Group 15 – Production Design for SafaPani – Final
Contents
1. eo jo auch es a papy gir ane a amm pamiboro paoyndori oq aT 3j mge py Jo OI ORY mo pep ric peony samy caia ap CTT OY prend seen pursson o pss sy Ape gum eqs seus veria ora prag omae nope meom oem was naang ta oo TAN TE j pesi se eus sq Wto sarssooxi pur quus T vogeuny 1507 JO t BAT me unpre qesema monsako megan seen punoguma a enw Sep imperi pox pas mcrae 1q pat mo poms Meu upper Sag NO peret Suppor Sendo sais PUT tps sumi sms sepioopursamd equ au phous s vare reps sepes Sarra Sor enemy sulted experts their specialty and the major takeaways imary of the const Table RI A sums Ri Appendix S CES Analysis Yield strength elastic limit ksi Povethvene raha ET 7 22. eve i i e notre Tomer t Rod Poker roam io asper hae Pme Foam o stone eae st Q AI Pamer Foam L0 Pome Foam Oo Figure 51 CES software plotting yield strength over price for materials that are good insulators have excellent durability when submerged in freshwater have excellent durability in a rural atmosphere have fair durability in UV radiation sunlight and less than 137 Ib ft density st
3. Note You must use a pin that is analog input capable SAEI ts ede I Parameters 0 1 varlabie to keep track of how bright the Yellow Variables dealing with the charge across the electrodes float cowiceb count 0 7 Coulombe of charge that have passed through the electrodes initialize at ficonst int coulomb limit 105000 4 TODO Determine Correct coulomb limit correlated to iron in water 1 Coulomb Limit that signale that Electrolysis is over Const int coulomb limit 8 TESTING VALUE Coulomb count setting to be able to test over short timeframe Coulombs const float Rehunt 0 057 00 Shane resistor value Ohms Used to convert Current sense output voltage to current 07 Variables to keep track of interval times const long Semple time 10000 Time interval between current calculation ne const long mix time 1000000 TODO Determine correct mix time Time to wait after electrolysis ms Sonst long battery sample time 20000 Time interval between battery voltage checks ee const int led pulse tine 1000 Time interval between Red LED toggles when the Battery is low 8 const int yellow led update time 60 Time interval between Yellow LED fade updates ee Variables used to keep track of time elapsed timesinceCurrentCheck 0 Variable to keep track of ho
4. 5127 19 Appendix Figure 8 Ideal moldable design and working prototype side by side Economics After the model was researched tested and finalized quotes for each component of our design were compiled in order to determine a final Bill of Materials for our device The summarized unit cost for each component in our final design is shown in Table 2 Reaction Ves Cont of molding te reaction veel excluding mold tooling peer T ce Cont of mang hellen mel tooling espere semi Cost of ofthe shelf enke ndr 1 sino ra ricas nnd pce et Sore eerie 2x Sin pie sis lows rescate ation vei and ad sr i clearer rent vnl 2 32 ote Secure holder 1o roaction vul 1 En jana Wats Salem hole 3n reaction veo 1 sia PETG Sheet Temofomed Lid Covers reaction vessel t prevent debris C E Sess rdg vessel E Dispersion Pine Dispersos water ovr sand fer to prevent PU Sand Holders Secures nd spars nylon fabrie sand fiter soa Fase Tabi wed for seen usxns spat suene mutter Tube Provides water sd far emen LIN ES Joss Ge desse used to onset an sen sand iren s ema oul cost for alt in device m Send wed fiter ion reip s sia Ciner boc Sand tonis device vo appropriate height 2 sis CRUS EI us
5. Cacus Table 2 A detailed Bill of Materials for the molded SafaPani device Unit pricing for a production run of 1000 devices in the U S and adjusted for Calcutta is shown Spee 1 In order to obtain estimates for injection molded parts we consulted with our manufacturing contact Dana Howe The additional tooling and assembly estimates are compiled in Table 3 to generate a total cost per 1000 units of our device We consulted with our established manufacturing contact in Calcutta Jaydeep Dasgupta for costs of parts purchased in India Our sponsor who has conducted business in Calcutta has approximated the cost adjustment factor of goods and services between the U S and Calcutta as 50 70 discount Thus we will convert our prices for production determined in USD by a factor of 0 6 to estimate the total cost in India Cost of Goods Sold 545 449 67 17 000 00 32200000 060 50 669 80 Labor rate Sour 5200 Labor time hours x assembly cost 37 000 09 Reaction vessel 20 09 Shell 480 00 packaging cost 760 00 51 429 80 Table 3 A breakdown of the costs associated with molding tooling and assembly By incorporating these costs with our Bill of Materials we can estimate the total cost of producing 1000 units of our device PATHWAY TO SOLUTION The process of developing building and validating the SafaPani design involved research consultations and m
6. and price M6 Appendix N Microprocessor Code 77 copyright VitlageTech Solutions wuw villagetechsolutions org 07 author Steve the Dartmouth 89 90 team Winter 2015 77 tiiename SataPanirirmearevi ino compile with Arduino IDE raid is used to program an Arduino Nano or an ATTinyBa to run an electrolysis process for the SafaPani household arsenic water filter Te aes MOSFET transistor to turn on and off current through two iron electrodes in water It also senses the current through the electrodes By using a MAXAUBD Current sense integrated circuit and shunt resistor E Minclude lt elapsedmiiiis h gt trom playground Arduino ce Code ElapsedMillis Not a built in Arduino library Must be installed in Arduino Library nove dlapsedHillis library must be pre loaded in the Arduino i 11 gee http p ayground arduino cc Code ElapsedMi11is to dounload the library JI and http arduino cc en Guide Libraries for instructions on snatalling the library in Arduino IDE Minciude lt gt 77 Library of fonctions to disable peripherals on the ATTinyBi to limit the power usage Minelude cavr aleep l JI Library of fonctions to put the ATTiny in different Levels Gf sleep to limit the power usage p j 7 ARDUINO NANO PIN LOCATIONS Enable when using the Arduino Nano as the microcontroller 1 Digital Pin Locations const int green ded 12 JI Pin to cont
7. We are providing VTS with the following four deliverables SolidWorks design of a custom molded design housing the SafaPani process Assembly instructions for design Working prototype of design that satisfies the identified specifications Pictorial user guide showing the user how to safely operate the SafaPani We have achieved all four of these deliverables with a main focus on the moldable design which is central to moving the device towards production and have done all we can to validate the effectiveness of our product inside the timeframe of ENGS 89 90 TABLE OF CONTENTS EXECUTIVE SUMMARY ACKNOWLEDGEMENT BACKGROUND PROCESS OVERVIEW ECONOMIC AND SOCIAL CONSTRAINTS PROBLEM AND NEED STATEMENT REQUIREMENTS AND SPECIFICATIONS DELIVERABLES SOLUTION ELECTRONIC CIRCUIT TESTING WORKING PROTOTYPE Economics PATHWAY TO SOLUTION NEXT STEPS WORKS CITED APPENDICES ACKNOWLEDGEMENTS We would like to thank the Class of 1980 from Dartmouth College for their generous donation and active support of the project We would like to also thank David Sowerwine and Skip Stritter at VillageTech Solutions VTS for all of their help Furthermore we are grateful of Professor Robbie s help as advisor on the project and for all the experts who graciously lent us their time and knowledge ii BACKGROUND Naturally occurring arsenic
8. degradation in the warm wet areas it would be placed Furthermore both polyester and polyester cotton fabric had high flow rate and durability however they both ended up retaining water and required a long time to dry Comparatively nylon not only had a high flow rate and durability but it also was quick drying Therefore we chose nylon fabric over a polyester or polyester blend fabric to use in our final design User Testing In order to validate and improve upon our design we asked volunteers to carry out tasks that would mimic both daily use and necessary maintenance In our first two tests the subjects were asked to replace the electrodes and change the sand In order to measure the ease of this maintenance we timed the users with the idea that shorter time correlates to ease of replacement Appendix K In our next test the subjects were asked to use our device as intended by following a pictorial step by step guide in order to obtain feedback on the clarity of our instructions Appendix L We also determined average flow rates for processing the 15 L of water that each subject filtered which is displayed in Figure 5 After completion the subjects were surveyed on the ease of maintenance intuitiveness ease of use and helpfulness of the pictorial guide Finally the subject was asked one way in which they would improve our device to not only narrow down problem areas in our design but also highlight any unique problem we may
