Industrial Sponsorship

Dynamic Performance Investigation of Base Isolated Structures

By Ather K. Sharif

Guidelines for the Measurement, Evaluation and Implementation of Base Isolation Systems to Attenuate Ground Vibration

Contents of Proposed Standard



1 Scope

2 Normative references

3 Definitions

4 Objectives of Base Isolation

5 Appointment of Specialist Engineer and Duties

6 Evaluation Procedure Flowchart

7 Measurements for sites subject to vibration from man made sources

7.1 Scope of Survey

7.2 Information required by Specialist Engineer

7.3 Measurement Parameters

7.4 Characteristics of Instrument Chain

7.5 Measurement Locations

7.6 Coupling of sensors

7.7 Measurement samples

8 Soil Dynamic Properties

9 Structural Options

10 Alternative Vibration Control Options

11 Prediction and Evaluation of Building Response (Traditional Foundations)

11.1 Evaluation in terms of effects on the Structure

11.2 Evaluation in terms of Human response

11.3 Evaluation in terms of Sensitive Equipment or Processes

11.4 Identification of required Attenuation

12 Factors to consider in the modelling of Base Isolated Buildings

12.1 Alternative Isolator Options and their Deployment

12.2 Wind Induced Motion Control Devices

12.3 Potential benefits/disbenefits of Base Isolation

13 Risk Analysis

14 Validation by Test Structures

15 Structural Design Implications

16 Architectural Design Implications

17 Services Design Implications

18 Outline Specification for the Supply and Testing of Isolators

19 Construction implications for Base Isolated Building

20 Cost Implications of Base Isolation

21 Validation Measurements for Quality Assurance

22 Building Notices and Documentation

23 Maintenance and Condition Monitoring

Annexe A (Informative): Bibliography

Guidelines for the Measurement, Evaluation and Implementation of Base Isolation Systems to Attenuate Ground Vibration

1 Scope

This International Standard gives guidelines for the measurement and evaluation of sites affected by vibration and groundborne noise from man made sources, with a view to control by Base Isolation. It specifies how the predicted levels should be evaluated, dealing with structure type, human comfort commensurate with nature of occupancy and with regard to sensitive equipment.

This standard encourages the assessment of other vibration control options, and guides their evaluation in terms of relative performance and cost to ensure that Base Isolation is adopted where it is technically appropriate and cost effective to do so.

Where a need for Base Isolation is established, this Standard gives guidelines for the prediction of building response, indicating limitations of traditional single-degree-of-freedom analysis, and the advantages of more involved multi-degree-of-freedom models. It provides guidance for the implementation of Base Isolation systems in new buildings, discussing the various isolators that may be used. It offers guidance on the evaluation and control of wind induced motion. This standard specifies how the performance characteristics of the Base Isolation system should be evaluated, in terms of the possible benefits and disbenefits along with guidance on risk analysis.

This standard gives guidance on the structural, architectural, building services, construction and cost implications of Base Isolation for the building. It specifies the level of site inspections necessary to limit risk of short circuits. It gives outline guidance for the supply and testing of isolators.

It provides guidance on measurements to validate performance for quality assurance and indicates requirements for ongoing condition monitoring, maintenance, building notices and documentation.

This standard does not indicate limit values for the structure or provide vibration and groundborne noise limits for human occupation or any equipment/process. It does not provide in detail the specification requirements for the supply and testing of isolators, or offer a specific method of dynamic analysis. It does not refer to the use of Base Isolation as a means of "earthquake" ground motion control.

2 Normative references

The following standards contain provisions, which through reference in this text constitute provisions of this International Standard. At the time of publication, the editions indicated were valid. All standards are subject to revision, and parties to agreements based on this International Standard are encouraged to investigate the possibility of applying the most recent editions of the standards indicated below.

BS6177:1982, Selection and use of elastomeric bearings for vibration isolation of buildings

ISO 4866:1990, Measurement of vibrations and evaluation of their effects on buildings

ISO2631-1:1997, Evaluation of human exposure to whole body vibration - Part 1

ISO2631-2:1989, Evaluation of human exposure to whole body vibration- Part 2: Continuous and shock-induced vibration in buildings (1 to 80Hz)

BS6472:1992, Evaluation of human exposure to vibration in buildings (1 to 80 Hz)

DIN 4150 part 2:1992,

BS EN 60068-2-64:1995, Environmental Testing, Part 2 Test Methods

ISO 3010:1988, Bases for design of structures - Seismic actions on structures

ENV 1991-2-4:1995, Eurocode 1, Part 2-4, Wind Actions on Structures

ISO8569:1989, Shock and vibration sensitive electronic equipment

ISO1996 part-1(1982), part-2(1987), part-3(1987) Description and measurement of Environmental Noise

ISO 10137:1992, Design of structures - Serviceability of buildings against vibration

3 Definitions

Base Isolation

Isolators placed between the foundation and superstructure of a building to attenuate the effect of vibration and/or groundborne noise from man made sources. This usually occurs at the base of a building, although the isolators may be introduced at other levels.


Resilient element introduced between un-isolated and isolated part of the building. Typically comprise natural/synthetic rubber bearings or steel coil springs.

Rigid Body Mount Frequency

Refers to a vertical natural frequency assuming the building to be a rigid body on the isolators, and its value may dictate the type of isolator used.

Insertion Loss

The benefit or disbenefit of Base Isolating a building compared with the performance of an un-isolated building, were it to have been built at the site.

Short Circuit

Bridging between an isolated building and any adjoining un-isolated structures or the ground, that may compromise performance of the Base Isolation system.

The Client

The Client may be the Developer, Owner Occupier, Tenant or their Agent (e.g. Architect, Structural Engineer), responsible for the development of a site effected by vibration and groundborne noise from man made sources.

The Specialist Engineer

The Specialist Engineer is responsible for advising the Client upon the effects of building on a site subject to vibration and groundborne noise from man made sources. This can include providing options for attenuation which may include Base Isolation.

Isolator Supplier

The Isolator Supplier is responsible for the supply of isolators to an agreed specification. The Isolator Supplier may undertake testing of the isolators to prove conformance with the specification, and may also be responsible for site installation.

Independent Testing Agency

An Independent Testing Agency may be responsible for the testing of isolators and/or undertake measurements on site to validate performance.

Main Contractor

The Main Contractor is responsible to implement the Base Isolation system in accordance with the plans and specifications produced by the Specialist Engineer.

4 Objectives of Base Isolation

Base Isolation may be implemented in buildings to attenuate groundborne vibration and groundborne noise due to man made sources (e.g. railways or machinery) [1,2,3].

NOTE Base Isolation for earthquake applications are not covered by this standard.

5 Appointment of Specialist Engineer and Duties

The Client may appoint a Specialist Engineer to undertake site measurements, evaluation and design to permit a development, fit for purpose in respect of vibration and groundborne noise, commensurate with the structure type and intended use.

Where Base Isolation is to be adopted, the Specialist Engineer should advise the Client and /or his Agents of implications that this may have on other disciplines (e.g. Architect, Structural Engineer, Services Engineer, Quantity Surveyor and Contractor). The role of the Specialist Engineer may include site inspections and testing to validate performance. The Specialist Engineer may pass on responsibility for particular aspects of design of the Base Isolation system, that may overlap with the duties of any Agent appointed by the Client, with their agreement.

