AutomaticRobotTrajectoryforThermal-Sprayed ComplexSurfacesdownloads.hindawi.com/journals/amse/2018/8697056.pdf · gramming method [7–9]. With the development of the ro-botics, the

Research ArticleAutomatic Robot Trajectory for Thermal-SprayedComplex Surfaces

Dandan Fang12 You Zheng 23 Botao Zhang4 Xiangbo Li5 Pengfei Ju6 Hua Li 4

and Cunnian Zeng2

1Aux Group Ningbo 315201 China2Wuhan University of Technology Wuhan 430000 China3SELF Electronics Co Ningbo 315201 China4Key Laboratory of Marine Materials and Related Technologies Key Laboratory of Marine Materials and Protective Technologiesof Zhejiang Province Ningbo Institute of Materials Technology and Engineering Chinese Academy of SciencesNingbo 315201 China5State Key Laboratory for Marine Corrosion and Protection Luoyang Ship Material Research Institute (LSMRI)Qingdao 266101 China6Shanghai Aerospace Equipment Manufacture Shanghai 200245 China

Correspondence should be addressed to You Zheng zhengyounbuteducn and Hua Li lihuanimteaccn

Received 15 January 2018 Revised 27 March 2018 Accepted 22 April 2018 Published 8 July 2018

Academic Editor Shuo Yin

Copyright copy 2018 Dandan Fang 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

Automatic trajectory generation for thermal spray application is highly desirable for todayrsquos automotive manufacturing Au-tomatic robot trajectory for free-form surfaces to satisfy the coating uniform is still highly challenging due to the complexgeometry of free-form surfaces-e purpose of this study is to present and implement a method for automatic generation of robottrajectory according to the given spray parameters on polygon profile and complex curved free-form surfaces such as torch speedspray distance spray angle and so on -is software development foundation is an Add-In programme of RobotStudio which isoff-line programming and simulation software of ABB Company -e experimental results show that the robot trajectory can begenerated rapidly accurately and automatically on the complex geometries by this method

1 Introduction

Robot manipulators are programmable mechanical systemsdesigned to execute many tasks in industry -e advancedrobot system has been widely used in thermal spray ap-plication in recent years [1ndash3] When a robot is used tra-jectory generation represents an important part of the workbecause the uniformity of spray thickness can significantlyinfluence the quality of coating and the robot trajectory hasa great effect on the uniformity of coating [4ndash6]

Robot trajectory tells the robot how to move around It iscomposed of a series of curves which are defined by the sprayparameters in thermal spray operations and there are severaltargets which are represented by orientation and Cartesianposition on each curve -e generation of robot trajectory onplane surface in thermal spraying is traditionally made by

such methods as point-by-point or teaching-playback -ismethod is tedious and time-consuming -e demand forcoating on curved surface has increased over the last fewyears It requires hundreds of points in the trajectory anddifferent orientation for each point so it is very difficult tocreate such a trajectory by manual point-by-point pro-gramming method [7ndash9] With the development of the ro-botics the robot manufacturers provide software for off-lineprogramming such as RobotStudio which is a simulation andoff-line programming software of ABB Company [10] Avirtual workshop can be elaborated and the robot programcan be prepared and simulated with RobotStudio

-e complete robot trajectory can be generated manuallypoint by point which is a very laborious process and notprecise in RobotStudio Furthermore the functions ofRobotStudio are too narrow to process some complex

HindawiAdvances in Materials Science and EngineeringVolume 2018 Article ID 8697056 11 pageshttpsdoiorg10115520188697056

workpieces for thermal spray applications since the fact thatspraying trajectory is composed of many lines or curves withcertain offset Consequently it is necessary to developa software tool to generate robot trajectory for free-formsurfaces according to the given spray parameters

Most previous researchers typically focused on the meshmethod for automated trajectory generation since suchmethod is capable of generating trajectories on all types ofcomplex workpieces [8 11] In the mesh method the meshof the CAD model is generated at first then the trajectoryfrom the mesh points and normal vectors to the surface arefound out according to graphics theories Unfortunately theoff-line programming software does not provide any ap-plication programming interface (API) for the meshmethod Most of the off-line programming software such asRobotStudio uses the bounding box method to represent 3Dmodels since the bounding box method is efficient in real-time collision detection algorithms In other words de-velopers need to develop various functions and programs ofthe off-line programming software independently For ex-ample developers must transfer the coordinates of points tothe robot system by geometrical transformations Althoughsome mathematical software such as MATLAB could helprealize the functions too much heavy labor is still involvedin software developing and porting

