tank. The size of the fish food (Feed) dispensed in the pipeline system varies from time to time. For the first 51 days, 4.5 mm diameter spherical pellets are dispensed in the culture tank. With growing fish size, 6.5 mm pellets are distributed in the tank for the next 51 days. Throughout this report, the calculations are done for the larger 6.5 mm pellets since they occupy more volume inside a hopper for a specified weight (for example, 3000 kg). The image of the 6.5 mm feed pellets is shown in Figure 4. The constituents of the feed pellets are Crude protein, Crude fat, Carbohydrates, Fibre, Ash and Phosphorus at different Proportions.

Figure 3 Hopper wall angle

Figure 4 Feed pellets

(18)

2.1.3 Hopper

The picture illustrated in Figure 5 represents a hopper which is a cylindrical tank-like structure used to store granular materials. It consists of conical bottom part attached to one or more cylindrical sections placed on top of each other. The hopper is constructed by bending and rolling of steel sheets ranging from 1 mm to 6 mm thickness into a hollow cylindrical structure and held together with the help of rivets. Support structures are provided to the hopper to prevent it from collapsing. The Hopper material changes from one manufacturer to another based on their structural requirements. In general, the thickness of the sheet metal in the hopper varies along its vertical dimension, i.e. the bottom cone part has a comparatively higher material thickness to withstand the weight of particle layers above it.

Though the flow of particles inside the hopper seems easy, it involves in the calculation of several parameters like powder's wall friction angle, semi-included angle, interparticle friction etc. These parameters are required to determine the type of flow inside the hopper. There are two different types of flow inside a Hopper, namely mass flow, and funnel flow (see Figure 6).

Conical Bottom

Outlet

Cylindrical tank

Figure 5 Schematic representation of hopper for storing granular materials

a) b)

Figure 6 a) Funnel Flow b) Mass flow

(19)

In mass flow during discharge as depicted in Figure 7, the first portion of the granular material entering the silo is the first to exit through the hopper outlet, i.e. it provides first-in, first-out flow with all the particles in motion during discharge. Figure 8 shows the funnel flow, where the central portion of material forms a channel and flows out first leaving behind stagnant zones inside the hopper. The mass flow has several advantages over the funnel flow. It can be observed from Figure 7; the motion of the particles is uniform and steady [6]. The Bulk density of the powder discharged remains constant and is independent of the height of the Hopper/Silo. As a result, the particle layers remain non-cohesive throughout the discharge operation taking place inside the hopper.

However, the interparticle friction and wall friction due to the continuous flow of powders erodes the hopper wall in the long run. In such cases, funnel flow or core flow is used. The funnel flow shown in Figure 8 has a predominant flow of particle stream through the central channel compared to a nearly stagnant flow at the hopper walls [6]. Note that a hopper giving mass flow with one type of particle necessarily does not provide mass flow with different particle. According to [7]

the flow of the particles inside the hopper depends on

 Semi-included angle

 Wall friction angle

 Effective angle of internal friction

 Powder's flow function

 Untapped Bulk density

In general, the material used for the hopper construction changes from one manufacturer to another depending on the requirements. These requirements are drafted based on the type of material to be stored in the hopper, its abrasiveness, and the maximum weight of the Bulk material it can hold.

The RAS system discussed in this report requires a durable hopper which is environment friendly and has zero adverse effects on the fish feed particles. On extensive research, two materials, namely Stainless steel AISI 316 [8] and Mild Steel coated with Aluzinc (DX51D+AZ) [9], were identified to be commonly used in the food industries for the transporting bulk materials. These

Figure 8 Funnel Flow of discrete elements in a hopper Figure 7 Mass Flow of discrete elements in a hopper

(20)

two materials are environment friendly with good corrosion resistance and mechanical properties.

The latter part of the report gives a detailed explanation of the tests carried out for evaluating the mechanical properties of the hopper material.

2.1.4 Screw conveyors

The Screw conveyor system transports the particles from inlet to outlet with the help of a screw auger. It consists of multiple hollow shafts supported at the ends and connected using a solid shaft with fastening elements. A motor drives the auger through the power transmission components.

To construct the screw auger, the inner diameter of the spiral helix is welded onto the outer diameter of the hollow shaft (see Figure 9 [10]). This auger setup is enclosed by a trough to move the materials in a predefined path. The helix face of the screw auger provides the necessary thrusting force to drive the feed pellets towards the outlet. The particles are sheared and tumbled continuously by rotary, and linear motion exerted on them by the screw auger. Each rotation of the screw auger transports the feed material present between screw auger pitch through continuous tumbling and shearing action.

There are distinct augers which are shaftless throughout its length except for end cap and the motor.

These screw augers provide comparatively higher mass flow rate due to larger volumetric space available between the conveying tube and the screw auger. The two ends of the shaft are supported by bearings to avoid eccentric turning. One of the significant advantages which can be observed from Figure 10 of shaftless screw augers is that they provide flexibility with length [11], enabling them to fit into pipes having a certain radius of curvature.

