You are on page 1of 73

17-19 Gladding Place

P O Box 76 134
Manukau City,
New Zealand

Phone: +64-9-262 2885


Fax: +64-9-262 2856
Email: structural@hera.org.nz
HERA Webpages: www.hera.org.nz

No. 71 December 2002/January 2003


The author(s) of each article in this publication are noted at the The slab panel design procedure detailed herein has been
beginning of the article. the subject of development and review by a number of
people. The effort and input of these people is greatly
appreciated.

Introduction In This Issue Page


On-line DCB Now Available 1
Almost all of this issue is devoted to the
presentation of the Second Edition of the Slab Second Edition of the SPM Design
3
Panel Method of Floor Design for Dependable Procedure
Inelastic Response in Severe Fires. Called the • Overview and General 3
SPM design procedure, the first edition was • Figures for SPM 15
officially released in DCB No. 60, 2001 and has • Appendix A: Detailed 36
undergone a thorough revision over the second Procedure
half of 2002, incorporating results and knowledge
gained from a landmark series of Slab Panel fire References 72
tests undertaken in mid-2002.

Preceding this are two very short articles. The first On-line Design & Construction
announces the new on-line DCB now available. Bulletin Is Now Available
Second is a rebuttal to the analysis of Santa
Claus - thus closing the debate on that fellow for Pages 6 to 19 of DCB No. 69 contain an article
Christmas 2002! that covers all the current guidance presented in
the DCBs from July 1995 to August/September
Before commencing with the first article, a long- 2002. Only current guidance is included; where
awaited announcement: later guidance on a topic supersedes earlier
guidance, either this is stated or, more typically,
AS/NZS 2312:2002 is available the earlier guidance is not mentioned.
After a longer than anticipated gestation period, The details presented in the ‘what’s current’ article
the replacement to the 1994 edition of the are grouped under the following topic headings:
steelwork corrosion protection standard is
available. • Contractual issues and quality
• Design examples, design queries and
AS/NZS 2312:2002 [1], Guide to the Protection of design concepts
Structural Steel Against Atmospheric Corrosion by
• Design for durability
the Use of Protective Coatings, represents a very
• Design for earthquake
significant advance in the specification of cost-
• Design for fatigue
effective corrosion protection solutions that will
meet the durability requirements of the NZBC [4] • Design for fire and behaviour in fire
and the client. • Design for serviceability
• Design of specific types of structures
More details on this new standard and the • Design of connections
advantages it offers will be given in DCB No. 72, • Design of structural hollow section members
February/March 2003. and connections
• Design of members not listed elsewhere

HERA Steel Design & Construction Bulletin Page 1 No. 71, December 02/ January 03
• Design of non-ferrous metals The analysis presented in DCB 70 applies
• Innovative and economical steel design Newtonian mechanics and philosophical
• Material properties and availability viewpoints to a problem nested in the relativistic
• Publications domain. Santa does not have to visit 100 million
• Research results homes in one night. Instead, by use of time
• Steel Structures Standard, NZS 3404 transformation techniques he simultaneously visits
• Composite construction all 100 million homes.
• Miscellaneous items A basis for this lies in the recent works of
Similsone on Clifford Algebras. This work is an
Each topic heading gives brief details of each item extension of the work in the mid '70's of Prigigione
covered under that topic and its location, with in theoretical thermodynamics. His analysis
the location specified in the format DCB No. xx demonstrates that ensembles or collections of
pp. a – b things exhibit a drop in entropy relative to an
uncollected state.
This material can also be accessed on-line, as of
the beginning of January 2003. To do this, users For example, the collection of neurons in the
simply access the HERA web-site at human biocomputer is one of the most organized
www.hera.org.nz, click on the On-Line Design and collections to be found. The entropy drop is
significant, and in fact life itself can be shown not
Construction Bulletin link in the lower left hand
corner and follow the instructions. to exist without this drop. The universe taken as a
collection exhibits an entropy drop nearly 100
Two levels of access are available. The first, million times as great as the human mind, leading
which is available at no charge, accesses just the to conclusions by theologians of a self-aware and
topic headings that are given above. The second fantastically intelligent universe.
level gives full access, which allows users to bring An alternative interpretation of this work is a
up the details of each item covered under a given distortion of time, folding not just space but also
topic, then to click on the DCB No. xx, pp. a – b time back upon itself so that a single point in time
link that will take them to the beginning of the can exist an infinite number of times. A further
selected article. interpretation is that all time may only be one
instant.
Please note that Full Access is available to DCB
Subscribers only. Current subscribers have This suggests that from our viewpoint, Santa visits
already been sent the required password. a lot of places instantaneously. From Santa's
viewpoint, however, he visits each home
The earlier DCB’s were not produced in a format individually in sequence. The time required for this
that could be accommodated on-line, so this on- is in the base time (i.e. based on the planetary
line access starts with DCB No. 38 and then rotation) so that Santa does not age appreciably.
covers all issues from Nos. 40 to 69. This covers Rumours coming from the Aurora Borealis Bar,
most of the material that is still current. If the indicate that in fact Santa visits this place daily
‘what’s current’ article refers to an earlier issue, after delivering to two dozen or so homes. He and
then a note appears saying that this material will Rudolph hang out there for several hours. The
be sent by post on request. other reindeer, piqued at the apparent favouritism
toward Rudolph, tend to go to the Southern Cross
The HERA Structural Engineer would like to (Bar names are notoriously illogical) where they
gratefully acknowledge the work of Diana drink J&B on the rocks. Despite his Germanic
Reichwein, from the Fachochschule Ravensburg - origins, Santa has a decided predilection for
Weingarten, who has developed this interactive Armagnac while Rudolph, loyal to his Scandinavian
facility. origins, drinks aqua vie. Any rumours of a prurient
nature are totally unfounded as Rudolph is often
Santa Claus – The Rebuttal!! seen leaving with one or more female reindeer.
Santa appears to be quite faithful to Mrs. Claus.
In a moment of pre-Christmas whimsy, the HERA Structural
Engineer presented an Engineer’s view of Santa Claus! Due to the concurrency of time at this time, it is not
Regretfully, this view (which he must admit came from an uncommon to see several hundred thousand
unknown source) concluded that Santa couldn’t exist! Santas and Rudolphs simultaneously. The Aurora
Showing the versatility for which New Zealand Academics are Borealis is the second most successful lounge on
famed, John Butterworth of the University of Auckland has sent earth, second only to the Southern Cross in fact,
in the following rebuttal, which is based on the observation that, where over a million flying reindeer are seen at any
as Santa does deliver the goods, he must exist!
given time.
This closes the issue of Santa (to be or not to be…) for another Kendall et al provide further evidence for the actual
year! existence of Santa. Their soon to be published
work in self-repairing DNA and RNA is also based

HERA Steel Design & Construction Bulletin Page 2 No. 71, December 02/ January 03
upon Prigigione's work. Under certain conditions, The first SPM procedure was published in DCB
induced in humans by universal love, humour and No. 60, February 2001, with the development work
generosity, the organism can repair damage to on which it was based published in that issue and
itself. There are references to this condition all in DCB No. 59. Since then, feedback from use of
through history in the literature starting with the the procedure has resulted in amendments and
epic of Gilgamesh (approx 4,500 BC) to as recently further research. The most significant
as "Lost Horizon" written in 1933. Some of the development since February 2001 has been the
research in Gaia may bear results in this area also. undertaking of a series of 6 slab panel fire tests,
details of which are given in [3, 36]. The results of
Thus it is seen that it is entirely possible for Santa
these tests have validated the concept and
to exist and that there is, in fact, substantial
enabled revisions to the procedure to give greater
evidence of his actual existence. Coupled with a
accuracy and to remove some elements of
suitable attitude and minor application of the time
conservatism. Details of these revisions are given
warp, Santa brings the goodies through to the kids.
in DCB No. 70, pp. 1-19.
Just a jump to the left…
While a considerable number of publications are
Design of Multi-Storey Steel referenced in this bulletin, when it comes to the
Framed Buildings With application the procedure is written in such a way
that it can be applied in accordance with
Unprotected Secondary Beams publications commonly available to a structural or
or Joists for Dependable fire design engineer.
Inelastic Response in Severe Naturally, this second edition of the SPM
Fires: Second Edition supersedes the first edition of the procedure and
the SPM computer program dated SPM0302.
This article and design procedure has been written by G
Charles Clifton, HERA Structural Engineer and Christopher
Beck, Undergraduate Student on Study Leave at HERA from
The new program is entitled SPM0103.
the Fachhochschule Offenburg.
1.2 Scope and content
1. Introduction and Scope
The basis of the design procedure is presented in
1.1 General background section 2. The structural performance that will
be delivered by the procedure is described in
This DCB presents the second edition of the Slab section 3.
Panel Method of Fire Emergency Design (SPM), The procedure is only applicable to buildings with
which allows for the use of unprotected secondary certain structural characteristics, which are
beams or joists supporting concrete floor slabs and described in section 4. As with any procedure
exposed to severe fire conditions. It is written for written to provide dependable inelastic behaviour,
application to multi-storey steel framed buildings of the detailing is as important as the design and the
Fire Hazard Category (FHC) 1, 2 or 3 and caters detailing requirements are presented in section 5.
for the maximum structural fire severity possible
from this range of building occupancies. (FHC is as Maintaining effective compartmentation is covered
defined in Comment to Paragraph 2.2 of [5]). in section 6, while controlling the onset of structural
damage is outlined in section 7. Section 8
The slab panel design procedure is fundamentally introduces the SPM0103 computer program, which
different in philosophy and procedure to traditional implements this second edition. Section 9 then
fire engineering design provisions used in New presents a design example including showing the
Zealand. This is because the procedure is written use of the program. Section 10 gives pointers
for application to large sub-assemblages in on application of the procedure for maximum
buildings subject to fully developed fire conditions cost effectiveness. This is followed by
involving unprotected steel members, where these acknowledgments, in section 11.
members may be subjected to very high
temperatures and considerable inelastic demand in Following the acknowledgments are all the figures
the event of fully developed fire. The extent of this for this issue.
inelastic demand is anticipated and a dependable
proportion of the additional reserve of strength The detailed steps of the procedure are presented
available from the building when undergoing this in Appendix A, which includes a commentary to
deformation is incorporated into the procedure. these steps.
The procedure is also written around the
temperatures that unprotected steel beams can Finally, the references to this issue are presented.
realistically reach in fully developed FHC 1, 2 and
3 fires.

HERA Steel Design & Construction Bulletin Page 3 No. 71, December 02/ January 03
2. Basis of Slab Panel Design Procedure The rest of section 2 provides a very brief overview
of some experimental testing and analytical
2.1 General modelling work that supports the basis to the
procedure.
Under ambient temperature conditions, the beams
in a composite floor system support the floor slab. 2.2 Ambient temperature testing
For example, with reference to Fig. 71.1 herein, the
load path involved in resisting dead and live loads As part of the development work for the tensile
under ambient temperature conditions is: membrane model, Bailey [9] undertook a large-
scale ambient temperature tensile membrane test,
Slab → secondary beams → primary beams → columns details of which are shown in Figs. 71.2 and 71.3.
(71.1)
Fig. 71.2 shows the general specimen set-up. This
Under severe fire conditions, when the secondary
involved constructing a slab panel with 9.5 metre
interior beams are unprotected, they lose most of
by 6.5 metre centreline to support dimensions,
their strength, such that the ambient temperature
formed by casting the floor system onto a profiled
load path in equation 71.1 cannot be maintained.
steel deck base that was removed prior to loading.
As a result, the beams form plastic hinges and the
The slab panel was reinforced only with A142
load-carrying mechanism changes to a two-way
shrinkage and temperature control mesh; the
system. Under severe fire conditions incorporating
standard mesh used in the UK for a 130 mm thick
unprotected secondary beams, two-way action
profiled slab. The bar diameter and pitch is shown
prevails, involving the region of slab and
in that figure.
unprotected secondary beams known as a slab
panel.
The slab panel reinforcement, thickness and
dimensions were chosen so that the self-weight of
The slab panel resists applied load by two-way
the slab panel equalled the calculated yieldline
action back to the supports, through a load path
load-carrying capacity of the panel, using the
involving:
actual mechanical properties of the mesh. This
meant that any additional load-carrying capacity
Slab panel → supporting beams →columns (71.2)
was due to tensile membrane action (plus a minor
contribution from strain hardening of the
This is illustrated in Fig. 71.1. The same concept
reinforcement).
is applicable to floor slabs supported on closely
spaced joists, such as the Speedfloor systems.
Fig. 71.3 shows the position of applied loads and
two patterns of cracking. The loading pattern and
The slab panel develops its load-carrying capacity
application was such as to induce effective uniform
in the deformed state through:
applied load, while allowing access to the top
Yieldline moment action, plus surface for observation and instrumentation.
Tensile membrane enhancement
The yieldline cracking pattern shown in that figure
The loads transferred from the slab panel into the had formed on the underside on removal of the
supporting beams (ie. as given by the tributary decking.
floor slab areas contained within the yieldline
patterns shown in Fig. 71.1) must be resisted by The imposed load was applied in increments, until
those supporting beams and transferred back to at an applied load of 0.8 times the yieldline load
the columns. carrying capacity, the cracking pattern associated
with tensile membrane action started to form. This
The basis of the design procedure has
pattern was fully developed at an applied load of
been described in two places; first in
0.96 times the yieldline load carrying capacity (ie. a
section 2.3, pp. 7-8, DCB No. 59 and secondly in
total load of 1.96 times the self-weight). The slab
the section Basis of Slab Panel Design Procedure,
panel did not collapse at that point; the loading
pp. 153-159 of [22]. The results from the
mechanism ran out of stroke, meaning that no
experimental fire test [3] have not changed the
further load could be applied.
validity of these two mechanisms (yieldline
moment action and tensile membrane
There was comprehensive monitoring of stresses,
enhancement), however they have shown that the
strains and deflections, as reported in [9].
mechanisms do not act sequentially in a slab panel
under severe fire attack, but rather act
These results and those of subsequent ambient
concurrently. This is elaborated on in section 5,
temperature tests from the UK confirmed the basis
pp. 17-19 of DCB No. 70.
of the design procedure under ambient
Readers are invited to study either of the above temperature, in terms of the two structural
two references for more details on the basis of the mechanisms involved and their sequence of
procedure. operation.

HERA Steel Design & Construction Bulletin Page 4 No. 71, December 02/ January 03
2.3 Fire testing of slab panels Fig. 71.10 shows the deformed shape of the D147
slab; very good agreement between predicted and
This comprises the six 4.15 metre x 3.15 metre experimental deflections have been obtained for
slab panels tested by Lim et.al. [3, 36] and this slab panel, as shown in Fig. 20 of [36].
overviewed in section 3, pp. 3-6 of DCB No. 70.
Having validated the computer modelling, the
Fig. 71.4 shows the test set-up section across the strain/stress outputs from that program can be
short span of the furnace, while Fig. 71.5 shows used to track and quantify the structural
the applied loading regime. Fig. 71.6 shows a slab mechanisms operating under fire conditions
in position on the furnace, prior to the test through the test. These details are shown in [36]
commencing. for the D147 slab. They show that the combination
of yieldline and tensile membrane action resists the
Figs. 71.7 and 71.8 show the cracking pattern on loading but in a sequence of development very
the top and bottom respectively of the D147 slab different to that under ambient temperature loading
panel (the 100 m thick panel with minimum to destruction. This is discussed further in section
reinforcement). 5, pp. 17-19, DCB No. 70.

The results of these tests have been used to These analyses also show the strain and stress
amend the first edition of the SPM procedure as state in unprotected elements of supporting
described in the Commentary to Appendix A. members (the Speedfloor joists in this instance) to
be very complex; details are outlined in sections
Only by fire testing can the performance of a slab CA4.1 and CA4.2.1 of Appendix A herein.
panel under the representative ultimate limit state
3. Structural Performance to be Delivered
environment of constant vertical load and severe
by Procedure
fire conditions of high temperatures and thermal
gradient development be experimentally 3.1 Under severe fire conditions
determined.
The structural performance that will be delivered by
In this regard, the fire tests [3, 36] have generated this procedure, in the event of fully developed fire
crucial information that the ambient temperature conditions, is as follows:
tests cannot provide.
(1) Slab and unprotected secondary beams
When it comes to verifying the internal structural may undergo appreciable permanent
mechanisms within the slab panel, however, these deformation.
can only be recorded through to failure under
ambient temperature loading, as done in [9]. The The maximum extent of this deformation is
same cannot be done in the fire tests, as the strain described in section 4.4 of DCB Issue No.
gauges all failed due to elevated temperatures at 70. In practice, the inelastic demand would
around 60-100 minutes into the tests, usually be less, for the following reasons:
corresponding to their limiting temperatures of
300oC and well short of slab panel failure. • Lower fire load
Therefore, it requires analytical modelling to track • Presence of shielding linings
strains and corresponding stresses through to slab • Non-fire rated enclosures reducing the
panel failure in fire conditions. fire size
• Fire service intervention
2.4 Analytical modelling of slab panels
(2) Support beams and columns will undergo
minimum permanent deformation, compared
This is the second major part of Linus Lim’s
to that within the slab panel
project. The results are not yet (end of 2002)
published, so brief mention here is made of the (3) Load-carrying capacity and integrity of the
analytical work using figures from the already floor system will be preserved.
published reference [36].
(4) Insulation requirements will be met for at
Fig. 71.9 shows the finite element discretisation of least the F rating times specified by C/AS1
one quarter of the slab panel. This modelling was [5].
undertaken using SAFIR; a heat-flow and structural
analysis in fire program described in DCB No. 59 (5) Local collapse and global collapse will be
and in [36]. dependably prevented.

One of the principal purposes of the analytical In practice, structural repair and reinstatement of a
modelling has been to validate the program using steel building designed to this method will be
the experimental test results. Details of the almost as straightforward as that for a building with
principal steps involved are given in [36]. all floor support beams protected, based on

HERA Steel Design & Construction Bulletin Page 5 No. 71, December 02/ January 03
experience from actual fires in modern, multi- Positions, covers are as specified in
storey framed buildings. Appendix A.

3.2 Maintaining effective compartmentation 4.2 Steel beams/joists


Effective compartmentation will be maintained, Typically all steel beams or joists will be composite
both between floors and between firecells on the with the floor slab.
same floor. The former is a consequence of the
floor system performance that will be dependably If these beams are not composite, then the
delivered by this procedure. The latter may require supporting beams around the sides of the slab
special detailing, which is covered in section 6. panel, as shown in Fig. 71.1, require shear studs to
3.3 Suppression of structural damage NZS 3404 Clause 13.3.2.3 (h); ie. a single stud at
a maximum spacing of 4to, where to = overall slab
Although this slab panel design procedure thickness.
anticipates and allows for significant floor system
deformation and associated structural damage, this Hot-rolled beams, welded beams, Speedfloor
can be minimised through: joists, beams with web openings are all suitable.
However, see commentary section CA4.2.1 for
• Shielding linings (limited effectiveness) guidance where the unprotected secondary beams
• Sprinkler protection (extremely effective) are either of non-uniform cross section or contain
web openings.
This is covered further in section 7.
See section A6 of Appendix A for the extent of
4. Building Structure Characteristics protection required to the slab panel supporting
Required for Implementation of Slab beams.
Panel Design Procedure
The design structural fire severity (FRR) to use for
4.1 Floor slabs supporting beams, when determining the extent of
passive fire protection required in accordance with
Floor slab characteristics required are;
section A6.2 step 3 option 3.1, is given by:
(1) Concrete is normal weight, (NWC)
• FRR = 1.0 teq for beams with Tl < 700oC
f c' ≥ 20 MPa.
• FRR = 0.9 teq for beams with Tl ≥ 700oC
Aggregate selection for slabs not cast onto Tl = beam limiting temperature to Clause 11.5 of
steel decking must be to section 4.7 of DCB [10].
No. 70, pp. 16, 17, to suppress spalling.
The procedure is readily expandable to light See DCB Issue No. 59 for the background to these
weight concrete (LWC); guidance on recommendations, except that the 0.8 factor given
therein has been increased to 0.9 as a result of the
determining temperatures for some slab
panel components using LWC floor slabs is fine-tuning of the method to the experimental fire
already included in Appendix A. test results [3].

(2) Mesh reinforcement Intumescent paint protection of the protected


beams is an option, however, for the high structural
As decribed in the notes to Fig. 71.12, all fire severities expected it would be expensive.
mesh reinforcement is to be grade 500L to There is also the issue of ensuring that the target
AS/NZS 4671 [24]. The mesh can be limiting temperatures are met for beams and
formed from deformed or plain bars. Mesh columns (section 4.3 below), where these limiting
bar spacing must be between 150 mm and temperatures are less than the typically 620oC
250 mm centres. limiting temperature for which intumescent paint
thicknesses are determined. Finally, there is an
Position, covers are as specified in Appendix issue with insulation using intumescent paint; see
A. the last paragraph in section 6.2.1 herein.

