Home

Interactive Cooling System

Contents

1. B a CHR OR GER BA 199 Page COPPER 07 INE i Ss _ _ Signy m do 2 t WSs ONI SIN31SAS 2 NH 7 E 514 8104 9345 L _ 250 00 520 T aiva Ag ONNS i ons MSN 4 4 a mo SA RO 140 12 10 CEP REM Com Wem Gem mag arim paS EEE 00 520 HET avd 491 NOISIA EN LA UGS ANJUTA SSI AS UGA Ia ODN 80 EYEL LAA A 22 VIA cee ee eee a eee 8 141 Page GOPPER 11
2. 89 7 89 eR ee 90 RT J M 90 Pump HTC EEUU TERT 90 1 8 90 COUGH i i rH TTE 91 misil epi a a 91 Boom rc 91 Malbtetialite 55525568 91 MEE 72 a a a NN 72 HR 72 CC 93 PTT a n 95 n 97 Appendix Custom OMG AUS 59 5 98 Appendix B RS LogiX 500 Programming RM IUE 108 Appendix C 126 Appendix D MATLAB Code for Plant Simulink 128 Appendix E ICS Finances and Bill of Materials 6 129 Appendix F Copper Tubing Components ots 130 146 Appendix H Connection Diagram for the eina 153 Appendix l Users t dee se 154 Appendix J Advanced Calculation Equations os ent eie Bes 177 Acknowledgements Spe
3. 27 Control gt DERE ED TU mmt 24 801811 Mode UNG NM RETE TENET 27 AE EEE 30 Data Acquisition and Signal Processing 30 Touchscreen and User 31 Secuon 51111 55 NM 33 e E 34 Display Cabinet Construction 34 Fiberglass Shell Mold MAKING 34 Frame ONS NEU 38 Fiberglass Shell Construction 40 Customized N Eae 42 43 Caster 45 Two Phase Cooling Construction 47 Ice M HH 47 Cas E TTE 4 8 Cotd Flate Housing 4 8 T E 52 Shroud COWS 53 Fanand shroud instalation RR m 54 m a 55 OWN 57 58 E 64 Logo FF RU UE 64 Graphic Art 65 Under CS TREO 65 66 Section I
4. 60914289905 08009 11 4 EVEN 11100761607 40 SVE 2 14 TIV CX Y LLOC zu CAL 4 AVAL ANV 8 80 01 0111480 LOM IU SSTWIMcO Li NOG BE OOV 162204 218404 112 KO 1 KLENNVH Di 24 SIVE NOQ SINI 802 LEE 0 2 COPPER T3 ILI VII ILL gt 144 Page KOH Duypo posa mG 50 SALSAS 3112 1 NOL BOO NLAINNYH HIM G 301511 87 IM iV 1954 aon gosto 102 122 Cri 1 RR ERASMI 1 WNOUYOCC3 XSSOO UT 16 COPPER 14 p y W S798 145 Page STRAINER SUB ASSEMBLY APPENDIX G SUB ASSEMBLIES G 1 pepojdx3 Q3 18WN3SSV 5 Q3Z2VH8 333
5. RING CERTUM 11 0 OL 02 040 FRDWAL HOC 3 100 4 CONDENSER PRINT ECL AL Sith OIW OL cagas 38 15303 77 69 101 Page A 5 CUSTOM PUMP BASE ISO VIEW ERIT M I PASSIVATE RIFTION ENDPLATE DLS APPLY AFTER rum Eu i gt 2 2 2 lt 2 5 LASER MARK CUSTOMER LOGO af Ful b 4 A LASER MARK ADS 00218 P M O R A L 4 1 PATET 4 LASER MARK S N f Page 102 6 1 02 Buuds ulisag de gt HIS 757 wWwON wks Nw Jee IPAS NF AB CUI seta amena 5 Se amas wo XII 0002 4354 2997 0 1 31925 V V NOILO3S 94820 12000044 2 s3001nv A8
6. 4 8c GIES 15149134 CH Cow 1968 U BEAD 8 EVE ACVB BIOM 281 OL TOE 5781 42 o o 103 A 7 PUMP SEAL A8 755 Tr oie 30 IS O3dSHSINIJ 620 0001 4358 S3903dHWHS Vv Noudnsos3d AMOLSIH NOISIA3M 0300009514 1 28 THLE lt LIVEE OAIN LOO LADADO Ga 104 8 PUMP MANIFOLD YSBIWON 154 1008 434 3125 3726 89 0 238 NOILO3S 091 05 ULT gx 22 o xc 1 TSS 88 000 871 000 100 3 117 00 13 891 eic 88 5 XY 054 1564 292 0 XC 100 3117 LIII LJ perm weeds TT LA gi 2 E d 207 T 105 Page A 9 PUMP MANIFOLD COVER PLATE
7. SF Cm dh Ras od a ER 132 Page 4 04 NY 7 301511 91 j O3dSHSINH v 69199955 520 VAT mT 1900 1 mi MORI WRT 133 F 5 05 i 3190 AG 81 535 HSINIJ Wits 19 9 20909 00 20 134 6 05 2 peg ede T ONMO A HOLIS SIH NO 0 WNOLUWOCO3 201 1 un xx mE Fn Di 123 mariri el CoS 1243 80 21 511 VE vee D9EYIS02 EEO XI Las Ls as IL 135 Page 7 05 3
8. UPC mt 5 YNWON MN puse Heu DANH auon 2999002 00 520 AND E Ln 1 D mil 1 LEER Tt MO do EV CL dsl LETS lis aai 1 Ar EYEL 136 8 06 PRL AVIV ING AOS angu pu g 4 28825 S MAIVA 40 1511 z 8314 2335 S3SWnNiNvd gt ao 0 co he wr m _ _ 1744 JHN _ 1381 Ne 030a 137 F 9 COPPER 07 WNOLLYDN0S WSSOOJITE NY es ___ 138 10 08 I 11 cm wsn TTL ESSWTN 1593 0 560 bocce ______ 1 49 ee 4
9. 50 100 150 200 250 300 350 400 Time seconds Figure 72 500 W at Steady State Page 68 EXTREMES TEST The objective of this test is to find out how hot or cold the cold plate can get at extreme operating conditions without the fan being on and then with the fan at full speed This gives an idea of the lowest temperature we can cool the electronics to and the highest temperature the electronics can get up to in our system Note An electronic safety shut off was implemented in the code so that any measured temperature does not go above 90 PROCEDURE Supply power to the ICS Select heat load to be 100 W Set pump speed to 14 Set fan to 100 Start data logging Wait 10 minutes or till steady state whichever is first Stop data logging Turn off power supply to 5 p eve RESULTS OF A The test results showed that the lowest temperature that electronics can be cooled to in our system is approximately 31 C This is with the lowest heat load and the highest fan speed See Figure 73 for results C C C 2 0 gt 0 200 300 400 Time seconds Figure 73 Extremes Test 100W Page 69 PROCEDURE B Supply power to the ICS Select heat load to be 1000 W Set pump speed to 64 Set fan to 0 Start data logging Wait 10 minutes or till steady state or until maximum allowed temperat
10. Son 1 411 11551501563 621 412 1130813125 42 4 55 412 134 1541 5570688 as 5 28 1146 0782811 24 414 40 135 1190 0270562 44 821 234 1 44 53 4 ms 413 ms 214 1158 709373 1 36 417 4 4 116 37 41 1157 540625 357 41 1171288125 509 Ms 422 206 134 11753229444 509 155 421 gt 113 118012244394 w 35 2 42 40 3 4 801 7412994 359 412 41 822 11930504684 159 425 411 42 234 1194 4079814 Figure 26 Sample of Data Collected 13 If the Meet the Designers button is chosen Figure 27 is the screen that will be seen Interactive Cooling System Ch istoph E x 2 4 Jessica Hunnicutt gt m 2 gt ro The 2 Advisors Dr Hosni Abu Mulaweh Dr Hossein Oloami Figure 27 Meet the Designers 14 If the Contact Information button is chosen Figure 28 15 the screen that will be seen 16 170 Contact Information vil 4 an Rose ve d li Figure 28 Contact Information 15 You can go back at any time to view different things or
11. d d Parker Figure 21 Verifying Heat and Temperature Inputs 12 Page 166 8 This page gives var ety of activities that the user do They include viewing temperature at the cold plate vs time graph Temperature Chart viewing the temperature and pressure at each node Summary Pagel viewing advanced calculations such as enthalpy entropy and so on Advanced Calculations starting or stopping the data logging process Start Logging viewing pictures of the ICS team Meet the Designers and viewing contact information for Parker Hannifin Contact Information After each option the user can press the BACK button in to choose another activity except for the data logging button which doesn t go to a different screen but just starts and stops the data logging process with each click See Figure 22 Watch it Cool n Temperature Chart 4 Start Logging Summary Page 4 Meet the Designers Contact information Figure 22 Cooling Screen 13 167 If the Temperature Chart button is chosen screen similar to what is shown in Figure 23 should appear The temperature readings at 0 seconds are the current readings You can increase the amount of data points viewed by clicking the button circled in red and typing the amount of data points you want Temperature Chart Figure 23 Temperature Chart Screen The blue l
12. 146 Page G 2 FLOW METER SUB ASSEMBLY Page 147 G 3 SAFETY VALVE SUB ASSEMBLY NEN LOND 08d WNOLLVOrTI3 4S3004f1V 25 Oi 1 355 937399 ____ NOI LOS fe 1580 1180 61400 40 LRT EY ov Z4 109 O4 21285 1021180 TES 11 DAEMGCDI so DE SOL 111 ee UTIs L OW 1 STS SIRO HERE toe ARTS 1 DE gt gt 2 LLLA 8 8 Lu M 8 6 o o o 148 Page 6 4 BOTTOM CYCLE SUB ASSEMBLY 0318N3SS Y 149 Page G 5 COLD PLATE SUB ASSEMBLY 150 Page G 6 ACCUMULATOR SUB ASSEMBLY 9 151 Page FINAL BRAZING ASSEMBLY ETUE ow oo 9 oV wo o ene
13. be approximated as the following non linear Equation 11 1 e D i 2 51 where eis roughness factor Notice that an iterative process has to be used in order to solve for So pressure difference between node 1 and node 2 can be approximated as the following defined in Equation 12 A 2 ee 12 The maximum pressure head of the pump is approximately 180 kPa at the flow rate of 428 ml min The pressure loss caused by the piping must not exceed 0 196 of the pump head AP ipemax lt 180 kPa The iterative process was carried out using the solver Page 20 command using Microsoft Excel for various standard pipe sizes at the refrigerant and piping properties summarized in Table 4 This produced the pipe sizing shown in Table 9 Table 4 Refrigerant and Piping Properties Property Value Units T TIN E 0 0001649 5 p 1155 kg m Q 1 66667E 05 m s e 0 000005 Table 5 Pipe Sizes and Properties Nominal Pipe Size Inside diameter m f Delta P Pa 1 8 6 833E 03 9305 0 04016 208 0 1 4 9 246 03 6876 0 03768 49 1 3 8 1 252 02 5077 0 03486 11 6 1 2 1 580 02 4024 0 03255 As can seen from Table 5 1 4 piping is the smallest pipe diameter that meets the pressure drop criteria and that was the chosen size PUMP The ICS will be equipped with a 1000 ml min pump The Diener brushless DC motor pump was ordered from the smart series w
14. 71297 YSN 6 AR LOSY 442102044 24 CO HOO 12941 12352302 RHL 106 10 HEATER BRACKETS bt 338 NN ONL H i Pa NOISID3Hd 71497 WSN N3AVH M3N _ NOISIAIQ JIVWTO ha ons ON N23 40114135534 oq U Sts 664 und 144 3 80 157911 v ew M Geen Se whe v uad uui XXX Wife 157 8218 2 _ NYHL 86522 OJAS 244391500 15515 511140044 902 30 149144803 8 13530034 2514 351 43H14 ONY 38 DU d HIS 5945 8 ONLLINONOD 3 10 3504404 4 0350 33H10 01035012510 031402 38 LON 51891 02 13141200 ONIONV1 330Nn NO SI NIJINNVH 01 AMY 1304 4084 ONY 7911 9013 09 9 ANI 1 44371200 SIMI 107 Page APPENDIX B RS LOGIX 5
15. VEW BRAZED amp ASSEMELE ee Page 152 APPENDIX H CONNECTION DIAGRAM FOR THE ICS oT 2 a _ 1 FUSE 8 9 1 9 DINAMITE 30 24 HEATERS 4 24VDC PS 10A e o0 10 24 VDC COMMON FAN l mnm CONTROL ANALOG VOLTAGE OUTPUT 1 24 VDC COMMON FLOW METER FLOW METER Lo DAQ SIGNAL ANALOG VOLTAGE INPUT DAQ ANALOG OUTPUT COMMON 0 13 2 24 VDC PUMP 10 9 e CONTROL O 1 1 VOLTAGE ot 4 11 ANALOG 5 6 DINAMITE lt 10 INPUT 2000 DA10 24F0 0000 5 NODE 1 PRESSURE 1 2 BLACK 121 CONTROL 2 NODE 2 PRESSURE BOARD RED 122 BLACK 5 2 T e NODE3PRESSURE e 123 e 7 BLACK 2 SEM1100 o NODE 4 PRESSURE 2 4 BLACK 5 2 e FED PUMP RED THERMOCOUPLE INP
16. Figure 46 Rubbing O Ring Grease over Entire O Ring Page 49 Figure 47 Place O Ring in Cold Plate Figure 49 Place Clean Glass over Top of Cold Plate 50 Figure 50 Place Cold Plate Housing Cover over the Top of Glass and Cold Plate Figure 51 Cold Plate Assembly Flipped Upside Down Figure 52 Cold Plate Housing Assembly with Continued Torque Applied Page 51 Figure 54 Assembled Cold Plate Housing with Heaters and Thermocouple CYCLE TRANSFER Once the entire cycle is assembled brazed and leak checked the entire system is transferred from the testing rig to the cabinet This requires multiple hands and strength Figure 55 shows the difficulty of rotating the whole two phase cooling cycle out of the testing rig and rotating it under the shelves in the cabinet before the IPS was attached to the cold plate and the accumulator can support the weight of the system without bending the copper tubing The pump manifold accumulator and cold plate were dry fitted prior to the cycle transfer to verify the components would indeed fit through the fiberglass cut outs Page 52 If tolerances did not allow each component to fit through the openings were filed down until they matched and allowed the components to push through without scraping Figure 55 Jessica Hunnicutt and Christopher Gerardot Placing Two Phase Cooling Cycle into Parker ICS Cabinet from Testing Rig not shown SHROUD CONSTRUC
17. the conceptual design stage of the ICS it was decided to use a fuzzy logic controller in order to control the fan This control method was not the method implemented in the final ICS because it was difficult to get an accurate model of the system to put in MATLAB Simulink The method that was implemented was similar to a Proportional Derivative PD controller This controller Is a function of heat load and temperature desired The pump has a set output for a given heat input whereas the fan must vary with the temperature difference of the cold plate with relation to the desired temperature These two control methods will be explained as well as a how they were implemented The controllers were implemented in the same programming language used in the data acquisition process RSLogix 500 The complete RSLogix 500 program can be seen in Appendix B The pump control was designed to maintain a refrigerant quality of less than 0 7 In order to do this MATLAB was used to generate a table that specifies the correct pump speed in both volumetric and mass flow rates These two tables can be seen below in Table 8 and Table 9 MATLAB code used to derive these tables is shown in Appendix C Table 8 Mass Flow Rate Page 24 Table 9 Volumetric Flow Rate 100W 37 21 38 73 4041 4231 4449 46 988 49 85 53 19 57 24 6252 70 20 83 08 109 18 182 15 200 W 0 001461 0 001503 0 001551 0 001606 0 001672 0 001749 0 001839 0 001944 0 002069 0 0
18. Manufacturer Number e Supplier e Supplier Number Use in Interactive Cooling System Unit Unit Price Quantity Total Price Quote Number Quote Date e Order Date e Parker Purchase Order PO Number e Received Date e Calculated Lead Time Once the supplies arrived the ICS began to take shape The construction of the unit is briefly described in the following sections display cabinet construction two phase cooling construction electronics assembly programming and finishing touches Each section Is further divided in to subsections and 1 explained below DISPLAY CABINET CONSTRUCTION FIBERGLASS SHELL MOLD MAKING To develop a fiberglass cabinet with the custom design shown in Figure 14 a fiberglass mold was created Hoosier Patterns Inc is a local company that specializes in mold making for industrial applications After an on site meeting and email correspondence with Keith Gerber and Dave Rittmeyer the president and CAD supervisor of Hoosier Patterns respectively it was agreed to CNC machine high density fine foam to construct the mold for the fiberglass shell Due to the size of the ICS in relation to the Page 34 maximum table size of Hoosier 5 machine the mold was created by machining nine different pieces that were then glued together to make the final mold See Hoosier Pattern s machining and mold making process in Figures 15 to 22 Figure 14 Cabinet 3D Rendering
19. Figure 15 Large High Density Foam Purchased For Fiberglass Mold Page 395 Figure 16 Small Pieces of High Density Foam for Fiberglass Mold Figure 18 Side View of CNC Machining Page 36 Figure 19 Top Plate Machining of Fiberglass Mold Ere 4 A CS necu Figure 20 Plate of Fiberglass Mold in Machine Figure 21 Bottom View of Fiberglass Mold in Assembly Process Page 37 Figure 22 Alternative View of Fiberglass Mold the Assembly Process ALUMINUM FRAME CONSTRUCTION A skeleton frame was designed to house and support the fiberglass skin as well as the touch screen computer DAG and two phase cooling cycle To accommodate this Parker s Electromagnetic Division s Industrial Profile System IPS was used to construct the entire display unit s frame The design of the aluminum frame utilized 40mm standard weight IPS with 28mm to support two phase cooling accumulator sight glass and cold plate To secure IPS to each other various methods of buttress connectors angle brackets and tapped holes for socket head cap screws were used See the base and frame pictures in Figures 23 to 26 for frame details The frame was divided into a base shelves and structure members to make the final assembly rendered in Figure 26 Figure 23 Assembled Aluminum Frame Base of ISC Page 38 SHELF 1 MAKE 2 ALL CONNECTIONS USE Figure 24 Shelf 1 A
