Simulation of Thermal Images Presented by Vijaya Priya
Simulation of Thermal Images Presented by Vijaya Priya Govindasamy Masters Student November 11 th 2003
Outline n Personal facts n Simulation of Thermal Images n n Segmentation of Thermal Images n n Objective Introduction to Mu. SES Models used for simulations and simulation results Objective Idea of Watershed Segmentation Preliminary results Conclusion 2
Personal Facts n n Bachelor of Engineering – Electrical and Electronics Engineering Communication Between Intelligent Systems Through Controller Area Network (CAN) 3
Applications of CAN n Passenger cars n n Factory Automation n n CAN is used as the in-vehicle network for the engine management, the body electronic like door and roof control etc CAN is used to interconnect machines, process control units and production sub-systems Industrial Machine control n CAN finds its application as an embedded network for industrial machine control, which links single devices as well as sub-systems http: //www. can-cia. de/applications/passengercars/tirepressure. html 4
Outline n Personal facts n Simulation of Thermal Images n n Segmentation of Thermal Images n n Objective Introduction to Mu. SES Models used for simulations and simulation results Objective Idea of segmentation Some preliminary results from Spring 2003 Conclusion 5
Motivation n n Simulate the infrared prediction of the under vehicle parts Investigate the presence of unusual parts in under vehicle scene, like the presence of cold objects near the muffler The thermal images are simulated using the software Mu. SES Develop under vehicle automotive parts using the software Rhino 3 d Simulate thermal images for the models created, downloaded CAD models, and scanned and reconstructed models 6
Need for Simulation n Reduces the time and cost of developing prototypes for test purposes n Gives more flexibility in controlling the parameters n Infrared prediction measurements are accurate and faster 7
Mu. SES n n Mu. SES – Multi-service Electro-optic Signature Mu. SES is a thermal modeling tool, used to model the steady state and transient distribution of heat over complex surface descriptions of component systems. Mu. SES models 3 -D conduction, convection, and radiation. The output from Mu. SES is the temperature map of the component system which can be viewed using the integrated post-processor 8
Mu. SES Interface Source: Mu. SES Manual 9
Poor Mesh n Mu. SES treats overlapping and adjacent vertices as unconnected parts n The polygons should be uniform with aspect ratio of unity n Overlapping elements should be embedded on the host surface Unconnected parts Source: Mu. SES Manual Overlapping elements 10
Meshing Requirements n The characteristics of a good quality mesh for import into Mu. SES are: n n n n All adjacent polygons share common vertices (equivalenced mesh) All polygons are 3 or 4 -sided (triangles or quads) All polygons are convex All polygons have an aspect ratio near unity(e. g. no long and skinny polygons) Polygons are spread uniformly across the surface (e. g. avoid fans of polygons) No overlapping or repeated facets Surface mesh only (e. g. thin plates represented by their exterior surface only) 11
Part Types n Assigned temperature parts n n n Temperature curves are assigned for such parts Surface condition and Bounding box condition are assigned Calculated Temperature parts n n Only initial temperature value is given Convective heat coefficient value is assigned Temperature values are calculated as a result of numerical solution Surface condition, Material properties and Bounding box condition are assigned 12
Properties n Material Type n n n Thickness n n n Does not mean anything with the geometry Used only for calculation Surface condition n Given for calculated parts Based on the type of material the numerical solution is carried out Defines the emmisivity property Paint codes are also used Convective coefficient n n The value of ‘H’ is given for calculated parts Used in numerical solution 13
Property Definition n Sample Property Definition n Engine Interior n n The temperature is very high due to combustion Engine Exterior n In real time the coolant fluid reduces the outside temperature Engine Temperature curve 14
Property Definition n Sample property definition n Muffler Interior n n Exhaust gas temperature Muffler Exterior n Outside heat loss to the environment – convection & radiation Muffler Temperature curve 15
Solution Analysis n Start time and End time n Step size n Tolerance slope and Tolerance n View factor rays n Number of iterations n Accuracy of the solution n Convergence parameter 16
Models Simulated n Simple rod model to understand the working of Mu. SES n Exhaust system model n Car underbody model n IRIS lab scanned Water neck and Muffler Models n Toyota Model 17
Rod Model of a simple rod n n n The model is designed and studied to get familiar with the working of the software Thermal properties are assigned to one end of the rod and the radiation and conduction effects are studied The left end temperature is set as 1000 degree Celsius Maximum temperature Minimum temperature 18
Results Rod simulation result n n The result is found to be exact Maximum temperature The bottom of the temperature is less than the temperature at the left end Minimum temperature 19
Exhaust System Model n Experimental setup n The model consists of the following parts: n n n n Exhaust System Floor pan Muffler Inlet and Outlet Pipes Manifold The model is created using Rhino 3 d and imported into Mu. SES as a wavefront (. obj) file The maximum temperature is assigned as 250 degree Celsius The simulation is for a period of 30 minutes The thermal properties and boundary conditions are assigned in Mu. SES 20