(3) Bar reinforcement 4.3 Columns


Interior support bars (ie. bars which span Bare steel columns are required to be passive
over a support into two adjacent slab panels) protected full height. The limiting temperature to
are to be deformed grade 500E to [24]. use and the FRR to apply are given in section 7.5
Edge and trimmer bars are to be deformed of DCB Issue No. 59, with this advice repeated
grade 500E or grade 500N. below.

HERA Steel Design & Construction Bulletin Page 6 No. 71, December 02/ January 03
The thickness of insulation material required for the most common examples are shown in
columns should be as follows: Fig. 71.11.

(1) For columns with cross sections complying 4.4.2 Connections to supporting edge beams
with the cross section geometry provisions
of NZS 3404 [10] Table 12.5 for category 1, These will be subjected to lower rotation and axial
2 or 3, use FRR = 1.0teq. (These columns force demands from fire, either because the beams
have a limiting temperature of 550oC) are protected or because they possess a very high
reserve of strength. However, they will typically be
(2) For columns with flanges that are more connections between supporting edge beams and
slender than the category 3 limit, use the columns (eg as shown in Figs 71.17, 71.18) and
following: therefore subject to earthquake design and
detailing requirements. These requirements will
• If profile protected, FRR = 1.25teq cover the fire condition.
• Box protected except with Promatect H,
FRR = 1.0teq Once again, connections given in [6] are suitable.
• Box protected with Promatect H,
FRR = 1.25teq 4.5 Extending passive protection into the
ends of unprotected supporting members
(These columns have a limiting temperature
of 400oC, as given below ) There are many instances where unprotected
secondary beams will be supported by protected
In both cases, the insulation material should be primary beams. In these instances, the passive
selected in accordance with the manufacturer’s fire protection on the primary beam must be
published data for the FRR, SF and limiting extended slightly into the end of the unprotected
temperature used. If the regression analysis data secondary beam, to avoid significant localised heat
is not available to allow the thickness to be altered flow from the secondary to primary beam. The
as a function of the limiting temperature, then the extent of this overlap is discussed on pp. 11, 12 of
data will have been derived for T1 = 550oC and the DCB No. 66 and the recommendations from there
use of 1.25teq is therefore necessary in instances, are as follows:
as given above, to meet the lower limiting
temperature of 400oC, which is applicable to (1) When an unprotected beam member is
columns more slender than category 3. connected into a protected beam or a
protected column member, the insulation
For unprotected, concrete-filled structural hollow material used on the protected member
section columns, see DCB No. 58, pp. 25-30 and should extend onto the ends of the
the referenced Canadian paper. Use FRR = 1.0teq unprotected member so as to:
for these unprotected, concrete-filled columns. (1.1) Completely cover the connection region, any
connection components and connectors, as
4.4 Connections
shown in Fig. 71.11, and
4.4.1 Connections from unprotected (1.2) Extend at least 200 mm into the span of the
secondary beams to supporting beams unprotected beam from its end, also as
or columns shown in Fig. 71.11.

Connections between unprotected secondary (2) The insulation material should be the same
beams and primary beams or columns are type and thickness as used on the protected
subjected to significant inelastic rotation up to the member and must be detailed to dependably
maximum temperature, then are subject to tension remain in place during a fire. (The latter is
force during the cooling phase. This is described important for boarded insulation materials, in
in more detail on pages 11, 12 of DCB Issue No. terms of how they are fastened at the ends
54 and illustrated in, for example, [12]. These of the otherwise unprotected members).
connections must be designed and detailed to (3) When an unprotected Speedfloor joist
retain their integrity during both the heating and the is connected to a protected beam (see
cooling phases. Fig. 71.16) the insulation material should
completely cover the area of the joist shoe
These requirements are similar to the severe through which the joist is connected to the
earthquake requirement to retain integrity under beam and the area of the joist web
seismic-generated inelastic rotation demand. The connected to that joist shoe. (Refer to the
same design and detailing concepts should be steel beam support detail in [14] for details
used. All the connections given in HERA Report of this shoe).
R4-100 [6] will be suitable in this regard. Two of

HERA Steel Design & Construction Bulletin Page 7 No. 71, December 02/ January 03
(4) Protected beams supporting Speedfloor (3) When unprotected concrete-filled structural
joists should use spray applied protection or hollow section columns are used
ceramic fibre blanket protection, which is
able to easily cover the surfaces involved in • FRR is provided to DCB Issue No. 58,
(3) and will dependably remain in place pp 25-30, and the Canadian method
(pinned where necessary) in a severe fire. referenced from that DCB
These requirements cannot be easily • For protected beams to these
achieved with board protection in that columns, treat the connection region
application. as shown in DCB Issue No. 42, Fig.
42.2, by running the passive
4.6 Overall structural stability protection over the column within the
depth of the connection region.
The slab panel design procedure is applicable to
all structural systems. There are no specific 6. Maintaining Effective Compartmentation
limitations on type or position of lateral load-
resisting system elements imposed by the use of 6.1 Floor to floor compartmentation
this method.
This is covered by the slab panel design
Lateral load-resisting systems will have protected procedure, which will ensure that the floors
columns, which will therefore be subject to function as fire separations throughout the fire.
negligible inelastic demand in fire, and either
protected beams or beams with a very high 6.2 Compartmentation between firecells on
reserve of strength in severe fire conditions. the same floor
These systems will therefore retain their integrity
under severe fire conditions. Fire separating walls on the same floor must cope
with the expected slab panel deformations. There
All gravity system columns will be protected and are two cases to consider; walls located under slab
hence subject to negligible inelastic demand in fire. panel support beams and walls located within the
slab panel.
Thus local and global structural stability will be
retained throughout the fire. 6.2.1 Compartment walls under slab panel
support beams
5. Detailing Requirements for Use With
the SPM Procedure These beams will not deform more than is
expected from a conventionally protected beam
As with any system designed to deliver a and no special detailing is required, as shown in
dependable level of inelastic response, the Fig. 71.19.
detailing is as important as the design.
Insulation requirements must be fulfilled and voids
(1) This especially relates to the floor slab, and service penetrations must be fire stopped.
where:
Beams protected with intumescent coatings require
• Decking must be fastened to beams
additional insulation because the temperature on
to NZS 3404 [10] Clause 13.3.2.4
the non-fire side is likely to exceed the temperature
• Mesh must be lapped to NZS 3101
rise associated with meeting the insulation criteria.
[11] Clause 7.3.21
• Bars must be lapped to NZS 3101 6.2.2 Compartment walls within the slab panel
Clause 7.3.17 region
• See Figs. 71.12 to 71.18 for detailing
of reinforcement These must be detailed to accommodate the
• Covers are important and must be expected slab panel deformation. A deformable
placed as specified in Appendix A and blanket with the necessary fire resistance (stability,
Figs. 71.12 to 71.18. insulation) must be used, as shown in Fig. 71.20.

(2) When passive fire protection is specified, it The deflection allowance recommended is for Lx/25
must be placed as specified, especially: over the middle half of the slab panel, reducing
linearly to zero at the supports. Apply this in both
• Full length of beams the x-direction and the y-direction, as appropriate.
• Full height of columns
• Into the ends of unprotected Deformable fire rated systems are available, as
secondary beams (section 4.5 and they are needed to maintain fire resistance across
Fig. 71.11) seismic gaps. An example is given on pp. 1, 2 and

HERA Steel Design & Construction Bulletin Page 8 No. 71, December 02/ January 03
Fig. 61.1, DCB No. 61. Contact HERA for details • In an unsprinklered building, the onset of
of systems and suppliers, if required. structural damage to the floor should be
delayed until a dependable period of time
7. Controlling The Onset of Structural after full fire development commences. This
Damage From Severe Fire period of time is set at 15 minutes. To
achieve this, the floor system needs to be
7.1 Establishing a satisfactory threshold for shielded from direct exposure to the fire by a
the onset of permanent floor system ceiling system capable of remaining intact
deformation for around 20 minutes of standard fire test
time. This does not have to be a fire rated
The SPM procedure takes account of the system, but one which can, on the basis of
dependable reserve of strength available from a standard fire test performance, be expected
floor system when it undergoes the significant to remain in place for the 20 minute of
deformation that will be developed by sustained standard fire test exposure. Guidance on
fully developed fire attack. It is an ultimate limit ceiling systems suitable for this are given in
state design procedure, similar in this regard to the section 7.3.
concept of design of a building for limited ductile to
fully ductile response to earthquake. • It must be remembered that the above is
simply to create a damage threshold against
This raises the very important question as to what floor system deformation commencing
is a satisfactory damage threshold for the design of immediately an commencement of fully
a floor system to the SPM procedure. This is an developed fire conditions in an unsprinklered
issue that we have considered carefully in the building. The radiation barrier is not
development of all HERA fire engineering design required to deliver the dependable floor
procedures for steel structures and our approach system response given by the SPM method.
has been as follows:
• The above approach exceeds the
• If fully developed fire conditions in a firecell requirements of the NZBC [4] and should
are allowed to develop, the damage to the meet any realistic client and/or insurer
firecell above will be extensive from smoke expectations that are in excess of those
and heat, not withstanding any structural from [4].
damage.
7.2 Sprinkler protection
• Given that even the onset of fully developed
fire conditions represents a damaging This is the preferred method; it suppresses
condition, what is the likelihood of full fire structural damage by suppressing full fire
development compared with the likelihood of development.
other damaging events such as severe
earthquake? The annual probability of fully developed fire
occurrence in a sprinklered multi-storey building of
• The answer to that is that, in an the type for which this procedure has been written
unsprinklered building, the fully developed has been assessed by Feeney and Buchanan [18].
fire condition is at least 5 times more likely to It is extremely low, as shown by the following
occur over a 50 year timeframe (the comparisons:
assumed lifetime of the building) than is the
severe earthquake, which is more likely than • Annual probability full fire development, no
any other structurally damaging event [18]. earthquake, sprinkler failure, is 1.2 x 10-5
• Annual probability full fire development
• In a sprinklered building, the probability of following earthquake, sprinkler failure,
fully developed fire occurring is very much is 1 x 10-4
lower than that of severe earthquake or any • Compare with annual probability of severe
other likely structurally damaging event. earthquake in the highest seismic regions,
which is 2.2 x 10-3.
• The approach we have taken, with the SPM
procedure, is that, in a sprinkler protected Sprinkler protection also brings many other
building, how quickly structural damage benefits with regard to life safety and property
would occur after the onset of fully protection.
developed fire is not important. This is
because damage will only occur in the very
remote event of sprinkler failure, as
elaborated on in section 7.2.

HERA Steel Design & Construction Bulletin Page 9 No. 71, December 02/ January 03
7.3 Shielding of the floor of an unsprinklered Indicative times for given types of tile for
building by a non-fire-rated radiation which a non-fire-rated suspended ceiling, with the
barrier cross-runner span ≤ 600 mm, will remain in place
for ≥ 20 mins are:
7.3.1 Overview
(i) For fire-resistant plasterboard tiles, using
Shielding does not suppress full fire development; 12.5 mm thick plasterboard: 25 minutes
it only provides a radiation barrier for a limited
period of time. This shields the floor system (ii) For fire-resistant plasterboard tiles, using 16
above the barrier from the fire generated radiation mm thick plasterboard: 30 minutes
below. Provided that the shielding system remains
in place for the period of high fire temperatures, it (iii) For steel backed tiles: 60+ minutes
will keep the temperatures of unprotected beams
above the barrier and the thermal gradient in the 7.3.3 Use of a non-fire-rated fixed ceiling or
floor slab much lower than they would otherwise lining as a radiation barrier
be, with correspondingly lower floor system
deformation. This has been graphically illustrated A fixed ceiling is one which comprises a monolithic
in Australian natural fire tests, eg [33], on well ceiling supplied and fixed in accordance with
ventilated enclosures with fire loads corresponding the manufacturers recommendations. The most
to FHC 2 and FHC 3. common product used in New Zealand is
Gib®Board, from Winstone Wallboards Ltd.
The effectiveness of shielding linings is very
dependent on the system used. For small fire The performance of the Gib®Board range of ceiling
cells, moderate structural fire severity fire loads systems has been well established under standard
(FHC 1), moderate to high levels of ventilation and fire test conditions and verified by its performance
fixed linings, the system is very effective. It is in numerous actual building fires. Winstone
routinely used in FED in such applications, eg. as Wallboards fire engineering personnel have
described in [21]. provided technical guidance on the time at which
each system can be considered to suffer breach of
For high structural fire severity fire loads (FHC
integrity under fully developed fire conditions.
2, 3) shielding linings will only be effective for part
of the time of fully developed fire occurrence These times for ≥ 20 mins in place are:
unless there is a high level of ventilation available.
Thus they will provide the desired damage (i) For a 12.5 mm thick Gib® Ultraline system:
threshold from section 7.1, allowing extra time for 20 mins
manual suppression of the fire. If this doesn’t
occur, then fully developed fire spread over at least (ii) For a 12.5 mm thick Gib®Fyreline system:
the firecell of origin is likely and the damage to the 27 mins
building (especially the non-structural damage) will
be severe. (iii) For a 16 mm thick Gib®Fyreline system:
40 mins
Selection of a suitable barrier system is covered in
section 7.3.2 for a suspended ceiling and section (iv) For a 2 x 16 mm thick Gib®Fyreline system:
7.3.3 for a fixed ceiling. Only barrier systems 70+mins
capable of meeting the 20 minute standard fire test
survival time are given. The material is taken from Note that standard Gib®Board is not suitable.
section 5.12 of [12].
The board must be installed in accordance with the
7.3.2 Use of a non-fire-rated suspended ceiling appropriate Gib®Board specification for that
as a radiation barrier product.

To achieve the desired 20 minute standard fire test 8. Revised SPM Program
survival time, the suspended ceiling must have the
following general characteristics; 8.1 New features
(1) Any support grid members directly exposed
The SPM Program has been upgraded to
to the fire are made from steel and not
implement the second edition of the procedure.
aluminium The new version is SPM0103. This upgrade has
(2) The layout of the support grid is important, also included additions to enhance the ease of
especially the distance between the simply use. An example of the input screen for this
supported cross-runners which span
program is shown in Fig. 71.21 and the calculation
between the continuous main runners. This
screen in Fig. 71.22.
distance should not exceed 600 mm.
Longer tile life in place is obtained by
reducing this span further to 400 mm.

HERA Steel Design & Construction Bulletin Page 10 No. 71, December 02/ January 03
Features of the SPM0103 version are as follows: The cover to the lower bar will be a mesh bar
diameter greater than that to the upper bar.
(1) The slab type options available are now:
The cover for the mesh is specified to the top of
• Trapezoidal (Hi-bond) the concrete, in accordance with the NZS 3404
• Flat slab (Secondary beams) convention for specifying mesh cover as top cover.
• Flat slab (Speedfloor)
• Flat slab (Traydec) Unless the designer specifies which way around
the mesh is oriented, both possibilities should be
The Trapezoidal (Hi-bond) option should be checked. The orientation giving the lowest value
chosen for any trapezoidal deck profile. The of wu is used in design. The difference will
rib height must be entered; this is 54 mm be small when the mesh bar diameter is 5.3
for Hi-bond [30]. It incorporates the thermal or 7.5 mm. However, if larger bar diameters are
insulation effect of the decking on the used, then the temperatures in the lower layer will
concrete and reinforcement temperatures be appreciably hotter than those of the upper layer
within the slab that is given by equation and the orientation of the mesh can influence wu by
71.A6.3. This would be applicable to other as much as 5-10%. The greatest differences will
Trapezoidal decking profiles of the same arise in slab panels of high aspect (Ly/Lx) ratio
overall shape. The specification of heat where bars are used as mesh, with bar diameters
flow path dimensions c1, c2 and c3 for deck 10 or 12 mm.
trough bars for Hi-bond are, c1 = c2 = 80 mm
for c 3 = 40 mm 8.2.2 For the interior support bars

The flat slab (secondary beam) option is These bars run in the x-direction and sit on
used for flat slabs not cast on decking. top of the mesh, as specified by Notes 2 and 4 to
Fig. 71.12. This means that they sit directly on
The flat slab (Speedfloor) option is the mesh y-direction bars. Thus their depth within
specifically for flat slabs incorporating the slab is known, without the need to specify a
Speedfloor joists and implements the system cover to these bars.
- specific details of section A5.
8.2.3 For the deck trough bars
The flat slab (Traydec) option is for a flat
slab on steel deck and incorporates the These are positioned a specified vertical distance
insulation effect of the decking given by off the base of the deck ribs, as shown in
equation 71.A6.2 into the concrete and Fig. 71.28. The cover, c3, is specified off the
reinforcement temperature predictions. bottom of the rib for these bars, along with the
covers to the sides of the rib, c1 and c2, measured
(2) For grade 500 mesh or bar from [24], perpendicular to the rib wall.
use fy20 = 500 MPa.
For a bar positioned 40 mm off the bottom of
(3) In the input screen, values greyed out are a Hi-bond rib, c1 = c2 = 80 mm. These are
not required and hence made non recommended standard values to use, with c3
accessible for the option chosen. measured off the bottom of the rib, not the dovetail
return in the base of the rib.
(4) The FLED is used for either one or two
operations, as required: Designers can select different values for c3, in
which case they will have to determine the
• It assigns the Fire Hazard Category corresponding values for c1 and c2.
(FHC) for all applications, using the
relationship given in COMMENT 3 to 8.3 How to obtain the program
Para 2.2.1 of C/AS1 [5].
• If the option to calculate the time The SPM Program is available as a single
equivalent is selected, it is used in this executable file entitled SPM0103. It will operate in
calculation through equation 71.A2.1. all Windows systems from Windows 95 onwards.

8.2 Application of mesh and reinforcement Those wanting a copy, which will be sent via email,
covers should contact Charles Clifton at
structural@hera.org.nz.
8.2.1 For the mesh reinforcement
The program can be installed simply by double-
As shown in Fig. 71.21, the program requires the clicking on the icon & following the instructions. It
mesh bar cover to the top of the concrete to be will be sent with a sample calculation file, being the
specified for the bars in the x and y directions. design example presented in the next section.

HERA Steel Design & Construction Bulletin Page 11 No. 71, December 02/ January 03
9 Design Example All other required input data is given in Fig. 71.24.

9.1 General description and scope of 9.3 Design adequacy of slab panel
example
This is shown in Fig. 71.25. In this instance, the
The floor system used in this design example is slab panel has satisfactory moment/tensile
shown in Fig. 71.23. It is from an actual 17 storey membrane capacity (ratio of w*/wu = 0.80) and
building, previously used in our fire research satisfactory shear capacity (ratio of v*/vu = 0.55).
programme, more details of which are given in
DCB Issue No. 48. It is also the same basic The reinforcement level for integrity is also
example used in the first edition of the SPM, sufficient, this check is to section A4.3 herein.
presented in section 8, pp. 9-14 of DCB No. 60.
9.4 Design of two supporting beams for the
However, some changes have been made as a
fire emergency condition
result of the SPM procedure revision and the new
reinforcement standard [24]. This section covers the design of two of the slab
panel supporting beams. One is a primary interior
The building was built in 1988 using Grade 250
beam, the other a secondary edge beam. The first
steel for the floor support beams. For the
beam illustrates application of section A6.2
purposes of this example, the floor beams have
been re-sized, using Grade 300 members, to give step 3 option 3.1 – ie. application of passive
smaller member sizes likely to be used in a current fire protection. The second beam illustrates
application of section A6.2 step 3 option 3.2
design. The dimensions, layout of the structural
– ie. demonstrating an adequate reserve of
systems (seismic and gravity) have not been
strength without passive protection, using an
altered.
elevated temperature capacity check.
The ratio of (Av/Af) for the firecell in this building in
9.4.1 Design of primary interior beam B3-C3
practice is 0.2; in this design example it has been
reduced to 0.125 to illustrate application to a This is a 530UB82 Grade 300 member, spanning
higher structural fire severity (teq). 8.2 metres. See Fig. 71.23. It is connected to the
The scope of this design example is as follows: columns with WP (simple) connections, eg. to [6].
(1) Fire emergency design load, simply
• Input data is given in section 9.2 supported moment (section A6.2, step 1)
• Checking the design adequacy of the slab
panel is covered in section 9.3 wu,pib = 0.5 x 4.57 (8.9 + 8.9) + 0.8
• Design of two of the supporting beams for = 41.5 kN/m
the fire emergency condition is given in (the 0.8 is the beam’s selfweight)
section 9.4
• Design of the columns for the fire w uL2
emergency condition is given in section 9.5 M ∗ss = = 349 kNm
8
• The overall outcome is summarised in
section 9.6 (2) Determination of rf (section A6.2, step 2(1))
The design example uses the procedure given in ∗
Mss 349
Appendix A, as implemented by the SPM0103 rf = = = 0.38
program. φfire Mpos 930

9.2 Input data φMsx 558


Mpos ≈ 1.5 = 1.5 x = 930 kNm
ö 0.9
The slab panel is shown in Fig. 71.23, being the
area of floor bounded by gridlines A, 3, D and 5.
(φMsx is obtained from published tables, eg.
The other slab panel (bounded by grids A, 1, D
[20]. The 1.5 factor takes account of
and 3) will have the same solution.
composite action (see DCB Issue No. 2)).
The applied load, w*, comprises:
φfire = 1.0 (NZS 3404 Amendment No. 1)
• Slab dead load, including ponding = 2.46 kPa
• Superimposed dead load = 0.5 kPa (3) Determination of Tl (section A6.2, step 2 (3))
• Self-weight of secondary beams = 0.21 kPa
• Long term live load = 0.4 x 3.5 = 1.4 kPa Tl = 905 – 690r f = 646oC
• Total load, w* = 4.57 kPa As Tl > 550oC, any of the passive fire
In the calculation of time equivalent, the thermal protection manufacturer’s published design
inertia is determined from Table 5.1 of [7] as data can be used directly for determining the
b = 1700 J/m2Cs0.5.