20. Personal appreciation is extended to Vanessa for being patient and encouraging Patricia for guidance and unwavering support Page 9 Abstract amp Summary Parker Hannifin s Precision Cooling Business Unit has sponsored the following capstone senior design project for five Indiana University Purdue University Fort Wayne engineering students The purpose of this project is to create an Interactive Cooling System ICS that demonstrates the versatility and capabilities of Parker s Precision Cooling two phase cooling technology The two phase cooling technology utilizes the heat of vaporization of a refrigerant in order to absorb excessive heat commonly generated by a higher powered electronic two phase cooling technology is safer and more efficient method of heat transfer that reduces the weight Increases power density and costs far less than the traditional heat sink or water cooling system The ICS is composed of the following components cold plate condenser fan pump accumulator piping pressure sensors temperature sensors flow meter R 134A refrigerant fiberglass shell aluminum frame power supply personal computer touchscreen monitor data acquisition system and control algorithm These components were researched analyzed modeled and selected to achieve specific performance criteria that are detailed further in Fall 2010 semester s report Report 1 The phase composition of the refrigerant at
21. Pump Speed Voltage Output 100 W 14 0 7 200 W 20 1 300 W 26 1 3 400 W 22 1 6 500 38 1 5 600 W 42 2 1 700 W 46 23 800 W 54 2 7 200 W 58 2 9 1000 W 64 22 With the new pump speeds the ICS stays within two phase region the thermal energy is more efficiently transferred from the cold plate to the refrigerant Without this changed the controller would take more time and would be highly inefficient at removing heat from the system A single control strategy was implemented in order to control the fan That general strategy can be seen below in Equation 13 Page 25 F V F4 V Tep C T C k V C 13 where Fan Output Voltage Fss Steady State Fan Speed Temperature at the Cold Plate T Temperature Desired k Constant Equal to 3 A separate Fss for each heat load was created in order to achieve the desired response The different steady state values are shown below in Table 11 Table 11 Fss for Given Heat Load Heat Load Steady State Fan Speed Equation 100 Fs 2 260E 03 T 2 938E 01 T 1 292 01 1 957 02 200 Fs 1 556E 03 T 2 425E 01 T 1 275E 01 T 2 278 02 300 Fss 1 184 03 3 2 007E 01 T 1 155E 01 T 2 276 02 400 Fss 2 002 03 3 3 534 01 2 2 094E 01 T 4 190 02 500 Fss 5 689E 04 T 1 136E 01 T 7 790E 00 T 1 854 02 600 Fss 1 315E 03 T 2 703E 01 T 1 864E 01 T 4 341 02 700 W Fss 4 7298
22. 3 11 258 length of piping 3 in inches 13 43 0 0254 length of piping 3 in meters u3 0 029 diameter of piping 3 in inches d3 u3 0 0254 diameter of piping 3 in meters A3 pi d3 2 4 cross sectional area of piping 3 in m 2 380 nominal coolant flow in ml min Fco b 1000000 60 Snominal flow rate L s A1 L1 A2 L2 A3 L3 14 285 time constant Page 128 ICS FINANCES AND BILL OF MATERIALS APPENDIX E esas 129 APPENDIX TUBING COMPONENTS COPPER 01 F 1 130 Page 131 2 02 SUIS DONIS 0 Lalo 1 PLI A8 Qxv 22250 IU 553525 denan memm cce ERIS 14 OG VR Fla F 3 03 8 4 _ rv on aa ge puag YSN NR Mw LJ 8 Peres 30 ISiT 811 7 Las 6997159905 020208 104 M
23. Different combinations of heat load and desired temperature were selected making sure that the whole range from 100 W to 1000 W was covered All heat loads were tested with different randomly chosen desired temperatures Also the heat loads were done in random order to make sure that the order of the tests was not a factor in the results since in the final application the order of customer inputs would be random PROCEDURE Supply power to the ICS Select heat load Select desired temperature Wait till the temperature at the cold plate 15 within 1 of desired temperature Choose next set of inputs Repeat steps 2 5 until all heat loads have been covered Turn off power to the ICS puc cue A Page 72 RESULTS The first test done had input values of 500 W for heat load and 60 C for desired temperature graph below in Figure 76 shows the temperature at the cold plate versus time temperature at the cold plate is initially about 55 C and since this is below the desired temperature the fan is initially off so the cold plate can heat up Once the temperature at the cold plate exceeds 60 C the fan turns on and starts to cool down the temperature at the cold plate This is the reason for the initial peak on the graph After a while the fan regulates itself and gets the temperature at the cold plate to settle to about 60 6 which is within 1 C of the desired temperature of 60 C Heat Load 500 W amp Des
24. Table 1 Variables Descriptions and Units Description Heat from Source Controlled Source Temperature State 1 Temperature Temp of Air at Condenser Output Ambient Air Temperature Enthalpy at State 1 Enthalpy at State 2 Enthalpy at State 3 Mass Flow Rate of R134a Mass Flow Rate of Condenser Air Thermal Resistance of Cold Plate Specific Heat of Air Value Units kW Vel kJ kg kJ kg kJ kg kg s kg s C kW kJ kg K To properly control the ICS the thermodynamic system was analyzed completely This analysis was required to properly size and select components such as the heat exchanger the pump the cold plate and the fan The control system will utilize the following thermodynamic analysis to control the temperature of the heat source Figure 1 shows a schematic of the thermodynamic cycle 10 Figure 1 Thermodynamic Schematic of Two Phase Cooling Technology VARIABLES INITIAL CONDITIONS AND ASSUMPTIONS There are several initial conditions and assumptions that must be made to properly analyze the thermodynamics of the system The first assumption that must be made is the thermal resistance of the cold plate Since the designed cold plate is similar to other cold plates that Parker uses the thermal resistance can be assumed to be the same Typical Parker cold plates that work with R134a refrigerant in two phase cooling have a thermal resistance of 30 C kW key a
25. ___ 1A Ad UJIJI dd iuum a LL 262 amnes 157 8114 jJo3dsHSiM3 ___ H38WON 40 580 sot e BAN UOA acl _ Fa Farhi E Cm P 142 Pela COPPER 12 gl ACH WRU 14 FT ow 8 4 S 30 ISI 81145 533 H38AnN 1593 umido GOS 209 19000539 WNOLYONCA 5300 ITY GIODH TCAC RTISILDECOR LXXTCO C2 T c EVE 381 2 DE 3 4 eee SY WIS 511 YWZ SOL WIE DETA gt ENG s os mmm home mun boh cum aoa m m p muc m 143 Page Burma WS p 80 IM 5345
26. signal to the appropriate signal needed by the heaters to provide the correct heat output The input signal for the pump is obtained from the DAG control system This signal is calculated based on the heat input and the quality of refrigerant that is desired The signal input to the fan is also obtained from the DAQ control system Based on the temperature input of the heat source given by the sensor the control system preforms the algorithm to output the corresponding signal to the fan which removes the appropriate amount of heat from the system Other sensor inputs such as pressure from all nodes 11 flow rate of the refrigerant and temperatures from all nodes go into the DAQ as well DATA ACQUISITION AND SIGNAL PROCESSING Data acquisition 15 the process of gathering physical readings from sensors and converting it into a usable analog signal then processing the signal Once the usable analog signal has been generated the data acquisition system filters the signals so that it be converted to a digital signal The output desired from the control system is then sent back to a controlled device via an analog output signal from the data acquisition system Page 30 the ICS the sensors include thermocouples for temperature measurement pressure transducers for pressure measurement and a flow meter to measure the flow rate of the moving refrigerant These sensors will output an electrical signal based on that specific transdu
27. the fan turns on and starts to cool down the temperature at the cold plate This is the reason for the initial peak on the graph After a while the fan regulates itself and gets the temperature at the cold plate to settle to about 74 3 C which is within 19 of the desired temperature of 75 This happened within 40 seconds Heat Load 1000 W amp Des 75 Ol C D Actual Temp D 5 U C2 e esired Temp 40 60 Time seconds Figure 79 Heat Load of 1000 W and Desired temperature of 75 C Page 76 The next test had input values of 400 for heat load and 65 for desired temperature The graph below in Figure 80 shows the temperature at the cold plate versus time The temperature at the cold plate is initially about 689 and since this 15 above the desired temperature the fan is immediately turned on so the temperature at the cold plate can reduce There is undershoot in the graph because the fan speed was initially too high and cooled the cold plate below the desired temperature When the fan speed reduced the temperature at the cold plate increased again and settled at roughly 65 C Heat Load 400 W amp Des Temp 65 C ce C Ol C D Actual Temp C gt C 5 U Q Temp NO 40 60 80 Time seconds Figure 80 Heat Load of 400 W and
28. Actual Temp esired Temp 150 200 250 300 Time seconds Figure 82 Heat Load of 200 W and Desired Temperature of 42 C Heat Load 100 W amp Des Temp 50 C Actual Temp Desired Temp 100 150 200 290 Time seconds Figure 83 Heat Load of 100 W and Desired Temperature of 50 C Page 779 C O C D C2 C C 2 U Q NO C C Temperature Heat Load 800 W amp Des Temp Actual Temp esired Temp 100 Time seconds Figure 84 Heat Load of 800 W and Desired Temperature of 65 Heat Load 300 W amp Des Temp Actual Temp esired Temp 50 100 Time seconds Figure 85 Heat Load of 300 W and Desired Temperature of 45 Page 80 oc 5 U t I C2 C Temperature C Heat Load 900 W amp Des Temp Actual Temp Temp 40 60 80 Time seconds Figure 86 Heat Load of 900 W and Desired Temperature of 85 Heat Load 600 W amp Des Temp Actual Temp Desired Temp 100 150 200 Time seconds Figure 87 Heat Load of 600 W and Desired Temperature of 55 Page 81 EASE OF OPERATION TEST The objective of this test is to figure out how easy it is for users to navigate the GUI Ease is measured by how many use questions are asked during a trial run use ques
29. ICS unit will be 1 The size will match the 3 5 by 7 dimensions of the cold plate John C Ernst the manufacturing company of the glass provides a chart to assist in ordering the proper thickness of glass to withstand the pressure of the working fluid it is displaying and the unsupported length of the glass A copy of the chart is displayed in Figure 4 Page 16 Minimum Recommended Pressure PSIG 1 2 3 4 5 6 7 B 010 11 12 13 Unsupported Diameter inches Figure 4 John C Ernst Pressure Chart for Sight Glass For tempered borosilicate glass the pressure ratings in Figure 4 are increased by 300 Since the maximum operating pressure of the refrigerant is approximately 2068 kPa 300 PSI the maximum pressure read from Figure 4 is 100 PSI The maximum unsupported length is 5 5 Referencing the Figure 4 the minimum thickness of glass is 0 75 thick To ensure that the potentially dangerous glass does not break a safety margin is added to this approximation The cold plate in the ICS will be covered 1 thick piece of glass MOUNTING The cold plate and sight glass are sealed together in the ICS using the assembly shown in Figures 5 and 6 An aluminum plate with counter bores squeezes the cold plate to seal directly to the glass by screwing bolts into a machined overlay aluminum cover The base plate includes mounting holes to secure the assembly to the exterior structure See Appendix A 2 and A 3 for the aluminum
30. adjusted by the automatic control system In future versions it may be desirable to allow it to be manually adjusted in addition to having it automatically controlled An additional manual button could be added to the GUI Several adjustments would have to be added to the ICS control to accommodate this option Page 89 GENERAL SIZE It was required that the 5 must be small enough to fit through a standard door This was achieved Recommendation The 1 5 team would recommend making the depth of the unit bigger for stability reasons REFRIGERANT Parker specified that a refrigerant must be used as the working fluid R134a was the refrigerant used in the 5 Recommendation Due to its low situation pressures it would also be the recommended refrigerant to use in future versions PUMP SPECIFICATION It was specified that the 5 use a pump that was similar to other Parker two phase cooling systems The ICS uses a 1000 mL min pump that is consistent with the requirement Recommendation The same pump 15 recommended on future versions TOUCHSCREEN A touchscreen was to be used to allow an operator to control the ICS In the ICS this was achieved A large HP touchscreen was used in conjunction with InteractX software The touchscreen displays thermodynamic information per the project requirements Recommendation The team would recommend the use of a touchscreen in future versions While the cost is high and the programming
31. after the tank has been opened to allow refrigerant to fill the hoses for a more accurate measurement of how much refrigerant has been moved into the ICS refrigerant flow valve Monitor the scale to determine how much refrigerant has left the tank Close the refrigerant flow valve when the correct reading has been achieved Closethe refrigerant tank You can now remove all hoses from the refrigerant tank oil injector and refrigerant flow valve You have successfully charged the ICS with refrigerant Note You may want to observe the level of refrigerant in the accumulator sight glass in orderto determine if the correct amount of refrigerant was put intc the ICS 22 176 APPENDIX ADVANCED CALCULATION EQUATIONS This appendix details the methods used to obtain the thermodynamic values displayed on the Advanced Calculations screen of the ICS GUI Assumptions liquid 15 a saturated liquid at the inlet of the cold plate Therefore the density p of the liquid refrigerant at the pump can be approximated kg 1167 50 S The pump is designed so that the control voltage is linearly proportional to the liquid volumetric flow rate The volumetric flow rate is determined by Equation 1 where 15 the pump 5 control voltage V 1 67 x 10 8 S r pump E 1 The mass flow rate is determined by Equation 2 m pV 2 At states 1 3 and 4 refer to Figure 1 for s
32. channel thermocouple input module to get pressure readings At the end of the row a terminal end cap 1 attached to close the circuit picture of the DAQ assembly is provided below in Figure 66 Figure 66 DAQ Assembly The Allen Bradley base unit was given inputs for L1 hot L2 neutral and ground from the 10 VAC power strip Next hot neutral and ground wires were also connected to the 24 VDC power supply and at the output of the power supply blue wires were connected for both 24 V and common Hot neutral and ground wires were also connected to the DIN A MITE which is the device that sends the current signal to the resistive heater blocks placed underneath the cold plate A 30A fuse was placed in series with the hot wire connection to the DIN A MITE in order to protect the circuit Hot and neutral wires were also connected to the STATUS SEM which is the device that was used to convert voltage to current that will be passed through the resistive heaters The DAQ outputs a 0 to 10 VDC signal for the heaters and sends this signal to the STATUS SEM This device has many configurations of use depending on how the internal DIP switches are set Since the input was 0 to 10V and an output of 4 to 20mA was desired the internal DIP switches 1 2 and 3 were set to 0010 00 and 0001 respectively The output of the STATUS SEM 4 to 20 mA signal is then sent to the DIN A MITE and then the DIN A MITE sends it to the heaters There are 2 he