EXAMPLE Static design aspects associated with the Base Isolation system may be passed to the Structural Engineer, provided there is clear distinction in responsibility.

NOTE Isolator Suppliers may offer advice on the use of their products, but may not all have expertise to advise upon the in-situ dynamic effects of using their products.

6 Evaluation Procedure Flowchart

7 Measurements for sites subject to vibration from man made sources -with a view to control by Base Isolation

This section describes the measurements needed to characterise the source to allow an evaluation of building response, with regard to control by Base Isolation.

7.1 Scope of Survey

A survey may be required for screening assessment, outline planning considerations or detailed design. The scope of the survey should be compatible with its intended use. Initial surveys provide useful data, which allows more detailed surveys to be better designed, both in terms of where and when measurements are taken, but also to influence the instrumentation used.

7.2 Information Required by the Specialist Engineer

Where possible, the Client should provide and the Specialist Engineer seek to obtain the following:

Site plans drawn to scale, showing the existing development site along with the location of existing or future vibration source(s). These plans should also provide the footprint of proposed or extended building(s), to allow distances to the source(s) to be estimated. They should indicate formation level of basements, foundations and type/depth of piles if they are envisaged. If the source is below ground level (e.g. a railway in a tunnel), the depth and alignment should be shown in section, along with proximity of proposed foundations of the new or extended building.

Site investigation report(s) that provide geology of the site. The site investigation report(s) should describe ground conditions and indicate relevant soil properties for an area encompassing the source(s) and the proposed building(s) to a sufficient depth, identifying changes in strata and position of water tables (see clause 8).

Details on the proposed structure, as is available at the time of planning the survey. Including anticipated or preferred options for; foundations, structural frame, floor construction, along with plans identifying floor use. This allows sensitive areas of a development to be identified, allowing appropriate allocation of survey resources.

Limit values for sensitive equipment or processes if envisaged.

Details from Local Planning Authorities, on any policy with regard to noise and vibration for developments in their area.

Foundation details for the existing or proposed source(s), including their dimensions and form of construction.

Information which is source specific.

EXAMPLE For a railway source, the number of tracks, whether they are jointed or welded, along with presence of any points and the nature of track sub-base, and the elevation which may be at-grade, in a cutting, on an embankment or in a tunnel. The railway operator should be consulted to identify current usage, especially freight or high-speed traffic and identify if there are any expected or planned changes to intensity, type of traffic, or the track type. The railway operator will not however be prohibited from making any changes in the future (see clause 13).

7.3 Measurement Parameters

Measurement parameter(s) should be appropriate to the target limit values.

Where vibration effects on the structure are to be evaluated, Peak Particle Velocity (PPV) is the preferred measurement (see ISO 4866:1990).

For evaluation to assess human occupation, acceleration is the preferred measurement although velocity can be used, to obtain running r.m.s values, KB values or Vibration Dose Values (see ISO2631-1:1997, ISO2631-2:1989, DIN 4150-2:1992, BS6472:1992).

Where the effect of vibration on sensitive equipment is evaluated, preferred parameters are often frequency dependent (tri-partite: displacement at low, velocity in the mid-range and acceleration criterion at high frequencies). The range of equipment and processes preclude guidance on any one measurement parameter. Guidance from the manufacturer should therefore be sought (see ISO8569:1989).

This standard does not explicitly provide guidance on the measurement of external airborne noise, but should be considered to establish the overall noise climate for the site (see ISO1996-1(1982)-2(1987)-3(1987)). It does however consider noise that arises due to vibration of the structure, referred to as groundborne noise. The preferred groundborne noise measurements are LAeq,T where time 'T' should be specified appropriate to the source/evaluation, and LAmax, with a stated time constant (slow/fast). The measurements should include a recording of LA90 (slow/fast) in the absence of the man made source (see ISO1996 parts 1,2 and 3 for definitions).

Analogue or Digital recording of time histories of vibration and where possible groundborne noise are preferred, to allow subsequent processing to derive any other parameter, and in particular their spectral characteristics.

7.4 Characteristics of Instrument Chain

The characteristics of the instrument chain, which may comprise transducers, amplifiers, signal conditioners, cables, data acquisition means and data storage medium should be fully understood. The frequency range should be appropriate to the source and evaluation required. The dynamic range should be sufficient to span the source magnitude, from ambient levels to the peak of an event. The phase characteristics should be known especially where time domain analysis is being undertaken.

The effects of ground loops, electromagnetic interference and triboelectric noise may be relevant to the measurement system and should be minimised.

The signal of interest should exceed noise levels, which may arise within the instrumentation itself or be due to ambient levels of vibration on the site. The noise inherent in instrumentation can be reduced by selection of appropriate instrumentation and appropriate site deployment. There is likely to be less control over ambient vibration levels on a site. If the ambient vibration level is significant, consideration should be given to how it may be reduced, and what implications the final level may have on analysing the signal of interest. Generally, the signal to noise ratio should not be less than 5dB, and ideally in excess of 10dB. This test can be applied to the time history, but applies across the relevant frequency range of the spectra.

7.5 Measurement Locations

The distribution and quantity of measurement locations depends upon the size of the site, footprint of the development, and proximity of the source(s).

They also depend upon the current conditions on site (e.g. presence of existing buildings, road surfaces etc.).

Measurements should be conducted on the nearest point(s) of the proposed building(s) to the source(s).

Additional measurements may be required at different sections along the site boundary to the source. Factors such as; variations in the structure, use of the proposed building, variation in ground conditions or source-related factors (e.g. position of points in the case of a railway) need to be considered and based upon expert evaluation.

The site can be large such that the foundations of the proposed building(s) may be at significantly different distances from the source. Sufficient measurements should therefore be taken to establish how vibration levels decay with distance from the source.

Factors such as the depth of basements, the formation level of foundations or piles, should be considered, as they may be placed closer to the source (e.g. to a railway in a tunnel or in a cutting). Or they may be placed in/on a geological stratum affected to a greater degree by the source.

Measurements may be taken in the base of boreholes at proposed formation level, or toe of piles. Soil removal down to formation level may affect subsequent vibration levels. Validity of such measurements to characterise inputs should be considered.

Foundations of existing building(s) may be used to take recordings upon, but it should be borne in mind that the free-field level when the building(s) are demolished will be altered owing to the absence of inertia and stiffening effect of the existing building(s).

Measurements in areas close to buildings may be in soil that was disturbed, during the original construction of the foundations. Existing building(s) can also affect local ground vibration levels, causing higher levels upstream due to wave reflection and lower levels downstream as building(s) take up and dissipate energy.

The orientation of horizontal measurements may be aligned to be radial or tangential to the source, or parallel to the building axis, and should be noted.

Noise measurements in existing building(s) may assist in the empirical evaluation of levels that may arise in the new or altered building. Measurement locations should consider that highest noise levels often arise in the basement/ground floor and reduce with storey height. Noise levels may be greater at off centre locations in a room due to normal modes. Unattended groundborne noise monitoring is not recommended.

7.6 Coupling of Sensors

Sensors should be coupled to the structure or ground in a manner that achieves faithful recording. Brackets where used should be sized to ensure that resonant modes with the coupled sensor are outside the frequency range of interest (see ISO4866:1990).

