experimentandpredictionofablationdepthinexcimerlaser...

Research ArticleExperiment and Prediction of Ablation Depth in Excimer LaserMicromachining of Optical Polymer Waveguides

K F Tamrin 1 S S Zakariyah2 K M A Hossain3 and N A Sheikh 4

1Department of Mechanical and Manufacturing Engineering Faculty of Engineering Universiti Malaysia Sarawak (UNIMAS)94300 Kota Samarahan Sarawak Malaysia2College of Engineering and Technology University of Derby Derby DE22 3AW UK3School of Electrical Mechanical and Mechatronics System University of Technology Sydney Ultimo NSW Australia4Department of Mechanical Engineering Faculty of Engineering and Technology HITEC University Islamabad Pakistan

Correspondence should be addressed to K F Tamrin tkfikriunimasmy

Received 27 October 2017 Revised 26 December 2017 Accepted 9 January 2018 Published 4 March 2018

Academic Editor Angela De Bonis

Copyright copy 2018 K F Tamrin et al -is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Extending the data transfer rates through dense interconnections at inter- and intraboard levels is a well-established techniqueespecially in consumer electronics at the expense of more cross talk electromagnetic interference (EMI) and power dissi-pation Optical transmission using optical fibre is practically immune to the aforementioned factors Among themanufacturing methods UV laser ablation using an excimer laser has been repeatedly demonstrated as a suitable technique tofabricate multimode polymer waveguides However the main challenge is to precisely control and predict the topology of thewaveguides without the need for extensive characterisation which is both time consuming and costly In this paper the authorspresent experimental results of investigation to relate the fluence scanning speed number of shots and passes at varying pulserepetition rate with the depth of ablation of an acrylate-based photopolymer -e depth of ablation essentially affects totalinternal reflection and insertion loss and these must be kept at minimum for a successful optical interconnection on printedcircuit boards-e results are then used to predict depth of ablation for this material by means of adaptive neurofuzzy inferencesystem (ANFIS) modelling -e predicted results with a correlation of 09993 show good agreement with the experimentalvalues-is finding will be useful in better predictions along with resource optimisation and ultimately helps in reducing cost ofpolymer waveguide fabrication

1 Introduction

Consumer electronics are experiencing uphill challenge toovercome high-frequency (gt10Gbs) transmission problemssuch as cross talk electromagnetic interference and powerdissipation especially with high interconnectivity [1] Rushfor miniaturization addition of more featuresapplicationsand communication at microscales (eg chip-to-chip) haveexposed the limits of copper-based transmissions in elec-tronic circuits To overcome this bottleneck an opticallyenabled printed circuit board is a viable solution as thistechnique has been successfully used for long haul commu-nications that is optical fiber -e optical interconnections(OIs) used in conjunction with copper connection thus

forming electric-optical bridge are usually made up of polymerwaveguides Various fabrication techniques of these polymerwaveguides have been proposed including laser direct writingphotolithography inkjet printing and laser ablation [2]Amongst all laser ablation is the preferred technique forfabrication primarily due to its proven use as well as availabilitywith PCB manufacturers [1 2]

Laser-matter interaction and its suitability for machiningvarious engineering materials including various polymers[3ndash5] have been well researched and documented in theliterature Micromachining using IR (infrared) laser sourcesparticularly using CO2 laser is common as it allows sig-nificantly higher rates of machining [6] A few studiesreport the use of CO2 laser process for polymer waveguide

HindawiAdvances in Materials Science and EngineeringVolume 2018 Article ID 5616432 9 pageshttpsdoiorg10115520185616432

fabrications In the pulsed mode CO2 laser machining ofPMMA (polymethyl methacrylate) waveguides was re-ported by Steenberge et al [7] Subsequently an extensivestudy on a continuous wave (CW) CO2 laser processing ofTruemodetrade polymer is detailed by Zakariyah et al [8 9]e study presented direct relationship between the depthof ablation and the scanning energy density while the widthof the track was seen to vary logarithmically with the scanningenergy density In another study submerged waveguides weredemonstrated by Ozcan et al [10] using a CW CO2 laser inplanar silica lms that showed reduction in the propagation loss

In PCBmanufacturing the use of a CO2 laser is generallypreferred owing to its lower cost and availability On theother hand an excimer laser oers certain advantages byallowingmachining of ne features at controllable ratesisprocess also ensures that there is less thermal damage on theablated materials with minimum heat-aected zone (HAZ)dictated by among other factors a very high-absorptioncoecient and short interaction between the laser and theablated sample e two key controlling parameters of thislaser class are short wavelength and pulse duration eformer reduces the thermal diusivity of the surface whilethe latter provides high-energy pulses Furthermore highmachining quality and ablation threshold are related to thepulse width duration of excimer laser [11] Typical operatingwavelengths of excimer lasers used for machining polymermaterials especially for fabrication of optical waveguides are193 nm (with ArF) and 248 nm (with KrF) [12]

Energy and pulse repetition rate (PRR) are two im-portant process parameters highlighted by Steenberge et al[7 13 14] for the fabrication of waveguides in ORMOCER Itwas observed that both parameters inuence the smoothnessof side walls of guides which remains crucial for the prop-agation loss In another study Peging et al [15] employedexcimer laser to fabricate single-mode optical waveguides inPMMA Following the method used for CO2 laser [10]Peging et al selected UV uences below the ablationthreshold of PMMA Another study conducted by Zakariyahet al [11] showed good surface characteristics of excimer lasercompared to 355 nm UV NdYAG and 106 μm CO2 lasers