Figure 10 Flexible Screw Conveyor Conveying tube

Motor

End cap

Figure 9 Conventional screw conveyor for transporting Bulk materials

(21)

2.1.5 Pipeline system

The pipeline system integrates the hopper-screw conveyor system with the culture tank. Figure 11 shows the mainline pipe connecting the feeder system split into three sections, namely the pump end, junction end, and the tank/nozzle end. All three parts use the pipework components based on the DN150 nominal size. The water is pumped to the culture tank at a specified pressure and velocity. This portion of piping until the screw auger system is called the pump section. The T-junction is connected to the screw conveyor to dispense the particles into the fluid stream by gravity. The final portion of pipeline extending from junction end till the nozzle at the tank is called the tank section. The pipe material used here is PVC, and the pipe standard followed is CEN/TC 155 [12]. Using the available pipeline parameters such as the density of the fluid, velocity of the fluid and the vertical height of fluid level from the ground, the minimum pressure required to pump the water to a defined head (in metres) can be calculated using Bernoulli's theorem [13]. The details of the pipe dimensions are provided in Table 1 and its visual representation in Figure 12.

Table 1 Pipe Dimensions according to CEN Standard

Pipe Specification: DN 150

Material PVC

Inside Diameter (ID) 154.08 mm

Outside Diameter (OD) 168.3 mm

Pipe Wall Thickness (t) 7.11 mm

Surface roughness, Ra 1.5 µm

Figure 11 Pipeline layout in RAS system

Pump end

Junction end

Water from pump

Culture tank T-Junction

Integration with delivery tube of the conveyor

Tank section Nozzle

(22)

Figure 12 Dimensions of DN150 based on CEN Standards

2.1.6 Control system

A control system directs and edits the parameters of a continuous time-variant system in such a way that the desired output is produced [14]. The two significant attributes of a control system are its stability and its ability to retain the desired output. The two types of control systems are

 Open-loop and

 Closed-loop control systems

An open-loop control system takes in the time-variant input given and gives the results as produced. There is no control over the value of the output produced irrespective of the desired value of the result. There are no external or internal disturbances encountered that affects the system. The components of an open-loop control system are shown in Figure 13.

In a closed-loop control system, as depicted in Figure 14, feedback control is incorporated that adjusts the parameters of the system to get the desired output. The difference between ideal value and measured value from the sensor is taken as an error and added as feedback. A closed-loop system minimizes the error between what is measured and what is desired iteratively. As a result, the desired result is obtained and maintained for successive cycles.

Figure 14 Process flow diagram of a Closed-Loop control system Figure 13 Process flow diagram of an Open-Loop control system

OD

ID t

(23)

A control system has different components like Reference or setpoint, Controller, Actuator, and Plant. The components are sequenced as illustrated in process flow diagrams. An actuator is a component that does the work to obtain the desired output. The object whose output must be controlled is called the plant. The reference or set point is the desired output to be produced from the system. The output signal measured using the sensor is compared with the desired output. The difference in value is termed as an error in the signal. The controller acts based on the computed error and sends the command to the actuator to correct its parameters to get the desired output. If the data from previously installed sensors for an intended application are not available, a logical solution is to model the sensor values as Gaussian distribution [15].

2.1.7 Bulk solid flow sensor

A Bulk solid flow sensor shown in Figure 15 is used for measuring the mass flow rate of Bulk solid flow inside an Open or closed channel [16] [17]. It accurately predicts the number of particles or the weight of the Bulk material flowing through channels. Bulk solid flow sensor works under the principle of physical doppler effect [18]. According to this principle, the sensor generates a uniform electromagnetic field in microwave frequency, and it calculates the number of particles passing through the field. Every particle passing through this electromagnetic field reflects the microwaves to the sensor's receiver. The main advantage of this sensor is that it can be integrated into the existing systems having a vertical channel or a horizontal channel effortlessly.

Figure 15 Bulk solid flow sensor application Particle flow

Bulk solid flow sensor

Open Channel Sensor Closed Channel

Sensor

(24)

3 IMPLEMENTATION

3.1 Methodology

3.1.1 Hopper selection

For selecting the Hopper, maximum weight/load to be stored is first determined. Then a safety factor is considered since the mass to be stored may vary. The commercially available hoppers are explored, and the number of particles that can be held within the given volume (𝑁_𝑝ℎ) is calculated using equation (1) where 𝑉_{ℎ𝑜𝑝𝑝𝑒𝑟} indicates the volume of the selected hopper and 𝑉_{𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒} indicates the volume of the feed particle. The total mass of all the particles that can be filled (𝑀_𝑝ℎ) is calculated using equation (2). For the mass experiment, 15 trials were conducted by measuring the mass of a single particle ( m_p) in a 25 ml container, as shown in Figure 16. The experimental observations are provided in under section 4.1.1.