HERA Steel Design & Construction Bulletin Page 12 No. 71, December 02/ January 03
product thickness required, as these will ∗
M ss = 0.1283 WL + wc1 L2/8 = 42 kNm
have been derived for Tl ≥ 550oC.
W = 27 x 4.45/2 = 56.6 kN
(4) Selection of passive fire protection product
(2) Determination of rf (section A6.2, step 2 (2))
Either a spray or board system should be
used. Given that beams forming part of a
M∗ss 42
perimeter type seismic-resisting system are =
unlikely to need passive protection in this φfire Mpos + φfire Mneg 1.0 x 620 + 1.0 x 620
building (this is confirmed for the MRF = 0.034
beams in section 9.4.2), the number of 558
beams on each floor requiring passive Mneg = Msx (rigid connections) =
0.9
protection is small. Therefore a board
= 620 kNm
system will be selected in this instance.
Mpos (conservatively take as non-composite)
Gib® Fireboard [35] is chosen in this
= Msx = 620 kNm
instance. It is one of a range of board
systems available; see details of this range (3) Determination of Tl (section A6.2, step
in HERA Report R4-89 [13]. 2 (3))
(5) Determination of thickness, number of layers Tl = 905 – 690 rf = 882oC
required.
As the limiting temperature is very high, use
Hp/A for 530UB82, top flange shielded, box step 3 option 3.2 to perform an elevated
protection = 15.3 x 7.85 = 120m-1 temperature moment and shear capacity
(This is obtained from eg. [20]) check on the beam.
From Gib® Fireboard catalogue [35]; (4) Determine the design elevated temperature
1 layer 20mm OK for 60 mins up to Hp/A of the beam and connection components
= 260m-1 (section A6.2, step 3.2.2).
1 layer 20mm OK for 90 mins up to Hp/A
= 112m-1 These are:
FRR = 1.0teq = 68 mins (see section 8.2.2) • For beam bottom flange and web
850 – 50 = 800oC
Limiting Hp/A for 68 mins • For beam top flange, 750 – 50 = 700oC
 90 − 68  • Connection components are not critical in
=   (260 − 112) + 112 = 221m
-1

 90 − 60  this instance, as the bolts and endplate


are within the protected region of the
Hp/Aactual = 120 < 221 √ OK column

Use 1 layer 20mm Gib® Fireboard on this (5) Determine the design elevated temperature
beam. moment capacity of the beam (section A6.2,
step 3.2.3)
9.4.2 Design of secondary edge beam A4-A5
fyθ/fy20, for 800oC = 0.110
This is part of the moment-resisting frame seismic- (see Tables 71.3 and 71.4)
resisting system along gridline A. See Fig. 71.23.
The member is a 530UB82 Grade 300, spanning Mneg, θ = Mpos, θ = Ms x 0.110 = 68.2 kNm
4.45m, with MEP Category 3 connections to [6] at (base on bottom flange temperature)
each end. (The size is that used in the original
design, however the grade, category and (5) Determine moment capacity through a
connection type have been changed to what would plastic collapse mechanism check (section
be more typically used under current design A6.2, step 3.2.4)
practice).
(φfire Mposθ + (φfireMneg, θ)
= 68.2 + 68.2
(1) Fire emergency design load, simply = 136.4 kNm
supported moment. M ∗ss = 42 kNm < 136.4 kN, √ OK
As seen in Fig. 71.23, the load from the slab Beam has adequate moment capacity
panel is triangular, varying from zero at end
A5 to a maximum value of L1 w* = 5.9 x 4.57 (7) Check the elevated temperature shear
= 27 kN at end A4. L1 comes from Fig. capacity (section A6.2, step 3.2.5).
71.25 and is 5.9 m. There is also a line load Beams have rigid end connections; OK
of 3 kN/m from the cladding.

HERA Steel Design & Construction Bulletin Page 13 No. 71, December 02/ January 03
Conclusion: 10. Guidance on Cost-Effective Application
Beam A4-A5 does not require passive fire of the Slab Panel Method
protection; it fulfills its slab panel support function
unprotected, by virtue of its high reserve of Design the floor and structural system for the
strength. gravity and lateral loading conditions. This
determines beam size, spacing, slab depth,
9.5 Design of columns for the fire emergency concrete strength, etc.
condition
Then commence the slab panel design for fire.
Minimum column size is W14 x 176
Step 1: Select the size of slab panel in
All columns satisfy NZS 3404 Table 12.5 Category accordance with section A2.2 on page
3 section geometry limit, therefore: 37. It is advantageous to keep Ly/Lx as
close to 1 as practicable.
Tl = 550oC (see section 4.3 herein)
FRR = 1.0 teq = 68 mins Step 2: Determine which of sides 1 and 3 can
W14 x 176 dimensions are (all in mm or mm2) carry negative moment.
d = 368 ; b = 372; tf = 23.8 ; tw = 14.4 ; A = 22,600
Step 3: Start with the following reinforcement
2b + 2d contents:
Hp/A, for four sided box protection, =
A
(1) Mesh : 665, comprising 5.3 mm
= 65m-1
diameter bars at 150 mm centres,
From [35], for FRR = 68 mins, 1 layer 20mm Gib cold-formed mesh to grade 500L,
Fireboard is satisfactory for all columns. fyr20 = 500 MPa

The columns must be protected full height. (2) Interior support bars : DH10 at 250
mm centres, hot-rolled bars, grade
9.6 Overall outcome 500E, fyr20 = 500 MPa
For the floor system shown in Fig. 71.23, the (3) No deck trough bars
requirements of this slab panel method for fire
emergency conditions are; Step 4: Input all other variables and do the first
check on the load-carrying capacity of
(1) For slab reinforcement the slab panel. If this is satisfied, check
shear capacity. If this is also satisfied,
• Mesh : 665 Large; 5.3 mm diameter
the design is complete.
bars at 150 centres each way
• Interior support bars : DH12 at 500 If the design load-carrying capacity is not
centres satisfied, then:
• Deck trough bars : 1 DH16/trough
• Edge and trimmer bars: as shown in (1) Add deck trough bars, until either
Fig. 71.12. All bars DH10; spacing of my = mx or the load-carrying capacity
edge bars is 250 mm centres is adequate. If it needs to be further
increased, then;
(2) Floor system beams, etc
(2) Increase the mesh area as required
• Sizes and layouts as shown in until the load-carrying capacity is
Fig. 71.23 and as determined for adequate.
ambient temperature design
• All beams are either composite or with If the load-carrying capacity is still not adequate
shear studs at not more than 480 mm and Ly/Lx ≥ 1.25, then designate one of the
centres(4to) unprotected secondary beams as a protected
support beam, in accordance with section A2.2,
(3) Extent of passive fire protection and repeat the above calculations.

• All columns If the load-carrying capacity is still not adequate


• Primary interior beams A3-B3, B3-C3, and Ly/Lx < 1.25, protect the secondary beams in
C3-D3 accordance with the C/AS1 [5] requirements.
• Primary edge beams A5-B5, C5-D5,
A1-B1 (in practice, beam C1−D1 is in Remember that the detailing requirements given in,
the stairwell region, with negligible fire or referenced from, section 5 must be applied.
load and part of a separate firecell) See Fig. 71.12 and the accompanying notes for
the detailing requirements to the slab panel itself.

HERA Steel Design & Construction Bulletin Page 14 No. 71, December 02/ January 03
11. Acknowledgments (3) Linus Lim, under the supervision of
Professor Andy Buchanan and Dr. Peter
The HERA Structural Engineer, principal developer Moss, University of Canterbury, for the Slab
of the SPM design procedure, would like to Panel fire tests and analyses contained in
acknowledge the contribution of all [3, 36].
persons/organisations involved in the development
of this procedure, with special mention of: (4) Professor Colin Bailey Professor of
Structural Engineering, UMIST, UK and Jef
(1) Robert Schmid and Christopher Beck, the Robinson, Honorary Professor of
two undergraduate students from Germany Construction Marketing at the University of
who have undertaken much of the analysis Sheffield, UK.
work and development of the SPM
programs. (5) The Foundation for Research, Science and
Technology, for providing past and on-going
(2) The UK BRE and Corus, for supplying the funding of HERA’s fire research programme,
fire test and slab panel load test data which has led to the development of the first
contained in [9, 17]. and second editions of the SPM design
procedure.

Fig. 71. 1 (from [21])


Reflected Floor Plan for Application of Slab Panel Fire Engineering Design Procedure to a
Concrete Slab on Profiled Steel Deck Supported on Primary and Secondary Beams

Notes to Fig. 71.1


• The beam positions shown are the centrelines.
• The exterior of the building shown is the edge of the concrete slab
• A two slab panel floor system is shown; the concept is applicable to larger floor plan areas.
• Lateral load-resisting systems are not explicitly shown, but their position is not restricted by the method
• The secondary interior beams are unprotected. The columns will have passive fire protection. The primary interior beams
and the edge beams may be unprotected if they have a suitably high reserve of strength; for example through being part of a
seismic resisting system.

HERA Steel Design & Construction Bulletin Page 15 No. 71, December 02/ January 03
Fig 71.2
Slab Panel Ambient Temperature Test; General Specimen Set-up (from [9])

Fig 71.3
Slab Panel Ambient Temperature Test; Position of Applied Loads and Pattern of Cracking From Yieldline
Development and From Tensile Membrane Fracture (from [9])

HERA Steel Design & Construction Bulletin Page 16 No. 71, December 02/ January 03
Fig 71.4
Slab Panel Fire Tests; Section Across the Short Span of the Furnace (from [3])

Fig 71.5
Slab Panel Fire Tests; Position of Applied Loads (from [3])

HERA Steel Design & Construction Bulletin Page 17 No. 71, December 02/ January 03
Fig 71.6
Slab Panel Fire Tests; Slab in Position for Test

Fig 71.7
Top Surface Crack Pattern of the D147 Flat Slab (from [36])

HERA Steel Design & Construction Bulletin Page 18 No. 71, December 02/ January 03
Fig 71.8
Bottom Surface Crack Pattern of the D147 Flat Slab (from [36])

Fig 71.9
Finite Element Discretization of a Quarter of the Slab Panel (from [36])

Fig 71.10
Deflected Shape of the D147 Slab Panel at 3 Hours (Magnified 1.5 times) (from [36])

HERA Steel Design & Construction Bulletin Page 19 No. 71, December 02/ January 03
Fig 71.11
Unprotected Secondary Beam to Protected Primary Beam Connection, Showing Extent of Passive Protection
Required over the Connection Region.

HERA Steel Design & Construction Bulletin Page 20 No. 71, December 02/ January 03
Fig 71.12
Floor Plan Showing Additional Reinforcement Required in the Slab Panels.
Notes to Fig. 71.12
1. Edge and trimmer bar reinforcement is to be deformed grade 500E or 500N to AS/NZS 4671 [24]. Interior support
bar reinforcement is to be grade 500E. Mesh reinforcement is to be either deformed or plain grade 500L (or grade
500N or 500E can be used).
2. The edge bars, trimmer bars and interior support bars are required in all instances. They are placed on top of the
slab mesh.
3. The area of the edge bars/m width in each direction (x or y) is to be not less than Amesh required in that direction.
4. The area of the interior support bars, which are in the x direction, is to be not less than Axmesh .
5. Anchorage of edge bars is to be by standard hook, to NZS 3101 Clause 7.3.14, around the trimmer bar. See Figs.
71.14 and 71.15.
6. The length of the edge bars into the slab panel, measured from the end of the hook, is to be 750 mm
7. Edge bars spacing to be at 250mm centres maximum and typical; less if required to satisfy the minimum area
requirement from note 3 for the bar diameter chosen
8. Interior support bars to extend 0.15Lx plus the bar development length past the centreline of the supporting beam
into each slab panel; spacing at 250mm centres maximum and typical. Size is typically 10 mm or 12 mm as
required.
9. The deck trough bars are optional and are used if necessary to increase the load-carrying capacity of the slab panel.
They are placed in the troughs as shown in Figs. 71.13 and 71.14. Size is 10mm to 16mm as required.
10. The centreline position of the primary interior beams is shown in this view, because the placing of the interior
support bars is based around this position. The rest of the floor support beams are not shown herein.
11. The trimmer bars are placed outside the centre-line of the shear studs relative to the slab panel.

HERA Steel Design & Construction Bulletin Page 21 No. 71, December 02/ January 03
Fig 71.13
Section A – A Through Slab Panels (Incorporating Steel Decking) at the Primary Interior Beam

Note: See Fig. 71.17 for location of this section in plan

Fig 71.14
Section B – B Through Slab Panel 1 (Incorporating Steel Decking) at the Primary Edge Beam

Note: See Fig. 71.17 for location of this section in plan

HERA Steel Design & Construction Bulletin Page 22 No. 71, December 02/ January 03
Fig 71.15
Section C – C Through Slab Panel 1 (Incorporating Steel Decking) at the Secondary Edge Beam

Note: See Fig. 71.17 for location of this section in plan

Fig 71.16
Section D – D Through Slab Panels (Incorporating Speedfloor Joists) at the Primary Interior Beam

Note: See Fig. 71.18 for location of this section in plan

HERA Steel Design & Construction Bulletin Page 23 No. 71, December 02/ January 03
Fig 71.17
Reflected Floor Plan for Application of Slab Panel Fire Engineering Design Procedure to a Concrete Slab on
Profiled Steel Deck Supported on Primary and Secondary Beams

Notes to Fig. 71.17


1. The beam positions shown are the centrelines.
2. The exterior of the building shown is the edge of the concrete slab.
3. A two slab panel floor system is shown; the concept is applicable to larger floor plan areas.
4. Lateral load-resisting systems are not explicitly shown, but their position is not restricted by the method.
5. The secondary interior beams are unprotected. The columns will have passive fire protection. The primary interior beams and the
edge beams may be unprotected if they have a suitably high reserve of strength; see details in section A6.2 step 3 option 3.2

HERA Steel Design & Construction Bulletin Page 24 No. 71, December 02/ January 03
Fig 71.18
Reflected Floor Plan for Application of Slab Panel Fire Engineering Design Procedures to a Concrete Slab
Supported on Speedfloor Joists and Primary Beams

Notes to Fig. 71.18


1. The beam and joist positions shown are the centrelines.
2. The exterior of the building shown is the edge of the concrete slab.
3. A two slab panel floor system is shown; the concept is applicable to larger floor plan areas.
4. Lateral load-resisting systems are not explicitly shown, but their position is not restricted by this method.
5. The Speedfloor Joists are unprotected. The columns will have passive fire protection. The primary interior beams will typically have
passive fire protection. The edge beam protection may be through radiation barriers where the beam design limiting temperature is
not exceeded. Alternatively, the primary interior beams and the edge beams may be unprotected if they have suitably high reserve
of strength; see details in section A6.2 step 3 option 3.2

HERA Steel Design & Construction Bulletin Page 25 No. 71, December 02/ January 03
Protection to
beam (spray
or board)
Normal
deflection
head

Compartment
wall
Fig 71.19
Fire Separating Walls Below Slab Panel Edge Support Beams (detail from [37])

Deformable blanket

Compartment wall
Fig 71.20
Fire Separating Walls Within Slab Panel Region (detail from [37])

HERA Steel Design & Construction Bulletin Page 26 No. 71, December 02/ January 03
Fig 71.21
Input Screen for the SPM0103 Program for the Speedfloor Fire Test Slab Panel

Fig 71.22
Calculation Screen for the SPM0103 Program for the Speedfloor Fire Test Slab Panel

HERA Steel Design & Construction Bulletin Page 27 No. 71, December 02/ January 03
Fig 71.23
Reflected Floor Plan of Telecom Towers Building, Showing the Slab Panel Used in the
Design Example of Section 9.

Note: The position of gridlines B and C are not drawn to scale in the horizontal dimension; the actual dimensions are as stated.

HERA Steel Design & Construction Bulletin Page 28 No. 71, December 02/ January 03
Fig 71.24
Input Screen for the SPM0103 Program for the Design Example in Section 9.

Fig 71.25
Calculation Screen for the SPM0103 Program for the Design Example in Section 9.

HERA Steel Design & Construction Bulletin Page 29 No. 71, December 02/ January 03
Fig 71.26
Reflected Floor Plan Showing Dimensions for Slab Panel Yield Line Pattern and Developed Moments

Notes to Fig. 71.26


1. The y-direction is parallel to the span of the decking; the x-direction is parallel to the span of the secondary beams.
2. Dependable negative moment resistance is able to be developed only over the primary interior beams and only through the use of
additional reinforcing (interior support) bars. For slab panel 1 shown above, the resistance is available only from side 3.

HERA Steel Design & Construction Bulletin Page 30 No. 71, December 02/ January 03
c ymesh
c xmesh
Mesh in y-direction
Mesh in x-direction

u1 u2 d mesh
Concrete
u3 t0
25 25

h rc

Fig 71.27
Position of Mesh and Heat Flow Paths for Mesh Temperature Determination
Note: For flat slabs, the dimensions u1, u 2 and u3 are measured from the soffit of the slab.

c1 c2
c4 c3 c5

Fig 71.28
Cross – Section Through Dimond Hi-bond Slab Showing Position of Deck Trough Bar and Heat Flow Paths
to this Bar (Dimensions in mm).

Note: The dovetail return in the base of the Hi-bond rib is not shown herein, but its effect is incorporated into the temperature
determination. The cover c 3 is measured to the bottom of the rib, not the top of the dovetail return.

θ θ

Fire exposed face

Fig 71.29
Resultant Compression Force in the Concrete Compression Zone Under Negative Moment Action.

HERA Steel Design & Construction Bulletin Page 31 No. 71, December 02/ January 03
Fig 71.30
Section E – E Through Region of Floor Slab Incorporating Steel Decking, Showing Development of Positive
Moment Capacity Using the Mesh Reinforcement and Secondary Beams

Notes to Fig. 71.30


1. See Fig. 71.26 for location of this section, which is taken across the x-direction looking along the y-direction.
2. The optional deck trough bars are not shown in this section.

HERA Steel Design & Construction Bulletin Page 32 No. 71, December 02/ January 03
Fig 71.31
Uniformly Loaded Rectangular Slab With All Edges Supported (from [19])

Notes to Fig. 71.31


_______ = slab positive yieldline moment
- - - - - - = slab negative yieldline moment in the x-direction (where applicable)
mx = positive moment in the x-direction per unit width along the y axis
my = positive moment in the y-direction per unit width along the x axis
mx´ = negative moment in the x-direction per unit width along the y axis
Side 1 = length AD, etc.
F = fixed edge supported (side 1 or 3, where applicable)
S = pinned edge support (sides 2 and 4, always)

Fig 71.32
Location of Critical Cross – Section for Shear Capacity Determination, Slab Panel Supported on Unprotected
Secondary Beams.

HERA Steel Design & Construction Bulletin Page 33 No. 71, December 02/ January 03
Fig 71.33
Application of Design Procedure to Concrete Slabs Supported on Speedfloor Joists and Primary Beams

Notes to Fig. 71.33


1. This figure incorporates the information presented in Fig. 71.18 with dimensions and moments required for application of the slab
panel fire engineering design procedure.
2. Tributary areas for shear capacity checks are also shown.

Bottom of slab

Fig. 71.34
Fig. 71.34
Cross – Section Through the Speedfloor Joist for Determining Joist Contribution to Slab Panel Yieldline
Capacity

HERA Steel Design & Construction Bulletin Page 34 No. 71, December 02/ January 03
Fig 71.35
Speedfloor Slab Panel Following Furnace Test Showing Buckled State of the Joists
(Especially the Central Joist).

Fig. 71.36
Plan View of Cardington Demonstration Furniture Test Fire Showing Gas Temperature Contours at Time of
Peak Temperatures
Note: The dotted lines are the floor support beams; horizontal are secondary beams, vertical primary beams

HERA Steel Design & Construction Bulletin Page 35 No. 71, December 02/ January 03
Appendix A to DCB No. 71

Design Procedure for the Slab Panel Method, Second Edition

A1 Introduction and Scope

This appendix presents the second edition of the Slab Panel Method (SPM) design procedure and
detailing requirements in step by step format. The layout used is as close as practicable to that used
for the first edition, presented in DCB No. 60 in 2001.
It commences, in section A2, with the definitions of critical components, variables and notation,
including defining the slab panel floor area to which the procedure is applied.
Section A3 covers the design loads and design structural fire severity required for implementation of
the procedure - i.e. the input design actions.