For long distance transmission of signal through thepolymeric waveguides total internal reection is an im-portant feature which is strongly inuenced by the depth ofablation As shown in Figure 1(a) it can be noted thatminimum depth of ablation is desirable although propaga-tion of optical signal can be achieved by having scenarioshown in Figure 1(b) However further increase in depth of

ablation can expose the cladding which means not only morelaser power is utilized for the fabrication but also undesiredthermal eects lead to increase the losses in the opticalchannel is scenario is depicted in Figure 1(c) with pos-sibility of severe damage to the FR4 substrate

Apart from internal reection insertion loss is anothermain criterion for the optical interconnection In fact theprime measure of any optical waveguide is insertion losswhich is strongly inuenced by the depth of ablation as wellas track width Both depth of ablation and track width aremainly dependent on the laser processing parameters in-cluding laser power rate of pulse repetition focal distanceand scanning speed Further investigation of these param-eters is presented in forthcoming sections

e ablation process for polymer waveguides usingexcimer laser is intricate involving several process param-eters and material properties Some researchers have sug-gested using techniques of soft computing to deal with thesecomplexities [6 8ndash10] ese techniques are good at dis-secting intricate problems with their features of nonlinearityand adaptability For instance articial neural network(ANN) and fuzzy logic have been successfully utilized topredict complex system properties [16] However ANN hasgained some untoward reviews due to its rather lsquolsquoblack boxrsquorsquonature On the other hand taking the advantage of ap-proximate reasoning modelling techniques based on fuzzylogic have gained some attention However main bottleneckfor this technique is the construction of membershipfunctions (MFs) While combining both fuzzy logic andANN adaptive network-based fuzzy system takes full ad-vantage of both techniques and eliminates their individualshortcomings [11] is approach known as adaptive neu-rofuzzy inference system (ANFIS) imitates human thinkingprocess and constructs input-output maps on humanknowledge as well as the data pairs

Several studies involving prediction of mechanicalproperties have found ANFIS predictions more reliablecompared to ANN For instance results reported by Zareand Khaki [17] and Karaagaccedil et al [18] showed that ANFISperformed better than ANN for predicting optimum valuesof their respective studies Related to the use of ANFIS forlaser micromachining no such study could be found inliterature especially for the fabrication of optical waveguides

In the light of above discussion this work has three mainobjectives (1) experimental investigation of the use of248 nm excimer laser for micromachining of optical polymerused in multimode polymer waveguides (2) ANFIS mod-elling in prediction of ablation depth from the experimentaldata and (3) comparison of ANFISmodelling with the resultgained from the experiment It is important to mention thatit will be a rst study of its kind to utilize ANFIS andcompares the results with the experimental observations foracrylate-based photopolymer using UV excimer

2 Experimental Method

21 Laser System For this work 7000 Series Exitech of 20 nspulse krypton uoride (KrF) excimer laser (λ 248 nm) wasused which could deliver up to 250mJpulse e size and

Core

Lower cladding Depth ofablation

(a) (b)Desirable Acceptable Undesirable

FR4 substrate

(c)

Figure 1 A schematic diagram showing three dierent types ofablation depths using laser beam with top-hat prole

2 Advances in Materials Science and Engineering

shape of the beam were primarily determined by the maskdimension placed after the beam exit which then passedthrough a 10x projection lens is allowed a focus beamspot to be obtained at the work piece for eective patterningof the samples In the reported experiments a TEM00 beammode from the laser was masked by a square beam mask(5times 5mm2) resulting in top-hat-like beam prole

22 Optical Polymer Material A UV-curable acrylate-basedoptical polymer from Exxelis Ltd (UK) commerciallyknown as Truemode was usedis is a light-yellow mixtureof monomers and liquid at room temperature Its low ab-sorption loss shown to be lt004 dBcm at 850 nm [15]supports its suitability for applications in optical in-terconnections It is also shown to be compatible withstandard printed circuit boards manufacturing processparticularly at the high temperature and pressure involvedFollowing several trials of various formulations available fromthe supplier authors selected EXX-core 37E and EXX-clad 277formulations with indices of refraction of 15560 and 15265respectively e refractive indices were measured at 850nmwavelength which is the same wavelength at which the opticalpropagation loss is measured for polymer waveguides

23 Fabrication Figure 2 shows the key fabrication phasesemployed in patterning a single layer optical waveguide(single- or mutimode) similar to what is reported in [19] Tosummarise the FR4 base substrate was decontaminatedfrom debris by cleaning in water (or methanol) and wasallowed to get fully dry in an oven Optical layers (core andcladding) were thereafter formed on FR4 by spin coatingTruemode at approximately 350 revmin for 30 sec elower cladding was spun and UV cured initially in N2 en-vironment under UV light for 200plusmn 40 sec is is followedby spin coating the core layer At this stage the whole sample

of FR4-clad-core was UV cured again for complete cross-linking and subsequently oven-baked at a temperature of100degC for 60 minutes e obtained thickness in each case(core or cladding layer) was 40plusmn 10 microns As the core andcladding layers are essentially made from the same materialswith a slight dierence in the refractive index as previouslyindicated their interaction with the laser and the absorptioncoecient is expected to be similar erefore nothingunusual was observed in the interface of these layersGenerally the fabrication of polymer waveguides was fol-lowed by an extensive system characterisation and as a re-sult the steps are same whether the ultimate goal was tofabricate or to characterise the depth of ablation is is toensure that data obtained during the characterisation can beimplemented for the fabrication phase without undue ad-justment to the setting