𝑁_𝑝ℎ = (𝑉_{ℎ𝑜𝑝𝑝𝑒𝑟}

𝑉_{𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒}) 𝑃𝑑 (1)

𝑀_𝑝ℎ = 𝑁_𝑝ℎ𝑚_𝑝 (2)

3.1.2 Hopper evaluation

The selected Hopper is evaluated for structural integrity to ensure that it does not collapse during its service time. The Hopper evaluation that was carried out in ANSYS® Workbench™

and

is explained in detail.

The steps involved in the Hopper design analysis are,

 Create and import a simplified CAD model in ANSYS

 Definition of material properties

 Discretizing the 3D model

 Setting up boundary conditions

 Defining the loads

 Simulation and Extraction of results (Stress and strain)

 Parametric analysis

Figure 16 Particles used for mass measurement

(25)

Figure 17 Translation of engineering drawing to 3D CAD model

Figure 18 Defining the fixed support for static structural analysis

The first step involved in the analysis is to transform the engineering drawing of the Hopper and Hopper bottom given by the supplier MAFA AB to 3D CAD model using Autodesk Inventor Professional 2020. From the dimensions given in the engineering drawing, the hopper assembly was reconstructed into a simplified model, as shown in Figure 17. The two components were then assembled and imported as a single rigid body in ANSYS to perform the structural analysis.

Hopper

Hopper bottom

Fixed support

(26)

In the Main window of ANSYS Workbench, Static structural analysis is selected from the toolbox.

Since two materials are used, namely Mild Steel and SS316, the analysis is first performed using SS316 as hopper material. Then the entire setup, including the parametric study, is replicated for Mild Steel. The properties of Mild Steel and SS316 are not available in the material library of ANSYS, so the physical properties of the material were imported from Outokumpu product catalogue [8]. After setting up the material properties, the simplified 3D CAD assembly of the Hopper was imported as a Step file.

In the Workbench editor menu, the material for the Hopper and the Hopper's bottom is assigned as SS316. All the components in the model are then meshed using an adaptive size meshing. The meshed model of the hopper assembly is shown in Figure 19. After meshing, the next step is to define the loads and boundary conditions. From the drawing shown in Figure 17Figure 18, it is evident that the entire load of fish pellets that is to be carried by the Hopper is supported at the base by horizontal structural members and cylindrical body using vertical structural members. The two surfaces highlighted in blue colour shown in Figure 18 were defined as fixed support.

Figure 19 Discretization of the imported CAD model in ANSYS

Liquid level

Hydrostatic pressure exerted throughout the hopper length

Hopper inner walls

(27)

Since ANSYS does not support discrete element simulations, a major assumption is made in this step by considering the discrete elements as a fluid of equivalent Bulk density. The approximated fluid is modelled as a hydrostatic pressure acting on the inner walls of the hopper assembly, having the density equal to the Bulk density of the feed pellets (see Figure 20). A factor of safety has been implemented, and the analysis is repeated for liquid with density 1000 kg/m³ (density of water). The analysis is made simpler by parameterizing the study (see Figure 21) with a density as the Design variable, and the results were consolidated based on maximum von-mises stress acting on the hopper walls for both the materials.

3.1.3 Selection and assessment of conveyor system

From the literature study, two system concepts were selected for the conveying system, namely the Conventional Screw Auger and Shaftless Screw Auger. Both have their advantages and disadvantages. In section 4.2, a Pugh’s evaluation matrix [19] is constructed for conventional screw conveyor and shaftless screw conveyor based on the following factors.

 Material requirement

 Design complexity

 Conveying efficiency

 Maintenance and service

3.1.4 Calculation of power and torque requirements

The screw conveyor system connects the Hopper with the pipeline system. The particles from the Hopper are transported to the pipeline system via the screw auger. Irrespective of the type of auger, the motor power is calculated for the equivalent commercially available auger with shafts (refer Figure 9 and Figure 27). This is a conservative approach as the Power required for driving the conventional screw conveyors is comparatively more than shaftless screw conveyors. From the conveyor manual referred (see appendix A), the calculation of power and torque requirements is done for the auger having the following dimensions as specified in Table 2.

Table 2 Dimensions of the screw auger from CEMC screw conveyor manual Screw Auger specifications

Diameter of the Screw, 𝑑_𝑠 152.4 mm

Diameter of the shaft, 𝑑_𝑐 25.4 mm

Pitch distance 152.4 mm

Radius of the feed particle, 𝑟_𝑝 3.25 mm

Figure 21 Parametrizing the design variables for simulation

(28)

The number of particles to be dispensed per pitch of the conventional screw auger is determined by the following formula,

𝑁_𝑝𝑝 = 𝜋

4(𝑑_𝑠²−𝑑_𝑐²) 4

3𝜋𝑟_𝑝³

(𝑃𝑑) (3)

Where,

𝑑_𝑠− 𝑆𝑐𝑟𝑒𝑤 𝑎𝑢𝑔𝑒𝑟 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 𝑑_𝑐− 𝑆ℎ𝑎𝑓𝑡 𝑜𝑢𝑡𝑒𝑟 𝑑𝑖𝑎𝑚𝑡𝑒𝑟

𝑟_𝑝− 𝑟𝑎𝑑𝑖𝑢𝑠 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑝ℎ𝑒𝑟𝑖𝑐𝑎𝑙 𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒

The packing density of the sphere in a cylindrical volume for random loose packing is 0.54 [4]

Based on the number of particles per pitch ( N_pp), mass per pitch (M_pp) can be determined by multiplying the weight of a single particle ( mp).

