Home

VDM RoAD Software - Deep Blue

1. 55500 TIT E Dreta oO e TIT mes A AE EE eee Wie ee oe ee a ae He yi Ne ee pi eS MARNAS ak Ks Le f 7 paper EGS v a h a EA X coordinate m se eee Figure 7 Example plots of simulation results Figure 8 shows four more plots from the same run Here are plots of the vertical tire forces and the vehicle roll angles for the tractor and semitrailer Vehicle speed is also shown overlaid with the target speed shown earlier as an input in Figure 5 VDM ROAD includes more than 30 predefined plot setups such as the eight shown in Figure 7 and Figure 8 Further users can define new plots involving any combination of the hundred or so variables stored on disk whenever a simulation is run Fz vert forces R side vs S WB 12 AR 3 R lane Wheel load N 500 1000 1600 Station number m a Fz wert forces L side vs WB 12 At3 R lane d Wheel load N Ruheel axe 1 if rho 3 Rune axle 2 f Runeel axe 3 i Ruheel axe 4 1 Longitudinal velocity kph Luhed ande PE 4 Lwheel axle 2 A 100 t 7 L wheel axe 3 E choc Lwheel ale 4 Tractor a Semitrailer i E gt H anataat iatan tires EE Vx vs Station WB 12 At3 R lane wF 30 l nE ot sod 1000 Station number Figure 8 More plots for a simulation run Software Architecture VDM ROAD is made up of four tightly integrated types of software that are shown conceptually
2. Technical Summary In the late 1980s the Federal Highway Administration FHWA initiated a multiyear program to develop an interactive highway safety design model IHSDM that systematically considers safety of roadway and roadside design elements in creating cost effective highway design alternatives Research was conducted at The University of Michigan Transportation Research Institute UMTRI under FHWA Contract DTFH 61 93 R 00142 to select vehicle dynamics models and begin the process of adding them to the IHSDM The existing TruckSim and AutoSim technologies were selected and extended for this project The main product of the research is a comprehensive software package called VDM RoAD Vehicle Dynamic Models Roadway Analysis and Design VDM RoAD was developed to simulate the vehicle dynamics behavior of cars trucks and two unit combination vehicles tractor semitrailers on 3D road designs This software runs under Windows 95 and Windows NT A full description of VDM RoAD along with instructions for its use is provided in a 350 page reference manual VDM RoAD User Reference Manual Version 1 0 VDM ROAD provides an advanced simulation of vehicle dynamics that works with the 3D road design data In just a few minutes an engineer can simulate the behavior of a vehicle driving over several kilometers of a new roadway design Results are viewed immediately in the form of X Y plots and wire frame animation The new software is
3. Y Y N Y X 3 INTEGRATION WITH ROAD DESIGN DATA This section provides an overview of the road geometry used in the VDM RoAD models and then describes how the road design data are integrated with the vehicle dynamics models The road design data are used to perform several functions in the VDM ROAD design 1 to provide a continuous 3D surface that supports the vehicle and interacts with the tires 2 to provide a horizontal path as a reference for the closed loop steering model and 3 to determine the location and orientation of the vehicle at the start of a simulation run The location and orientation are part of the vehicle initial conditions Using road design data together with a vehicle dynamics model requires the information in the roadway coordinate system to be compatible with the vehicle coordinate system The following subsections describe the roadway and vehicle coordinate systems and the general methods used to transform information between them After the roadway coordinate system has been described the section concludes by describing how the above three functions are accomplished Coordinate Systems The describing dynamical equations for a vehicle are solved by the simulation programs in VDM RoAD to provide coordinates of critical points in the vehicle based on a global X Y Z coordinate system The road design data from the IHSDM data file is also tied to a global X Y Z coordinate system The global sy
4. design CAD and finite element analysis FEA programs into FHWA s research and design activities An integrated design process was conceived that would use modern computer technology to simultaneously take into account significant factors of performance and safety 1 The road design is stored in a commercial CAD file format such as Intergraph s MICROSTATION or Autodesk s AUTOCAD As shown in Figure 1 other modules are envisioned to analyze proposed designs and provide the designer with rapid feedback to identify possible problems Design Benefit Cost Ree on Ge Roadside Safety Module Structure Module Consistency Vehicle Dynamics Module Module Commercial CAD Package Policy Review Module Driver Performance Fi IS Module Accident Predictive Traffic Module Revised Module Alternatives Figure 1 Concept of the IHSDM This report documents the research conducted at The University of Michigan Transportation Research Institute UMTRI under FHWA Contract DTFH 61 93 R 00142 The objectives of the research were to select vehicle dynamics models and link them through a common database to standard commercial roadway CAD programs The software architecture is designed for flexibility to enable linking with a driver performance model in the near future The design also allows integration with finite element and other commercially available modules TOR tate Project Deliverables The project involved the following
5. PAR The PAR files transfer user settings from the SGUI to the animator The animator program in VDM RoAD is also used in other software packages and is updated often The most recent version can be downloaded via the Internet from the web site http www trucksim com animator html ay Origin of global coordinate System ane re y A H e lt n Origin of moving coordinate system wet we we we r e F TA E Camera A point Ca le n Figure 13 Geometry of the camera point and the look point PAR files ERD file Animator set up and shape Motion information from information from database simulation programs Animator Figure 14 Animator input files N y e The Plotter Plots of the VDM ROAD simulation results are viewed with a program called Windows Engineering Plotter WinEP WinEP is a versatile tool that can be used to plot any two variables against each other It can also overlay data from the same file or different files WinEP has a workspace defined by a main window with a menu bar The main window contains plot windows each showing plots made with a single set of X Y axes Figure 15 shows the main WinEP window with three example plot windows pu aie 523 gt WB 20 Alt 3 R lane Ay vs Station WB 20 Alt 3 R lane Lateral acceleration g s Y vs X trajectory WB 20 AIt3 R lane Y position m 117400 ese eneenees 117000 i ARA ARAS esebe
6. Q 8 a So SO N Q s a 3 0 N gt gt uw o O DD ce ae 3 O hand port b user can view poin PEKAN Se SONANDO E PAPP aidi LESS AENA A a aadd sa RET A ANNAR NAAMAA A a aa i k ABIA A a g AE DO Si s WENER O EEIN ervei te ER chek weri moreg ERA MM cde 6s ap ign anita POAC ais ponia gos P mm ELTON o EON a IA da CT w siana ina naene is Ole A FOARREN TAO A ESA iy PARRE ne neare NN j sari it pashan peep onan AEE k ping erro ae poe ares ee a T 15 MI PU PY AAA jm ico cer DA RN AAN uae Figure 6 Animation of simulation results The user clicks the Plot button also located in the right hand portion of Figure 3 to view engineering plots of the simulation results For example Figure 7 shows four plots made this way The upper left plot shows lateral acceleration experienced by the tractor and trailer along with the lateral acceleration defined by the road geometry and vehicle speed See Appendix A for a discussion of the lateral acceleration due to the road The upper right plot shows the steering wheel angle needed to navigate the road with the specified vehicle speed and lateral position The plot underneath lower right shows the tracking of each wheel in the simulated vehicle This plot summarizes the tracking requirements for the combination of vehicle and speed for the entire section of road station O to 1900 PF Wink La
7. deliverables in addition to this report e The main deliverable is a comprehensive software package called VDM RoAD Vehicle Dynamic Models Roadway Analysis and Design VDM RoAD was developed to simulate the vehicle dynamics behavior of cars trucks and two unit combination vehicles tractor semitrailers on 3D road designs This software runs under Windows 95 and Windows NT e A full description of VDM ROAD along with instructions for its use is provided in a 350 page reference manual VDM ROAD User Reference Manual Version 1 0 2 e Two copies of the commercial simulation program TruckSim were delivered TruckSim supports vehicle dynamics studies that focus more on the vehicle and less on the road design properties e Two HP workstations were purchased for the project One was shipped to the Turner Fairbank Highway Research Center in May 1995 and the other has been used at UMTRI as an Internet server for road profiles and public domain software maintained by UMTRI e UMTRI prepared presentations software demonstrations and an exhibit for the TRB FHWA 3D Transportation Visualization and Simulation Symposium Workshop that was held in Houston during May 1995 e A UNIX program called RAPID COMBINE was created and delivered in 1995 to combine vehicle response properties with road design data and speeds e A UNIX version of the VDM RoAD was created However the UNIX version is less developed than the Windows version O
8. in Figure 9 1 Data screens serve as the primary interface to VDM ROAD They contain vehicle model parameters control inputs run settings and links to IHSDM files with road geometric data The data screens are part of a database with many libraries of related data sets 2 Simulation programs numerically solve equations of motion 1 e mathematical vehicle models to calculate output variables The process of performing these calculations is called making a simulation run These programs are applied automatically from the data screens when you click a run button 3 An animator shows the computed vehicle motions using wire frame shapes for the vehicle and road The user can view the simulated motions zoom in and out with a virtual camera and interactively move around the simulated vehicle to change point of view Y Oss 4 The Windows Engineering Plotter WinEP creates plots of vehicle variables as functions of time or as cross plots of output variables This tool is used to view any of the hundreds of variables computed by the simulation models om VDM ROAD Startup Vehicle Dynamics Solver Programs Lat Pos vs Station All Axles A BUS Alt 3 A lane Te ene Oe Wire frame animator Engineering Plotter Figure 9 Four parts of VDM RoAD Most user interactions with VDM RoAD involve preparing data sets for simulation runs or viewing outputs of simulation runs In either case the user deals exte
9. in cars and heavy trucks The kinematical and dynamical equations are valid for full nonlinear 3D motions of rigid bodies VDM RoAD uses closed loop controller algorithms to steer the vehicle along the specified path and to drive at the prescribed speed he simulations run quickly given the level of detail in the models On Pentium Pro computers the programs run in real time or faster In other words a run simulating a two minute test will finish in less than two minutes ght button he ri te int ima trols located he An ing t click left part of screen S x seas ee LS EOS gt Resse Rees SE SS E ae seiner ee T See 3 ES PAURE y 900 1000 800 500 600 700 Station number m 400 300 icle speed Target veh 5 Figure iewed us For example the results are v gt f the screen see Figure ing con 3 f the run as shown in Figure 6 From within the animator the imation O ider control lower Other controls allow the user to change A sl 10r in real time le behav 1C 200 Closed Loop Speed Control 100 0 the veh 10n O Input Target forward speed ih After a run has been made fl ngs up an an x z CERES E ERES SS PERCE Ewe sense rece eS Se A RA NIDO e PER TAE CARA RARA En Az pa a 3 2 O wn ES g Bo Ss o pa S gt o Q S O Se i 2 a SS 28 gt E