3 Prediction of Depth of Ablation by Means ofAdaptive Neurofuzzy InferenceSystem (ANFIS)

In this study an adaptive neurofuzzy inference system(ANFIS) is adopted to predict the depth of ablation asa function of ve input variables using experimental datalisted in Table 1 An ANFIS is an adaptive network whichcombines fuzzy logic and neural network system to developthe prediction model based on the input knowledge eANFIS was rst introduced by Jang in 1993 [20] andfunctionally it is similar to the TakagindashSugeno-type inferencemodel To diligently adjust the membership function of thefuzzy inference system (FIS) the ANFIS employs the leastsquares method and back-propagation gradient descentmethod from neural network [21] Figure 3 illustratesa general ANFIS architecture assuming two inputs namelyx and y and one output namely f e architecture typicallyhas ve dierent layers characterized by adaptive and xed

FR4 substrate FR4 substrate FR4 substrate

(a) Stage 1

(d) Stage 4

(b) Stage 2 (c) Stage 3

FR4 substrate FR4 substrate

Lower cladding Lower claddingCore

Lower cladding

(e) Stage 5

Figure 2 Key fabrication phases of laser ablation of multimode waveguides (a) a base substrate of FR4 (b) formation of lower cladding byspin coating (c) formation of the core layer (d) laser ablation of the resulting sample from the core layer into the cladding layer and (e) spincoating of the (optional) upper cladding

Advances in Materials Science and Engineering 3

nodes As highlighted in Figure 3 the output of each layer(layers 2ndash5) is used as the input to the succeeding layer

-e first layer is the adaptive node layer or commonlyknown as the fuzzification layer where membership func-tions (generalized bell Gaussian or triangular) are desig-nated In this layer bell-shaped membership function can beselected with the maximum value of 10 and minimum valueof 00 -e function can be written as

μAi(x) 1

1 + ( x minus ci1113857( ai1113857( 11138572

1113876 1113877bi

(1)

where ai bi and ci are a set of parameters and microAi(x) is themembership function corresponding to the input parameterx Accordingly the function microAi(x) changes as ai bi and civary

Each node in the second layer is nonadaptive (fixed) andit acts as the fuzzy ldquoANDrdquo operator in the premise section ofthe fuzzy if-then rule In other words it multiplies the in-coming signals and the output represents the firing (op-erating) strength of each fuzzy rule For example

Wi μAi(x)μBi(y) i 1 2 (2)

Table 1 Experimental layout and results

ExperimentProcessing parameter Response

Number ofshots

Fluence(Jcm2)

Scanning speed(mmmin)

Pulse repetitionrate (Hz)

Number ofpasses

Maximum depth(μm)1 shot

Mean depth(μm)1 shot

1 1 0086 120 20 1 0799 06782 2 0086 60 20 1 186 1653 3 0086 40 20 1 279 2494 4 0086 30 20 1 353 3125 5 0086 24 20 1 456 4046 6 0086 20 20 1 573 5157 7 0086 171 20 1 630 5538 8 0086 15 20 1 708 629 9 0086 133 20 1 759 66110 10 0086 12 20 1 881 76311 12 0086 10 20 1 975 8512 14 0086 857 20 1 120 10413 16 0086 75 20 1 131 11414 18 0086 667 20 1 149 13115 20 0086 6 20 1 166 14516 22 0086 545 20 1 193 1717 24 0086 5 20 1 200 17518 26 0086 462 20 1 218 18919 28 0086 429 20 1 229 20220 30 0086 4 20 1 242 2121 10 003 6 10 1 0252 017322 10 0058 6 10 1 0516 042523 10 0086 6 10 1 0797 069524 10 0113 6 10 1 111 098825 10 0141 6 10 1 146 13526 10 0169 6 10 1 179 16627 10 0197 6 10 1 196 18228 10 0224 6 10 1 219 20429 10 0252 6 10 1 23 21730 10 028 6 10 1 25 23631 1 0086 120 20 3 279 24332 1 0086 120 20 4 39 3133 1 0086 120 20 5 41 34234 1 0086 120 20 9 86 75835 1 0086 120 20 10 100 84736 1 0086 72 12 1 107 0967


whereWi is the weight function for the next layer and microBi(y)is the membership function corresponding to the inputparameter y

e third layer is also a nonadaptive layer which is alsoknown as the normalized layer It serves to calculate the nor-malized ring strength in (2) using the following equation [22]

Adaptive node

Fixed node

Layer 1

x

y

W1

W2

sum

f

Layer 2 Layer 3 Layer 4 Layer 5

A1

A2

B1

B2

W1

W2

W1 f1

W1 f2

Figure 3 Typical ANFIS architecture for two inputs and one output

02 025 03 035 04 045 05 055 06 065 07 075 015010050

Maximum depth1

11 microm103 microm

Depth

0206 mmMean depthWidth

ndash1ndash075ndash05ndash025

0025

05075

(microm

)

1

(mm)

Figure 4 Screenshot of depth measurement using Talysurf CLI scanning topography-measuring instrumente instrument gives values ofmaximum and mean depths of ablation and width of ablation

(a) (b)

Figure 5 SEM images of ablated depth at (a) 250x and (b) 650x magnications (sample number 61 number of shots 45uence 01 Jcm2 scanning speed 33mmmin and pulse repetition rate 25Hz) Measured maximum and mean depths are 1002 μmand 0740 μm respectively


Wlowasti

Wi

W1 + W2 i 1 2 (3)

-e fourth layer consists of adaptive nodes -is layerestimates the contribution of each input parameter towardsthe overall output -is is carried out by the assignment ofweights to minimize the deviation between outputs anddesired results -e relation between inputs x and y andoutput f can be defined as

Wlowasti fi W

lowasti pix + qiy + ri( 1113857 i 1 2 (4)

where pi qi and ri are the linear parameters or consequentparameters and they can be adjusted in the training process

