Structural Use of Glass in Buildings (Second Edition) by Chris O'Regan Beng (Hons) Ceng Mistructe

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

Structural Use of Glass in Buildings (Second Edition)

By Chris ORegan BEng(Hons) CEng MIStructE


Worked example for floor plate design
A glass floor plate is to be installed that sits over a basement. The support is via a steel frame with a 1.5m x 1.5m grillage of
beams. This frame provides a continuous support and the following analysis and design of the glass floor plate provisionally
assumes it to be infinitely stiff. The following calculation determines the thickness of the glass required for the floor plate.

The below floor loadings are typical for domestic use and should be used for design:
Characteristic variable action qk = 1.5kN/m2
Characteristic point action Qk= 3.0kN
The glass will not be sandblasted and will be enamelled on the wearing side. This will not impact on the design of the glass
as the enamelling is on the compression side of the floor plate.
The glass is to be made up of three layers of glass laminated with PVB, It is assumed that the top sheet will be toughened
and not included in the stress calculations although will be considered for deflection. To provide the most effective post
breakage behaviour the lower two sheets will be heat strengthened.
Two action durations apply: permanent and short.

Permanent action condition


Determine design strength of glass using Appendix C.
fg;d =

kmodksp fg;k
M:A

kv(fb;k - fg;k )
M:v

For load duration > 50 years, kmod = 0.29

kmod = 0.29, ksp = 1.0,fg;k=45 N/mm2 fb;k = 70N/mm2, kv = 1.0,


fg;d =

M:A = 1.6 and

M:v

= 1.2.

0.29 1.0 45N/mm


1.0(70N/mm - 45N/mm )
+
= 29N/mm2
1.6
1.2

Variable action condition i.e. short duration load


For load duration of 5 hours, for pedestrian action; kmod = 0.60
fg;d =

0.60 1.0 45N/mm 1.0(70N/mm - 45N/mm )


+
= 37.7N/mm2
1.2
1.6

Determine the effective thickness of the glass required based on design strength

Page | 1

Structural Use of Glass in Buildings (Second Edition)


By Chris ORegan BEng(Hons) CEng MIStructE
Try two 12mm thick plies and one 12mm thick sacrificial ply. The laminate is a PVB based material and is 0.76mm thick.
In order to calculate the effective thickness of the glass pane, the shear interaction between the plies of the laminated glass
via the PVB interlayer must be accounted for.
The simplified method of determining stress of glass within each ply is as follows.
The effective thickness of the glass when considering deflection due to bending is:

hef;w =

where:
hk
hm,k

3
hk + 12

hk hm,k

is the thickness of each ply


is the distance between the middle of the ply to the centre of the laminated glass pane.

When calculating stress the equivalent thickness is calculated from:

; ;

where:
hj
hm,j

+2

is the thickness of each ply


is the distance between the middle of the ply to the centre of the laminated glass pane.

Assuming a value of 0.1 for when considering variable actions and 0 when assessing effects due to permanent actions,
two effective thicknesses need to be calculated.

Effective thickness in permanent action condition


The effective thickness of the laminated glass with respect to permanent action is as follows:
hef;w =

123 mm + 123 mm = 15.1mm

Note that the variable equates to 0 hence it is not included in the calculation. Additionally the toughened glass top ply is
not included when determining the overall depth of the glass as it is considered to be sacrificial.

Effective thickness in Variable Action Condition


The effective thickness of the laminated glass with respect to variable action is as follows:
3

hef;w =

12 mm + 12 mm + 12 0.1 (12mm
= 16.7mm
6.42 mm + 12mm 6.42 mm)

This is the effective thickness of the laminated glass that is to be used for deflection calculations.
The effective thickness of the laminated glass for bending stress is as follows:

; ;

16. 7 mm
= 18.7mm
12mm + 2 0.1 6.4mm

Note how the thickness has increased from 15.1mm to 16.7mm due to the inclusion of shear interaction within the PVB
based interlayer, which is then increased to 18.7mm when calculating stress

Page | 2

Structural Use of Glass in Buildings (Second Edition)


By Chris ORegan BEng(Hons) CEng MIStructE
Bending stress check
There are two action conditions to check for: permanent and variable (short). With respect to the permanent condition, the
following applies:
Permanent Actions due to self-weight = 0.036m x 25kN/m3 = 0.9kN/m2
(Note this includes the 12mm thick toughened glass top ply which is not included when determining the effective thickness
of the glass.)
Partial factor for permanent action

q=1.35

UDLult = 0.9kN/m2 x

x 1.35 = 1.22kN/m2

g = 0.9kN/m2

The applied bending stress to the plate during the permanent condition is defined in Table 11.4 of Roarks Formulas for
Stress and Strain 8th Edition as:

max

qb2

= 0.287

The variable is dependent upon the ratio of geometry of the plate. In this instance the ratio of the dimensions of the plate
is 1; therefore = 0.287

max

0.287 1.22kN/m 1.52 m


0.01872 m

2.3N/mm2 < 29N/mm2 OK

= 2.310 kN/m2 = 2.3N/mm

The applied bending stress to the plate during the variable action condition is:
1.22kN/m2 + 1.5 kN/m2 x

= 1.22kN/m2 + 1.5 kN/m2 x 1.5 = 3.47kN/m2

There are two effects to consider for the variable action condition: that of the UDL and the point load.