To blend into the off-line programming software thecutting method was proposed and applied to generate robottrajectories in thermal spraying With this method everypoint of curves can be obtained by some special APIs to realizeobject Boolean operations and the trajectories data also canbe calculated and simulated in the off-line programmingsoftware directly It means that the entire developmentprocess with the orthogonal cutting method can be simplerand faster than the process with the method of mesh gen-eration -ermal Spray Toolkit (TST) is one such applicationwhich was developed based on the RobotStudio software forthermal spraying to provide solutions to generate automatictrajectories on various workpieces according to the right sprayparameters and trajectory parameters Some details of TSTforthe simple workpiece were introduced in [12] In this paperwe try to answer the question how to generate robot trajectoryfor complex surfaces automatically

2 Development Environment

21 RobotStudio Software Off-line programming is the bestway to generate robot trajectory on complex geometryABBrsquos simulation and off-line programming softwareRobotStudio allow robot programming to be done on a PCin the office without shutting down production It alsoenables robot programs to be prepared in advance whichincreases overall productivity RobotStudio provides thetools to increase the profitability of a robot system byperforming tasks such as training programming and op-timization without disturbing production -is providesnumerous benefits including risk reduction quicker start-up shorter changeover and increased productivity [10]

RobotStudio is built on the ABB Virtual Controller anexact copy of the real software that runs the robots in

production It thus allows very realistic simulations to beperformed using real robot programs and configuration filesidentical to those used on the shop floor RobotStudio allowscreating Add-In programs that enable users to customizeand extend its functionality -ermal Spray Toolkit is anAdd-In program to extend RobotStudiorsquos functions

Currently there are two different methods for Add-Indevelopment One is creating an Add-In in Visual StudioTool for Applications (VSTA) and the other is creating anAdd-In based on RobotStudio API in Visual Studio

22 RobotStudio API An Application Programming In-terface (API) is an interface implemented by a softwareprogram which enables it to interact with other software Itis similar to the interaction of user interface which facilitatescommunication between humans and computers An API isimplemented by applications libraries and operating sys-tems to determine their vocabularies and calling conven-tions and used to access their services It may includespecifications for routines data structures object classesand protocols used to create communications between theconsumer and the implementer of the API

RobotStudio API shows the RobotStudio Object Modeland explains the methods properties and events for eachobject In the development of -ermal Spray Toolkit theRobotStudio API is used

23Create anAdd-InwithVSTA Add-Ins are programs thatadd optional commands and features to RobotStudio Beforeusing an Add-In RobotStudio must be installed on thecomputer and then loaded in the memory Loading an Add-In program makes the feature available in RobotStudio andadds any associated commands to the appropriate menus

Visual Studio Tool for Applications (VSTA) is includedin RobotStudio and can allow create Add-Ins in a simpleway but there are some limitations when using VSTA tocreate the Add-Ins which are caused by Microsoft ProxyGenerator -ese things will not work

(i) A method that has an array type as parameter(ii) A method or property that references the System

Drawing or System Windows Forms namespaces(iii) A method or property that references the type(iv) An event that is declared static

Since Add-Ins created with Visual Studio do not use theproxy these limitations do not apply

24 Create an Add-In in Visual Studio To be able to use theentire RobotStudio API without limitations Visual Studiocan be used to create an Add-In All the codes and interfacecan be developed in Visual Studio (C or VBNET) andRobotSutdio can load the Add-In program automaticallywhen it is opened -ermal Spray Toolkit was developed inenvironment C which is more powerful than VBNET-especific steps for creating an Add-In in Visual Studio arelisted in Appendix

2 Advances in Materials Science and Engineering

3 Autogeneration of Trajectory

31 3D Model Computer-aided design (CAD) is a well-known computer-based tool that assists engineers architectsand other design professionals in their design activitiesModel-based object representations are known from computergraphics and CAD RobotStudio is based on a 3D model-e model of workpiece can be imported to RobotStudioand trajectory can be generated directly based on the CADmodel in RobotStudio [13] But RobotStudio meets thechallenges in thermal spray application -e function ofRobotStudio is not powerful enough to deal with theBoolean operation of a 3D model especially when con-fronted with the complex surface It is difficult to get a seriesof curves on the complex surface according to the sprayparameters such as the robot velocity the spray distance thespray angle the step between each pass and the over-length-e development of-ermal Spray Toolkit must be based onthe API of RobotStudio As a result -ermal Spray Toolkithas a great flexibility to deal with complex surfaces

32 Parameters of Cold Spray Process -e kinematic pa-rameters of the cold spray process are demonstrated inFigure 1 Spray distance is the distance from the nozzle exitof spray gun to the substrate surface scanning step is thedistance between two neighbor scanning passes spray angleis the angle between the centerline of the spray gun and thesubstrate surface gun velocity is the relative velocity betweenspray gun and substrate [14ndash16]