33. click at any time to go back to the start screen to start over or to exit the application 171 Exiting Run time Environment 1 Click HOME from any screen lor BACK till you get to a screen with a HOME button and then click HOME to go back to the start screen 2 Click the STOP button on the welcome page This turns off the fan and resistive heaters See Figure 29 recision Cooling Syster 1 Click Stop 8 Figure 29 Home 3 Now the EXIT button shows up Click it to exit the run time environment alternatively if a keybcard is connected hit Alt F4 See Figure 30 ENG NEERING SUCCESS Figure 30 Exit Screen 18 Page 172 4 Close the Machine Shop Suite window it is open usually happens if option 2 is used to get to run time environment 5 Double click the Shutdown icon on the desktop as shown in Figure 31 Figure 31 Shutdown ICS 6 After the computer has shut down flip the power entry module switch to the OFF position 7 Unplug the power fram the wall outlet 19 Page 173 How to Add Remove Refrigerant from the ICS How to Remove Refrigerant from the ICS 1 10 11 12 13 Turn off the recovery unit 13 16 17 First you need to gather three things before starting the physical evacuation of refrigerant from the ICS You will need to get 2 refrigerant hoses a recovery tank and a recovery unit Note All hoses and connections will re
34. connected to the computer include the computer power cord the cable coming from the touchscreen the audio cable coming from the touchscreen and the USB cable coming from the touchscreen the USB extension cables coming from their holders mounted beneath the touchscreen and the serial cord connected to the DAQ system The computer is oriented to place the wires entering exiting the computer into the installed raceway track to maintain an organized appearance and function Figure 64 Computer Orientation and Ratcheting Straps Page 957 ELECTRICAL WIRING In wiring the ICS all electrical wiring codes were followed This means that blue wires were used for DC signals black wires for hot 120 VAC signals white for neutral and green for ground The gauges of the wires were taken into account when installing in order to avoid overheating of the wires whole ICS unit gets its electrical power supply from a standard wall outlet 120VAC A power entry module is utilized in the unit for safety reasons and also to make it easier to turn the unit power on and off The power entry module is a device with a switch in it so that even if the 5 is plugged into the wall as long as the switch isn t pressed on the ON side there would be no power to the unit 12 A picture of the power entry module is provided in Figure 65 A power strip with several outlets for 120 VAC is then connected to the power entry module inside the unit so that it ca
35. grease on both the copper pipe and thermocouple tip placing the thermocouple on the pipe and then taping them tightly together with insulation tape and then securing the whole thing with electrical tape There was a separate thermocouple to measure ambient temperature For this a hole was drilled in the side of the fiberglass shell about 3 feet from the ground and then the thermocouple was pushed through this hole and secured in place The thermocouple at the base of the cold plate was placed in between the cold plate which had been rubbed Page 61 with thermal grease and the two resistive heaters forming sandwich Figure 68 1 below shows a thermocouple that was used in the 5 Figure 68 A thermocouple An electrical connection diagram of the ICS is provided in Appendix H COMMUNICATION BEIWEEN PROCESSOR AND TOUCHSCREEN First the PC was set up so that it doesn t go to sleep mode or log off due to inactivity This keeps the screen always active The software that was used to program the user interface is called InteractX 3 0 This software is used in manufacturing plants and factories for Human Machine Interface and controls Two different applications were set up during the course of this project Application A called Testing3 21 gms to provide an interface to carry out the testing of the ICS and Application B called ICS gms be used in the finished product where the user can select their in
36. hfc1 34a push pdf EBM Papst 220 FTD Series Tubeaxial Fan Available http www ebmpapst us allpdfs 2200ftd pdf Johnson Controls P499 Series Electronic Pressure Transducer online Available http cgproducts johnsoncontrols com MET_PDF 12011190 pdf Ohmite Heat Sinkable Planar Resistor online Available http www ohmite com cgi bin showpage cgi product tap1000_ series OMEGA 7 type Thermocouples online Available http www omega nl Temperature pdf 5LSC 5SRTC pdf Protieus Industries Inc 8000 Series Liquid Flow Meter online Available http www proteusind com 8000 8000DS pdf Rockwell Automation Flow Meter Input Module 1769 F4 online Available http literature rockwellautomation com idc groups literature documents td 1769 td006 en p pdf Rockwell Automation solated Analog Output Module 1769 OF4V online Available http literature rockwellautomation com idc groups literature documents in 1769 in076_ en p pdf Rockwell Automation Pressure Transducer Input Module 1769 F8 online Available http samplecode rockwellautomation com idc groups literature documents in 1769 in067_ en p pdf Schurter Electronic Components Power Entry Module online Available http www schurter com var schurter storage ilcatalogue files document d atasheet en pdf typ_EC11 pdf Sola Power Supplies 24V DC Power Supply online Available http www solahd com products powersupplies sdn sdnC htm 96 Ap
37. industrial cleaner to remove all fingerprints and residue before assembly see Figure 44 for cleaning Second the cold plate was thoroughly and meticulously cleaned with scotch bright to remove all fingerprints and unwanted discoloration of the copper cold plate See Chris Gerardot cleaning the cold plate in Figure 45 Third step is to apply O ring grease to the entire cold plate O ring to insure a tight face seal shown in Figure 46 The next step Is to Insert the O ring and then place the cold plate into the bottom of the cold place housing plate as shown in Figures 47 and 48 respectively Place the cleaned tempered glass on top of the cold place with a cloth to ensure there are no fingerprints as seen in Figure 49 Place the top cover of the cold plate housing assembly over the glass and cold plate without scraping the glass as demonstrated in Figure 50 Gently flip the whole assembly over Figure 51 and begin inserting 20 x 1 inch socket head cap screws SHCS into each threaded hole in the cold plate housing cover through the clearance holes in the bottom housing plate Use 7 ft lbs to torque each screw down as seen in Figure 52 completed cold plate housing is modeled by James Stoller in Figure 53 The heaters were applied with thermal paste and one thermocouple was placed in the middle of the second heater closest to the exit of the cold plate The assembly is shown in Figure 54 Page 48 Figure 44 Cleaning of Borosilicate Glass
38. ounces per year which is the sensitivity limit of the equipment Recommendation There is no further recommendation Page 92 Conclusion In conclusion the ICS system is visually striking and attracts interest of prospective customers at trade shows employees at the Parker Hannifin New Haven facility and students at IPFW Once people are interested in the display unit the ICS allows the user to adjust the heat load and desired temperature of the heat source The automatic control system then accurately brings the electronic device we are cooling to the desired temperature in a minimal amount of time They are able to easily navigate through the GUI on the touch screen to see different properties of the system and information on the two phase cooling technology The ICS team believes that the system has met and exceeded the expectations of the desired display unit for Parker Hannifin s Precision Cooling Business Unit Parker Hannifin is looking forward to showing their capabilities and versatility of the two phase cooling technology with the aid of the ICS Page 394 References 1 2 3 4 6 7 8 9 10 11 12 13 Diener Precision Pumps Gear Pump Silencer 1000 Series online Available http www dienerprecisionpumps com en gear pump silencer1000 html DuPont Refrigerants 734 Refrigerant lonline Available http www2 dupont com Refrigerants en US assets downloads h45945
39. part of the thermodynamic analysis to aid the control system is the flow rate of the refrigerant When the cycle is started a user will input a heat load The system will recognize this load and will set the refrigerant s mass flow rate to a specific value This value is determined by an energy balance equation over the cold plate First the desired temperature of the refrigerant at State 1 must be calculated This temperature is dependent on the controlled temperature input temperature of the heat source and the heat load The desired temperature of the refrigerant is determined from Equation 1 1 Pressure and temperature sensors at State 1 will measure thermodynamic data before the refrigerant travels into the cold plate The refrigerant at State 2 is a mixture of liquid and vapor refrigerant with a quality of 0 7 and has the same temperature and pressure as State 1 Using thermodynamic tables the enthalpy of each of these states can be determined Since it is assumed that the heat out of the source is equal to the heat transferred to the refrigerant over the cold plate Equation 2 can be used to determine the mass flow rate of the fluid Ihe pump speed will be adjusted so that the refrigerant will flow at this rate After the pump speed is adjusted it will constantly provide this flow rate until the user changes the heat load input Q hi 2 Page 12 DETERMINING THE FAN SPEED The rest of the thermodynamic ana
40. that the heat source can be held at steady state at especially in the higher heat load cases 800 is not large For example the device can bring the source to between 70 and 85 at 1000 W This is only a range of 15 The team s recommendation would be to use either a better cold plate a better condenser or a better fan so the ICS would be capable of cooling a 1000 W source to cooler than 70 9C or to design the ICS so that It 15 able to operate at higher temperatures INSTRUMENTATION The 5 was required to use thermocouples and pressure sensors to measure thermodynamic data of the refrigerant Using this information it was to shut itself down when the pressure or temperature exceeded the limits of operation It was also required to be able to export data to a PC Recommendation Each of these criteria was met and there are no further recommendations HUMAN INTERACTION 5 was to have an appropriate amount of human interaction This included the ability to set a heat load and also the ability to set an achievable temperature for the system to cool the source to These options are available for the user to adjust in the GUI The user must also have different viewing options so he or she can observe several different types of data This parameter was adequately achieved Recommendation n the current completed version of the 5 the user does not have the ability to manually adjust the fan It is only
41. the power entry module in the ON position Figure 3 Power Button Switch 3 Openrear cabinet doors Page 156 4 Press power button on CPU See Figure 4 below Press the Power Button Figure 4 CPU Power Button 9 Close rear cabinet doors 6 Go tothe front of the ICS and type in password on screen if prompted 7 Double click on the black and white Parker icon on the desktop as shown in Figure 5 Figure 5 Desktop View Page 197 8 If any license window pops up as shcwn in Figure 6 just close it Figure 6 InteractX Run Startup Service Window 9 Theapplication should be running now See Figure 7 ENGINEERING SUCCESS Figure 7 Application is Now Running 158 Getting to Run time Environment Option 2 1 Follow option 1 from steps 1 through 5 2 Double click on the Machine Shop Suite icon on the desktop as shown in the red circle in Figure 8 f 1 rA gt Figure 8 Desktop View 3 Click the icon corresponding to Open an Existing Machine Shop Project See Figure 9 Figure 9 MachineShop Welcome Window 4 Make sure the Project location displayed is C Program Files Machine Shop Suite Projects See Figure 10 Page 159 Figure 10 Open Project Window 5 Click on CS gms and select See Figure 11 Note If a Project Recovery window appears press Resume ltappears that the last session for ICS terminated
42. unexpectedy onere sin nerd pem rm back to the previously saved project onem Figure 11 Project Recovery Page 160 6 In the Project Browser window expand the tree if needed and double click CS with the pink diamond icon which is under nteract X under kA under 5 See Figure 12 Figure 12 ICS Project File Welcome Screen Note If the Project Browser window is not visible go to View and select Application Browser 7 If any license window pops up like Figure 13 below just close it Start License Service Run As 1 Hour Runtime Figure 13 InteractX Run Startup Service Window Page 161 8 Menu Bar select Tools gt Run Application alternatively click the green forward arrow circled in red the task bar See Figure 14 Figure 14 Starting Application 9 Click Connectea not Offline 10 Again if a license window pops up as in Figure 13 above close it 11 The application should be running now Page 162 Demonstrating the 2 phase precision cooling technology 1 Clickthe START buttan on the screen to start the process See Figure 15 ENGINEERING Figure 1 Home Screen of ICS Application 2 Click BEGIN See Figure 16 Click Begin M 4 Jarker ENGINEERING SUCCESS Figure 2 Beginning Application 163 3 Select the device
43. you would like to cool based the heat load and desired temperature descriptions given A ternatively select CUSTOM in order to specify your heat load and desired temperature inputs See Figure 17 Select Cooling Needs Figure 17 Selection of Cooling Needs 4 the preset devices has been selected or the CUSTOM button was selected the CONTINUE button will show up otherwise it will not show up Click CONT NUE to continue See Figure 18 for how the screen locks after CUSTOM has been selected If one of the preset devices was selected skip to step 8 Select Cooling Needs Figure 18 Selecting Custom Setting 10 Page 164 5 If CUSTOM was selected you have click the heat load corresponding what you want to cool This value shows up in the display window Figure 19 shows the screen when a heat load of 500 W was selected If satisfied click CONTINUE Custom Heat Load Figure 19 Select Custom Heating Load 6 Select your desired temperature This value shows up in the display window Figure 20 shows the screen when a desired temperature of 55 C was selected If satisfied click CONTINUE ire Figure 20 Selecting Desired Temperature 11 Page 165 7 This page shows you the parameters you selected If satisfied click CONT NUE if not click BACK See Figure 21 Get Ready to Cool ran 3 or 4 fi Ira 6 26 C 57 7
44. 00 PROGRAMMING 0000 LES LES ES 0001 Les Than A B Less Than Less Than lt Lern Than A B Source A Source A Source F2 32 Source 531 25 531 25 250 25 0 Source B N73 Source H73 Source B N73 Source H73 2258 5e 224 22 LES LES ES ES 0002 Than Les Less Than A B Less Than A B Less Thani AB Source F834 Source 8 35 Source 28 35 Source Source F8 38 21 5 291 34 297 321 Source Source H74 Source Source H74 Source 80 Boe 80 lt Ble LES Less Than A H Source 38 Source 7 4 te LES 0003 Lest Than Soue 00039 lt Source 0 5 0004 Page 108 0005 0007 0008 109 0010 0011 0012 0013 110 0014 0015 0016 0017 0019 0031 0035 113 0045 0045 Source 1292 1292 114 0047 0050 0051 0052 0053 0054 0055 115 0057 0058 0059 0061 0062 Multiply Source Page 00012 0 0012 116 0054 0067 0069 0071 0072 0073 0074 0075 0076 0078 0079 Source 18541 185 41 Source 28 42 155 6607 42 155 6607 Source 43 0 0 Source 44 46 106 60