The procedure is presented in sections A4 and A5. Section A4 covers use of the procedure for a
concrete slab on profiled steel decking supported on a network of primary and secondary composite
steel beams. Section A5 covers use for a concrete slab on Speedfloor Joists, supported by primary
beams.
Section A6 covers design of the supporting beams for the edges of the slab panel. These beams may
require protection from direct fire exposure. Section A7 covers design of the supporting columns,
which will require protection from direct fire exposure. The figures used in this appendix are included
immediately prior to the appendix.

For each section, and where appropriate, a commentary to the design procedure provisions is given.
The commentary is presented following sections A1 - A7; not all sections have a commentary. The
appendix and commentary relate only to the detailed procedure itself; for a general introduction to the
procedure, overview of its development and related aspects, refer to the details proceeding this
appendix.

A2 Definitions of Critical Components, Variables and Notation

A2.1 Definition of slab panel


For ambient temperature design, the beams support the floor. Under severe fire conditions, the
unprotected beams lose their load-carrying capacity and the floor slab system ends up supporting the
beams. This support is provided by a region of floor slab acting in two way yieldline/tensile membrane
action, transferring the applied loads back to the supporting effectively rigid members.
The unit of floor slab system that develops this capacity to resist the fire emergency design loads is
termed the slab panel.
A slab panel is defined as the area of floor slab and unprotected beams or joists that spans, under
two-way action, between beams that remain effectively undistorted under severe fire conditions
relative to the peak downwards deflection expected within the slab panel region. The slab panel has
dimensions Lx and Ly - i.e. length in the x - direction and length in the y - direction, e.g. as shown in
Figs. 71.26 or 71.33.
Supports which remain largely undistorted under severe fire conditions are:
• Beams which are shielded from direct exposure to the fire in order to limit their temperature in
accordance with NZS 3404 [10] Clause 11.5. These include primary interior and edge beams,
secondary edge beams and any secondary interior beams which are protected to limit the
length of the slab panel (see section A2.2)
• Beams which form part of the lateral load-resisting system and which are required to be
protected or have a very high reserve of strength under fire emergency conditions (see e.g.
section A6.2 Step 3.2 herein)
• Supports formed by a separate firecell and hence which will remain relatively cold at the time
the slab panel region is over 600oC, such as those around a lift core or stair well, or an adjacent
apartment in a series of apartments on a floor in a building.

HERA Steel Design & Construction Bulletin Page 36 No. 71, December 02/ January 03
A2.2 Determining the size of the slab panel
These are the dimensions Lx and Ly (see Figs. 71.26 or 71.33).
The length Lx is the distance between adjacent primary beams.
The length Ly is the distance between adjacent points of effective slab panel edge support, such as the
edges of the building (see Figs. 71.26 or 71.33), or between a building edge and an internal core. Its
determination will be clear cut in most instances, but will be more a matter of designer judgement in
others. For example, in application to an apartment building floor containing multiple apartments, Ly
can be conservatively taken as the length of two adjacent apartments.

A lower limit on Ly is Ly = Lx (i.e. a square panel).


An upper limit of Ly = 30 m is recommended for an open plan firecell.
A further recommendation is that Ly/Lx ≤ 2.5, however this is a recommendation only.

One support for the slab panel, for determining Ly, will always be available - e.g. at the edge of the
building. If a second support is not obvious, due to the building layout, within a length of 2.5Lx, then
this support can be provided by choosing a suitably positioned secondary interior beam and designing
and protecting it accordingly.
This secondary interior support beam will then split the floor system into two adjacent slab panels and
it will support a triangular load from both slab panels under severe fire conditions. This loading could
be considerably more than the design load for ambient temperature conditions, requiring the beam
size to be increased for the fire emergency design case, in the manner as is documented in section
A6.2 step 1 for the secondary edge beams. Because of this, a secondary interior beam used as a
slab panel support, to reduce Ly, should always be a beam spanning directly onto columns at each
end, rather than onto primary beams at either or both ends.
The bars shown in Fig. 71.12 would also need to be provided over this secondary interior support
beam, but spanning only 750 mm into each slab panel. They would be placed into each slab panel, in
the same way as interior support bars, in this instance.

A2.3 Notation used


A2.3.1 Normal font notation

∆ = Deflection of slab
θ = gas temperature
φ = strength reduction factor
γ = Coefficient for heat transfer to the reinforcement
θf = bolt temperature
φfire = strength reduction factor for fire, from NZS 3404 Amendment No. 1 Clause 11.5
θs = steel temperature
A = area
a = depth of concrete compression stress block
Af = floor area of firecell
Ar = area of reinforcement
Arx = area of reinforcement in x-direction
Ary = area of reinforcement in y-direction
Asb = area of secondary beam element (top flange, bottom flange or web)
Av = area of vertical openings (fixed plus variable) available to an enclosure
b = width of slab
bf = width of beam flange
Csθ = Coefficient used in the calculation of the reinforcement temperature
c = cover to reinforcement, from top or bottom of concrete as specified
d = depth of beam
d1 = clear depth of beam between flanges
dv = effective slab depth for shear capacity
E = modulus of elasticity
e = depth from centroid of tension force to top surface of concrete slab in compression
ec = depth from concrete surface under maximum compression to centroid of concrete in
compression

HERA Steel Design & Construction Bulletin Page 37 No. 71, December 02/ January 03
er = depth from centroid of bar to top of concrete
erx = depth from centroid of tension action to top of concrete, for positive moment in the x-direction
ery = depth from centroid of tension action to top of concrete, for positive moment in the y-direction
fc′ = nominal (specified) cylinder compression strength of concrete (typically 28 day) (MPa)
FHC = fire hazard category
FLED = fire load energy density (MJ/m² floor area)
FRR = fire resistance rating
fu = nominal tensile strength of steel, bolt, weld etc.
fy = nominal yield stress of steel
G = dead load
H = height of enclosure
hrc = rib hight for trapezoidal profile
i = storey i or side i, as appropriate
kb = thermal conductivity coefficient
L = dimension of slab panel
M = moment (kNm)
M∗ = design moment
mx
' = negative moment in the x-direction per unit length along the y axis
Mrc = Composite beam positive moment capacity
mx = positive moment in the x-direction per unit length along the y axis
my = positive moment in the y-direction per unit length along the x axis
nb = number of bolts in secondary beam to primary beam connections
Qu = reduced live load (for fire emergency conditions)
Rcc = internal compression force from concrete
rf = reduction factor
Rts = internal tension force from steel
Rtsb = internal tension force from steel secondary beam element
Ru = nominal capacity
S∗ = design action
Ssb = spacing of secondary beams
teq = Equivalent standard fire test time for a reinforced concrete or an insulated steel component
tf = thickness of beam flange
Tl = limiting temperature of structural component
to = total slab thickness
tw = thickness of beam web
u1,u2,u3 = distances from centroid of reinforcement to fire exposed faces of slab
v* = design shear load per metre width at edge of floor slab panel
V* = design shear load over a defined length of floor slab panel
vc = nominal shear stress available from concrete
vc = nominal shear stress in concrete
Vfn = nominal shear capacity of bolt, threads included in shear plane
vu = design shear capacity of floor slab per metre width
Vu,sb-pb = design shear capacity of secondary beam to primary beam connection at elevated
temperature
Vv = nominal shear capacity of steel beam
∗ = design load on floor slab panel under fire emergency conditions, including selfweight
w
wf = ventilation factor
wu = ultimate load - carrying capacity of floor slab panel
wylθ = yieldline load carrying capacity of floor slab panel at elevated temperature

HERA Steel Design & Construction Bulletin Page 38 No. 71, December 02/ January 03
A2.3.2 Subscripts

θ = value at elevated temperature


1 = relating to slab panel 1
2 = relating to slab panel 2
20 = value at ambient temperature (assumed to be 20°C)
A–E = section A – E of the Speedfloor joist
bf = bottom flange (of secondary beam)
bolt = bolt
deck = profiled composite steel decking
dtb = deck trough bar
f = fastener or floor
isb = interior support bar
max = maximum
mesh = mesh reinforcement in slab
min = minimum
pb = primary beam
r = reinforcement
s = steel
sb = secondary beam
sfj = Speedfloor joist
slab = Slab
ss = simply supported (slab panel - i.e. no negative moment capacity on any side)
tf = top flange (of secondary beam)
w = weld; or web (of secondary beam)
x = value in x axis direction
y = value in y axis direction

A3 Design Loads and Determination of Structural Fire Severity

A3.1 Design loads


These are the vertical loads given by NZS 4203 [15] Clause 2.4.3.4, Load Combination (7) for fire
emergency conditions. This combination is repeated herein as equation 71.A1:

w* = G + Qu (71.A1)

where:
G = gravity dead load on slab panel, including the self-weight of slab and unprotected
beams/joists
Qu = ΨuQ = long-term live load from [15].

A3.2 Design structural fire severity

This is the time equivalent, teq, or a lower value if allowed by [4] and specified by the user.

t eq = FLED ⋅ kb ⋅ w f (71.A2.1)

where:
FLED = design fire load energy density (e.g. from COMMENT 3 to Para.2.2 of C/AS1 [5])
kb = thermal inertia factor for the bounding elements (from Table 5.1 of the Fire
Engineering Design Guide [7])
wf = ventilation factor
 6   Av  
0. 3 4

=   0 .62 + 90 0.4 −  > 0.5 (71.A2.2)
 H    Af  

H = inside hight of firecell
Av = area of vertical openings
Af = floor area of firecell
Av
= must be in the range of 0.025 to 0.25
Af

HERA Steel Design & Construction Bulletin Page 39 No. 71, December 02/ January 03
A4 Slab Panel Design; Floors Incorporating In-Situ Concrete Slabs on Profiled Steel Decking or
Flat Slabs Without Decking, Supported by Primary and Secondary Composite Steel Beams

A4.1 Temperatures of slab panel components and associated mechanical properties

A4.1.1 Mesh within the concrete slab

A4.1.1.1 X-direction mesh bars


The approach used is modified from section 6.4.2 of [8], on the basis, firstly, of section 7.3 of DCB
Issue No. 59 and, secondly, section 4 of DCB No. 70. To elaborate:

Step 1: Determine the design FRR

FRR, mesh = 0.9 teq (71. A3)

Step 2: Determine the distances (u1, u2 and u3)xmesh from the mesh to the top of the ribs, for trapezoidal
decking, or to the soffit of the slab for a flat slab or for flat slab decking.

This is undertaken in accordance with Fig. 71.27.

u3,xmesh =t0 – hrc – cxmesh – dmesh/ 2 (71.A4.1.1)

where:
to = total depth of the slab (mm)
hrc = rib height for trapezoidal profile (mm)
= 0 for decking profile giving a solid slab, or for a solid slab (no decking)
cxmesh = top cover to mesh bars in the x-direction (mm)
dmesh = diameter of bar (mm)

Having calculated u3x,mesh;

u1,xmesh = u 2,xmesh = u 3,xmesh2 + 25 2 (71.A4.1.2)

Step 3: Determine the heat transfer coefficient, γ

This is given by equations 6.2 and 6.3 of [8]. Equation 6.2 is repeated below as equation 71.A5.1. The
calculation uses the value of u1,xmesh, u2,xmesh and u3,xmesh from step 2.

1 1 1 1
= + + (71.A5.1)
γ u1 u2 u3

The distances are expressed in mm.

Step 4: Determine the design temperature of the mesh bars.

For slabs without decking, use the equation 71.A6.1, otherwise use the equations 71.A6.2 or 71.A6.3.

θs = Cs,θ - 350γ (71.A6.1)

Cs,θ has the following values:


• 900oC for FRR = 22 mins
• 1019ºC for FRR = 34 mins
• 1175oC for FRR = 60 mins
• 1285oC for FRR = 90 mins
• 1370oC for FRR = 120 mins
• 1455oC for FRR = 180 mins
• 1545ºC for FRR = 240 mins

HERA Steel Design & Construction Bulletin Page 40 No. 71, December 02/ January 03
For the Traydec [25] flat slab: (71.A6.2)

ès = (774 .1 − 191.7ã ) ln(1.125t eq) +


11090.8 − 2508ã
+ (0.2667ã − 0 .539 )teq + (663 .1ã − 2759 .9 )
teq

For the Dimond Hi-bond [30] trapezoidal slab: (71.A6.3)

ès = (844.3 − 178.7ã ) ln(1.125t eq) +


712.6ã + 7379
+ (0.507ã − 1 .189 )teq + (514.1ã − 2873.4 )
teq

where:
θs = temperature of the steel bar
γ = is determined from equation 71.A5.1
teq = time equivalent from equation 71.A2.1 or as specified by the user

Step 5: Determine the elevated temperature design yield stress in the mesh bars, fyrθ,xmesh.

This utilises the relationship given in EC 2 Part 1.2 [28]. The mesh can be made from either cold-
worked or hot-formed bar, with different expressions given for each. These relationships are changed
from those given in DCB No. 60, due to an error in the tables in EC2 [28] which assigns the wrong
reinforcement type to each set of design values.

The correct details are as follows:

For hot-formed mesh bars, the relationships between fyrθ and fyr20 are as follows;
fyrè
ès ≤ 400 o C, = 1.0 (71.A7.1)
fyr20
fyrè
ès = 500 o C, = 0.78 (71.A7.2)
fyr20
fyr è
ès = 600o C, = 0.47 (71.A7.3)
fyr20
fyr è
ès = 700 o C, = 0.23 (71.A7.4)
fyr20
fyrè
ès = 800 o C, = 0.11 (71.A7.5)
fyr20

Linear interpolation is used between these points.

These values are appropriate to mesh comprising grade E or grade N bars to AS/NZS 4671 [24].

Deformed bars, when used as mesh, interior support bars or deck trough bars, use the hot-formed
relationship.

For cold-worked welded mesh, i.e. mesh bars made to grade L of AS/NZS 4671 [24], the relationships
between fyrθ and fyr20 are as follows:

fyrè
ès ≤ 350 o C, = 1.0 (71.A7.6)
fyr20
fyrè  6650 - 9ès 
350o C < ès ≤ 700 o C, =  (71.A7.7)
fyr20  3500 

HERA Steel Design & Construction Bulletin Page 41 No. 71, December 02/ January 03
fyrè  1200 - ès 
700 o C < ès ≤ 1200 o C, =  (71.A7.8)
fyr20  5000 

At the end of step 5 for the x-direction, the elevated temperature design yield stress in the x-direction
mesh bars, fyrθ,xmesh, is known.

A4.1.1.2 Y-direction mesh bars

The same five step approach and requirements as for the mesh in x-direction are used, with
appropriate distances (u1, u2, and u3)ymesh . Details are as follows:

Step1: Determine the design FRR.

FRR, mesh = 0.9teq (71.A3)

This is the same equation as for the x-direction mesh bars.

Step 2: Determine the distances (u1, u2, and u3)ymesh from the mesh to the top of the ribs, for
trapezoidal decking, or to the soffit of the slab for flat slab decking, or for slabs without decking.

This is undertaken in accordance with Fig. 71.27.

u3,ymesh =t0 – hrc – cymesh – dmesh/ 2 (71.A4.2.1)

where:
to = total depth of the slab (mm)
hrc = rib height for trapezoidal slab (mm)
= 0 for decking profile giving a solid slab, or for a solid slab (no decking)
cymesh = top cover to mesh bars in the y-direction (mm)
dmesh = diameter of bar (mm)

Having calculated u3,ymesh;

u1,ymesh = u 2,ymesh = u3,ymesh2 + 25 2 (71.A4.2.2)

Steps 3-5: as for steps 3-5, section A4.1.1.1, using the value of FRR,mesh and (u1, u2 and u3)ymesh from
steps 1 and 2 above and equations 71.A7.1 to 71.A7.8 for fyrθ / fyr20.

At the end of step 5 for the y-direction, the elevated temperature design yield stress in the y-direction
mesh bars, fyrθ,ymesh, is known.

A4.1.2 Interior support bars

The same five step approach and requirements as for the mesh are used, with appropriate distances
u1, u2, u3 from the fire-exposed top of the rib. Details are as follows:

Step 1: Determine the design FRR.

FRR,isb = 0.9teq (71. A3)

This is the same equation as for the mesh in the x and y direction.

Step 2: Determine the distances u1, u2, u3 from the interior support bars down to the top of the ribs, for
trapezoidal decking, or down to the soffit of the slab for flat slab decking.

disb
u3,isb = t0 − hrc − cymesh + (71.A8.1)
2

HERA Steel Design & Construction Bulletin Page 42 No. 71, December 02/ January 03
where:
t0 = total depth of the slab (mm)
hrc = rib height for trapezoidal slab (mm)
cymesh = top cover to mesh bars in y-direction (mm)
disb/2 = diameter of the interior support bar (10mm to 16 mm diameter, as required)

u1,isb = u 2,isb = u3,isb2 + 25 2 (71.A8.2)

Steps 3-5: as for steps 3-5, section A4.1.1.1, using the value of FRR,isb and (u1, u2, u3)isb from steps 1
and 2 above and equations 71.A7.1 to 71.A7.5 for fyrθ / f yr20.

At the end of step 5 for the interior support bars, the elevated temperature design yield stress in these
bars, fyrθ,isb, is known.

A4.1.3 Deck trough bars

The deck trough bars (dtbs) comprise 1 bar per deck trough of a profiled steel deck, placed typically in
every trough, when used. They are optional to the design and are only used where necessary to
increase wu for the slab. They are placed most commonly at 40 mm cover off the deck trough, as
shown in Fig. 71.28. Their size will typically be 10 mm to 16 mm diameter.

Step 1: Determine the design FRR

FRR,dtb = 0.9teq (71. A3)

For deck trough bars in Dimond Hi-bond, steps 2 – 5 are as follows:

Step 2: Determine the distances (u1, u2, u4, u5)dtb.

• c3 = 40 mm, or as specified by the user (see Fig 71.28)


• u4 = u5 = (c3 + 0.5 ⋅ ddtb )2 + 25 2
• c 1 = c2 = 80 mm, provided that the bar is placed in the centre of the trough and c3 = 40mm.

If c3 ≠ 40mm, then c1 and c2 must also be specified by the user.

• u1 = c1 + 0.5 ddtb
• u2 = c2 + 0.5 ddtb

Step 3: Determine the heat transfer coefficient, γ

1 1 1 1 1
= + + +
ã u1 u2 u4 u5

Step 4,5: as for step 4,5, section A4.1.1.1, using the value of FRR,dtb from step 1 above, γ from step 2
above, θs from equation 71.A6.3 and equations 71.A7.1 to 71.A7.5 for fyrθ / fyr20.

At the end of step 5, the elevated temperature design yield stress in the deck trough bars in Dimond
Hi-bond, fyrθ,dtb, is known.

For deck trough bars in Traydec, steps 2 – 5 are as follows:

Step 2: Determine the distances (u1, u2 and u3)dtb

ddtb
u3,dtb = c3 +
2
u1,dtb = u2,dtb = u 3,dtb2 + 25 2

HERA Steel Design & Construction Bulletin Page 43 No. 71, December 02/ January 03
c3 = as specified by the user, being the cover from the bottom of the slab to the underside of the
bar

Steps 3 – 5: as for steps 3 – 5, section A.4.1.1.1, using the value of FRR,dtb and (u1, u2 and u3)dtb from
steps 1 and 2 above, θs from equation 71.A6.2 and equations 71.A7.1 to 71.A7.5 for fyrθ / f yr20.

At the end of step 5, the elevated temperature design yield stress in the deck trough bars in Traydec,
fyrθ,dtb, is known

A4.1.4 Concrete and decking

The maximum concrete & decking temperatures have been determined from the wide range of natural
fire SAPPHIRE/SAFIR analyses described in DCB Issue No. 59 and which are consistent with the
results from the slab panel experimental tests [3]. For normal weight concrete, trapezoidal decking;

• Temperature across top of ribs = 700ºC for FHC 1


= 750ºC for FHC 2
= 800ºC for FHC 3
(FHC 1 to FHC 3 are as defined in Comment to Para.2.2 of [5]).

• Average temperature over full deck width = 700ºC for FHC 1


= 750ºC for FHC 2
= 800ºC for FHC 3

For normal weight concrete, flat slab base;


• Temperature = 750ºC for FHC 1
• Temperature = 810ºC for FHC 2
• Temperature = 870ºC for FHC 3

The decking strength is taken as zero at these temperatures, in accordance with Table 4.1of [12]. The
concrete strength at elevated temperature is given by R4-82 [8] and NZS 3101 [11] as :

'
fcè
èc ≤ 350 o C, '
= 1.0 (71.A9.1)
fc20

'
fcè  910 - è c 
350 o C < èc ≤ 910 o C, '
=  (71.A9.2)
fc20  560 

'
fcè
910o C < èc , '
=0 (71.A9.3)
fc20

In practice, when the concrete reaches its peak temperature and hence minimum compression
strength at the exposed face, layers deeper into the slab are cooler and stronger. To account for this,
in negative moment and support shear calculations, requires analysis of successive layers of concrete
to find the extent of the compression zone. This concept is shown in Fig. 71.29. This is undertaken in
Fig. 10 and Table 6 of [8], with the results tabulated for ease of use. The results are presented for
times under the standard fire test in [8], thus requiring the equivalent standard fire test times (te) for the
slab temperatures given above to allow the data in [8] to be used. These times are:

FHC 1: te = 60 minutes
FHC 2: te = 75 minutes
FHC 3: te = 90 minutes

From these times, values of ecθ are determined; see the details following equation 71.A21.