9. sand screen to contain the sand For the most part these pieces are independent of the overall form and were therefore optimized independently The electrode holders were designed to be simple and strong while also allowing for easy removal of the electrodes The sand screen s main design challenge lay in designing a simple solution that would let water but not sand to pass Figures 12 and 13 show the main design progressions used to reach the final elegant solutions 11 Figure 12 Progression of Electrode Holder design Angled laser cut rectangles with feet designed by VTS with 2 mum center section U shaped with separate 2 mm shim Trapezoidal clips with 2 mm nub and perpendicular stabilizers Velero straps screwed into floor of reaction vessel and electrodes spaced by 2 mm diameter O ring Figure 13 Progression of Sand Screen design Four types of tubing tested silicone A35 on hardness sc A40 rubber latex A35 and neoprene A61 Two structural support options disk with 1 5 diameter holes or X Fully assembled sand screen option with rubber latex A35 and X style support These iterations led to our final deliverable a moldable nested bucket design that conforms to the unique pressures and constraints of life in rural Nepal The progression of our work from the beginning of ENGS 89 and to the end of ENGS 90 leaves the SafaPani as a manufacturable product NEXT STEPS Dartmouth students have worked on various aspects of
10. the front view of the reservoir Appendix E Manufacturing and Assembly Instructions Figure E1 The reaction vessel can be molded in a 2 part mold Draft analysis shows the reaction vessel has appropriate draft angle for molding Figure E2 The shell can be molded in a 2 part mold Draft analysis shows the shell has appropriate draft angle for molding Figure The reservoir tank be molded in a 3 part mold analysis shows the tank has an appropriate draft angle for molding E1 Appendix F Manufacturing and Assembly Instructions E 2 0 E EH E SECTONAA Figure F1 Engineering drawing of the reaction vessel Fi 9 pak s Figure F2 Engineering drawing of the reservoir tank Reservoir Tank 5 3 F2 7 c p o NH BN 18 29430 Figure F3 Engineering drawing of the dispersion plate F3 2 o Uh ij d pw li dJ i 3 5 Figure F4 Engineering drawing of the electrode F4 Sandscreen Ring Figure FS Engineering drawing of the sandscreen ring F5 zm wi ueeios puos 88 SSS lt Hia Figure F6 E
11. to filter the water The Kanchan Arsenic Filter consists of a top bucket full of rusty nails and brick chips that is suspended over a sand filter with a piping assembly at the bottom of the bucket The filler uses the pressure in the system to pump water into an outside collection container While both designs successfully filter arsenic they both have shortcomings The Kanchan filter not only requires many parts and complex assembly but also suffers from an incredibly slow flow rate that continues to decrease as the sand filter quickly becomes clogged with precipitates Additionally both filters were often found abandoned by users after small protruding valves or piping acquired minor damage from minor accidents in the home Disregarded maintenance procedures and inefficient subsidies by both ENPHO and FFF have given no incentive for users to maintain or repair their filter nor the opportunity to purchase new ones Piersma 2013 Therefore a robust inexpensive design coupled with a well designed marketing and business model is needed to successfully implement the SafaPani technology into Nepali households at Appendix R Expert Consultations GGUSUIS USES FERT jo poems jo Nea ICI j O00 S venom mem sorsepr ip operae emu emp mung sgn epejpros wp zo eed
12. Cook Engineering Design Center Final Design Review Paper Submitted in partial fulfillment of the requirements for ENGS 90 Engineering Design Methodology and Project Initiation Production Design for SafaPani March 4 2015 Sponsored by VillageTech Solutions Project Team 15 Scott Hansen Stephen Jenkins Jamie Potter Julia Zaskorski Faculty Adviser Peter Robbie EXECUTIVE SUMMARY Nepal is facing an arsenic epidemic Millions of its citizens unknowingly drink water contaminated with arsenic at levels higher than ten times the World Health Organization WHO standard of safety As one of the poorest countries in the world Nepal lacks the resources and infrastructure necessary to treat its water supplies on a large scale In response VillageTech Solutions VTS and ENGS 89 90 teams have been developing the SafaPani filter The electrocoagulation process of the SafaPani has been shown to lower concentrations of arsenic common in Nepal to below the WHO safety standard However the SafaPani currently lacks a manufacturable model that incorporates the process into a housing designed specifically for families in developing countries Over the course of ENGS 89 90 we have focused on design for manufacturing creating an elegant device and minimizing the cost to produce and assemble the full product With guidance from plastics manufacturers DFM experts and contacts in Nepal the final SafaPani appliance has been uniquely designed
13. ET To measure charge passed to the water we will monitor the current in the circuit To check if the battery is approaching its lower limit we will need to monitor the source voltage The microcontroller uses a reference voltage to convert an analog signal to a 10 bit digital value To perform accurate voltage and current measurement we need to have a constant voltage source However the battery we will be using can vary from 13 5 11 0 volts during its operation cycle For this reason we will also need to include a voltage regulator that provides a lower constant output voltage regardless of source voltage The choice of voltage regulator also depends on operational voltage of microcontroller Summary of Inputs to the Controller Battery voltage reading Electrolytic current reading Summary of output signals from the Controller Red LED e Green LED e Yellow LED MOSFET to control electrolysis MOSFET to control stirring note stirring may not be a feature of the first designs but the electronics should be designed to accommodate stirring in future versions if needed 1 Process Flow The flowchart in figure K1 describes the method used by the microcontroller to oversee the electrolysis process The circuit periodically checks the battery voltage and shuts off the electrolysis process if the battery is too low It also checks the current and updates the total charge that has crossed the electr
14. Level Requiremei The controller circuit controls the current through the electrodes submerged in the rea and uses LEDs to communicate its internal state to the user Charge through the electrodes will be monitored by measuring the current in the circuit at equal time intervals and then performing discrete integration of current with respect to time The outputs from the LEDs are as follows n vessel Green Tum ON when mixing time is over OK to open valve to sand filter Yellow Slow blink when the electrodes are powered Fast blink during mixing time Red Blink for low battery The power source will be a 12 volt deep cycle battery When the battery approaches the minimum voltage range the Red LED will blink to signal the user The controller will go to sleep if the battery gets past this minimum voltage The controller will be turned on and off with a switch plac power wires NOTE user testing and focus group input is needed to refine the blinking and on off sequence of the LEDs to be sure that they clearly indicate to a non expert user in another culture what is happening gt gt in series with the Implementation Requirement An 8 bit microcontroller will be used as the brain of the circuit Since the electrolytic load will be variable the microcontroller will control the current with logic level MOSFET If a Stirrer is added in the future to increase mixing efficiency it will also be controlled with a second MOSF
15. Punch out of 1 8 high density polyethylene sheet 5272 sand Purchased 1 20 Sand screen Cut 1 8 high density polyethylene ring and disk Cut 1 8 rubber ring Cut nylon fabrie circle Polyurethane glue 5851 Shell Injection molded high density polyethylene 545 Reservoir tank Rotomolded 51000 density polyethylene Stand Purchased cinder blocks 5150 Figure DI Manufacturing method and cost for each component Di Figure Detailed view of our device Clockwise fom top Front view op down view cutaway view and side cutaway view D2 Figure D3 The center of mass of device without reservoir tank and stand is shown The center of mass is on the axis in the center of the shell and in the lower half of the shell within the sand D3 Figure D4 More detailed views of the shell clockwise from top right isometric cutaway view front view and close up cutaway view of the raised spokes on the bottom of the shell Figure DS Detailed views of the reaction vessel clockwise from top le isometric view top view and side isometric cutaway view Detailed side view of the lid CO Figure D6 Detailed views of the sand screen clockwise from top lefi Sand screen support disk Completed sand screen Side angled isometric cutaway view 05 Figure DT Isometric view and
16. al pieces described above have been synthesized into the circuit schematic shown in figure L2 The figure shows a high level diagram of the detailed Eagle schematic ireuit with its individual components as well as a m Figure M2 Circuit high level diagram and circult schematic showing the A Ttiny 4 microprocessor and the peripheral components used to make measurements control the current and signal the user The circuit has been prototyped on a breadboard shown to work and designed as a printed circuit board in EAGLE Figure L3 shows the board layout for the 1 5 x 1 5 circuit board The bottom holes are for the ground and 12V wires from the battery The top two holes are for the wires going to the electrodes M4 Figure M3 Board layout of the SafaPani control circuit on a 1 5 x1 5 board Holes have connections to allow wire straps to add strain relief to the wires placed near the wire M5 Full Parts List for the Custom PCB Lowry 0009 0 0001 n E meos Ll Gun sped wn ped este morie veio ae uD Ot we Table MI A complete summary of all parts used to construct the SafaPani s PCB The information includes the part quantity manufacturer part number