10. no user interface other than the display of text and the acceptance of keyboard entries Figure 12 shows ay ee that they have the same appearance as a plain DOS text program However the solvers are technically 32 bit Windows programs axle tractor with i axle trailer TruckSim braking and handling model Version 4 0 2 JIN LB Units Generated by AutoSim June 27 1997 isimfile found batch operation Input file with parameter values C TRUCKSIM TNARUNSA492 LPI Echo file with initial conditions C TRUCKSIM IN RUNSS492 LPO0 Echo file with final conditions C ATRUCKSTM IN RUNSS492 LPF Output file with time history data C TRUCKSIM IN RUNSS492 ERD purpur file with log of nested inputs C TRUCKSIM INARUNSA492 LOG Figure 12 Screen display when solver program is running The solver programs contain equations of motion for the vehicle model in the form or ordinary differential equations The equations are far too complex to solve in closed form so numerical integration is used The source code for each simulation program is generated using the AutoSim commercial multibody program 7 All of the equations related to the kinematics and dynamics of rigid bodies are derived and assembled automatically However equations for the tires spring driver model and road geometry are contained in hand written subroutines that are linked with the AutoSim generated code This approach allows highly complex mechanical systems to
11. on Intel PCs equipped with Windows 95 or Windows NT It is self contained requiring no additional programs or tools to function The most striking feature of VDM RoAD is the ease of use The software can be used by engineers who are no experts in simulation or even vehicle dynamics This contrasts with older simulation packages Vehicle dynamics programs have traditionally offered mechanical engineers the ability to perform virtual tests on the computer to study the dynamics of vehicles in limited test conditions Considerable expertise was needed in the fields of vehicle dynamics and numerical analysis Further the time needed to become proficient with simulation software has always been so lengthy that only the most determined engineers would become productive Overview of VDM RoAD Operation VDM ROAD can be used to evaluate highway designs as they would be experienced by different vehicles traveling over a variety of speeds VDM ROAD includes twelve example vehicles from the AASHTO Green Book 3 It also includes data sets for a car Ford Taurus and heavy truck Those data sets were based on extensive laboratory measurements to support research at the National Highway Traffic Safety Administration NHTSA involving the validation of computer models of vehicle dynamics Figure 2 shows a screen dump of the main screen display in VDM RoAD The pull down menu lists some of the example runs that are installed A user can browse through the datab
12. sets in the library aneen E pa Ho Pra on tax E gdp le Data Screen A view of one data set though the graphical user interface Figure 10 Contents of a VDM ROAD library Data Screens Figure 11 shows an example data screen Each data screen has three kinds of elements with which the user interacts ns ee Yellow fields These contain data that can be edited directly For example to change wheelbase find the yellow field with the current wheelbase value 3 click on the field and change the value using the mouse and keyboard 2 Buttons All of the library screens include buttons at the top to quickly navigate through the data sets in the library and to go to other libraries and programs in VDM ROAD see the buttons near 2 3 Pull down menus Some of the buttons are marked with a triangle that pressing them with the left mouse button brings up a pull down menu The yellow field in the upper left corner of the screen is the title of the data set AASHTO 3 Axle Tractor 1 The title is simply text chosen by the user to identify this data set in the library It has no significance to the simulation Figure 11 Example data screen The Simulation Solver Programs VDM ROAD contains programs that solve the equations of motion for vehicle models predicting motions forces and other variables The solver programs are so called console applications Each behaves as a terminal console with
13. 40 0 00 0 00000 253 257 54376 719 117196 281 41 029 0 00 0 00000 277 836 54398 111 117184177 40 377 0 00 0 00000 283 059 54402 657 117181 605 40 239 155 00 44 50016 303 260 54420 596 117172 928 39 704 00 44 50016 312 861 54429 842 117169 506 39 472 00 44 50016 373 641 54489 751 117161 885 39 00 44 50016 443 54519 100 117166 3 00 0 938 54587 192 117185 000000000 Figure 4 Road geometry Another simulation input is a speed for the vehicle As shown in Figure 5 the speed is not necessarily a constant The user can specify speed as a function of station number To make a simulation run in VDM ROAD the user picks a vehicle speed and road using the links in the left region of the Runs screen The controls and settings used to run the simulation are in the middle of the screen see Figure 2 There are fields for entering a starting and stopping station number For the example run represented in Figure 2 the vehicle is initially located at station 0 and the run ends when the vehicle reaches station 1900 A button with the caption Run Simulation is used to make the virtual test on the computer simulating the specified vehicle running down the specified road under the conditions defined by the links on the screen The mathematical vehicle models in VDM ROAD involve detailed nonlinear tire models and nonlinear spring models They include the major kinematic and compliance effects in the suspensions and steering systems
14. AD specifies a starting and stopping station number for ach simulation run For example in Figure 2 on page 4 the run starts at station O and stops at station 1900 The starting and stopping station numbers are used for three purposes in VDM ROAD 6 The starting station number designated S is used to locate the vehicle at the start of the run The stopping station number is used as a criterion for stopping the run The starting and stopping station numbers are compared to determine the direction the vehicle is traveling increasing station or decreasing station More specifically the initial conditions of the vehicle are set as follows The vehicle reference point center of front axle is located at the starting station number S specified on the Runs screen The vehicle reference point is located at the lateral position L specified as the target location at the given starting station L S 28 The heading of the vehicle is set to match the heading of the road at S when vehicle is traveling in the direction of increasing station When traveling in the opposite direction the heading is set opposite that of the road The pitch angle of the vehicle is set to match the longitudinal slope of the road at S When traveling in the opposite direction the negative slope is used The roll angle of the vehicle is set to match the superelevation of the road at S and L S When traveling in the opposite direction
15. S e A A IO ENERGIA E AO E ee Pai e a e Ra Suben ales ce E 3 E ES iy f oe sas Si aistin i EE E t Me RES A RN AA md AARASIAASINAOD ts 1092 E oe i do a f E j i 3 y Oo 200 400 600 800 1000 1200 1400 7600 1800 2000 a a o Staton number at front axle We Sas Ore oe 2 VG ge te Laera acceleralon 98 w pi p s x 5 j 7 A E i i a Kon ad i j fu li EE J F es cease enone sotesete posea cee caca aries fect eaten OLOCAU eto ataeevcnipemecrscteicec a a NS AMIA Etre a BB E 39 i i t ES i Fo 3 H i A 3 AS PA rien a DETE AA PNET once A A A A A A FES 7 i de ee i od A i E ES a pes i x E ae 5 pE a NS AN A OS GO fe NA a A O AS PR era ES pf i i 5 7 4 i ee H E wa t E A i H h i E EN 4 fect H i t ae B ha aveaeeeRavdeacsectuveroeutansy ey sorsra G on eah a Porron E ET E E E T R TTS TTT Be H H H A i z 3 z E 4 i A a i oe BA 3 p h gt Hee PS neo He Le Bora OO DEALS me eee gt feuoecavecadteey d E y H H gets R BES Y iv 5 413 eo res COPIAN cheer Naren rrnesmipecros dz crecieron bt ME eri i H shai AA A A e Y A e AS A ET ig p f Hi as 3 ni EUS thee E B g ho 3 f i A e gd g ag og 1 ites TE i E de E SEENE Te 5 i E PENN a Dea l p 3 A JEN E ee fo E OARS A SES AEA ce s ats E asaanivareveph aoe insert m A E prdenseninnnbenntpnsisdinernt
16. Sim technologies were selected and extended for this project The connection to roadway CAD packages is made through a standard IHSDM interchange file that is created as an output by CAD packages such as GEOPAK and read as an input by VDM ROAD The VDM ROAD provides an advanced simulation of vehicle dynamics that works with the 3D road design data In just a few minutes an engineer can simulate the behavior of a vehicle driving over several kilometers of a new roadway design Results are viewed immediately Vehicle motions can be viewed with wire frame animation The software can be used by engineers who are not experts in simulation or even vehicle dynamics This contrasts with older simulation packages where considerable expertise was needed in the fields of vehicle dynamics and numerical analysis Further the time needed to become proficient with simulation software has always been so lengthy that only the most determined engineers would become productive The software runs on Intel PCs equipped with Windows 95 or Windows NT With Pentium Pro or Pentium II machines the simulations run faster than real time The new software is almost ideal for simulating what would happen if a particular vehicle is operated on a particular road with a given input path lane position as a function of station number and speed also specified as a function of station number We recommend that the software be used for this purpose The UNIX Platform Whe
17. TA o coooocccononnnnnononcnnnanicnncanocanannnonoss 2l Coordinate Syste MS arciere A N a 21 Coordinate System Transformations ressonon eaa 23 Applications of the Road Coordinate SySteM oooocooooonocnonncncnonnnrncanonncananonancnnnos A 24 4 INTEGRATION WITH A DRIVER PERFORMANCE MODULE oococccocccnoonos 30 Overview of Man Machine Simulation a 30 Integration of DPM and VDM scscssssssssnssessesesssssnernsecsesessacesenaseeoneaes 31 5 CONCLUSIONS AND RECOMMENDATIONS concocccioccnconccnannnonnnonanannnnanananonnos 35 VDM ROAD SOL WATER E A 35 The UNIX Platform ribald eeu daha E unt eanernans Mio Integration with Road Design Data cscsesseeseeseeseeeeeteeteereeeeneeteeeeees O Integration of Vehicle and Driver Performance Models oooonconccnnonocononoconocnornonos 37 Use of Vehicle Dynamics Simulation in Roadway DeslgO ooooonoonnonnnnoncononnonass 37 APPENDIX A LATERAL ACCELERATION o ccoonccnoonnccnnanonacannaracnccananonancnrononos 41 The Basic Equidad we 41 The Cause of Lateral AcceleraOMialaaa 42 The Significance of Lateral Acceleration sonia 42 Linking Vehicle Dynamics to Road Design occocccioccnanonncnonnonanonononannconnccnr conos 44 REFERENCES atenemos 47 111 List of Figures Figured Concept of the MSDM psurcri aio 1 Figure 2 Example runs in VDM ROAD posisine con nne cra crnonaconross 4 Figure 3 Example combination vehicles in VDM ROAD coooonn
18. Tye dl 4 Jo pW qw SN 64 ey WA ul UU zW U wy uy oqui s E EE s Uawemnseey jo wejs s LUO LUIJU OU 10 OQUI S 9YI S IS ze Bunoesgns 1948 6 5 98x3 FHNLVYSdWAL eunyesodwe snis 9o9 enyjejeduJa 1ayuaJyej do Wi uj UMOYS 8q jeus 1 0004 ue 39 2015 SAWNIOA 3LON p qn seJjow s9Z0 0 spied o1qnod paqno seljow 82 0 0 yaa 391qn9 s si G8z e suojjeb s314111110 96 sasuno pini 3NNTOA sweJ6eDau 06 0 Qi 000Z suo poys swesbojiy SPO spunod sweib Se sz seouno 4619m SSV s8Jej994y 26 0 saJoe pajenbs s newojy 65Z sa yus esenbs pesenbs sejjau 9 88 0 spje esenbs pejenbs seJjeu 6260 0 399 esenbs pajenbs s an wpjju Z Gy9 seyouj sjenbs VIV s newojiy both seu SeJjaus y16 0 spied sesjoul 8yo 0 190 s n w u yS E Sayou HLONSA Puls OL Ag Aidyinw MOUY NOA UEUM SLINN IS OL SNOISUJANOS JLVINIXOUdAdV SHOLOVA NOISHSANOD 18 DIHLZMN P el yeb ZO I qi ZO ve Jul P zl gU ju pA y ul oqu s Table of Contents INTRODUCTION cis 1 Project Deliverable Soinua 2 Organization Of this REDOM css 2 2 THE VDM ROAD SOFTWARE sida ip a Overview of VDM RoAD Operation coooooccooncnnnncnonnnnnnononconanonannanronono nono cnn anno conos 3 Software Architect A arsine 10 Comparison of VDM ROAD and TruckSim Features oococcoconononoononnronarnrnenaanannanos 18 3 INTEGRATION WITH ROAD DESIGN DA