A4.1.5 Unprotected secondary beam components

The unprotected secondary beams span the short dimension, Lx, as shown in Fig. 71.26.

HERA Steel Design & Construction Bulletin Page 44 No. 71, December 02/ January 03
Being unprotected, they are assumed to directly exposed to the fire and therefore reach very high
temperatures. It has been shown from the Cardington tests [12,17] that the bottom flange and web
reach 95% of the peak fire temperature and the top flange, with its greater shielding and proximity to
the heat sink of the concrete slab, remains 150oC below the peak fire temperature.

Table 71.3
Design Temperatures of Unprotected
Secondary Beam Elements

Bottom Flange Web Top Flange


FHC 1, NWC 800 800 700
FHC 2, NWC 850 850 750
FHC 3, NWC 900 900 800
FHC 1, LWC 900 900 800
FHC 2, LWC 950 950 850
FHC 3, LWC 1000 1000 900

Notes to Table 71.3


o
1. All temperatures are C
2. NWC = normal weight concrete; density ≥ 2300 kg/m
3

LWC = light weight concrete; density ≤ 1900 kg/m


3

3. FHC = fire hazard category, as given by [5]

The peak fire temperatures have been obtained as described in the commentary section CA4.1.5.,
with the relationship described above then used to obtain the peak steel beam element temperatures.
These temperatures are a function of the Fire Hazard Category, type of concrete used in the floor
slabs and position of beam element. They are as given by Table 71.3.

The variation of yield stress with elevated temperature for the unprotected secondary beam elements
is taken from EC 3 Part 1.2 [29]. The values are given in Table 71.4 and are those for a strain > 2%,
this being appropriate for the magnitude of deformation developed by the slab panel and unprotected
secondary beam elements in severe fire conditions.

Table 71.4
Relationship Between fysbθ and fysb20 for
Unprotected Secondary Beam Elements
Bottom Flange Web Top Flange
FHC 1, NWC 0.110 0.110 0.230
FHC 2, NWC 0.085 0.085 0.170
FHC 3, NWC 0.060 0.060 0.110
FHC 1, LWC 0.060 0.060 0.110
FHC 2, LWC 0.050 0.050 0.085
FHC 3, LWC 0.040 0.040 0.060

Notes to Table 71.4:


1. These values apply to the secondary beam elements away from the connections.
2. These values are input into equations 71.A11.2 to 71.A11.4 through the constant Csb,θ , which is used in equation 71.A15.

A4.1.6 Secondary beam to primary beam connections; bolts and secondary beam web

Based on analyses of the Cardington test temperatures, the maximum bolt temperature,θf ,is taken as
that for the unprotected secondary beam top flange, from Table 71.3. This is conservative.

The same temperature is used for the beam web at the connection.

Given the bolt design temperature, the variation of bolt tensile strength with temperature for high
strength structural bolts has been established by UK testing, as detailed in section 4.3.3 of [12]. This
variation is as follows, for bolts in the relevant temperature range:

fufè
= 0.170 - (θf - 680) 0.5312 x 10 -3 (71.A10)
fuf20

for 680oC < θf ≤ 1000oC

HERA Steel Design & Construction Bulletin Page 45 No. 71, December 02/ January 03
A4.2 Design adequacy of slab panel; concrete slabs on profiled steel decking

A4.2.1 Yieldline load-carrying capacity of slab panel

(1) Determination of mx

This is developed by the mesh reinforcement, in the x-direction, and by the secondary beam elements,
as shown in Fig. 71.30. Designer judgement is required on the secondary beam contribution when
there are only one or two secondary beams within the slab panel area. The determination of mx is
based on reinforced concrete bending theory, thus:

Rtsx,mesh = Arx,mesh fyrθ,xmesh (71.A11.1)


Rtsbx,tf = (tf bf ) fysbθ,tf / Ssb (71.A11.2)
Rtsbx,w = ((d - 2tf ) tw ) fysbθ,w / Ssb (71.A11.3)
Rtsbx,bf = (tf bf ) fysbθ,bf / Ssb (71.A11.4)
Rtsx,total = ΣRts from mesh, tf, w, bf (71.A11.5)

erx = [(cxmesh+dmesh/2) Rtsx,mesh + (to + 0.5tf ) Rtsbx,tf


+ (to + 0.5d) Rtsbx,w
+ (to + d - 0.5tf ) Rtsbx,bf ] / Rtsx,total (71.A12)

R tsx, total
ax = (71.A13)
0.85 fc',20 b
mx = φfire [Rtsx,mesh (cxmesh + dmesh/2 - 0.5ax)
+ Rtsbx,tf (to + 0.5tf - 0.5ax)
+ Rtsbx,w (to + 0.5d - 0.5ax)
+ Rtsbx,bf (to + d - 0.5tf - 0.5ax)] (71.A14)

where:
φfire = 1.0, for the reasons given in Clause 11.5 of [10] (introduced through Amendment No. 1)
Arx,mesh = area/m width (mm2/m) of mesh in x-direction
≥ the area of mesh in the x-direction required for integrity; see section A4.3 for the
minimum mesh area requirements for integrity
fyrθ,xmesh = elevated temperature yield stress of mesh reinforcement, from section A4.1.1.1 (MPa)
tf , bf , d, tw = secondary beam element dimensions

fysbθ = Csb,θfysb20 (71.A15)


'
fc,20 = ambient temperature concrete 28 day specified compression stress (MPa)
b = 1000 mm
ax = depth of concrete compression stress block generated by Rtsx,total
Rts = elevated temperature tension capacity of the nominated steel element
cxmesh = cover from top of concrete to mesh bars in the x-direction, from section A4.1.1.1
dmesh = diameter of mesh bar
erx = distance from top of concrete to centroid of tension force, for x moment
dsb = depth of secondary beam

All dimensions are in mm.

(2) Determination of my

This is developed by the mesh reinforcement in the y-direction, plus the contribution of the deck trough
bars, if installed. Details are as follows:

Rtsy,mesh = Ary,mesh fyrθ,ymesh (71.A16.1)


Rtsy,dtb = Ary,dtb fyrθ,dtb (71.A16.2)
Rtsy,total = Rtsy,mesh + Rtsy,dtb (71.A16.3)

 
R tsy,mesh cymesh + 2  + Rtsy, dtb (t0 − u 3,dtb )
 dmesh 
   
ery = (71.A17)
R tsy, total

HERA Steel Design & Construction Bulletin Page 46 No. 71, December 02/ January 03
R tsy, total
ay = '
(71.A18)
0.85fc,20b

  d ay   ay 
my = φfire R tsy, mesh c ymesh + mesh −  + Rtsy, dtb  t0 − u3,dtb −  (71.A19)
  2 2  2 

where:
Ary,mesh = area/m width (mm2/m) of mesh in y-direction
≥ the area of mesh in the y-direction required for integrity; see section A4.3 for the minimum
mesh area requirements for integrity
Ary,dtb = area/m width (mm2/m) of deck trough bars
fyrθ,ymesh = elevated temperature yield stress of mesh reinforcement, from section A4.1.1.2
ay = depth of concrete compression stress block generated by Rtsy,total
to = full depth of slab
u3,dtb = distance from centroid of deck trough bar to base of decking
φfire = 1.0
b = 1000 mm
cymesh = cover from top of concrete to mesh bars in the y-direction, from section A4.1.1.2
dmesh = diameter of mesh bar
All dimensions are in mm.

(3) Determination of m 'x

This is the negative moment capacity in the x-direction, per unit length along the y-axis. It is only
applicable at an interior support over a primary beam into an adjacent slab panel, where it is
generated by the tension capacity of the interior support bars. Details of determining m 'x are as
follows:

Rtsx, isb = Arx,isb fyrθ,isb (71.A20)


 d 
m 'x = φfireR tx, isb t0 − hrc − cymesh + isb − ecè  (71.A21)
 2 
where:
ecθ = 14 mm for FHC 1
= 17 mm for FHC 2
= 19 mm for FHC 3
Arx,isb = area of interior support bars, expressed as mm2/m length of interior support bars in the
x-direction
fyrθ,isb = elevated temperature yield stress of the interior support bars, from section A4.1.2
cymesh = top cover to mesh bars in y-direction, from section A4.1.1.2
disb = 10 mm to 16 mm, as required
All dimensions are in mm.

(4) Calculation of yieldline load-carrying capacity

This is the elevated temperature load-carrying capacity for the actual support conditions, wylθ (ie. sides
1 and / or 3 developing negative moment, where applicable). It is obtained from the yieldline equation
for a general rectangular slab with either pinned or fixed edge supports, as given by [19], and shown in
Fig. 71.31. The equation is;

6 mx ì s X 2
w yl è = 2
(71.A22.1)
2 
L    Y 2  Ly 
2
L2x  x      + 3 ì s    -
Y
 Ly 
    X   Lx   X
 

where:
w ylè = yieldline load carrying capacity at elevated temperature

HERA Steel Design & Construction Bulletin Page 47 No. 71, December 02/ January 03
mx = positive moment in the x-direction per unit width along the y axis

µs = mx/my (71.A22.2)
my = positive moment in the y-direction per unit width along the x-axis
X = 1 + i1 + 1 + i3 ≥ 2 (71.A22.3)
Y =2 (71.A22.4)
i1 and i3 are defined in Fig. 71.31, i2 and i4 = 0

A4.2.2 Yieldline load-carrying capacity of simply-supported slab panel


This is determined from equation 71.A22.1, by setting X = Y = 2 (i.e. ignoring negative moments along
sides 1 and/or 3). The result is w ylθ,ss .

A4.2.3 Limiting deflection of slab panel

The limiting deflection, ∆ max, is given by equation 71.A23.3 as the lesser of equations 71.A23.1 and
71.A23.2. It gives the maximum allowed vertical deflection of the slab panel along line CD (see Fig.
71.26) relative to the adjacent supports.

4.27 x 10- 4 Lx
2
 0.9 fyr20,mesh  3Ly2
∆1 = +   (71.A23.1)
h  E  8
t0 − rc  20,mesh 
2
2
4.27 x 10 -4 Lx L
∆2 = + x (71.A23.2)
hrc 25
t0 −
2
where:
all dimensions are in mm
fyr20, E20 are in MPa
∆max = min (∆1, ∆2) (71.A23.3)

A4.2.4 Determination of tensile membrane enhancement


This is undertaken through equations 71.A24 to 71.A42. The process is sequential and builds on the
values of variables already determined. The equations look daunting but are easy to apply
and readily programmable into a spreadsheet. (Alternatively and preferably, use the HERA -
developed software; see section 8).

 2 
L   L x m y  3m L m 
L = x    ≤ 0.5L
y y
+ - x (71.A24)
1 2   L m  m L m  y
  y x  x y x 

Ly
a= (71.A25)
Lx

L1
n= (71.A26)
Ly

 4n a2 (1- 2n) 
k=  +1 (71.A27)
 4n a + 1 
2 2

1  1   Lx  1  L  1  1   L 
2 2 2
A=   - - 1  (nLy )2 + x  -   (nLy )2 + x  (71.A28)
2  1 + k   8n  2n   4  3  1 + k   4 
 

1  k 2   nLy
2
 2 
k (nLy )2 + Lx 
B=    - (71.A29)
2  1 + k   2 3 (1 + k)  4 
  

HERA Steel Design & Construction Bulletin Page 48 No. 71, December 02/ January 03
2
Lx
C= (k - 1) (71.A30)
16 n

 Ly   Ly nLy 
D =  - nLy   -  (71.A31)
 2  4 2 

2
1.1Lx
b= (71.A32)
8 (A + B + C - D)

 2 ax   2 ay 
gox = 1 -  ; goy = 1 -  (71.A33.1) ; (71.A33.2)
 erx   ery 
 

2g ox 2g oy
áx = ; αy = (71.A34.1) ; (71.A34.2)
3 + g ox 3 + g oy

1 - g ox 1 - goy
βx = ; βy = (71.A35.1) ; (71.A35.2)
3 + gox 3 + g oy
4b  ∆max   n (3k + 2) nk 3 
e1mx =    (1- 2n ) + - (71.A36)
3 + gox  3 (1 + k )2 3 (1+ k )2 
 erx  
4b  ∆max   2 + 3k k3 
e 2my =   -  (71.A37)
3 + g oy  ery   6 (1+ k )2 6 (1 + k )2 
   
 α b β b2 2 
e1bx = 2n 1 + x (k − 1) − x
2 3
( ) ( (
k − k + 1  + (1 − 2 n ) 1 − α x b − β x b 2 )) (71.A38)
 
á yb â yb 2
e 2by = 1 + (k - 1) - (k 2 - k + 1) (71.A39)
2 3
e1 = e1mx + e1bx (71.A40)
e2 = e 2my + e 2by (71.A41)
( e1 - e 2 )
e = e1 - (71.A42)
(1 + 2a 2 )
wu = w ylθ - w ylθ,ss + w ylθ,ss e (71.A43)

In equations 71.A24 to 71.A43;


L1 = distance from sides 2 and 4 to the intersection of the centre and diagonal yield lines (see
Fig. 71.26)
e = enhancement in simply supported slab panel load-carrying capacity due to tensile membrane
action
wu = design load-carrying capacity of slab panel when at maximum deflection along yieldline CD (see
Fig. 71.26)
A4.2.5 Check on moment/tension membrane adequacy of slab panel

If wu (equation 71.A43) ≥ w* (equation 71.A1), then panel moment/tension membrane capacity is


adequate.

If wu < w*, then moment/tension membrane adequacy must be increased.

A4.2.6 Determining the shear capacity of the slab panel

This is undertaken over the tributary area shown in Fig. 71.32. Shear capacity is developed through
load transfer from slab to supporting edge beams and also from load transfer from slab to secondary
beams and hence into the supporting primary beams. The shear capacity available from each mode
is given in (1) and (2) below.

HERA Steel Design & Construction Bulletin Page 49 No. 71, December 02/ January 03
(1) Shear capacity available through the slab
The critical location is at SS shown in Fig. 71.32.

vu,slab = φfire vcdv (kN/m width) (71.A44)


dv = t0 – hrc – cymesh + disb / 2 – 0.67ecθ (mm) (71.A45)
vc = 0.17 fc' (MPa) (71.A46)
φfire = 0.75/0.85 = 0.89 for shear

where:
vc = as given by NZS 3101 [11] Clause 9.3.2.2
cxmesh = top cover to mesh bars in x-direction, from section A4.1.1.1
cymesh = top cover to mesh bars in y-direction, from section A4.1.1.2
disb = 10 or 12 mm, typically
ecθ = as given by section A4.2.1 (3); immediately following equation 71.A21.
fc ’ = nominal concrete compression strength at, typically, 28 days (MPa)
(2) Shear capacity available through the secondary beam to primary beam connection
This is given by equation 71.A47.3, which is the lesser of equation 71.A47.1 and equation
71.A47.2.

Vu,sb-pb = φfire nb φ Vfn rf x 1.25 (71.A47.1)


Vu,vsb = φfire φ Vv Csb,θ,tf x 1.11 (71.A47.2)
Vu,sb = min (Vu,sb - pb ; Vu,vsb) (71.A47.3)
where:
φfire = 0.80/0.85 = 0.94 for equation 71.A47.1
φfire = (0.9/0.85 ≤ 1.0) = 1.0 for equation 71.A47.2
nb = no. of bolts in the connection between secondary beam and primary beam
φVfn = design ambient temperature bolt shear capacity, threads included, from eg. [20]
φVv = design shear capacity for secondary beam, from eg. [20]
fufè
rf = f uf20
, as given by equation 71.A10 for the temperature given in Table 71.3 for the
top flange element
Csb,θ,tf = as given by Table 71.4 for the top flange element
1.25 = 1/0.8; converting φVfn to Vfn
1.11 = 1/0.9; converting φVv to Vv

A4.2.7 Determine design shear per metre width at the supports


Lx
v* = w ∗ (kN/m width) (71.A48)
2
where:
w* = as given by equation 71.A1 (kPa)
Lx = short span of slab panel (m)
A4.2.8 Check on shear adequacy

Vu,sb
If v∗ ≤ vu,slab + , then the slab shear adequacy is ok.
S sb
If this check is not satisfied, consider either (1), or one or more of (2) and (3), to gain compliance:

(1) Design and detail the secondary beams so that their bottom flanges sit inside and above the
bottom flange of the supporting primary beam, eg. as shown in Fig. 71.11; or

(2) Increase the concrete strength, f c' , to increase the slab shear resistance, vu,slab; and/or

(3) Increase the number and or diameter of bolts in the secondary beam to primary beam
connection to increase Vu, sb-pb, if this is the limiting factor in equation 71.A47.3.

HERA Steel Design & Construction Bulletin Page 50 No. 71, December 02/ January 03
A4.3 Mesh area required for the preservation of integrity
Integrity is the ability of the floor slab to resist the passage of flame or hot gases [5]. For the floor
systems covered by this design procedure, the system will be expected to preserve integrity for the
specified structural fire severity period (teq).
With a concrete slab cast onto steel deck, integrity is retained even when a full depth crack greater
than around 1mm in width develops in the slab, because the decking will seal the base of the crack
against the passage of flame or hot gasses.

With a concrete slab not cast onto steel deck, the preservation of integrity has to be through limiting
the maximum width of any full depth crack to 1mm.

One of the principal purposes behind the slab panel fire tests [3] was to determine the slab mesh
reinforcement requirements for integrity directly from experiment, in lieu of the provisions of SPM 1,
stipulated in section CA5 of DCB No. 60, which were based on the strong crack control requirements
of AS 3600 [27] and expected to be conservative.

As described in section 4.5, pp.15–16 of DCB No. 70, those provisions are shown to be very
conservative and have been able to be significantly relaxed to the following:

For slabs not cast onto steel decking:

Arx, mesh; Ary, mesh ≥ 200 ⋅ A1(xory) ⋅ A2 (mm² / m width) (71.A49.1)

For slabs cast onto steel decking:

Arx, mesh; Ary,mesh ≥ 147 ⋅ A1(xory) ⋅ A2 (mm² / m width) (71.A49.2)

smesh(xory)
A1(xory) = ≥1 (71.A49.3)
150

t0 − hrc 2
A2 = ≥1 (71.A49.4)
110

150 mm ≤ smesh(xory) ≤ 250 mm (71.A49.5)


where:
A1 = factor relating to mesh bar spacing (no units)
A2 = factor relating to slab effective depth (no units)
smesh = mesh bar spacing (mm)

A5 Slab Panel Design; Floors Incorporating In-Situ Concrete Slabs on Speedfloor Joists
Supported on Primary Beams

Section A5 follows the same format and approach as section A4. Where details are identical, the
relevant provisions of section A4 are referenced, rather than the provisions being repeated.

A5.1 Temperatures of slab panel components and associated mechanical properties

A5.1.1 Mesh within the concrete slab

As shown in Fig. 71.16, the mesh sits on top of the upstand of the Speedfloor truss, which is
embedded in the concrete slab. This upstand height from the base of the slab is 38 mm [14] for both
joist sizes made.

Step 1: Determine the design FRR

FRR, mesh = 0.9 teq (71.A3)

HERA Steel Design & Construction Bulletin Page 51 No. 71, December 02/ January 03
Step 2: Determine the distances u1, u2 and u3 from the mesh to the slab soffit

For mesh placed over Speedfloor joists, the y-direction mesh bars run over the top of the joists and
the x-direction bars are taken in this procedure as being on the underside of the y-direction bars.

Step 2.1: x - direction mesh bars

d mesh
u3,xmesh = 38mm − (71.A50.1)
2
u1,xmesh = u2,xmesh = u3,xmesh2 + 252 (71.A50.2)

Step 2.1: y - direction mesh bars

dmesh
u3,ymesh = 38mm + (71.A50.3)
2
u1,ymesh = u2,ymesh = u3,ymesh2 + 25 2 (71.A50.4)

Step 3: Determine the heat transfer coefficient, γ

This uses equation 71.A5.1 from section A4.1.1.1, with the values of u1, u2 and u3 for each direction as
given above.

Step 4: Determine the design temperature of the mesh bars

This uses equation 71.A6.1 from section A4.1.1.1, FRR,mesh from step 1 above and γ from step 3
above:

θs = Cs,θ - 350 γ (71.A6.1)

where:
θs = temperature of the steel bar
γ = as determined from step 3 above
Cs,θ has the following values:

• 900oC for FRR = 22 mins


• 1019ºC for FRR = 34 mins
• 1175oC for FRR = 60 mins
• 1285oC for FRR = 90 mins
• 1370oC for FRR = 120 mins
• 1455oC for FRR = 180 mins
• 1545ºC for FRR = 240 mins

Step 5: Determine the elevated temperature design yield stress in the mesh bars, fyrθ,mesh.

