Вы находитесь на странице: 1из 153

Orion Training

Flat Slabs
Copyright CSC(UK) Ltd. 2006
Flat Slabs
CONTENT
Introduction
Verification Model
Comparison with the Code Approach
FE Analysis and Design Process
Initial Sizing
Adjusting model properties for deflections
Verifying analysis results
Determining As requirements
Checking Deflections
Moment Transfer at Edges
Punching Shear
Flat Slabs - Introduction
Why Flat Slab and why FE?
Perhaps you can tell us, is it:
For flat slab:
Reduced Structural Zones?
Potential for Fast Formwork Erection and Construction?
Un-obstructed service zones?
For FE:
A potential to accommodate more irregular column
positioning?
Only way to deal confidently with more complex isolated
load conditions?
Only way to deal confidently with significant openings?
Flat Slabs - Introduction
Why Flat Slab and why FE?
Whatever the reasons, we are certainly
seeing more and more demand for an
FE based flat slab solution within the
context of a building modelling
environment.
Orion Training
Orion Training
Flat Slabs
Verification Model
Copyright CSC(UK) Ltd. 2006
Flat Slabs Verification Model
Where should we start?
We are quite often faced with questions about highly
irregular layouts where the engineer is questioning
the results they are not what he expects, but he
has no strong basis for suggesting what the correct
results should be.
High Hogging Moments ?
Sagging Moments Small (What about Pattern Loading ?)
SO - The place to start has to be with a regular
structure where we can compare the expected results
based on longstanding and proven code idealisations
with the results from FE analysis.
The Example Model
A regular model.
5 bay by 5 bay.
8m bays throughout.
Top Floor columns
below only 400square
section.
300mm thick slab.
G = 8.7kN/m
2
Q = 5.0kN/m
2
Ult Load = 20.18kN/m
2
Flat Slabs Verification Model
The following accepted techniques can
be run on this model to give us some
expected results
The Yield Line Method
The Code Approach (BS8110)
Flat Slabs Verification Model
The Yield Line Method
The basic expectation is that yield line
approach needs less reinforcement than is
required using an FE approach. It should be
more efficient where structural regularity
allows it to be applied.
Although FE is conservative in comparison, it
can deal with more irregular cases.
Flat Slabs Verification Model
A Yield Line Calc is available in CSCs TEDDS
software.
Flat Slabs Verification Model
The Code Approach
(BS8110)
The code provides alternative simplified
methods of dealing with something this
regular.
The simplest method is outlined in cl 3.7.2.7
and allows strip design moments to be
determined using the coefficients in table
3.12.
The more complex option requires an
idealised frame analysis, more tailored to the
geometry of each specific building - we will
only look at this option in this example.
The Code Approach
(BS8110)
This method requires you to idealise and
analyse a flat slab structure:
The idealisation depends on being able to identify
column strips and middle strips.
You analyse the sub-frame as you would any
other beam and column sub-frame model.
You need to cater for pattern loads or apply
moment re-distribution to the results based on the
case with all spans fully loaded.
You proportion the resulting forces between
column strips and middle strips in an empirical
fashion.
The Code Approach
(BS8110)
The Analysis Model
Clause 3.7.2.2 indicates the idealised 2D frame analysis should be
based on gross section properties of the columns and slab strips
(where strip width equals full panel width - 8m in this case). The
strips are loaded with the full slab load.
We can analyse this simple frame in any general analysis package,
e.g. S-Frame.
The Code Approach
(BS8110)
The Basic Analysis Results
The results for the case with all spans fully loaded are
shown above.
Since these are from a linear elastic analysis, clause 3.7.2.1
(refers you to 3.5.2.3) applies and a 20% redistribution from
supports to the spans is required.
The Code Approach
(BS8110)
Results after Redistribution
627
664
550
846
877
73 Mmts after
redistribution
Percentage
change
Mmts before
Redistribution
+36%
-20%
+52%
-20%
+15%
-20%
461
830
363
1057
763
91
Span 3
Suppt 3
Span 2
Suppt 2
Span 1
Suppt 1
The Code Approach
(BS8110)
BS8110 Table 3.18 indicates how
design moments should then be
proportioned between Column Strips
and Middle Strips.