-e fifth layer is the defuzzification (output) layer andessentially the summation of all fourth layer signals -eoutput f can be computed by the following equation

f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)

4 Results and Discussion

Typical and essential stages involved in multimode polymerwaveguide fabrication were followed and effect of processparameters on the prepared samples were analysed andmeasured Before measurement the ablated samples werefirst sectioned (using Buehler IsoMetreg low-speed saw) andthen polished using abrasives (1 μmdiamond paste) Ablateddepths of channels were measured using the SmartScoperegFlashtrade 200 optical measuring device -e device has a tra-verse length of 200times 200times150mm with a resolution of05 μm Topographic measurements were carried out usinga single point sensor gauge of the Talysurf CLI scanner whichcan travel up to a maximum speed of 20mms -e latterinstrument took measurements across the channel at equalintervals along the channel and it automatically calculatedthe respective mean and maximum depths of ablation -eexperimental results are given in Table 1 A total of 36experiments were designed It should be noted that the effectof all individual processing parameters on the responseparameters were studied separately while keeping otherparameters constant

As shown in Figures 4 and 5 the ablated section resemblesthe top-hat beamprofilewith some tapering effect on both edges(slightly tapered sidewalls) Onemethod to lower the tapering athigh fluence (sim2 Jmiddotcmminus2) as suggested in [23] is by increasingthe number of shots However despite the investigation noconclusion can be drawn on the significance of these factors onthe tapering effect seen in the ablated profile Furthermoresome of the observed tapering effects can be linked to the focusposition of laser as evident in laser system characterisation

As earlier it is pointed out that the use of excimer laser ispreferred as it is readily absorbed by the polymers and causesminimum thermal load and damage to the surroundings In thisstudy UV pulses of 10ndash35ns in length were used which pro-duced relatively clean ablation with little thermal damage tosurroundings as shown in Figure 6(a) On the other hand highprocessing speed of its counterpart IR CO2 laser resulted in largeHAZ which is undesirable and may influence the propagationlosses in the waveguide (Figure 6(b)) However slow speed andhigh operating costs of excimer laser and the use of fluorine gas(toxic in nature) are main reasons of its lesser acceptance in thePCB industry -e last factor can be easily overcome by puttingin place engineering and administrative controls while the effectsof others can also be minimised

For the ANFIS architecture five inputs for the network areused Number of shots fluence scanning speed pulse repetitionrate and number of passes are the processing parameters whilethe depth of ablation is the measured output In the presentwork the experimental data were divided into two groups forthe purpose of training and subsequent testing-e first 27 datafrom Table 1 were chosen to be the training samples while theremaining data were used to be the testing samples (9 samples)

In the beginning a triangular-shaped membership functionwas selected for each input variable in ANFIS architecture

Side wall

Track bottom

(a)

Track bottom

Side wall

(b)

Figure 6 Surface roughness comparison of ablation tracks machined using (a) excimer (sample number 116 line 5) and (b) CO2 (sample A1)lasers

Table 2 Mean prediction error for different types of membershipfunctions

Membership function Mean prediction error ()Triangular (trimf) 4194Generalized bell (gbellmf) 1991Gaussian (gaussmf) 2183Trapezoidal (trapmf) 3233


Subsequently dierent types of membership function wereevaluated that is generalized bell Gaussian and trapezoidal ASugeno-type FIS structure was chosen to initialize the mem-bership function parameters for ANFIS training and the outputmembership function was set to be linear

Equations (4) and (5) were used to predict the outputswhich help in training the ANFIS model Weights andthresholds at dierent layers were adjusted using the back-propagation algorithm relying on the gradient searchtechnique Weights were adjusted in order to minimize themean-squared error between the experimental and predictedresults e training phase required some initial estimationof ANFIS parameters such as type of membership functionsfor each input variable and fuzzy rule For instance Table 2shows results of the mean prediction error for four dierenttypes of membership functions Due to its lowest percentageof error the generalized bell function was chosen for bothtraining and testing (Figure 7) After the training of theAFNIS model is completed the model can be utilized to

predict the depth of ablation A complete comparison of thepredicted and experimentally measured depth of ablation isshown in Table 3 and Figure 8

It is important to mention that the mean relative errorfor the present ANFIS is comparable to the ndings in theliterature For instance Jia et al [24] acquired 523 meanrelative error for 30 sets of experiments while Zare andKhaki [17] managed to obtain much lesser mean relativeerror of 162 for 10 sets of experiments ANFIS predictionfor the present work can be considered comparatively accuratesince the training data extensively cover the complete range ofparameters Based on Figure 8 experiment numbers 6 and 33show considerably large prediction errors of 7934 and16371 respectively Nevertheless the predicted values re-main below themeasuredmaximumdepths of ablation Figure 9graphically compares experiment and predicted depth of ab-lations based on regression analysis It is clear that the predictedANFIS results compare well with the experimental results withthe correlation coecient (R2) value up to 09993

5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)

Figure 7 Membership function diagrams based on generalized bell for all inputs (a) number of shots (b) uence (c) scanning speed(d) pulse repetition rate and (e) number of passes


5 Conclusions

is paper experimentally investigates the eects of varyinglaser processing parameters on the depth of ablation ofoptical polymer waveguides Laser ablation using excimerlaser results in a cleaner and smoother ablated structureswith minimal thermal damage to the surroundings Howeverdue to slow speed and high operating costs of the excimer laserthere is a need for a mechanism to accurately predict andestimate the ablation depth of the optical waveguides forvarious input parameters Here an adaptive neurofuzzy

inference system (ANFIS) model is proposed and applied toa series of experimental datasets e trained ANFIS modelpredicts the ablation depth with an accuracy of R2 09993 anda mean prediction error of 1991 It has been shown that withthe aid of the ANFIS a multiple inputsingle output fuzzyinference system can be created that accurately characterizesthe ablation depth of optical polymer waveguides

