Composite beams
NSC 23
Technical Digest 2018
Case Methodology (90 minutes of fire exposure) θc,top Kc
1 Room Temperature 20 1.00
2 EN 1994-1-2 Annex D (heff = 100 mm) 160 0.97
3 BS 5950-8 with EN 1994-1-2 Annex D (heff ≥ 100 mm) 160 0.97
4 Medium value according to NCCI (weighted average) 224 0.93
5 Ignoring Ribs According to EC (heff = 70 mm) 246 0.90
6 Ignoring Ribs According to BS 5950-8 (heff = 70 mm) 260 0.89
7 Ignoring Ribs According to NCCI (heff = 70 mm) 244 0.91
8
Assuming 40% of steel top flange temperature (θtop flange = 620°C) EN 1994-1-2,
4.3.4.2.5 (2) – for shear studs resistance.
for assessing the sagging bending resistance of composite slabs and beams
is suggested, as, in general, only a modest depth at the top of the slab will
be necessary to obtain section equilibrium at elevated temperature. Thus,
an example assuming room temperature in the slab will also be considered
(note that if floor screed is considered for the minimum insulation thickness,
the temperature in the top concrete fibre can be slightly higher).
For 90 minutes of fire exposure, the minimum insulation thickness
according to EN 1994-1-2 Annex D would be heff ≥ 100 mm (note that the
profile falls outside the scope of Annex D of EN 1994-1-2, which limits l3 to
115 mm, compared to the actual value of 125 mm). According to the NCCI, a
minimum thickness of h1 ≥ 70 mm is imposed, while BS 5950-8 suggests h1 ≥
70 mm for lightweight concrete and h1 ≥ 80 mm for normal weight concrete.
In Figure 2, for 90 minutes of fire exposure, the different temperature
distributions in the concrete flange according to the three different UK
resources can be found for normal weight concrete. Slab depth is measured
from the face exposed to fire.
For 90 minutes of fire exposure, the temperatures in the top (Table 2)
and the bottom (Table 3) fibres of the concrete flange above the steel sheet
can be obtained (i.e. X = 130 mm, and X = 60 mm, respectively, according
to Figure 1). Eight possible
approaches are presented. Once the
temperatures have been obtained,
the respective concrete resistance
reduction factors (Kc) according
to Table 3.3 of EN 1994-1-2 can
be obtained. In the top concrete
fibres, according to EN 1994-1-2
and BS 5950-8 approaches, the
top temperature is in fact close to
140°C (Cases 2 and 3). Even with
conservative approaches (Cases 5,
6 and 7), the temperature in the top
concrete fibre is generally below
250°C, so no concrete strength
reduction would be needed for the
top concrete fibres. On the other
hand, for lower concrete fibres,
the strength reduction can be up
to 29 % for Cases 2 and 3 and 83
% for Case 6. Thus, depending of
the depth of the concrete flange
required for section equilibrium, the
concrete resistance may have some
significant reductions.
To evaluate the impact of
different temperature distributions
in the slab, the critical steel
temperatures shown in Table 1
were assumed as fixed. The plastic
bending resistance under fire, for
each slab profiles temperatures
(Cases 1 to 8) were then evaluated,
and are presented in Table 4 and
Table 5 (overleaf ) for the two
worked examples. The degree
of shear connection (η) can vary
between 0 and 1 in a composite
beam. Results for different degrees
of shear connection are presented
in steps of 0.25 between those two
extreme cases, obtained through
a stress block analysis. Partial
interaction curves are presented for
both worked examples in Figure 3,
for 6 m and 12 m worked examples.
Conclusions
1. The UK NCCI gives temperature profiles at/above ribs and between ribs for
composite slabs; in the paper, a weighted average temperature is suggested to
assess the sagging bending resistance of the composite beams design under fire.
2. The temperature distribution profile in the composite slab has generally
minimal impact in the composite beam sagging plastic bending resistance
because: (i) only the top concrete strips are usually needed to obtain
section equilibrium, which are not significantly affected by the slab
temperature; (ii) differences in the position of the plastic neutral axis are
usually small between the approaches; (iii) as the concrete flange tends to
be more resistant at elevated temperature than the steel, even if the slab
temperature is actually higher than considered, only small changes in the
neutral axis are expected, as a small increase in the assumed slab depth
increases considerably the slab resistance.
3. For assessing the resistance of the slab, generally no reduction in
strength is needed (ambient temperature may be assumed). An alternative
often used, which is to assume the slab temperature is equal to 40% of the
steel top flange temperature (a rule used to assess studs resistance under
fire), can be seen as a conservative solution.
Figure 2 - Temperature distribution according to different UK resources.
248 0.90
Table 2 – Top concrete fibre temperature according to different approaches (X = 130 mm).
Case Methodology (90 minutes of fire exposure) θc,top Kc
1 Room Temperature 20 1.00
2 EN 1994-1-2 Annex D (heff = 100 mm) 428 0.71
3 BS 5950-8 with EN 1994-1-2 Annex D (heff ≥ 100 mm) 430 0.71
4 Medium value according to NCCI (weighted average) 559 0.51
5 Ignoring Ribs According to EC (heff = 70 mm) 738 0.24
6 Ignoring Ribs According to BS 5950-8 (heff = 70 mm) 790 0.17
7 Ignoring Ribs According to NCCI (heff = 70 mm) 747 0.23
8
Assuming 40% of steel top flange temperature (θtop flange = 620°C) EN 1994-1-2,
4.3.4.2.5 (2) – for shear studs resistance.
248 0.90
Table 3 – Bottom concrete fibre temperature according to different approaches (X = 60 mm).