This page contains a Flash digital edition of a book.
BUILDINGS, STRUCTURES, MODELLING & TESTING Dynamic response of tall timber buildings A


n MSc research project conducted at Imperial College London by Ishan Abeysekera, funded by the Institution of Structural Engineers, has examined the dynamic response of tall timber buildings of different configurations subjected to Tornado and Downburst wind loading. Four buildings incorporating: i) solid cross-laminated-timber (CLT) shear- wall systems, ii) glued-laminated (Glulam) frames, and iii) hybrid systems, were studied.


ortance as a sustainable alternative to concrete and steel buildings. bility the performance of tall timber buildings under transient wind loads are


mance of 4 configurations of timber buildings under transient wind loads by the comfort criteria set out in ISO10137[1].


Building 2 11-storey


ral modes along weak axis – Period (s)


ode 1: 1.17s ode 4: 0.28s


Mode 1: 1.24s Mode 4: 0.38s


) building with


as rigid frames valent springs. e most of the mes from the


nstraints with els.


Buildings of 7, 11, 20 and 30 storeys were modelled and analysed in the Finite Element software SeismoStruct. Particular attention was given to the accurate representation of the connection behaviour as well as material and structural damping. Wind velocity series were generated following typical Tornado and Downburst distributions and extensive response history analyses were performed. The results were discussed in terms of peak acceleration levels attained at selected locations throughout the height of the building considering varying frequency ranges. The software SeismoSignal was employed for post-processing. A comparison against international performance criteria for user comfort as set out in ISO10137 revealed the inability of the buildings studied here to satisfy the codified acceleration limits. Therefore, additional analyses employing tuned-mass-dampers (TMD) tailored to the most demanding frequency response in each case were carried out. It was found that TMDs with an active mass of 5% of the total building mass are able to produce up to a 50% reduction in peak acceleration levels. The models proposed as part of the present study constitute a fundamental step towards the assessment of alternative response modification strategies as well as the development of numerical tools for the future optimisation of TMD designs. A poster giving further details of this project and details of the Institution of Structural Engineers Grant Scheme are available at: http://www.istructe. org/education/scholarships-grants- and-bursaries/msc-research-grants. Diagrams and their explanations featured in this article have been taken from the poster.


Building 3 20-storey


Building 4 30-storey


Mode 1: 1.76s Mode 4: 0.54s


Building 2


• This building employs Glued-Laminated (GluLam) beams and columns, CLT cores and a CLT slab flooring system.


• Floors are modelled as rigid constraints, with all the building mass concentrated at the floor levels.


• CLT cores modelled with equivalent elastic beam-column sections.


Buildings 3 and 4


• Based on the research project by Skidmore Owings and Merrill, SOM [2].


For further information about this research please contact the project supervisor, Dr Christian Málaga-Chuquitaype, Imperial


Link element s


• These buildings use a composite framing system with concrete and timber elements.


• Walls and columns are modelled as elastic frame elements.


• Full-strength connections assumed for beam- column joints.


panels 4


• Floors are modelled as rigid constraints with floor masses and vertical loads simulated as lumped mass elements.


College London (02075 946003; E-mail: c.malaga@imperial. ac.uk).


0.01 0.1 Frequency (Hz) 1 6.Conclusions 10 0.01 0.1 Frequency (Hz) 1 10 0.01 0.1 Acknowledgements Frequency (Hz) 1 10


• Downburst loading appears to be more critical for lower buildings (Buildings 1 to 3) whilst tornado loading is more critical for tall buildings (Building 4).


• Downburst loading appears to be more critical for lower buildings (Buildings 1 to 3) whilst tornado loading is more critical for tall buildings (Building 4).


• When subjected to downburst action, all buildings fail to meet the ISO10137 criteria for the full range of frequencies analysed.


• TMDs with an active mass of 5% the total building mass are able to reduce in nearly 50% the peak accelerations. Further studies should be performed leading to optimized designs of TMDs.


Innovation & Research Focus Issue 101 MAY 2015


• TMDs with an active mass of 5% the total building mass are able to reduce the peak accelerations by nearly 50%. Further studies should be performed leading to optimized designs of TMDs.


References 2.Modelling Buildings modelled


Building 1 7-storey


Building 2 11-storey


Building 3 20-storey


Building 4 30-storey


Dynamic Respon Supervisor: D


Timber buildings have well established economic, social and environmental advantages over other structural options. In particular, their low carbon footprint and high strength to weight ratio makes tall timber construction an attractive solution for satisfying the pressing housing demands in densely populated areas at minimum environmental costs. However, due to the low mass and flexibility typical of tall timber construction, concerns regarding their dynamic behaviour have been raised.


1.Introduction • Tall timber buildings are gaining importance as a sustainable alternative to concrete and steel buildings.


• Due to their low mass and high flexibility the performance of tall timber buildings under transient wind loads are under question.