19. Y coordinates 2 Determine the Z coordinate of the design centerline at S based on vertical geometry data Also determine the longitudinal slope dZ 0S 3 For the current station S calculate change in Z as a function of L due to design superelevation Also determine the lateral slope 0Z dL 4 Add the superelevation effect step 3 to the centerline elevation step 2 to obtain the total height of the road surface at the wheel contact point ER ae 5 Transform the two slope values dZ dS and dZ dL to slopes with respect to the global X and Y directions dZ dX and dZ dY Use the sine and cosine of the road heading angle to accomplish this The complete details for computing Z and the two slope quantities are presented in Appendix H of the VDM RoAD Users Manual The application of these quantities to the vehicle model is described in Appendix G Interactions With the Closed Loop Steer Controller In order to simulate the driving of a vehicle over a 3D road design the vehicle models in VDM RoAD include a closed loop steering controller Figure 20 shows a simple flow chart of the controller logic target path driver model Full response Full vehicle dynamics simulation Xy Yy Vx VY Y Y UL Figure 20 Closed loop steer controller The driver model takes as inputs a target path and some feedback variables from the full vehicle model The controller works by looking ahead by some amount of time
20. ahnenstparcnnansaisanisantabarpaabsarnnsosite T T TTN san iria ADDRPOAPD er PANAI ISR cri iniaprirbnro A AI PRO ane ES SN A H y A ES y E EIC END GS E PANA oe seaman tan CANON q SACA eipcercorescrecro EDELMAN EEE Coid addins e A wd bene 4 _ o 200 400 E 300 1000 1200 EI 1600 1800 2000 e Maa y es i Staion nue m Acc tte eee RONAN OIR A A CREE ED EAE ECE EEA PA Meine tees eee HPR RRHH att es H Het ee per CAE ETA ERC HCE REACCIONES EA EERE HEU Oe TERR SEFE EEEIEE Ne E EEEH EEEE EEEE ERE ER EEA EE EEEE EE ER Figure 25 Lateral acceleration output from VDM RoAD Figure 26 shows a detailed view of the lateral acceleration for a small portion of the run At about station S 284 the road goes from a tangent section to a constant radius turn For a constant speed profile the Road acceleration would rise to a constant level of about 0 32 g s The variations seen in the plot are due to changes in vehicle speed The figures show that the lateral acceleration of the vehicle is close to the value that would be computed using only the road geometry and the speed The differences are due to transient effects during changes in road geometry One source of the transient effects is the driver steering which results in a small amount of lateral tracking error Another is the offtracking inherent in the vehicle Because all points in the vehicle do not track the target position perfectly especially at the transition there is a small differe
21. ailed vehicle model adds little if any function for ordinary design evaluation Friction Utilization Friction utilization is the ratio of tire shear force divided by the tire vertical force As long as the utilization is less than the available friction the tire grips the road as expected by the driver The lateral acceleration predicted by equation 5 in Appendix A is equivalent to friction utilization for the simplest possible vehicle model a point mass Detailed vehicle models such as those in VDM ROAD give the friction utilization for each tire and can provide more accuracy when the utilization is close to the frictional limit The acceleration limits from the Green Book are much lower than friction levels associated with sliding about 0 1 g for highway speeds This means that sliding will only be a problem for travel speeds well above the limit associated with the Green Book design guidelines Friction utilization is not going to be an issue for any highway vehicle if the acceleration limits from the Green Book are met It is only possible to exceed the road friction by traveling at speeds much higher than the design speed For design purposes a good predictor of the friction utilization is the lateral acceleration predicted by equation 5 in Appendix A The equation can be manipulated to compute a maximum speed V 39 In general the speed associated with a specification of friction utilization will be higher than a speed ass
22. almost ideal for simulating what would happen if a particular vehicle is operated on a particular road with a given input path lane position as a function of station number and speed also specified as a function of station number Research is now underway in separate FHWA contracts to develop a driver performance module DPM The report recommends a simple strategy for integrating the future DPM with the existing VDM RoAD software When JHSDM was envisioned it was thought that full simulation of vehicle dynamics would be needed to evaluate the suitability of a road design for use by a variety of vehicles Based on our understanding of the concerns of the designer a conclusion of the research is that full simulation capability as provided by VDM ROAD is not needed by the road designer Consequently future integration of VDM RoAD within JHSDM is not recommended During the project it became apparent that some users were expecting the vehicle dynamics module to predict lateral acceleration An appendix is attached to the report to explain why lateral acceleration is largely unrelated to vehicle dynamics Using simple equations for computing effective lateral acceleration as a function of radius speed and superelevation it is possible to give the designer simple measures of the comfort rollover danger and frictional requirements Comprehensive models such as those in VDM ROAD are not necessary for these purposes Vehicle offtracking and hi
23. and determining the steering wheel angle shown in the figure as u that will cause the vehicle path to best match the target path over the preview interval The steering from the driver is added to a steering due to other factors mechanical kinematic and compliance properties of the suspension shown in the figure as u The driver model includes parameters control logic and even a simplified vehicle dynamics model The target path is defined with a table of L versus S values as shown for an example lane change in Figure 21 In the figure the input path would start at the design centerline changing to a point 3 35 m left of the centerline after station S 135 m 26 ara SAR one 4 ARANA see ERRE eae raat PFE one ean hats ES 100 150 200 Station distance along designated path m Figure 21 Target path input The driver model must determine the lateral position of the target path relative to a coordinate system based on the moving vehicle Figure 22 illustrates the geometry Inertial Y Driver model Y axis Inertial Mass center Target path Origin front axle as ie Predicted path ZA constant steer DHS Xv Yv Driver model X axis inertial coordinates Figure 22 Coordinate system of driver model The driver model algorithm makes the following coordinate system transformations yy a The global coordinates of a reference point Xy Yy are converted to the roadwa
24. anvases non caducan Kgrensenoss ancora rovesanfeatennoned rovnueandhtacssomtocacarnedesionscthens forn foriranenajoracr einer es E H f 3 A H H 7 z A Y H sd H Fl gt t BY t i 13 1 t i 1 A 3 3 4 j 1 i 3 i RE E i fo ot a eee S i EI TN PR PR A i 4 1 H i 3 H i H t H H NA i y H A veng reuvesiet remera e 4 dress enn OS 104 Hees eocetnae tewedti en lirica beta Ng coe Bev tersed dete Biter ghee PA E E y H E i 3 F i p E H 4 ti z j 3 i E i h l r 3 a t a i t t H t E KS aera rerep ereenn Di OO IA sagen alg anves sevens 0 CEIT TATON A t H H i H t 4 H A 3 i i f j H F i i H y 3 H H H i H T appese sensetyerennted vseneseahrenaeeea alrcnoran cons Al CAOS O A IO progra cooler creta denorcasgarnmoc force narun O H H H k H i H t i i t i anetercarnnreten I PE ronca A a aaar i a T es E E toi tol i A ESE citada i E f i SS sassadeerscnedpaee a E A naco rre a cian E E H i f ti f 4 y H i E gt 2 2 i 4 H 4 F A Seares A AES PENE E EREE SEY O TAS e pN s F t a g B hi 3 y p H i 4 H i f p r A H f t r H i F rpgs arenkennonneectnney E H i tf h i a 1 t i 1 y renfe Peres eee rinin rr gas tes nens ki i i 7 1 3 i A H i i concentraron cea i i i H i a 4 i i i i 3 i H E 4 i i H aaraaropnoanara nenatdyeanaanunfanaenes ie B
25. appearing on page 142 is the same as equation 3 with the substitution of f for Ayeff V 2 mph t 15 Re 5 Here f is a dimensionless friction factor based on the assumption of a point mass following the radius R at speed V A slightly more complex formula is given on page 141 of the Green Book and is repeated below f e Vmph ef 15 R 6 Equation 6 includes a factor 1 ef that accounts for extra shear force and reduced normal force due to the difference between the plane of the turn perpendicular to vertical and the instant plane of the road inclined by superelevation For small levels of acceleration and superelevation equations 5 and 6 are interchangeable The Cause of Lateral Acceleration Lateral acceleration is experienced by a moving object following a curved path As indicated in equations 1 3 1t is determined by the speed and radius of the turn For a moving vehicle the input can be viewed as the set of control actions taken by the driver throttle steer and braking If the driver is viewed as the source of input then the speed of the vehicle and the path followed by the vehicle are outputs along with all results of the path including radius and lateral acceleration However another view is that the driver controls the vehicle to follow the road at a target speed From this point of view the lateral acceleration is fully determined by the choice of speed and the road
26. arget path and the drive torque to cause the vehicle to maintain the target speed Goals Y Path Command Speed Command VDM RoAD A Action Figure 24 Simplified view of driver control The DPM VDM system shown in Figure 24 would function in two steps 1 The DPM module would take as inputs the road geometry driver parameters traffic conditions and possibly a small number of vehicle parameters Vehicle parameters might not even be needed As output it would produce the target path lateral position as a function of station number and target speed also as a function of station number 2 VDM ROAD takes the target speed and lateral position as inputs and generates the steering wheel angle and drive torque to follow them taking into account the full dynamic response of the vehicle In this design the second part VDM ROAD is already complete The first part DPM is being developed The rationale for this modeling approach is that the driver behavior involving knowledge and rules does not depend on details of the vehicle dynamics It is assumed that when a driver decides what path to take it is without regard to details of vehicle behavior In planning a target path or speed the assumption is that the driver will do whatever it takes to control the vehicle to hit the target If the vehicle is sluggish the controller in VDM RoAD will generate larger steering wheel angles If the vehicle is responsive the cont
27. ase of VDM RoAD or use the menu to jump right to the results for a specific simulation run For example in the figure the selected run is named WB 12 Alt 3 R Lane This title was composed by a user and has not special significance to the simulation programs within VDM RoAD The title indicates that the run involves a vehicle called WB 12 running in the right hand lane of a road design called Alt 3 WB 12 is a combination vehicle in the AASHTO Green Book made of a 12 foot tractor towing a 12 ft semitrailer IS E ee ee eRe oes Bae RMN EEE neon a Figure 2 Example runs in VDM RoAD The Runs screen is shown again in Figure 3 with a different pull down menu showing The window contains three categories of user controls inputs the run and outputs On the left under the caption Model Parameters amp Inputs are four links to various kinds of input data a vehicle in the figure the vehicle is called WB 12 Combination aroad in the figure the road is called Alt 3 Design target speed defined as a function of station number and target lateral position also defined as a function of station number These four links are made by the user with pull down menus For example Figure 3 shows the screen appearance when the button next to the Vehicles link is clicked A pull down menu appears listing all of the vehicle data sets that are available in a library of vehicles In this case the library contain