Results n n n The heat distribution is studied Exhaust system simulation result The radiation and conduction effects is visible The inlet pipe is at the maximum temperature 21
Car Model n n n Under body Engine Muffler Transmission Inverter Radiator Catalytic converter Starter Motor Batteries Intake and Exhaust Manifold Wheels Upper body 22
Simulation Result of Car Model 23
Assumptions n Construction of model n Reasonable temperatures n Medium complexity n Construction of the model Tolerance slope of solution convergence - 5 e-007 24
Solution Justification Simulation result of Upper Body & Under body 25
Temperature Curves Catalytic converter Engine Muffler 26
Thermal Images of Upper Body Car Upper body Infrared prediction View 1 View 2 27
Thermal Image of Under Body Infrared Prediction of Car Under body 28
Simulation of Laser Scanned 3 d Models Model of a Waterneck System n n n The water neck is connected between the radiator and the engine block The water neck is a part of the cooling system The maximum temperature is assigned to the engine block as 200 F and the radiator box is assigned 100 F 29
Result of Simulation Waterneck model simulation result n n The simulation duration : 10 minutes The conduction occurs at the end of the water neck embedded on the engine block 30
Scanning and Reconstruction of Muffler Original color image of a car muffler Front view Range image of a car muffler Experimental setup of IVP range scanner Back view Superquadric fitted 3 D model of the car muffler Courtesy: Umayal 31
Result for Scanned Muffler Real time heat distribution of a Muffler 32
Toyota Tundra Undercarriage Source: http: //www. 3 dcadbrowser. com 33
Simulation Result 34
Simulated Thermal Images 35
Toyota Tundra Engine Source: http: //www. 3 dcadbrowser. com 36
Toyota Engine Model n n The model had 1104 parts when downloaded With the help of Yohan, the 1104 parts were extracted out as VRML files The VRML files were merged according to the various part definitions Meaningful parts like exhaust manifold, cylinder blocks and heads, piston, injectors were formed 37
Engine Model Toyota Engine Model Crank Oil pan Exhaust Manifold 38
Bioheat transfer n Transportation of thermal energy in a living tissue is a complex process involving Head Model n n n n Conduction Convection Radiation Metabolism Evaporation Phase change Various methods developed to determine thermal conductivity and diffusivity of biomaterials Source: http: //imaging. utk. edu/~rangan/3 D%20 Database/iris. htm 39
Bioheat Transfer n Human thermal model n n Head Model in Mu. SES Passive System: Described by equations resultant from the application of heat and mass balances to a tissue control volume. Temperature control system: Responsible for the maintenance of the human body’s temperature. 40
Heat Transfer within the Tissue n Heat transfer equation (Pennes) n n n n n tissue density (kg/m 3) c tissue specific heat (J. kg-1. o. C-1) T tissue temperature (o. C) t time (s) k tissue heat conductivity (W. m-1. o. C-1) b blood density (kg/m 3) b blood perfusion rate (s-1) Cb blood specific heat (J. kg-1. o. C-1) Tar, i arterial blood temperature inside the ith cylinder (o. C) q is metabolic heat production (W/m 3) 41
Example of Simulation http: //www. thermoanalytics. com/services/biothermal. html 42
Outline n Personal facts n Simulation of Thermal Images n n Segmentation of Thermal Images n n Objective Introduction to Mu. SES Models used for simulations and simulation results Objective Idea of segmentation Some preliminary results from Spring 2003 Conclusion 43
Objective n Segment meaningful automotive parts from thermal images of under vehicle scene n Review some of the common segmentation procedures n Implement the algorithms for thermal image segmentation 44
Watershed Transformation n Visualize an image in three dimensions: two spatial coordinates versus gray levels. We consider three types of points: n n n Points belonging to a regional minimum Points at which a drop of water, if placed at the location of any of those points, would fall with certainty to a single minimum --- Catchment basin/Watershed Points at which water would be equally likely to fall to more than one such minimum --- Divide lines/Watershed lines Source: Digital image processing Gonzalez and Woods 45
Watershed Transformation Original Image Watershed transformation Source: http: //cmm. ensmp. fr/~beucher/wtshed. html Topographic Image Segmented Image 46
Preliminary Results From Matlab Original Image Source: http: //imaging. utk. edu/safer Segmented Image 47
Watershed Transformation n The algorithm is decomposed into two steps: n n The sorting step: Initial sorting of the pixels in the increasing order of their gray values The flooding step: Involves the progressive flooding of the catchment basins of the image n Currently developing the algorithm in C++ n Results available by the end of this semester Source: Watersheds in digital spaces: an efficient algorithm based on immersion simulations – L. Vincent and P. Soille 48
Conclusion n n Simulation of thermal images eases the analysis of the automotive parts under different temperature conditions Limitations: n n Meshing Simulation duration Segmentation of thermal images helps in highlighting the under vehicle automotive part Future work is to concentrate a particular method to segment thermal images 49
Thank You 50
- Slides: 50