17. and daily use criteria for a product manufactured in Calcutta and used in Nepal First it must be cheaply manufacturable in order to be practical to distribute to the villagers of Nepal Second it must be easy to use and aesthetically pleasing to our intended users so that it is adopted quickly into the daily task of collecting water Third it must require minimal maintenance and resist wear in order to maximize device longevity Table 1 summarizes the identified requirements and associated specifications necessary to meet these design objectives Requirement Minimize cost of manufacturing 1 Cost of Goods Sold Easily Transportable 2 System weight without water or sand in Calcutta Materials are sourceable Function i Powered 4 Battery powered 12V battery z 5 Can withstand a force applied from the side or above gt 150N 1003 Filtration sand height 0 195 gt Dally Use Safe 7 Reliably filter arsenic ET i Hold a family s daily usage S Reservoir size 10 Time required to change sand lt 10 min M E 11 Time required to change electrodes lt 10 min mecs Nom 12 Time required to teach how to use cnin 13 Clearance under device to fil up water containers gt 03m 14 Long lifespan EM obese 15 Mintanence needed Monthly 11 Fits wellin Nepalese home 16 Footprint E 12 Water appears clean 17 Water is free of particulates smell and discoloratio
18. around these three objectives The SafaPani consists of three custom molded plastic containers a reaction vessel to house the electrocoagulation process an outer shell that will hold the reaction vessel and filtration sand and a reservoir tank to retain the clean water Estimated die costs for the molds of these three containers in the U S are over 50 000 However manufacturing costs after producing a mold approach material cost and in quantities of 10 000 devices or higher the cost of the mold factors minimally into the unit cost of a device Testing using a fully functioning prototype built from off the shelf components confirmed that the SafaPani design satisfies each identified requirement and associated specification Finite element analysis on a computer model of the SafaPani confirmed the structural integrity of the product and user testing has shown that it is intuitive and simple to operate Looking to the future there is still room for optimization of our ideal design but the first step to be taken is to perform arsenic testing on our prototype to ensure that it effectively filters arsenic to below the WHO standard of 10 ppb Successful filtration results will be the final hurdle in validating the function our design Building a small number of our working prototype units will allow VTS to quickly and cheaply test the design in Nepal and acquire usability data to further validate and perhaps enhance the design
19. at GW Plastics a company that specializes in the injection molding of small complex parts for the automotive and healthcare industries Mr Howe showed us on the plant floor to watch the injection molding process in real time and help give more context and understanding to the manufacturing process Reaction vessel and shell should be inj because need inner die anyway Our store bought working prototype blue chemical container was injection molded o Text indentations incorporated into mold Depressions on bottom for strength Reservoir needs to be blow molded or roto molded because tion molded no considerable savings with blow molding s fully enclosed Incorporate screw holes into mold don t drill after the fact Adding threaded neck increases blow molding cost by about 30 Blow molded bottles begin as an injection molded tube with threaded neck vs Star shaped cuts in base of shell Both accomplish increase in drainage surface area to eliminate drainage mat How to package and ship o Expensive to ship air Make sure they nest well into each other Criteria to nest outside diameter of bottom must be smaller than inside diameter of top Reaction vessel flared lip will help outer shell hold its shape Will inject off ition vessel to avoid flapper valve Flow of material will be asymmetric nter for Tolerance of wall thickness 2 5mm 0 2mm All m
20. ckly and cheaply test the design in the real world while acquiring usability data to further validate and perhaps enhance the design Ideally VTS will bring these prototypes to Nepal in the summer of 2015 In parallel with field testing DHE should begin to apply for grants or fundraise for the relatively large cost of molding and tooling for molding the three main pieces of the design and the initial production run Successful small scale tests of the working prototype in Nepal can provide VTS with the confidence to pay for the prototype molds necessary to create the fully custom SafaPani design The production of these molds permits VTS to cheaply fabricate the reaction vessel shell and reservoir and begin the process of bringing the SafaPani into the hands of the people who need them most 13 Works CITED Arsenic in drinking water seen as threat Associated Press August 30 2007 Water related Diseases WHO Accessed October 13 2014 http www who int water_sanitation_health diseases arsenico enl Nepal World Bank Accessed October 13 2014 http data worldbank org country nepal Husband Dan Baskin Jeremy Piersma Deutsch ENGS 89 Final Design Review SafaPani Arsenic free Water 2012 14 Appendix A Electrocoagulation Theory Electrocoagulation can be viewed as a four step process once a current is placed through the electrodes 1 Iron is oxidized at the anode and released into solution Fey
21. ctual battery voltage O 7 the setup routine runs ence the device is turned on or reset is pressed setup 0 delay 1000 Wait one second for transient currente to die out 17 Ensure no floating pins to help save power ATTINYEA Specific tortine ied 412 7 tee 4 OUTPUT Haigitaiirice i LOM Configure the pins of the microcontroller as input and output pinmode voltage divider pin INPUT BinMode current pin INPUT PinMode green 184 OUTPUT inode yellow ied OUTPUT PinMode red 194 OUTPUT PinMode MOSFET pin OUTPUT I Power saving settings pour ade disable 17 Disable the ADC Hill be enabled when used ipower ues disable J Disable universial serial bus Mpower timeri disable J Disable timer 1 as it is not used Serial begin 9600 sersal printini time Cure Volt Cul end setup OHHH man oop 17 The loop routine runs over and over again forever void loop 0 07 Battery Monitoring when low battery is detected pulse red LED every 2 seconds 17 Check the battery voltage every battery sample time seconds batterycheck digitalWrite MOSFET pin HIGH 00 Turn on the current to electrodes float instantaneous current check current Calculate the instantaneous current through the electrodes Amperes 17 integrate the current to update Coulomb count 14 Fix bug to correctly calculate the amount of co
22. determine max led brightness 20 seams to give a qualitative max brightness yellow led fade amount yellow led fade amount D Set the Yellow LED to new brightness analognrite yellow led yellow led brightness D end update yellow led 11 if the battery is low and red led pulae time has elapsed toggle the Red LED void update red ied 0 it iowBatt 44 timesinceLouBattLEDUpdate gt red led pulse tine 0 timesincelowBattlEDUpdate red ied pulse time Reset the timeSinceLouBattLED counter 17 Toggle the Red led Le digitainead red ied Low alkrive red 1ed HIGH 7 Write red_1ed 101l end IF end update red ied 1 7 Calculate the instantaneous current through the electrodes float check currenti _ _ 1 1 Enabie the ADC float current val snalogRead current pin 17 Read the value from the current sense ampliter power ade disable Disable the anc current val current val and vas vun bitt 0 Vss Map the binary value read to between current val current val 1000 f Convert the current val fron my to float current current wal Rehumt 60 0 17 Calculate the current using the shunt resistor and an amplifier scalar of 60 J 7 end check current 07 Check the battery level every battery sample time seconds f Set iowBatt if battery ia low and power off f battery is very lou void batte
23. e data obtained through user testing K1 4 Was the device intuitive to 2 Was the device easy to use use B 5 p 3 How helpful was the 4 How easy were the electrodes instruction guide to replace A totaal sven 5 Was the sand easy to access Figure K2 These five graphs show the ranked on a 1 5 scale responses of users to each of our five survey questions Times to Change Electrodes of Working Prototype foreach Subject gt am 1 3 4 5 s n Subject Figure K3 Times for each subject to replace electrodes plotted with a line showing the average time to change electrodes over all testing Flow Rate of Working Prototype for each Subject ex entente esce E irs 3 ose r Sus Bass EE Fm os 8 Subject Figure Flow rate for each individual test subject plotted with the average flow rate over all testing Appendix L Pictorial Instruction Guide and Steps to Perform TY Oo The steps in words are as follows Close flapper valve Pour in water Close lid Switch on Wait ven light will tum on Pull cord and hook it to wall of outer bucket Switch off and open tank valve 12 Appendix M Detailed Electronics Description and Parts List SafaPani Controller Circuit Manual High