45. 02223 0 002430 0 002740 0 003291 0 004563 V 109 00 11340 11813 123 43 129 45 136 34 144 23 153 35 164 11 17748 195 61 223 41 273 32 385 30 143 75 14942 155 54 162 36 170 08 178 90 189 00 200 62 214 22 23080 252 61 284 74 339 51 4153 09 IW 177 64 184 60 192 06 200 29 209 57 220 16 232 28 246 19 262 36 28176 306 62 341 88 399 21 510 37 OOW 21073 21897 22770 237 26 247 99 260 20 274 16 290 17 308 68 33060 358 05 395 66 45417 561 11 700 243 03 252 53 262 51 273 33 285 40 299 09 314 73 332 66 353 29 37750 407 20 446 66 505 54 507 62 OW 274 55 285 31 296 51 308 54 321 86 336 90 354 08 373 74 396 32 42261 454 32 495 29 554 13 551 27 OW 305 29 317 30 32972 342 92 357 41 373 71 39226 413 50 437 85 46604 499 58 541 86 600 46 592 94 335 27 348 52 362 15 376 50 392 11 409 56 429 37 452 03 477 98 50790 543 12 586 58 644 91 733 14 These values were then converted to a voltage using a scaled parameter within the RSLogix 500 code that is programmed into the Allen Bradley Micrologix 1500 base processor These values were relatively inefficient at keeping the quality of the refrigerant less than 0 7 the refrigerant would lose its thermal conductivity and cause the fan control to take time in excess of three minutes to reach its desired temperature A new pump speed table was generated to implement a pump speed greater than that of the previous two tables The new values are shown below in Table Z Table 10 Corrected Pump Speed Values Heat Load
46. 04 1 1375E 01 T 9 2583E 00 T 2 5668E 02 800 Fss 2 6188bE 05 T 2 2100E 02 T 3 9510 00 1 6430 02 200 Fss 7 3317E 01 T 6 0186 01 1000 Fss 6 4591E 01 T 5 4659 01 where Fss Steady State Fan Speed V T Temperature Desired Using the control strategy and Fss the ICS is able to control the system within 1 C The value the control strategy was manipulated in order to achieve a more aggressive change In temperature relative to the temperature difference of the cold plate and the desired temperature Equation 13 was then multiplied by ten in the control strategy so Page 26 that the output to the fan would be a percentage as this was how the RSLogix 500 code was implemented SIMULINK MODELING A Simulink model was created in order to see how close the theoretical model would be to the actual results gathered from the built system In Figure 10 below you can see the plant used in the Simulink model The Simulink model in Figure 11 is designed to show the system cooling to a desired delta temperature the temperature difference between the desired temperature and ambient this model the desired temperature was set to 10 C above ambient temperature which has been normalized to 0 C This Simulink model shows how the system responds when it is cooled from a starting delta temperature of nearly 22 59 and cooled down to a delta temperature of 10 over a period of roughly tw
47. 3218 02 m dot Q 99 T 19 Q 1000 delta density 8E 09 T FLUID 6 2E 06 T FLUID 5 0 0002 T FLUID 4 0 01 T FLUID 3 0 2329 T FLUID 2 1 2427 T FLUID 1293 4 v dot Q 99 T 19 2m dot Q 99 T 19 density 1e6 60 end end The delta h is basically a difference between the saturated enthalpy and the 70 quality enthalpy The equations for that and the density were developed Jo after we plotted a whole bunch of data points in Excel The m dot is the mass flow rate The v dot is the volume flow rate Basically removing the unit conversion factors v dotem dot density where density is a function of the temperature of the fluid shown line 10 m dot Q delta where delta h is a function of the temperature of the fluid in line 7 dot Q delta h density As you can see both temperature and heat input play role in the pump speed Note The MATLAB code generates values for T from 209 to 100 C and heat loads from 100W to 1000W The tables in the control section only show the values for 35 C to 100 C in increments of 5 C and heat loads from 100W to 1000W in increments of 100W Equations came from DuPont R 134a thermodynamic SI tables 2 Page 126 2 EES PROGRAM Heat In dot 1 Critical Temperature 45 Cold Plate Dimensions A 0 0087870792 Heat Transfer Equations cp 50 R134A Flow Rate m dot 134 0 dot h 2 1 State 1 x 120 T 1 dot RH cp
48. 34 0082 0083 0085 0086 0087 0089 ECT Wy A LUE ADD 0 1138 0 1138 FLOAT VALUE ADD Add Source Page 119 0092 0093 0094 0095 iply Source A 3951 3951 Page 120 0100 0101 0102 0103 0104 0105 0107 0108 0109 0110 0111 0112 0113 0114 0115 0116 0117 0118 0119 0120 0121 0122 0123 0124 0125 Lew Than Soume F8 13 140 lt Source 10 0 123 0127 0128 00 lt Source 100 0 100 0 lt 0129 0130 Input Min Inpat Max Sealed Min Scaled 14 0 00 00 1000 100 0 00 00 lt 5000 0 5000 0 013 Seale w Paramaten Input N72 500 Input Min 0 0 00 Input N76 1000 Sealed Min 00 004 Sealed 8 55 10000 04 Output 0 11 Input Min Input Max Sealed Min Sealed Max Output 100 0 28 89 00 lt 00 004 100 0 100 D 00 0 0 10000 0 10000 0 12 1000 124 0131 0132 Page 125 APPENDIX C PUMP PROGRAMMING C 1 MAILAB for 0 100 1000 for T220 100 FLUID Q 1000 15 T delta h 1 529647E 09 T FLUID 6 3 970166E 07 T FLUID 5 3 929848E 05 T FLUID 4 1 794693E 03 T FLUID 3 4 042677E 02 T FLUID 2 2 905297E 01 T FLUID 1 58
49. 60 C C C Ol C D Actual Temp oO C Temp 0 50 100 Time seconds Figure 76 Heat Load of 500 W and Desired Temperature of 60 C Page 73 The next test had input values of 200 W for heat load and 70 for desired temperature The graph below in Figure 77 shows the temperature at the cold plate versus time temperature at the cold plate is initially about 58 and since this is below the desired temperature the fan is initially off so the cold plate can heat up Once the temperature at the cold plate exceeds 70 the fan turns on and maintains the temperature at the cold plate at around 70 C In this graph there is no overshoot in temperature there is a smooth transition to the desired temperature and the temperature is maintained Heat Load 200 W amp Des Temp 70 C Actual Temp Temp 2 5 150 200 250 300 Time seconds Figure 77 Heat Load of 200 W and Desired temperature of 70 C Page 74 The next test had input values of 500 W for heat load and 609 for desired temperature again This test was repeated because we wanted to find out if the system would still reach the desired temperature even if the initial temperature at the cold plate was higher than the desired temperature as opposed to the previous time when the initial temper
50. Desired Temperature of 65 Page 77 The next test had input values of 700 W for heat load and 60 for desired temperature The graph below in Figure 81 shows the temperature at the cold plate versus time The temperature at the cold plate is initially about 63 C and since this is above the desired temperature the fan is immediately turned on so the temperature at the cold plate can reduce When the fan speed settled to its final value the temperature at the cold plate ended up being little lower than 59 This does not meet our requirement of within 1 C but we believe this has to do with different environmental conditions such as ambient temperature and initial temperature of the cold plate Heat Load 700 W amp Des Temp 60 C C Ol C D Actual Temp C2 C Temp O 0 gt 40 60 80 Time seconds Figure 81 Heat Load of 700 W and Desired Temperature of 60 The following figures Figures 82 to 87 show the results of several other test points that were chosen each case the desired temperature was always met to within 1 C This shows that the control system designed is capable of achieving an input desired temperature within the temperature range possible for each heat load Page 78 Temperature 2 5 Heat Load 200 W amp Des Temp 42
51. Indiana University Purdue University Fort Wayne Department of Engineering YS Un M W NN ENGR 410 411 Capstone Senior Design Project Report 2 Project Title Interactive Cooling System Team Members Alex Derickson Christopher Gerardot Jessica Hunnicutt James Stoller Omobola Thomas Faculty Advisors Dr Hosni Abu Mulaweh Dr Hossein Oloomi Date May 4 2011 ____ 4 E me 6 Section Detailed Conceptual Design Tc 9 10 Variables Initial Conditions ANd 5 11 Determining the Flow Rate of the Refrigerant 12 1 2 13 Other 0 1 ET 13 9 14 FER ARa V O N 14 ___________ 16 E 17 Peels c 18 PH 17 E 19 me E 21 ge
52. InteractX 3 0 MachineShop Page 7 Tests were conducted in order to develop empirical relationships between the heat source temperature and fan speed in order to develop a control algorithm working condition test After the control algorithm was in place tests were conducted in order to see how quickly and accurately the heat source temperature reached the desired temperature control systems operation test An ease of operation test was conducted in order to qualtatively assess the CUI Testing revealed that the phase composition of the refrigerant is held at a steady state quality of no greater than 0 7 The maximum weight of the display is less than 500 The display fits through a common 3 x7 door The exterior of the display unit is safe to the touch The ICS is able to dissipate heat Loads from 100 to 1000 W The ICS is able to hold a specific desired temperature to within 1 C of the user input In general the goals set out at the beginning of the project were either met or exceeded Future testing should be done in order to design a controller that is able to handle the transient effects from switching from a higher heat load to a lower heat load Page 8 Section Detailed Conceptual Design THERMODYNAMIC ANALYSIS Variables descriptions and units to aid in the discussion of the thermodynamic analysis are summarized in Table 1 Variable Q source T T air out h h hg 11 134 Mair eq
53. T 1 Pressure R134A T2T 1 1 h 1 Enthalpy R134A T2T 1 1 s 1 Entropy R134A T2T 1 xex 1 rho 12Density R134A 1 1 State 2 O D p 21 2 Pressure R134A T 2 2 Enthalpy R134A T2T 2 2 Entropy R134A T2T 2 2 3 oi mE lea co AF 51 1 101 Enthalpy Air T2T 5 e e uu I 0 Enthalpy Air 6 Fan Flow Rate m dot dot R134A h 1 h 2 h 5 h 6 Flow Rate to ml min dot RH134A rho 1 1e 6 60 Note The Heat Input and Critical Temperature may be varied as necessary to find the corresponding volumetric pump flow rate The actual flow rate sent to the pump is sent in the form of a voltage signal 5 V corresponds to 1000 mL minute The relationship is linear down to 0 V which corresponds to 0 mL minute Page 127 APPENDIX D MATLAB CODE FOR PLANT SIMULINK MODEL 5 parameters oe 1 62 225 length of piping 1 in inches 11 41 0 0254 length of piping 1 in meters 1 0 029559 cross sectional area 1 in in 2 Al al 0 0254 2 cross sectional area 1 122 2 8 length of piping 2 in inches L2 q2 0 0254 length of piping 2 in meters 2 0 073062 cross sectional area 2 in 1172 A2 a2 0 0254 2 cross sectional area 2 in m 2
54. TION The fan shroud was constructed from mild steel sheets using the pattern made in advance The templates were plot in 1 1 scale and cut out appropriately Using traditional duct work bending methods the edges made on the shroud were bent and glued into place using duct caulking See Figures 56 thru 58 for pictures showing duct construction Figure 56 Metal Bending Page 53 Figure 58 Final Shroud FAN AND SHROUD INSTALLATION Due to the tight proximity of the condenser shroud and fan final placements they were assembled individually in the cabinet The condenser is verified of placement in relation to the condenser cut out on the fiberglass The shroud is then maneuvered into place beneath an 5 member and caulked into place on the top and bottom lips The flanges are then inserted and twisted into place with a bead of duct caulk attaching it to the shroud and to the manifold of the condenser The IPS holding the fan is then inserted from the bottom and secured to the aluminum frame of the cabinet The fan is then gently but forcefully placed inside the shroud opening The fan is then twisted to align 3 holes along the tracks of the IPS and secured with IPS drop in T nuts The fan cowl and shroud is then foamed to minimize remove all air leaks that could cause the fan to blow air other than to the condenser Page 54 ELECTRONICS ASSEMBLY The touchscreen for the interior unit is a remounted HP touchscreen monitor The m
55. UT AMBIENT lt 5 14 0 2 NODE 1 lt 41 2 2 NODE 2 lt 14 2 NODE 3 lt 5 14 3 2 NODE 4 lt 144 e COLD PLATE lt A 4 5 153 APPENDIX USER S MANUAL 2011 Interactive Cooling System User s Manual Created By Alex Derickson Chris Gerardot Jessica Hunnicutt James Stoller and Omobola Thomas Page 154 Table of Contents aO 2 Enyironment Option 1 rr 2 Getting to Run time Environment Option 2 5 Demonstrating the 2 phase precision cooling 9 Exiting Run time 18 How to Add Remove Refrigerant from the 20 How to Refrigerant 20 How to Vacuum the ICS prior to Adding Refrigerant 20 Adding Refrigerant to the IOS 22 1 Page 155 ICS User Manual Running the ICS Getting to Run time Environment Option 1 1 Plug in female end of cord to power entry module on the left back side of unit and the male end of cord to a standard wal outlet See Figure 1 and 2 Figure 1 Plug Power Cord into Standard Outlet Figure 2 Plug Female End of Power Cord into Power Entry Module of ICS 2 Flipswitch to ON position on power entry module Figure 3 shows
56. al Temperature 2 200 200 6 Figure 89 Before Transient Response 900 W d 80 C to 600 W 4 55 C The refrigerant is able to enter a superheated state during the transient operation when going from a higher heat load and high temperature to a lower heat load with a low temperature he pump immediately slows its speed per the lower heat load pump control However since the system 15 still running at a high temperature the pump 15 not running fast enough to get the refrigerant over the cold plate before it completely vaporizes solve this problem the design team agreed to re design the pump control to account for this transient operation Page 83 To solve this problem the pump speed increases when the system changes from a high temperature to a low temperature while the fan is operating at 100 This higher pump speed forces the refrigerant through the cold plate fast enough to prevent it from vaporizing before the exit See the transient response of the ICS going from 900 W at 809 to 600 W at 55 C after the control system modification in Figure 90 Transient Response After Control Modification Desired Temperature Actual Temperature LI 100 150 6 Figure 90 After Transient Response 900 W d 80 C to 600 W d 55 C Compare Figure 89 with 90 note the time it takes to achieve the desired te
57. aters used and they are connected in series with each other A black wire comes straight from the DIN A MITE to one end of one of the heaters the other end of the first heater is connected to one end of the second heater with another black wire and finally a white wire goes from the other end of the second heater back to the DIN A MITE to complete the circuit Page 959 The also outputs a 0 to 10 VDC control signal to the fan The fan gets 24 from the positive side of the 24 VDC power supply and the common lead of the fan is connected to the negative side of the DC power supply The positive terminal of the output channel for the fan is connected to the control lead of the fan while the negative terminal of the output channel for the fan is connected to common The fan is capable of 550 cfm at full speed 10 V and this corresponds linearly to the 0 to 10V control input The fan is manufactured by EBM Pabst and was discussed in Section 1 Detailed Conceptual Design The DAQ outputs a 0 to 5 VDC control signal to the pump The pump is a silencer smart series pump manufactured by Diener The pump gets 24 VDC from the positive side of the 24 VDC power supply and the common lead of the pump is connected to the negative side of the DC power supply The positive terminal of the output channel for the pump 5 connected to the control lead of the pump while the negative terminal of the output channel for the pump is connected to common The pum
58. ature was 55 C The graph below in Figure 78 shows the temperature at the cold plate versus time The temperature at the cold plate is initially about 65 C and since this is above the desired temperature the fan is immediately turned on in order to cool the cold plate This happens but then the temperature at the cold plate starts to increase again to about 71 C This is due to the refrigerant in the cold plate losing its two phase cooling effect because the fan was at the highest speed thereby causing the pressure of the refrigerant to drop significantly Eventually when the refrigerant is back in the two phase region about 2 minutes from the start of the test the temperature at the cold plate settles to 60 Therefore the desired temperature was achieved but it just took a longer time and a curvy route Heat Load 500 W amp Des Temp 60 2 Actual Temp O gt U Temp 100 150 200 Time seconds Figure 78 Heat Load of 500 W and Desired Temperature of 60 2 Page 75 The next test had input values of 1000 W for heat load and 75 C for desired temperature The graph below in Figure 79 shows the temperature at the cold plate versus time The temperature at the cold plate is initially about 65 C and since this is below the desired temperature the fan is initially off so the cold plate can heat up Once the temperature at the cold plate exceeds 75 C