33 9e Methodology of Generation of Trajectory As men-tioned above the methodology of generation of trajectoryproposed in this paper is based on the cutting methodKnown by the characteristics of thermal spray the uniformscanning step could keep a stable TCP velocity so that theuniformity of coating can usually be guaranteed -at iswhen designing with cutting method the uniformity of thedistance of two neighbor scanning passes should beconsidered

With a polygonal substrate the key of algorithm is tocreate a plane orthogonal with the polygon profile to equallydivide the substrate In this algorithm to create an initialplane Π a point on the polygon should be selected at firstUsing the theory of differential geometry [17] the equationof the polygon is described as

x x(t)

y y(t)

z z(t) αle tle β

(1)

P0(x(t0) y(t0) z(t0)) isin L is the selected pointBased on the differential geometry [17] the tangent

vector τrarr at this point should be

τrarr xprime t0( 1113857 yprime t0( 1113857 zprime t0( 11138571113864 1113865 (2)

And the unit normal vector of the polygon can begenerated easily by the vector product of two noncollinearvectors on the polygon

In the RobotStudio the scroll bar tools can indicate allthe points around the surface for selection the API functionEdgeGetTangent(middot) is able to get the tangent vector on pointP0 and the API function FacesGetNormalToSurface(middot) canget the unit normal vector of a polygon

Particularly when one edge of the polygon is a straight linethe direction vector on the line is just right to be the tangentvector of the point on the line Simply connect the two end-points of that line and the tangent vector can be generated

When the tangent vector on the point of that curve andthe unit normal vector is available the normal vector ofinitial orthogonal plane Π is supposed to be

βrarr

τrarr times nrarr

a b c (3)

-us the initial orthogonal plane equation involvingP0(x(t0) y(t0) z(t0)) isin L on the curve should be

a x(t)minusx t0( 1113857( 1113857 + b y(t)minusy t0( 1113857( 1113857 + c z(t)minus z t0( 1113857( 1113857 0

(4)

After that just move the plane up or down with equaldistance and the trajectory with equal scanning step can begenerated

For plane surface the robot trajectory is simple because theunit normal vector on every point of the surface is the same Butfor curved surface there are many targets on each curve forkeeping the same spray distance Some curved surfaces are verycomplex and unsymmetrical For example as shown inFigure 2 the product is a propeller blade and its spraying areaconsists of small surfaces which have different curvatures (theblue lines represent the normal direction of small surfaces andthe red lines represent the tangent direction of small surfaces)

-e unit normal vectors on surface vary with the surfacecurvature which brings a great challenge to cuttingmethods To keep the uniformity of scanning step as much aspossible first of all select two borders on the substrate andget some equidistant points on these two borders -enchoose a curve with the largest curvature from the surface onthe directions of the two edges and divide it equally to getsome equidistant points -e cutting plane can be created bythe three corresponding points on these three lines Since ona smooth surface the longest curve is usually the curve withthe largest curvature So the curves on the same direction ofthe longest curve can be divided as equally as possible if thelongest curve is divided Also it is difficult to choose

Spray angle

Substart

Scanningstep

Spray gun

Gun velocitySpray distance

Overlength

Figure 1 -e kinematic parameters of cold spray process

Advances in Materials Science and Engineering 3

the curve with the largest curvature In previous research themanual selection method was widely used However thismethod is not convenient or accurate enough Based on thecontour line theory in GIS a method to choose a curve withthe largest curvature automatically with the help of thedensity of contour lines is presented in this paper Accordingto the contour line theory the denser the contour thesteeper the slope in the area and the thinner the contour thesmaller the gradient of the terrain As shown in Figure 3from C1 height to C9 height the contour density along thedirection of L1 is higher than that along the direction of L2and L3 and the path along the direction of L1 is also theshortest In other words the denser the contour alonga certain direction the shorter the path along this direction

In the following section the density analysis method isreplaced with the path analysis method to find the line withthe greatest curvature

Consequently starting from the contour line on the topthe path to the edgewith the denser is the projection of the curvewith the largest curvature For instance there is a hemispheremodel with a radius of 1m in Figure 4 And a plane which isparallel to the substrate is used to cut the sphere in every 01minterval so 11 contour lines can be generated -e distributionof contour lines is as shown in Figure 5

-e result of calculation shows that the length of theshortest path starting from the contour line a is dmin n1 +

n2 + middot middot middot + n11 along with the radius direction of the sphere Itis because the shortest path between two contour lines is alongthe direction of radius As shown in Figure 5 starting frompoint b the length of n2 on the radius direction is shorter thanany other path (eg pathm2) Using the shortest path and theunit normal vector of plane at the bottom a new plane orsurface can be created-e intersection line of the plane or thesurface is the curve with the largest curvature As shown inFigure 6 the curve n is the largest curvature selected on thehemisphere in Figure 4

