Enabling numerical Simulations in semantic 3D City Models using CityGML

Enabling numerical Simulations in semantic 3D City Models using CityGML

13th South East Asian Survey Congress, Singapore

Arne Schilling Stefan Trometer

virtualcitySYSTEMS GmbH CADFEM GmbH

Berlin, Germany Grafing, Germany

Arne Schilling, virtualcitySYSTEMS , SEASC 2015 2

Geoportal Solutions

What is Urban Simulation?

Challenge

Project 1 – Urban Air Flow

Project 2 – Urban Blast Simulation

Discussion

Outline

3 Arne Schilling, virtualcitySYSTEMS , SEASC 2015

Geoportal Solutions

Points of Interest


Geoportal Solutions

Integration of OGC Services


Geoportal Solutions

Spatial & Semantic Queries



Simulation of physical phenomena in urban environments using virtual mockups / 3D city models



Mechanical simulations of building structures

Acoustic studies / blast simulations

Wind field simulations

Simulation of flood events

…


GIS is not designed for performing complex physical computations

• Strengths of GIS are data management, spatial analysis and visualization

• 3D city models are created using remote sensing and automatic feature extraction methods, not using CAD software

Challenge


Challenge




Physical simulations are frequently done in product design

• Using Computer Aided Design (CAD) and Computer Aided Engineering (CAE) software such as ANSYS

• Based on the Finite Element approach -> Numerical Simulation


Challenge




Physical simulations are frequently done in product design

• Using Computer Aided Design (CAD) and Computer Aided Engineering (CAE) software such as ANSYS

• Based on the Finite Element approach -> Numerical Simulation

Technological gap between GIS and CAE Worlds

• CityGML cannot be loaded by CAE software (e.g. ANSYS). Geometrical representations are very different -> we need conversion tools

• Requirements on data quality are different

• Simulation results are mostly stored in proprietary formats


Challenge

Create a workflow from 3D Geographic Information Systems to Simulation frameworks and back


Simulating the wind flow in urban environments for calculating pedestrian comfort levels

Collaboration with IDAC Ltd.

Part of Environmental Impact Assessment

Project 1 - Urban Air Flow

Partners

Pedestrian comfort at ground level (streamlines)

Velocity contours at a height of 7m and 150m

Novosibirsk State University of Architecture and Civil Engineering


Data set: Frankfurt a.M. Germany: many high rise buildings

Approach: Computational Fluid Dynamics (CFD) analysis

Typical wind loads described by velocity profiles were applied

The Navier-stokes equations were solved along with the 𝑘−𝜀 turbulence model.

The analysis was performed using ANSYS CFX “high-performance, general purpose fluid dynamics program”


13

From http://www.ansys.com


Data Link - Geometry Processing Format Conversion from CityGML to STEP

Recreation of geometry and topology

Error detection and Geometry healing

• no self intersections

• no non-manifold geometries

• no interior surfaces

• no holes (“watertight”)

• no stray elements

Defeaturing


14

Geometry healing, e.g. removing inner surfaces (top) and resolving self-intersections (bottom)


Numerical Computations Meshing -> ca. 43 Mio. elements

Setting physical properties (e.g. air density)

Boundary conditions

Running CFD solver


15


Presentation Cross sections

Coloring 3D city model

Charts (velocity contours) for reporting


16


Background Unexploded bombs from WW2

Safety perimeters must be set up quickly

Critical infrastructure facilities in the area must be determined

Project 1- Urban Blast Simulation

Safety perimeters (red: evacuation, blue: curfew)

Recovery of defused bomb

Disposal of defused bomb


Approach Data exchange using STEP

Geometry processing (healing)

Shockwave propagation using AUTODYN and APOLLO solvers

Reintegration of simulation results in 3D city model


Partners

Funding

Web-Interface for visualization of numerical results



Partners

Funding

Fast solution by the APOLLO Blastsimulator in cooperation with the Fraunhofer EMI

Objectives Easy accessibility to decision-makers (fire fighers, police, bomb

disposal units etc.)

Integrated and automatic workflow for scenario definition, numerical simulation and presentation of results

Derivation of damage indicators (e.g. broken windows, masonry)



Partners

Funding

Derived detailed building models from city models

Analysis and optimization of detailed structural members

H

h

R

x

z

Source: GEORES (www.geores.de)

Further Research Investigate vulnerability of buildings to blasts (simulate structural

damages) -> planning purposes

Increase the resilience of facilities to terroristic threats


Restrictions • CityGML LOD 2 city models (no windows, balconies etc, roof information), no

information on interior structures

• Simulated area limited to 1x1km (limitation in ANSYS)

• STEP CAD models are not geo-referenced, extension required

Discussion

Steady-state air flow LOD3 model

Super-detailed structural simulation


Discussion

Benefits • Solution can be added to existing Spatial Data Infrastructures

• Solution relies on open standards (CityGML, OGC services) and industry standards (STEP)

• Data is stored and managed in a central data repository (Oracle, PostgreSQL)

• Will be possible to reintegrate simulation results into 3D city model database (damage indicators, wind loads)


Current Technology Partners

Urban Simulation Network

Urban Simulation

&

Urban Blast Protection

Flooding & Tsunami Protection

Urban Air Flow

Smoke & Pollution Development

Crowd Movement

Novosibirsk State University of Architecture and Civil Engineering


Conceptual Approach

ETL processes

Open architecture • Other applications can be

plugged into / connected to the 3D SDI

virtualcitySYSTEMS 2015

http://www.businesslocationcenter.de/wab/maps/main/



Thank You Arne Schilling Stefan Trometer

virtualcitySYSTEMS GmbH CADFEM GmbH

[email protected] [email protected]

http://www.virtualcitysystems.de http://www.cadfem.de