59. cer s properties For the ICS the thermocouples will output a voltage the pressure transducers will output a current between 4 and 20mA and the flow meter will output a voltage signal based on the flow rate of the system Figure 13 is a diagram of the system showing where the different sensors will be placed Mode 3 Legend X0 Y X Temperature measurement Pressure measurement Flow meter CONDENSER RESERVOIR 4 2 Mode 1 COLD PLATE PUMP X0 HEAT SOURCE Figure 13 Schematic of the ICS with Sensor Placement Once the sensor data has been acquired the DAG will then condition the electrical signals to filter out the noise and erroneous data This will be done by the data acquisition hardware These filtered signals will then be converted to a digital signal using an analog to digital converter in the DAG Once the controller has determined the desired fan speed for the signal the controller will output an electrical signal to the DAQ DAQ then sends an analog voltage to the fan TOUCHSCREEN AND USER INTERFACE The touchscreen selected for the 5 is the HP Compaq L2105tm monitor The 21 5 inch optical touchscreen monitor was lower in price after more investigation using the Parker company discounts monitor will be attached to a personal computer using a USB port and cable connection The monitor s included stand will be removed Page 31 and instead t
60. cial thanks and consideration goes out to Parker Hannifin s Precision Cooling Department for providing us with project support facilities and funding IPFW Engineering Department for providing us meeting space computing capabilities and a firm engineering foundation Dr Hosni Abu Mulaweh and Dr Hossein Oloomi for providing us technical and theoretical guidance A special thanks to the follow individuals who have helped us build our senior design project Hoosier Patterns Inc Keith Gerber Dave Rittmeyer HyTech Fiberglass Inc Rick Witzigreuter IPFW Dr Donald Mueller Luvata St Louis Dale Hotard Parker Hannifin Headquarters Steve O Shaughnessey Scott Gill Dale Thompson Tim Louvar Brandon Wegmann Matthew Snyder Elizabeth Garr Chris Gorman Kim Ellis RandyBell Fred Pilon Hank Gilbert Brad Bearman Richard Carissimi Andy Muskin Parker Hannifin President CAD Design Sales Marketing CAD Specialists Supervisor Owner Vice President Associate Professor of Mechanical Engineering and Department Chair Senior Sales Engineer Marketing Communications Program Manager Business Development Marketing Director Project Engineer Design Engineer Test Engineer Project Engineer Prototype Technician Senior Lab Technician Engineering Lab Technician Lab Leader Electrical Design Engineer Senior Manufacturing Engineer Senior Visual Communication Designer Automation Technician
61. cover and aluminum base detailed drawings respectively Page 17 Figure 5 Cold Plate Assembly Figure 6 Two Exploded Cold Plate Assembly Views CONDENSER The condenser for the ICS was designed and purchased from Luvata is a microchannel design with approximate dimensions of heat transfer area of 8 x 10 with 1 depth condenser has a maximum dissipation of 1 038 W at 370 cfm air flow from the fan and a refrigerant flow rate of 0 84lbm min See the Luvata dimensioned drawing in Appendix 4 Several trials of the condenser were performed using Luvata s proprietary software An initial evaluation of the results is shown in Table 3 The delta air temperature is the Page 18 change of in temperature between the incoming air temperature and the exiting air temperature Table Condenser Data Leaving Air Temperature Air Temperature Refrigerant Flow Rate Refrigerant Flow Rate Lom min Lom min 462 5 4625 3 711 4 711 5 611 HEATER To create the varying heat loads of the user inputs a heat sinkable planar resistor heater was selected for the ICS 6 Two 1000 W heaters with 3 inch square surfaces will be adhered to the underneath side of the cold plate using the recommended compound A variable voltage will be connected to both leads of both heaters to create the effects of heat production from a heat source such as an processor or Silicon Control
62. e Recommendation The team s recommendation is to re evaluate this requirement at a future date Page 91 AESTHETICS The ICS was to be as aesthetically pleasing as possible because it will often be displayed in public places and the overall appearance represents IPFW and Parker Hannifin The team believes that the ICS is adequately aesthetically pleasing During construction and testing many observers commented on the device leading the designers to believe that it is able to draw interest in people as they pass Recommendation fiberglass shell either a new design or the current design would be recommended again in future versions of the ICS SAFETY The ICS was required to be safe to all operators and observers This required the unit to not have sharp edges not blow overly hot air out the front be unable to burn someone and be vertically stable Each of these requirements was met Recommendation While the unit is sufficiently stable it is recommended that the base be designed to be larger in future versions to add to the stability The unit is not easy to tip over but it can be done if an observer would purposely or accidently add a horizontal force near the top of the unit ENVIRONMENTAL INTEGRITY The unit was to have a leak rate of less than 0 1 ounces per year Using very precise leak checking equipment the unit was carefully tested for leaks was determined by an expert that the unit does not leak more than 0 1
63. e saturation enthalpy divided by the latent enthalpy at this temperature At ICS operating conditions the latent enthalpy ranges from 165 kJ kg to 175 kJ kg In the quality calculation it will be assumed to be 170 kJ kg Equation 6 estimates the quality at state 2 2 x h2 r7 6 The entropy values for the 134 refrigerant are found using a method similar to used to find the enthalpy Again the refrigerant was approximated as a saturated liquid at states 1 3 and 4 To find the entropy at states 1 3 and 4 as well as the saturated liquid entropy at state 2 a cubic best fit equation was developed and is shown as Equation 7 51 saturation 3 4 1 915 X 107 2 084 1079 7 4 014 X 10 5 T1234 1 585 To find the entropy at the 2 phase state the latent entropy must be assumed For the ICS operating conditions the latent entropy ranges from 0 53 kJ kg K to 0 58 kJ kg K For the state 2 entropy calculation it will be assumed to be 0 555 kJ kg K The entropy at state 2 is then approximated by Equation 8 kJ S2 0 555 kg K Ssaturation 8 Note These methods yield accurate approximations ONLY when the system has reached steady state Page 178
64. ear View of Finished Cabinet Delivered to Parker s Engineering Lab IPFW Unit Figure 41 Side View of ICS at Parker s Engineering Lab Page 46 TWO PHASE COOLING CONSTRUCTION TESTING RIG In order assist the assigned brazier of the ICS a testing rig was built in order to house the two phase cooling system pieces and parts during the assembly brazing process See Figure 42 for the isometric view of the testing rig Figure 43 Testing Rig with Installed Two Phase Cooling Cycle Page 47 BRAZING Brazing was performed by the in house Parker professional brazier Chris Gorman He first took the specified drawings of each copper tubing bend see Appendix F for all copper tubing segments made and their xyz corner bend locations programmed them into his copper tubing bending equipment and cut them to size He then cleaned and began assembling each sub assembly according to the schematics shown Appendix Each sub assembly was individually leak tested with helium The final assembly included attaching the pump to the pump manifold and the encasing the cold plate with the custom designed and machined cold plate housing See Appendix A for the pump manifold and cold plate housing detailed dimensions COLD PLATE HOUSING ASSEMBLY The most complex and important assembly of the two phase cooling cycle 15 the cold plate and cold plate housing assembly The first step is to clean the 7 3 5 x 1 tempered borosilicate glass with
65. ects a heat load from 100 W to 1000 W in 100 W increments his is achieved using value buttons A numeric display button also shows what input was selected by the user At the next screen the user selects the desired temperature by a similar process Here the screen is programmed to only display options of desired temperature that the device be cooled to allowed maximum temperatures based on the heat load input that was chosen After pressing the Continue button the screen shows the user the custom inputs that were selected so the user can review before proceeding If the user is not satisfied they can use the Back buttons to make changes To exit the run time environment the user just has to click the Home action button which takes them to the welcome page where the whole adventure began The user must click the Stop button before the Exit button appears This feature is that is avallable in InteractX where certain parts of the screen do not appear until a certain condition is met After pressing Exit that is the end of the interactive session FINISHING TOUCHES LOGO PLACEMENT The logo for Parker was ordered as a black vinyl decal 0 25 thick piece of acrylic was machined and routed to place through the 1 8 thick fiberglass and glued into place A scrap piece of IPS 40mm extrusion was mounted to the back side of another piece of acrylic in which a white LED rope light was secured in a serpentine pattern for even lig
66. ed by pressing the start stop button The ICS provides options to the users to either simulate the cooling of 3 devices called Device 1 Device 2 and Device 3 respectively or to create their own custom device which they intend to cool Device 1 has a heat Load of 200 W and a set desired temperature of 42 C Device 2 has a heat load of 500 W and a set desired temperature of 60 C Device 3 has a heat load of 800 W and a set desired temperature of 65 C This selection of pre defined heat loads and desired temperatures is achieved using discrete buttons When a Continue action button is pressed the screen shows a page where the Page 63 user can make different selections On this page the user can start the logging process press a button to view advanced calculations such as enthalpy and entropy at each node press another button to view a temperature vs time graph of the temperature at the cold plate press a button to view the temperature and pressure at each node and so on The data logging was set up so that data Is saved every second from the time the data logging discrete button 15 pressed to the time 15 pressed a second time to stop the data logging The data is stored in a Microsoft Access file on the desktop called Data mdb When this file is opened the data is found in a table called Mytable This data can be exported to Microsoft Excel and then manipulated from there If the custom setting is chosen the user sel
67. en tested 0 Turn off the ICS FH 30 ONO RESULTS From Figure 75 below it can be seen that at higher fan speeds the temperature at the cold plate decreases for each heat load from 100 W to 1000 W The equations derived from each heat load curve were used in designing the control system and they can be found in the Section Detailed Conceptual Design in Table 11 At 100 W the temperatures at the cold plate are much lower about 30 degrees lower than the temperatures at the cold plate when the heat load is 1000 W The ranges of the temperature at the cold plate for each heat load can also be deduced from the graph For instance from the graph we can tell than one cannot cool a 100 W heat load below 30 C and we cannot cool 1000 W heat load below 65 because at the highest fan speed that is the lowest temperature that can be achieved Page 71 Cold Plate Temperature vs Fan Settings O 100 W 200 W 300 W gt 400 W 900 W 9 600 W 700 W 800 W 900W 1000 W NO OO O o o 4 5 5 O 4 O O 6 Fan Speed Figure 75 Summary of Cold Plate Temperatures with Given Fan Speed and Heat Load CONTROL SYSTEM OPERATION TEST The objective of this test is to make sure that the control system works properly i e the desired temperature at the cold plate is reached within 190
68. fan 5 manufacturer given data The fan also was required to have the capability to exceed the pressure drop that will occur in the condenser According Page 22 to Luvata 5 air pressure drop as air flows through the fins the maximum flow rate is 140 Pa Figure 8 shows that the fan is capable of overcoming this pressure Figure 9 shows a dimensioned schematic of the fan Table 7 EBM Papst Fan Data Voltage Volt Power Noise Max Ambient Bearing Wot Part m s Features DC Range 4 Temp C 155 2214F 2TDHHO 0 261 24 16 to 30 54 66 70 Ball Leads 1 0 Air Flow cubic meters hour 340 680 1019 1359 1599 2039 1 00 248 2 m o 5 3 e 1 5 80 199 2 2 60 149 9 gt 2 9 40 99 20 50 200 400 600 800 1000 1200 Air Flow cubic leet mnulal Figure 8 Manufacturer Pressure Drop Data 3 11 0 280 Figure 9 Dimensioned Schematic of Fan Page 23 CONTROL SYSTEM In the ICS there are two controllers to be designed and implemented The first controller will set the correct voltage output for the pump in order to maintain a refrigerant quality of less than 0 7 and the second controller will control the voltage output to the fan in order to achieve a desired temperature to be maintained at the cold plate Two different methods of controlling the fan and the pump were created because their design goals were much different
69. he monitor will be secured directly to the aluminum extrusion frame of the 5 using a quick release VESA 100mm bracket The user will be asked to choose a device to be cooled or create a custom scenario by pressing the desired option on the touchscreen If a predefined electrical component 15 selected pre programmed input values will be used for the control system If a custom scenario Is selected the user will be asked to enter in a heat load and desired temperature of the device being cooled enter in the heat load and desired temperature the user will input these values by touching the corresponding options on the touchscreen After the inputs have been entered the system will begin working to meet the desired temperature While the control system is in operation the pressure and temperature at each node will be displayed The flow rate of the refrigerant temperature of the heat source and ambient temperature will also be displayed Page 32 Section ll Building Process PROCUREMENT To complete the senior design conceptual design a detailed list of components material tools and supplies were composed and ordered through Parker Hannifin This bill of materials was continuously updated and adjusted during changes in production of the ICS units and quotes consultations with suppliers The bill of materials for the display unit is documented in Appendix E This Ust includes the following Description Manufacturer
