The UK Wind Engineering Society's view on the use of experimental and computational methods in wind engineering

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

Outline

• Introduction
• Wind engineering tools
• Pedestrian level wind studies
• Structural wind load and façade wind pressure studies
• The future of wind engineering
• Conclusions and recommendations

Introduction

Introduction

• Principal Authors:

Stefano Cammelli [WSP]

Francesco Dorigatti [RWDI]

Daniel Hackett [RWDI]

Alex Ping To [Arup]

Bernardo Vazquez [Buro Happold]

​

• Additional Contributors:

Gordon Breeze [BRE]

David Hargreaves [University of Nottingham]

Zheng-Tong Xie [University of Southampton]

​

• Freely available from the UK Wind Engineering Society website [www.windengineering.org.uk]

Introduction

• Prepared by the UK Wind Engineering Society to provide a brief overview of the state of the art of both experimental wind tunnel testing and computational fluid dynamics (CFD) methods routinely used in the industry
• A brief document aimed at a wider audience of non-experts who are required to interpret the output of wind engineering studies produced by experts in the field
• Focused on: buildings in the context of an urban environment
• Focused on: pedestrian level wind studies; structural wind load studies; and façade wind pressure studies

Introduction

Proceedings of the Institution of Civil Engineers, Volume 19 Issue 4, August 1961, pp. 449-472.

Wind engineering tools

Physical model wind tunnel testing

• Tests conducted in atmospheric boundary layer wind tunnel facilities
• Cross-section: typically ranging between 2m and 5m in width and 1.5m and 3m in height
• Facilities featuring a 15m to 20m long up-stream fetch equipped with flow-conditioning devices
• Working section normally equipped with a 360°-rotating turntable
• Proximity models covering generally a radius of 300m to 600m from the project site
• Typical model scales in the 1:200–1:500 range

​

​

• Publications that set the standards for best practice in the execution of wind tunnel testing are provided

Numerical modelling

• DNS: prohibitively expensive and impractical for most engineering applications
• Scale-Resolving Simulations (SRS) and non-Scale-Resolving Simulations are introduced
• Pros and cons of RANS, LES and “hybrid” approaches such as DES are presented and discussed
• Publications that set the standards for best practice in the execution of CFD simulations are provided

Pedestrian level wind studies

Pedestrian level wind studies

• Conducted either during concept design – if the planning application is submitted at the end of RIBA Stage 2, or otherwise during RIBA Stage 3
• Assessment methodology, approaches and techniques for the execution of pedestrian level wind studies (including suitable criteria to be used) are presented
• The "City of London wind microclimate guidelines" are introduced: the rules set within this document align the work of wind

engineering specialists, wind tunnel testing

laboratories and CFD practitioners; reduce the scope for discrepancies in the

approach; and facilitate the work of peer reviewers

Pedestrian level wind studies

• Physical model wind tunnel testing

Pedestrian level wind studies

• Numerical modelling

Pedestrian level wind studies

• Physical model wind tunnel testing vs. numerical modelling

Structural wind load and façade wind pressure studies

Structural wind load and façade wind pressure studies

• Typically conducted post-planning either during RIBA Stage 3 or during RIBA Stage 4
• Assessment methodology, approaches and techniques for the execution of structural wind load and façade wind pressure studies are presented
• Wind tunnel techniques (HFFB, HFPI and aeroelastic) to be used at different stages of the design are introduced

• The use of CFD in the context of structural wind load and façade wind pressure studies largely remains in the research phase (e.g. for simple geometries in isolation, discrepancies in terms structural wind loading between wind tunnel testing and CFD simulations employing SRS techniques have been reported to be of the order of 5%; despite this, in more realistic scenarios of tall building in an urban context, greater discrepancies have been shown in the literature)

• Physical model wind tunnel testing

Structural wind load and façade wind pressure studies

• Numerical modelling

Structural wind load and façade wind pressure studies

The future of wind engineering

The future of wind engineering

Physical model wind tunnel testing

• Advancements in the speed, accuracy and level of details of wind tunnel model construction
• Increased efficiency through further process automation
• Better spatial resolution / greater number of measurement locations
• Active turbulence / gust generation to improve the quality of inlet turbulence for large length scales
• Increased accuracy in wind speed measurements at challenging locations such as balconies / terraces
• Combination of pedestrian level wind assessments with other environmental factors
• Increased synergy with CFD

Numerical modelling

• Increased synergy with wind tunnel testing
• Optimisation algorithms for a broader range of variables that affect the comfort of pedestrians
• CFD to begin to be used for aerodynamic optimisation and form finding studies of tall buildings
• Multi-fidelity simulation techniques will become more common and they will help reducing computational costs whilst maintaining the high accuracy required by the design process
• Machine learning will also play a crucial role in augmenting CFD tools
• Quantum computing as a potential enabler for DNS within commercial applications?

The future of wind engineering

Conclusions and recommendations

Conclusions

• The successful use of either experimental or computational methods is subject to the knowledge and experience of the user (or – in other words – an inexperienced user could get incorrect results out of either a wind tunnel test or a CFD simulation)
• This point is especially relevant for computational methods where the low barrier to entry and ease of manipulating the basic physics makes it particularly susceptible: the sole ability to correctly and proficiently perform a CFD simulation does not in fact qualify a CFD user as a wind engineer
• Considering the current state-of-the-art in pedestrian level wind studies in the UK, both experimental and computational methods are viable and useful for the purpose of assessing wind microclimate conditions
• Given the advantages and the disadvantages of the two approaches, there are situations where one method is more appropriate than the other – but equally – there are plenty of cases where the two methods could be used interchangeably without sacrificing the accuracy of the assessment
• There are often good reasons to make use of both experimental and computational methods on a project to take advantage of the inherent benefits of one method whilst offsetting its disadvantages

Conclusions

• Both experimental and computational methods should comply with the respective best practice guidelines, which state the minimum requirements to be met by the chosen methodology for its results to be robust
• Despite encouraging results in the research of SRS techniques for structural wind load and façade wind pressure studies, validations of CFD simulations in this context remain limited to a rather restricted number of simplified case studies and applications

Recommendations

• Pedestrian level wind studies

- Both experimental and computational methods are viable and useful tools for the

purpose of assessing wind microclimate conditions. Whilst these tools are – in

general – largely interchangeable, the reader should be mindful that non-SRS

techniques such as RANS simulations have the potential to underpredict the level

of wind comfort in low wind speed regions (often dominated by ‘gusts’)

- An integrated ‘CFD-wind tunnel testing’ approach can be particularly

advantageous, especially when CFD is used during the early stages of a project to

inform the massing (including the need of a podium, façade set-backs, etc.) as well

as the orientation of the proposed building(s)

​

• Structural wind load and façade wind pressure studies

- Structural wind load and façade wind pressure studies should only be performed

experimentally, i.e. in a boundary layer wind tunnel facility

- The sole use of CFD simulations (SRS techniques) in support of structural and

façade engineering design is – at the time the present document is published –

not recommended

Thank you!