By cutting planes (polygon cases) or objects (surfacecases) a series of lines could be generated -en find outthe endpoints of each line and calculate the three-dimensional coordinate (x y z) and quaternion-vector(q1 q2 q3 q4) of these endpoints in the world coordinatesystem by some functions -e quaternion-vector (q1 q2q3 q4) is used to represent the orientation of endpoint inthe three-dimensional space

Figure 3 -e contour line map

109080706050403020101 08 06 04 02 0ndash02ndash04ndash06ndash08 ndash1

1080604020ndash02ndash04ndash06ndash08ndash1

Figure 4 A hemisphere model with a radius of 1m

(a)

(b)

Figure 2 Complex curved surface model


Finally the spray parameter over-length is defined-eoretically the over-length is equal to the distance duringwhich the robot accelerates from a halt to a necessary sprayspeed In such a way the scanning speed on the workpiececould be constant and the overheating on the edge could beavoided Furthermore it means obviously a great amount ofpower savings -us the best solution is to keep a constantover-length of each point on the edge of the workpiece Sincedifferent types of robot have different kinematic parametersthe over-length depends on a specific robot type It is verydifficult to find the exact position of these over-lengths onreal workpiece in the work cell even in off-line pro-gramming under RobotStudio

For clarity of description of the algorithm above twoexamples using an arbitrary polygon and a curved surface togenerate trajectory are used for illustration respectively

34 Automatic Generation on a Polygon Profile -e methodused for automatic generation on a curved-edge polygon

profile is as follows and the sketch map is shown inFigure 7

(1) Click on the surface of CAD model workpiece in thegraphics window

(2) Pull the scroll bar which indicates all the pointsaround the surface the direction of tangent of thispoint is the direction of trajectory and then choosethe trajectory starting from left or right

(3) Create a large plane involving two points to intersectthe spraying surface and ensure the large plane andthe spraying surface are orthogonal Create the firstintersection line

(4) According to the value of step move down the largeplane and create a series of lines

(5) Create targets on both the ends of series of linesaccording to the interlength Calculate the robotvelocity and orientation for each target

(6) Add all the targets to trajectory in order

-is function is not only suitable to round and ellipsebut also suitable to many irregular curved-edge polygonsurfaces With the straight-edge polygon profile thestraight edge is selected instead of the point on the edge Inthis way autogeneration of trajectory could be simpler-e method used for automatic generation on a straight-edge polygon profile is as follows and the sketch map isshown in Figure 8

(1) Click on the surface of the CAD model workpiece inthe graphics window all the edges of the surface willbe displayed in the list

(2) Select one edge as the direction of trajectory andchoose the trajectory starting from left or right

(3) Create a large plane on the edge to intersect thespraying surface and ensure that the large plane andthe spraying surface are orthogonal Create the firstintersection line

(4) According to the value of step move down the largeplane Create a series of lines



35 AutoGeneration for Curved Surface -e followingpresents the method of autogeneration for a curved surfaceand the sketch map is shown in Figure 9

(1) Create two edges of the complex 3D coating surface(2) -e two edges are divided into equal parts according

to the spray parameters(3) Find the line with the greatest curvature of the

complex 3D coating surface on the directions of thetwo edges and divide this line equally to get someequidistant points Create a large plane through thecorresponding points on the edges and the line with

10908070605040302010108 06 04 02 0ndash02ndash04ndash06ndash08 ndash1


n

Figure 6 -e largest curvature n selected on the hemisphere

08

06

04

02

0

ndash02

ndash04

ndash06

ndash08

ndash1 ndash08 ndash06 ndash04 ndash02 0 02 04 06 08 1

a

n1

n2m2

n3

n4n5n6n7n8

b

Figure 5 -e contour lines of the hemisphere model


the greatest curvature to intersect this 3D modelCreate the first closed intersection curve -is step issuitable for a smooth curved surface For an irregularcurved surface manual adjustment is necessary toensure that these closed intersection curves areequidistant

(4) Offset the large plane to the next correspondingpoints on the edges and the line with the greatestcurvature to create a series of closed intersectioncurves

(5) Choose the part of closed intersection curves whichare located on the spraying area

(6) Create targets on the series of selected curvesaccording to the interlength Calculate the robotvelocity and orientation for each target Add all thetargets to trajectory in order

36 Calculation of Velocity Robot velocity plays a signifi-cant role in coating quality A robot usually has a continual

motion and robot velocity is planned as constant as possibleto obtain a uniform thickness if the step is constant But withthe methodology used in thermal spray toolkit the step isvaried so the robot velocity should change with the step Intheory the robot velocity is inversely proportional to thestep v2v1 d1d2 Figure 10 shows the meaning of char-acter v1 v2 d1 and d2 -e robot velocities are calculated inthe Add-In program and assigned to corresponding targetsin order to keep the uniformity