Conflicts of Interest

e authors declare that they have no conicts of interest

Acknowledgments

e initial result of the work reported here was part ofthe Integrated Optical and Electronic Interconnect PCB

Table 3 Comparison of experimental and predicted mean depthsand their relative errors

Experiment Experimentalmean depth (μm)

Predicted meandepth (μm)

relativeerror

(gbellmf )1 0678 0702 34962 165 1651 00553 249 2474 06554 312 3184 20355 404 3979 15056 515 4741 79347 553 5437 16898 62 6139 09819 661 6782 260410 763 7442 246111 85 8752 296412 104 10168 223413 114 11721 281814 131 13333 177915 145 14914 285216 17 16398 354417 175 17740 137218 189 18976 040419 202 20142 028720 21 21250 119021 0173 0174 040522 0425 0428 061223 0695 0691 057624 0988 0988 004025 135 1356 043726 166 1652 049427 182 1851 172028 204 2017 113229 217 2184 064130 236 2350 041531 243 2348 337032 31 3164 206833 342 3980 1637134 758 7556 032235 847 8489 021836 0967 0967 0000Mean relative error 1991

0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number

ExperimentPredictionError ()

Figure 8 Graph comparing experiment and predicted depth ofablation including error prediction

25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10

Experimental depth of ablation (microm)15 20 25

Predicted = 10013 times (Exp) + 00056R2 = 09993

Figure 9 Experimental and ANFIS model-predicted depth ofablation plotted for the best model developed using ve inputvariables


Manufacturing (OPCB) IeMRC Flagship Project FS060101financially supported by the UK Engineering and PhysicalSciences Research Council (EPSRC) through the InnovativeElectronics Manufacturing Research Centre (IeMRC) -emodelling work is supported by the Ministry of Higher Edu-cation Malaysia under Grant no F02FRGS16192017 -eauthors are also grateful for the technical input of the 8 col-laborating companies and three universities (LoughboroughUniversity University College London and Heriot-WattUniversity) who were partners in the consortium and spe-cial thanks are due toKhadijahOlaniyan for helpful discussions

References

[1] M C Gower ldquoIndustrial applications of laser micro-machiningrdquo Optics Express vol 7 no 2 pp 56ndash67 2000

[2] K Yung D Zeng and T Yue ldquoXPS investigation of Upilex-Spolyimide ablated by 355 nm Nd YAG laser irradiationrdquoApplied Surface Science vol 173 no 3-4 pp 193ndash202 2001

[3] K Tamrin Y Nukman and S Zakariyah ldquoLaser lap joining ofdissimilar materials a review of factors affecting jointstrengthrdquo Materials and Manufacturing Processes vol 28no 8 pp 857ndash871 2013

[4] K Tamrin Y Nukman N A Sheikh and M Z HarizamldquoDetermination of optimum parameters using grey relationalanalysis for multi-performance characteristics in CO2 laserjoining of dissimilar materialsrdquoOptics and Lasers in Engineeringvol 57 pp 40ndash47 2014

[5] K Tamrin Y Nukman and N Sheikh ldquoLaser spot welding ofthermoplastic and ceramic an experimental investigationrdquo Mate-rials andManufacturing Processes vol 30 no 9 pp1138ndash1145 2015

[6] K Tamrin Y Nukman I A Choudhury and S ShirleyldquoMultiple-objective optimization in precision laser cutting ofdifferent thermoplasticsrdquo Optics and Lasers in Engineeringvol 67 pp 57ndash65 2015

[7] G Van Steenberge P Geerinck S Van Put and P Van DaeleldquoIntegration of multimode waveguides and micromirrorcouplers in printed circuit boards using laser ablationrdquo inProceedings of the Photonics Europe International Society forOptics and Photonics Strasbourg France April 2004

[8] K Tamrin S Zakariyah and N Sheikh ldquoMulti-criteria op-timization in CO2 laser ablation of multimode polymerwaveguidesrdquo Optics and Lasers in Engineering vol 75pp 48ndash56 2015

[9] S S Zakariyah P P Conway D A Hutt K Wang andD R Selviah ldquoCO2 laser micromachining of optical wave-guides for interconnection on circuit boardsrdquo Optics andLasers in Engineering vol 50 no 12 pp 1752ndash1756 2012

[10] L Ccedil Ozcan F Guay R Kashyap and L Martinu ldquoFabricationof buried waveguides in planar silica films using a direct CWlaser writing techniquerdquo Journal of Non-Crystalline Solidsvol 354 no 42 pp 4833ndash4839 2008

[11] S S Zakariyah P P Conway D A Hutt et al ldquoPolymer opticalwaveguide fabrication using laser ablationrdquo in Proceedings of the11th Electronics Packaging Technology Conference EPTCrsquo09Singapore December 2009

[12] Y-T Chen K Naessens R Baets Y-S Liao and A A TsengldquoAblation of transparent materials using excimer lasers forphotonic applicationsrdquo Optical Review vol 12 no 6pp 427ndash441 2005

[13] NHendrickx G Van Steenberge P Geerinck and P VanDaeleldquoLaser ablation as enabling technology for the structuring of

optical multilayer structuresrdquo Journal of Physics ConferenceSeries vol 59 pp 118ndash121 2007

[14] G Van Steenberge P Geerinck S Van Put et al ldquoMT-compatible laser-ablated interconnections for optical printedcircuit boardsrdquo Journal of Lightwave Technology vol 22 no 9pp 2083ndash2090 2004