70. he painter This custom automotive paint was taken to local hardware company and shook to re mix the color pigments Hobby craft brushes were then cleaned and meticulously used to touchup every imperfection found Figure 71 shows Jessica Hunnicutt correcting and scuff to the 105 from shipping Figure 71 Paint Touch Ups Page 66 Section Testing STEADY STATE TEMPERATURE AND TIME CONSTANT TEST The objective of this test is to generate a temperature vs time plot of the temperature at the cold plate at the nominal operating conditions 500 W heat load 38 pump speed and 5096 fan speed in order to obtain the steady state temperature and the time constant of the system PROCEDURE Supply power to the ICS Set the heat load to 500 W Set the pump speed to 38 Set the fan speed to 50 Start the data logging until temperature at cold plate reaches steady state Stop the data logging Graph temperature vs time plot Scale off the x axis crossing of the initial slope 10 Scale off steady state temperature 11 Turn off power supply to ICS c ee ae M a RESULTS The test results showed that the steady state temperature is about 49 and the x axis crossing of the initial slope is about 14 seconds The ambient temperature was about 22 C during this test See Figure 72 for charted results of the test 900 W Steady State CI OI FP Oo D
71. hich includes the pump controller integrated into the assembly From the thermodynamic calculations the maximum volumetric flow rate is determined to be 427 7 ml min 7 13E 6 2 5 Using Figure 7 the available pressure the pump can deliver at 2000 rpm is 1 8 bar 180kPa Page 21 Typical Performance Curve Water at Room Temperature 1200 1 800 1000 Flow Rate ml min 2 e e 5 Amperage amps Flow 000 Flow 2000rpm Flow 3000rpm Flow Full Speed Amps 1000 rpm 0 1 2 3 4 n Amps 2000 rpm Amps 3000 rpm Differential Pressure Bar Amps Full Speed Figure 7 Performance Curve of Diener Silencer Series Pump 1 The capacity of the pump was verified after reviewing the maximum pressure drops across the various 5 components included in Table 6 See the custom engraved pump base plate in Appendix A 5 Table 6 Component Pressure Drops and Gains Max AP Drop Across Condenser 2267 Max AP Drop Across Cold Plate 309 Max AP Drop Across Pipe Pa AP Gain Across Pump at Max Flow 180 000 FAN The fan that was selected for the ICS condenser is an EBM Papst 2214F 2TDHHO fan With the data from Luvata a fan with a volumetric flow rate of 0 175 3 5 370 ft min is needed to eject 1000 Watts of heat from the system The fan also needs to a diameter smaller than the height of the condenser face 8 in The selected fan meets both of these requirements Table 7 presents the
72. ht distribution Once the acrylic glued to the fiberglass cured the vinyl cut out Page 64 decal of the Parker logo was gently smoothed with the aid of a plastic scrapper to remove air bubbles See Figure 69 for a picture of the final logo Figure 69 Parker Backlit Logo GRAPHIC ART PLACEMENT The final graphic design finalized with the aid of Richard Carissimi was sized and ordered on matte vinyl Once received the fiberglass was gently marked with pencil for centerlines on the face and on the decal transfer paper in order to assure correct vertical and horizontal placement A level was used to verify that the vinyl decal was not tilted in either direction See the vinyl art also shown in Figure 69 UNDER CABINET LIGHTING The cabinet was accented with under lighting This was done with the use of plastic P clips mounted to the underneath side of the IPS base and centered around the entire perimeter of the unit The power cord was then zip tied and plugged into the power strip of the unit See Figure 70 for underbody light reflection Figure 70 Parker ICS minus Touchscreen with Underbody Lighting Reflection Shown Page 65 PAINT TOUCH UPS Due to the unstable and accident prone nature of any project small nicks and scratches were a result of transporting the unit to and from the fiberglass manufacturers the paint shop Parker facility etc To correct or hide the damage extra paint was collected from t
73. ine shows the actual temperature at the cold plate while the red line shows the desired temperature selected 55 C 14 Page 168 10 If the Summary Page button is chosen a screen similar to what is shown in Figure 24 should appear The temperature anc pressure at each node as well as ambient desired and cold plate temperature will be displayed Summary Page Figure 24 Summary of Sensor Data 11 If the Advanced Calculations buttor is chosen a screen similar to what is shown in Figure 25 should appear Enthalpy and Entropy at each node state as well as volumetric flow rate of the refrigerant and the quality of the refrigerant will be displayed Advanced Calculations Figure 25 Advanced Thermodynamic Properties Calculated 12 If the Start Logging button is chosen the screen still remains the same but the data logging will begin After one has exit the run time environment the data will be stored on 15 Page 169 the desktop in a Microsoft Access file called Data To stop data logcing just click the Start Logging button again which by this time will read Stop Logging Figure 26 shows screen shot of some sample data collected Yi g r SS 4 8 T e Xo UM 4 5 214 12029571473 500 10 54 7 n 113 2057811 5 53 ma 18 134 x 41 ms 19
74. l UUMM 67 Steady State Temperature and Time Constant 68 Fale E 68 68 ep bathe ST 69 n 69 E 69 a 70 PRO SUL SO 70 Working 011586 ER 70 PrO COUE X LUE IHE DM M 71 p c 71 Control System TeSt 72 Licx E Uu MMC x LI MM Ld e ML 42 FS SO TE 73 FASC Ol Operation 82 FOC 6 1 gm 82 8 82 Controt Design a aa 83 Section IV Evaluation amp RecommendationS exstare na vest ot uio 86 Phase COMBOS MOM 87 EL T E 87 c 88 Exterior Operating Temperature 88 Per ormnmaN 88
75. l input is ten percent of the fan working capacity 1 10V control input to the fan Gris the relationship between pump speed and the desired heat load his value is calculated to 5 479 107 m3 J Figure 12 Simulink Response As can be seen above the Simulink response is rather precise and accurate demonstrates 5 given a desired temperature of 10 can achieve this temperature within three minutes Page 29 ELECTRICAL CONNECTIONS The ICS will be getting all of its electrical power from a standard wall outlet 120VAC power entry module is plugged directly into the wall outlet and this is there to isolate the system from the main power supply for safety reasons lt also helps to be able to cut off power to the 5 when needed This delivers the 120VAC straight from the wall without modifying it 24VDC power supply is connected to the power entry module and it outputs 24V for various DC components in the ICS 13 The PC and monitor also get their 120VAC power from the power entry module The Data Acquisition DAQ hardware and the DIN A MITE which controls the resistive heaters heat source also get their power 120VAC from the power entry module The 24VDC power supply supplies voltage to the pump the fan the flow meter and the pressure sensors DIN A MITE supplies current to the heaters which give off heat corresponding to the signal from the user input The DIN A MITE converts this input
76. led Rectifier SCR PIPE SIZING The following is an analysis of how the size of the pipes that will be used in the ICS was determined Reynolds number can be expressed as the following relationship in Equation 4 Res ES 4 where density 15 p velocity 15 v pipe diameter 15 D and viscosity 15 Mass flow rate is m is defined in Equation 5 m pQ 5 where volumetric flow rate is Q The velocity of the refrigerant is also equal to the following shown in Equation 6 6 40 4m A Page 19 where pipe cross sectional area 1 A Reynolds 5 number may also be expressed as the following as defined in Equation 7 4 411 The Bernoulli s equation relates the pressure drop between any two points in single path pipe system see Equation 8 This is used to calculate the pressure losses from the pump outlet node 1 throughout the entire refrigerant cycle piping and back to the pump inlet node 2 P 2 2 az X him 8 For 1 5 the change in kinetic energy and the change potential energy throughout the cycle are negligible and the minor losses are also negligible so the energy balance 15 reduced to the following as shown in Equation 9 P4 2 _ 2 2 2 Xn 9 Each major loss in pressure head is defined in Equation 10 L v 10 where friction factor for turbulent Reynolds number Re gt 2300
77. lysis aids the controller by determining and adjusting the fan speed to properly cool the ICS With State 2 known State 3 was analyzed under ideal running conditions the refrigerant should be a saturated liquid at this state It also was assumed that the pressure drop in the condenser was equal to 80 kPa Using these two properties the temperature and enthalpy of the refrigerant were determined from published thermodynamic tables Since the enthalpies of State 2 and State 3 are known as well as the flow rate of the refrigerant inside the condenser Equation 3 can be utilized to determine the flow rate of the air 1 3 While the Equation 3 holds valid it does now allow a solution of the mass flow rate of the air because the temperature of the air leaving the condenser is also unknown Since the refrigerant flow rate is pre determined the values may have to be interpolated from the condenser data Once the flow rate is determined the values of the air flow rate and the leaving temperature must be adjusted until Equation 3 becomes balanced This will give an initial air flow rate to communicate to the fan The controller will monitor the temperature of the cold plate and other thermodynamic properties of the system If the system does not become steady with the characteristics that match the thermodynamic model the controller will adjust the fan speed until the system Is running p
78. mperature is much less after the modification The control system was tested rigorously after the modifications were made in order to ensure performance across the heat and temperature inputs see Figure 91 Note the absence of unstable peaks like in Figure 89 the two phase refrigerant composition was not lost Page 84 Figure 91 Screen Shot of Control System Test Page 85 Section lll Evaluation amp Recommendations In the evaluation of the ICS it was necessary to determine if the device adequately meets each of the requirements and specifications listed in the problem statement The problem statement was created by the team prior to design of the ICS The following subsections will examine the performance of the ICS in regard to each of these requirements and specifications as well as listing the recommended course of action if a future 1 5 was created PHASE COMPOSITION The ICS was designed so that refrigerant would change phase as it passes over and cools the cold plate It was specified that this two phase process must be visible and the refrigerant must not enter a superheated state he desired quality at the exit of the cold plate was set at 0 70 Testing concluded that the ICS meets the first part of this requirement is easy for an observer to see the change of state of the refrigerant as passes over the cold plate second part of this requirement was also met After the cont
79. n converts it to flow measurement in gallons per minute gpm 8 9 The flow meter was rated for 0 05 to 0 3 and since this range was tight the units were converted to ml min and that was displayed on the screen instead A 24 VDC signal was connected to the flow meter as well as a common signal output signal of the flow meter was then connected to the positive terminal of the input module channel and the negative terminal was connected to analog common the DAQ Analog common on the DAQ was then connected to the negative terminal of the 24 VDC power supply The thermocouples used were T type thermocouples made by Omega 7 They came in 36 lengths with one end of the dissimilar metals soldered together and the other end in a plastic terminator Since some of the thermocouples were going to be placed more than 36 away from the thermocouple input module their lengths had to be extended First a rough estimate of how long each of the six thermocouples had to be was made and then extra thermocouple wire had to be obtained cut to length and then terminated with a corresponding plastic terminal so they could be attached to the purchased thermocouples Ihe thermocouples were hooked up with the positive sides to the positive terminal of each channel in the DAQ and the negative sides to the negative terminal of each channel in the DAQ There were four thermocouples attached to each node in the thermal cycle by first placing thermal
80. n provide power for the various components that need 120 VAC such as the CPU the touch screen monitor the Allen Bradley DAQ base unit processor the DIN A MITE the STATUS SEM and the 24 VDC power supply Figure 65 Power Entry Module All of the electronic components were mounted on DIN rails which were then secured to the internal aluminum skeleton of the ICS The DIN rails provided great support and flexibility They were convenient because all of the components were shaped in a way that they could be mounted on DIN rails and so they could be slid from side to side in order to make room or create adequate spacing between components DIN rails also provide grounding capabilities Terminal blocks or DIN nectors also proved very helpful In connecting several wires together without all the connections being not organized The Allen Bradley DAQ consists of a base module and if needed several 1 0 modules connected to the base processor which must then be terminated with an end cap The base unit used is the Micrologix 1500 LRP series and it contains the processor supplies 24 VDC to the other attached modules has LED indicators and contains some 1 0 ports The expansion 1 0 modules used include 4 channel output module to send control signals to the fan pump and heater an 8 channel analog input module to get readings Page 58 from the pressure sensors 4 channel analog input module to get readings from the flow meter and finally a 6
81. nd more fiberglass panels and resin are used to build a total of 1 8 thick fiberglass skin once it is cured See Figure 30 for resin application Figure 30 Resin Application Once the aluminum frame is briefly dry fitted to verify the dimensions the fiberglass manufacturers continue to spray a back lip onto the inside of the mold after attaching a barrier to make the back of the unit more easily accessible with the creation of reinforced fiberglass doors Once the whole fiberglass skin still in the mold cures it is then painted on the inside of the fiberglass to make the final display cabinet easier to assembly and maintenance on the inside The aluminum frame is then inserted from the bottom of the mold and placed appropriately in the mold Selected members of the aluminum frame are then fiber glassed into place to make the final bond between the frame and the fiberglass shell The frame and fiberglass assembly is then hoisted out of the mold by an overhead lift and blown out of the mold through the blowholes to overcome the suction effect see Figure 31 Page 41 Figure 31 5 Removed from Mold Doors are made from reinforced fiberglass and epoxied in place to the of the back of the fiberglass shell with the aid of piano hinges The fiberglass cabinet is then one step closer to completion CUSTOMIZED CUT OUTS To accommodate and showcase the two phase cooling technology the unit undergoes selective cutting Using a pr
82. nted cured and buffed the masking comes off including the strip covering up the pinstripe The unit is then clear coated and buffed twice to unveil a car body finish See Figures 34 to 38 for photos taken during the painting process of the IPFW donation unit Same process was repeated for the Parker 1 5 unit Page 43 Figure 34 Pinstripe Painted Figure 35 Top and Bottom Painted Figure 36 Mid Section Painted Figure 37 3D Borders Painted Page 44 Figure 38 Rear View of Finished Fiberglass Cabinet with Clear Coats CASTER ADJUSTMENTS Before the unit is set up right after the unit arrives from Hy Tec Fiberglass the casters are removed and cleaned thoroughly Due to the dirty dusty environment of the fiberglass manufacturing the display cabinet is covered in fiberglass dust that needed to be removed After the casters were removed from the base of the unit the four bolts holding them in place were applied a generous amount of Lock Tite thread glue and screwed back into place This ensures that rigorous travel and vibrations from moving the unit on uneven surfaces does not cause the caster bolts to loosen and eventually fall out leading to a catastrophic collapse of the unit The final IPFW cabinet at Parker s Engineering Lab is showcased in Figures 39 to 41 T i Figure 39 Finished Cabinet Delivered to 5 Engineering Lab Unit Page 45 Figure 40 Finished R