24. e output valve Spec 13 The full device is 0 75 m tall and will sit 0 3 m off the ground with the stand bringing the full height to about 1 m which is slightly above the hip level of an average adult Minimizing the height of the device has been a driving objective as Nepalese citizens are required to fill the reaction vessel from the top with 15 L 15 kg of water Safety and reliability are paramount to the successful use of our device A household water filter has the potential for accidental consumption of arsenic and thus demands precision and care when operating the device As with any manufactured device there are associated risks and reliability issues which we have addressed in Appendix I When operated according to the pictorial user manual though our device successfully takes in contaminated water and outputs clean water that is free of debris color and odor Spec 17 Beyond meeting our design requirements and specifications the manufacturing costs are minimized due to utilizing industry standards for molding methods and using common shapes and materials for cut pieces The light weight plastic containers make the device easily transportable when empty Spec 2 and also are durable for the 5 year lifespan of the device Spec 14 Every manufacturing step we have deemed necessary is possible in Calcutta Spec 3 The device is battery powered and holds a family s daily volume of water Spec 4 The shape and supports bu
25. e used for sand secca mess sos Neoprene Tubing Provides watertight seal for sand screen ler 738 contia Give Adbesive sed construct and seal sand screen asm Total cos fr all electronic components in device sant Send ned o fer arsenic iron precipitare soins sie Toul 512719 Figure P1 Bill of materials for the working prototype P1 Appendix Q State of the Art Analysis Figure Q1 Left SONO Filter set up in a Nepalese home Right Kanchan Filter being used by a Nepalese woman There are currently many different arsenic filtration products on the market varying in filtration method and size Examples of filters currently on the market include the household sized SONO Filter ALCAN Filter READ F Filter Kanchan Arsenic Filter and the community sized SIDKO plant However after consulting with previous SafaPani team member Chad Piersma and Dartmouth Humanitarian Engineering leader Meili Eubank both of whom have conducted a field study of the heavily affected Nawalparasi district of southern Nepal two main competitors of the SafaPani device stood out the SONO Filter by Filters for Families FFF and the Kanchan Filter by the Environmental and Health Organization in Nepal ENPHO Both filters use an arsenic absorption process through the means of a composite iron material The SONO Filter is a two bucket system that uses bed of iron fillings to react with arsenic and a bed of sand
26. ese are farmers welding is not a common skill Cinderblocks and bricks are available and could prop up our d 1 Moff ground is fine to fill Nepalese containers Children aren t allowed to use the device is most cases The device is seen as too valuable to be entrusted to children s Still pour water into filter whether get it from water tower or straight from well 12 Volt batteries are not common in households Would need to buy battery solar panels and recharge system which doubles cost Some Nepalese are on the grid hence no need of battery m Grid shuts off daily at known intervals for load issues sometimes for 12 hours Color preference of Katie and Jeremy Pastel colors Light blue light orange light green Figure C1 Water collection at a community well tap with few people lining up to fill buckets containers c2 Figure C2 Our friend Nandu standing next to a Kanchan filter that is currently being used as a planter Figure C3 Unused SONO filter sitting in the comer of a home in Kunuwar ca Appendix D Detailed View of our Molded Design lis Thermoformed 0 06 polyethylene terephthalate 5285 Electrode holders Cut strips of purchased Velero Purchased plastic screws plastic nuts rubber washers 5310 Electrodes Cut from stock iron 5029 Flapper Purchased 5300 Reaction Vessel Injection mold high density polyethylene Dispersion plate
27. etric units for international manufacturing Potential limitations Food safe approvals Must validate material and once selec ed very difficult to deviate el might sink into reservoir m Might need to support with internal column pipe with holes out of plastic e matweb com o Reaction ves Shows regions that material is available Usually not effective to import HDPE for reaction vessel shell reservoir Not necessarily dimensionally stable Polystyrene or polyester for clear thermoforming FEA analysis to see if 80Ibs can be supported by reservoir Might need to increase height so that more support is directly underneath Add ridges for strength they don t add too much complexity to the mold Draft angles no less than 2 degrees Stamp dispersion plate and sand screen out of a roll of polypropylene or polyethylene G1 Class 6 material HDPE12450N DOW FDA approved Material pricing 1 31 lb o Smaller quantities for prototype 3 4 Ib Considerations for injection molding Weight of material Cycle time 1 cavity tool o 650 ton machine Quality control 0 25 operators per machine but would probably be higher for our pi Packaging cost 25 cents a piece for inner bucket 43 cents a piece for shell Ifthe shell is too heavy for the tank to support we can have a column made from a tube with holes in the sides on end placed under shell for support Think of how the de
28. gallon bucket for flapper valve and drill two small holes for nylon screws to secure velcro for electrode holders Order of assembly screw 1 rubber washer velero put through hole in bucket 1 rubber washer finish with nylon nut Next place one rubber o ring around each iron bar and then strap the pair of bars in with velcro 5 File two channels in rim of reaction vessel One for electrode wires and the other for the flapper pull cord Figure O1 Assembly instructions for the working prototype o1 Appendix P Working Prototype Bill of Materials Working Prototype Bill of Materials 15 Gallo Chemical Container 5 Gallon Horse Feed Bucket Cost of off the shelf tonk ia nin eril needed for eleocoagulaion preces Secures electrodes Allows resealable connection between reaction vessel and sand filet Secures electrode holder to reaction vessel Secures electrode holder to reaction vessel Attach wires to electrodes Seal serew hole in reaction vessel sza ons PETG steer Thermofomed Lid Covers reaction vessel to prevent debris ety oori 5285 suing Secures lid when refilling reaction vessel soa Dispersion Plate Disperses water over sand fiter to prevent channeling Sora OE ee eee ERE ETE sized Provides stable porous platform to support the sad filter sme soa 2 Bulkhead Adapter Connection between the shell and reservoir 1 sus Fabric Fabr
29. gt Fe ag 27 2 Water is separated into hydrogen gas and soluble hydroxide ions through electrolysis at the cathode 203 2 amp Hag 20H oq 3 is oxidized further while in solution With the OH being made at the cathode this reaction is kept at a reasonably neutral pH which is ideal for the formation of Fe ions 4 H200 14 Ong Fe Fe aq 4 Iron ions react with soluble hydroxide in solution to form an insoluble precipitate These flocs increase in size with time as the Fe OH 3 continues to precipitate onto existing flocs Throughout this s arsenic is also trapped and bound to these larger Fe OH flocs and therefore removed from the Fe OH 3s Fe OH 3 AsO e Rate of reaction for the oxidation of iron is directly related to Faraday s First Law of Electrolysis p edt Li where dnis the rate of moles of iron dissolved is current through the anode Fis Faraday e of the iron Experimental literature observes closely followed reaction rates b a theoretical value for 2 2 and Zis the valen We can also define a charge loading parameter seen below Ist q where q is the amount of charge through solution per volume treated Combining charge loading with Faraday s Law above gives q z d Fe This equation is useful for modeling an electrochemical batch reactor since charge loading can be seen as directly related to
30. have overlooked Appendix K Ultimately subjects thought that our device was fairly intuitive and easy to use Flow Rate of Working Prototype for each Subject UO Y 4 T 4 Figure 5 Flow rate results of the working SafaPani prototype The average flow rate over all subjects was 0 676 Limin WORKING PROTOTYPE Our prototype represents the ideal moldable design constructed to the best of our ability without utilizing custom molding Since the estimated die costs for injection and rotomolding are over 50 000 we decided that we would produce a highly representative and fully functioning prototype with off the shelf components Appendix O In this way we avoided spending the substantive cost of tooling the mold before user testing revealed any necessary changes in the design In Figure 8 below the ideal model and the working prototype are displayed side by side As one can see the general shape of the molded design is conserved in the prototype The reaction vessel still nests within the shell with the curved lip of the reaction vessel extending over the edge of the shell Additionally the shell in the working prototype contains the dispersion plate sand screen and sand found in the moldable design To replicate the topography of the bottom of the shell in the molded design marbles were used However the overall height of the prototype is taller than that of the molded design The price of the working prototype is