This uses equations 71.A7.1 to 71.A7.5 if the mesh is grade E or N (hot-formed) to AS/NZS 4671 [24]
or equations 71.A7.6 to 71.A7.8 if the mesh is grade L (cold-formed) to [24]; see details in step 5,
section A4.1.1.1 and A4.1.1.2.

A5.1.2 Interior support bars

As shown in Fig. 71.16, these sit on top of the mesh reinforcement.

Step 1: Determine the design FRR

FRR,isb = 0.9 teq (71.A3)

This is the same equation as for the mesh.

HERA Steel Design & Construction Bulletin Page 52 No. 71, December 02/ January 03
Step 2: Determine the distances u1, u2 and u3 from the interior support bars to the base of the slab.

The interior support bars sit over the top of the mesh y-direction bars at a specified top cover of 25mm.

u3,isb = to – cisb – disb / 2 (71.A51.1)

cisb = 25 mm (from [14])


disb = 10 mm to 16 mm, as required

u1,isb = u2,isb = u3,isb2 + 25 2 (71.A51.2)

Steps 3 – 5: as for steps 3 – 5, section A5.1.1, using the value of FRR,isb and (u1, u2, u3)isb from steps 1
and 2 above and equations 71.A7.1 to 71.A7.5 for fyrθ / f yr20

At the end of step 5, the elevated temperature design yield stress in the interior support bars, fyrθ,isb, is
known.

A5.1.3 Speedfloor Joists

As described in the commentary to this section, the proportion of the Speedfloor Joist that is
embedded within the concrete slab contributes a small but significant component to the slab panel
capacity, through acting as additional reinforcement in the x-direction. This contribution has been
determined by dividing the lengths of joist embedded within the concrete into 5 individual sections,
then determining the tensile contribution of each section in the x-direction and hence its contribution to
the development of positive moment in the x-direction per unit length along the y-axis. The dimensions
of each length and its distance from the base of the concrete are set by the shape of joist [14] and the
temperature variation with time has been determined from the experimental tests. See Fig 71.34. The
details are as follows:

(
Section A: R tsx, A = 42mm 2 ⋅ fyrèrA Ssfj )
fyrèrA = 416MPa ⋅ (Cs,è − 700º C)
Section B: ( )
R tsx, B = 64.8mm 2 ⋅ fyrèrB Ssfj
fyrèrB = 416MPa ⋅ (Cs,è − 600º C)
Section C: ( )
R tsx, C = 62mm 2 ⋅ fyrèrC Ssfj
fyrèrC = 416MPa ⋅ (Cs,è − 500º C)
(
Section D: R tsx, D = 64.8mm 2 ⋅ fyrèrD Ssfj )
fyrèrD = 416MPa ⋅ (Cs,è − 430º C)
Section E: ( )
R tsx, E = 81mm 2 ⋅ fyrèrE Ssfj
fyrèrE = 416MPa ⋅ (Cs,è − 380º C)

E
Rtsx, sfj = ∑ Rtsx,n (71.A52)
n= A

where:
Cs,θ = from section A5.1.1 step 4
Ssfj = spacing between Speedfloor joists
Rtsx,sfj = elevated temperature tension capacity of the Speedfloor Joist

A5.1.4 Concrete
The floor slab has a flat concrete base.

For normal weight concrete;

• Temperature = 750ºC for FHC 1


• Temperature = 810ºC for FHC 2
• Temperature = 870ºC for FHC 3

HERA Steel Design & Construction Bulletin Page 53 No. 71, December 02/ January 03
Following the same approach as described in section A4.1.4, these temperatures are converted into
standard fire test times, being te = 60 mins for FHC 1, te = 75 mins for FHC 2 and te = 90 mins for
FHC 3.

A5.1.5 Speedfloor Joists to primary beam connections

The Speedfloor Joist to primary beam connection is considered to have zero elevated temperature
shear or axial capacity.

A5.2 Design adequacy of slab panel; concrete slabs on Speedfloor Joists

A5.2.1 Yieldline load-carrying capacity of slab panel

(1) Determination of mx

This is developed by the mesh reinforcement and embedded portion of the Speedfloor joists in the
x-direction (see Figs. 71.33 and 71.34), thus:

Rtsx,mesh = Arx,mesh fyrθ,xmesh (71.A53)

Rtsx,total = Rtsx,mesh + Rtsx,sfj (71.A54)

R tsx, total
ax = '
(71.A55)
0 .85 fc20 b

  d mesh  
R tsx,mesh cxmesh + 2  + Rtsx,A (t0 − u A ) + R tsx,B (t0 − u B )
   
+ R tsx,C (t0 − u C ) + R tsx,D (t0 − uD ) + Rtsx,E (t0 − uE ) 
erx = (71.A56)
R tsx,total

  d mesh  ax   a   a 
R tsx,mesh  cxmesh +  −  + R tsx, A t0 − u A − x  + R tsx,B  t0 − uB − x 
mx = φ fire   2  2  2  2 
(71.A57)
 
+ R tsx, C t0 − uC − ax  + R tsx, D t0 − uD − ax  + Rtsx.E t0 − uE − ax  
  2  2  2 

where:
Arx,mesh = area/m width (mm2/m) of mesh in x-direction
≥ the area of mesh in the x-direction required for integrity; see section A5.3 for the
minimum mesh area requirements for integrity
Rtsx,sfj = elevated temperature tension capacity of the Speedfloor joist from section A5.1.3
to = thickness of floor slab (mm)
fyrθ,mesh = elevated temperature yield stress of mesh reinforcement from section
A5.1.1 (MPa)
φfire = 1.0 (see NZS 3404 Amendment No. 1 Clause 11.5)
'
fc20 = ambient temperature concrete 28 day specified compressive stress (MPa)
b = 1000 (mm)
ax = depth of concrete compression stress block generated by Rtsx,mesh (mm)
uA-E = bottom cover to centroid from section A5.1.3, Fig. 71.34
cxmesh = t0 – 38mm (the embedded height of the Speedfloor joist is 38 mm)

(2) Determination of my

This is developed by the mesh reinforcement in the y-direction (see Fig. 71.33), thus:

Rtsy,mesh = Ary,mesh fyrθ,ymesh (71.A58)

HERA Steel Design & Construction Bulletin Page 54 No. 71, December 02/ January 03
Rtsy,mesh
ay = '
(71.A59)
0.85 fc20b

my = φfire Rtsy,mesh (t0 – 38 – 0.5dmesh – 0.5ay) (71.A60)

where:
Ary,mesh = area/m width (mm2/m) of mesh in y-direction
≥ the area of mesh in the y-direction required for integrity; see section A5.3 for the minimum
mesh area requirements for integrity
to = thickness of floor slab (mm)
fyrθ,mesh = elevated temperature yield stress of mesh reinforcement from section A5.1.1(MPa)
φfire = 1.0 (see NZS 3404 Amendment No. 1 Clause 11.5)
b = 1000 (mm)
ay = depth of concrete compression stress block generated by Rtsy,mesh (mm)

(3) Determination of m 'x

This is the negative moment capacity in the x-direction per unit length along the y-axis (see
Fig. 71.33). It is only applicable at an interior support over a primary interior beam into an adjacent
slab panel, where it is generated by the tension capacity of the interior support bars.

It is determined as follows:

Rtsx,isb = Arx,isb fyrθ,isb (71.A61)

m 'x = φfire Rtsx,isb (to – er,isb – ecθ) (71.A62)


where:
ecθ = 14 mm for FHC 1
= 17 mm for FHC 2
= 19 mm for FHC 3
Arx,isb = area of interior support bars, expressed as mm2/m length of these bars along the y-axis in
the x-direction
fyrθ,isb = elevated temperature yield stress of the isbs from section A5.1.2
er,isb = distance from top of concrete down to centroid of interior support bar
= cisb + disb/2
cisb = 25 mm (from [14])
disb = 10 mm to 12 mm, as required

(4) Calculation of yieldline load-carrying capacity

This uses equations 71.A22, with mx, my and m 'x from sections (1), (2) and (3) above.

A5.2.2 Yieldline load-carrying capacity of simply-supported slab panel

This is determined from equation 71.A22.1 by setting X = Y = 2 (ie. ignoring negative moments along
sides 1 and 3). The result is wylθ,ss.

A5.2.3 Limiting deflection of slab panel

As determined from section A4.2.3, but setting hrc = 0 in equations 71.A23.1 and 71.A23.2.

A5.2.4 Determination of tensile membrane enhancement

As determined from section A4.2.4, equations 71.A24 to 71.A43, using the relevant values calculated
above for the floor system on Speedfloor joists, as shown in Fig. 71.33.

The end result is wu, from equation 71.A43, which in this instance is the design load-carrying capacity
of the slab panel when at maximum deflection along gridline CD (see Fig. 71.33).

HERA Steel Design & Construction Bulletin Page 55 No. 71, December 02/ January 03
A5.2.5 Check on moment/tension membrane adequacy of slab panel

If wu (equation 71.A43 but applied to the Speedfloor system) ≥ w* (equation 71.A1), then slab panel
moment/tension membrane capacity is adequate.

If wu < w*, then moment/tension membrane adequacy must be increased.

A5.2.6 Determining the shear capacity of the slab panel

This is undertaken over the tributary areas shown in Fig. 71.33. Shear capacity is developed through
load transfer from the slab to the supporting edge beams. Any shear capacity in the Speedfloor joist
to beam connections under fire conditions is ignored.

The slab is a solid slab, so the critical location is at the edge of the primary beam flange.

vu,slab = φfire vc dv (kN/m width) (71.A63)

dv = to – er,isb – 0.67ecθ (mm) (71.A64)

vc = 0.17 fc' (MPa)

φfire = 0.75/0.85 = 0.89 for shear

where:
vc = as given by NZS 3101 [11] Clause 9.3.2.2
ecθ = as given by section A5.2.1 (3); immediately following equation 71.A62
er,isb = distance from top of concrete down to centroid of interior support bar
= cisb + disb/2

A5.2.7 Determine the design shear per metre width on the tributary area at the supports

Lx
v* = w* (kN/m width) (71.A65)
2

where:
w* = as given by equation 71.A1 (kPa)
Lx = short span of slab panel (m)

A5.2.8 Check on shear adequacy of the slab panel

If v* ≤ vu,slab, then the slab shear adequacy is ok.

If this check is not satisfied, then consider one or both of the following:

(1) Increase the concrete strength, fc' , to increase vc; and/or


(2) Increase the reinforcement at the edges (ie the edge bars and the interior support bars) until vc
is increased through the additional reinforcement in accordance with NZS 3101 [11]
Equation 9-3.

A5.3 Mesh area required for the preservation of integrity

Integrity is the ability of the floor slab to resist the passage of flame or hot gases [5]. For the floor
systems covered by this design procedure, the system will be expected to preserve integrity for the
specified structural fire severity period (teq).

With a concrete slab cast onto steel deck, integrity is retained even when a full depth crack greater
than around 1mm in width develops in the slab, because the decking will seal the base of the crack
against the passage of flame or hot gasses.

HERA Steel Design & Construction Bulletin Page 56 No. 71, December 02/ January 03
With a concrete slab not cast onto steel deck, the preservation of integrity has to be through limiting
the maximum width of any full depth crack to 1mm.

One of the principal purposes behind the slab panel fire tests [3] was to determine the slab mesh
reinforcement requirements for integrity directly from experiment, in lieu of the provisions of SPM 1,
stipulated in section CA5 of DCB No. 60, which were based on the strong crack control requirements
of AS 3600 [27] and expected to be conservative.

As described in section 4.5, pp.15–16 of DCB No. 70, those provisions are shown to be very
conservative and have been able to be significantly relaxed to the following:

For Speedfloor slabs (which are not cast onto steel decking):

Arx, mesh; Ary,mesh ≥ 200 ⋅ A1(xory) ⋅ A2 (mm² / m width) (71.A49.1)

smesh(xory)
A1(xory) = ≥1 (71.A49.3)
150

t 0 − hrc 2
A2 = ≥1 (71.A49.4)
110

150 mm ≤ smesh(xory) ≤ 250 mm (71.A49.5)


where:
A1 = factor relating to mesh bar spacing (no units)
A2 = factor relating to slab effective depth (no units)
smesh = mesh bar spacing (mm)

A6 Design of Supporting Beams

A6.1 Scope

The supporting beams are those around the slab panel edges that will remain effectively undistorted in
fully developed fire conditions.

For example, in the slab panels shown in Figs. 71.17 and 71.18, they comprise the primary beams
(interior and edge) and the secondary edge beams.

To ensure that they remain effectively undistorted unless they are very greatly overdesigned for fire
emergency conditions, they must be shielded from direct exposure to fire by either:

• passive fire protection; or


• radiation barriers that operate for a suitably long period of time, as detailed in section 7.3 herein.

Application of these provisions is detailed in section A6.2.

A6.2 Application

The design of these beams for the fire emergency condition is as follows:

Step 1: Determine the applied fire emergency design load


This is the load w* from equation 71.A1 applied over the tributary area given by the slab panel design,
plus the beam self-weight and any applied dead load directly on the beams.

The primary interior beams are loaded by the area BCDF from slab panel 1 (see Figs. 71.26 or 71.33),
plus the corresponding area from slab panel 2, plus the beam’s selfweight. It is either accurate or
conservative to design these beams for a line load given by;

wu,pib = 0.5 w* (Lx1 + Lx2) + wselfweight (71.A66)

The primary edge beams are loaded by the area ACDE from slab panel 1 (see Figs. 71.26 or 71.33)
plus the beam selfweight and the weight of any cladding supported by the beam.

HERA Steel Design & Construction Bulletin Page 57 No. 71, December 02/ January 03
The secondary edge beams are loaded by the tributary area ACB or EDF from slab panel 1, plus the
beam selfweight and the weight of any cladding supported by the beam. The slab panel loading is
triangular, varying from zero at the ends to L1 w* at midspan. L1 is as given by equation 71.A24. Note
that this loading pattern is considerably different to and may be more severe than that for ambient
temperature design. Where the secondary edge beams are gravity carrying only, this slab panel fire
emergency condition may determine the beam size, because of the increased area of tributary
loading.

Step 2: Determine the limiting temperature on each supporting beam


This is undertaken in accordance with NZS 3404 Clause 11.5 and normal plastic design philosophy. It
involves calculating the factor rf, which is undertaken as follows:

(1) For beams with simple end connections


Mss
rf = (71.A67)
φfire Mpos

where:


M ss = simply supported design bending moment from the step 1 loads
Mpos = Mrc = midspan nominal bending moment capacity of the composite beam
φfire = 1.0, as introduced via NZS 3404 Amendment No. 1

(2) For beams with rigid or semi-rigid end connections

 Mss∗ 
rf =   (71.A68)
 φfire Ì 
pos + φfire Mneg 

where:
Mpos = Mrc
Mneg = for rigid connections, the support nominal negative moment capacity of the
beam, typically Ms
= for semi-rigid connections, the negative moment capacity of the connection
(e.g. see DCB No. 58 for the semi-rigid Flange Bolted Joint and DCB No. 68 for the
semi-rigid Sliding Hinge Joint).
φfire = 1.0

(3) Determine Tl from NZS 3404 [10] Clause 11.5

(4) For the secondary edge beams, if necessary increase the beam moment capacity so
that Tl ≥ 5500C

Step 3: Determine the extent of protection required to the beam. There are two options
available here.

Option 3.1: Using passive fire protection

(1) Determine the FRR required, in accordance with section A3.2 of this appendix and section 4.2
herein.

(2) Select an appropriate material and thickness of insulation required, eg. by using HERA Report
R4-89 [13] and the appropriate product supplier’s design charts. The most cost-effective
material and system should always be used. For beams in interior environments and hidden
from view, use a sprayed cementations coating, or use a board system where the number of
beams requiring protection is small. Relative costings are given in HERA Report R4-96 [23].

HERA Steel Design & Construction Bulletin Page 58 No. 71, December 02/ January 03
Option 3.2: Perform an elevated temperature moment and shear capacity check

This option is particularly suited to the beams of seismic-resisting systems, which are sized for
earthquake and often have a high reserve of strength in fire. It should also be considered for any
beam for which the limiting temperature from step 2 (3) above exceeds 750oC. These members may
not need passive fire protection, however the provisions of NZS 3404 Clauses 11.5 and 11.6 are very
conservative in such instances. A more accurate approach is to calculate the elevated temperature
moment and shear resistance of the beam, as an unprotected member, and to check if this is
adequate, as follows:

Step 3.2.1: Determine the design actions

These are M ∗ss and V*, determined using the loads from step 1 of this section.

Step 3.2.2: Determine the design elevated temperature of the beam and connection
components

For interior beams, use the design temperatures given in Table 71.3 for the unprotected secondary
beam elements.

For edge beams, use 50oC less than the design temperatures for the unprotected secondary beam
elements given in Table 71.3. This applies to all edge beams, whether adjacent to external openings
or not.

The same temperatures are used for the connection components to those interior or edge beams, as
appropriate.

Step 3.2.3: Determine the design elevated temperature moment capacity of the beam

This involves first determining the reduction in yield stress, due to the elevated temperature from step
3.2.2, using Table 71.4 or using Table 3.1 of EC3 Part 1.2 [29] if the appropriate value is not given in
Table 71.4.

The ambient temperature moment capacities, Mpos and Mneg (see step 2 above) are then multiplied by
the reduction factor for the bottom flange to give the elevated temperature moment capacities, Mpos,θ
and Mneg, θ.

Step 3.2.4: Determine moment adequacy through a plastic collapse mechanism check

For a beam with rigid or semi-rigid connections at both ends, this involves satisfying equation 71.A69


(φfire Mpos,θ + φfire Mneg, θ ) ≥ M ss (71.A69)

where:
φfire = 1.0

Step 3.2.5: Check the elevated temperature shear capacity

For beams with rigid end connections, this can be taken as satisfactory.

For beams with semi-rigid or simple connections, check the elevated temperature shear capacity of
the connection, applying the principles given in section A4.2.6(2) with the connection component
temperatures given in step 3.2.2 above. The capacity must equal or exceed V* from step 3.2.1 of this
section.

Step 3.2.6: Conclusion

If steps 3.2.4 and 3.2.5 are satisfied, the supporting beam may be left unprotected.

HERA Steel Design & Construction Bulletin Page 59 No. 71, December 02/ January 03
A7 Design of Supporting Columns

A7.1 Scope

The supporting columns are required to be protected in order that they remain effectively undistorted
in fully developed fire conditions. This will be through the use of passive fire protection.

Application of these provisions is detailed in section A7.2.

A7.2 Application

The design of the columns for the fire engineering condition is as follows:

Step 1: Determine the design structural fire severity (FRR) to use on the columns

This is given by sections 7.5 or 7.6 of DCB Issue No. 59 and repeated in section 4.3 herein.

Step 2: Select an appropriate material and thickness of insulation required

Use published guidance, such as HERA Report R4-89 [13] and the appropriate product supplier’s
design charts.

The most cost-effective material and system should always be used. Where the column is hidden
from view, a sprayed cementations coating is preferred. Where the column is exposed, consider a
boarded system. Boarded systems come in a variation of products and prices, with the price being
strongly influenced by the fixing system. This should be as rapid to assemble as possible.

Commentary to Appendix A
CA1 Commentary on Introduction and Scope

The design procedure is written for application to two general types of systems:

(1) In-situ concrete flat slabs and in-situ concrete slabs on profiled steel deck, with the slab
supported on secondary and primary beams. For the slabs on steel deck, there are several
decking profiles available; two of them (Hi-bond [30] and Tray-dec [25]) were included in the
slab panel fire test series [3] undertaken at BRANZ during mid – 2002. Where aspects of the
design procedure are specific to one deck profile, this is stated herein.

(2) Concrete slabs on Speedfloor joists. These are closely spaced, cold-formed joists and are
proprietary product, details of which are given in [14]. They are the only system of their type
currently available in New Zealand and the procedure is specifically written around the
characteristics of that product, such as height of beam ribs. They were also part of the slab
panel fire test program [3].

CA2.1 Commentary on definition of slab panel

For application of this procedure, in a severe fire the supports of the slab panel must remain
effectively undistorted, relative to the peak deformations developed within the slab panel region.

The limiting temperature concept given in NZS 3404 [10] is based on the maximum temperature
reached by an individual (beam) member when at the point of failure under standard fire test
conditions. These test conditions involve unrestrained supports and a load level such that a failure
deflection of around beam span/25 (Lb/25) is reached at a limiting temperature of 5500C [13].

When the same beam is placed into a building, with simple supports (such as WP or FE
connections from [6]) and is heated in a natural fire to 5500C, the actual deflection that will be
reached is very much lower, even for a beam carrying the maximum ratio of (design fire emergency
load) to (design load-carrying capacity at room temperature). The range of deflections reached in
this situation have been determined from realistic fire tests, eg. as summarised in [16] – see
especially Tables 5.1 and 5.2 therein, or given by test 1 from [17]. They are no more than Lb/100,
with an expected maximum of Lb/75. These deflections will typically be reached in the interior

HERA Steel Design & Construction Bulletin Page 60 No. 71, December 02/ January 03
beams, edge beam deflections will be lower. Over half of this deflection is due to thermal effects
and is recovered on cooling.