4 Program Implementationand Experimentations

To validate the methodology and visualize the trajectories anAdd-In programme with interface has been developed in theC environment and it can be loaded automatically whenopening the RobotStudio -e function of the program hasbeen tested using a different 2D model and 3D modelFurthermore the real experiment has been implemented toevaluate the effect of autogeneration of trajectory

(1) (2) (3)

(4) (5) (6)

Figure 7 Autogeneration on a curved-edge polygon profile

(1) (2) (3)

(4) (5) (6)

Figure 8 Autogeneration on a straight-edge polygon profile


41 Implementation on a Straight-Edge Polygon Profile-e following figures provide some typical examples toillustrate the function of autogeneration of trajectory ona polygon profile According to the thermal spray char-acteristics it can be clearly seen that the first line and thelast line of trajectory are located outside the workpiece inorder to ensure the uniformity of coating Figure 11 showsthe autogenerated trajectory on a type of regular polygonsand Figure 12 shows the autogenerated trajectory ona type of irregular polygons

42 Implementation on a Curved-Edge Polygon Profile Asseen from Figure 13 the autogenerated trajectory on dif-ferent kinds of curved-edge polygons can be realized by TST-e beginning part and the end part of the trajectory areprolonged to a certain distance as the start and end of torchrespectively

43 Implementation of on Free-Form Surfaces Figure 14shows the program implementation of autogeneration oftrajectory on free-form surfaces As shown in the figure thefunction can process various complex 3D workpieces evenpropeller blades

44 Experiments and Analysis It is well known that thescanning speed is the most important parameter for thecoating homogeneity during the deposition So the ex-periment is implemented to analyse the TCP velocityaccording to the selected robot trajectory As a professionalrobot off-line programming platform RobotStudio is ableto supply the kinematics data in simulations which arehighly close with the data generated in actual motions ofrobots -e simulation process of RobotStudio enablesusers to obtain the real-time value of TCP position speedand the rotation angle of robot axis and to implement thecollision detection between robot torch and workpiece Avirtual experimental environment which is completelyconsistent with the real experimental environment is

(1) (2) (3)

(4) (6)(5)

Figure 9 Autogeneration for complex curved surface

v1d1

d2

v2

Figure 10 Meaning of character v1 v2 d1 and d2


(a) (b) (c)

Figure 11 Autogenerated trajectory on a type of regular straight-edge polygons

(a) (b) (c)

Figure 12 Autogenerated trajectory on a type of irregular straight-edge polygons

(a) (b) (c)

Figure 13 Autogenerated trajectory on a type of curved-edge polygons


created as shown in Figure 15 which simulated the sprayprocess to get the record of TCP velocity -e shape of thesubstrate surface is described in Figure 16 A series ofcurves and several targets in the robot trajectories gener-ated by TST are shown in Figures 17(a) and 17(b)respectively

AF4-MB type torch is guided by a robot of IRB4400 -escanning step is 0022m and the over-length is 003mDuring the simulation process the real-time TCP velocityand position are recorded on each trajectory point -esimulated TCP velocity and the calculated TCP velocity forrobot trajectory are shown in Figure 18 respectively

When passing through the complex curved surface eachpass has 5 samples and the starting point is on the left of themodel As mentioned in the calculation of velocity sectionthe TCP velocity on first pass is the reference velocity whichequals to 300mms in this experiment -eoretically the

(a) (b) (c) (d)

Figure 14 Autogeneration of trajectory on free-form surfaces

Figure 15 -e virtual experimental environment

Figure 16 -e shape of the substrate surface

(a)

(b)

Figure 17 A series of curves (a) and several targets (b) in the robottrajectories


robot velocity is inversely proportional to the scanningstep Since the left border of the model is wider than theright border of the model the TCP velocity of the passfrom right to left should be increasing In the same way theTCP velocity of the pass from left to right should be de-creasing -e simulated TCP velocity (blue line) in Figure 18is basically close to the calculated TCP velocity (red line)which is inversely proportional to the scanning step It meansthat the method for autogeneration trajectory works wellfor thermal spray application -e simulated TCP velocitydecreased by about 10 less than the calculated TCP ve-locity that corresponds to the passage of torch at therounded area of the workpiece where the rotation is themost difficult

5 Conclusion

Autogeneration trajectory is needed to operate the sprayprocess on a free-form surface but no off-line programmingsystem has yet been developed to meet this need In thispaper -ermal Spray Toolkit was developed based onRobotStudio it is flexible and robust enough to deal with anycomplex 2D and 3Dmodel and the robotic trajectory can begenerated automatically according to the spray parameterswith high efficiency

Appendix

(1) Create a new project with Class Library as yourtemplate

(2) To use the RobotStudio API you need to reference theABBRoboticslowastdll assemblies -ese are as follows