31. ilt into the device make it safe and stable Spec 5 The simplicity of the design aids in user friendliness and ease of maintenance Specs 10 11 12 15 The wide open top allows users to easily pour water from their collection bucket into the reaction vessel The flat side of the tank allows the device to be put against a wall maximizing storage capacity while still minimizing footprint Spec 16 Finally we have confirmed the compatibility of our full size design with the aluminum molds required for injection molding with draft analysis in SolidWorks Spec 3 see Appendix E ELECTRONIC CIRCUIT The SafaPani system req process The following list outl ss an electronic circuit to control the electrocoagulation ies the objectives of the circuit 1 Controls the current to the two iron electrodes submerged in the reaction vessel 2 Measures current through the electrodes and shuts the process off at a designated coulomb limit Doing so enables the SafaPani to introduce a precise amount of iron into the water in order to consistently filter Measures the battery voltage and shuts off the system when the battery is low 4 Signals to the user when the reaction vessel is undergoing electrolysis when it waiting for the iron to mix with the arsenic and when the process is over This lets the user know when it is time to open the valve from the reaction vessel to the sand filer Figure 6 shows the circuit schema
32. in groundwater is an issue on every continent A study in 2007 revealed over 137 million people in more than 70 countries are probably affected by arsenic poisoning of drinking water Arsenic 2007 Ingestion of high levels of arsenic leads to a disease called arsenicosis Symptoms can take years to appear and include color changes in skin painful hard patches palms and soles of feet and some forms of cancer Water related Diseases 2014 While arsenic ingestion is a serious problem worldwide it is a catastrophe in Nepal Some areas of Nepal have levels of arsenic in their groundwater over 500 parts per billion ppb or 50 times the World Health Organization s standard of safety As one of the poorest countries in the world Nepal lacks the resources and infrastructure necessary to treat its water supplies on a large scale Therefore Nepal s citizens need a widespread cheap and low maintenance solution to remove arsenic from their contaminated water Process OVERVIEW VillageTech Solutions VTS has been developing the SafaPani a low cost household arsenic filtration appliance that utilizes electrocoagulation to filter arsenic from water Appendix A During electrocoagulation two iron electrodes release iron ions into the water through electrolysis which react with hydroxide to form iron III hydroxide Arsenic then complexes with the iron hydroxide and forms a precipitate which can be filtered out through sand With this che
33. ioned drawings for manufacturing can be found in Appendices D and F The containers are designed such that the reaction vessel nests inside the shell to decrease footprint and add stability Spec 16 The inner reaction vessel sits on the built in shelf of the shell and the reservoir holds the shell firmly upright with four raised braces Each container is shown in Figure 3 Manufacturing and assembly instructions can be found in Appendix F The 15 L reaction vessel is a wide bucket with a lip that curves over the edge of the shell Spec 18 The lip of this bucket fits over the rim of the shell so if the vessel is overfilled or water is spilled the device will shed the water and not contaminate the other tanks Spee 7 The lip has the added benefit of creating a handle with which the reaction vessel can be lified out of the shell to access the sand for maintenance The reaction vessel is best manufactured through injection molding due to its wide open top and inner complexities Dana Howe an experienced engineer at the injection molding company GW Plastics reviewed our design and confirmed that injection molding is the optimal method of manufacture Appendix He recommended a minimum draft angle of 2 and a wall thickness of 2 5 mm with a tolerance of 0 2 mm which we have integrated in our design _ Figure 3 The three main containers of the SafaPani the reaction vessel shell reservoir flapper valve in the b
34. iron dosage in solution However oxide buildup on the degrade the quality of the reaction and dissolve less iron than theoretically predicted trodes over time can Ly Appendix B Previous ENGS 89 90 Testing and Optimization Final Ife Initial As 200pph 13 Wel Added oom Figure BI Final Fe and As at initial Fe 15 20 25 30 ppm amp As 200 ppb IAs at Various Initial Fe E ELI E 1 1 2 Figure B2 Final at P 1 2 4 ppm at various initial Fe Initial As 200 ppm Sand Filter Depth vs Final As Sand Filter Depth vs Final Fe ee Mm i 15 20 25 is 20 mo Sand Filter Depth em Sand Filter Depth cm Figure B3 Comparison of effluent Fe amp As at 15 20 and 25em filter depths Initial As 200 ppb and Initial Fe 20 ppm Bi Appendix C Jeremy and Katie Skype Meeting 2 21 15 The following are notes from a video conferen call on Feb 21 2015 with Jeremy Baskin and Katie Zhang two Dartmouth Humanitarian Engincers who spent the winter 2015 abroad in Nepal researching the current water situation and preparing for the potential distribution of the SafaPani system Some Nepalese villages have overhead tanks which distribute to community taps Began only 3 4 years ago but look long term Government is building 4 5 tanks per year Each tank can serve 1 2 hundred households Each ta
35. ltage return battery voltage Y 11 Battery ia too low to continue electrocoagulation 77 Turn the Red LED and wait until device 12 shut down CURRERNTLY NOT USED Should we Jump here if battery is too low to continue low battery 0 Turn off the current to electrodes write MOSFET pin LOM NS Inrite gresn led LOW Turn off Green LED should already be off igitaiwrite yellow led LOM Turn off Yellow LED 07 Turn Red LED on digitalweitelred led HIGH Wait until device is shut off Mait for shutdown and battery 177 10 Electrolysis is done Turn off current to electrodes wait mix time while blinking the Yellow LED and then turn on the Green LED and wait for device to be shut off void electrolysis process conplete 0 digitalWrite NOSPET pin LOM Reaction over Turn off the current to electrodes yellow led fade amount 2 Double the speed of the Yellow LED fading Nait and mix time Wait for iron and arsenic to mix and react Mixing over Light the Green LED and shut off the Yellow and Red LEDS digitainrite yellow led LOMI Turn off Yellow LED digitaiWrite red led 104 Turn off Red LED MigitaiWrite gresn 1ed HIGH Turn on Green LED Wait until device is shut off uate for shutdown 0707 Bnd Siectrolysis process complete 0 3000 Data recording 17 Record data values for later anal
36. mistry at its heart the full SafaPani process is summarized in Figure 1 Previous ENGS 89 90 groups have validated electrocoagulation as an effective means of arsenic filtration Appendix B however their system lacked a manufacturable housing that is cheap simple and durable enough to succeed in a rural Nepalese home a D Figure 1 A flow chart of the SafuPani process transforming contaminated water into clean water ECONOMIC AND SOCIAL CONSTRAINTS The median income in Nepal is around 700 year and is even lower in our targeted region Nawalparasi Nepal 2014 Additionally our end users are very self sufficient using their limited resources to purchase essentials such as kerosene fertilizer and batteries Therefore minimizing the production cost of our product is a pivotal factor in generating demand Furthermore our target users live in very rural environments and have limited opportunities to obtain new parts for repairs or routine maintenance Appendix C As a result our product is durable and reliable PROBLEM AND NEED STATEMENT VTS has proved that their unique electrocoagulation method is both simple and effective at reducing arsenic concentrations to safe levels However VTS needs an inexpensive user friendly and manufacturable design to house its arsenic filtration method that will succeed in Nepal REQUIREMENTS AND SPECIFICATIONS The design of SafaPani must meet the required manufacturing function
