9543103a2878dfb8ffff83b7ffffd524

Please download to get full document.

View again

of 6
7 views
PDF
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Strategies for Minimum Energy Operation for Precision Machining Nancy Diaz , Moneer Helu , Andrew Jarvis , Stefan Tönissen , David Dornfeld , Ralf Schlosser 1 Laboratory for Manufacturing and Sustainability, University of California, Berkeley 2 WZL, RWTH Aachen Abstract The development of green machine tools will require novel approaches for design, production and operation for energy savings and reduced environmental impact. We describe here work on three projects: i. influence of process par
Document Share
Documents Related
Document Tags
Document Transcript
  © 2009 The Proceedings of MTTRF 2009 Annual Meeting  Strategies for Minimum Energy Operation for Precision Machining Nancy Diaz 1 , Moneer Helu 1 , Andrew Jarvis 1 , Stefan Tönissen 1 , David Dornfeld 1 , Ralf Schlosser  2   1 Laboratory for Manufacturing and Sustainability, University of California, Berkeley 2 WZL, RWTH Aachen Abstract The development of green machine tools will require novel approaches for design, production and operation for energy savings and reduced environmental impact. We describe here work on three projects: i. influence of processparameters on power consumption of end-milling using force and process time models with experimental verification.Process parameters are chosen to minimize process time since power consumed by a machine tool is essentiallyindependent of the load and energy per unit manufactured decreases with process time; ii. KERS (kinetic energyrecovery system) for machine design and modeling the integration of a recovery system into a machine tool tocalculate the amount of energy that could be recovered, and whether the environmental benefits are significant; andiii. evaluation of interoperability solutions, such as MTConnect, as tools enabling a standardized plug-and-play platform to integrate sensors with a unified monitoring scheme to achieve improved energy performance.   Keywords: Power consumption of machine tools, KERS, Interoperability   1 INTRODUCTION Manufacturing processes carried out on machine tools areenergy intensive. As machine tools have become moreadvanced, their degree of automation has risen by addingcomponents such as tool change mechanisms or additionalaxes. Given the general trend of increasing power demandof machine tools the cost that companies have to expendon electrical energy will rise in the future. Furthermore, theexternal costs on the environment rise, since currently themajority of electrical power is obtained from burning fossilresources. A foreseeable shortage of fossil resources and agrowing demand to include the external cost of environmental damage in product prices are likely toincrease the cost of electrical energy for companies evenfurther. Therefore, in order to maintain competitiveness andlower costs, companies have to identify ways to decreasethe energy consumed during manufacturing for a givenproduct.Figure 1 shows four approaches to modifying a product life-cycle ranked by the potential impact on a given objectiveand the depth of analysis of the existing state. This paper applies the idea of product life-cycle modification to theobjective of lowering the energy demand of machine toolsfor a given product. Section 2 summarizes researchconducted in which process parameters of end-milling arevaried (level 3). Section 3 discusses the potential of installation of kinetic energy recovery systems (KERS) inmachine tools. The installation of such devices may beconsidered a level 1 process redesign of a product life-cycle. Section 4 concludes with a general evaluation of energy monitoring and interoperability standards. Level 1Productand ProcessDesignLevel 2Productand ProcessPlanningLevel 3ProcessParameter SelectionandOptimizationLevel 4Post-Processing    D  e  s   i  g  n  o   f   N  e  w   P  r  o  c  e  s  s  e  s   /   S  y  s   t  e  m  s   A  n  a   l  y  s   i  s  o   f   E  x   i  s   t   i  n  g   P  r  o  c  e  s  s  e  s     Level 1Productand ProcessDesignLevel 2Productand ProcessPlanningLevel 3ProcessParameter SelectionandOptimizationLevel 4Post-Processing    D  e  s   i  g  n  o   f   N  e  w   P  r  o  c  e  s  s  e  s   /   S  y  s   t  e  m  s   A  n  a   l  y  s   i  s  o   f   E  x   i  s   t   i  n  g   P  r  o  c  e  s  s  e  s   Level 1Productand ProcessDesignLevel 2Productand ProcessPlanningLevel 3ProcessParameter SelectionandOptimizationLevel 4Post-Processing    D  e  s   i  g  n  o   f   N  e  w   P  