ABBRoboticsMathmdashVector and matrix mathABBRoboticsRobotStudiomdashGeneral and top-level(eg nonstation specific) classesABBRoboticsRobotStudioEnvironmentmdashFormanipulating the GUI for example adding menusbuttons and so onABBRoboticsRobotStudioStationsmdashFor manipu-lating stations and their contents

ABBRoboticsRobotStudioStationsFormsmdashGUIcontrols used by RobotStudioYou can add a new reference to your project byright click on References in the Solution Explorerand choose Add ReferenceGo to the Browse tab and browse to your Robot-Studio installation directory -is is where you willfind the dll

(3) Add the following lines to the top of your code

CopyCusing ABBRoboticsMathusing ABBRoboticsRobotStudiousing ABBRoboticsRobotStudioEnvironmentusing ABBRoboticsRobotStudioStationsusing ABBRoboticsRobotStudioStationsForms

(4) Add the following code to the class

CopyCpublic static void AddinMain()-is is where you write your code

-e AddinMain()-method is the entry point foryour Add-In

(5) Go to the properties of your project (right click onyour project in the Solution Explorer and selectProperties) and select the Build Events Add thefollowing line to the Post-built event command line

XCopyy ldquo$(TargetPath)rdquo ldquoCProgram FilesABBIndustrial ITRobotics ITRobotStudio 512BinAddinsrdquoAssuming that this is the directory for yourRobotStudio installation-is will copy your Add-In dll-file to the Robot-Studio Add-In directory when building your Add-In

(6) You could also add the path to RobotStudioexe toStart external program under Debug in Properties soRobotStudio will start every time you want to debugyour Add-In

(7) Write your code (or copy and paste one of the ex-amples into AddinMain())

(8) Select Start Debugging from Debug menu (or pressF5) -is will start RobotStudio and autoload yourAdd-In Since your Add-In will load and executedirectly when RobotStudio starts (even though there isnot active station) it will probably throw an exceptionIt might be a good idea to add a menu button to startyour code Have a look at the AddMenus and ButtonsExample for how to create menus and buttons

(9) If you go to the Add-Insmenu you can find your Add-In under the General folder If you right click on yourAdd-In you can select whether or not it shouldautoload and you can load it if it is not already loaded

RobotStudio Add-Ins can be created in VBNETor C

500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated

Figure 18 -e simulated TCP velocity and the calculated TCPvelocity for robot trajectory


Data Availability

-e data used to support the findings of this study are notavailable from the corresponding author upon request dueto the commercial interests in the research

Conflicts of Interest

-e authors declare that there are no conflicts of interestregarding the publication of this article

Acknowledgments

-is work was supported by the National Natural ScienceFoundation of China (31500772 21705158 51401127 and51771122) Key Research and Development Program of Zhe-jiang Province (2015C01036 and 2107C01003) Zhejiang Pro-vincial Natural Science Foundation of China (LQ14F010001LY15F010006 LY16F10014 LQ15F020004 and LY18F020025)Key Scientific and Technological Projects of Ningbo(2014B92001) Postdoctoral Science Foundation of ZhejiangProvince (BSH02020) Ningbo Natural Science Foundation(2017A610109 and 2015A610132) Research Fund of State KeyLaboratory for Marine Corrosion and Protection of LuoyangShip Material Research Institute (LSMRI) under the Contractno KF160409 International Scientific and TechnologicalCooperation Project of Ningbo and Science and Tech-nology Commission of Shanghai Municipality (Grant nos16QB1401100 and 15QB1401500)

References

[1] R Gadow and M Floristan ldquoManufacturing engineering inthermal spraying by advanced robot systems and processkinematicsrdquo in Future Development of 9ermal SprayCoatings pp 259ndash280 Elsevier New York NY USA 2015

[2] N Espallargas Future Development of 9ermal Spray Coat-ings Types Designs Manufacture and Applications ElsevierNew York NY USA 2015

[3] S Deng and C Chen ldquoGeneration of robot trajectory forthermal sprayrdquo in Robot Kinematics and Motion PlanningNova Science Publishers Hauppauge NY USA 2015

[4] M Floristan J A Montesinos J A Garcıa-Marın A Killingerand R Gadow Robot Trajectory Planning for High Quality9ermal Spray Coating Processes on Complex Shaped Compo-nents ASM International Geauga County OH USA 2012

[5] D Fang S Deng H Liao and C Coddet ldquo-e effect of robotkinematics on the coating thickness uniformityrdquo Journal of9ermal Spray Technology vol 19 no 4 pp 796ndash804 2010

[6] R Gadow A Candel and M Floristan ldquoOptimized robottrajectory generation for thermal spraying operations andhigh quality coatings on free-form surfacesrdquo Surface andCoatings Technology vol 205 no 4 pp 1074ndash1079 2010

