Self-Powered Wireless Vibration-Sensing System for Machining
Monitoring
Tien-Kan Chung*
a, Hao Lee
a, Chia-Yung Tseng
a, Wen-Tuan Lo
a, Chieh-Min Wang
a,
Wen-Chin Wang
b, Chi-Jen Tu
b, Pei-Yuan Tasi
b, and Jui-Wen Chang
ba
Department of Mechanical Engineering, National Chiao Tung University, Hsinchu 30010, Taiwan.
bPrecision Machinery Research and Development Center, Taichung 40768, Taiwan.
ABSTRACT
In this paper, we demonstrate an attachable energy-harvester-powered wireless vibration-sensing module for milling-process monitoring. The system consists of an electromagnetic energy harvester, MEMS accelerometer, and wireless module. The harvester consisting of an inductance and magnets utilizes the electromagnetic-induction approach to harvest the mechanical energy from the milling process and subsequently convert the mechanical energy to an electrical energy. Furthermore, through an energy-storage/rectification circuit, the harvested energy is capable of steadily powering both the accelerometer and wireless module. Through integrating the harvester, accelerometer, and wireless module, a self-powered wireless vibration-sensing system is achieved. The test result of the system monitoring the milling process shows the system successfully senses the vibration produced from the milling and subsequently transmits the vibration signals to the terminal computer. Through analyzing the vibration data received by the terminal computer, we establish a criterion for reconstructing the status, condition, and operating-sequence of the milling process. The reconstructed status precisely matches the real status of the milling process. That is, the system is capable of demonstrating a real-time monitoring of the milling process.
Keywords: energy harvester, electromagnetic, power generator, vibration, machining monitoring, self-powered, wireless sensor
1. INTRODUCTION
Recently in the society of wireless sensors network, researchers demonstrated an innovative wireless sensing/monitoring application for machining monitoring [1]. In the application, the wireless sensing system is used to sense, analyze, and monitor the temperature of the milling cutter in the machining process. This real-time monitoring prevents the machining failure caused by over-heating in the process and subsequently enhances the reliability of the machining tool. However, the maintenance becomes difficult due to the battery replacement issue in the case of numerous sensors used in a sensor-network. To address this issue, an energy-harvester-powered (i.e., toward self-powered) wireless sensing system is a preferred solution. Nowadays, solar cells are comprehensively used as the energy harvesters for powering the wireless sensing system. However, solar-cell-powered wireless sensing system is not suitable for indoor applications, especially inappropriate for monitoring the machines in a factory. Therefore, researchers use vibrational energy harvesters [2] (such as piezoelectric, electromagnetic, and electrostatic energy harvesters) instead of the solar cells to generate power for the indoor applications. However, the circuit for piezoelectric energy harvesters is more difficult than electromagnetic energy harvesters. Thus, electromagnetic energy harvesters are comprehensively utilized by the researchers to power the wireless sensing systems for the indoor applications. That is, more appropriate for monitoring the machines in a factory. More recently, researchers demonstrated an electromagnetic energy harvester powering a wireless sensing system for health monitoring of a spindle [3]. However, both the harvester and wireless sensing system have to be embedded into the spindle of the machine. Consequently, the spindle must be modified (i.e., customized/non-standard spindle). This causes serious issues in design and manufacturing of the customized/non-standard spindle to both spindle- and machine- manufacturers. More seriously, the customized/non-standard spindle may not be fully compatible with the existing machines. This results serious problems in machining. Therefore, an attachable (i.e., nondestructive for the spindle) self-powered wireless sensing system is needed. Hence, in this paper, we demonstrate an attachable energy-harvester-powered wireless vibration-sensing system for machining monitoring.