28. ata vertical geometry design data and superelevation design data Integration of Vehicle and Driver Performance Models Section 4 presented a framework for describing man machine interactions at three levels based on 1 knowledge 2 rules and 3 skill It is only in the third category where rapid driver control skill interacts with the vehicle dynamics such that the vehicle is controlled as needed to follow commands in the form of target speed and lateral position within the lanes of the roadway The part of the driver model that determines the target path and speed is largely unaffected by details of the dynamic response of a specific vehicle Therefore we recommend that the DPM be developed using a simple point mass representation of the vehicle to determine the target speed and lateral position Perhaps the driver model would not even need a point mass It might be that the target speed and path can be determined based only on kinematical considerations involving speed point of view of the road geometry traffic and environmental factors VDM ROAD already includes the skill part of a driver model and can read the position and speed outputs generated by the DPM Use of Vehicle Dynamics Simulation in Roadway Design The IHSDM project is an ambitious undertaking and one in which the direction has changed as progress is made When this project started an objective was to select the best existing vehicle dynamics models and inte
29. be simulated with high confidence that most of the computer source code is free of common human errors Also the equations generated by AutoSim have been shown to be highly efficient generally running faster than equations generated by other means 8 When a simulation program runs it creates summary files that list each parameter value One of these files is created before the run LPO and the other is created at the end of the run LPF Both contain all parameter values In addition the LPO file contains the initial conditions for the state variables in the simulation The information in the LPO file is sufficient to exactly repeat a run The LPF file is nearly identical except ios that instead of initial values 1t contains final values of the state variables By modifying the start and stop times an existing run can be continued possibly using a different kind of simulation program The main purpose of each simulation program is to calculate time histories of variables of interest Those time histories are stored in a binary data file with the extension BIN The BIN files contain numerical data organized by channel number and sample number similar to test data recorded on a multitrack recorder A companion file with extension ERD describes the layout of the BIN file and contains labeling information for each variable It also contains the information that would normally be put into a log sheet summarizing the data including text n
30. blished in which the equations fit on a single page even for combination vehicles such as double trailer vehicles 3 As with other types of design criteria we recommend that candidate designs be evaluated only for the worst case vehicle For offtracking this is usually a single unit vehicle with the largest wheelbase allowed on the road A simple program that simulates only low speed offtracking should be developed for this purpose 40 APPENDIX A LATERAL ACCELERATION This appendix describes the relationship between vehicle dynamics road design and lateral acceleration This relationship is of interest to road designers because lateral acceleration is related to perceived performance in at least three ways 1 The lateral acceleration experienced by road users might be high enough to cause unease or discomfort 2 The required friction between tire and road that is needed for a vehicle to stay on the road at a given speed might exceed the available friction causing the driver to lose control and possibly leave the road 3 The lateral acceleration experienced by some commercial vehicles might be high enough to cause a rollover accident The Basic Equation There is a simple relationship between geometry speed and lateral acceleration Ay v2 Ay RG 1 where Ay has units of g s V has units of m s R has units of m and G is acceleration due to gravity 9 80665 m s This equation has nothing to do with mechanic
31. ccocannnnnonoonoonornonornrnrnconoros 5 Bisure4 Road CeOMe Ly escurre 6 Moure S Target vehicle Speed is 7 Figure 6 Animation of simulation results scssi e 8 Figure 7 Example plots of simulation results ccccsessesssesseetseseessesseeeeessereeseesaeenes 9 Figure 8 More plots for a simulation TUN ressorsa nano ro nanncnrnanarrnrnness 10 Figure 9 Four parts of VDM ROAD csccsccssecsssessseseneosssearecsntanecsersneceenreneesteasees 11 Figure 10 Contents of a VDM ROAD library oonnonnncnonnoncnnnnonnonncononnnronoonorancononenrnnanoos 12 Figure 11 Example data SCreensennginincs ie a ea aienteiioes te 13 Figure 12 Screen display when solver program is running ococoononncnnoncnonononnnnnnnnnoncnnnnos 14 Figure 13 Geometry of the camera point and the look pOlDt oonoocnonconinnononorononrarononos 16 Fieure 14 Animator input files riada a 16 Figure 15 The WInEP Works pacesacniand ais ald Figure 16 Batch run Control cscssssssecsseesersssensenecssennsasenrsensassnerssconsnseasnensesconeenes 18 Figure 17 The S L Z coordinate System ooconnononcnnnenornonnnenoncnornnononnonnnnnornnnanancnncnnanannos 22 Figure 18 Computation of X Y from S L cocoococonccccnonnonccnnnnnononnonnornnornnenonnonannonornnnnanccnos 23 Figure 19 Directions of tire action Vectors oononnoconcononononcannannononcononnonnnnaronana canon o ranonnonos 25 Figure 20 Closed loop steer controller oooomononoononnonnornornnnaracan
32. d Figure 19 shows vectors that describe the four main actions on the vehicle generated by the tire longitudinal force F lateral force F vertical force F2 and aligning moment M The three forces apply at a point defined by the intersection of the wheel plane with the 3D road surface This point labeled W in the figure has X and Y coordinates that are known approximately from the vehicle model The Z coordinate is determined by the road design data oa eee Figure 19 Directions of tire action vectors The road normal vector is shown in the figure and labeled r The road normal is a vector whose direction is perpendicular to the road surface If the road is flat and level r points straight up in the direction opposite of the gravity vector However in general it is tilted slightly due to vertical geometry and superelevation The road normal is used in the vehicle dynamics models together with the direction of the spin axis of the wheel to define the vectors t and t that define the directions in which the tire longitudinal and lateral forces are applied In order to calculate the directions of the vectors it is necessary to determine the road normal at each tire location W Starting with the X and Y coordinates of point W the vehicle simulation programs in VDM RoAD compute Z and two slope values The process is as follows 1 Determine the S L coordinates associated with the point by transforming the known X
33. d that items 5 and 6 be predicted with a different type of model than used in VDM RoAD For hill climbing a simple point mass representation with a parameter specifying the power weight ratio is sufficient For hill descending a thermal model to predict brake temperature and fade would be needed The first four items can all be addressed with models much simpler than the ones in VDM ROAD Following are the recommendations for providing feedback to the designer in IHSDM for these four items Lateral Acceleration Comfort Lateral acceleration is defined solely by the geometry of a path taken and the instant speed The equation is gt Ay RG 1 where V is forward speed and R is the instant radius of the path The radius is a combination of the roadway geometry and the relative lateral position taken by the vehicle within the road The dynamic properties of the vehicle have only a minor effect at transition points see Appendix A for a more detailed discussion We recommend that equation 1 be used to compute lateral acceleration as a function of the target speed generated by the DPM A detailed vehicle model is of no use for this purpose 38 Vehicle Rollover A vehicle navigating a turn is subject to centrifugal force acting on its center of mass and causing the vehicle to roll away from the turn e g the vehicle rolls to the right in a turn to the left If the vehicle rolls so much that the inner wheels lift off the gro
34. e X Y Z system The other coordinates S and L are based on the design centerline To locate point P in the road coordinate system one would 1 start on the projection of the road design line into the X Y plane at station S 0 and then travel along the projection of the design line S meters 2 then go laterally L meters along a line in the X Y plane that is perpendicular to the projection of the design line and finally 3 goup Z meters parallel with the Z axis The right hand rule requires that L be positive for positions to the right of the centerline when facing upstream in the direction of decreasing S If facing in the direction of increasing S then points to the left of the centerline are positive and points to the right are negative The sign convention for L defines a right handed coordinate system when positive Z is up Only right handed coordinate systems are used in complex commercial 3D software packages Thus a right handed system was defined for VDM ROAD Coordinate System Transformations The Z coordinates from the X Y Z and S L Z systems are the same Therefore transformations between the two systems involve only the two dimensional X Y and S L systems The process of transforming S L coordinates into X Y values is direct For any pair of S L values there is a single pair of X Y coordinates Figure 18 shows the geometry for an arbitrary point X Y the centerline and a design point on the centerl
35. eed Allowable f values range from about 0 10 for 70 mi h to 0 21 at 10 mi h These levels are mainly set by occupant comfort and anxiety The use of the term friction utilization is misleading because humans sense the effective lateral acceleration equation 7 rather than friction utilization The data in the Green Book really provide guidelines on comfort not friction use Vehicle properties can affect the correlation between the anxiety of a driver and speed in at least two ways First the roll angle in equation 7 depends on the vehicle properties Some vehicles roll more than others at the same level of lateral acceleration Second the anxiety of the driver depends in part on an assumption of the vehicle limits The driver of a truck should be troubled by a level of lateral acceleration that would not bother the driver of a sports car Vehicle Rollover A vehicle navigating a turn is subject to centrifugal force acting on its center of mass and causing the vehicle to roll away from the turn e g the vehicle rolls to the right in a turn to the left If the vehicle rolls so much that the inner wheels lift off the ground it is unstable and rolls over The primary vehicle property that influences the propensity to roll over is half the ratio of the track width to the height of the mass center Roughly this ratio corresponds to the lateral acceleration at which the vehicle rolls over For example if the track distance between tire center pla
36. eeded for preparing graphical plots of the data By convention ERD and BIN file pairs are simply called ERD files The name ERD is used because the format was created by the Engineering Research Division at UMTRI where it has been used since the early 1980s Data processing programs for ERD files obtain most of the information needed from the file itself For example the high level of automation in the animator and proline exists because both were designed to work with ERD files VDM ROAD can also be made to produce simple text output files for export to spreadsheet programs and other analysis software The Animator VDM ROAD includes a program for animating wire frame figures to visualize vehicle motions The animation is accomplished by drawing images similar to what would be seen with a video camera and updating the images many times per second to show motion The view depends on where the camera is located It is also determined by how the camera is aimed Figure 13 shows the basic geometry and the relationships between the camera point a look point the 3D system being animated and the 2D image that is recorded The animator program reads two kinds of input files see Figure 14 Motion variables are read from an ERD file created by the simulation programs whenever a run is made Other information such as program settings definitions of parts and shape information are described with keyword based text files with extension