The maximum expected deformation within the slab panel will occur along the yielding length CD,
in Fig. 71.26. This could be as high as Lx/12.5 (comprising Lx/25 of mechanical effects and Lx/25 of
thermal-induced rotation). An example is shown in Fig. 71.10.

Thus the peak deflections of the slab panel supporting members will be no greater than 1/6th of the
peak slab panel deflections, which meets the “effectively undistorted” criterion.

CA2.2 Commentary on determining the size of the slab panel

No commentary explanation is needed with respect to Lx.

The lower limit on Ly = Lx is applied principally for ease of application of the design procedure, and
also because it represents a realistic lower limit in most instances. In the rare instance where
Ly < Lx, then the procedure can be applied with the orientation of the dimensions Ly and Lx reversed
from that shown in Figs. 71.26 or 71.33. Alternatively a conservative answer may be obtained by
setting Ly = Lx.

The upper limit on Ly = 30m is used because of the migrating fire behaviour that will occur in large
firecells of multi-storey buildings. This means that only part of a large open slab panel will be
subjected to fully developed fire conditions at any one time, thus a limit on slab panel long
dimension (Ly) is reasonable to apply. The same issue was faced in developing the separation
distance criteria of C/AS1 [5]. The BIA Working Group 20 (of which the HERA Structural Engineer
was a member) considered that using a 30 m maximum emitter width for determining separation
criteria was an appropriate allowance for the migrating fire effect and the same concept and figure
is used here. This maximum length, for Ly, is applicable to slab panels that are open in terms of fire
spread (e.g. an open plan office). For buildings containing full height walls and especially fire
and/or sound rated walls, fire spread will be much reduced. For an apartment building, a
conservative slab panel width would be the width of two adjacent apartments.
In practice, Lx = 12 m would be a typical upper limit on the shorter panel dimension, thus Ly = 30
denotes a panel with Ly/Lx ≥ 2.5. The calculation of wu (see section A4.2.5) is not very sensitive to
Ly/Lx, once Ly/Lx ≥ 2.5, which means any unconservatism inherent in limiting Ly to 30 m is negligible
and in practice will be more than compensated for by ignoring the benefits of non-uniform fire
conditions and heating of floor components within the slab panel region.

The slab panel load-carrying capacity is developed by two-way action, however, and as Ly/Lx
increases, the advantages of two-way action diminish. At Ly/Lx = 2.5, the yield-line capacity for a
slab panel with mx = my is 1.58 times that for one-way action along Lx alone, while at Ly/Lx = 5 it is
only 1.26 times the one-way capacity. It is therefore recommended that Ly/Lx ≤ 2.5 is used, in
practice. This can be achieved by designing selected secondary interior beams as slab panel
edges, where required. See details in section A2.2. This decision has load-carrying capacity
implications for those beams, which may need to have their strength under fire emergency
conditions increased accordingly.

CA3.2 Commentary on design structural fire severity

Note that, in a sprinklered building, a sprinkler reduction factor is not used in calculating teq for this
procedure. This is because the design procedure is based on design for the event of sprinkler
failure and subsequent full fire development.

This procedure is intended for application to multi-storey office buildings and other applications with
Fire Hazard Category (FHC) of 1 – 3 (see C/AS1 [5] Table 2.1). When these buildings are
sprinklered, the probability of a fire reaching full development in any one building is assessed at 1.2
x 10-5 or less [18]. This is an extremely small probability; some 150 times less than the probability
of a building being subjected to any other ultimate limit state event [18]. Even the probability of fully
developed fire following severe earthquake in a sprinklered building is an order of magnitude less
than the probability of occurrence of the earthquake itself [18].

HERA Steel Design & Construction Bulletin Page 61 No. 71, December 02/ January 03
CA4 Commentary on Slab Panel Design Method; Floors Incorporating In-Situ Concrete Slabs
on Profiled Steel Sheeting, Supported by Primary and Secondary Composite Beams

CA4.1 Temperatures of slab panel components and associated mechanical properties

In reporting [9] on the ambient temperature tensile membrane test undertaken to verify this design
methodology, Bailey advises that for structural fire severities (ie. time equivalent periods)
exceeding around 30 mins, the elevated temperatures of the mesh reinforcement and the
appropriate reduction in ambient temperature mechanical properties need to be taken into account.

This New Zealand application of this tensile membrane procedure [2] is developed principally for
use in office buildings with normal weight concrete floors, where the structural fire severity will
typically be not less than 50 mins and could be higher than 180 mins. In such instances, it is
important that as realistic assessment of the elevated temperatures of the structural components
as possible is made and that appropriate elevated temperature mechanical properties are used.

As part of the research undertaken in developing the first edition of the SPM procedure, HERA
undertook a broad range of advanced heat transfer analyses in order to determine the
temperatures of the concrete, decking and reinforcement for a typical slab an profiled steel decking
subjected to a wide range of natural fire conditions. Summary details were presented in DCB No.
59 and the results were used in the SPM first edition. The slab panel fire tests [3] have provided a
comprehensive set of experimental data with which to compare all the analytical results and make
adjustments where appropriate. This has been done; the differences between the predicted
temperatures from this second edition and those of the first edition are presented in DCB No. 70.
This has further enhanced the accuracy of the temperature prediction provisions for the critical slab
panel components, which are:

• reinforcing mesh within the slab (section A4.1.1)


• additional bar reinforcement (sections A4.1.2, A4.1.3)
• concrete and decking on the fire exposed underside of the slab (section A4.1.4)

A background to the elevated temperature mechanical properties used for these components is
given in CA4.1.1 - CA4.1.3.

As can be seen from Figs. 71.17, 71.18, 71.12, there are more components involved in a slab
panel than the four mentioned above. Background to the temperature/mechanical property
requirements for the rest of those components is now given:

• The trimmer bars and edge bars (see Fig. 71.12) are present to ensure there is continuity of
tension force transfer from the mesh reinforcement across the top of the slab panel supporting
beams and into these beams through the shear studs. They also assist in shear resistance.
However, an explicit calculation of tensile force from these bars is not required, so their
temperature and mechanical properties do not require determination.

• The shear studs on the supporting beams are important to the overall system performance
when the supporting beams are composite, however their elevated temperature capacity is not
required to be determined. Observations from the Cardington fire tests [17] show that they
maintain full shear connection between the concrete slab and supporting steel beams under
severe fire conditions.

• The unprotected secondary beams (Fig. 71.17) reach very high temperatures, however they still
make the dominant contribution to the slab panel positive moment in the x-direction per unit
length along the y-axis, compared to the reinforcement within the slab. Derivation of the design
temperatures and elevated temperature mechanical properties used for these secondary beam
components is given in CA4.1.5. The slab panel fire tests [3] did not include a slab panel with
unprotected supporting secondary beams, however the Speedfloor test slab panel included
three unprotected Speedfloor joists, as shown in Fig. 71.35. The central joist, in particular, made
a significant contribution to the stiffness of the slab panel, increasing the rate and magnitude of
deflection in the initial 30 mins and then decreasing the rate of deflection with time from 90
minutes onward. This can be seen from comparing the deflection curves for the 661 flat slab
and the Speedfloor slab in Fig. 70.4, DCB No. 70. This contribution was made despite the
significant buckling experienced by the joists, especially the central joist, as shown in Fig. 71.35.

HERA Steel Design & Construction Bulletin Page 62 No. 71, December 02/ January 03
Unprotected secondary beam I–sections undergo only minor local buckling away from the
supports in severe fire conditions, as shown for example in Fig. 59.2 in DCB No. 59 and
throughout the Cardington fire tests [17]. Their contribution to slab panel strength and stiffness
will be significantly greater, can be dependably calculated and is incorporated into the design
procedure as described in section A4.1.5 and Figure 71.30.

• The Speedfloor joists (Fig. 71.18, 71.16) make an appreciable contribution to slab panel positive
moment in the x-direction, however significant local buckling of the joist elements beneath the
concrete (see Fig. 71.35) makes the contribution from these elements impossible to quantify for
design. However, the contribution from the elements embedded in the floor slab can be
dependably quantified, as detailed in section A5.1.3.

• Derivation of the temperatures in the bolts of the secondary beam to primary beam connections
and in the web of the secondary beam at the connection to the primary beam is covered in
CA4.1.6.

CA4.1.1 Mesh within the concrete slab; elevated temperature mechanical properties

The provisions for determining component temperature given in the first edition have been modified
for greater accuracy on the basis of the experimental testing. These modifications are fine-tuning
and involve:

• Changes to the design heat paths used to calculate component temperatures


• Changes to the coefficient Cs,θ, which was previously obtained from [8]
• These modified provisions from [8] apply to flat slabs without decking. Their accuracy,
compared with the experimental results, is shown as the SPM 2 values in DCB No. 70 Figs.
70.8, 70.9 and 70.12 – 70.15 inclusive. Also shown is the corresponding results from the first
edition provisions, termed the SPM 1 values.
• For a flat slab on steel deck, the deck provides significant insulation at lower time equivalent
values, with this influence diminishing as the time equivalent (structural fire severity) increases.
A specific equation has been developed for this, based on the Traydec test results. This is
equation 71.A6.2; its accuracy is shown in Figs. 70.12 to 70.15, DCB No. 70.
• For the trapezoidal (Hi-bond) profile, another specific relationship between θs, teq and the
position of reinforcement has been developed. This is equation 71.A6.3; its accuracy is shown
in Figs. 70.12 to 70.15.

Having obtained the mesh temperature, the variation of mechanical properties with temperature
needs to be determined. This can be obtained from a number of sources, such as NZS 3101 [11],
however the values given in EC2 Part 1.2 [28] are more appropriate to use, for the following
reasons:

• They differentiate between cold-worked and hot-formed reinforcement, which is desirable, given
that the properties are different and the mesh could be of either sort. However, the table
headings given for each form of mesh in [28] are the wrong way around. The correct
assignment of properties is made herein.

• Values are given for strains ≥ 2%, which is appropriate to the magnitude of deformation
developed by the slab panel in severe fire conditions at the ultimate limit state

The relationships for cold-worked bars are given in Table A.3 of [28], while those for hot-formed
bars are given in Table A.4 of [28].

In applying them to this procedure, the hot-formed bar relationship is presented as a set of data
points, with linear interpolation between each point. These are equations 71.A7.1 to 71.A7.5. The
cold-worked relationships of Table A.3 of [28] can be expressed in linear equation form, without
loss of accuracy, and this approach is used herein.

The cold-worked provisions cover grade L mesh to AS/NZS 4671 [24], while the hot-formed
provisions cover grade E or N mesh or bars to [24].

HERA Steel Design & Construction Bulletin Page 63 No. 71, December 02/ January 03
CA4.1.2 Interior support bars; elevated temperature mechanical properties

These bars are hot-formed grade E (typically) to [24], hence the variation of mechanical properties
with temperature are taken from Table A.4 of [28].

CA4.1.3 Deck trough bars; elevated temperate mechanical properties

Determination of temperature of these bars depends on the decking profile and uses the profile –
specific equations developed from [3].

These bars are hot-formed grade E (typically) to [24], hence the variation of mechanical properties
with temperature are taken from Table A.4 of [28].

CA4.1.4 Concrete; elevated temperature mechanical properties

These are taken from NZS 3101 [11], rather than EC2, for three reasons:

• The values in [11] have been developed for concretes incorporating New Zealand aggregates,
while the values in [28] are for concretes with European aggregates

• The values in [11] have been used to develop the table in HERA Report R4-82 [8] from which
ecθ is obtained

• Any inherent conservatism in the value of fcè


'
has minimal influence on the design capacity

CA4.1.5 Unprotected secondary beam components; temperatures reached and elevated temperature
mechanical properties

The unprotected secondary beam components are exposed to the fully developed fire and reach
correspondingly high temperatures. It is important, in developing this design procedure, to make a
suitable determination of the design temperatures for these elements. The process that has been
gone through to derive the design temperatures given in Table 71.3 is as follows:

Step 1: Determine the maximum average fire temperatures reached, at any time, from
experimental real fire testing, around the secondary beams over an area affecting more than one
beam.

The test used for this determination has been the Cardington Demonstration Furniture test, which
generated the highest temperatures of all such tests reported in the literature. This real fire test
involved a 135m2 enclosure, fire load of 46kg/m2 floor area (920MJ/m2), including 20% plastic, and
the position, extent of ventilation arranged so as to generate the maximum structural fire security
[17]. The fire generated peak gas temperatures of over 1100oC, with a heat release rate of 40MW.

For a period of around 10-15 minutes, gas temperatures reached a maximum average of around
1000oC over an area of the enclosure encompassing two supporting secondary beams. Average
temperatures over the enclosure as a whole remained below 1000oC at all times, while spot
temperatures reached just over 1100oC. See details in Fig. 71.36. The minimum area to which
this slab panel procedure would be realistically applied is to a square of length and width equal to
the secondary beam span, which is 60% of the Demonstration Test enclosure area. Therefore an
appropriate upper bound gas temperature to take, as being applied uniformly over the slab panel
area, for this fire, is 1000oC. This is used as the peak temperature for FHC 2 and light-weight
concrete (LWC). (The test fire load, at 920 MJ/m2 floor area (80% cellulose, 20% plastic),
corresponds to FHC 2 (800 MJ/m2 floor area) and the concrete used in the floor slabs was LWC,
density = 1900 kg/m3).

Step 2: Extend the result to normal weight concrete (NWC) and to FHC 1 and 3 fuel loads.

This was undertaken by running the fire model described in DCB No. 59 (see section 3, pages 8-11
therein) for enclosures with appropriate differences in thermal inertia and fire load. The different
values of enclosure thermal inertia used were b = 1200 J/m2s0.5C for the Cardington test which is
appropriate for the concrete density (1900 kg/m3) used, compared with b = 1700 J/m2s0.5C for NWC
construction. The different fire loads were 400MJ /m2 floor area for FHC1, 800 MJ/m2 floor area for
FHC2 and 1200 MJ/m2 floor area for FHC3.

HERA Steel Design & Construction Bulletin Page 64 No. 71, December 02/ January 03
These results showed the following:

• ∆θg = + 100ºC in going from NWC → LWC


• ∆θg = + 70ºC in going from FHC2 → FHC3
• ∆θg = - 50ºC in going from FHC2 → FHC1

These values were used to adjust θg = 1000oC for (FHC2, LWC) to the five other cases.

Step 3: Obtain the beam element temperatures from the gas temperatures.

The relationships used are taken from analysis of the Cardington test data and are as detailed in
HERA Report R4-90-DD-Rev 2 [12]. They are:

• beam bottom flange and web; θs = 0.95 x θg


• beam top flange; θs = θg – 150oC

The values so obtained were rounded to the nearest 50o C, to give the steel beam element
temperatures presented in Table 71.3.

These temperatures apply to the secondary beams away from their ends. At the beam ends
adjacent to the primary beams, even when the primary beams are unprotected, the temperatures
are consistently lower, with the beam web and bolt temperatures being slightly less than the top
flange temperatures given in Table 71.3.

Having obtained the design temperatures for the secondary beam elements, the next step is to
obtain elevated temperature mechanical properties. Given the high temperatures and strain levels
involved, it is important to use published values that are the most appropriate to this application.

The most accurate set of data is that from Poh [31], which covers stress/strain/time characteristics
for grade 250/300 steel beams, for temperatures up to 1200oC. However, given the development
of this procedure for design rather than as a research tool, it was considered desirable to use data
from an established standard. In this case, EC3 Part 1.2 [29] Table 3.1 was used to give
(fysbθ/fysb20) for each beam element. The results are given in Table 71.4.

CA4.1.6 Secondary beam to primary beam connections; bolts and secondary beam web

Even when the primary and secondary beams are both unprotected, the connection region is
cooler than the regions away from the connections [17]. This is allowed for in the procedure by
applying the beam top flange temperature given in Table 71.3 to the beam web and bolts at the
connection region.

Where the primary beam is protected, which will typically be the case, considerably lower
connection component and connector temperatures will occur, eg. as given in HERA Report R4-
DD-Rev 2 [12] Section 4.5.3.2. This will be ensured by the requirement, from section 4.5 and Fig.
71.11, to extend the protection through the connection region and slightly into the secondary beam.

The elevated temperature relationship for bolt tensile strength presented in equation 71.A10 is
based on UK test data, as described in section 4.3.3 of [12]. The data has been derived from tests
on UK high strength structural bolts (equivalent to our property class 8.8 bolts).

CA4.2 Design adequacy of slab panel; concrete slabs on profiled steel decking

CA4.2.1 Yieldline load-carrying capacity of slab panel

The minimum area of mesh required for integrity has been determined from the experimental
testing [3] and is given in section A4.3.

Determination of all yieldline moment capacities follows standard reinforced concrete theory [19].

HERA Steel Design & Construction Bulletin Page 65 No. 71, December 02/ January 03
For the positive moment capacities (mx and my), the concrete compression block (see Fig. 71.30) is
at the top of the slab and so is taken at ambient temperature concrete strength. The tension force
developed by all components is reduced by elevated temperature as previously described.

The unprotected secondary beam contribution within the slab panel to mx is treated as a line load,
by dividing the contribution from each beam by the beam spacing. This is sufficiently accurate
where there are at least three such beams within the panel length, Ly. (Fig. 71.26 shows five
beams, for example). Where there are only 1 or 2 beams, then this approach may overestimate
the secondary beam contribution and designer judgement will be required to determine if the
contribution needs modifying from that used herein.

While the tension/compression force distribution shown in Fig. 71.30 provides a realistic basis for
determining mx moment for yieldline capacity determination, the strain state leading to these
stresses under severe fire conditions is very different to that which would develop under ambient
temperature conditions. This is because of the complex effects of fire-induced heating on the slab-
secondary beam system. In practice, the unprotected secondary beams will be subject to rapid
heating, leading to compression pre-straining into the inelastic range due to restraint of thermal
expansion of the beam by the slab. This differential heating also causes the slab panel to deform
towards the fire, as shown in Fig. 70.4, DCB No. 70.

With increasing heating and weakening of the system, the buildup of compression strain in the
beam elements slows and then reverses, due to the influence of the applied loading moment-
induced tension on the beam elements. This generates tensile stress in the beam elements
associated with inelastic tensile straining. This condition occurs as the slab panel deformation is
increasing towards its maximum value and is the condition represented by the stress state
in Fig. 71.30. This is seen from the Speedfloor slab panel test reported in [3].

Thus the strain history of the secondary beam connected to the slab via shear studs and subject to
severe fire is complex, with a regime of inelastic compression strain development, followed by
inelastic tension strain development in the midspan region, where the loading-induced positive
moment effect is greatest.

With the slab itself, the initial period of rapid vertical deflection due to thermal effects causes the
slab reinforcement to go into tension. After around 20-30 minutes of standard fire exposure, the
tensile strains peak and reduce, then becoming compressive. This is despite the vertical
downwards deflection continuing to increase. See details in [3].

The unprotected secondary beam contribution included herein also assumes the following:

(i) The secondary beams are of uniform cross section. If beams of varying cross section along
their length are used, the secondary beam contribution must be amended (or, conservatively,
ignored). For tapered beams of triangular cross section (minimum at ends, maximum at centre),
use the dimensions at the third point along the beam span.

(ii) There are no openings in the webs of these beams.

If there are web openings and they are reinforced, then the web contribution can be determined
as given by equation 71.A11.3.

If there are web openings and they are unreinforced, then either the web contribution can
conservatively be ignored or a reduced web area used. This will only be required where the
openings are within the middle two quarters of the secondary beams (which is their typical
location).

Where more than one source of tension force development exists within the cross section, the
centroid of that force from the top of concrete must be determined. This is erx or ery, given by
equations 71.A12 and 71.A17, respectively. This value is used in determining the tensile
membrane enhancement arising from displacement of the centre of the slab panel relative to the
supports, in equations 71.A36 and 71.A37.

( )
For the negative moment capacity m 'x , the concrete compression block is on the fire-exposed
side and is therefore affected by elevated temperature. The loss of concrete strength due to

HERA Steel Design & Construction Bulletin Page 66 No. 71, December 02/ January 03
elevated temperature is accounted for, through the values of ecθ given in section A4.2.1 (3). The
concept is shown in Fig. 71.29 and the values given have been derived as follows:

(1) The maximum temperatures developed in the fire exposed concrete from all the natural fire
test cases given in DCB Issue No. 59 Table 59.1 have been determined. These
temperatures are listed in section A4.1.4 herein.

(2) The times taken to reach the same temperatures under standard ISO fire curve heating have
been determined; these times are also listed in section A4.1.4.

(3) The results from (1) and (2) have been checked for consistency with the experimental test
results.

(4) The tension force that can be developed by the maximum area of interior support bar
reinforcement has been determined, by taking this reinforcement at full yield stress.

(5) The depth of concrete in compression requied to balance this tension force, for the
appropriate standard fire test time, has been determined from HERA Report R4-82 [8]
Table 6. See Fig. 71.29 for this concept.

(6) The resulting value of eccentricity from that table is the value of ecθ used in the calculations
for m 'x . That value accounts directly for loss of concrete strength due to elevated
temperatures.