M_pp = N_ppm_p (4) The single-particle mass used is the average mass of 15 particles. The experimental data for finding the single-particle weight is given in section 4.1.1. This approach to finding the number of particles in the volumetric space between the pitch is for the general purpose only. Actual experimentation must be done by filling the space with particles experimentally or by using the discrete element simulations to find out the mass occupied in that space. The performance of screw auger depends on numerous factors such as percentage of conveyor loading, bearing factor, friction horsepower, material horsepower etc. The formulas governing screw auger calculations according to [20] are given below.

𝑅𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 =𝑚𝑎𝑠𝑠 𝑡𝑜 𝑏𝑒 𝑐𝑜𝑛𝑣𝑒𝑦𝑒𝑑 𝑝𝑒𝑟 ℎ𝑜𝑢𝑟

𝑏𝑢𝑙𝑘 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 (5) 𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 (𝑓𝑡³

⁄ ) = (𝑅𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦)𝐶𝐹ℎ𝑟 ₁𝐶𝐹₂𝐶𝐹₃ (6) Where,

𝐶𝐹₁− 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑝𝑖𝑡𝑐ℎ 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟

𝐶𝐹₂− 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑓𝑙𝑖𝑔ℎ𝑡 𝑚𝑜𝑑𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 𝐶𝐹₃− 𝑆𝑝𝑒𝑐𝑖𝑎𝑙 𝑠𝑐𝑟𝑒𝑤 𝑚𝑖𝑥𝑖𝑛𝑔 𝑝𝑎𝑑𝑑𝑙𝑒 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟

The capacity per rpm is selected from the screw conveyor design manual. The total power required by the motor to drive the screw auger is the sum of power needed to overcome friction (𝐻𝑃_𝑓) and power required to transport the material multiplied (𝐻𝑃_𝑚) by overloading factor (𝐹₀).

𝐻𝑃_{𝑡𝑜𝑡𝑎𝑙} =^(𝐻𝑃^𝑚^+𝐻𝑃^𝑓^)𝐹^𝑜

𝑒 (7) 𝐻𝑃_𝑚 =𝐿𝑁𝐹_𝑑𝐹_𝑏

10⁶ (8) 𝐻𝑃_𝑓 =^𝐶𝐿𝜌^𝑏^𝐹^𝑚^𝐹^𝑓^𝐹^𝑝

10⁶ (9)

(29)

𝐶 − 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦 𝑟𝑒𝑞𝑢𝑖𝑟𝑒𝑑 𝜌_𝑏− 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝑎𝑠 𝑐𝑜𝑛𝑣𝑒𝑦𝑒𝑑 𝐹_𝑚− 𝑀𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝐹𝑎𝑐𝑡𝑜𝑟 𝐹_𝑓− 𝐹𝑙𝑖𝑔ℎ𝑡 𝑚𝑜𝑑𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝐻𝑝 𝑓𝑎𝑐𝑡𝑜𝑟

𝐹_𝑝− 𝑃𝑎𝑑𝑑𝑙𝑒 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟 𝐹_𝑑 − 𝐶𝑜𝑛𝑣𝑒𝑦𝑜𝑟 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟 𝐹_𝑏− 𝐻𝑎𝑛𝑔𝑒𝑟 𝑏𝑒𝑎𝑟𝑖𝑛𝑔 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟 𝐹₀ − 𝑂𝑣𝑒𝑟𝑙𝑜𝑎𝑑 𝐻𝑃 𝑓𝑎𝑐𝑡𝑜𝑟

𝑒 − 𝐷𝑟𝑖𝑣𝑒 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 𝐿 − 𝐿𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑐𝑟𝑒𝑤 𝑎𝑢𝑔𝑒𝑟 𝑁 − 𝑆𝑝𝑒𝑒𝑑 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑐𝑟𝑒𝑤 𝑐𝑜𝑛𝑣𝑒𝑦𝑜𝑟

Detailed information on the different factors and the instructions for selecting them are mentioned in the screw conveyor design manual and the Matlab code in Appendix A.

3.1.5 Control system

The control system for the feedback control is modelled in Simulink® plugin in Matlab [21]. The primary objective of the control loop in this application is to maintain a constant mass flow rate by tweaking the rpm of the motor. The entire structure of the control system modelled in Simulink is shown in Figure 22. Figure 23 illustrates that the bulk solid flow sensor is fit through a hole on the surface of the vertical pipe. The sensor setup acts as a closed channel bulk flow sensor, and it estimates the mass flow rate of the discrete particles falling through the pipe. For getting a clear view of how the closed-loop control works, the mass flow rate is assumed to be 5000 kg/hr. A simple relation is defined to state the relation between the rps (revolutions per second) of the motor and the amount of fish feed dispensed for a single revolution of the motor shaft connected to the auger. From the assumption, 1.39 kg of feed must be dispensed every second (See Appendix E).