[15] W Pfleging M Przybylski and H Bruckner ldquoExcimer lasermaterial processingmdashstate of the art and new approaches inmicrosystem technologyrdquo in Proceedings of the SPIE EuropeJuly 2006

[16] A Hossain A Hossain Y Nukman et al ldquoA fuzzy logic-based prediction model for kerf width in laser beam ma-chiningrdquo Materials and Manufacturing Processes vol 31no 5 pp 679ndash684 2016

[17] M Zare and J V Khaki ldquoPrediction of mechanical propertiesof a warm compacted molybdenum prealloy using artificialneural network and adaptive neuro-fuzzy modelsrdquo Materialsamp Design vol 38 pp 26ndash31 2012

[18] B Karaagaccedil M Inal and V Deniz ldquoPredicting optimum curetime of rubber compounds by means of ANFISrdquoMaterials ampDesign vol 35 pp 833ndash838 2012

[19] S S Zakariyah P P Conway D A Hutt et al ldquoFabrication ofpolymer waveguides by laser ablation using a 355 nmwavelength Nd YAG laserrdquo Journal of Lightwave Technologyvol 29 no 23 pp 3566ndash3576 2011

[20] J-S Jang ldquoANFIS adaptive-network-based fuzzy inferencesystemrdquo IEEE transactions on systems man and cyberneticsvol 23 no 3 pp 665ndash685 1993

[21] H B DemuthM T HaganM Hudson Beale andO De JesusNeural Network Design Martin Hagan 2014

[22] S S Roy ldquoDesign of adaptive neuro-fuzzy inference systemfor predicting surface roughness in turning operationrdquoJournal of Scientific and Industrial Research vol 64 no 9p 653 2005

[23] N Bityurin ldquo8 Studies on laser ablation of polymersrdquo AnnualReports Section ldquoCrdquo (Physical Chemistry) vol 101 pp 216ndash2472005

[24] W Jia W Zeng Y Han J Liu Y Zhou and Q WangldquoPrediction of flow stress in isothermal compression of Ti60alloy using an adaptive network-based fuzzy inference sys-temrdquoMaterials amp Design vol 32 no 10 pp 4676ndash4683 2011


CorrosionInternational Journal of

Hindawiwwwhindawicom Volume 2018

Advances in

Materials Science and EngineeringHindawiwwwhindawicom Volume 2018


Journal of

Chemistry

Analytical ChemistryInternational Journal of


ScienticaHindawiwwwhindawicom Volume 2018

Polymer ScienceInternational Journal of



Advances in Condensed Matter Physics


International Journal of

BiomaterialsHindawiwwwhindawicom

Journal ofEngineeringVolume 2018

Applied ChemistryJournal of


NanotechnologyHindawiwwwhindawicom Volume 2018

Journal of


High Energy PhysicsAdvances in

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

TribologyAdvances in



ChemistryAdvances in


Advances inPhysical Chemistry


BioMed Research InternationalMaterials

Journal of


Na

nom

ate

ria

ls


Journal ofNanomaterials

Submit your manuscripts atwwwhindawicom

fabrications In the pulsed mode CO2 laser machining ofPMMA (polymethyl methacrylate) waveguides was re-ported by Steenberge et al [7] Subsequently an extensivestudy on a continuous wave (CW) CO2 laser processing ofTruemodetrade polymer is detailed by Zakariyah et al [8 9]e study presented direct relationship between the depthof ablation and the scanning energy density while the widthof the track was seen to vary logarithmically with the scanningenergy density In another study submerged waveguides weredemonstrated by Ozcan et al [10] using a CW CO2 laser inplanar silica lms that showed reduction in the propagation loss

In PCBmanufacturing the use of a CO2 laser is generallypreferred owing to its lower cost and availability On theother hand an excimer laser oers certain advantages byallowingmachining of ne features at controllable ratesisprocess also ensures that there is less thermal damage on theablated materials with minimum heat-aected zone (HAZ)dictated by among other factors a very high-absorptioncoecient and short interaction between the laser and theablated sample e two key controlling parameters of thislaser class are short wavelength and pulse duration eformer reduces the thermal diusivity of the surface whilethe latter provides high-energy pulses Furthermore highmachining quality and ablation threshold are related to thepulse width duration of excimer laser [11] Typical operatingwavelengths of excimer lasers used for machining polymermaterials especially for fabrication of optical waveguides are193 nm (with ArF) and 248 nm (with KrF) [12]

Energy and pulse repetition rate (PRR) are two im-portant process parameters highlighted by Steenberge et al[7 13 14] for the fabrication of waveguides in ORMOCER Itwas observed that both parameters inuence the smoothnessof side walls of guides which remains crucial for the prop-agation loss In another study Peging et al [15] employedexcimer laser to fabricate single-mode optical waveguides inPMMA Following the method used for CO2 laser [10]Peging et al selected UV uences below the ablationthreshold of PMMA Another study conducted by Zakariyahet al [11] showed good surface characteristics of excimer lasercompared to 355 nm UV NdYAG and 106 μm CO2 lasers

For long distance transmission of signal through thepolymeric waveguides total internal reection is an im-portant feature which is strongly inuenced by the depth ofablation As shown in Figure 1(a) it can be noted thatminimum depth of ablation is desirable although propaga-tion of optical signal can be achieved by having scenarioshown in Figure 1(b) However further increase in depth of

ablation can expose the cladding which means not only morelaser power is utilized for the fabrication but also undesiredthermal eects lead to increase the losses in the opticalchannel is scenario is depicted in Figure 1(c) with pos-sibility of severe damage to the FR4 substrate