r  o  c  e  s  s  e  s   /   S  y  s   t  e  m  s   A  n  a   l  y  s   i  s  o   f   E  x   i  s   t   i  n  g   P  r  o  c  e  s  s  e  s   Figure 1: Strategies of green manufacturing 2 PROCESS PARAMETER OPTIMIZATION2.1 Theory The research goal of this project is to analyze the impact of process parameter selection on the energy consumptionper part manufactured for an exemplary manufacturingprocess. End-milling was selected because of its wide usein the industry.The energy per unit manufactured is determined by boththe power demand of the machine tool during machiningand the processing time (see Figure 2). Energyper unitFeedrateSpindleRPMFeed per toothNumberof flutesTime per unitPower demand   Energyper unit   FeedrateSpindleRPMFeed per toothNumberof flutesTime per unitPower demandEnergyper unit   FeedrateSpindleRPMFeed per toothNumberof flutesTime per unitPower demandTime per unitPower demand   Figure 2: Influence of process parameters on the energyper unit manufacturedThe power demand of a machine tool may be divided into aconstant and a variable component [1]. The constant power can be attributed to the computer, fans, lighting, etc. of themachine tool. This component of the total power demand isindependent of process parameter selection. The variablepower demand, though, is dependent on processparameter selection and can be attributed to the spindle or the drives of the table axes.The processing time per unit manufactured is determinedby the feed rate. Cutting conditions at a specific feed ratedepend on the selection of the number of revolutions per minute of the spindle, the feed per tooth and the number of flutes.Figure 3 identifies two opposing effects on the energy per unit manufactured. First, as the feed rate increases theprocessing time is reduced. Therefore, the contribution of the constant power demand of the machine tool to theenergy per unit manufactured decreases. Second, anincrease in the feed rate demands more power from themachine tool with or without adjustment of the cuttingspeed. Depending on which effect prevails, three differentmachining regions may be found.  © 2009 The Proceedings of MTTRF 2009 Annual Meeting    e  n  e  r  g  y  p  e  r  u  n   i   t  m  a  n  u   f  a  c   t  u  r  e   d feedrateEnergy consumptionof components withconstant power demandEnergy consumptionof components withvariable power demandTotal energyRegion 2Region 1Region 3       e  n  e  r  g  y  p  e  r  u  n   i   t  m  a  n  u   f  a  c   t  u  r  e   d feedrateEnergy consumptionof components withconstant power demandEnergy consumptionof components withvariable power demandTotal energyRegion 2Region 1Region 3   Figure 3: Regions of machining processThe sum of both energy contributions results in a parabolic-total energy plot also shown in Figure 3. In Region 1 thedecrease due to a shorter processing time dominates theincreased variable power demand. In this region the feedrate will be chosen as fast as technically possible. InRegion 2 the energy per unit manufactured is fairlyconstant, whereas the increase of the variable power demand dominates in Region 3. If the process is located inRegion 3, slower feed rates would lead to lower energy per unit manufactured. 2.2 Results Experimental studies were conducted to determine theregion of the machining process in Figure 3. Initialexperiments (i.) kept the feed per tooth constant byincreasing the spindle speed proportionally to the feed rateand (ii.) varied the feed per tooth at a constant spindlespeed. Subsequently, the energy per unit manufactured of conventional cutting with a 2-flute uncoated carbide end-mill was compared to the energy per unit manufactured of high speed cutting with a 2- and 4-flute TiN-coated end-mill.Slot cutting experiments of a low carbon steel (AISI 1030)were conducted on a Mori Seiki NV1500DCG with an 8mmuncoated carbide end-mill and a depth of cut of 2mm to beconsistent with   [2]. For each process parameter combination, 52 slot cuts were performed in order to studythe wear of the tool. The power demand of the machine toolwas recorded using a WattNode Modbus power meter viaan MTConnect monitoring system (see Section 4). 