[7] B Zhou X Zhang ZMeng and X Dai ldquoOff-line programmingsystem of industrial robot for spraying manufacturing optimi-zationrdquo in Proceedings of the 33th Chinese Control Conferencepp 8495ndash8500 Nanjing China July 2014

[8] A Candel and R Gadow ldquoTrajectory generation and couplednumerical simulation for thermal spraying applications oncomplex geometriesrdquo Journal of 9ermal Spray Technologyvol 18 no 5-6 pp 981ndash987 2009

[9] D Hegels T Wiederkehr and H Muller ldquoSimulation basediterative post-optimization of paths of robot guided thermalsprayingrdquo Robotics and Computer-Integrated Manufacturingvol 35 pp 1ndash15 2015

[10] ABB Robotics RobotStudiotrade Users Guide ABB RoboticProducts Zurich Switzerland 2015

[11] P N Atkar A Greenfield D C Conner H Choset andA A Rizzi ldquoUniform coverage of automotive surfacepatchesrdquo International Journal of Robotics Research vol 24no 11 pp 883ndash898 2005

[12] S Deng H Liao and C Coddet ldquoRobotic trajectory auto-generation in thermal sprayingrdquo in Proceedings of the9ermalSpray Connects Explore Its Surfacing Potential Basel SwitzerlandMay 2005

[13] R Holubek D R Delgado Sobrino P Kostal and R RuzarovskyldquoOffline programming of an ABB robot using imported CADmodels in the RobotStudio software environmentrdquo AppliedMechanics and Materials vol 693 pp 62ndash67 2014

[14] K Wen M Liu K Zhou et al ldquo-e influence of anode innercontour on atmospheric DC plasma spraying processrdquo Ad-vances in Materials Science and Engineering vol 2017 ArticleID 2084363 12 pages 2017

[15] P Ctibor M Kasparova J Bellin E Le Guen and F ZahalkaldquoPlasma spraying and characterization of tungsten carbide-cobalt coatings by the water-stabilized system WSPrdquo Ad-vances in Materials Science and Engineering vol 2009 ArticleID 254848 11 pages 2009

[16] L Pawlowski 9e Science and Engineering of 9ermal SprayCoatings John Wiley amp Sons Hoboken NY USA 2nd edition2008

[17] V Y Rovenski Differential Geometry of Curves and SurfacesPrentice-Hall Inc Upper Saddle River NY USA 2004


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

workpieces for thermal spray applications since the fact thatspraying trajectory is composed of many lines or curves withcertain offset Consequently it is necessary to developa software tool to generate robot trajectory for free-formsurfaces according to the given spray parameters

Most previous researchers typically focused on the meshmethod for automated trajectory generation since suchmethod is capable of generating trajectories on all types ofcomplex workpieces [8 11] In the mesh method the meshof the CAD model is generated at first then the trajectoryfrom the mesh points and normal vectors to the surface arefound out according to graphics theories Unfortunately theoff-line programming software does not provide any ap-plication programming interface (API) for the meshmethod Most of the off-line programming software such asRobotStudio uses the bounding box method to represent 3Dmodels since the bounding box method is efficient in real-time collision detection algorithms In other words de-velopers need to develop various functions and programs ofthe off-line programming software independently For ex-ample developers must transfer the coordinates of points tothe robot system by geometrical transformations Althoughsome mathematical software such as MATLAB could helprealize the functions too much heavy labor is still involvedin software developing and porting

To blend into the off-line programming software thecutting method was proposed and applied to generate robottrajectories in thermal spraying With this method everypoint of curves can be obtained by some special APIs to realizeobject Boolean operations and the trajectories data also canbe calculated and simulated in the off-line programmingsoftware directly It means that the entire developmentprocess with the orthogonal cutting method can be simplerand faster than the process with the method of mesh gen-eration -ermal Spray Toolkit (TST) is one such applicationwhich was developed based on the RobotStudio software forthermal spraying to provide solutions to generate automatictrajectories on various workpieces according to the right sprayparameters and trajectory parameters Some details of TSTforthe simple workpiece were introduced in [12] In this paperwe try to answer the question how to generate robot trajectoryfor complex surfaces automatically

2 Development Environment

21 RobotStudio Software Off-line programming is the bestway to generate robot trajectory on complex geometryABBrsquos simulation and off-line programming softwareRobotStudio allow robot programming to be done on a PCin the office without shutting down production It alsoenables robot programs to be prepared in advance whichincreases overall productivity RobotStudio provides thetools to increase the profitability of a robot system byperforming tasks such as training programming and op-timization without disturbing production -is providesnumerous benefits including risk reduction quicker start-up shorter changeover and increased productivity [10]