37. ette Cae iter a A O IS ESSE SORDOS ORGIAS Rannan A 4 j 3 i r H i H r H E 4 0 q 3 0 500 1000 1600 2000 0 00 1000 1500 A a Y Station number m jon number axe m aN at wo Lal Pos vs Station All Axles wB 12 AR 3 R lane iris ane ae AA AAP RSA A IRAP REAR S DD A A APR RENDON MEP NL RIN NN RA AR POT CARCASA INO COOOL ISO COPY APTANA I Lat Pos vs Station All Axles WB 12 AR3 R lane Lateral position m Target a L wheel axle 1 L wheel axle 2 L wheel axle 3 vee enojes rocdor non joe nang H H 4 Ajena card jac uon ni gueon en pon rmondena cano cana nod ane AS naran i anehi on E sets a e A Pb OS o e rae ae te wheel ade 4 Rwheel axle 1 4 heredes erre dadas i dArenaadute laednocaatrestndarn rado dom R wheel axle 2 E 4 E en De es E R wheel axles duo noendanrenos R wheel axle 4 exar pee Soe on oe O te Se So ea ote ee natn iiaa paren res nae me ce A AD A A A A wade Meceavenen romania A EE aves nao nba moon nro ne aaa nenn ajoriacadcanianas E nesese TERET speed i tees A a IO IICA IA Sea e penas o sone bwntrans esorta eeren eretti O O iiei eee heer dite E SA E VESAK pretender regar iia Jeorrobrarirdacos xo qeveeeqerenes serraje A ween tbe aed coe e ett eanguacendeasenafetnevieenee geeen we
38. f this project was to select among existing vehicle dynamics simulation programs and link them through a common database to commercial roadway CAD programs The existing TruckSim and AutoSim technologies were selected and extended for this project to create a software package called VDM RoAD Vehicle Dynamic Models Roadway Analysis and Design VDM RoAD simulates the vehicle dynamics behavior of cars trucks and two unit combination vehicles tractor semitrailers on 3D road designs The software runs under Windows 95 and Windows NT and shows results graphically with plots and animations The software can be used by engineers who are not experts in simulation or even vehicle dynamics The new software was developed with the intention of adding it to an interactive highway safety design model IHSDM that systematically considers safety of roadway and roadside design elements in creating cost effective highway design alternatives Although VDM ROAD is designed to work with other parts of the IHSDM further integration is not recommended because most factors of performance that are of interest to roadway designers do not require vehicle dynamics simulation Vehicle related performance measures are largely defined by lateral acceleration which depends only on the kinematics that link roadway geometry and travel speed 17 Key Words 18 Distribution Statement Computer simulation vehicle dynamics No restrictions This document is available to rollover r
39. functionality in the UNIX version of the graphic interface as was possible in the Windows version We are not sure it 1s even possible Before the project concluded FHWA changed the platform of choice for IHSDM from UNIX to Windows NT We recommend that all UNIX products developed in this program be abandoned If there is a need in the future to run the vehicle dynamics simulations under UNIX the source code for the vehicle models can be easily compiled However we would not recommend attempting to reproduce the other parts of VDM ROAD unless improved development tools become available Integration with Road Design Data In its current version VDM ROAD is integrated with IHSDM through the interchange file format that summarizes the entire 3D road geometry A portion of an interchange file is shown in Figure 4 on page 6 The road design data are presented in tabular form with 37 columns Each row in the table represents a change in some property of the design For example a new row is needed at the beginning or end of a turn to specify properties of the new radius spiral or tangent A new row is also needed at any transition in the vertical geometry or any transition in the superelevation in any of up to 14 lanes The interchange file contains three kinds of design information 1 horizontal geometry of the roadway centerline 2 vertical geometry of the roadway centerline and 3 transverse profile information width and slope for 13
40. geometry It is not understood among all civil engineers that the same lateral acceleration is obtained for any two vehicles following the road if their speeds are the same The individual dynamic behavior of the vehicle has no influence whatsoever on the lateral acceleration experienced when the vehicle follows a road at constant speed The Significance of Lateral Acceleration Although vehicle dynamics have no influence whatsoever on the lateral acceleration experienced by a vehicle following a road at constant speed the significance of the lateral acceleration does depend on the individual vehicle properties MEE y PUE Recall that there are three links of interest between lateral acceleration and user comfort and safety The mechanics of each link are discussed below to provide the basis for recommendation made later Occupant Discomfort Lateral acceleration is directly sensed by drivers and passengers However humans cannot easily distinguish between true lateral acceleration and the effect of leaning due to road superelevation and vehicle roll Equation 3 includes the effect of superelevation To add the apparent lateral acceleration sensed by the occupant of a vehicle with some body roll the equation should be y2 Ayeff RG E 7 where is the roll angle of the vehicle radians with respect to the road The AASHTO Green Book shows research results from the 1930s and 1940s that correlated friction utilization f with sp
41. grate them into the IHSDM so they could be used to run a fleet of vehicles over a candidate 3D roadway design VDM RoAD was developed and meets this objective However just because vehicle dynamics programs can be used by designers does not mean they should be used e yee When IHSDM was envisioned it was thought that full simulation of vehicle dynamics would be needed to evaluate the suitability of a road design for use by a variety of vehicles As a minimum the fleet of simulated vehicles would match the 15 examples from the AASHTO Green Book Upon considering the objectives of a designer we do not recommend the continued integration of VDM RoAD within IHSDM for improving the design process We conclude that the concerns of the designer are so specific that a more efficient avpreaen should be used Some specific concerns that involve vehicle responses are 1 Will the lateral acceleration experienced by drivers exceed limits found in the AASHTO Green Book Those limits involve low levels of acceleration and indicate a threshold of discomfort 2 Will any vehicles run into frictional limits and risk running off the road due to skidding Will any vehicles risk rolling over 3 4 Will large vehicles have difficulty navigating sharp turns due to offtracking 5 Will hill climbing challenge underpowered vehicles to maintain high speed 6 Will long descents challenge the ability of heavy trucks to maintain low speed We recommen
42. imum f values of the individual tires Two vehicles with the same tires and weights can have substantially different limits where loss of control occurs due to friction limits The limit condition depends primarily on lateral acceleration However friction varies with speed especially on wet surfaces so the limit condition for a given pavement depends on both lateral acceleration and speed such that higher lateral accelerations can be achieved at lower speeds Lateral acceleration is the primary variable predicting the limit and speed is a secondary variable Linking Vehicle Dynamics to Road Design The preceding material is intended to make two points 1 The lateral acceleration experienced by a vehicle following a road with given radius and superelevation depends mainly on the speed Lateral acceleration is defined by the motion not the physical properties of the vehicle experiencing the motion 2 Limits associated with vehicle dynamics can be characterized by the lateral acceleration at which the limit occurs One is the point at which a driver feels unsafe the second is rollover and a third is when a loss of control occurs due to insufficient friction between tire and road During the program described by this report a program called RAPID COMBINE was created to combine road design data with outputs obtained from detailed vehicle dynamics models The calculations involved nothing more than equations 3 and 4 During a review of the p
43. ine X Y r A Y X Centerline Figure 18 Computation of X Y from S L The process for computing X and Y is as follows I Find the index i for the last design point preceding S This is done using a table lookup procedure Compute the S distance of the current point past the last design point This change is shown in Figure 18 as AS Compute changes in X and Y in a rotated coordinate system where X is tangent to the centerline at the design point These changes are shown as AX and AY in Figure 18 The details of the calculation depend on whether the road centerline is a straight b has a constant radius or c is on a spiral Using the heading angle of the centerline at the design point transform the quantities AX and AY to obtain X and Y changes in the global coordinate system AX and AY Add AX and AY to the coordinates of the design point X and Y This gives the X and Y coordinates at a point on the design centerline at S Determine the heading angle at S Different methods are used depending on whether the road is straight on a constant radius turn or on a spiral Use the heading angle to determine changes in X and Y due to L TE ae 7 Add the X and Y offsets from step 6 to determine the X and Y coordinates for the point at S L Additional details of the transformation are provided in Appendix H of the VDM ROAD User Manual 2 Converting from X Y coordi
44. l time step Values of 0 01 to 0 1 sec are more typical Running DPM and VDM Sequentially In viewing the control flow chart in Figure 23 consider the computation requirements for the three loops The outer loop knowledge is so abstract it might be beyond the scope of DPM Abstract thinking does not involve rapid coupling with vehicle dynamics because the thinking process generally covers a time scale involving minutes or hours The intermediate loop rules 1s probably the main part of the DPM Taking into account traffic conditions perceived road geometry and the current speed and location of the driver s vehicle the driver model generates commands in the form of a target path and target speed The time scale involves seconds The inner loop skills involves the close coupling of the driver control and the vehicle response Given a target path and target speed this level of control determines how the driver controls the throttle brakes and steering wheel to achieve the targets This time scale covers tenths of seconds Given that only the inner loop skills involves a short time scale high frequencies it is possible to separate the driver control into two parts as shown in Figure 24 In this ee figure the DPM is the part of the simulation that produces commands the target path and target speed The VDM ROAD software includes the skill controller algorithms for generating steering wheel angle to cause the vehicle to follow the t
45. ll climbing performance are not related to lateral acceleration and cannot be simulated with simple algebraic equations However the associated vehicle models are extremely simple and do not require the complexity of a comprehensive simulation package such as VDM ROAD VEHICLE DYNAMICS PROGRAMS FOR ROADWA Y AND ROADSIDE STUDIES Final Report UMTRI 98 20 1 Prepared for Federal Highway Administration Contract DTFH 61 93 R 00142 Michael W Sayers April 1999 Technical Report Documentation Page 1 Report No 2 Government Accession No 3 Recipient s Catalog No FHWA RD 98 169 4 Title and Subtitle 5 Report Date VEHICLE DYNAMICS PROGRAMS FOR ROADWAY 6 Performing aa eg Code AND ROADSIDE STUDIES 04 07 4 8 Performing Organization Report No 7 Author s Michael W Sayers UMTRI 98 20 1 10 Work Unit No TRAIS 9 Performing Organization Name and Address The University of Michigan Transportation Research Institute 2901 Baxter Road Ann Arbor Michigan 48109 12 Sponsoring Agency Name and Address Office of Engineering Research and Development Federal Highway Administration 6300 Georgetown Pike McLean Virginia 22101 2296 15 Supplementary Notes Leonard C Meczkowski COTR HSR 20 11 Contract or Grant No DTFH 61 93 R 00142 13 Type of Report and Period Covered Final Report Dec 1993 Nov 1997 14 Sponsoring Agency Code 16 Abstract The main objective o