Apart from internal reection insertion loss is anothermain criterion for the optical interconnection In fact theprime measure of any optical waveguide is insertion losswhich is strongly inuenced by the depth of ablation as wellas track width Both depth of ablation and track width aremainly dependent on the laser processing parameters in-cluding laser power rate of pulse repetition focal distanceand scanning speed Further investigation of these param-eters is presented in forthcoming sections

e ablation process for polymer waveguides usingexcimer laser is intricate involving several process param-eters and material properties Some researchers have sug-gested using techniques of soft computing to deal with thesecomplexities [6 8ndash10] ese techniques are good at dis-secting intricate problems with their features of nonlinearityand adaptability For instance articial neural network(ANN) and fuzzy logic have been successfully utilized topredict complex system properties [16] However ANN hasgained some untoward reviews due to its rather lsquolsquoblack boxrsquorsquonature On the other hand taking the advantage of ap-proximate reasoning modelling techniques based on fuzzylogic have gained some attention However main bottleneckfor this technique is the construction of membershipfunctions (MFs) While combining both fuzzy logic andANN adaptive network-based fuzzy system takes full ad-vantage of both techniques and eliminates their individualshortcomings [11] is approach known as adaptive neu-rofuzzy inference system (ANFIS) imitates human thinkingprocess and constructs input-output maps on humanknowledge as well as the data pairs

Several studies involving prediction of mechanicalproperties have found ANFIS predictions more reliablecompared to ANN For instance results reported by Zareand Khaki [17] and Karaagaccedil et al [18] showed that ANFISperformed better than ANN for predicting optimum valuesof their respective studies Related to the use of ANFIS forlaser micromachining no such study could be found inliterature especially for the fabrication of optical waveguides

In the light of above discussion this work has three mainobjectives (1) experimental investigation of the use of248 nm excimer laser for micromachining of optical polymerused in multimode polymer waveguides (2) ANFIS mod-elling in prediction of ablation depth from the experimentaldata and (3) comparison of ANFISmodelling with the resultgained from the experiment It is important to mention thatit will be a rst study of its kind to utilize ANFIS andcompares the results with the experimental observations foracrylate-based photopolymer using UV excimer

2 Experimental Method

21 Laser System For this work 7000 Series Exitech of 20 nspulse krypton uoride (KrF) excimer laser (λ 248 nm) wasused which could deliver up to 250mJpulse e size and

Core

Lower cladding Depth ofablation

(a) (b)Desirable Acceptable Undesirable

FR4 substrate

(c)

Figure 1 A schematic diagram showing three dierent types ofablation depths using laser beam with top-hat prole









(a) Stage 1

(d) Stage 4




Lower cladding

(e) Stage 5





μAi(x) 1

1 + ( x minus ci1113857( ai1113857( 11138572

1113876 1113877bi

(1)






Number ofshots

Fluence(Jcm2)



Number ofpasses



1 1 0086 120 20 1 0799 06782 2 0086 60 20 1 186 1653 3 0086 40 20 1 279 2494 4 0086 30 20 1 353 3125 5 0086 24 20 1 456 4046 6 0086 20 20 1 573 5157 7 0086 171 20 1 630 5538 8 0086 15 20 1 708 629 9 0086 133 20 1 759 66110 10 0086 12 20 1 881 76311 12 0086 10 20 1 975 8512 14 0086 857 20 1 120 10413 16 0086 75 20 1 131 11414 18 0086 667 20 1 149 13115 20 0086 6 20 1 166 14516 22 0086 545 20 1 193 1717 24 0086 5 20 1 200 17518 26 0086 462 20 1 218 18919 28 0086 429 20 1 229 20220 30 0086 4 20 1 242 2121 10 003 6 10 1 0252 017322 10 0058 6 10 1 0516 042523 10 0086 6 10 1 0797 069524 10 0113 6 10 1 111 098825 10 0141 6 10 1 146 13526 10 0169 6 10 1 179 16627 10 0197 6 10 1 196 18228 10 0224 6 10 1 219 20429 10 0252 6 10 1 23 21730 10 028 6 10 1 25 23631 1 0086 120 20 3 279 24332 1 0086 120 20 4 39 3133 1 0086 120 20 5 41 34234 1 0086 120 20 9 86 75835 1 0086 120 20 10 100 84736 1 0086 72 12 1 107 0967




Adaptive node

Fixed node

Layer 1

x

y

W1

W2

sum

f


A1

A2

B1

B2

W1

W2

W1 f1

W1 f2


02 025 03 035 04 045 05 055 06 065 07 075 015010050

Maximum depth1

11 microm103 microm

Depth



0025

05075

(microm

)

1

(mm)


(a) (b)



Wlowasti

Wi

W1 + W2 i 1 2 (3)


Wlowasti fi W




f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)







Side wall

Track bottom

(a)

Track bottom

Side wall

(b)









5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls











(a) Stage 1

(d) Stage 4




Lower cladding

(e) Stage 5





μAi(x) 1

1 + ( x minus ci1113857( ai1113857( 11138572

1113876 1113877bi

(1)






Number ofshots

Fluence(Jcm2)