This is a proportional model wherein the mass dispensed is directly proportional to the number of revolutions of the motor. The feedback gain used in closed-loop feedback control is to converge the results and to stabilize the overall system. The value of the gain can be tuned for quicker convergence of the results.

Figure 22 Simulink Model of Closed-Loop control system

(30)

The bulk solid flow sensor data is modelled as white Gaussian noise with the mean value equal to the ideal mass flow of feed dispensed per revolution of screw auger for 1 hour[15]. The sensor value mimicked is brought close to the real-time situation. The command used for generating the sensor data and the histogram plot is given in Figure 24. The difference in the desired output value and the sensor value is computed and sent through the feedback controller, which converts the deducted mass flow rate back into rps (revolutions per second). The value is then added as feedback to the motor controller and will either add or reduces the speed of the motor shaft to get the desired mass flow rate in kg/s.

Figure 24 Histogram of simulated sensor data Figure 23 Mass flow rate control setup Vertical pipe

Motor feedback controller

Data generated by bulk solid flow sensor Sensor fitted

through drilled hole

Conveying tube

(31)

4 RESULTS

4.1 Experimental results

4.1.1 Hopper selection

In the case of the fish feeder, the maximum weight of pellets to be stored is 2520 kg for 51 days.

Since the weight may vary, a safety factor with 20% excess load has been considered, and the hopper is selected for holding a maximum weight of 3000 kg. The lab-scale experiment was set up to find individual particle mass is shown in Figure 25. The results of the experiment conducted according to section 3.1.1, are consolidated in Table 3.

Table 3 Bulk mass measurements for 6.5 mm diameter spherical pellets Mass experiment

Trial Number Mass with the glass container

×10^-3(kg)

Mass of the empty container

×10^-3(kg)

Mass of individual Particle ×10^-3(kg)

Density of single particle (kg/m³)

1 36.86 20.38 0.16 1113

2 36.5 20.38 0.18 1252

3 37.6 20.38 0.19 1321

4 37.42 20.38 0.16 1113

5 37.57 20.38 0.15 1043

6 37.79 20.38 0.17 1182

7 37.54 20.38 0.18 1252

8 37.87 20.38 0.2 1391

9 36.52 20.38 0.18 1252

10 36.3 20.38 0.2 1391

11 36.19 20.38 0.17 1182

12 37.68 20.38 0.17 1182

13 37.35 20.38 0.18 1252

14 37.05 20.38 0.2 1391

15 37.76 20.38 0.17 1182

Average = 0.18 1233

Standard Deviation = 0.015 110

Figure 25 Lab-Scale Bulk mass measurement setup

(32)

The smallest commercially available hopper provided by MAFA AB, which was previously used for the aquaculture application had a minimum volume of 6.3 m³. The specifications of the selected hopper are shown in Table 4.

Table 4 Specifications of the commercially available hopper Hopper specifications

Manufacturer MAFA AB

Type UNS-6

Material Mild Steel coated with Aluzinc

Semi included angle 21⁰

Volume 6.3 m³

Height 5.33 m

Diameter 1.88 m

Weight of hopper 380 kg

The maximum weight of feed pellets that can be stored in the 6.3 m³ volume is calculated using equation (2). The maximum weight of feed pellets which can be stored in the 6.3 m³ volume hopper was found to be 4260 kg.

4.1.2 Hopper structural analysis

As discussed in section 3.1.2, the results of the parametric structural study on the hopper carried out in ANSYS® Workbench™ using two different Mild Steel and SS316. The deformation plot for Stainless Steel AISI 316 (SS316) hopper with a hydrostatic pressure exerted by a liquid having a density of 700kg/m³ is shown in Figure 26. The results of the similarly performed simulations

Figure 26 ANSYS result showing total deformation in mm along the hopper bottom

(33)

Table 5 Results of the parametric analysis on hopper deformation Material Yield strength

(Ys) of the material

(Mpa)

Density of approximated

feed pellets (kg/m³)

Approximate weight of fluid

in 6.3 m³ hopper (kg)

Maximum Equivalent

stress (Mpa)

Maximum deformation

(mm) Mild Steel

coated with Aluzinc

370 700 4410 168 2.77

1000 (water) 6300 240 3.96

Stainless Steel AISI

316

270 700 4410 173 2.73

1000 6300 247.24 3.90

4.2 Selection of conveyor system

Based on existing literature [22], two design choices were available for integration of the screw auger with the hopper bottom: 1) the conventional screw auger based on design manual (see Figure 27) and 2) the shaftless screw auger (see Figure 28). The evaluation matrix for selecting the screw auger is shown in Table 6.