37. n 13 Fits well into how Nepali people 1q pegction vessel size m gather water Table 1 A breakdown of the requirements and associated specifications our design must meet to satisfy the project objectives DELIVERABLES We are providing VTS with the following four deliverables SolidWorks files of a custom molded housing for the SafaPani process Assembly instructions for the design Working prototype of the design that satisfies the identified specifications Pictorial user guide showing the user how to safely operate the SafaPani Delivering these four items will enable VTS to begin the process of bringing electrocoagulation to Nepalese citizens as a means to filter their arsenic contaminated water We have achieved all four of these deliverables with a main focus on the moldable design which is central to moving the device towards production SOLUTION Throughout the course of the project we have worked with manufacturing design and molding experts to create a device capable of meeting all of our requirements and specifications The final product is shown in Figure 2 Figure 2 Assembled device and exploded diagram detailing each component of the SafaPani design Our design consists of three molded plastic containers a reaction vessel that houses the electrocoagulation process an outer shell that holds the reaction vessel and the filtration sand and a 65 L reservoir tank that holds clean water Alternative views and dimens
38. ngineering drawing of the sand sereen 2 1 cx B a 1 88 E ES Ba d ia j H 4 BUE ts Figure F7 Engineering drawing of the sandscreen support disk F7 D Figure FS Engineering drawing of the shell F8 Manufacturing and Assembly Instructions Injection mold reaction vessel and shell and rotomold reservoir out of high density polyethylene Pot electronics in epoxy Screw electronics board into wall of reaction vessel with nylon screws rubber washers and nylon muts Insert nylon screw through rubber washer velcro strip and inside of reaction vessel for each of el two holes ire with rubber washer and nylon nut on outside of reaction ves Install flapper valve in depression in bottom of reaction vessel Connect wires to electrodes us Place o rings 2mm cross s Secure electrodes in velcro straps in upright position Place reaction vessel inside shell Place lid on top of reaction vessel Attach valve to reservoir tank spout Place tank on cinder blocks Nest shell into place on top of reservoir tank t 8 32 steel screws and nuts tional diameter around individual electrodes F9 Appendix G Dana Howe Meeting Notes 2 13 15 The following are notes from an onsite meeting on February 13 2015 with Dana Howe at GW Plastics located at 239 Pleasant Street Bethel Vermont 05032 Mr Howe is a Systems manager and Cost Engineer
39. odes This value is proportional to the amount of iron ions released into the water and is used as a measure to end the electrolysis process After this charge limit is re amount of time to let the iron ions react with the arsenic lt lt NOTE Testing is needed to determine the amount of charge and the mixing time necessary to react with common arsenic levels seen in Nepal gt gt Een Tum on curant ed the controller simply waits for a specified Figure 1 SafaPani microprocessor flowchart showing the steps used to control the electrolysis process Circuit Description Microcontroller The controller will need to control three LEDs take current and voltage measurements and level MOSFET We had initially thought of using ATtiny 85 for the project because of its tiny form factor and low cost However the ATtiny85 has only 6 usable 0 pins and we will need at least 4 output pins and 3 input pins For development and testing purpose we will also need to have enough memory on chip to collect field data for later analysis Therefore it would be best to use a development platform like Arduino Nano for development platform and use a bigger microcontroller like ATtiny84 for the wide scale production device control current to the electrolysis process via a logi Here are the features of Arduino Nano and ATtiny84 that will be most important to our project Arduino Nano Has inbuilt voltage regula
40. of 100N applied from the side to the top of the device would be required to push over the shell Figure 4 FEA analysis of the reservoir tank Left Maximum stress of 12 50 MPa Right Maximum deflection of 1 501 em as a result of the applied load of a full reaction vessel and shell Sand Screen Fabric One of the main design challenges lay in creating a method to contain the sand while allowing water to pass and be strong enough to withstand the weight of a 20 cm high column of sand With these two design requirements in mind we examined the sand permeability and flow rate performance of cotton polyester polyester cotton nylon agricultural Weed Block fabric and mesh screening In order to test a fabric s ability to contain sand a pouch was constructed out of each fabric and a constant amount of sand was added Water was passed through each pouch of sand and collected to determine if any sand escaped In order to decide which had the optimal flow rate 1 L of water was timed while it was poured through each sand pouch Appendix J The agricultural fabric was the strongest but had the slowest flow rate The Weed Block fabric had the fastest flow rate however it was prone to inelastic deformation under stress so it was also rejected Therefore the only materials that satisfied our two design requirements were cotton nylon polyester and polyester cotton fabric However cotton is a naturally grown fiber and is subsequently prone to
41. olonbe that have flown across the coulomb count instantaneous current fleat sample time 1000 TESTING 7 enough charge has passed stop the electrolysis process and start waiting aix tine for mixing if coulomb count gt coulomb Limit electrolysis process complete Wait mix time and then signal process is complete 0 N3 0 7000 Record the time current and cumulative charge record data current Wait sample time seconde from the the previous time through the loop to save battery 11 Blink the yellow and Red LED appropriate Wait and blinkt sample time J 17 end loop HMH Banton 1 1t t LB Fade Yellow LED according to yellow led fade amount to signal that the process is active 77 Fade faster during mixing as the process almost over void update yellow ledt 7 enough time has passed Update the Yellow LEDS fade amount if itimesincevellowiEDUpdete gt yellow led update time Reset Yellow LED Update tine VimesincevellowhED pdate yellow led update timer Update LED brightness Yellow led brightness yellow led fade amount 7 Fix any out of bounds problems Sf yellow ied brightness gt 20 yellow 184 brightness 20 else if yellow ied brightness lt 0 yellow led brightness 0 17 Reverse the direction of the fading at the edges of fade voltages if yellow led brightness 0 yellow led brightness 20 TODO
42. onverter has only 10 bits and the voltage resistor the current resolution would be 50 mA unit which is insufficient to perform proper charge calculation For this reason we decided to use a MAX4080 current sense amplifier The SAUS MAX4080 chip measures voltage drops across a shunt resistor 0 05 Ohms and magnifies that value by a gain of 60 More information about the current sense amplifier can be found from the link datasheets maximintegrated com en ds MAX4080 MA X408 pd Voltage Measuring Voltage Measurement is done using a simple voltage divider with two resistors in parallel to the circuit and therefore power consumption will be of not a great concern The resistors 22 Ik and 10k drop the maximum battery voltage of 12V down to 3 74V a value that can be read by the ATtinyS4 s ADC Voltage Supply to Microcontroller The circuit is powered with a 12 Volt deep cycle battery However the ATtiny84 microprocessor requires a 5 Volt power supply Therefore a 7805 Voltage regulator is used to supply 5 Volts to the microcontroller In System Programmer In order to manufacture thousands of these circuits a simple and robust method of programming the ATtiny84 is required A dual inline 6 pin header is soldered to the PCB to allow an AVRISP MKii device to quickly program the ATtiny84 and the circuit is designed in such a way that in system programming is possible Circuit Schematic and Board Layout The individu
43. ost importantly design iterations We began by researching the problem of arsenic contamination in Nepal in depth including the state of the art already designed to solve it The shortcomings of the Sono Filter and Kanchan Filter two popular competitors were readily apparent Appendices C and Q For example both have brittle and weak features which protrude from the device and could shear off Previous ENGS 89 90 teams and Dartmouth Humanitarian Engineering DHE have done a lot of work in solving these issues The most recent iteration of the Dartmouth design used two large nested buckets Considering these designs it was apparent that plastic molding would be large part of the project Asa result part of our research involved investigating plastic manufacturing processes specifically injection molding blow molding and rotational molding Each process has its own advantages general practices and common materials Consulting specialists such as Dana Howe and Professor Ulrike Wegst gave us expert opinions and guidance on these topics Appendix R Using the Cambridge Engineering Selector CES software we were able to evaluate over 100 materials based on their relevant properties Appendix S Our requirements whittled the list to just 19 and helped us select HDPE as the cheapest options for many of our parts In parallel with researching manufacturing processes we followed a regimented design strategy to develop the form and function of each