Ground motion should preferably be recorded in free-field conditions. For practical reasons it may be necessary to make use of existing ground slabs and foundations, or cast test foundations. Resonances and impedance mismatch with the surrounding soil should be considered when evaluating the faithfulness of this medium.

Sensors can be buried, although characteristics of the soil (e.g. particle size (coarse aggregate/boulders) in relation to sensor), quality of backfill and variability in workmanship (e.g. tamping) are factors that can affect the quality of 'free-field' measurements. These factors can also affect the repeatability of measurements if deployed repeatedly.

Where downhole measurements are necessary, the diameter of the borehole/size-of-plug and how it is cased should relate to the soil characteristics expected. The plug at the base of the borehole should ideally be free of the casing, and risk of water ingress considered.

7.7 Measurement Samples

The source(s) should be sampled sufficiently, noting nature of the source (Deterministic: periodic/non-periodic, Random: stationary/ergodic, non-stationary) and requirements for statistical accuracy [4,5], which are a matter of engineering judgement.

For a railway source, different train events should be recorded and identified according to train type (Local-Passenger, Inter-City or Freight), indicating if they are abnormally slow or fast. In the case of machinery, which item or items of plant are operating, effect of variations in load cycle or difference in work piece are sampling considerations.

When train events are infrequent, or item of plant is operational over restricted periods, this will affect the number of events that can reasonably be sampled at each measurement set up. The survey should either be extended to allow more representative results, or the risk of analysing less representative data borne in mind (see clause 13).

Where it is known that a track supports freight traffic, the survey should include samples of this type. Often freight traffic occurs during the night, at which time they have the greatest potential to disturb residents. Overnight recording is therefore recommended. Sufficient number of freight pass-bys should be sampled to increase the likelihood of including measurements of laden wagons.

Where continuous unattended monitoring is necessary, simultaneous monitoring should ideally be undertaken at two locations to allow correlated peaks to confirm ground motion is from the source of interest.

Samples of background vibration level, in absence of the vibration source(s) of interest should be recorded where this is possible or can be arranged.

Time histories should be obtained for all 3 axes in at least one of the measurement locations.

It is known that vibration is often strongest in one of the horizontal axes, typically radial to the source, but omission of the tangential axis should only be adopted if there is strong experience relevant to the source and site particulars to support this decision.

Measurements to establish variation of vibration levels along the site boundary, decay with distance across the site, or variation with depth, should only be based upon simultaneous measurements, to avoid comparison of different events, (e.g. different train pass-bys, or different load cycles of a machine). Simultaneous measurements may be limited to the principle axis in which vibration is of concern.

If the source is known to be constant or nearly so during the survey period, measurements at different locations need not necessarily be based upon simultaneous recording. However, it is recommended that at least one monitoring point be left as a reference whilst sampling is undertaken at different points using the spare sensor(s).

8 Soil Dynamic Properties

In some cases it may be necessary for the Specialist Engineer to influence the scope of a geotechnical survey, or specify/undertake surveys to establish soil dynamic properties.

NOTE 1 It should be borne in mind that some ground treatments (e.g. vibrocompaction, dynamic compaction, grouting) may alter the dynamic characteristics of a site.

NOTE 2 Extreme environmental conditions may also affect wave propagation characteristics in the ground, which should be considered (e.g. permafrost).

9 Structural Options

A structural option favoured by the Client or his Architect/Structural Engineer may have been conceived without regard to vibration control considerations. The Specialist Engineer may judge the given option to be unlikely to perform adequately in which case alternative options may form the starting point in the prediction and evaluation process (see Figure 1).

10 Alternative Vibration Control Options

Alternative vibration control options, directed at the source, propagation path and receiver should be considered. The range of options may provide viable cost effective alternatives to Base Isolation, and may in some cases be used as supplementary measures to a Base Isolation proposal. No significance should be attached to the order of the alternative options listed, and the list is not exhaustive.

At Source


isolate source

undertake maintenance on plant

alter design of plant

adopt different work practice.

procure new plant


smoother running surface

alter vehicle unsprung mass

alter line speed

resilient rail fixings, booted sleepers, track ballast mats or a floating track slab.


improve quality of running surface

remove potholes refit loose manhole covers

Propagation Path

increase propagation path, site building or sensitive parts far away from the source

Trenches in path between source and building

concrete wall built into the ground between source and receiver (Impedance Wall)

ground treatment to stiffen a layer of soil (Wave Impedance Block)

At Receiver

Rearrangement of plan (e.g. locate building far from source, switch location of car parks and landscaping into nearby areas)

Reconsider use of site (commercial may be less sensitive than residential).

Avoid floor resonance with dominant peaks in ground vibration spectrum (de-tune)

Solid ground bearing slabs may be preferential to suspended slabs (e.g. bungalows in place of two storey dwellings)

Unavoidable resonances targeted at frequencies of least Human sensitivity, usually requires high frequency (bandwidth may be broad, increasing risk of tuning with source)

Low natural frequency floors may avoid tuning with source (also a narrower bandwidth), but a risk of footfall induced vibration (see ISO 10137:1992).

A floor may constructed on isolators off a floor below (floating floor)

Isolate sensitive areas (box-within-a-box)

Isolate individual items of sensitive equipment

A dynamic system added to the primary system (Dynamic Vibration Absorber), to neutralise motion (at a specific frequency, where de-tuning cannot be achieved)

Select structural form for optimum damping (e.g. concrete in preference to steel)

Constrained layer floor damping treatments

Foundations taken into strata with less vibration, decouple from soil near surface

Sensitive equipment on ground rather than suspended floors, or equipment on foundation bearing deep in soil, and decoupled from building or soil near surface.

Deploy traditional structural materials in a way that achieves a rigid body mount frequency comparable to that obtained with Base Isolation

Increase path length of vibration source to increase damping (e.g. suspend floors from tops of columns rather than supported from columns directly off the ground)

Irregular construction patterns and discontinuities in the construction

Heavier forms of construction

Increase background noise levels to mask intrusive noise

Active vibration control using electromechanical or hydraulic actuators (unlikely to be a viable option in all but special circumstances)

11 Prediction and Evaluation of Building Response (Traditional Foundations)

The building response may be predicted using empirical methods, theoretical methods or a combination of the two (semi-empirical), (see ISO 10137:1992).

Empirical methods, which involve a collation of field measurements/observations, can be used to make predictions of groundborne noise and vibration in similar situations, and caution must be exercised in situations that are not comparable to the database.

Theoretical models vary in the extent of the problem modelled and differ in the nature of the solution, either solved in Closed Form, Finite Difference, Finite Element, Boundary Element or Statistical Energy Analysis techniques.

Traditional foundations for a building are often treated as rigid, but this is not always a reasonable assumption. The actual stiffness of foundations and their damping capacity may need to be based upon site specific analysis, either theoretical or experimental.

The complexity of a model and degree of confidence in the prediction depend upon the application of the model. For outline planning purposes a lower level of accuracy in the prediction may be acceptable, but a higher level is necessary for detail design. According to the risk analysis (see Clause 13), a parametric study may be needed to determine the effect of variations in the assumption of model parameters, which should begin with the most likely values.

The parameters to be predicted, for vibration and/or groundborne noise, must be compatible with the limit values that need to be achieved.