Storage and !ctified Circui 1 Wireless TrJnsmittir
Milling
er Vibra Wirel 1g 2omputer MillingC 1 viprat ition Sensor less Module ut te r's :ion F ] Self- powere Sensing .d Wireless System MilliegCu Rotatio 111 I I I I LRE I lcr's ergy Harvest The conceptu a milling-pro accelerometer on the machin rotate with th produced betw to the inducta inductance. T the milling c electrical ene in order to st vibration sen The system is After the term to indicate theAccording to the test setup) assembled as devices we fa [shown in fig microcontroll fabricated by attached on th combined wi system is Ana the machinin vibration gen ual approach o cess is shown r, and wireles ne while sever he spindle w ween the rota ance. Due to th That is, a perio cutter’s rotati ergy. Furtherm teadily power sing system. M s able to sens minal compute e actual status Figure the conceptua ) is shown in s the energy-h abricated for gure 3(a)], ME ler/RF-antenn utilizing an in he spindle. Th ith voltage re alogy Device, ng process. Th nerated from th of the attachab n in figure 1. T ss module [4-7 ral permanent while the indu ated magnets a he electromag odic alternatin ion during th more, through r both the ME Meanwhile, v e the vibratio er received the s/condition/op 1. The concept 3 al approach in figure 2. Acco harvester-powe the system is EMS accelero na transmitter) nductance wit he energy-stor gulator, as sh , AXDL 330. he accelerome he cutting too 2 ble energy-ha The self-powe 7]. The harves t magnets are uctance is fix and fixed indu gnetic-inductio ng power outp he milling pro an energy-sto EMS accelerom ibration occur n and transmi e vibration da perating-seque tual approach o 3. FABRIC n figure 1, the ording to the p ered (i.e., self
shown in fig meter (i.e., vi ) [shown in th th NdFeB mag rage/rectificat hown in the l This accelero eter is glued t ol while millin 2. DESIGN arvester-power ered system co ster consists o attached on th xed on the ma uctance, i.e., on, the period put is generat ocess and su orage/rectifica meter and wi rs while the m it the vibratio ata, the data ar ence of the mi of the self-powe CATION AN e correspondin practical appr f-powered) w gure 3. The sy ibration senso he right side o gnets. The ind tion circuit of left side of fig ometer is sensi to the over ar ng. The wirel N red wireless v onsists of an e of a magnet an he spindle of t achine. Due the magnets p dic/varied mag ted. This achi ubsequently co ation circuit, th ireless module milling cutter on signal to th re analyzed th lling. That is,
ered wireless vib
ND TESTING
ng practical ap roach, the core
ireless vibrati ystem consists or) [shown in f figure 3(c)]. ductance is fix the energy ha gure 3(c). Th itive to low-fr rm of the mil ess module ut vibration-sens electromagneti nd inductance. the machine. D to this, a con produces a pe gnetic flux ind eves harvestin onverting the he harvested e e. This achiev and work piec e terminal use hrough a criter a real-time-m bration-sensing G pproach (also r e devices are f ion-sensing sy s of an electro figure 3(b)], a The electrom xed on the mac arvester is mo he MEMS acc requency vibra lling machine tilizes a low-p sing system us tic energy harv . The inductan During milling ntinuous relat eriodic/varied duces a voltag ng mechanica e mechanical electrical ener ves a self-pow ce are physica er (i.e., termin rion we establ monitoring of t g system referred as an fabricated and ystem. A phot omagnetic ene and wireless m magnetic energ chine while th odified from a celerometer w ation frequent e in order to e power-consum sed to monito vester, MEMS nce is attached g, the magnets tive-motion is magnetic flux e output in the al energy from energy to an rgy is rectified wered wireless ally contacted nal computer) lished in orde the milling. n illustration o d subsequently tograph of the ergy harvester module (i.e., a gy harvester is he magnets are bridge-circui we used in the tly occurred in experience the mption ZigBee r S d s s x e m n d s d. ). r f y e r a s e it e n e e
(a) rott Inductor
l
>rk PieceTable
(b) AcceleYJfYIE'`LE Y meter Signal Amplifier tectifiedI I N Circuit Il I\r
lo: _ :_ (c) En-E Magnet Milli Cutt(y)
Wireless Module I L Signal Amplifier 2rgy 'Rectified cuit Over Arr )v \Ill'..'
communicatio transceiver fo computer). A order to distin before transm the swarf gen module, an en The illustratio figure 4(a) is spindle is rot side view of selected as th Figure 2. milling pr test setup) Figure 3. storage/re on-protocol [ or wireless d A Microchip P nguish the sen mitted to the w nerated from m nergy-harvest on and photog s chosen to b tated with a rothe thick plat he machining t The correspo rocess through s ). The photograp ectified circuit. [8-9] based w data-communic IC18F2520 is nsor’s signal t wireless modu milling. Finall er powered w graph of the te e monitored. otational spee te is shown in tool shown in onding practica sensing/analyzi ph of the (a) e wireless trans cation (i.e., tr s used in the m to noise, the o
le. The wirele ly, through in wireless vibrati esting setup is The testing c ed of 1650 rpm n figure 4(b) figure 4(d). al approach of ing the milling
energy harveste
sceiver (Pixie ransmitting th module for pro output signal o ess module is ntegrating the ion-sensing sy s shown in fig condition we m. The work and 4(c), resp f the self-powe cutter’s vibratio er, (b) MEMS e Lite, Flexip he vibration s ocessing signa of the sensor i
set beside the energy harve ystem is achiev ure 2 and 4, r decided in th piece is a SK pectively. An ered wireless v on (the approac accelerometer panel) with C signal from th als and contro s amplified by e column of th
ster, MEMS a ved and ready espectively. A his paper is d KD-11 tool-ste 8-mm tungste ibration-sensing ch is also referre r, and (c) wirel CC2420 ZigB the module to olling the RF t y an instrume he milling ma accelerometer y for testing. A milling mac described as f eel thick plate ten carbide m ng system mon ed as an illustra eless module w Bee-ready RF o the termina transceiver. In ented amplifie chine to avoid r, and wireless chine shown in following. The e. The top and illing cutter is nitoring the ation of the with energy F al n r d s n e d s
(a) reless ModulE UIagc/I\Cl,lliic -larvester & Energy (b) (d)
5 cm
(c) 1 cm5 cm
Vibration Wii c+ JL Ç Sensor Energy I Figure 4. the millinThe test resu steel thick pl represents a r of the figure) (that is, the b output voltag experiences th the accelerom Because we seconds, ther received by th correlation b received by t oscillated vol produced by of the produc frequency of out of the ran vibration occu Through anal with different time monitor flow chart of received whi voltage-ampl result in erro capable of in monitoring, th amplitude dat is “Stopped” data are rangi
The photograp g cutter.