46. mework for providing three levels of abstraction involving human control of machines The three levels are called skills rules and knowledge Figure 23 shows how these levels of performance can be related to the modeling of driver vehicle interactions Goals DRIVER Symbols Knowledge Plans Strategy Tactics Commands Driver Skills Perception Controls Sight Execution Sound Vehicle Dynamics SYSTEM Ti Y Figure 23 Driver control behavior Action The big picture of man machine interaction as described by Rasmussen covers extremely abstract notions such as the analysis of unfamiliar situations and development of general plans or strategies based on the knowledge of the human The next level is Aes less abstract It involves tactical decisions based on rules developed from past experience The outcome of the knowledge and rule based behavior are commands In the context of the driver of a vehicle the commands are a target speed and lateral position of the vehicle on the roadway The third level of behavior involves the driver controlling the vehicle through the steering wheel throttle and brakes The objective of the driver is to use previously learned skills to execute the commands 7 Integration of DPM and VDM The DPM is presently in the early stages of development and details of its operation have not been finalized However it is probably safe to assume that a
47. mics and the effects of changes in vehicle design On the other hand VDM RoAD is designed for civil engineers to evaluate candidate road designs and to see the effect of changes in road design Accordingly the two software packages have some significant differences as summarized in Table 1 In order to make new runs with TruckSim or CarSim the user must have a hardware key called a DONGLE that attaches to the parallel port of the PC FHWA has been provided with two copies of TruckSim along with two hardware keys All rights to VDM RoAD have been transferred to the government The VDM RoAD software does not require a hardware key and can be distributed as the government chooses The source code for VDM RoAD has been provided to FHWA pa poe Table 1 Comparison of TruckSim and VDM ROAD features Vehicle Model Features Vehicle Model Features Alternate support for STI tire model 1994 version Full 3D tire road interface slope cross slope Number of truck models Number of car models Unit systems Tire kinematics lateral slip inclination etc Suspension spring and damper forces and deflections VDM RoAD No es JL es es es es es es No No Vvs S S T Roll V Vehicle road Y Y Y Y Y Y Y TruckSim Y Y Y Y metric or English es es es es No No Y vs X V vs T Steer vs T Pressure vs T No es es es No No O es es XYZ angles T Roll V Vehicle Y
48. n this project started in late 1993 the vehicle models and supporting programs existed only on the Windows platform Yet most States with road design CAD systems were using UNIX As the project developed one of the contract modifications included a task of porting the software from Windows to UNIX A partially functional version of VDM ROAD exists on UNIX e The simulation solver programs are generally platform independent The code is written in Fortran 77 and runs on any machine with a suitable compiler Most of the code also exists in C the exception is the interface with the IHSDM 35 interchange files There were no problems generating executable binary UNIX files e The plotter and animator are C programs that make use of the graphics and user interface libraries of specific operating systems The Windows versions make use of well developed libraries and development tools that only exist for Windows The UNIX versions are functional but would require considerably more development effort to reach the same level of utility as the Windows versions Performance of the animator was disappointing it runs slower on a workstation than the Windows version on a low end PC e The graphical user interface on Windows is programmed in ToolBook 4 A similar product for UNIX exists that 1s called MetaCard However we found that our productivity with MetaCard was much less than with ToolBook We were never able to program the same level of
49. nates to determine S L values is not as direct as the above transformation One complication is that multiple values of S L can exist for a single point in X Y space For example the center of a design curve can be equally well defined using any value of S along the curve with the magnitude of the L value being equal to the radius The method is described in Appendix H of the user manual It requires knowledge of the design point S that is associated with X and Y Separate transforms are presented for tangent and constant radius turns The geometry has not yet been worked out for spirals A procedure is followed to compute AS and L The calculated value of AS is then inspected to see if the assumed design point was correct For example if the design point is located at the start of a curve covering 200 meters the valid range of AS is 0 to 200 If the computed value of AS is greater than 200 then the calculated values of S and L are not correct because the point of interest is past the end of the curve In this case the procedure would be repeated using the next design point Applications of the Road Coordinate System As noted at the start of this section there are several applications of the transformations in VDM RoAD Tire Road Interactions The most essential application of the road geometry is to define the 3D ground surface under each wheel of the vehicle such that the interactions between the ground and the tires can be compute
50. nce in the lateral acceleration Another source of the difference is due to vehicle roll The reference point in the vehicle is above the ground at the mass center and follows a slightly different path than a point at ground level when the vehicle rolls Although differences exist the conclusion to be drawn from the figures is that the simple calculation of A labeled Road in the figures is very close to the lateral acceleration in the vehicle If the geometry includes the effect of a driver starting a turn earlier the transient effect would be reduced 45 RESF Me et Se sas oe Ay vs Station P AS Riane Lateral acceleration g s ee ee A a ee ee ee a S AOU ee oe ate af ag We at ats io ere Bae ee ye a Vehicle in a Road iio Eat o2 pa ae A os E wie An 240 260 280 300 320 340 360 380 400 420 440 460 430 eT Cake oe Steton number m os Figure 26 Detail view of lateral acceleration le 10 11 12 13 REFERENCES Reagan J A The Interactive Highway Safety Design Model Designing for Safety by Analyzing Road Geometrics Public Roads Summer 1994 pp 37 43 VDM ROAD User Reference Manual Version 1 0 Prepared by UMTRI and submitted to FHWA December 1997 350 p A Policy of Geometric Design of Highways and Streets 1994 AASHO Green Book American Association of State Highway and Transportation Official
51. nes for a car is 5 ft and the mass center is 2 ft above the ground then the lateral acceleration needed to roll the car over on flat ground is approximately 1 2 5 2 1 25 g s As another example a truck with a wheel spacing of 6 ft and a mass center height of 6 ft would roll over with a lateral acceleration of roughly 1 2 6 6 0 5 g s Rollover accidents that occur because of a combination of high speed and a short radius turn usually involve vehicles with a high ratio of half track over mass center height 43 True Friction Utilization Friction utilization is defined as the ratio of tire road shear force divided by normal force The utilization is usually different for each tire on a vehicle due to transfer of load from the inside wheels to the outside the nonlinear relation between shear force and load and the suspension and chassis properties of the vehicle The friction utilization is typically highest for the unloaded wheels on the inside of the turn e g the left wheels when the vehicle is turning to the left and lowest for the loaded wheels on the outside of the turn The friction utilization f described in the Green Book is based on the assumption of a point mass following the road geometry at speed V There is no simple relationship between the point mass friction utilization and the friction utilization of the individual tires other than the fact that the point mass f value will lie between the minimum and max
52. nrnrannoncanonanananaanarnnonos 26 Figure 21 Target path input ooonnnncnoconnnnnnoncononnonannoranncnnonnannoonnnnoc corno n anne na roca racaonnasansos 21 Figure 22 Coordinate system of driver MOdEL cesses ssseesserseseteesssseseesssseseeseees 21 Figure 23 Driver control behavior esssssssresesserererersresererereneresterneseneneresesrererererereres 30 Figure 24 Simplified view of driver contrOl c csseseseeees ea 33 Figure 25 Lateral acceleration output from VDM ROAD ooooconccconnnnannnnncnnnonininanannanacnnos 45 Figure 26 Detail view of lateral acceleration ocoonconnconocnonnrnrnnnancononncnananancrnnonnonnrnnos 46 pee y y A 1 INTRODUCTION In the late 1980s the Federal Highway Administration FHWA assigned a high priority to research and development of better highway design practices and criteria The goal of this multiyear program is to develop an interactive highway safety design model IHSDM that systematically considers safety of roadway and roadside design elements in creating cost effective highway design alternatives The computing environment is intended to serve several purposes First it provides a data structure for the convenient development of input and output data libraries Second 1t enables data to be shared between the different modules of this computing environment Third it encourages the integration of commercially developed and maintained vehicle dynamics computer aided
53. nsively with the VDM RoAD database The database was programmed using a software package called ToolBook available from Asymetrix Inc 4 ToolBook is an scriptable object based software development tool primarily used for preparing educational and multimedia material It is also commonly used for prototyping graphical user interfaces for complex software projects In addition to VDM RoAD other simulation packages have been prepared using the architecture shown in Figure 9 The architecture is called the simulation graphical user interface SGUI and includes a standard extension to ToolBook developed at UMTRI 5 6 The database can be considered at four levels Figure 10 shows a graphical view of the first three 1 Data screens Any screen display in VDM RoAD that has editable fields or other forms of user settings is called a data screen 2 Data sets A data set is the information you provide and can edit in a data screen minus the screen itself The data screen is a view of a data set as seen through the SGUI the VDM RoAD graphical interface 3 Libraries A library is a collection of one or more data sets plus the information needed to provide a view in the user interface The next figure shows the contents of a library multiple data sets plus a graphical view of one data set at a time 4 The VDM ROAD database VDM ROAD includes about 50 libraries ERA E A 4 a TD th ey OLR AENA RAT Y e Individual data
54. oadway design geometry the public through the National Technical IHSDM Information Service Springfield Virginia 22161 19 Security Classif of this report 20 Security Classif of this page 21 No of Pages 22 Price Unclassified Unclassified 46 Jo P ed eb ZO qi ZO DL gitu zli zU pu pA y ul oqui s Oo de gt Oo 00 08 Q o oy oY cle 9 86 ce do Jo eunyesodwe yoyuesyey ze ppe in esadw9a usyy 6 98ex 34N1V43SdWAL 3 spsed 91qno goet peqno seljew 398 9 qn9 SLE SE p qn sejjeuw suojje6 y9z 0 seul seouno pini yEOO SOIR EW AWNIOA suo poys cor 1 Bx 000 1 swesbebow spunod coz 2 swego seouno eseo o sweJ6 14619m SSVI soJoe esz zw 000 OL 910 994 sej u asenbs 6 0 pesenbs SeJja8uJo JA yee ejenbs v9OL OL pejenbs seJjjew seyou ajenbs 91000 pajenbs seljowsyjjw vVayHV sep 1z79 0 SO1 9WO gt spied 60 seJjow 199 ez seJj9uw seyou 6 0 0 sejjaowpp1U HLIIN31 Pula OL Ag Aid OW MOUN NOA veuM SLINN IS OL SNOISHSANOO SLVWIXOUddV Y 10615 18PJO YMHJ JO Juevesmbes ey 0 WIOJUOS 810398 S UL snisi99 Do sul Pa uu Ow 64 ey Uy zUJUJ wy w a ww oqu s inches wd pl P 1 z iu ale i rm NN alla le Ti L AU me err vt et zt tt 01 6 iu St an de a re ili ain al An nk an he ar al el ah 2 TEF rn me