Factors listed under Clause 12 are particularly relevant to an analysis of base isolated structures and may also be relevant to buildings with traditional foundations.

11.1 Evaluation in terms of effects on the Structure

The predicted building response levels can be evaluated in terms of their effects on the structure, with reference to ISO4866:1990, and appropriate National Standards.

Vibration damage thresholds are not known precisely, but when treated in a probabilistic way are often in excess of human tolerance levels. Complaints of alleged damage may however arise at relatively low levels.

11.2 Evaluation in terms of Human Response

The predicted building response levels can be evaluated in terms of Human response by reference to ISO 2631 part 1:1997,part 2:1989, and taking note of limit values in appropriate National Standards, or in planning conditions.

Vibration that is just perceptible can evoke adverse comment in an environment where high quality is expected, although vibration of higher levels may be tolerated. Groundborne noise impact is often related to how much the noise from the source exceeds background levels. A degree of intrusion may be commercially acceptable in some cases, noting that control measures have both technical limitations and financial consequences. Hospital operating theatres require vibration to be imperceptible when they are operational. Consideration can be given to isolating individual areas (e.g. box within a box type construction) as an alternative to Base Isolation of the entire building.

It should be noted that halving of vibration levels and/or a reduction of 5dBA in groundborne noise are considered to be changes that are subjectively perceived as significant and should form the basis of judging whether Base Isolation is of value.

11.3 Evaluation in terms of Sensitive Equipment or Processes

Equipment and processes may be sensitive to vibration, but need not be assumed to be sensitive unless guided so by the manufacturer. Proper siting or isolation of individual equipment/processes may prove more appropriate than Base Isolation of the building. The Specialist Engineer should consult with equipment manufacturers to identify the nature of vibration limits, exact consequences of excursions above the limit, and any uncertainties or factors of safety, as this can affect overall risk analysis (see Clause 13).

11.4 Identification of Required Attenuation

Comparison of predicted building response using traditional foundations with limit values identifies the amount of attenuation necessary. Sensitivity analysis of variables may be used to identify confidence limits associated with the required attenuation.

Any improvements necessary may well be achieved using basic vibration control techniques, which should be studied alongside the option of Base Isolation (Clause 10).

12 Factors to consider in the modelling of Base Isolated Buildings

Whilst there is field data to support empirical prediction for buildings with traditional foundations, there is currently insufficient field data to support empirical prediction for Base Isolated buildings. Theoretical models are therefore necessary, either as stand alone, or which make some use of empirical models (semi-empirical modelling).

NOTE 1 This standard recommends measurements to validate performance of Base Isolated buildings, which should in the future provide data for empirical models.

NOTE 2 A vibration survey of existing Base Isolated building(s) comparable to a new situation is encouraged as an aid to indicate expected performance/behaviour.

An important aspect in the theoretical modelling of buildings is the characterisation of the inputs to a model. Field measurements may pertain to free-field conditions, but there will arise changes due to interaction with the unloaded foundations. The superstructure will also affect foundation response and this will depend upon the nature of its coupling, either traditionally coupled or isolator coupled to the foundation (see Figure_2).

The nature of the source should be considered, to establish if steady state, random or transient response analysis is appropriate. The modelling of inputs in a spatially extended model should consider the effect of correlation between the inputs, as this can have significant influence on the results [7]. The analysis may need to consider the effect of starting transients along with the steady state solution. A finite time is needed to induce full resonant response, which may be relevant, particularly in systems with low natural frequency and low damping. Analysis may need to consider horizontal and vertical components of the source, either in some combination or separately. Variations between static and dynamic properties of materials should be noted

The dead loads assumed for a structure should be unfactored. The imposed loads should also be unfactored and realistic appropriate to use, including any finishes and services.

Single-degree-of-freedom (SDOF) models which treat the building as a rigid body on isolators are inappropriate. Field measurements show that the resonance characteristics of the supported structure can dominate the results. Multi-degree-of-freedom (MDOF) models are therefore necessary. Such models may need to include dynamic characteristics of the foundation or structure below the isolators. Complexity of the analysis depends upon the structure and the confidence level required (see Clause 13).

A basic model of a framed building can comprise a single column, and where appropriate the model should lump a portion of the storey mass per storey height onto the column, and recognise that columns may become more slender with height. The model can consider both fixed and resilient supports, and where appropriate model the finite resilience of the foundations bearing on the soil. For tall framed buildings it is possible that the fixed-free axial mode of a column, modelled with lumped masses per storey height, may yield a frequency that is close to the intended rigid body mount frequency of the isolation system, thereby lowering the frequency for this mode [8]. The applicability of this basic model may be restricted to low frequencies and will only treat a single input. An extended model of a building will be necessary to treat the dynamic interaction of floors, adjoining columns/walls and accommodate multiple input, either in two or in some cases three-dimensions. A small squat building of brick and block walls should not be assumed to behave as a rigid body, unless supported by theory.

A model should be sufficiently detailed to identify natural modes of the structure within the frequency range of interest. These should be compared with the spectral characteristics of the source, to identify the risk of resonances.

There is a possibility that floors may act as Dynamic Vibration Absorbers (vibration neutraliser) and a high level of motion may arise in parts of the building intended to be best protected (the degree of correlation of inputs can effect the outcome [7]).

The magnitude of resonant responses that could arise should be investigated. The nature of the damping model assumed can be critical. Damping may be modelled as viscous, although the increase in damping with frequency does not relate well to field experience of structures. It may be appropriate to use alphad and betad Rayleigh damping constants to achieve near constant damping in a certain frequency range, or use a hysteretic damping model. Where damping is localised (e.g. rubber bearings) it is mobilised in a mode dependent way, which requires careful consideration especially when the localised damping is high. Damping values taken from the literature should be at strain conditions applicable to the design circumstances.

Extreme environmental affects on material properties may need to be considered (e.g. increase in stiffness of rubber bearings at extreme low temperature).

Groundborne noise may be predicted using empirical relationships with the magnitude of vibration expected in the structure. This may also be calculated theoretically, taking into account the sound pressure developed by a vibrating surface [3]. The analysis may be repeated at relevant frequencies, and will involve factors such as the vibration velocity of the radiating surface, radiating area, and room absorption. Acoustic coupling may need to be considered in very thin cavities between isolated and un-isolated areas.

Short circuits can compromise performance of a Base Isolation system, and therefore their avoidance in design and construction is critical (see Clauses 15,16,17 and 19)

12.1 Alternative Isolator Options and their Deployment

Isolators for use in Base Isolation applications are commonly made from natural/synthetic rubber (see BS6177:1982), or formed using steel coil springs. The possibility of other types cannot be excluded.

Rubber bearings are typically formed as discrete blocks. Rubber bearings should be deployed in a manner that will ensure that they are not subjected to tensile loading. Whilst resilient mats can be used, they generally produce high stiffness outside the range normally considered suitable for Base Isolation. Steel coil springs can provide the lowest rigid body mount frequency and can also be supplied with pre-stressing assemblies, that can allow easy replacement or fine-tuning. When evaluating different isolators, it should be noted that their dynamic performance depends upon dynamic stiffness and damping, and not on how specifically these characteristics are achieved.