lt of the self-p late milled by real operating-). The red das base line is co
ge signals of he vibration p meter are tran set the wirele e are total 56 he terminal co etween the o the terminal c ltage signals w the milling do ced output volt 2500 Hz to re nge of 1.0 V-
urred in the m lyzing the vib t operating-se ring). For the f the criterion le milling ha itude data in f r. As we men ndicating the he actual mill ta). That is, th if the termina ing from 1.0V
h of (a) the tes
powered wire y the 8-mm -sequence of t sh-line repres orresponded to f the accelero produced by m nsmitted from ess module to data received omputer are f operating-sequ computer is o with a magnit oes oscillate t tage signals a ecord/analyze 1.3 V. That is milling. That is bration data w equence, millin milling proce is shown in f ve no signific figure 5 as a c ntioned in per vibration occ ling sequence/ he self-powere al computer ha V to 1.3 V. Du ting setup, (b) 4. RESUL eless vibration tungsten carb the milling (i. sents the base o the bias-volt ometer are os milling. Subseq m the system o send signal d by the termin fitted by the b uence of the obvious. Whil tude either hig the accelerom re far apart fro e the vibration
s, most voltag s, the system s e obtained, w ng status, and ess with a typ figure 6(a). B cant differenc criterion to dir rvious section curred in mil /status is indic ed wireless vi ad received tw ue to this, the front view of th LTS AND DI n-sensing syst bide cutter) is .e., idle, millin e line of the s
tage for activ scillated abov quently, the o to the termin ls from the s nal computer black curve in milling and le the cutter m gher than 1.3V meter resulting om the base li n signals. The ge signals (at successfully s we establish a d milling cond pical operating Based on the e ce with those rectly indicate n, the oscillate lling status. T cated by the c ibration-sensin wo continuous probability o he work piece, ISCUSSION tem monitorin s shown in fi ng, and then id sensing signal
ating the acce ve and below output voltage nal computer ystem to the during the mi the figure. A the output vo mills the work V or lower tha g in oscillated ine. Therefore analyzed resu least 70% of enses the vibr criterion capa ditions (note: t g-sequence (i. experimental d received wh e/distinguish th ed output volt To reduce the riterion with a ng system det data which th f false-detecti (c) side view o N ng the milling igure 5. In fi dle; the status of the capac elerometer). D
the base lin signals (i.e., v by the wirel terminal com illing lasting f ccording to th oltage signals k piece, the t an 1.0 V. This output voltag e, we use an os ult shows at le the signals) ar ration occurred able of reconst this is extreme .e., whether m data shown in ile idling. Th he real-time m tage signals h e chance of f a consequent d termines the s he amplitude o ion is significa
of the work pie
g process (the igure 5, the b s is referred to citive MEMS Due to this bia ne while the vibration data less module o mputer once p for 110 secon he data, line, s (indicating terminal comp s result shows ge signals. Ho scilloscope w east 70 % of t are capable of d in the millin tructing the m ely important milling the wo n figure 5, som hus, in this ca milling status/ have at least 7 false-indicatin data set (more status of the m of the voltage antly decrease ece, and (d) SKD-11 tool blue solid-line o the right axis accelerometer as-voltage, the acceleromete a) produced by of the system per every two
ds. These data and curve, the the vibration puter received s the vibration owever, not al with a sampling
the signals are f indicating the ng. milling process toward a real ork-piece), the me of the data ase, using one condition may 70% of chance ng/detection in e than just one milling process signal of both ed to 10%. On -e s r e r y m. o a e ) d n ll g e e s -e a e y e n e s h n
(a) tri -d 1.3V? o s Voltage 1.3V? No
.