55. ociated with comfort or rollover We do not recommend the use of detailed models such as those in VDM ROAD as a design tool involving friction utilization If a designer wishes to know the maximum speed possible before a specified level of friction utilization is reached we recommend that equation 5 from Appendix A be used to compute the speed as a function of station number according to changes in design radius Vehicle Offtracking Highway vehicles with front wheel steering systems have the characteristic that when turning the wheels of the rear axles track inboard from the paths followed by the front wheels This behavior is of concern when designing intersections truck loading areas bus stop areas and any roadways or pavements where large vehicles must turn The vehicle models in VDM RoAD can be used to determine tracking requirements for any geometry that is described using the IHSDM interchange file format However the behavior can also be simulated using much simpler means Decades ago a drafting tool called a tractrix integrator was used to draw the paths followed by the wheels of trucks making sharp turns Graphic templates are included in the AASHTO Green Book for the 15 design vehicles The AASTHO templates are probably the most commonly used tool for evaluating intersection designs In the 1980s FHWA supported the development of software on a personal computer to perform the same function 2 A simpler formulation was later pu
56. rganization of this Report This report is intended to describe the project for which VDM RoAD was developed The full documentation for VDM ROAD is provided in the user reference manual VDM RoAD User Reference Manual Version 1 0 2 Section 2 describes the VDM RoAD software It has example applications and a short overview of the architecture Section 3 covers the topic of how the vehicle dynamics models are integrated with road design data Section 4 discusses how VDM RoAD can be integrated in the future with the driver performance module currently in development Section 5 presents conclusions and recommendations for future integration of vehicle dynamics models within IHDSM An appendix is provided that discusses the topic of lateral acceleration During the project it became apparent that some users were expecting the vehicle dynamics module to predict lateral acceleration This appendix was prepared to Gapian why lateral acceleration is largely unrelated to vehicle dynamics 2 THE VDM ROAD SOFTWARE VDM ROAD is a marriage of vehicle dynamics simulation technology with roadway CAD technology In just a few minutes an engineer can simulate the behavior of a vehicle driving over several kilometers of a new roadway design Results are viewed immediately Plots of important variables can instantly identify trouble spots station number where excessive motions occur Vehicle motions can be viewed with wire frame animation The software runs
57. roject in 1996 FHWA consultants determined that the simple calculation of lateral acceleration as a function of road design and vehicle speed was not sufficient the full simulation capabilities of VDM RoAD were required The capabilities of RAPID COMBINE have been integrated within the vehicle models in VDM ROAD For example Figure 25 shows the lateral acceleration as a function of station The figure also shows the vehicle speed used to compute A 44 A A tA ES ANSE AS vericte longitudinal velocity P alt 3 R lane A eo Vehicle mane vey P as Rine verie rg velcty ce B AAA rar nene eet Aran nme nmd ane nen eaten nud rencor 7 s 7 ae bancas pias a a A E TEE A gt a A a A E H f ES 3 E H i E Be oe doo E bo i 3 3 i ae E H A A Beobeveisennneseennipebsinasaninietnsfiniensicteaeanitifeserbatesursnsbsiedinenicisssontsininfravtaninenteateedbdseasveriesee sons B idear Sn rome i y 2 g E i a R 7 he A Eos i 4 i g E s AS i ie 1 an H 2 ES aacezeannsncsausnoiperasstienee iaiki eniri cvemeroscci n RS DORA E TN ALANINE ee A aE OT atea oia arecmerocsrehprormeriorrosciprcericriros cinc cmo cinc PA ES 1 i ES i 3 i oe 4 i f ANNANLA T ee ANNO OLE NTERLENOERER VALIA CONAN e neoerboneda rei LOCOS NADAN Net ee ee oboe A DOS ANEEL EE R correos ne g A e i H i He i ion ee f 3 A A O A bee o A S
58. roller in VDM RoAD will generate less control The DPM can perform all of the planning without concern for the details VDM RoAD will then do whatever it takes to execute the plans Of course in extreme cases the vehicle will crash When the controller tries to follow a path that is beyond the physical capabilities of the vehicle and road an accident occurs in the form of a rollover or of the vehicle skidding off of the road In these cases the simulated driver behavior will be wrong pure and simple However this limit in accuracy is not necessarily a limit in the usefulness of the software The basic objective is to identify potential problems involving road design vehicle type driver behavior and signing If the simulation shows a large amount of corrective steering or tire loads reaching zero then a problem has been identified and the software has provided useful information The details of how a driver and vehicle interact after the problem occurs is not of interest if an alternative design can be developed a AER 5 CONCLUSIONS AND RECOMMENDATIONS This section presents the main conclusions drawn from the project and presents specific recommendations for the continued development of the IHSDM VDM RoAD Software The main objective of this project was to select among existing vehicle dynamics simulation programs and link them through a common database to commercial roadway CAD programs The existing TruckSim and Auto
59. s more nen anet 116300 E daran noo ny ee ae 116600 54000 55000 X coordinate m Figure 15 The WinEP workspace Each plot window contains a graphical representation of one or more X Y data sets An X Y data set 1s a series of X and Y values obtained from a data file The X Y data sets can come from the same file or from different files The X values in each data set do not have to be the same and the data sets do not have to contain the same number of points WinEP is used in other software packages and is updated often The most recent version can be downloaded via the Internet from the web site http www trucksim com winep html N o peer Batch Control The VDM ROAD database includes vehicle descriptions target paths lateral position as a function of station number target speeds speed as a function of station number and IHSDM data files with road geometry The database also includes descriptions of runs plots and animations For example the Runs screen see Figure 2 and Figure 3 shows a data set for a run and the Runs library contains the data sets for numerous runs VDM ROAD also contains several data sets that include groups of runs and or plot settings For example Figure 16 shows a screen used to make a batch of simulation runs The user uses the Add and Remove buttons to create a list of runs in the lower left part of the screen under the heading Data Sets to Run Each line in this lis
60. s Man and Cybernetics Vol SMC 13 No 3 May June 1983 pp 257 266 Sayers M W Vehicle Off tracking Models Transportation Research Record 1052 National Research Council Washington D C 1986 pp 53 62 Sayers M W Exact Equations for Tractrix Curves Associated with Vehicle Offtracking Vehicle System Dynamics Vol 20 No 5 1991 pp 297 308 ae 1 eee
61. s Washington D C 1994 Asymetrix ToolBook User Manual Asymetrix Corporation Bellevue WA 1994 Sayers M W and Mink C A Simulation Graphical User Interface for Vehicle Dynamics Models SAE Transactions 1995 Vol 104 Society of Automotive Engineers Warrendale PA 1996 pp 58 67 Sayers M W and Mink C Integration of Road Design with Vehicle Dynamics Models Through a Simulation Graphical User Interface SGUD Presented at the Transportation Research Board 74th Annual Meeting Washington D C January 22 28 1995 AUTOSIM Reference Manual Version 2 5 Mechanical Simulation Corporation Ann Arbor MI 1998 Sayers M W Symbolic Vector Dyadic Multibody Formalism for Tree Topology Systems Journal of Guidance Control and Dynamics Vol 14 No 6 Nov Dec 1991 pp 1240 1250 Sayers M W and Riley S M Modeling Assumptions for Realistic Multibody Simulations of the Yaw and Roll Behavior of Heavy Trucks Society of Automotive Engineers SAE Paper 960173 1996 Sayers M W and Han D A Generic Multibody Vehicle Model for Simulating Handling and Braking Segel L ed The Dynamics of Vehicles on Roads and on Tracks Proceedings of the 14th IAVSD Symposium Lisse Swets and Zeitlinger 1996 pp 599 613 Rasmussen J Skills Rules and Knowledge Signals Signs and Symbols and Other Distinctions in Human Performance Models IEEE Transactions on System
62. s friction or physical properties of a vehicle It is derived purely from the mathematical definitions of velocity and acceleration When V has the units of mi h and R has units of ft the equation can be written with the appropriate units conversions and value of G G 32 17 ft sec2 ee Vmph _ Vimph gt y T496Rg 15 Rg 2 Equation 2 appears with slightly different notation in the AASHTO Green Book on pages 141 and 142 in a discussion of side friction 3 It is important to recognize that the above equations have nothing to do with vehicle properties They apply for cars motorcycles heavy trucks or anything that follows a path with radius R at speed V If the road and user are tilted by a superelevation e then an effective lateral acceleration can be defined v2 Ayeff FRG e 3 where e is superelevation m m into the turn This is the lateral acceleration that would be experienced by an observer leaning with the road Equation 3 can be solved for speed a E eae V Ayeff RG 2 4 In the above equations lateral acceleration is made nondimensional by dividing by gravity On a flat surface the ratio of lateral acceleration to gravity for a body is identical to the ratio of lateral force divided by the weight of the object For a vehicle supported by the road the lateral force divided by weight is called the friction utilization The Green Book has several equations for friction utilization One
63. s combination vehicles involving trailers A different vehicle can be selected from the library by clicking on the button to display the menu then choosing a vehicle from the menu ES eee a DS S AN I te iori lat 2 Combir AS y SS See een RRR NN ses oe Vee oe X o S Se So S A 3 Bae Fr ST eee STE SETI SESE EERTSESTR SESS SERRE na RRA AR 1S he t ata set in t ne data 1 d data are T The data screen 1C the d e lation The road geometr in the IHSDM database format imu example the s e h to the pathname les that are used to draw the road geometry t VDM RoAD in in 4 n 1 icles ion h t igure 1 ll be used le 1 F 1 f In add 1n On ve t e ina th hown 1S S data 5 THM t 10n fi ining imat Example comb the road geometry that w 3 pathname nes le conta ks to an fi the road geometry fi full ROAD ROADS ALT3 igure in F ink de fines text 11 imator program e 10Na in a h t bed shows 1 The next screen that de VDM ludes opt frame an descr D inc wire Job Number tra lt Chain Name ALT3 Initial Heading Number Regions 1 Start Station 0 000 End Station 14950 Chane Station ets paga dae de ae Y ae y E Radius Dangle 00 Do 00 00 oo 7 00 d 0 000 54156 295 117320 990 47 7
64. sections four lanes parts of the median and parts of the outside ditches The file also contains redundant information X Y Z coordinates and heading angle 36 Because all three kinds of information are written in the same file 1t is necessary to put a row in the table for any station number where any of the three kinds of information start a transition There are two difficulties caused by this format First the large number of columns 37 and the redundant information make the interchange file difficult to create by hand for debugging and performing quick what if analyses It is only practical to create the files by computer even though the information contained might be very simple The second problem is that the software reading the file must deal with extraneous rows of data that are unrelated to the information of interest For example a program dealing only with vertical geometry must read every row of data involving changes in horizontal geometry or superelevation The program must include logic to determine that those lines from the input file are not relevant and should be ignored The three kinds of design data are independent Software that reads design data could be programmed more efficiently if each kind of information were stored in a separate file Therefore we recommend that a new interchange format be designed for future work The new format should use separate files to store horizontal geometry design d