UDL
max =

0.287 3.47kN/m 1.52 m


= 6.410 kN/m2 = 6.4N/mm
0.01872 m

6.4N/mm2 < 37.7N/mm2 OK

Concentrated point action


The area over which the action is applied is 50mm x 50mm = 2500mm2
equivalent circle area radius ro =r if r

Bending stress at point of concentrated load:

0.5t

=
Where:

W is the design concentrated point action


is Poissons Ratio

Page | 3

=
3

; ;

2500mm

(1 + ) ln

= 28.2mm

Structural Use of Glass in Buildings (Second Edition)


By Chris ORegan BEng(Hons) CEng MIStructE
is a variable that is dependent upon the ratio of geometry of the plate. See Table 11.4 of Roarks Formulas for Stress and
Strain 8th Edition.
In this instance the ratio of the dimensions of the plate equate to 1, therefore =0.435
Partial factor for variable concentrated action
point load =

Q=1.5

3 3000 1.5
2 1500mm
+ 0.435 = 29.1N/mm2
(1 + 0.22) ln
2 18.72
28mm

including self-weight

oa

=29.1N/mm2 + 2.3N/mm2=31.4 N/mm2

31.4N/mm2 < 37.7N/mm2 OK

Deflection check
It Is safe to assume all three panels of glass participate in restricting deflection
Equivalent thickness of 3 layers is as follows:
3

hef;w =

Equivalent I value =

123 mm + 123 mm + 12 mm
12.82 mm + 12mm 0 mm

12 0.1 (12mm
= 21.5mm
12mm 12.8 mm)

= 1.24 10 mm

Taking the point load as the worst case condition:


Serviceability state deflection =

3000N 1500 mm
= 2.4mm
48 70000N/mm 1.24 10 mm

Assuming a span/depth ratio of 250, the maximum allowable deflection is 6mm > 2.4mm therefore the floor plate passes
both serviceability and strength checks.

Post-failure check
As all three sheets of glass could fail the only truly reliable test for the proposed design is to undergo physical testing of the
design to confirm its suitability.
Some schools of thought suggest a design method where one sheet is assumed to remain intact. This can offer some
measure of comfort but does not cover the case where all three sheets are broken. This approach is illustrated below.
Two action durations apply: permanent and short.

Permanent action condition


The strength of the glass is the same for the intact floor plate.

Bending stress check


In the post-failure condition there remain the two action conditions to check for: permanent and variable (short). With
respect to the permanent condition, the following applies:
Actions are similar to intact condition i.e. 0.9 kN/m2
Partial factor for permanent action g=1.0 due to accidental condition

Page | 4

Structural Use of Glass in Buildings (Second Edition)


By Chris ORegan BEng(Hons) CEng MIStructE
UDLult = 0.9kN/m2 x

g = 0.9kN/m2 x 1.00 = 0.9kN/m2

When taking into account the two way spanning action of the floor plate, the applied bending stress to the plate during the
permanent condition is:
=

qb2

, = 0.287
h2ef:w
0.287 0.9kN/m 1.52 m
= 4.010 kN/m2 = 4.0N/mm
=
max
0.0122 m

max

4.0N/mm2 < 29.0N/mm2 OK

The applied bending stress to the plate during the variable action, short duration condition is:
Partial factor for varible action

q=1.0 due

UDLult = 0.9kN/m2 + 1.5kN/m2 x

to accidental condition

q = 0.9kN/m2

+ 1.5kN/m2 x 1.0 = 2.4kN/m2

There are two effects to consider for the variable action condition: that of the UDL and point load.

UDL
max =

0.287 2.4kN/m 1.52 m


0.0122 m

including self-weight

oa

= 10.810 kN/m2 = 10.8N/mm

=10.8N/mm2 + 2.3N/mm2=13.1 N/mm2

13.1N/mm2 < 37.7N/mm2 OK

Point load check


Although this is an accidental condition, the point load check still applies.
Partial factor for variable concentrated action

concentrated action

3 3000 1.0
2

2 12 mm

including self-weight

oa

Q=1.0

(1 + 0.22) ln

2 1500mm
+ 0.435 = 47.1N/mm2
28.2mm

=47.1N/mm + 2.3N/mm =49.4 N/mm

49.4N/mm2 > 37.7N/mm2 fails


If the two bottom sheets remain as 14 mm and the top sheet is dropped to 8mm then the overall thickness will remain the
same and the stress calculated above will reduce to 36.9N/mm2, which is within the allowable design bending stress.
No deflection check required as it is a post-failure condition.
The overall thickness of the glass is 37.5mm, which correlates closely to the span/depth ratio of 40.

Page | 5

You might also like