Table 6 Pugh's Evaluation matrix for selection of conveyor system

Pugh's Evaluation matrix Parameters Weight (0-5)

Points

Design method 1 Design method 2

Material requirement 5 4 2

Design complexity 4 4 3

Conveying efficiency 4 2 3

Maintenance & Service requirement

3 4 3

∑(𝑤𝑒𝑖𝑔ℎ𝑡𝑎𝑔𝑒 × 𝑝𝑜𝑖𝑛𝑡𝑠) 56 43

The power required to drive the conventional screw auger is larger than the shaftless auger due to rotational inertia of components. The power and torque requirements are calculated according to equations (3) to (8). These equations are based on the screw conveyor manual [20]. A Matlab code attached in Appendix A provides the calculation for relatively larger auger, whose results are provided in Table 7. The units of the calculation are converted from customary units to SI units.

(34)

Figure 27 Conventional screw auger setup for Bulk material transport

Figure 28 Shaftless Screw auger setup for Bulk material transport

Delivery tube

(35)

Table 7 Calculated Drive requirements Calculated Screw auger parameters

Diameter of the Screw 152.4 mm

Diameter of the shaft 19 mm

Speed of the motor 1200 rpm

Length of the Auger 2400 mm

Efficiency (spur gear drive) 0.88

Total power required 0.5 Hp

Total torque required 41 Nm

The specifications of the selected motor for the driveshaft, calculated using equations (3) to (9) is presented in Table 8. The motor selection is made based on the torque demand to move the material through the conveying tube. The calculations for the torque and the horsepower required is provided in Appendix A.

Table 8 Specifications of the geared motor attached to the driving shaft Motor Specifications

Manufacturer Nord drive systems AB

Designation Standard Line Gearmotor SK 01XZ - 80SH/4

TF

Motor type 3-phase Asynchronous

Power 0.55 kW

Rated speed 1420 rpm

Supply frequency 50 Hz

Number of poles 4

Gear ratio 11.6:1

Max speed at the output shaft 123 rpm Max torque at the output shaft 43 Nm

(36)

The motor designated for conventional screw auger is used for driving the shaftless screw auger as a conservative approach. The dimensions of shaftless screw auger corresponding to design method 2 are illustrated in Figure 29. The specifications of the screw auger are provided in Table 9.

Table 9 Specifications of MAFLEX-skruv 90/75 Shaftless Screw auger Specifications

Designation Maflex-skruv 90/75

Inside diameter (ID) 41.4 mm

Outside diameter (OD) 60.45 mm

Pitch, mm 41.4 mm

Figure 30 Dimensions of the conveying tube

The product “Maflex – Skruv 90/75” provides a set of standard screw auger and the conveying tubes compatible with the Hopper's bottom and the stainless-steel tube (refer Figure 32). The cross- sectional image of the conveying tube is displayed in Figure 30, and its specifications are provided in Table 10.

Table 10 Specifications of the conveying tube for MAFLEX skruv 90/75 Conveying tube Specifications

Outside diameter, A 96.1 mm

Inside diameter, B 89.3 mm

Inside diameter, C 82.2 mm

Outside diameter, D 89.0 mm

ID OD

Pitch

Figure 29 Dimensions of Shaftless Screw auger

(37)

The flexible screw auger runs from the hopper’s bottom (driven end) to the motor shaft (driving end) inside the conveying tube which is fixed to the stainless-steel delivery tube (a standard component provided by MAFA AB). The stainless-steel delivery tube connects the motor assembly to the right with the help of fasteners, and the conveying tubes to the left with the help of pipe clamps. The location lid and flanged housing design are adapted to the stainless-steel tube (see Figure 31).

Once the hopper, auger and motor have been selected, the next step is to design the integrating subsystems like driving shaft assembly and driven shaft assembly that connects the components mentioned above to complete the feeding conveying system.

Figure 31 Cross section of delivery tube integrated with location lid and flanged housing

Stainless steel Delivery tube Location lid

Flanged housing Connection for

conveying tubes

(38)

4.3 Feed conveying system design

In this subsection, the design of integrating components between the hopper and the conveyor is carried out. The integrating components are split into two parts, one for the hopper and driven shaft integration and the other for the motor assembly. The hopper assembly consists of the hopper body and hopper bottom fastened together. The driven shaft sub-assembly is attached to the hopper via a flanged housing. The other end of the hopper is connected to the stainless-steel delivery tube via plastic pipes. The 3D CAD of the finalized fish feeding mechanism modelled in Autodesk®

Inventor Professional 2021 is shown in Figure 32.

4.3.1 Driven shaft system design

The screw auger is fitted to the hopper bottom with a flanged housing. This housing accommodates the driven shaft, ball bearing (SKF 61906) and two oil seals. The seals used in the system are located inside the bore with an interference fit with outside diameter. The driven shaft (see Figure 33 and Figure 34) is connected to the screw auger at one end and is resting against the bearing’s inner ring on the other end. The outer diameter of the sleeve provides a sliding fit with auger’s inner diameter. The auger is held tightly in place with two hook screws (or hook bolts) and nut setup passing through the hole drilled in the shaft and sleeve surface. The hook screws present in Figure 34 is just a representation to realize its function in the assembly. The ball bearing in the flanged housing supports the stepped diameter of the driven shaft and is sealed from the

Figure 32 3D-CAD model of the Fish feeding system

Driving shaft Assembly

Driven shaft Assembly

PVC- Conveying tubes Hopper/Silo

Hopper Bottom

Direction of feed dispense Delivery tube

(39)

ring. The axial locking of the shaft is taken care of by the snap ring present on the shaft groove.