• This study aims to assess the performance of 4 configurations of timber buildings under transient wind loads by means of numerical analysis against the comfort criteria set out in ISO10137[1].


Dynamic Response of Tall Timber Buildings Ishan K Abeysekera


Supervisor: Dr Christian Málaga-Chuquitaype Natural modes along weak axis – Period (s) 3.Wind Loading Building 1


Two types of wind profiles are considered: i) downburst, and ii) tornado (see Figure on the right). Generated wind load profiles were applied to all 4 building under consideration in the direction parallel to the strong axis (i.e. causing bending about the weak axis).


Mode 1: 0.68s Mode 3: 0.16s


Based on experimental data the analysis was run for various values of damping ranging from 3 to 13%of the critical value.


4. Tuned Mass Damper (TMD)


Mode 1: 1.17s Mode 4: 0.28s


• Cross Laminated Timber (CLT) building with steel connections.


The effectiveness of Tuned Mass Dampers (TMD) to mitigate the high levels of floor acceleration was examined. Up to a 50% reduction in the peak acceleration values was possible in most cases by employing TMD with a mass equivalent to 5% of the building mass. No design optimization was performed.


5.Results


Rigid panel


1


• Floor modelled as rigid constraints with concentrated masses at floor levels.


• The CLT panels were modelled as rigid frames and the connections as equivalent springs. This is justified by the fact the most of the flexibility of the building comes from the connections (shown below). Results and conclusions


Normalized Downburst wind velocity profile


Wind loading and Tuned Mass Damper (TMD)


Two types of wind profiles are considered: i) downburst, and ii) tornado. Generated wind load profiles were applied to all 4 buildings under consideration in the direction parallel to the strong axis (i.e. causing bending about the weak axis). The effectiveness of Tuned Mass Dampers (TMD) to mitigate the high levels of floor acceleration was examined.


0.0 1.0 2.0 3.0 4.0


• This building employs Glued-Laminated (GluLam) beams and columns, CLT cores and a CLT slab flooring system.


• Floors are modelled as rigid constraints, with all the building mass concentrated at the floor levels.


• CLT cores modelled with equivalent elastic beam-column sections.


0.0 0.2 0.4 0.6 0.8 1.0


0 0.5 1 Normalized wind speed Buildings 3 and 4 1.5 0.0 0.5 1.0 Normalized wind speed


The floor acceleration histories were post-processed in order to compare the building response with codified criteria. Results are presented below for the top floor in all buildings under study with approximately 10%viscous damping.


The floor acceleration histories were post-processed in order to compare the building response with codified criteria. Results are presented below for the top floor in all buildings under study with approximately 10% viscous damping.


Building 1 - Downburst


w/o TMD with TMD Office Criteria Residential Criteria


0.1


Rigid panel


0.01 0.1 1


Figure: Modelling of CLT panels Building 1 - Tornado


Frequency (Hz) 1


w/o TMD with TMD Office Criteria Residential Criteria


0.1 1


Node Rigid bar


10 1 Bulding 2 - Downburst 0.1


Link element s


w/o TMD with TMD Office Criteria Residential Criteria


Building 3 - Downburst


• Walls and columns are modelled as elastic frame elements.


w/o TMD with TMD Office Criteria Residential Criteria


• Full-strength connections assumed for beam- column joints.


0.1 0.01 0.1 Frequency (Hz) Building 2 - Tornado


w/o TMD with TMD Office Criteria Residential Criteria


0.1 1


w/o TMD with TMD Office Criteria Residential Criteria


0.1


• Floors are modelled as rigid constraints with floor masses and vertical loads simulated as lumped mass elements.


0.01 1 10 0.1 1 Frequency (Hz) Building 3 - Tornado 10 1


• Based on the research project by Skidmore Owings and Merrill, SOM [2].


• These buildings use a composite framing system with concrete and timber elements.


1.5 Building 2


Mode 1: 1.24s Mode 4: 0.38s


Normalized Tornado wind velocity profile


Mode 1: 1.76s Mode 4: 0.54s


• When subjected to downburst action, all buildings fail to meet the ISO10137 criteria for the full range of frequencies analysed.


The financial support of the IStructE through a MSc Research Grant for the research described in this poster is gratefully acknowledged.


[1] ISO 10137:2007 Bases for design of structures - Serviceability of buildings and walkways against vibrations.


[2] Skidmore Owings & Merrill. Timber Tower Research Project. Available from: http://www.som.com/ideas/research/timber_tower_ research_ project [3] SeismoStruct v6.5 available from http://www.seismosoft


www.innovationandresearchfocus.org.uk


Peak accelration top floor (m/s2)


Peak accelration top floor (m/s2)


Peak accelration top floor (m/s2)


Peak accelration top floor (m/s2)


Normalized height


Peak accelration top floor (m/s2)


Peak accelration top floor (m/s2) Normalized height


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8