The diameter of mesh bar used for cold-drawn mesh to AS/NZS 4671 [24], ie. dmesh, ranges
from 5.3 mm to 11.2 mm.

Mesh cover is 30mm for slabs on profiled steel deck, as specified by NZS 3404 Commentary
Clause C13.2.2(a).

The additional bar reinforcement shown in Fig. 71.12 sits on top of the mesh reinforcement for
slabs on profiled steel deck.

The equation for yield-line load carrying capacity of a rectangular slab panel, equation 71.A22.1, is
from [19].

CA4.2.2 Yieldline capacity of simply-supported slab panel

The BRE method [2] has been developed for a simply supported slab panel, as it is assumed
(correctly) that cold-worked mesh reinforcement across any interior supports will fracture under the
negative curvatures developed. Hence the interior support bars are grade 500E deformed bars to
[24] rather than grade L mesh. The tensile membrane enhancement, given by equation 71.A42,
has been normalised with respect to the simply-supported load-carrying capacity [2] and it is
therefore applied only to that capacity, in equation 71.A43. Thus the simply supported load-
carrying capacity and the higher load-carrying capacity developed by including the interior support
bar contribution must both be determined.

CA4.2.3 Limiting deflection of slab panel

The limiting deflection equations are in the form of those proposed by Bailey [2], but modified as a
result of the analytical (DCB No. 59) and experimental fire testing [3] work undertaken. The first
term is the deflection due to thermal effects; the second term the deflection under applied loading.

The thermal effects are determined from the peak temperature difference between the top and
bottom of the slab. Bailey used 770oC, based on standard fire tests; we have used 770oC based
on the thermal analyses under natural fire conditions reported in DCB Issue No. 59. This value is
also consistent with the experimental test results [3].

Bailey comments [2] that, comparing the results of the method with the Cardington tests, there is a
“factor of safety” of around 2.4 in these equations, in terms of the limiting deflection allowed
compared with the deflection necessary to generate the full width crack at mid-span that is taken
[2] as the failure point for slab panel tensile membrane action. The slab panel experimental tests of

HERA Steel Design & Construction Bulletin Page 67 No. 71, December 02/ January 03
June / July 2002 [3] provided the opportunity to test a range of slab panels to the same failure point
as reached in the UK ambient temperature slab panel test reported in [9]. The key points that we
wanted to determine from these experimental tests were:

• Were the deflection limits given in DCB No. 60 appropriate?


• Were limits on slab panel behaviour set by the ambient temperature ductility capability of the
mesh required under severe fire conditions?
• What level of reinforcement is required for integrity?
• What is the behaviour over time and the final mode(s) of failure under severe fire conditions
compared with those determined for ambient temperature loading?
• At what time and at what deflection does the second edition of the SPM procedure predict
failure in relation to the experimental results?

All these questions were able to be answered from the experimental tests and the results of those
tests [3] have been used to modify the slab panel deflection limits as appropriate. To elaborate:

(i) The deflection term due to thermal gradient effects was consistent with the test results [3], so
that term (the first term in each equation) is unchanged from that in DCB No. 60, except for
replacing he with (t0 – 0.5hrc).
(ii) The second term in equation 71.A23.1 was originally based on developing a limiting strain of
0.5åy,20 in the longitudinal (y) direction [9]. This was a mesh ductility limit and was raised to
0.7åy,20 in DCB No. 60, based on a conservative assessment of the additional ductility
available from grade L mesh at temperatures of 300ºC to 700ºC. The slab panel tests [3]
confirmed this temperature range as realistic but showed the 0.7åy,20 limit to be still very
conservative. It has been raised to 0.9åy,20 which is still well within the capability of the grade
L mesh to develop in severe fire conditions.
(iii) The second term in equation 71.A23.2 was originally Lx/30, as recommended by Bailey [2].
This was also justified by the low ambient temperature ductility capability of the cold-drawn
wire mesh, as described in CA4.2.3 of DCB No. 60. However, the slab panel fire tests [3]
have shown the mesh to have considerably greater ductility in fire conditions, thus allowing
the Lx/30 limit to be raised to Lx/25.
(iv) The approach taken in determining refinements to these deflection limits has been as
follows:

• Determine what limits on deflection and load carrying capacity for the D147 flat slab,
which developed the full tensile membrane failure mechanism, put the full tensile
membrane failure point within the region of stable behaviour and before any observed full
depth cracks developed.
• Propose modifications to the deflection limits
• Determine the predicted failure point and compare with the experimental behaviour for
the 5 other test slab panels.
• Propose final changes once these steps are successfully completed

This process is described in section 4.4, p. 15 of DCB No. 70, and the results are shown in
Figs. 70.16 and 70.17 therein

When calculating ∆ max from equations 71.A23, it is found that the thermal and mechanical
contributions to deflection are generally similar. This means that the maximum vertical deflection of
the slab panel at CD relative to the primary interior beams could be Lx/13. The realistic worst case
in terms of negative rotation of the slab over the primary interior beams involves one slab panel at
full (thermal plus mechanical) deflection and the adjacent slab panel at full mechanical deflection
(having been subjected to fully developed fire and in the cool-down stage). Applying the same
approach for bar extension as that used above for the mesh shows that Grade 500E reinforcement
to AS/NZS 4671 [24], which has a minimum specified ambient temperature elongation of 10%, will
be able to accommodate this rotation without fracture, however grade 500L mesh bars will be
expected to fracture over the interior supports, as assumed in [2] and seen in [17].

CA4.2.4 Determination of tensile membrane enhancement

Equation 71.A24 is from [19] and is required to calculate dimension L1. For slab panels with
slightly flexible supports, as is the case in this instance, it is important to determine L1 based on the

HERA Steel Design & Construction Bulletin Page 68 No. 71, December 02/ January 03
critical variables mx, my, Lx, Ly, rather than assume a yieldline angle of intersection with the corners
of 45o [19].

The derivation of equations 71.A25 to 71.A42 is given in [2], except for notation changes to suit
New Zealand practice and the generalising of the terms to allow for multiple sources of tension
capacity in developing mx and my. An error from [2] in equation 71.A38 for e1bx , which is described
in DCB No. 66, p. 12, has been corrected.

Users should note that there will be instances where equation 71.A37 gives a negative value for
e2my . This is not an error; it simply shows that the enhancement due to displacement of the slab
panel centre relative to the supports has been overestimated for element 1 and this needs to be
compensated for.

The limits on tensile membrane enhancement proposed in DCB No. 66, pp. 10–11, have been
shown [3] not to be necessary and are removed in this second edition of the SPM procedure.

CA4.2.5 Check on moment/tension membrane adequacy

If the slab panel moment/tension membrane capacity is not adequate, options for increasing this
capacity include:

(1) Place deck trough bars to increase my.


(2) Increase the mesh reinforcement, noting the restriction on mesh pitch and grade in the
accompanying article.
(3) Increase the size of the supporting unprotected secondary beams
(4) Increase the concrete thickness and mesh cover to increase mx, my (however this also
increases w*).

See section 10 on page 14 herein for further guidance on obtaining a cost-effective solution.

CA4.2.6 Determining the shear capacity of the slab panel

(1) Shear capacity through the slab.

There are five points here to note in the commentary;

(i) As shown in Fig. 71.32, the slab shear check is undertaken along the primary support
beams rather than the secondary support (edge) beams. There are two reasons for
this. First is that the depth of concrete resisting shear is greater onto the secondary
beams (the average depth, he, rather than cover slab depth, to – hrc). Secondly, the
decking itself contributes to the shear resistance onto the secondary edge beams, as
the webs of the deck can develop some shear resistance, even at elevated
temperature, in that direction.

(ii) The loading on the secondary edge beams is triangular rather than uniform, making
the primary support beams more uniformly loaded and more critical for shear.

(iii) The depth of concrete, dv, used to resist shear is measured from the effective extreme
compression fibre of concrete to the centroid of longitudinal tension reinforcement
[11]. Over the supports, the centroid of the longitudinal tension reinforcement can be
taken as the centroid of the interior support bars, or edge bars, as appropriate. These
sit on top of the mesh and so increase dv. However, the extreme compression fibre is
not at the fire exposed concrete face, because of strength loss due to heating. For the
concrete temperatures involved, the effective extreme compression fibre can be taken
as 0.67ecθ from the heated face of the concrete. This is derived from [8] and the
determination of ecθ.

(iv) For slabs, the minimum concrete contribution is set by NZS 3101 Clause 9.3.2.2 as
0.17 fc' . For a flat slab or a slab on profiled steel deck, this minimum value will
apply for any practiable level of tension reinforcement, so the concrete slab shear
contibution is independent of the actual level of slab bar reinforcement supplied at the
supports.

HERA Steel Design & Construction Bulletin Page 69 No. 71, December 02/ January 03
(v) The strength reduction factor for concrete shear in fire, φfire = 0.75/0.85 = 0.89. This
comes from applying φfire = (φ/0.85) ≤ 1.0, in conjunction with φc = 0.75 for shear in
concrete, from NZS 3101 [11].

(2) Shear capacity available through the secondary beam to primary beam connection.

The slab and secondary beam are an integral unit, so for the slab to suffer a shear failure at
the primary beams, the secondary to primary beam connection will also have to undergo a
shear failure.

The minimum elevated temperature shear capacity at peak heating can be realistically
assessed as the minimum elevated temperature capaicty of either the bolt group in single
shear or the beam web in shear. Both modes are checked and the minimum capacity
selected; see equations 71.A47.1 to 71.A47.3; see equations 71.A47.1 to 71.A47.3. Where
the supporting primary beam is protected, the bolts connecting the secondary beam to
primary beam will also be protected, as described in section 4.5 and Fig. 71.11. However,
the reduced capacity of the unprotected beam web, given by equation 71.A47.2, will still
apply.

(3) Shear resistance during cool-down.

During the cool-down period, the sides of the slab panel, including the secondary beam to
primary beam connections, are subjected to increasing tension force. However, the
mechancial capacity of these components is also increasing as their temperature decreases.

The shear resistance and integrity will be maintained during this period provided that:

(i) The connections can accommodate the tension force without fracture of the bolts or
connection components, and

(ii) The ability to transfer longitudinal tension force through the slab panel edges into their
supports is not lost.

Connections designed and detailed to HERA Report R4-100 [6] are specifically configured to
avoid fracture under either inelastic rotation or fire-induced tension under cool-down; thus
suppressing failure from (i). One of the important functions of the edge and the interior
support bars is to suppress failure from (ii).

CA4.2.8 Check on shear adequacy

Obviously, option (1) is not possible in such instances as where the secondary beams are deeper
than the primary beams, unless the top of steel of each beam is at suitably different levels.

CA4.3 Mesh area required for the preservation of integrity

See DCB No. 70, section 4.5, pp. 15–16, for the background to these requirements.

CA5 Commentary on Slab Panel Design Method; Floors Incorporating In-Situ Concrete Slabs
on Speedfloor Joists Supported on Primary Beams

Application of the slab panel method to concrete slabs on Speedfloor joists supported on primary
beams follows much the same procedure as for concrete slabs on steel deck supported on
secondary and primary beams. The floor system layout is as shown in Fig. 71.33; the additional
reinforcement shown in Fig. 71.12 also applies.

Minor differences in terms of the position of mesh within the slab and position of additional bar
reinforcement are due to physical differences in the systems, in particular the protrusion of the top
of the Speedfloor joists into the slab and the manner in which the slab is cast. For those unfamiliar
with the Speedfloor system, refer to these details in [14].

There are three differences between the Speedfloor system and a composite floor system including
a concrete slab cast onto steel decking. These are now addressed insofar as they effect the design
procedure:

HERA Steel Design & Construction Bulletin Page 70 No. 71, December 02/ January 03
1. Preservation of integrity.

The lack of a steel deck to seal any through-depth cracking has an influence and the minimum
mesh area required to preserve integrity. This is covered in section A5.3, with the commentary
given in DCB No. 70, section 4.5, pp. 15–16.

2. Contribution of Speedfloor joists to moment capacity in the x-direction

In the first edition of the SPM design procedure, the contribution of the Speedfloor joists to the
development of mx was not included, however an indication of how this might be changed in a
failure revision was given in the commentary.

The experimental tests [3] showed that the Speedfloor joists do make a significant contribution to
the slab panel strength and stiffness, as described in section CA4.1 and Fig. 71.34. They also
provided sufficient data to allow the contribution from some of each joist, namely the elements
embedded in the concrete slab, to be included in the determination of mx. This has been done, in
section A5.1.3. A description of the process used is given in that section.

3. Resistance of shear at primary beams

The Speedfloor joists are not assumed capable of developing any shear resistance under severe
fire conditions and thus assisting in developing shear capacity at the slab panel to primary beam
supports. This is different to a conventional secondary beam and would be a worthwhile topic for
further research. The slab panel tests [3] did not provide any new data to enable changes to be
made in this regard.

CA5.3 Mesh area required for the preservation of integrity

See DCB No. 70, section 4.5, pp. 15–16, for the background to these requirements.

CA6 Design of Supporting Beams

CA6.2 Application

Note that the positive moment capacity of the supporting beams is the composite moment capacity,
Mrc.

When determining the adequacy of the supporting beams, two options are given. These are to suit
the beam conditions that will be encounted in practice.

The first option (option 3.1) involves applying passive fire protection in accordance with traditional
practice. This will be required for support beams, which are simply supported and designed to
carry gravity loading only.

The second option (option 3.2) is intended for lightly loaded beams, especially those that are part
of seismic-resisting systems and which therefore have a high reserve of strength in fire. It takes
into account the cooling effect which occurs at the edge of an enclosure, irrespective of whether
that edge is adjacent to openings or not. This effect is seen in all the Cardington enclosure tests
[17] and is quantified for design purposes, conservatively, in [37]. That recommendation is used in
step 3.2.2 herein.

The determination of moment adequacy through a plastic collapse mechanism check, in step 3.2.4,
follows established theory. Further guidance on applying this to different beam support conditions
can be found in section 6.1 of HERA Report R4-82 [8].

For beams with rigid connections, elevated temperature shear capacity is satisfied without the need
for an explicit check. With semi-rigid connections, this will also typically be the case; with simple
connections a check from first principles will be needed, using the minimum of the bolt group
elevated temperature shear capacity and the beam web elevated temperature shear capacity.

HERA Steel Design & Construction Bulletin Page 71 No. 71, December 02/ January 03
References (13) Barber, DJ; HERA Fire Protection Manuals
Sections 7 and 8, Passive / Active Fire
Protection of Steel; HERA, Manukau City,
(1) AS/NZS 2312:2002, Guide to the Protection
1996, HERA Report, R4-89.
of Structural Steel Against Atmospheric
Corrosion by the Use of Protective Coatings;
(14) Speedfloor Design Manual; Speedfloor
Standards New Zealand, Wellington.
Holdings Ltd, Auckland, 2001.
(2) Bailey, CG; Design of Steel Structures With
Composite Slabs at the Fire Limit State; UK (15) NZS 4203:1992, General Structural Design
Building Research Establishment, Watford, and Design Loadings for Buildings;
England, 2000, Report No. 81415. Standards New Zealand, Wellington, New
Zealand.
(3) Lim, L and Wade, C; Experimental Fire
Tests of Two-Way Concrete Slabs; (16) Clifton, GC and Forrest, E; Notes Prepared
University of Canterbury School of for a Seminar on Design of Steel Buildings
Engineering, Christchurch, 2002, Fire for Fire Emergency Conditions; HERA,
Engineering Research Report 02/12. Manukau City, 1996, HERA Report R4-91.

(4) NZBC:1992, New Zealand Building Code; (17) Kirby, BR; The Behaviour of a Multi-Storey
Building Industry Authority, Wellington. Steel Framed Building Subject to Fire Attack
- Experimental Data; British Steel Swinden
(5) C/AS1: 2001, Acceptable Solution for Fire Technology Centre, United Kingdom, 1998.
Safety; Building Industry Authority, Also data from BRE, Cardington, on the
Wellington, New Zealand. Corner Fire Test and Large Compartment
Fire Test, 1996.
(6) Hyland C; Structural Steelwork Connections
Guide Incorporating Amendment No. 1; (18) Feeney, MJ and Buchanan, AH; Accounting
HERA, Manukau City, New Zealand, for Sprinkler Effectiveness in Performance
1999/2001, HERA Report R4-100. Based Design of Steel Buildings; University
of Canterbury, Christchurch, 2000, Report
(7) Buchanan, AH (Editor); Fire Engineering No. 2000/15.
Design Guide, Second Edition; Centre for
Advanced Engineering, University of (19) Park, R; Ultimate Strength Design of
Canterbury, Christchurch, 2001. Reinforced Concrete Slabs, Volume 2;
University of Canterbury, Christchurch, 1970
(8) Barber, DJ; Calculation of the Fire (approx).
Resistance of Composite Concrete Slabs
With Profiled Steel Sheet Under Fire (20) Design Capacity Tables for Structural Steel,
Emergency Conditions; HERA, Manukau Third Edition, Volume 1: Open Sections;
City, 1994, HERA Report R4-82. Australian Institute of Steel Construction,
Sydney, Australia, 2000.
(9) Bailey, CG; The Tensile Membrane Action of
Unrestrained Composite Slabs Simulated (21) Clifton, GC and Feeney, MJ; Fire
Under Fire Conditions; UK Building Engineering Application to Multi-Storey Steel
Research Establishment, Watford, England, Structures; The Inaugural New Zealand
2000, paper accepted for publication in Metals Industry Conference, Rotorua, 2002,
Engineering Structures. Paper No. 14; HERA, Manukau City, 2002.

(10) NZS 3404: 1997, plus Amendment No. 1: (22) Clifton, GC et.al.; Design of Multi-Storey
2001, Steel Structures Standard; Standards Steel Framed Buildings With Unprotected
New Zealand, Wellington, New Zealand. Secondary Beams or Joists for Dependable
Inelastic Response in Severe Fires;
(11) NZS 3101:1995 (Including Amendment Nos. Proceedings of the Second International
1 and 2, 1997), Concrete Structures Workshop on Structures in Fire, 2002;
Standard; Standards New Zealand, University of Canterbury, Christchurch,
Wellington. 2002, pp. 151 – 174.

(12) Clifton, GC et. al.; Draft for Development: (23) Hyland, C; Structural Steelwork Estimating
Revision 2: Design Procedure for the Guide; HERA, Manukau City, 1998, HERA
Inelastic Floor System/Frame Response of Report R4-96.
Multi-Storey Steel Framed Buildings in Fully
Developed Natural Fires; HERA, Manukau
City, 2000, HERA Report R4-90-DD-Rev 2.

HERA Steel Design & Construction Bulletin Page 72 No. 71, December 02/ January 03
(24) AS/NZS 4671: 2001, Steel Reinforcing (37) Newman, GM et.al; Fire Safe Design – A
Materials; Standards New Zealand, New Approach to Multi-Storey Steel Framed
Wellington. Buildings, The Steel Construction Institute,
Ascot, England, 2000, SCI Publication
(25) Traydec 300 Specification; Forgan Jones P 288.
Structural Ltd, Silverdale, 1996.

(26) EC1-1-2/59:2001 (Third Draft) : Eurocode 1


– Actions on Structures Part 1-2: General
Actions – Actions on Structures Exposed to
Fire; CEN, Brussels, Belgium.

(27) AS 3600, 2001, Concrete Structures;


Standards Australia, Sydney, Australia.

(28) DD ENV 1992-2 : 1996, Eurocode 2: Design


of Concrete Structures, Part 1.2 General
Rules – Structural Fire Design (together with
United Kingdom National Application
Document); BSI Standards, London,
England.

(29) DD ENV 1992-2 : 1996, Eurocode 3: Design


of Steel Structures, Part 1.2 General Rules –
Structural Fire Design (together with United
Kingdom National Application Document);
BSI Standards, London, England.

(30) Hi-bond Design Manual; Dimond Structural,


Auckland, 1999, Manual No. 7.

(31) Poh, KW; Modelling Elevated Temperature


Properties of Structural Steel; BHP
Research, Melbourne, Australia, 1996,
Report BHPR/SM/R-055.

(32) Clifton, GC and Robinson, J; Notes


Prepared for a Seminar on The Behaviour
and Design of Multi-Storey Steel Framed
Buildings for Severe Fires, Revised June
2001; HERA Manukau City, 2001, HERA
Report R4-105.

(33) Thomas, IR et. al.; Fire Tests of the 140


William Street Office Building; BHP Co Ltd,
Melbourne, Australia, 1993.

(34) Beck, C and Clifton, GC; SPM0103 Design


Program, HERA, Manukau City, 2003.

(35) Gib® Fireboard Structural Steel Fire


Protection; Winstone Wallboards Ltd,
Auckland, 1995.

(36) Lim, L et.al.; Experimental Testing and


Numerical Modelling of Two-Way Concrete
Slabs Under Fire Conditions; Journal of the
Structural Engineering Society, Vol. 15, No.
2, 2002, pp. 12-26.

HERA Steel Design & Construction Bulletin Page 73 No. 71, December 02/ January 03

You might also like