83. o and one half minutes Figure 12 shows the response of the system as a function of time See Appendix D for MATLAB code of Simulink plant values 9 Transport Integrator Delay d Gain Figure 10 Simulink Plant The plant can be modeled as seen in Equation Y below where the steady state gain and is equal to 1 Tp the time constant of the plant and is equal to 14 285 Note the delay of the system is equal to 6 2907 Page 27 Figure 11 Simulink PID Controller where 10 G H 4E 05 9000 5205 and it is denoted as Transfer C s 5245 Function 5 47 09 94 10 Gi These values were measured or calculated from the physical system and then implemented in the Simulink model in order to produce results that would mimic the results of the actual system had the heat load and desired temperatures been input into the physical system Grand were derived from the properties of the thermocouples This relation involves both temperature C and Voltage mV and be seen below in Table 12 Page 28 Table 12 Temperature to Voltage Relationship for Thermocouple Temperature C Voltage mV 20 0 79 30 1 196 40 1 612 50 2 306 60 2 468 70 2 909 80 3 358 90 3 814 100 4 277 From this table a relationship of 1 C to 4 0E 05 Volts was produced Gris the fan gain constant was it was arbitrarily chosen to be 1 10 because one volt of the fan contro
84. of the GUI takes up a significant amount of time it definitely adds to the aesthetic appeal of the ICS It also allows for an organized way to control the 5 and also display information Page 90 COLD PLATE A copper cold plate was to be designed to evenly disperse the refrigerant flow through micro channels The cold plate in the ICS does an adequate job at this During operation it is evident that the refrigerant is flowing through each micro channel fairly evenly Recommendation There is no further recommendation HEAT EXCHANGER A heat exchanger was to be designed or selected to reject heat from the system In the 5 a refrigerant to air condenser was selected condenser coil was sized to fit a single controllable fan Recommendation This type of heat exchanger would be recommended in future versions COST The two ICS units together were to cost no more than 30 000 This cost limit was exceeded but no more than 10 of the overall cost In future versions it may be desirable to reduce cost If this is the case a significant amount of the cost could be reduced by not using a custom fiberglass shell or reusing the current shell Recommendation The 5 team recommends quotes be received from more suppliers manufacturers during the procurement process to reduce costs LOW MAINTENANCE The ICS was to require as little maintenance as possible At the time of this report the 5 has not yet required maintenanc
85. onitor is first laid down with the touch screen facing downward A Philips screw driver was used to remove the monitor stand Figure 59 included in the shipment of the monitor A Quick Release mount was purchased and the included instructions of the bracket were followed accordingly The attachment of the back plate is shown in Figure 60 Using drop ins for 40mm IPS aluminum extrusion the front plate of the Quick Release mount is placed on two pieces of the IPS Figure 61 Once secured the placement is tested Figure 62 and placed in the cabinet The touchscreen is then attached to the IPS by sliding the Quick Release Bracket together as demonstrated in Figure 63 Wires required for the touchscreen include the monitor power supply USB cable for touchscreen capabilities speaker wire and cable These are organized in a raceway track and connected to the appropriate terminals on the back of the computer Figure 59 Remove Monitor Stand from Monitor Page 55 Bas Figure 62 Check the Alignment and Screw down IPS to Designated IPS Pieces that are Pre Installed in Cabinet Page 56 Figure 63 Temporarily Engaging the Quick Release Mounting COMPUTER INSTALLATION To attach the computer to the inside of the fiberglass unit ratcheting straps were utilized in securing the front and back of the computer as shown in Figure 64 The excess nylon straps were cut off approximately one inch from their exit of the ratchets The wires
86. ovided template of the size and placement of the fiberglass the following components required cut outs Power Entry Module Fan Air Intake Grill ouch Screen USB Ports Accumulator Pump Manifold e level Indicator Sight Glass Cold Plate Assembly See the following Figures 32 and 33 for views of the final fiberglass cut out size and placement of the primed cabinet Page 42 Figure 32 Rear View of Primed ICS Cabinet with Fan Air Intake Grill Figure 33 Front View of Primed ICS Cabinet with Essential Two Phase Cooling Cycle Cut Outs PAINTING AND FINISHING The unit is first primed and allowed to cure and then sanded to a smooth finish before painting The first coat is the color for the pin striping bands above the top 3D border and below the bottom 3D border After it is allowed an ample amount of cure time it 15 buffed smooth The next stage of painting 1 the top and middle section of the unit The unit is masked off to cover the exact pinstripe and the middle section These sections are also allowed amble amount of time to cure and then buffed smooth The next coat of color is the middle section The finished top and bottom sections are then cover before the middle is painted Curing time and buffing follows The last color is the 3D borders entire unit minus the area of the borders is masked off The paint cures and is buffed appropriately After all the sections are pai
87. p is capable of 1000 ml min at full speed 5 V and this corresponds linearly to the 0 to 5V control input The pressure transducers used were the 499 series electronic pressure transducers manufactured by Johnson Controls 4 A pressure transducer was placed at each of the 4 nodes in the thermal cycle They output 4 to 20 mA current proportional to a pressure of 0 to 500 psi These pressure transducers have 3 leads red supply white output black common but since their output is current they are only used in a 2 wire configuration red and black The red wires of all the pressure transducers were connected to 24 VDC and the black wires were each connected to the positive terminal of channels 1 through 4 on the 8 channel analog input module Each negative terminal of channels 1 through 4 were connected to analog common on the input module which are all internally connected and then the negative side of the 24 VDC power supply was connected to one of the analog commons The current produced by the pressure transducers will be in the range of 4 to 20 mA and this current reading will be transferred to the base processor and converted to a pressure reading based on the configuration of the module picture of the pressure transducers used in shown below in Figure 67 Page 60 Figure 67 Pressure Transducer The flow meter used was a Proteus 8000 series liquid flow meter and it outputs a 0 to signal to the DAQ input module which the
88. pendices Page 97 APPENDIX A CUSTOM COMPONENTS COLD PLATE A 1 ESI Ia UU 2 2 OL See 72 36 22 22 gp Temas 224 met 1 PE s ee no oe 98 A 2 COLD PLATE COVER Ew B Y 2 gt 71231 SW31SAS 31 12 Si 1103 X OSC 574 p 01 IE 07011240 14 LOM LI 04 YIN AEL IHY 1 I VIF EE MO 11134220 1 AHL KOU VARET ERAH 01 CUES WULAN QU 02 99 Page 3 COLD PLATE BASE M3AO2 WOLLOB3lYld 0700 rire WS KORO 51431545 317v TO NOH PV SJ 82 INN 02 11 055 4 1 050 579 21 052 4 1 NYHL 8920 TUO 0
89. puts of heat load and desired temperature and the 5 automatically regulates itself InteractX makes use of devices called tags These can be likened to variable names in regular programming like C programming These tags contain values like integers floats strings and even binary information InteractX also has numeric displays to display numerical values which could be values stored in tags There are also action buttons available mostly these enable the user to move on to another panel screen Then there are discrete buttons used primarily for turning certain bits on and off There are a wide range of useful tools like this in InteractX which are located in the Tool Bin Page 62 InteractX is designed to work hand in hand with the DAQ systems The applications created were linked to the Allen Bradley equipment used by creating and setting up the channel and device and connecting the Allen Bradley base unit to the PC using the cable provided InteractX can also display sensor values if these values are set up so that they are stored in these tags For the InteractX application A the inputs were heat load fan speed and pump speed The values to be displayed were the temperatures pressures flow rate of refrigerant and quality of refrigerant at the exit of the cold plate Numeric Display tools were used to display all of these In addition buttons were created to increment or decrement the inputs so that certain test points could be crea
90. quire a 3 8 charge port or seal rite The hoses are typically 4 8ft in length and should be able to hold at least 500psi burst pressure Attach seal rites to the end of the refrigerant hoses This is so you will not spray refrigerant or Krytox oil upon conrecting or disconnecting the hose from the ICS recovery unit or the recovery tank Attach one seal rite to the ICS charge port located in the center of the mechanical system Attach the other end of the same hose to the recovery unit Make sure that this hose is attached to the connection of the recovery unit denoted inlet Attach one end of the second hose to the last remaining connection of the recovery unit denoted outlet Attach the other end of the same hose to the connection on the recovery tank denoted liquid If there is one port on the recovery tank this distinction will not be made and connect to the only port on the tank Open the valve on the top of the recovery tank Check to make sure that both dials located near the connections denoted inlet and outlet of the recovery unit are set to the closed position Turn on your recovery unit First open the outlet dial of the recovery unit to the most open position Slowly open the inlet dial of the recovery unit This is done slowly because at some point you might run into a situation where refrigerant and oil move too quickly through the recovery unit and it will make a loud noise If you hear this noise turn
91. r temperatures of all of the parts of the 5 were not to be hot to the touch While the exterior aluminum of the cold plate assembly and the copper of the accumulator can get fairly warm there was never a point in the testing where the surfaces became uncomfortably hot to touch with a hand Recommendation f a device similar to the ICS is created and is designed to run at temperatures higher than what the current ICS can run at the surfaces may become too hot to touch It would be recommended to either situate the components in a way in which the operator cannot touch them or add an outer insulating material so that the metal surfaces are guarded and untouchable PERFORMANCE was determined that the 5 must be able to deal with heat loads from 100 W 1000 W and it must be capable of cooling the source to within 1 C The ICS was designed to meet both of these criteria During testing it proved to cool the heat source within the required accuracy for the entire range of heat loads specified he system also proved to cool the system to a specified temperature within the required 3 minute time interval The system actually cooled slightly faster than the theoretical calculations Figure 92 shows the actual system response compared to the theoretical response in the mathematical model Figure 92 Comparing Theoretical Response to Actual Response Page 88 Recommendation While the requirement was met range of temperatures
92. ributers is to equally distribute the refrigerant to the micro channels so that the flow through the channels is uniform The flow distributers were optimized for uniform flow by using a trial and error method The model was adjusted using SolidWorks software until a uniform flow through the channels was accomplished The analysis was performed using the FEA simulation tools included in SolidWorks After the flow distributors were designed the channel layout was designed Since the channels needed to be large enough to easily view the flow of the refrigerant and the evaporation it was determined that the cold plate should contain no more than 30 channels Three different channel layouts were compared using FEA The first layout contained 10 channels the second contained 20 channels and the third contained 30 channels Using ANSYS software the thermal resistance of the cold plate was determined for each of these cases using two different convective heat transfer coefficients he values of these convective heat transfer coefficients were selected based on a range of published values for boiling refrigerant Table 2 shows the results of the FEA Page 14 Table 2 FEA Results of Channel Optimization h 1 500 W m K 100 000 W m K n channels Tmax C Trefrig C Reg C W Tmax C Trefrig C Req C W After performing the analysis it was determined that a cold plate with 30 channels was the most effective and therefore
93. rn on the vacuum pump Wait until the micrometer is below 100 microns If the micrometer will not go below 300 400 microns then the ICS is leaking and you will need to perform a leak test the ICS is below 100 microns you can remove the seal rite thet is attached to the ICS Turn off the vacuum pump Turn off the micrometer Remcve the seal rite attached to the vacuum pump You may now safely remove the hoses from both ends of the micrometer You have successfully pulled a vacuum on the ICS Note At this point you should be ready to charge the ICS with refrigerant 21 175 Adding Refrigerant to the ICS 1 19 12 13 14 15 16 17 First you will need to gather 9 things You will need to two refrigerant hoses a tank of clean R 134a refrigerant a scale capable of measuring precisely how much refrigerant is removed from the refrigerant tank an oil injector because the ICS charge port is not upright Krytox oil or comparable a valve to open close the flow of refrigerant two 3 8 male to male adapters like a double sided charge port a 3 8 plastic sealing cap and possibly a hand help electric heater Note Clean refers to refrigerant without oil This refrigerant should be pure Attach seal rites to the end of the refrigerant hoses This is so you will not spray refrigerant or Krytox oil upon conrecting or disconnecting the hose from the ICS recovery uni
94. rol system was completed there was no observed case in which the refrigerant was completely vapor at the cold plate s outlet Due to the design of our cold plate the quality is generally between 0 35 and 0 5 at the cold plate s outlet which is significantly lower than the specification of 0 70 This exit quality is lower than specified because the pump Is required to run faster than it was originally designed for to prevent the unstable condition described in the Control Design Modification section of this report Recommendation n future versions of the ICS it may be desired to use a smaller cold plate or develop a more intelligent pump control to prevent the unstable state described in the Control Design Modification section of this report If these issues were fixed the system could run at a higher quality which would yield more efficient cooling FAN SPEED second requirement was that the 5 had to have a variable fan speed that was controlled by feedback from the operating conditions Recommendation This requirement was met here is no further recommendation Page 87 MAXIMUM WEIGHT The ICS was not to exceed more than 500 lbs This specification was set so that the display could be moved easily The ICS unit was weighed at 363 lbs on an industrial scale It also proved to be easily transportable Recommendation This specification was met here is no further recommendation EXTERIOR OPERATING TEMPERATURE The exterio