44. ottom of the reaction vessel is used to drain the water into the sand filter after the electrocoagulation process is complete It is recessed in the bottom of the bucket to ensure all of the water drains The user must pull the nylon cord attached to the flapper to release the water into the sand filter once the reaction is complete The electrodes are secured to the bottom of the reaction vessel with strips of Velcro screwed into the bottom of the bucket with 8 32 nylon screws and nuts with rubber washers The Velcro securely holds the iron bars vertically while still providing access to remove and replace them so that the hydrogen bubbles produced between the electrodes can escape upwards Spec 11 O rings with a cross sectional diameter of 2 5 mm are placed around the electrodes to keep them properly spaced apart even with degradation of the iron over time The electronics are potted in epoxy and screwed to the inside wall of the reaction vessel with 8 32 nylon screws nuts and rubber washers The electronics are placed near the rim of the reaction vessel so the LEDs can be easily viewed through the clear lid to notify the user Two wires connect the potted electronics to the electrodes and two wires lead from the electronics to the battery external to the device The shell is a tall bucket in which the reaction vessel nests The bottom section is a smaller diameter than the top to create a ledge on which the reaction vessel can rest The sand is s
45. p serves 4 5 houses Registration fee to use Allowed to collect unlimited amount twice a day Water from overhead tanks is not contaminated 50 ppb arsenic m Currently being tested 40 000 60 000 upfront costs supplied by 80 government and NGOs international donors m 1956 villagers in terms of unskilled labor 190 cash from users Lifespan relatively long term Wealthier have spigot nearer to their homes live closer to highways e Kanchan Filter 75 100 for household filter with green pretty tiles for wealthy people 5800 rupees for cheap kanchan filters 4 subsidized price from NGO entrepreneurs sell them to the NGOS Looks nice easy to clean Large square shape fit well in home They like the concrete Looks more durable Put in corner next to counter 1f by 18 footprint of a microwave Wealthier paid full price for Kanchan filter 876 100 unsubsidized m Green tile added to bas m Place in comers of kitchen 1ft x Ift footprint x 4ft tall Proud of their filters c filter to make it look ni New Trunz filter on the market Sell 20L bottled water like water coolers Pay to refill Easy to buy clean water Expensive to buy 60 000 Use filtered water for drinking and cooking but not bathing Daily water usage for family household 5 people is 20 40L c1 Bath at tube well straight from well not an issue Most nepal
46. piece of the SafaPani Figure 9 summarizes the steps involved in our process The arrow from testing back to brainstorming highlights the importance of the many iterations we have made to arrive at our final solution 10 Define the Break designinto Draw up each feasible solution on ts own 8 5 Rank solutions Figure 9 The high level design method we used to develop each component of the SafaPani Each piece of the SafaPani underwent a design progression improving with input from our sponsor expert consults and testing results Figures 10 and 11 show the many overall SafaPani forms and corresponding prototypes used to test specific pieces of them The volume footprint and material of the design changed immensely throughout the project as we became more knowledgeable about the culture and needs of our user For example reducing the overall batch size and reservoir volume to 15 L and 65 L respectively better fits into the average Nepalese household daily use requirement Figure 10 Progression of SolidWorks modeling 2012 ENGS 89 90 final model image from Stefan Deutsch our PDR SolidWorks model our CDR SolidWorks model and our final SolidWorks model Figure 11 Progression of physical prototypes DHE 2013 prototype VTS 2014 summer prototype our PDR prototype and our final working prototype model Each form shown above also requires a holder for the electrodes and a water permeable
47. rol the Gr procesa complete battery is low sonat int MOSFET pin 2 J Pin to control the MOSFET to switch on or off current through the electrodes to signal when 00 Pin to control the Red LED to signal when the 07 Analog Pin locations conse int yellow ted 82 00 Pin to control the Yellow LED to signal when process is running 10 1 changed pin to analog output for fading Choose an analog pin for the Yellow LED const int voltage divider pin AT 10 Pin to read the battery voltage with a voltage const iat current pin ESI Pin to read the current through the electrodes with a current sense amplifier pA E iui p ibis JI ATIINYBG PIN LOCATIONS Enable when using the ATTiny8d as the microcontroller 77 Digital Pin Locations const int green ded 02 00 Pin to control the Green LED to signal when Process ie complete Sonst int red ded 17 Pin to control the Red LED to signal when the battery is lox N1 const int MOSFET pin 92 Pin to control the MOSFET to switch on or off current through the electrodes Analog Pin locations const int yellow 1ed 5 Pin to control the Yellow LED to signal when process ie running Sonst int voltage divider pin 5 Pin to read the battery voltage with a voltage wove You must use pin that is analog input const int current pin n Pin to read the current through the electrodes with a current sense amplifier
48. rycheek J it timeSinceBattCheck gt battery sample tine 1 Check battery every battery sample time timesincenattcheck battery sample time Reset the time since the battery was battery voltage 0 5 double battery voltage 0 5 double readvee Update Vee reading with smoothing Le battery voltage lt low battery limit 1 threshold set at low battery Limit for low voltage indication 0 lowBatt trues Set threshold while battery voltage lt power off battery limit threshold set at power off battery limit to power off if battery too low TODO Should we just shut the whole iie jump te low battery Battery voltage readvec D eise false 7 TE the battery voltage is above low battery limit set lowbatt as false 7 D 1 7 batteryCheck 07 Caleulate the battery voltage by scaling the value read from the voltage divider double readicel 0 ado enabie 11 Enable the ADC int sensor val analogRead voltage divider pin 11 Read the voltage divider andiog pin ade disable 11 Disable the double Voltage divider voltage map sensor val 0 bitt 0 VSS Calculate voltage Gv at the voltage divider Voltage divider voltage voltage divider voltage 1000 0 11 convert from double battery voltage voltage divider voltage voltage divider ratio Convert voltage divider voltage to battery vo
49. the SafaPani project since 2009 Our contribution to the system has been to create a manufacturable design to house the validated electrocoagulation process We have succeeded in this goal and validated the effectiveness of our product during ENGS 89 90 However there is always room for optimization Possible further investigations include 1 Recessing the valve in the reservoir tank to shield from shearing 2 Manipulating the form of the top of the reservoir tank to inhibit contaminants from entering the tank such as with a ridge or trough 3 Manufacture the printed circuit board with holes for screws to eliminate an assembly step 4 To improve the seal on the sand screen an oversized ring of rubber sheeting could be inserted between the two plastic disks such that the overhanging rubber edge deforms against the walls of the shell to create a seal 5 Completely minimize the cost of the device The next step that needs to be taken to move towards production is to perform arsenic testing on our prototype to ensure that it effectively filters arsenic to below the WHO standard of 12 10 ppb DHE is planning to conduct such a test in the coming spring of 2015 Successful arsenic filtration results will fully validate the functionality our design Another step is to perform field testing in Nepal and to learn more about how the SafaPani will fit into a villager s home Building a small number of our working prototype units will allow VTS to qui
50. through and sand is saturated Pour 1 L of water into bucket and begin timer at first contact of stream with sand Stop timer when drips of water are 5 seconds apart 22226 Cotton Polyester Polvester Cotton Nylon Aericultur No No No No No No Ye Yes Yes Ye Yes Yes 0 0400 0 0250 0 0263 00313 00200 0 0357 and through through Table J1 This table summarizes the results ofthe fabric testing Because of its fast flow rate and durability nylon was chosen as the optimal fabri J1 Appendix K User Testing Results and Questionnaire User Questionnaire Subject name Date and time of test How easy were the electrodes easy to replace Circle one 1 not at all 5 very 2345 Was the sand easy to access 2345 Was the device intuitive to use 2345 Was the device easy to use 2345 5 How helpful was the instructional guide 12345 Name one way in which you would improve this device For Group 15 only Time to change electrodes Overall running time for device T aux Figure KI This figure shows the questionnaire given To test subjects upon completion of testing The questions seek to determine both how intuitive our device is to use and how to improve on the design a ion veel pr touc Le prat dose Jelosiandvina cm Inyo of ket perenni ne ellis tol and con y mcer re Table K1 This table summarizes th