65. stem is needed by the underlying CAD system to perform calculations and to prepare graphic views of the design The location of an arbitrary point P in 3D space requires a coordinate system with three independent coordinates For the X Y Z system the three coordinates are X Y and Z each defined as the distance from the origin along a line parallel with the named axis To locate point P one would 1 start at the origin and then go X meters along the X axis 2 then go Y meters parallel with the Y axis and finally 3 go Z meters up parallel with the Z axis The global X Y Z coordinate system is theoretically sufficient for locating any point of interest in the road and any point of interest in the vehicle at a specific instant of time However it is desirable to locate the vehicle using station number S and lateral distance L from the design centerline rather than global X and Y coordinates The data in the IHSDM file define a continuous curved reference design line typically the centerline in 3D space The 2D horizontal geometry is defined by projecting the 3D line onto the global X Y plane as shown in Figure 17 Station number is defined as the distance along the 2D projection N Design line in 3D space SS rojection of design line into X Y plane Figure 17 The S L Z coordinate system The road coordinate system in VDM ROAD involves three coordinates named S L and Z The Z coordinate is the same as in th
66. t Pos vs Station All Axles WB 12 Ak 3 R lane ARR o Ay vs Station WB 12 At 3 Rane Steer SW us Staton WB 12 At3 R lane Lateral acceleration g s 0 4 Steering wheel ae deg 100 B 4 3 i 4 z i H 3 Bene H i t j i H H E H H H i i f a 1004 se 10M aa a a a a PAPI IRA A o demo a aa A DH n h H ri H 5 H E H E H i i H 1 H i i i t i i H Al 4 A A A CN O ODA do Eo A A SA H i H A H y h F H f 4 H i A i E t svsencelnencersochenens capt onosuenedfentersasnensonnafereenenanhorrsensesotsseredrantser rod snenase taba vactacsbeastonseQuceeesttbenscenerbencesesetenen tered i i 4 X H 3 i 3 a 3 i i Semitraller BO l i 0 NS p H creerme POSO DO R A GOOD an nc P 3 H i H A 4 0 2 ati peres TS TE jor so 50 ve T mgun H t H rf r p i t H i t t p H f i 5 i lanar dera ccn ro operaron rrdparen nar Junonnnergeradinocciamanciadonirnsn daria crnebanorio es p t H s t 4 H F 4 i i f z A A s i 4 p H A i i H j i f i i E i i anaoariprrencasedoronoaalorercnran Ss es a A A arn deaarsseeniassensssarec nad napetscs gt i 3 A r 4 fi i 4 de incre SEET river rie hen iesen crec es 4 i x H H y H 3 3 H H 7 H c
67. t is the name of a data set from a Runs library After the list is created the user can update all runs with a single button click This is convenient for example if the roadway design has been changed and a large number of vehicle responses are of interest for the new design 2 Runs Batch SEN lt Alt 3 Road Design 000 A BUS At 3 Rlane 0 BUS ASRianey Hil Cliimb Test MH AR3RLane MHIB AR3R lane ooo A Test Truck Alt 3R Lan SPASE jane i i P At3Rlane PA AL3R Iane lt Alt 3 Road Design A BUS AM 3 Rlane BUS AL3R lane 0 MH A 3R Lane MHB AR 3 Rlaney NHTSA Test Truck At 3R Lan Figure 16 Batch run control Comparison of VDM RoAD and TruckSim Features VDM ROAD uses mathematical models similar to those in the commercial programs TruckSim and CarSim The models are based on research conducted over the past few decades at UMTRI 9 0 The TruckSim and CarSim software packages were further 18 developed and commercialized by Mechanical Simulation Corporation MSC a private company in Ann Arbor Michigan MSC licenses maintains and supports TruckSim and CarSim TruckSim and VDM RoAD use the same graphical interface and database the same animation program and the same plotter The vehicle models have the same input parameters and the same equations are used to simulate vehicle dynamics However TruckSim is designed for mechanical engineers interested in studying vehicle dyna
68. t least the two inner loops shown in Figure 23 will be included in the IHSDM simulation of driver rules for choosing paths and speed and simulation of driver steering and speed control There are several methods by which the DPM and VDM modules can be integrated Three possibilities are mentioned below 1 The DPM and VDM programs can be merged into a single program that simulates both the vehicle dynamics and the driver control behavior 2 The DPM and VDM programs can be modified to run concurrently exchanging information as they run 3 The DPM and VDM programs can be run sequentially The DPM would be used to generate target paths and speed and then the VDM program would be run to generate the vehicle specific response to those commands Of these three methods the third is recommended Before describing how the DPM and VDM would be run sequentially the rationale will be presented for rejecting the first two approaches Problems with Running DPM and VDM Simultaneously Merging the DPM and VDM parts of IHSDM would involve taking the computer source code C Fortran etc for the two modules and combining the code into one single program module Although this approach is conceptually simple it is not practical The VDM RoAD simulation programs are large complex programs that are designed as stand alone programs They are not designed to be merged with other software and there is no documentation that describes how this wo
69. the negative superelevation is used The tire deflections are set to provide the static loads that would be expected on a flat level surface at zero speed Suspension deflections are set to provide the static spring forces that would be expected on a flat level surface at zero speed Forward vehicle speed is set to match the user specified input V S The tire and suspension deflections are generally not exactly equal to the equilibrium values although they are close For the car model aerodynamic forces and moments are neglected For all models a number of secondary effects are neglected For example load transfer is neglected between wheels due to small pitch and roll angles Consequently the first second or so of the simulation usually includes some bouncing pitching and rolling of the vehicle as it settles into an equilibrium condition on the 3D road surface in the presence of aerodynamic effects e DOs 4 INTEGRATION WITH A DRIVER PERFORMANCE MODULE A major part of the planned IHSDM system is a driver performance module DPM The DPM interacts with vehicle dynamics simulations from VDM ROAD or an alternative vehicle dynamics module to predict the motions of a driver vehicle system on the road This section describes how the DPM can be integrated with VDM ROAD Overview of Man Machine Simulation Models of man machine interactions can involve multiple levels of human behavior For example Jens Rasmussen provides a fra
70. ulation technologies developed under the U S Army Automotive Research Center ARC headquartered at the University of Michigan see the web site http arc engin umich edu ArcSim includes simulations of a _ military tractor semitrailer with a mathematical model similar to the ones in VDM RoAD The ArcSim program was modified and made to run concurrently with a separate simulation program integrated using the commercial simulation program SimuLink SimuLink is sold by The MathWorks Inc as an add on to its MATLAB package We found that two programs that both ran quickly by themselves were slowed down considerably when they were integrated A run that would take less than a minute for either program by itself took hours when the programs were connected The slowdown was due partly to different time scales associated with the models However the more significant problem was that the interprocess communication was extremely inefficient The problems of interprocess communication are in our view solvable However the amount of effort needed to solve the problem is not known The problem of having different time scales also applies to the one of integrating VDM and DPM The vehicle models include the high frequency dynamics of subsystems that require small integration time steps on the order of 0 002 sec Removing the high frequency dynamics reduces the accuracy of the simulation and is not easily accomplished The DPM does not require such a smal
71. uld be done Considerable expertise would be needed by the programmers to understand their design and modify them to work together with the DPM It should also be noted that the vehicle simulation in VDM RoAD is not performed by a single program VDM RoAD includes five separate vehicle simulation programs each customized with equations of motion for a specific class of vehicle cars and four configurations of heavy trucks At least five separate VDM DPM programs would have to be constructed Of course it is also theoretically possible to merge the five separate VDM programs and then merge that program with the DPM code This possibility is discounted because it involves even more work and requires even more expertise Modifying the VDM and DPM programs to run concurrently and exchange information is more feasible but is still not recommended There are at least two problems One is that considerable expertise is still needed to modify the programs to properly exchange information The programmers would have to master the design of the existing programs before they could modify them The second problem involves efficiency When multiple programs are run concurrently there can be a great deal of overhead and inefficiency introduced resulting in long run times UMTRI researchers have experience in how software can slow down when several simulation programs are made to run concurrently A software package called ArcSim was developed to showcase sim
72. und it is unstable and rolls over The propensity of vehicles to rollover is characterized by a rollover threshold the maximum lateral acceleration that can be sustained by the vehicle before rollover occurs Rollover threshold is measured experimentally with tilt tables that simulate the effect of lateral acceleration The rollover threshold can be obtained with detailed vehicle simulation models such as those in VDM ROAD or TruckSim The rollover threshold of a vehicle is generally not considered to depend on the geometry of a roadway The assumption underlying a tilt table is that the threshold found from the test is a valid indicator of the rollover stability under any condition A roadway designer should be concerned with the rollover threshold of the worst case vehicle that will be allowed on the road Given a rollover threshold equation 4 in Appendix A determines the maximum speed that should be allowed for a given radius of turn In general rollover thresholds are above 0 35 g s for most heavy trucks and above 0 25 g s for nearly all conventional highway vehicles These levels are much higher than the limits specified in the Green Book about 0 1 g which implies that the maximum speed for rollover considerations will be higher than the design speed assuming a realistic worst case rollover threshold We recommend that equation 4 be used to compute the maximum safe speed based on a single worst case rollover threshold A det
73. y S L system to determine the station number for the reference point Using the current vehicle speed and direction increasing station or decreasing station station numbers are computed for 10 equally spaced target points lying ahead of the vehicle The distance to the 10th target point is the preview time sec multiplied by the current vehicle speed m sec At each of the 10 target points the lateral coordinate of the target path see Figure 21 is determined by table lookup The S L coordinates for the 10 target points are transformed to the inertial X Y coordinate system The lateral location in the driver model coordinate system is calculated for each target point by first getting the inertial X and Y coordinates of the path as functions of the station at the target location Starg and then applying the transformation Ytarg Y Starg Yv cos y X Starg Xv sin y where y is the vehicle heading angle The above strategy is valid for any location of the vehicle where a station can be computed At zero speed the look ahead is set to a small distance to avoid numerical problems Complete details of the control algorithm are described in Appendix I of the VDM ROAD Users Manual Initialization of the Vehicle Position At the start of each simulation run the vehicle position and orientation are set to match the road geometry initial station number target speed and target lateral position The user of VDM RO

VDM RoAD Software - Deep Blue

Contents

Download Pdf Manuals

Related Search

Related Contents