95. roperly OI HER CONSIDERATIONS The highest pressure in the system will occur at the outlet of the pump which is at State 1 Since the temperature and the quality of State 1 are both known the pressure can be determined from published thermodynamic tables system will be designed to withstand this pressure with a high safety factor at the highest heat load that the source will provide This maximum pressure is approximately 2 000 kPa Page 13 COLD PLATE FEA ANALYSIS In the 5 the copper cold plate is the device that exchanges heat from the heat source to the fluid refrigerant The cold plate is mounted directly to the heat source Thermal grease 15 applied between the cold plate and heat source to reduce the contact thermal resistance custom cold plate was designed by the team for the ICS The first characteristic of the cold plate that was determined was the overall size In the current Cool Cube the cold plate is large enough to remove the necessary amount of heat from the source but it is not large enough adequately display the phase change of the refrigerant Therefore the team decided to design a larger cold plate to use on the ICS The overall dimensions were selected to be 0 18 m 7 by 0 09 m 3 5 After the overall dimensions were selected the flow cavity was designed The flow cavity consists of an inlet and outlet orifice flow distributers and micro channels purpose of the flow dist
96. ssembly and Dimensions for Aluminum Frame of ICS SHELF 2 MAKE 6 ALL CONNECTIONS USE BANJOS Figure 25 Shelf 2 Assembly and Dimensions for Aluminum Frame of ICS HT E TT Figure 26 Assembled Frame of ICS 39 FIBERGLASS SHELL CONSTRUCTION Once the machined mold arrives at Hy Tec Fiberglass Inc facility the mold was gel coated to insure that cured fiberglass can be removed and the mold can be reused in the future Due to the size of the mold blowholes are drilled in key locations of the mold to assist the manufacturer when it the ICS fiberglass is removed The technic used was to blow compressed air inside the bottom of the mold to release the large flat surface that is prone to sticking Gel coating is shown in Figure Figure 27 Gel Coating of Fiberglass Mold Once the mold is gel coated it is then sanded and finished to a clean surface This helps the final outside finish of the fiberglass to ensure it is smooth as possible Finishing is shown in Figure 28 The red marks are the covered blowholes Figure 28 Mold Finishing The next step in the fiberglass process was to begin to lay key panels of thick fiberglass panels cut to mimic the contours of the mold See Figure 29 for fiberglass laying Page 40 Figure 29 Fiberglass Panel Then a thick layer of resin it painted over the top of the fiberglass panel in generous quantities to completely saturate the panel More a
97. ssumption that must be made in the 5 thermodynamic system 15 phase state of the entering refrigerant The R 134a refrigerant must be in a saturated liquid state or sub cooled when it enters the cold plate For simplicity it will be considered a saturated liquid In the problem statement as detailed in Report 1 from Fall 2010 stated that the refrigerant should be about 7096 vapor by mass as it leaves the cold plate Therefore it can be assumed the quality of the R 134a is 0 70 at the exit of the cold plate be assumed the refrigerant at the exit of the condenser will be a saturated liquid For simplicity it can be assumed that there are no heat losses or pressure drops in the copper pipes However there will be a significant pressure drop that occurs in the condenser It will be assumed that the pressure drop that occurs in the condenser Is 80 kPa which is typical in other Parker systems that are similar to the ICS Page 11 Since the 5 system will only operate indoors it can be assumed that the ambient temperature of the air is 22 It can be assumed that at steady state the air condenser ejects an identical amount of heat from the system as the heat source adds to the system which is a conservative assumption can also be assumed that all of the heat that is generated by the heat source is transferred through the cold plate and then into the refrigerant DETERMINING THE FLOW RATE OF IHE REFRIGERANT The first
98. t or the recovery tank Attach the male to male adapter to cne side of the oil injector Cap off one end of the oil injector with a plastic or brass cap Plastic is preferred Fill you oil injector with the desired amount of oil while keeping the capped off end pointing down Attach one end of the of the oil injector hose to the last remaining male to male adapter Attach the male to male adapter to cne side of the refrigerant flow valve Make sure the refrigerant flow valve is closed Attach the other end of the refrigerant flow valve to a seal rite connected to a hose Attach the other end of the hose with seal rite to the ICS charge port Now you can turn the capped off end of the oil injector in the upright direction This will allow the oil to move toward the refrigerant flow valve Remcve the plastic cap attached to the oil injector 5 male to male adapter Attach a seal rite to the male to male adapter of the oil injector Attach the other end of the hose with seal rite to the clean refrigerant tank Make sure the tank is still in the closed position Inspect the refrigerant tank to determine which side needs to be upright in order to fill the ICS with liquid Place the refrigerant tank in the correct upright position upon the scale Next you will open the refrigerant tank You will need to zero the scale in orcer to determine how much refrigerant has been removed from the tank and put into the ICS This will need to be done
99. tate identification the liquid can approximated as a saturated liquid While it may be slightly sub cooled in reality the saturated enthalpy value is still a good approximation of the actual enthalpy value for ICS operating conditions at 5 C sub cool the error will not exceed 796 Equation is a best fit quadratic expression of the saturated liquid enthalpy 4 of R134a in kJ kg at temperature This equation is used to find the enthalpy at states 1 3 and A 0 005 T444 1 1287 202 2 3 Since the 5 is not capable of measuring quality of the 2 phase state after the cold plate state 2 the heat added to the fluid over the cold plate must be known to calculate its enthalpy The heat added to the fluid is assumed to be the heat input by the heat source 0 It is important to note that this assumption is valid only when the system IS operating at steady state The enthalpy of the R134a refrigerant at state 2 can then be estimated as shown in Equation 4 4 Page 177 To determine the quality at the 2 phase state the liquid saturation enthalpy at that state must be known Equation 5 is used to find this saturation enthalpy where 72 15 the temperature at that state Asaturation 0 005 T2 1 1287 T 2022 5 The quality of the refrigerant at the exit of the cold plate can be approximated as the quantity of the difference between the enthalpy at this state and th
100. ted As soon as an Input Is incremented the new value is displayed on the screen and also written to the tag This information 1 sent to the DAG and the processor sends the right current voltage to the output module of the DAQ and the equipment connected to the corresponding channel on the output module acts accordingly 10 A discrete button was also added on the screen to start and stop the whole system in toggle mode After it is pressed then any changes to the input values will take effect After testing 15 completed in order to turn the heat and fan off one can either decrement the inputs to zero and or just press the start stop discrete button and it turns back to green waiting for the system to be started again The InteractX application B which was created for the final product was more complicated and more in depth for one it was comprised of many screens with some depending on the previous one and also with the ability to go back to previous screens The user manual which is provided in Appendix gives a detailed description on how to navigate these screens The first screen allows the user to start the system for the developed run time environment The start stop discrete button described earlier is present here After this button is pressed then an action button 1 used to advance to the next screen There is also an exit button on this screen to exit the run time environment but this only shows up when the system has been stopp
101. the dial slightly to the closed position until the noise goes away Wait until the gauge above the inlet dial starts to read in a pressure below Opsi or in micrameters Remcve the seal rite attached to the ICS Close the recovery tank You may now safely remove the hoses from both inlet and outlet of the recovery unit You have successfully recovered the ICS of refrigerant and oil Note At this point you may safely remove components from the ICS if you wish How to Vacuum the ICS prior to Adding Refrigerant 1 First you will need to get three things before you can start the vacuuming of the ICS You will need to get two hoses preferably hose clean of oil i e a hose rot used in the recovery 20 Page 174 or charging of the ICS You may want to use one standard length hose and one may be less than one foot a vacuum pump and a micrometer Attach seal rites to the end of the refrigerant hoses This is so you will not allow the vacuumed pressure to be released from the ICS after you have properly pulled a vacuum Attach one seal rite to the ICS charge port located in the center of the mechanical system Attach the other end of the same hose to one connection of the micrometer Attach one seal rite of the second hose to the other connection of the micrometer Attach the other end of this hose to the only connection on the vacuum pump Turn on the micrometer if it is digital If it a physical gauge then you will skip this step Tu
102. the exit of the cold plate was required to be no greater than 70 gas The maximum weight of the display was set at 500 lbs The display must fit through a common three foot by seven foot door exterior of the display is safe to the touch The ICS should dissipate heat loads from 100 to 1000 W The ICS will achieve a specified desired temperature to within 1 C of the user input The fiberglass display cabinet was designed and fabricated to match a computer aided design model A negative of this model was computer numerical controlled machined to create a mold to fabricate the fiberglass shell This was secured onto a custom designed aluminum frame that was fabricated at Parker condenser pump and fan were selected and purchased The cold plate cold plate housing pump cover and pump manifold were machined off site The fan to condenser shroud and piping assemblies were designed assembled and brazed at Parker testing rig was made in advance to hold all the two phase cooling cycle during final brazing and leak testing Electrical wiring consisted of distributing power from the power entry module to the 24VDC power supply the DAQ the PC the touch screen and the DIN A MITE heater controller Low voltage wiring had to be distributed from the 24VDC power supply to the pump fan and sensors The DAQ was initialized and programmed with the control algorithm and the graphical user interface GUI designed with Parker s
103. tion is a question that tries to figure out how to do a particular function on the GUI PROCEDURE During a trial run 0 3 questions means easy 4 6 questions means medium and 7 10 questions means difficult Five students were randomly selected to interact with the ICS and Figure 88 below shows the result of the test RESULTS As the figure shows only one person out of the five asked more than one use question This shows that the GUI 15 easy to use Ease of Use Test Results 52 OO OO Number of GUI Use Questions Drew Cameron Christina Eric Isaiah Tested Volunteers Figure 88 Ease of Use Test Results Chart 82 CONTROL DESIGN MODIFICATION A problem was discovered during testing of the control system that required an innovative solution to fix The team noticed that when the user was going from one trial at a higher heat input and higher desired temperature to another trial with a lower heat input and lower desired temperature that the system would become temporarily unstable This instability was a result of the refrigerant leaving the two phase region and vaporizing See the transient response of the ICS going from 900 W at 80 to 600 W at 55 C before the control system modification in Figure 89 Note the instability that occurred when the two phase refrigerant composition was lost Transient Response Before Control Modification 7 Desired Temperature Actu
104. ure occurs whichever is first Stop data logging 8 Turn off power to the ICS S cs RESULTS OF B The test results showed that the highest temperature that electronics can get up to in our system is approximately 90 at this point the safety feature kicks in and turns off the heat and then the temperature starts to drop See results charted in Figure 74 O 5 U 60 80 100 120 Time seconds Figure 74 Extremes Test 1000 W WORKING CONDITIONS TEST The objective of this test is to generate a plot of the fan control voltage versus temperature at the cold plate for each heat load Using thermodynamic theories the pump speed that corresponds with an exit refrigerant quality of 0 7 7096 vapor was found for each heat load and the pump was manually set to this value Lines were fit through the derived curves and the equations for fan steady state speed voltage at each heat load were extracted Page 70 PROCEDURE 1 Supply power to the ICS 2 Set heat load to 1000 W 3 Set the fan to 10 V 100 4 Wait 5 minutes or until the system reaches steady state or goes over 80 degrees Record the heat input fan speed pump speed and the temperature at the cold plate Reduce the fan speed by 1 10 Repeat steps 4 6 until a fan voltage of 1 V has been tested Reduce the heat input by 100 W Repeat steps 3 7 until each heat Load has be
105. would be used in the ICS It is known from previous research at Parker that the actual thermal resistance Is similar to the results obtained when using a convective heat transfer coefficient of 1 500 W m K therefore this coefficient was used in other simulations and calculations Using the ANSYS tools a temperature distribution was mapped in the cold plate with 30 channels Figure 2 displays the temperature distribution throughout a cold plate when it is supplied with 1000 W of heat 0 070 m Figure 2 Temperature Distribution through Cold Plate with 1000 W Heat Supply of 50 Page 15 After selecting a cold plate with 30 channels the team decided to analyze the pressure drop of the refrigerant throughout the device ANSYS was used to perform this analysis Figure 3 displays the results of the pressure distribution throughout the cold plate The fluid used in the simulation was water that had a flow rate equal to the systems flow rate when 1000 W of heat is supplied results of the experiment ylelded a maximum pressure drop of about 310 Pa which should be similar to the maximum pressure drop of the actual system This pressure drop is adequate for use in the ICS system Figure 3 Pressure Distribution through Cold Plate See Appendix 1 for Cold Plate Dimensioned Drawing GLASS COVERING The sight glass covering for the cold plate is made of tempered borosilicate glass thickness of glass for the

Download Pdf Manuals

Related Search

Related Contents

  Ewent EW1272 holder  SERVICE MANUAL - Pdfstream.manualsonline.com  Microdata User Guide National Survey of Giving, Volunteering and  245U-E wireless Ethernet modem user manual  5-axis CNC Breakout board User Manual  

Copyright © All rights reserved. Failed to retrieve file