2.2.1 Constant feed per tooth The initial cutting conditions used a recommended feed per tooth of 0.125mm/tooth and a spindle speed of 800rpm togenerate a feed rate of 200mm/min. Figure 4 shows thatthe energy consumed by the machine tool per unitmanufactured decreases over feed rate at a constant feedper tooth of 0.125mm/tooth. However, tool wear increasessignificantly over feed rate. The tool consistently brokebefore having cut 52 slots at a feed rate of 500mm/min. Parameter  :DOC = 2mmWOC = 8mmf  t = 0.125mm/tooth Material: AISI 1030 Tool: UncoatedCarbideD = 8mmz = 2 05101520253035404550200250300350400450500 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   K   J   ]   Parameter  :DOC = 2mmWOC = 8mmf  t = 0.125mm/tooth Material: AISI 1030 Tool: UncoatedCarbideD = 8mmz = 2 05101520253035404550200250300350400450500 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   K   J   ]  Figure 4: Average energy per unit manufactured versusfeed rate 2.2.2 Constant spindle speed The feed per tooth was varied between 0.025mm/tooth(feed rate of 40mm/min) and 0.15mm/tooth (feed rate of 240mm/min). Preliminary studies showed that increasingthe feed per tooth beyond 0.15mm/tooth results in toolbreakage.Figure 5 shows that by lowering the feed per tooth(analogous to lowering the feed rate) the energy per unitmanufactured increases. However, surface qualityimproves creating a trade-off between surface quality andenergy consumption during machining. Furthermore, thetool wear effectively decreases as the feed rate is reducedsince the load on the tool is smaller. Parameter  :DOC = 2mmWOC = 8mmn spindle = 800min -1 Material: AISI 1030 Tool: UncoatedCarbideD = 8mmz = 2 0204060801001201401604080120180240 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   K   J   ]     Parameter  :DOC = 2mmWOC = 8mmn spindle = 800min -1 Material: AISI 1030 Tool: UncoatedCarbideD = 8mmz = 2 0204060801001201401604080120180240 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   K   J   ]  Figure 5: Average energy per unit manufactured versusfeed rate at constant spindle speed 2.2.3 Conventional versus high-speed machining High-speed cutting with coated end-mills involves greater cutting speeds at a lower feed per tooth. The feed rateincreases since the increase in cutting speed is greater than the decrease in the feed per tooth compared to cuttingat conventional speeds. In this study a 4-flute TiN coatedend-mill was used and the energy consumed wascompared to the 2-flute uncoated end-mill.  © 2009 The Proceedings of MTTRF 2009 Annual Meeting  Table 1 summarizes how the number of flutes ( z  ), spindlespeed ( n spindle ), feed per tooth ( f  t  ), and feed rate ( v  f  ) werevaried across the three experiments.Table 1: Summary of conventional versus high-speedcutting parameters Units Conv. High-Speed Coating  - - TiN z  - 2 4 n spindle   rev/min 800 7334 f  t    mm/tooth 0.125 0.033 v  f    mm/min 200 968 t  cut  s 28.8 6.0Figure 6 shows that there is a dramatic decrease in energyas the cutting parameters change from conventionalspeeds to high-speed values. This decrease in energy per unit manufactured is primarily a result of the decrease inprocessing time. 05101520253035404550200484968 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   k   J   ] Parameter  :DOC = 2mmWOC = 8mmLength= 100mm Material: AISI 1030 Tools: D = 8mm z = 2,t process = 28.8sn spindle = 800min -1 f  t = 0.125mm/toothz = 2,t process = 12.1sn spindle = 7334min -1 f  t = 0.033mm/toothCoatedCarbidez = 4, t process = 6.0sn spindle = 7334min -1 f  t = 0.033mm/tooth05101520253035404550200484968   Parameter  :DOC = 2mmWOC = 8mmLength= 100mm Material: AISI 1030 Tools: D = 8mm 05101520253035404550200484968 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   k   J   ] 05101520253035404550200484968 Feed rate [mm/min]    E  n  e  r  g  y  p  e  r  u  n   i   t   [   k   J   ]   Parameter  :DOC = 2mmWOC = 8mmLength= 100mm Material: AISI 1030 Tools: D = 8mm z = 2,t process = 28.8sn spindle = 800min -1 f  t = 0.125mm/toothz = 2,t process = 12.1sn spindle = 7334min -1 f  t = 0.033mm/toothCoatedCarbidez = 4, t process = 6.0sn spindle = 7334min -1 f  t = 0.033mm/tooth0510152025303540455020048496805101520253035404550200484968   Parameter  :DOC = 2mmWOC = 8mmLength= 100mm Material: AISI 1030 Tools: D = 8mm  Figure 6: Energy comparison of conventional cutting versushigh-speed cuttingEven though the feed rate during high-speed cutting wasincreased, tool wear reduces significantly after cutting 52slots compared to the conventional cut (see Figure 7). UncoatedCarbide -z = 2n spindle = 800rev/minv f  = 200 mm/minf  t = 0.125mm/toothCoatedCarbide -z = 4n spindle = 7334rev/minv f  = 968 mm/minf  t = 0.033mm/toothCoatedCarbide -z = 2n spindle = 7334rev/minv f  = 484mm/minf  t = 0.033mm/tooth   CoatedCarbide -z = 2n spindle = 7334rev/minv f  = 484mm/minf  t = 0.033mm/toothUncoatedCarbide -z = 2n spindle = 800rev/minv f  = 200 mm/minf  t = 0.125mm/toothCoatedCarbide -z = 4n spindle = 