When the conveyor is in action, the feed pellets get accommodated between the screw auger and the conveying tube. As a result, the pellets support the driven shaft from overhanging and keeps the bearing misalignment within acceptable limits. All the components in the system are standard parts except for the flanged housing and the cap.

Hook Screws + Nut Setup

Screw auger Shaft seals

Snap ring

Shaft Sleeve Deep Groove Ball

bearing

Housing cap

Driven shaft

Figure 34 Cross sectional view of driven shaft Sub-Assembly Figure 33 Driven shaft assembly integrated with hopper bottom

Hopper Bottom

Shaft Shoulder

Deep groove ball bearing

(40)

4.3.2 Driving shaft system design

Motor Shaft Key/Keyway

Oil seals Location Lid

Locknut

Coupling Shaft (Driven shaft)

Adapter Plate

Geared Motor

Deep Groove Ball bearing

Figure 35 Components of the Motor Sub-Assembly

Delivery tube

(41)

The motor assembly is connected to hopper bottom by a series of plastic pipes which is supported at various points from the roof using clamps. Figure 35 and Figure 36 shows the delivery tube carrying the entire weight of the motor sub-assembly. The opening at the bottom of the stainless- steel tube is integrated into the plastic T-junction of the main pipeline using pipework components.

In the motor assembly, one end of the driving shaft act as a coupling with an open keyway to transmit the motor torque. The sliding fit provided at the interface of the key/keyway prevents the dragging of the motor shaft due to auger weight and pellet load. At the same time, the other end of the coupling shaft is connected to the screw auger using a shaft sleeve with hook screws and counter locking nuts. The geared motor is fastened to the double-flanged housing with an intermediate adapter plate. The housing accommodates the deep groove ball bearing (SKF W6007), driveshaft, locknut, and a lid. Table 8 gives the specification of the motor used to drive the shaft and the screw auger. The flanged housing and lid locate the outer ring of the ball bearing, whereas the inner ring is located against the shaft step using a locknut (SKF KMFE 7 L35). This sub-Assembly inherently prevents the shaft from moving in the axial direction away from the motor. Radial seals are provided in the housing on either end for sealing the lubrication.

(42)

4.4 Detailed Design of components

4.4.1 Key/keyway

The key/keyway is used for transmitting the torque from the motor to the screw auger via the driven shaft. The dimensions of key/keyway are selected from the standards SMS 2305 [23], and SMS 2306 [24] is shown in Figure 37.

With the input parameters such as shaft diameter and torque to be transmitted, the key dimensions are calculated using according to Appendix B is presented in Table 11

Table 11 Calculated key dimensions based on SMS 2305 Key design parameters

Shaft diameter, D 20 mm

Material for key AISI 1020 CD Steel

Maximum shaft torque 60 Nm

Width of the key, B 6 mm

Depth of the key, H 6 mm

Minimum key length 12 mm

Depth of key on shaft, C1 3.5 mm

Depth of key on the hub, C2 2.8 mm

4.4.2 Bearings

There are two deep groove ball bearings present in the system, one at the hopper side and the other at the motor side. These bearings are selected based on the radial and axial loads, amount of space available for packaging of components and life of the bearing. As a conservative estimation, it is assumed that both the bearings (driven shaft and driving shaft) are assumed to carry the load of the auger and the pellets in the axial direction. In the radial direction, only the shaft weight is carried by the bearings in static conditions. During conveying operation, the feed pellets occupy the volumetric space between the auger and conveying tube. The flexibility of the shaftless auger makes the pellets present inside the conveying tube to share the radial load. Stainless steel shafts were used as shaft material [25]. The weight of the auger per meter length, 𝑊_𝐴𝑔 is 1.4 kg considering Mild Steel as material. The total length of the auger 𝐿_𝐴𝑔 passing through the conveying tube is approximately 4.15 m. The pellet loads are calculated by assuming the entire conveying

3 (𝑤 ). The volume of the Bulk Figure 37 Reference key dimensions based on SMS 2305

(43)

material and required auger length was found using features available in 3D CAD software. The mass of the heaviest shaft (𝑊_{𝑠ℎ𝑎𝑓𝑡}) (2.2 kg assuming stainless Steel as material) is taken as the radial load for both the bearings. For further clarification on determiningthe radial and axial load, see appendix D. Figure 38 depicts load acting on the motor shaft bearing in radial and axial direction.