RobotStudio is built on the ABB Virtual Controller anexact copy of the real software that runs the robots in

production It thus allows very realistic simulations to beperformed using real robot programs and configuration filesidentical to those used on the shop floor RobotStudio allowscreating Add-In programs that enable users to customizeand extend its functionality -ermal Spray Toolkit is anAdd-In program to extend RobotStudiorsquos functions

Currently there are two different methods for Add-Indevelopment One is creating an Add-In in Visual StudioTool for Applications (VSTA) and the other is creating anAdd-In based on RobotStudio API in Visual Studio

22 RobotStudio API An Application Programming In-terface (API) is an interface implemented by a softwareprogram which enables it to interact with other software Itis similar to the interaction of user interface which facilitatescommunication between humans and computers An API isimplemented by applications libraries and operating sys-tems to determine their vocabularies and calling conven-tions and used to access their services It may includespecifications for routines data structures object classesand protocols used to create communications between theconsumer and the implementer of the API

RobotStudio API shows the RobotStudio Object Modeland explains the methods properties and events for eachobject In the development of -ermal Spray Toolkit theRobotStudio API is used

23Create anAdd-InwithVSTA Add-Ins are programs thatadd optional commands and features to RobotStudio Beforeusing an Add-In RobotStudio must be installed on thecomputer and then loaded in the memory Loading an Add-In program makes the feature available in RobotStudio andadds any associated commands to the appropriate menus

Visual Studio Tool for Applications (VSTA) is includedin RobotStudio and can allow create Add-Ins in a simpleway but there are some limitations when using VSTA tocreate the Add-Ins which are caused by Microsoft ProxyGenerator -ese things will not work

(i) A method that has an array type as parameter(ii) A method or property that references the System

Drawing or System Windows Forms namespaces(iii) A method or property that references the type(iv) An event that is declared static

Since Add-Ins created with Visual Studio do not use theproxy these limitations do not apply

24 Create an Add-In in Visual Studio To be able to use theentire RobotStudio API without limitations Visual Studiocan be used to create an Add-In All the codes and interfacecan be developed in Visual Studio (C or VBNET) andRobotSutdio can load the Add-In program automaticallywhen it is opened -ermal Spray Toolkit was developed inenvironment C which is more powerful than VBNET-especific steps for creating an Add-In in Visual Studio arelisted in Appendix







x x(t)

y y(t)

z z(t) αle tle β

(1)








βrarr

τrarr times nrarr

a b c (3)



(4)




Spray angle

Substart

Scanningstep

Spray gun


Overlength













(a)

(b)


























n


08

06

04

02

0

ndash02

ndash04

ndash06

ndash08


a

n1

n2m2

n3

n4n5n6n7n8

b











(1) (2) (3)

(4) (5) (6)


(1) (2) (3)

(4) (5) (6)







(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls









x x(t)

y y(t)

z z(t) αle tle β

(1)








βrarr

τrarr times nrarr

a b c (3)



(4)




Spray angle

Substart

Scanningstep

Spray gun


Overlength













(a)

(b)


























n


08

06

04

02

0

ndash02

ndash04

ndash06

ndash08


a

n1

n2m2

n3

n4n5n6n7n8

b











(1) (2) (3)

(4) (5) (6)


(1) (2) (3)

(4) (5) (6)







(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls














(a)

(b)


























n


08

06

04

02

0

ndash02

ndash04

ndash06

ndash08


a

n1

n2m2

n3

n4n5n6n7n8

b











(1) (2) (3)

(4) (5) (6)


(1) (2) (3)

(4) (5) (6)







(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls



























n


08

06

04

02

0

ndash02

ndash04

ndash06

ndash08


a

n1

n2m2

n3

n4n5n6n7n8

b











(1) (2) (3)

(4) (5) (6)


(1) (2) (3)

(4) (5) (6)







(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls












(1) (2) (3)

(4) (5) (6)


(1) (2) (3)

(4) (5) (6)







(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls








(1) (2) (3)

(4) (6)(5)


v1d1

d2

v2



(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls




(a) (b) (c)


(a) (b) (c)


(a) (b) (c)






(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls







(a) (b) (c) (d)




(a)

(b)




5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls





5 Conclusion


Appendix

















500

450

400

350

300

2501 6 11 16 21 26 31 36 41 46

Trajectory point

TCP

velo

city

(mm

s)

SimulatedCalculated



Data Availability




Acknowledgments


References





















Advances in



Journal of

Chemistry















Journal of





Volume 2018









Journal of


Na

nom

ate

ria

ls




Data Availability




Acknowledgments


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




Documents

AutomaticRobotTrajectoryforThermal-Sprayed ComplexSurfacesdownloads.hindawi.com/journals/amse/2018/8697056.pdf · gramming method [7–9]. With the development of the ro-botics, the