Number ofpasses



1 1 0086 120 20 1 0799 06782 2 0086 60 20 1 186 1653 3 0086 40 20 1 279 2494 4 0086 30 20 1 353 3125 5 0086 24 20 1 456 4046 6 0086 20 20 1 573 5157 7 0086 171 20 1 630 5538 8 0086 15 20 1 708 629 9 0086 133 20 1 759 66110 10 0086 12 20 1 881 76311 12 0086 10 20 1 975 8512 14 0086 857 20 1 120 10413 16 0086 75 20 1 131 11414 18 0086 667 20 1 149 13115 20 0086 6 20 1 166 14516 22 0086 545 20 1 193 1717 24 0086 5 20 1 200 17518 26 0086 462 20 1 218 18919 28 0086 429 20 1 229 20220 30 0086 4 20 1 242 2121 10 003 6 10 1 0252 017322 10 0058 6 10 1 0516 042523 10 0086 6 10 1 0797 069524 10 0113 6 10 1 111 098825 10 0141 6 10 1 146 13526 10 0169 6 10 1 179 16627 10 0197 6 10 1 196 18228 10 0224 6 10 1 219 20429 10 0252 6 10 1 23 21730 10 028 6 10 1 25 23631 1 0086 120 20 3 279 24332 1 0086 120 20 4 39 3133 1 0086 120 20 5 41 34234 1 0086 120 20 9 86 75835 1 0086 120 20 10 100 84736 1 0086 72 12 1 107 0967




Adaptive node

Fixed node

Layer 1

x

y

W1

W2

sum

f


A1

A2

B1

B2

W1

W2

W1 f1

W1 f2


02 025 03 035 04 045 05 055 06 065 07 075 015010050

Maximum depth1

11 microm103 microm

Depth



0025

05075

(microm

)

1

(mm)


(a) (b)



Wlowasti

Wi

W1 + W2 i 1 2 (3)


Wlowasti fi W




f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)







Side wall

Track bottom

(a)

Track bottom

Side wall

(b)









5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls






μAi(x) 1

1 + ( x minus ci1113857( ai1113857( 11138572

1113876 1113877bi

(1)






Number ofshots

Fluence(Jcm2)



Number ofpasses



1 1 0086 120 20 1 0799 06782 2 0086 60 20 1 186 1653 3 0086 40 20 1 279 2494 4 0086 30 20 1 353 3125 5 0086 24 20 1 456 4046 6 0086 20 20 1 573 5157 7 0086 171 20 1 630 5538 8 0086 15 20 1 708 629 9 0086 133 20 1 759 66110 10 0086 12 20 1 881 76311 12 0086 10 20 1 975 8512 14 0086 857 20 1 120 10413 16 0086 75 20 1 131 11414 18 0086 667 20 1 149 13115 20 0086 6 20 1 166 14516 22 0086 545 20 1 193 1717 24 0086 5 20 1 200 17518 26 0086 462 20 1 218 18919 28 0086 429 20 1 229 20220 30 0086 4 20 1 242 2121 10 003 6 10 1 0252 017322 10 0058 6 10 1 0516 042523 10 0086 6 10 1 0797 069524 10 0113 6 10 1 111 098825 10 0141 6 10 1 146 13526 10 0169 6 10 1 179 16627 10 0197 6 10 1 196 18228 10 0224 6 10 1 219 20429 10 0252 6 10 1 23 21730 10 028 6 10 1 25 23631 1 0086 120 20 3 279 24332 1 0086 120 20 4 39 3133 1 0086 120 20 5 41 34234 1 0086 120 20 9 86 75835 1 0086 120 20 10 100 84736 1 0086 72 12 1 107 0967




Adaptive node

Fixed node

Layer 1

x

y

W1

W2

sum

f


A1

A2

B1

B2

W1

W2

W1 f1

W1 f2


02 025 03 035 04 045 05 055 06 065 07 075 015010050

Maximum depth1

11 microm103 microm

Depth



0025

05075

(microm

)

1

(mm)


(a) (b)



Wlowasti

Wi

W1 + W2 i 1 2 (3)


Wlowasti fi W




f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)







Side wall

Track bottom

(a)

Track bottom

Side wall

(b)









5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls






Adaptive node

Fixed node

Layer 1

x

y

W1

W2

sum

f


A1

A2

B1

B2

W1

W2

W1 f1

W1 f2


02 025 03 035 04 045 05 055 06 065 07 075 015010050

Maximum depth1

11 microm103 microm

Depth



0025

05075

(microm

)

1

(mm)


(a) (b)



Wlowasti

Wi

W1 + W2 i 1 2 (3)


Wlowasti fi W




f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)







Side wall

Track bottom

(a)

Track bottom

Side wall

(b)









5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls




Wlowasti

Wi

W1 + W2 i 1 2 (3)


Wlowasti fi W




f sumiWlowasti fi

sumiWlowasti

i 1 2 (5)







Side wall

Track bottom

(a)

Track bottom

Side wall

(b)









5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls








5 10 15 20 25

0

02

04

06

08

1

Number of shots

in1mf1 in1mf2

Deg

ree o

f mem

bers

hip

(a)

01 015 02 025

0

02

04

06

08

1

Fluence

Deg

ree o

f mem

bers

hip in2mf1 in2mf2

(b)

10 20 30 40 50 60 70 80 90 100 110 120

0

02

04

06

08

1

Feedrate

Deg

ree o

f mem

bers

hip

in3mf1 in3mf2

(c)

10 11 12 13 14 15 16 17 18 19 20

0

02

04

06

08

1

Frequency

in4mf1 in4mf2

Deg

ree o

f mem

bers

hip

(d)

Deg

ree o

f mem

bers

hip

1 2 3 4 5 6 7 8 9 10

0

02

04

06

08

1

Number of passes

in5mf1 in5mf2

(e)



5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls




5 Conclusions





Acknowledgments





relativeerror


0

5

10

15

20

25

30

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35

Dep

th o

f abl

atio

n (micro

m)

Experiment number



25

20

15

Pred

icte

d de

pth

of ab

latio

n (micro

m)

10

5

00 5 10






References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls





References





























Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls






Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls




experimentandpredictionofablationdepthinexcimerlaser...

Documents