The location chosen for the isolators depends upon many factors such as areas to be isolated, the dynamic characteristics of the isolated building and practicalities of providing access for inspection and replacement where necessary.

Isolators are often located in the base of a building, between the foundations that bear on the soil and the isolated building. Such isolator locations usually require greater provisions for inspection/replacement, and retaining walls to ensure that the part of the isolated building, if below ground level, does not short circuit with the ground.

Isolators can however be installed at any level, typically on the tops of enlarged columns. The resonant characteristics of the structure below the isolators may be more relevant in this case. At such locations fire protection is critical and fail safe provisions are more difficult to accommodate. The isolators may in some cases have to be deployed at different levels where affect of differential settlement may be more critical.

12.2 Wind Induced Motion Control Devices

Base Isolated buildings require an assessment of wind induced motion (see ENV 1991-2-4), and may for some structures require devices for their control. The devices should be deployed at optimum locations for their effectiveness (e.g. acting on stiff parts of a building, such as lift shafts and staircase cores).

Wind loading may be analysed for some buildings using a rigid body model in six-degrees-of-freedom. A thin cavity over a large area between isolated and un-isolated areas may act as an air spring and contribute to the overall stiffness.

Wind induced motion control devices may incorporate sliding joints to allow the building on isolators to freely deflect during construction and permit long term creep. Dowel pins may provide transverse restraint, and should be resiliently sleeved in either the isolated or un-isolated section, and allow room for elastic deflection and creep.

Resilient bearings for wind induced motion control devices may utilise non-linear bearings. Soft at low strains to avoid compromise in dynamic performance under normal conditions, but stiffer at higher strains associated with extreme wind loading.

Ultimate design loads for wind induced motion control devices shall be compatible with the return period used in the structural design of the building. Wind induced motion control devices should address fatigue from dynamic effect of wind over the service life.

Permissible motion of the structure with regard to structural integrity, human perception, or sensitive equipment/processes, should consider the probability distribution of wind speeds. Providing tighter limits for the majority of the lifetime of exposure, and broader limits where appropriate for the extreme wind loading situations.

12.3 Potential benefits/disbenefits of Base Isolation

The dynamic performance of Base Isolated buildings have not been broadly investigated or reported to allow a clear judgement to be made about the value of this technique as a means of vibration control, except by a few researchers [6,7,8]. The limited research shows that benefits of Base Isolation are not in line with common predictions made using SDOF models, which are fundamentally inappropriate. Base Isolation performance should be judged against what might have arisen had traditional foundations been used. This is not practical to determine on site, in all but special circumstances (see Clause 21), but may be examined theoretically [7].

Free-field measurements on a site are subject to change resulting from interaction with unloaded foundations. The response of this foundation is further modified by the presence of the superstructure and depends upon the nature of its coupling, which may be traditionally or isolator coupled (see Figure 2) [8].

Base Isolated buildings may loose some of the advantages enjoyed by an un-isolated building. Traditionally coupled foundations allow motion of some modes of the structure to include a participating soil mass which increases inertia for these modes, and would also allow energy to radiate back into the soil. A Base Isolated building is decoupled from the soil by the isolators. The lower support stiffness of the isolators makes the building more susceptible to wind induced motion. Ambient levels of vibration in a Base Isolated building, due to building services plant, are sometimes found to be higher than would arise in a traditionally founded structure [7,8].

Structural resonances can negate the potential performance of a Base Isolation system. Buildings will have many modes of vibration, some involve the foundations and others are ‘localised’, independent of foundation coupling. A large structure is likely to have a higher modal density than a small discrete structure, which in the theoretical limit approaches a simple SDOF. A high modal density of relevant modes in the dominant frequency range of the ground vibration spectrum could compromise the potential effectiveness of Base Isolation [8].

The calculation of benefit/disbenefit depends upon the datum situation referred to, and is therefore a relative rather than absolute measure. Base Isolation may prove to be effective at reducing groundborne noise, and would tend to be effective to a lesser extent in attenuating perceptible vibration. Attenuation would tend to increase for a lower rigid body mount frequency and damping. Performance is direction dependent, tending to be more effective in the horizontal direction [8].

Base Isolation should not however be viewed as a conservative solution, in that it will always improve the situation, since research shows that Base Isolation in certain circumstances may be detrimental to the dynamic response of the building [8]. It is therefore necessary to weigh up the advantages and disadvantages and thereby justify a decision to Base Isolate, as there are a variety of alternative control measures that may prove more effective and/or be more economical (see clause 10).

13 Risk Analysis

A decision to Base Isolate should be based upon appropriate models using input parameters with appropriate confidence bearing in mind the value and sensitivity of the building, level of assurance required in achieving limit values, and therefore likely cost benefit of the analysis.

Risk analysis is necessary to investigate effect of confidence limits on input parameters and weigh up benefits of detailed models. Input parameters to a model may need to be assumed, rather than determined by measurement. There may be limits to the extent of modelling possible, based upon hardware resources, or ultimately the state of knowledge. Detailed models and improved confidence in input parameters have cost and programme implications, which should be considered against cost implications of a failure to meet limit values, dependent upon the use of a building.

For example, failure to achieve limit values for sensitive equipment or processes may render the building unfit for purpose, and although alternative less demanding uses may be options for the site, there are likely to be serious programme and financial implications given that a serviceable site would still be required. The level of risk would be different in a building intended for residential occupation, as it may make the difference between achieving a high quality or just an ordinary habitable environment, affecting realisable rents or sale value.

Risk analysis should therefore consider the importance of achieving limit values, recognising consequential losses, possible remedial measures or alternative uses that may be available, along with technical or practical limitations in the modelling.

The Client should recognise an element of risk associated with dynamic performance of Base Isolated buildings, and the Specialist Engineer and/or Isolator Supplier should convey the confidence level associated with any benefits/disbenefits expected.

14 Validation by Test Structures

Prediction of building response can be complex and there are many unknowns/uncertainties associated with the dynamic analysis. There is a greater problem associated with the prediction process for Base Isolated buildings, because there are fewer field measurements to guide analysis. The Specialist Engineer should recognise limits of current knowledge, and advocate appropriate tests on model structures if it is necessary to improve confidence in predictions (see Clause 13).

15 Structural Design Implications

The Client's Structural Engineer should be responsible to note and implement points that have been identified as relevant to this discipline by the Specialist Engineer, who may be required to comment upon and/or approve such details.

There are structural design implications of incorporating Base Isolation into a building. Isolators alter support stiffness in vertical, horizontal and rotational degrees of freedom, which can have implications on the stability of the structure or a structural element.

Wind loading shall be investigated, and the demarcation of responsibility between the Structural Engineer and the Specialist Engineer clearly defined (see Clause 12.2).

Affects of differential settlement should be investigated, which may arise during construction, in relation to sequence of construction and storage of materials. It can also arise in service due to variations in loading pattern, isolator stiffness and creep.

Isolator failure scenarios should be considered and fail safe systems may need to be provided to limit differential movement, to avoid damage to the structure and/or damage to curtain walling/fabric of the building. A fail safe snubber system should not come into play during the service life of a building as a result of normal settlement by creep.

The structural concept should simplify implementation of the isolators, minimise risk of short circuits, and where they may arise allow for inspection and removal. For example, short circuits can arise from permanent or temporary shuttering, which may be difficult to remove in confined areas, or gaps can become filled with debris, in a way that they can neither be detected or removed.

It may be appropriate to provide in the design a means of replacing any faulty isolators.

Structural detailing should make provision for inspection of isolators during the construction period, and for the service life of the building. Such provisions can be based upon the use of a fibre optic endoscope.

Fire rating of isolators should be compatible with that used for other load bearing structural elements. This should be achieved by either strategic positioning of the isolators, or via external fire protection. If external fire protection is used, it may need to be validated by fire testing, either by the Isolator Supplier or an Independent Testing Agency. The fire protection if used shall not rigidly bridge the isolated and un-isolated parts of the structure. Its effect on impeding inspection of the bearings should be considered. The possibility of damage to the fire protection during construction, either accidental or through vandalism should be considered, as should means of repair.

There should be adequate drainage to avoid the isolators being permanently submerged. Steel coil isolators may contain viscous fluids, or comprise separate viscous dampers, in which case flooding should be prevented to avoid contamination of the damping fluid.

The bearing surfaces (top and bottom) for isolators should be smooth and level, typically to ±1mm in 1m for rubber bearings and ±2mm in 1m for steel coil springs. This ensures near parallel bearing surfaces are achieved. The top bearing surface may be provided by precast concrete or steel beams which have a pre-camber that is taken up as dead loads are applied, which should be considered. The bearing surface may be achieved using self-levelling cement based grouts. If epoxy resins are used, their susceptibility to flow under the temperatures induced in a fire should be considered. If a concrete (or epoxy) plinth is used to provide the bearing surface, adequate edge clearance or steel reinforcement should be provided to avoid spalling.

There should be adequate space around rubber bearings to allow free bulging of its faces. Some steel coil isolators require space to install hydraulic jacks and permit leverage of spanners to release pre-stress. The building weight itself can be used to release pre-stress, but access may still be required to post-stress these isolators for replacement or adjustment purposes (e.g. introduction or removal of nested coils).

The isolators and any associated wind control devices should have a service life compatible with that of the main building (typically 50 years). This requires consideration of long term affects in rubber bearings, and corrosion protection for steel coil isolators. Access for replacement of isolators may be necessary if a longer service life for the building is anticipated, or alterations/extensions are envisaged.

It may be necessary to review the provisions and requirements for lightning protection, as certain isolators may impede the electrical grounding.

NOTE Earthquake loading may need to be considered at certain sites, even if the Base Isolation system was not designed to cater for this source (see ISO3010:1988).

16 Architectural Design Implications

The Client's Architect should be responsible to note and implement points that have been identified as relevant to this discipline by the Specialist Engineer, who may be required to comment upon and/or approve such details.

Where building heights are limited due to planning considerations, allowance for isolator height, but more significantly the space provisions of double foundations (i.e. below and immediately above isolators) may be critical.

Provision should be made to allow for the changing level of an isolated building as it is constructed to that of any adjoining un-isolated areas. This can either be accommodated in the setting out of the structure, or be hidden by false floors and false ceilings.

Detailing of architectural features, such as curtain walling joints may need to anticipate static dead load deflection of an isolated building where it adjoins an un-isolated building, to ensure that the details do not promote the visual detection of any misalignment. Alternatively, setting out of such components can anticipate movements, or be installed at a time when the majority of dead load deflection has taken place.

Adequate gaps should be provided to adjoining un-isolated buildings to allow for thermal movements, settlement, creep, and wind induced sway.

Gaps should where necessary be sealed with flexible materials, ensuring that the material remains flexible with age, so as not to significantly compromise Base Isolation performance. The flexible material may need to meet fire rating requirements.

Architectural details (e.g. cladding) or landscaping should be specified to avoid the creation of short circuits with the ground or un-isolated buildings. Any feature that might lead to a short circuit should be detailed in a way to permit inspection and access provisions for remedy, according to the level of risk.

If rubber bearings are used, risk of vermin attack should be considered, taking advice from the manufacturers, and implementing pest control or protection where appropriate.

Sub-basement voids formed to permit inspection of isolators should be made with confined access safety provisions, such as ventilation, lighting and emergency escape.

17 Services Design Implications

The Client's Services Design Engineer should be responsible to note and implement points that have been identified as relevant to this discipline by the Specialist Engineer, who may be required to comment upon and/or approve such details.

Services that enter or leave an isolated building should ideally be connected when the majority of static dead load deflection has arisen. There should be adequate flexibility, either of the service element or with flexible couplings to accommodate static displacement due to elastic deflection, creep, and in some cases under isolator failure situations. The flexibility should be sufficient to avoid significantly affecting dynamic performance of the Base Isolation system.

Lifts may require an energy absorbing safety buffer in the lift pit (to arrest free fall) and lift guides may traverse part of an isolated and un-isolated structure. These should be detailed to avoid short circuits between the un-isolated and isolated parts of the building and accommodate any movements expected.

Plant rooms where possible should be sited in un-isolated areas, as a Base Isolated building can respond more adversely than a traditionally founded building.

Liquid storage tanks (e.g. fuel) should be provided with bund walls to meet the volume capacity of the tank, to avoid spillage into areas which might affect isolators.

18 Outline Specification for the Supply and Testing of Isolators

The Specialist Engineer should prepare a specification for the isolators, outlining the major engineering and dimensional requirements. It should have regard to design load and factor of safety for ultimate load, define required dynamic stiffness at a defined working load and specify the required damping properties at a defined strain level. The expected service life should be specified, and the Isolator Supplier should note the environment within which the isolators will operate.

The specification may extend beyond overall engineering properties of the isolator, to matters of detail concerning their design, fabrication and testing. Reference where possible should be made to applicable National and International Standards.

EXAMPLE 1 Specification for laminated natural rubber bearings may include minimum thickness for steel plates, define minimum cover for corrosion protection, and to avoid heat conduction into the bearing by steel plates in a fire. It may deal with the fabrication process, placing limits on any out of parallelism in the steel plates during vulcanisation.

EXAMPLE 2 Specification for steel coil isolators may include requirements for resilient noise stops to avoid conduction of audio frequencies, fluids to damp axial modes of the coil itself, and corrosion protection.

The Specialist Engineer should consider implications due to the Isolator Supplier offering standard products that do not meet the initial specification.

The Isolator Supplier should define height of the isolators, as either free height or loaded/pre-stressed height under specified load conditions. Any reference to vertical/horizontal stiffness should be qualified as dynamic or the static tangent stiffness at a defined working load. Some isolators (eg. rubber) can exhibit a large dynamic/static stiffness ratio, which should be quoted when the static tangent stiffness is provided.

The specification should define a permissible tolerance level on each parameter to allow for fabrication tolerances both on dimensional and engineering properties.

Requirements for the identification and testing of isolators should be agreed between the Specialist Engineer and the Isolator Supplier.

Identification on each isolator should include the Isolator Supplier’s name, address, unique serial number, and date of manufacture. Identification should also make clear which way up the unit should be installed when relevant. Identification should be such that it remains clear and legible during installation and for the service life.

Load testing may be required and in some cases essential. For example rubber bearings may require load testing in excess of design working load to precondition the bearings. In other cases sample testing of isolators may prove acceptable. In exceptional cases, for example steel coil isolators, reference to a library of test results for a given product may be acceptable. Where random sampling methods are used for testing, identification of substandard isolators may warrant an improved sampling rate on the product, if rejection rate or nature of rejection is considered abnormal.

Where load testing is undertaken, the test should exceed design load by an agreed margin, with particular emphasis on identifying instability. Structural freedom of the bearing surface may affect stability of the isolator. The test set-up should therefore replicate conditions to be met on site. Testing should include vertical and horizontal loads which may need to be combined (e.g. for wind induced motion control bearings). Testing should provide load deflection curves and details on any load cycles for records.

The effect of long term loading on an isolator should be considered to ensure that the major engineering properties are not subject to significant change. A long term load test should also be undertaken to provide the creep rate for the isolator when relevant.

The dynamic stiffness or dynamic/static stiffness ratio should be based upon dynamic testing of isolators, using a test rig for which the dynamic characteristics are known and taken into account where relevant. It may not be practical to undertake dynamics tests on isolators with large load capacity, and therefore indirect determination of dynamic stiffness may need to be considered

If isolators are made of component parts (e.g. steel coil isolators), these should be delivered to site pre-assembled from the factory. Where there are exceptional reasons for site assembly, the Isolator Supplier should provide appropriate instructions and inspect the works. Site assembly of wind induced motion control devices are more likely and may involve the Specialist Engineer in the supervision process.

The Isolator Supplier should make clear any precautions necessary for the correct handling and installation of the isolators on site, provisions for their protection, and maintenance. The terms and conditions of any warranty on the product and results of any test data should be made available to the Specialist Engineer.

19 Construction implications for Base Isolated Building

The Client's Main Contractor should be experienced with, or take independent advice, on the implementation of Base Isolation. The Main Contractor may agree method statements with the Specialist Engineer, although should take final responsibility for safe construction to the specification and drawings. The Main Contractor should be responsible for his sub-contractors, which may include the Isolator Supplier.

Implementation of a Base Isolation system will have programme implications, which differ, depending upon whether the isolators are installed at the time the superstructure is to be built upon them, or whether they are introduced at a later date.

NOTE Some steel coil isolators can be pre-stressed, allowing them to be released in-situ. When pre-stress is in excess of building weight, the stored energy may be used to raise a building, allowing release of temporary supports. This would allow the insertion loss of a Base Isolation system to be determined (see Clause 21).

Provisions should be made for safe storage of isolators on site. For example, rubber bearings may need protection from direct sunlight to avoid ozone attack, or corrosion protection of steel coil springs may be vulnerable to mechanical damage. Protection should be provided in accordance with the Isolator Supplier's instructions, with due regard to damage that may arise by accident or vandalism. Isolators should be delivered to site at the time they are required for installation, noting that lead times may be long.

Isolators will require a smooth level bearing surface (see Clause 15). This may be formed using self levelling grout, observing the manufacturer’s installation requirements, with particular regard to environmental limits on their placement.

Isolators should be handled on site in accordance with the Isolator Supplier's instructions. Where isolators have to be placed a certain way up, this should be observed using the markings displayed on the isolator. The Contractor should check that if unique identification markings are used to distinguish isolators, they are deployed at the specific locations indicated by the Specialist Engineer. It should be noted that isolators may appear identical, even though the material properties or internal make up can vary.

Isolators may differ in height, and this may also be due to fabrication tolerance. They may require gaskets placed on top, or below the unit. Gaskets shall be to the approval of the Specialist Engineer, and have regard to factors such as durability. If gaskets are used, the thickness should be selected to minimise the overall quantity at each location.

The isolators and wind induced motion control device bearings may need to be installed using adhesives, which should be to the approval of the Isolator Supplier, and used on site in accordance with the manufacturer's instructions.

In some cases the isolators or wind induced motion control devices may require mechanical fixings. Stainless steel fixings should not be used to secure mild steel elements without regard to the risk of bimetallic corrosion.

Welding should not be permitted close to isolators, or bearings used in wind induced motion control devices, unless expressly permitted by the Isolator Supplier.

Wind induced motion control devices may utilise sliding bearings, to permit building settlement due to dead load deflection and creep, whilst the device remains in contact to provide horizontal restraint. Contact areas of such sliding bearings may need to be kept clean and free of debris, during their installation and service life.

It may be necessary to monitor building settlement during construction. The resolution of the monitoring system would depend upon expected dead load deflection and tolerance limits for differential settlement. The Contractor should agree the monitoring requirements with the Specialist Engineer and/or the Client's Structural Engineer.

Limits on differential settlement, to avoid stressing the structure unduly, may require construction to proceed evenly, avoiding storey heights in one area to be developed far ahead of another or impose limits on storage of materials on the structure.

Safety of temporary structures shall be considered, such as scaffolding that may rely for their support on the isolated structure, which will settle as construction progresses.

During construction it may be necessary to inspect isolators for movement and damage to ensure that they are in acceptable condition before any further loading or construction makes it more impractical to remedy the situation. For example, isolators may be dislodged or damaged by deployment of precast concrete or steel beams, installation/fabrication of shuttering or contaminated by grout leaks in an in-situ pour.

Construction should be undertaken in a way that minimises risk of short circuits arising from the use of temporary / permanent shuttering, debris remaining within any gaps, or grout leaks. Grout or debris should not be allowed to enter gaps between any fail safe devices. Care should be taken to avoid ingress of grout via isolator inspection holes.

Gaps should be checked frequently and cleared regularly, as construction progresses. Where necessary, gaps should be flexibly sealed to the approval of the Specialist Engineer as soon as possible to prevent further debris collecting in such gaps during construction. Gaps should only be sealed after an inspection has been undertaken to confirm that the areas behind the area to be sealed are clear of short circuits.

The avoidance and elimination of short circuits is the most critical aspect in the ultimate performance of a Base Isolation system. The Contractor should therefore appoint specific member(s) of his staff to continually check this during construction, and seal off any sensitive areas to further access by other tradesman.

Compressed air jets used for cleaning, or removal of polystyrene shuttering, may pose a risk of sand sized particles being blown onto steel coil isolators, and damage corrosion protection by impacting particles, or contaminate viscous fluid where used.

Any changes in the design or approved construction method shall not be undertaken without first consulting with the Specialist Engineer.

Any damage to the isolators should be reported to the Specialist Engineer and Isolator Supplier for advice.

If wind induced motion control devices are used, it may be necessary to make final adjustments after the dead load deflection of the building has taken place.

When construction is complete, the height of isolators where accessible should be measured. Dead load deflection of isolators can then be determined, as a check that isolators are under design load conditions, and as a record to identify future changes.

Gaps between fail safe devices should be recorded, and in some cases a final level survey of the structure may be required to act as a record of any future movements.

20 Cost Implications of Base Isolation

The Client's Quantity Surveyor should provide the Client with cost implications of Base Isolation, which are not just cost of isolators (often a small fraction of overall costs) but additional materials and design/construction cost implications of Base Isolation.

21 Validation Measurements for Quality Assurance

As part of quality assurance procedures the Specialist Engineer or an Independent Testing Agency should undertake measurements to ascertain the dynamic performance characteristics of the Base Isolated building when complete.

The survey should establish if limit values for vibration and groundborne noise were achieved in respect of the intended use, reporting any shortfalls in performance and proposals for remedial measures where appropriate.

The survey should where possible establish the insertion loss of the Base Isolation system (see Clause 12.3).

Figure 2 shows transmissibility measurements across the isolators [B] and to other parts of the structure [C], (e.g. to tops of columns), which can be used to highlight the effect of structural resonances for comparison with MDOF models. The merits of using total and/or direct transmissibilities in these comparisons should be considered [6].

Measurements across the isolators do not strictly provide the insertion loss of a Base Isolation system, since this is defined as the benefit or disbenefit of Base Isolating a building, compared with the performance of an un-isolated building were that to have been built at the site.

The transmissibility [A] of Figure 2 between a free-field datum and the un-coupled foundation indicates the effect of dynamic soil structure interaction.

The transmissibility between a free-field datum and a structure includes some system characteristics (e.g. soil, foundations and intervening distance) which are of no direct interest. The difference in transmissibility to un-isolated and isolated states of the building provides Base Isolation insertion loss [8]. The actual insertion loss is therefore determined by the difference [Etc-Eic] to any identical points in the building, or to a point above the isolators this can be expressed as [B+Dic-Dtc].

The difference in response of the structure below the isolators [Dic-Dtc] arises because the interaction of building mass and stiffness is dependent upon its coupling with the foundation, either traditionally coupled or isolator coupled, and this effect is frequency dependent. Measurements across the isolator [B] may therefore be used as a basic estimate of insertion loss, and an error arises due to the size of the difference in the terms [Dic-Dtc]. The difference [Ntc-Nic] of Figure 2 refers to the insertion loss on groundborne noise.

The actual insertion loss could be determined by measurements in the unusual circumstances depicted by Figure 2, of the identical structure in isolated and un-isolated states. For example, it can be estimated from measurements of a building constructed on temporary 'rigid' blocks, which may be subsequently replaced by isolators at a later date in construction [8]. This may be a realistic option when pre-stressed steel coil isolators are used to raise a building, (see note to Clause 19) or if jacks can be used economically.

NOTE 'Rigid' temporary supports will in fact have a finite resilience, which should be considered in the evaluation of the un-isolated state.

There may be options to estimate insertion loss by comparing measurements to part of the development consisting of similar but un-isolated structures. There may also be buildings on neighbouring sites, which are comparable, providing a means to broadly estimate the effectiveness of a Base Isolation system [8]. In such cases, simultaneous measurements of isolated and un-isolated structures should be compared. Expert judgement is required to evaluate effects of other factors that may influence the results.

Measurements can be taken using the source for which the Base Isolation system was intended to cater for.

Insertion loss may be estimated using an impact or shaker source, applied on the ground or on the structure below the isolators. Such a point source is likely to display three dimensional decay characteristics in a way that will overestimate insertion loss, compared to the situation that may arise with the real source. This is because the real source may excite an entire area of the site, leading to numerous inputs, whereas a point source will be more localised.

Impact or shaker testing can be used to determine modal characteristics of the Base Isolated building, including estimates of modal damping.

Where impact testing is undertaken, a measurement of force should be included. The period of the impulse should be adequately short in relation to the lowest period of the structure. This is to ensure that a short sharp disturbance is applied, so that the structure's transient response is then not complicated by the removal of the disturbance, and to ensure excitation of the structure over an adequate frequency range. A sufficient number of impacts should be sampled, usually five.

Shaker tests should also include the force, either by measurement or theory. The shaker may be set to operate at different frequencies, in increments over the frequency range of interest. The shaker test should continue at each frequency for sufficient time, before recording data for analysis, to ensure that a steady-state condition is reached. The increments of frequency used in the shaker test should be small in relation to the bandwidth of the resonant modes that are of interest.

Modal characteristics can be inferred from ambient response testing, although insertion loss should not be estimated under these circumstances [8].

The source signal at all measurement locations used for transmissibility calculations should ideally be an order of magnitude above ambient vibration levels. This may necessitate measurement to be taken when all significant services plant is shutdown, and ideally late at night when there are no occupants (see Clause 7.4).

22 Building Notices and Documentation

Notice(s) should be displayed within appropriate location(s) in the building notifying occupants responsible that the building is Base Isolated. The notice(s) should indicate that no change of use or alterations to the building should be undertaken without regard to professional advice. They should indicate clearly if there are maintenance requirements and refer the occupant to relevant documentation. Where possible simple diagrams should accompany the notice(s), indicating the concept of Base Isolation, and the need to avoid creating any short circuits during the service life of the building.

Documentation should be prepared by the Specialist Engineer and /or the Client's Structural Engineer, indicating the restrictions that Base Isolation will have on any changes of use (loading), and upon any alterations to the structure. The documentation should include drawings showing the deployment of all components of the Base Isolation system, together with details on the specification, and addresses of relevant suppliers to enable any future owner to resource such products should they ever become necessary. If the design incorporated a specific means of inspection, maintenance plan or procedures for replacement, then details on such procedures should be provided. Documentation should include contact details for the original design team and Main Contractor. Documents should be updated when any changes are made.

The documentation shall include any datum measurements taken upon completion of construction, for subsequent comparisons, and be updated whenever future measurements are taken. Where possible, the documentation should also include relevant photographic records taken during construction and at completion.

If the Isolator Supplier provides a warranty, this shall be attached to the documents, together with the terms and conditions of the warranty. Such documents should accompany the title documents for the site. The insurance company should be notified where the building incorporates Base Isolation.

23 Maintenance and Condition Monitoring

A maintenance plan shall be developed by the Specialist Engineer, with due regard to any recommendations offered by the Isolator Supplier. The Specialist Engineer and/or the Isolator Supplier should define the intervals for any maintenance. This may be short in the early life of the building, increasing to longer intervals in the latter part of the service life.

Maintenance may be required, and will usually be associated with adjustments to any wind induced motion control devices, to ensure that they remain effective, and are not unduly strained as a result of normal elastic and creep deflection of the building. Condition monitoring may include requirements for visual examination of isolators, measurement of isolator heights/ or gaps between fail safe devices, a level survey of the building, or vibration/noise measurements. A maintenance log should be maintained, to update the documentation (see clause 22).

Annexe A (Informative) Bibliography

[1] Wallar R. A. Building on Springs, Pergamon Press, 1969

[2] Crockett J.H.A. Early attempts, research and modern techniques for insulating buildings. Proc. Conf. on Natural Rubber for Earthquake protection of buildings 1982

[3] Grootenhuis P Structural Elastomeric bearings and seatings. Chapter 9, Polymers and polymer composites in Construction, Thomas Telford 1989

[4] Bendat J.S. and Piersol A.G. Engineering applications of correlation and spectral analysis John Wiley and Sons, 1993

[5] Newland D.E. An introduction to random vibrations, spectral and wavelet analysis, third edition, Longman, 1993

[6] Newland D.E. and Hunt HEM Isolation of buildings from ground vibration: a review of recent progress, Proc Instn Mech Engrs Vol 205, 1991

[7] Cryer D.P. Modelling of vibration in buildings with application to base isolation, PhD Thesis, Cambridge University, 1994

[8] Sharif A.K. Dynamic Performance Investigation of Base Isolated Structures, Research Reports Vol I, Vol II, Civil Engineering Dynamics, UK, 1999