51. tic and board layout of the printed circuit board as well as an overview of the microcontroller process used to accomplish the four tasks listed above The board is encased in a clear epoxy that waterproofs the circuit and allows the user to see the signal LEDs on the board A more detailed description of the circuit design including a user manual and the microprocessor code can be found in Appendices M and N respectively t EEE SESE Figure 6 Microprocessor Howchari circuit schematic and printed circuit board layout TESTING To validate that the SafaPani design described above satisfies each required specification we performed qualitative and quantitative tests The following section describes each test and its results Finite pod Analysis A finite element analysis was conducted in SolidWorks to validate the structural integrity of the design The results can be found in Figure 4 The maximum weight of the full device with 15 L of water in the reaction vessel plus 20 cm of wet sand in the shell weighs about 45 kg and creates a load of 506 N on the reservoir tank Under these loading conditions the reservoir will deform 1 5 cm downwards in the center of the top but will not fail The maximum stress endured by the tank is about 12 5 MPa With a yield strength of HDPE is 26 MPa our design has a factor of safety of 2 08 With the weight and shape of the device a force
52. tor which will make early development process easier Input Voltage limits 7 12 volts Operation logic level voltage 5 Analog Input pins 8 Digital lO pins 14 M2 Analog to Digital Converter 10 bits KB of flash memory useful for logging data 32 USB port of power data stream Note voltage from USB might be unreliable ATtinye4 e 5150 BKBofnon volatile FLASH memory for code 5128 EEPROM for non volatile data 512 B SRAM for run time data Programmable in Arduino Environment Input Voltage 27 5 Analog Input pins Digital VO pins ADC channels 8 at 10 bits Timers Has sleep functionality with significantly reduced power consumption Analog to Digital converter 10 bits ATtiny85 e Sxxx 8 KB of non volatile FLASH memory for code 512B EEPROM for non volatile data e 512 B SRAM for run time data Programmable in Arduino Environment Input Voltage 1 8 5 5V Analog Input pins Digital pins 6 channels 4 e Timers 2 Has sleep functionality with significantly reduced power consumption Analog to Digital converter 4 at 10 bits of flash memory for program Current Measuring Initially we had planned to measure the current by measuring the voltage drop across a known resistor value However this wastes a significant amount of energy through heat For example the AD solution would be about Smv unit If we use a 0 1 ohm c
53. tored in the base of this container beneath the reaction vessel so that the sand is easily accessible to the user by lifting the reaction vessel out of the shell Spec 10 With the height of sand constrained to minimum 20 cm minimizing the diameter of the sand decreases the overall weight Mr Howe recommended that the shell be manufactured through injection molding as well with 2 5 mm thick walls and a tolerance of 0 2 mm It has the minimum draft angle of 2 to make the most stable shape that is still injection moldable The shell is 42 cm tall and with an upper diameter of 17 5 cm and smaller bottom diameter of 15 cm with the decrease in diameter occurring 18 cm from the top A slope is molded into the bottom of the shell to allow the water to drain out casily A raised ring along the inside of the bottom of the shell with spokes pointing to the drain will provide support for the sand screen while also increasing surface area for the water to drain out of the sand The sand screen rests on the bottom of the inside of the shell and retains the sand while allowing the filtered water to pass through The sand screen fabric is 100 nylon with a rip stop weave The fabric is held between two disks of high density polyethylene HDPE and fixed with urethane based glue One piece of the HDPE is shaped as an open annulus to allow the least impedance to water flow while the second piece is a disk with many 3 cm diameter holes to provide structural s
54. upport for the sand A low cost FDA approved rubber is adhered with glue around the disk to form a tight seal to the inner walls of the shell The 20 cm of sand rests on top of the sand screen Spec 6 This 20 cm height was determined optimal for filtration by the previous ENGS 89 90 team Appendix B On top of the sand rests a dispersion plate manufactured by being stamped out of a sheet of HDPE A transparent domed lid fits over the top of the reaction vessel and snaps into place There are grooves in the lip of the reaction vessel to provide a channel for the wires and flapper valve pull cord to exit the device without affecting the ability of the lid to close securely The dome shape of the lid discourages users from placing objects on top of the device and also sheds water The lid will be thermoformed out of clear polyethylene terephthalate The shell rests on the rotomolded reservoir tank Our tank design is similar to an existing rotomolded tank found in Calcutta Appendix H albeit with a 65 L capacity and a semi cylindrical shape to fit against a wall Spec 8 It is 30 cm tall with a radius of 30 cm and has a wall thickness of 3 mm It has horizontal ribbing for structural support as well as four angled supports to brace the shell A valve at the bottom of the reservoir allows the user to dispense the water when needed The reservoir tank is raised on a cinderblock stand to allow for adequate spacing to fill a large cooking pot underneath th
55. vice might not be set level in the home so our design must accommoda that through the angles of our design Adding a spout to the design is cheap in injection molding Put as much as possible of the design manufacturing into the mold The changing the overall shape of the design can add more support than changing the thickness timates per 1000 units 2144 73 material 493 00 machine 52 17 quality control 217 00 packaging 46 58 die maintenance 3144 63 or about S3 per piece excluding mold 17 000 for mold Shell estimates with similar break down 5152 38 or about S5 per piec 522 24 000 for mold G2 Appendix H Rotomolded Tank from Calcutta Figure H1 Example rotomolded water tank Calcutta Image courtesy of Jaydeep Dasgupta Appendix 1 Failure Modes and Effects Analysis and Safety Risk Assessment SRA Table I1 FEMA analysis results genio alos aj i pe gusanos puso oq ssec Table 12 SRA analysis results Appendix J Fabric Testing Protocol and Results Mesh Fabric Testing Protocol Procedure 1 14 diameter circle out of each fabric to be tested Place one piece of fabric into 1 5 gallon bucket should bunch and look like a coffee filter Pour 750g of Home Depot play sand sand into the bucket with filter Pour 1 L of water through the sand this step is to saturate the sand with water Allow water to run through until all water has run
56. w long since the Glapsedwiliis timesinceYellowiEDUpdate 0 Variable to keep track of how long since Yellow TED fade amount has been updated elapsedmi11is timesinceBatiCheck 0 Variable to keep track of how long since the Battery has been checked elapsedimiliis tinesinceLowBal has been updated zEDUpdate 0 Variable to keep track of how long since Red LED Variables to convert ADC values into actual voltages const int biet 1026 J of bits output from the ADC const int vss 46007 di Max voltage used to scale the ADC output to a voltage v 01 Check if this value is correct I found this value in the code sent to the Dartmouth 39 90 Winter 2015 team Battery Voltage Parameters ficat Battery voltage 12 Declare a variable to hold the current battery Voltage Volta initialize to 12 to make smoothing work const float low battery limit 11 5 TODO Determine low battery Limit Low battery Voltage limit voita const float power off battery limit 11 TODO Determine absolute lowest battery Limit Power off battery voltage limit Volta boolean lounatt false Initialize boolean to signal when the battery is getting iow below low battery limit Sonst float voltage divider ratio 3 21 Rl 10kOhme R2 22 1kDhme Battery Ratio DLvider Ri RI R2 Divider gt Battery Ratio RIsR2 RI 3 21 7 This value ia used to scale the voltage read from the voltage divider to the a
57. ysis old record datat int current 0 float time time tine 1000 0 Sersal peine time J Serial print current Serial printit Iz Serial print coui mb counti Seriai_print 1 07 end record data0 Wait time to wait seconds while still blinking the yellow LED Void wait and biink long time te wait Wait for time to wait to pass while Fading the yellow LED while timesinceCurrenttheck lt time to wait Fade the Yellow LED if enough tine has passed update yellow led J if the battery is low and red led pulae time has elapsed toggle the Red LED update red led 11 Reset timesincecurrentcheck timesincecurrentcheck 0 P77 end wait and biink Simply loop until the device is shut down void wait for shutdown Turn on the appropriate LEDs Wigitaiwrite yeilow led LOM Turn off Yellow LED N6 digitaiWrite green led 10W Turn off Green LED pe zed l d 16M Turn on Red LED power ali dissble JI Disable a11 peripherals TODO Go to sleep here We atill want the Red LED lit up hile J JI wait for a full reset Y end wait for shutdown N7 Appendix 0 Working Prototype Assembly 1 Empty 14 gallon taper sided chemical container 2 Fill bottom with 1 of marbles to fill in depressions and to increase surface arca for water flow 3 Add sandscreen and 8 inches of sand 4 Cut 2 5 diameter hole in 4
Download Pdf Manuals
Related Search