1
Consider Milling P Proceeding i r Tinse uSer (b) IsCï_'S`_; - - Baseline-U- Vibration Sigr
-Operating Sei J ial luence of Milling RE Re con: .al ;tructed Receive a VoIta@ 1 1.0 V, < Voltage < Yes
Check the Previou
Data V < Voltage < Yes V Consider Milling Stopped the other hand
system determ criterion, we comparison o capable of de Figure 5. cutter and Figure 6. The comp d, because the mines the stat e successfully of the reconstr emonstrating a
The test result d work piece is t
(a) The flow ch parison of the re
e data far awa tus of the mill y reconstruct
ructed and re a real-time mo
t of the self-po the 8-mm tungs
hart of the crite eal and reconstr
y from the ran ling process a the milling al milling sta onitoring of th
owered wireless sten carbide cut
erion for indicat ructed (criterion nge of 1.0 V -as “Proceeding status/condit atus/condition/ he milling proc s vibration-sens tter and SKD-1
ting the status, n-based) milling - 1.3 V only o g” after receiv tion/operating /operating-seq cess. sing system mo 1 tool-steel thic condition, and o g status, conditi ccurred while ving this sort g-sequence sh quence [figure
onitoring a mill ck plate, respec
operating-seque ion, and operati
e milling the w of data once. hown in figu e 6(b)] shows lling process. T ctively.
ence of the mill ing-sequence.
work piece, the . Based on the ure 6(b). The the system is The milling ling. (b) e e e s
5. CONCLUSION
We have successfully demonstrated the energy-harvester-powered (i.e., self-powered) wireless vibration-sensing system for read-time monitoring of a milling-process. The electromagnetic energy harvester harvests the mechanical energy from the milling and subsequently converts the mechanical energy to electrical energy to power the MEMS accelerometer and wireless module. The result shows the system successfully senses the vibration occurred in the milling and subsequently transmits the vibration data (voltage signals) to the terminal computer. Furthermore, based on the criterion we established through analyzing the vibration data received by the terminal computer, the operating-sequence of the milling machine is successfully reconstructed. This achieves a real-time monitoring of the milling process toward an intelligent machining-monitoring system. In the future, we will investigate and develop more advanced smart-machining functions for the self-powered sensing system.
ACKNOWLEDGMENT
Support for this work was obtained from the Taiwan National Science Council (NSC Grant No.101-2625-M-009-008 and NSC Grant No. 101-2221-E-009-022).
REFERENCES
[1] Wright, P. K., Dornfeld, D. A., Hillaire, R. G. and Ota, N. K., "A wireless sensor for tool temperature measurement and its integration within a manufacturing system," Transactions of the North American Manufacturing Research Institute of SME. 63-70 (2006).
[2] Chung, T. K., Lee, D. G., Ujihara, M. and Carman, G. P., “Design, simulation, and fabrication of a novel vibration-based magnetic energy harvesting device,” Transducers '07 & Eurosensors XXI, Digest of Technical Papers 1, U438-U439 (2007).
[3] Chang, L. C. and Lee, D. S., "The development of a monitoring system using a wireless and powerless sensing node deployed inside a spindle," Sensors 12(1), 24-41 (2012).
[4] Beeby, S. P., Tudor, M. J. and White, N. M., "Energy harvesting vibration sources for microsystems applications," Measurement Science & Technology 17(12), R175-R195 (2006).
[5] Lei, W., and Yuan, F. G., "Energy harvesting by magnetostrictive material (MsM) for powering wireless sensors in SHM," Proc. SPIE 6529(2), 652941 (2007).
[6] Beeby, S. P., Torah, R. N., Tudor, M. J., Glynne-Jones, P., O'Donnell, T., Saha, C. R., and Roy, S., "A micro electromagnetic generator for vibration energy harvesting," Journal of Micromechanics and Microengineering 17(7), 1257-1265 (2007).
[7] Kulkarni, S., Koukharenko, E., Torah, R., Tudor, J., Beeby, S., O'Donnell, T. and Roy, S., "Design, fabrication and test of integrated micro-scale vibration-based electromagnetic generator," Sensors and Actuators a-Physical 145-146(1-2), 336-342 (2008).
[8] Yick, J., Mukherjee, B. and Ghosal, D., "Wireless sensor network survey," Computer Networks 52(12), 2292-2330 (2008).
[9] Sung, W. T. and Hsu, Y. C., "Designing an industrial real-time measurement and monitoring system based on embedded system and ZigBee," Expert Systems with Applications 38(4), 4522-4529 (2011).