7334rev/minv f  = 968 mm/minf  t = 0.033mm/tooth   CoatedCarbide -z = 2n spindle = 7334rev/minv f  = 484mm/minf  t = 0.033mm/tooth   CoatedCarbide -z = 2n spindle = 7334rev/minv f  = 484mm/minf  t = 0.033mm/tooth  Figure 7: Tool wear comparison after 52 slots 3 ENERGY RECOVERY IN MACHINE TOOLS3.1 Background and Method Aside from taking into consideration process parameter selection as a method of reducing energy consumption in amachine tool, the recovery of energy through the use of akinetic energy recovery system (KERS) would be a possiblemethod of improving production efficiency within themanufacturing industry.By creating a computer model of the NV1500DCG’s spindlemotor and table, evaluations of the possible reductions inpower usage were made, as well as the cost benefits sucha system could bring. These components were chosen tomodel since they possess varying levels of kinetic energyduring workpiece manufacture. The use of experimental  © 2009 The Proceedings of MTTRF 2009 Annual Meeting  results obtained from the NV1500DCG for the power requirements to drive the spindle at various angular velocities, allowed the computer model results to be refinedto better match the behaviour of the actual machine tool.Figure 8: Comparison of experimental and model dataThree possibilities where energy recovery could occur wereinvestigated; when decelerating the spindle to stationary,during air cutting and when decelerating the table mass tostationary. Under optimum cutting conditions it was foundthat approximately one million deceleration events of thetable would be required to recover the same amount of energy that one spindle deceleration event would recover.For this reason, the installation of KERS onto the tabledrives was dismissed as requiring more effort andresources than the potential benefits.Investigation of the reduction in energy needed during air cutting using the NV1500DCG’s spindle motor (Fanuc α B80S/20000i) concluded that keeping the spindle speedconstant would be the most efficient strategy under allcutting speed and air cutting duration conditions. However,models of other motors did show a reduction in energyusage, suggesting that energy recovery during air cuttingfor some spindles may increase machine efficiency.Therefore spindle deceleration to stationary was found torecover the largest amounts of energy in the NV1500DCG.A system was modeled that stored the recovered energy insupercapacitors. This form of energy storage was usedmainly for its power density and longer life expectancy thansecondary batteries. A supercapacitor bank was thendefined that could store the energy from one decelerationevent from 20000rpm using a maximum voltage of 1kV anda charge / discharge efficiency of 90%.An environmental evaluation of multiple workpieces wascarried out through the use of a Monte Carlo simulationwhere tool size selection was based on a normaldistribution having a mean cutter diameter of 5mm, and astandard deviation of 2mm in increments of 0.5mm. Thesevalues were chosen to simulate high speed machining. 75parts at each different combination of time to machine apart (varied from two to five minutes) and tools used per part (varied from two to five) where simulated in order todetermine trends and magnitudes of expected power savings. 75 parts were simulated and averaged at eachcombination to lessen the effect that the pseudorandomnature in which the tools were selected would have onresults. 3.2 Results A supercapacitor bank made up of 400 350Fsupercapacitors connected in series, for a cost of $7200was defined [3]. The overall recovery efficiency of thesystem, as defined by equation 1, was found to be roughly74% for most angular velocities. (1) Spindle Of EnergyKinetic Total Motor  ToBack SuppliedEnergyTotal cov = eryre !   0 0.5 1 1.5 2x 10 4 01020304050607080Angular Velocity (rpm)Efficiency Energy Recovery Efficiency Versus Angular Velocity  Figure 9: Recovery efficiencyThe results for the environmental analysis show that power savings between 5 and 25% for the whole machine couldbe expected with a KERS under the simulated conditions(see Figure 10)Figure 10: Expected power savingsA simulation of a single part that used three tools (5mm,2.5mm and 4mm) and had a part time of 2 minutes wascompleted to evaluate the potential power savings and costreduction. A power saving of 20.41% was determined witha decrease in energy of 49.6kJ per part. A full breakdown of power usage within the NV1500DCG with and withoutKERS is shown in Figure 11.
Similar documents
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks