Differential movement between the brick cladding and timber frame of the TF2000 building
Grantham, R.1, Enjily V.2
ABSTRACT
Revision of the building fire resistance requirements in England and Wales in 1991allowed construction for timber
framed buildings beyond four storeys height. This has opened up new markets for timber frame construction but has also
raised a number of technical issues which require solutions for safe and economic use of medium-rise timber frame
buildings. The TF2000 project aims to provide solutions to some technical barriers for 6-storey buildings through
research conducted on a full scale building.
One of the core programmes for this project addressed the problems of differential movement between the brickwork
cladding and timber frame with increased storey height. To determine the actual movements, instrumentation was
installed at the start of construction to monitor loads and deflections as well as hydrothermal changes. Cross-grain
shrinkage, compression and bedding-in movements were separated from observations on the building with the aid of
sample testing. The results of this work provide guidance for the design of timber frame medium-rise buildings clad with
masonry.
INTRODUCTION
In 1991 the Building Regulations for England and Wales were modified, allowing for the first time, the construction of
medium-rise (4 to 8 storeys) timber frame buildings. The fire safety requirements in this revision extended the market
base for manufacturers of timber frame buildings but it was not until 1995 that the potential market was identified(6) as
part of the Timber Frame 2000(5) project (TF2000). This indicated that the percentage of multi-storey buildings suitable
for construction using timber frame in the UK were:
6 storeys
5%
5 storeys
15%
7&8 storeys
2%
3 storeys
45%
4 storeys
33%
Timber frame construction for low rise housing in Scotland already has a dominant market share of almost 50% with a
30% growth in new constructions identified for 2000-2001 by the Timber Frame Industry Association (TFIA), formerly
the Scottish Consortium of Timber Frame Industries (SCOTFI). Although the market share for timber frame in England,
Wales and Northern Ireland is presently, much lower than that of Scotland, a similar growth in this market has been
predicted. This has mainly been accredited to the economies of construction due to the level of prefabrication and speed
of erection, the excellent green audit of timber frame buildings, and a change in the public perception of timber frame
buildings.
1
2
Timber Engineer, Center for Timber Technology and Construction (CTTC), Building Research Establishment (BRE),
Garston, Watford, UK
Head, Timber Engineering, CTTC, BRE, Garston, Watford, UK
To help the timber frame industry extend their market base to include medium rise buildings (5-8 storeys) an
experimental building, the TF2000 building, was constructed as part of a collaborative project to address any regulatory
or technological issues concerned with extending the basic principals of construction used in low-rise timber frame to that
of medium-rise timber frame. One of the issues addressed through this project was the amount of differential movement
that may be expected between different components of the building structure. Since the amount of differential movement
between the timber frame and brickwork cladding increases with respect to the number of storeys, alternative methods to
those currently adopted may be required for providing connections between the timber frame and cladding and other
building components. This is the main issue to be addressed in this paper along with other construction tolerances for
tying the timber frame to other structural building components.
Description of the experimental timber frame building
An initial architectural brief to design a block of flats that represented the most extreme case in terms of regulatory and
technical constraints was modified to produce a more realistic ‘commercial’ brief comprising:
•
•
•
•
•
•
Six storeys;
Four flats per storey to have two bedrooms and up to 60m2 floor area;
A plan aspect ratio for the building not exceeding 2:1;
Platform timber frame type construction;
Communal access to floors by single timber stairs or lift (both shafts, protected timber structure).
Brick cladding.
The building, which has a footprint of 310m², has been designed(9) to the requirements of current British Standards for
domestic use and is shown with and without the brickwork cladding in figures 1 and 2.
Figure 1: The TF2000 building without cladding
Figure 2: The TF2000 building with brickwork cladding
MONTORING OF THE TF2000 BUILDING
At the start of construction of the TF2000 building, instruments were installed to monitor the vertical displacement of
timber wall and floor components, timber moisture contents and hydrothermal atmospheric conditions in and around the
building. Instrumented areas of the building were heated after completion of the timber frame construction and
plasterboard lining to induce shrinkage in timber components of the building. Complementary laboratory tests on timber
samples taken from the building were used to separate the shrinkage and compression components of tangential
movement observed in the building.
Measurements of the height of each storey during construction and the offset of the timber frame and brickwork cladding
from a vertical plumb line provided data for the assessment of construction tolerance.
Construction tolerance
The height of each floor was first measured using a laser, with an accuracy of ±1mm, shortly after its construction. Since
the effects of shrinkage and compression were negligible at the time of measurement, this provided a good comparison
between the design height and constructed height to identify the vertical construction tolerances shown in table 1.
Even though there is no actual guidance on the overall height tolerance for the constructed timber frame, BS 5268: Part 6:
Section 6.1(1) does specify the required dimensional accuracy and flatness of timber members and wall panels. The
tolerances should be ±1mm for timber members and +0, -5mm for the wall panel size. The measured tolerances, shown in
table 1, for the TF2000 building are greater than the those specified for timber frame components, which shows that there
is some construction lack-of-fit in the building.
Tolerances for the verticality of construction prescribed in BS 5268: Part 6: Section 6.2(2) allow a maximum deviation of
12mm from vertical for timber wall panels. This helps to maintain the function of the cavity as a moisture, thermal and
noise barrier as well as allowing the use of proprietary fixings for connecting the timber frame to the cladding.
Measurements of the building from a plumb line showed that 80% of both the brickwork and timber frame did not deviate
beyond ±12mm from vertical. On average, the cavity width remained within 10mm of design and was 13mm less than the
design width at the worst recorded location.
Table 1: As-built construction tolerance for the building storey height
Building location
Storey
height2, m
Design
height2, m
Average constructed
height1, m
Cumulative
tolerance, mm
Progressive
tolerance, mm
Fifth floor ceiling
2.428
16.026
16.031
+5
-5
Fifth floor
2.839
13.598
13.608
+10
-2
Fourth floor
2.680
10.759
10.771
+12
+3
Third floor
2.680
8.079
8.088
+9
+4
Second floor
2.680
5.399
5.404
+5
+4
First floor
2.719
2.719
2.720
+1
+1
0
0
0
0
0
Ground floor
Notes:
1. The mean from measurements taken at 16 locations around the building after each floor was constructed.
2. Determined from the actual installed components.
MOVEMENT OF THE TIMBER FRAME
In addition to manual measurements of the floor height during construction of the TF2000 building, vertical movement of
first storey wall and floor components was monitored using displacement transducers. This provided detailed information
for the movement of the joists, wall panel top rails, bottom rails and sole plate and studs as shown in figure 4 for the stud
location supporting the highest observed vertical load. The vertical movement of timber components perpendicular to the
grain can generally be described as a combination of the following:
•
•
•
•
Shrinkage
Bedding-in
Compression
Delayed compression (creep)
Plasterboarding starts
Timber Frame
and Roof finnished
7.0
6.0
5.0
4.0
Summation of 1. To 4.
4. Joists
3.0
1. Sole plate and bottom rail
2.0
2. Studding
1.0
3. Top rail
23-Nov
24-Oct
24-Sep
25-Aug
26-Jul
26-Jun
27-May
27-Apr
28-Mar
26-Feb
27-Jan
28-Dec
28-Nov
0.0
29-Sep
•
8.0
29-Oct
•
9.0
Displacement, mm
•
•
Grade and sectional size of timber
Initial moisture content of timber prior to
construction.
Site conditions during construction.
Micro-climatic conditions around the timber
frame throughout its service life.
Proportion of timber in compression
perpendicular to the grain within the structural
height of the building.
Construction tolerances and imperfections.
Heating starts
10.0
•
•
Plasterboarding
finished
Measurements taken on the TF2000 building contain all of these constituents of movement although the magnitude of
each will generally depend on:
For the TF2000 building, the total downward
movement of the first storey was about 8mm after
Time
completion of the building and internal heating.
Figure 4: History of vertical movement in the timber frame
Since other storeys of the building were left unheated
and the full design load was not applied to the whole
building, detailed data collected from the first storey wall and floor panels had to be extrapolated to account for the total
movement expected in an occupied building of this type. To extrapolate the basic data lab tests were conducted on timber
samples for determining the contribution of shrinkage and compression to the overall movement observed.
Vertical movement due to shrinkage
Although the amount of movement is highly dependant on species, the per cent value for structural timber in the TF2000
building will fall within the range 1.5 – 2.2 for tangential and 0.7 – 1.0 for radial(4). This assumes a change in the timber
moisture content approximately between 20% and 12%. To ratify these theoretical values and determine the shrinkage
versus moisture content relationship for timber used in the TF2000 building, over 60 samples of the joists and panel rails
were conditioned to different moisture contents and measured. Dimensional changes affecting the in-service movement
6.0%
Shrinkage/Nominal depth, %
Moisture content, %
20.0
18.0
16.0
14.0
12.0
10.0
Period of continuous heating
24-Oct
23-Nov
24-Sep
25-Aug
26-Jul
26-Jun
27-Apr
27-May
28-Mar
26-Feb
27-Jan
28-Dec
29-Oct
28-Nov
29-Sep
30-Aug
8.0
y = 0.2423x
2
R = 0.90
5.0%
4.0%
3.0%
y = 0.1654x
2.0%
2
R = 0.59
1.0%
0.0%
8.0%
10.0%
Bottom Rail
Stud
14.0%
16.0%
18.0%
20.0%
22.0%
Moisture content, %
Time
Sole Plate
12.0%
Top Rail
Joist
Figure 5: Historical change in average moisture content
All
Top and bottom rails and sole plates
Figure 6:
Joists and Rim beams
Dimensional change of timber for
varying moisture content
were recorded to produce the relationships shown in figure 6. The observed relationships predict similar amounts of
movement to the quoted theoretical values.
A record of the moisture contents for wall and floor panels during and after the construction period provided a basis for
predicting the contribution of shrinkage to the overall downward movement. The readings in figure 5 show that the wall
panels and floor joists had an average of 17.5% and 12.5% moisture content respectively, prior to erection.
Preconditioning of the joists to a target value of 12% was conducted to reduce the amount of tangential shrinkage in-
service thereby reducing the differential movement to be considered in design to 5mm per storey. Normally a design
value of 6mm per storey would be assumed without the use of super dried timber.
After installation of the plasterboard lining in the TF2000 building, a period of continuous heating ensued in ground floor
instrumented compartments. Temperatures of around 20°C during this period reduced the moisture content of the studs,
top rail and joists to about 9.5%, whilst the bottom rails and sole plates reached about 12.5%.
Vertical movement due to tangential compression
For the consideration of compression stiffness, over 40 tests were conducted on samples of the joists and panel rails
according to the recommendations of BS EN 1193: 1998(3). To provide a direct comparison with measurements taken on
the TF2000 building, sample sectional dimensions were kept the same as components in the building (38mm x 225mm
for joists and 38mm x 89mm for panel rails). This affected the failure modes for panel rails and joist samples due to the
slenderness ratio, which produced out of plane buckling failure of the joist samples and crushing of the panel rail samples.
Examples of the test results are provided in figure 7. The relationship determined between compression stiffness and
moisture content, as shown in figure 8, was modified for the enhancing effect of adjacent unloaded parts of the timber rail
or joist. Enhancements of 15% to 59%, over the basic test compression stiffness, were obtained using the
recommendations in section 6.1.5 of the working draft for EN 1995-1-1(7). The enhanced values for compression stiffness
were used to identify this component in the overall movement observed on the TF2000 building.
20.00
Panel rail sample
Applied load, kN
16.00
14.00
12.00
10.00
8.00
Joist sample
6.00
4.00
Compressive stiffness, N/mm2
180.0
18.00
2.00
y = -334.2x + 171.1
2
140.0
R = 0.19
120.0
100.0
80.0
y = -394.1x + 137.5
60.0
2
R = 0.31
40.0
20.0
0.0
0.00
0.00
160.0
8.0%
3.00
6.00
9.00
12.00
10.0%
15.00
Example compression test results
14.0%
16.0%
18.0%
20.0%
Moisture content, %
Compressive displacement, mm
Figure 7:
12.0%
Joists
Figure 8:
Panel rails
Change in the compressive stiffness with
varying moisture content
Bedding-in and Delayed compression
Although the contribution of delayed compression (creep) to the overall movement has been identified as potentially large
from tests conducted on small specimens(8) in the past, laboratory tests have shown that the majority of movement on the
TF2000 building is attributable to shrinkage and elastic compression. The remaining 39% of observed movement on the
building must be attributable to both bedding-in and delayed compression (creep).
DIFFERENTIAL MOVEMENT OF THE TF2000 BUILDING
Using the results of the shrinkage and compression tests, a prediction of the timber frame vertical movement has been
determined in figure 9. This negates any movement prior to occupation and is based on the following assumptions:
•
•
•
Compression movement, both elastic and creep will occur in the timber frame from the imposed occupancy load
only. This assumes that the plasterboard lining is complete at this stage or its’ load is supported inside the building.
The building is not heated to induce shrinkage until occupancy of the building.
Bedding in movement will have occurred prior to erection of the brickwork cladding
Solid timber joists and rim beams are used throughout the
building for floor panels and have been pre-conditioned to about
12% moisture content or less –typically that observed on the
TF2000 building.
The prediction shows that the vertical movement of the timber frame
at eaves level would be about 19.5mm for consideration of differential
movement. If the floor joists and rim beams had not been
preconditioned to about 12% moisture content, the total vertical
movement would have been doubled.
In consideration of the differential movement between the brickwork
cladding and timber frame, both thermal and moisture movement of
the clay brickwork was considered. Although a practical assessment
of the cladding movement could not be made as part of this project
because of the protected environment in which the building was
constructed, theoretical bounds provide a good estimate of the
anticipated movement. Three types of clay brickwork were adopted on
the TF2000 building with different expansion rates due to moisture;
this is generally more dominant than thermal movements.
Consideration of both the reversible and irreversible movement for the
brickwork gives values for the expected expansion at eaves level
ranging from 8.4mm to 12.3mm. This in combination with the timber
frame movement predicts a differential movement of 27.9mm to
31.8mm (4.6 to 5.3mm per storey).
6
5
4
Storey of building
•
3
2
1
Compressive movement
Shrinkage movement
Total = compression + shrinkage
0
0
5
10
15
20
25
Vertical movement, mm
Figure 9: Predicted movement of the TF2000
timber frame due to occupancy
CONCLUSIONS
Differential movement between the timber frame and masonry cladding is an important consideration for the design of
medium-rise timber frame buildings. When designing for the downward movement of the timber frame, both shrinkage
and compression should be considered due the effects of occupancy heating and variable loads. The use of super dried
timber for joists and rim beams with target installed moisture contents of 12% or less can be used to effectively reduce the
amount of tangential shrinkage in-service. This reduces the total amount of downward movement, which may also be
achieved by using engineered wood products such as LVL, Glulam, Parralam, I-beams, etc., which are manufactured at
low moisture content.
The TF2000 building incorporated a common type of flexible wall tie for attaching the brickwork cladding to the timber
frame. Although the specification for this type of wall tie states that it is only able to accommodate 30mm of vertical
movement, a suitable design for differential movement may be achieved using this type of wall tie provided that:
•
•
•
•
Measures are taken to reduce the amount of cross-grain shrinkage (i.e. Use of Super-dried timber or engineered
timber products).
Timber is suitably protected from the elements to ensure little deviation from the target moisture content for
construction on site.
Construction of the roof is completed and the full weight of plasterboard lining is supported by the timber frame prior
to construction of the brickwork cladding.
Brickwork with a low expansion rate due to moisture is adopted in the design of the building.
REFERENCES
1.
British Standards Institution. 1996. Structural use of timber: Part 6, Code of practice for timber frame walls: Section
6.1, Dwellings not exceeding four stories. BS 5268: Part 6: Section 6.1: 1996. BSI. London.
2.
British Standards Institution. 1998. Structural use of timber: Part 6, Code of practice for timber framed walls: Section
6.2, Buildings other than dwellings. BS5268: Part 6: Section 6.2: 1998 BSI. London.
3.
British Standards Institution. 1998. Timber structures: Structural timber and glue laminated timber: Determination of
shear strength and mechanical properties perpendicular to the grain. BS EN 1193: 1998. BSI. London.
4.
Desch,H.E., and Dinwoodie,J.M. 1996. Timber: structure, properties, conversion and use. 7th edition, MacMillan
Press Ltd.. London. pp 91-94.
5.
Enjily,V., and Mettem,C.J. 1995. Medium-rise timber frame buildings: Disproportionate collapse and other design
requirements. Joint BRE/TTL publication. UK.
6.
Enjily,V., and Palmer,S. 1996. Timber frame 2000, Phase I: Summary of commercial and technical findings, Joint
BRE/TTL publication. UK
7.
European Committee for Standardisation (CEN). 1999. Eurocode 5, Design of timber structures: Part 1.1, General
rules and rules for buildings. First draft EN 1995-1-1. Document CEN/TC 250/SC 5: 124. Trätek. Stockholm.
8.
Morlier, P.(ed). 1994. Creep in timber structures. RILEM report 8. E & FN Spon.
9.
Steer,P.J. 1998. Design of the TF2000 building. COST-E5 workshop. BRE. Watford.