𝑇𝑜𝑡𝑎𝑙 𝑎𝑥𝑖𝑎𝑙 𝑙𝑜𝑎𝑑 (𝑇𝐴𝐿) = (𝑤_𝐴𝑔× 𝐿_𝐴𝑔 + 𝑤_𝑝𝑐) 𝑇𝑜𝑡𝑎𝑙 𝑟𝑎𝑑𝑖𝑎𝑙 𝑙𝑜𝑎𝑑 (𝑇𝑅𝐿) = 𝑊_{𝑠ℎ𝑎𝑓𝑡}

The load parameters (displayed in Table 12) are kept constant for both bearings for bearing life computation. Assuming to be grease lubricated and temperature not exceeding 70⁰C, the bearing life was calculated using the SKF bearing calculator, which is based on the rolling bearing catalogue [26]. A factor of safety of 2 has been implemented for both radial and axial loads. Small internal clearance is assumed between two bearings races to reduce the radial play.

Table 12 Bearing life Calculation results by SKF online bearing life calculator Bearing life calculation

Total Axial Load (𝑇𝐴𝐿) 540 N

Total Radial load (𝑇𝑅𝐿) 440 N

Lubrication Grease SKF: LGFP2 (Food compatible)

Internal Clearance Normal

Equivalent dynamic load, N 990

SKF W6007 Deep groove ball bearing (Motor side) Dimensions (ID×OD×Width), mm 35×62×14

L10h, (hours) >2×10⁵

SKF W61906 Deep groove ball bearing (Hopper side) Dimensions (ID×OD×Width), mm 30×47×9

L10h, (hours) 33500

Deep groove ball bearing Axial load

(Weight of the pellets)

(Radial Load) Weight of Shaft

Keyway

Double flanged housing Screw auger

Driving shaft

Figure 38 Loads acting on the Motor shaft bearing

(44)

4.4.3 Flanged housings

The housings present in the motor end and the hopper end accommodates all the packaging components present in the system like the bearings, shafts, seals, locknuts, and snap rings. The flanged housings are designed to pack components mentioned above within the sub-assembly tightly. The images of the housings used in the system are attached in Figure 34 and Figure 36.

4.4.4 Locating lid

The function of the locating-lid is to push the outer ring of the bearing against the double-flanged housing. The inner bore of the lid acts as a housing for holding the oil seal via an interference fit.

Sliding direction Gap(𝑔₁)

Oil seal Driving Shaft

Location lid

Figure 39 Gap (g1) provided by locating lid for axial sliding

Figure 40 Gap (g2) provided between housing and snap ring for axial sliding Gap(𝑔₂) between

snap ring and housing

Hopper Bottom

Snap ring

Flanged housing

(45)

From Figure 39 and Figure 40 and, the sum of two gaps 𝑔₁ (Between the lid and outer ring of the bearing) and 𝑔₂ (Between snap ring and housing) is set to be less than 1 mm. The gap provides sliding motion between two shaft assemblies to prevent the driveshaft pulling motor shaft directly with the weight of auger and pellets.

4.5 Pipeline lift height

The following calculations based on Bernoulli's equation discussed in section 2.1.5 give a good understanding of how the theory is implemented in the system. The inner pipe diameter is constant throughout the pipeline system (refer to Table 1), which indicates that the velocity of the water is constant. The pipe calculations are done by assuming that the pipe’s inner surface is smooth, and there is no frictional head loss throughout the pipeline.

𝑃₁+1

2𝜌𝑉₁²+ 𝜌𝑔ℎ₁ = 𝑃₂ +1

2𝜌𝑉₂²+ 𝜌𝑔ℎ₂

Due to data confidentiality, the velocity of water flowing through the pipe is not disclosed in this report. As a result, when point A is raised to height h2 (= 2 meters), the level of water in the pipe equalizes with the water in the tank. The feed particles, when dispensed at this time, will directly fall into the fluid stream without stagnating on the sides of the T- junction pipe.

Applying Bernoulli's principle between 1 & 2,

𝑃₁ = 𝑃₂+ 𝜌𝑔ℎ₂

By applying the Bernoulli principle (see Figure 41) and reverse calculating the pressure 𝑃₁ at the pump end to maintain the water head of 2 meters in the culture tank is found to be 0.2 bar without considering the losses due to friction and pipe bends.

Figure 41 Schematic representation of Pipeline layout

T - Junction

ℎ₁ = 0

3 h₂= 2 m

Water from pump

Culture tank

1

2

A B

𝑃₁

𝑃₂ = 1 𝑎𝑡𝑚

𝜌 = 1000 𝑘𝑔/𝑚³ 𝑔 = 9.81 𝑚/𝑠²

(46)

4.6 Control of mass flow rate

From the feedback control loop discussed in section 3.1.5, input parameters were initialized, and graphs from the system model indicating the flow control and motor speed control are plotted in Figure 42 and Figure 43. The reference mass flow rate shown in Figure 42 was calculated to be 1.39 kg/s.

Figure 42 Change in mass dispensed per second

Figure 43 Corrected RPS to achieve targeted mass flow rate Ideal revolutions per second (15 rps)

15

Ideal mass dispensed per second (1.39 kg/s)

(47)

DESIGN OF FISH FEEDING MECHANISM FOR RECIRCULATION AQUACULTURE SYSTEM (RAS)