Clause 3.7.3.1 also requires the central
half of the column strip be designed for
2/3 of the hogging moment.
Distribution of Moments Explained
The Code Approach
(BS8110)
Distribution of Design Moments
55%
86kNm/m
50%
166kNm/m 55%
76kNm/m
50%
212kNm/m 55%
120kNm/m
Special
provisions
apply
Column Strip
Central 2m
Ditto
25%
83kNm/m Ditto
25%
106kNm/m Ditto
0 Column Strip
2 Edges (2m)
627kNm
664kNm
550kNm
846kNm
877kNm
73kNm Design Mmts
after redn
Middle Strip
4m wide 45%
71kNm/m
25%
42kNm/m 45%
62kNm/m
25%
53kNm/m 45%
99kNm/m
0
Span 3
Suppt 3
Span 2
Suppt 2
Span 1
Suppt 1
The Code Approach
(BS8110)
Relative magnitudes of design moments (after
moment redistribution and proportioning).
Sagging Moments
These vary from span to span highest moments in the
end spans.
For this regular layout, internal span moments about
70% of the end span moments.
Hogging Moments
These intensify rapidly as you approach the support (4
times higher than in middle strips)
We can utilise these expectations when determining
reinforcing requirements in FE.
Comparisons with FE
We have established some expected results using
the codes idealised approach.
We can now create FE models and make
comparisons
The FE Approach
Creating the FE Model
In accordance with published recommendations,
stiffness adjustments should be applied so the
serviceability limit state can be considered.
This can be achieved by applying a factor of 0.25 to
the short term elastic modulus, E.
Using this method the slab stiffness multiplier can be
left as 1.0.
Creating the FE Model
Include Column and Wall
Sections :
If not included, the slab is
meshed up to the column
centreline.
If included, the slab is
meshed up to the physical
perimeter of the column
and a frame of rigid
elements is created within
this boundary.
Example Comparing FE Modelling Options
Meshing
Try to aim for 6-8 segments between
column/support positions.
For a simple model like this the suggested default
of 2500 plates produces 8 segments.
Hogging Moments
Simplified Model More Sophisticated Model
Md1-top: Md1-top:
Max: 434kNm/m Max: 297kNm/m
The FE Approach
Sagging Moments
Simplified Model More Sophisticated Model
Md1-bot: Md1-bot:
Max: 108kNm/m Max: 111kNm/m
The FE Approach
Deflections
Simplified Model More Sophisticated Model
Max : 57mm Max: 52mm
The FE Approach
The FE Approach
Design Hogging Moments
Using the simpler FE model where column and
wall sections are not included we expose peak
centroidal results at the column heads of
434kNm/m.
The more sophisticated FE model inherently allows
us to deal with moments at the column face, and
the design values reduce to 297kNm/m.
The latter model is considered to be more realistic.
The FE Approach
Design Hogging Moments
A problem we are still faced with is that the
contour diagrams display nodal values.
These are very localised.
They suffer from being mesh sensitive -
increasing the mesh density will not achieve
convergence.
They take no account of yielding that would
occur in the real model.
The FE Approach
Design Hogging Moments
However, these peaks can be reduced by averaging the moment across
the width of the design strip.
These average moments are quite similar to the code method of
proportioning the design moments.
Column Strip
Inner
Col Strip
Middle Strip
Average moments
Code moments
Middle Strip
Example Comparing FE Modelling Options
Design Hogging Moments
Strips can be cut on the fly in the FE Post-Processor
In this view a 2m wide strip is cut across the column heads.
Within this strip the peak hogging moments are still 297kNm/m
(Note how results are not created within the column boundaries).
Example Comparing FE Modelling Options
Design Hogging Moments
To average the moment change the calc method to Integral
The design moment then reduces to 229kNm/m
You have the option to convert this into a required
area of steel.
Comparing FE with the Code Approach
Design Hogging Moments Comparison
The code approach suggests peak design moments of 212
kNm/m (265 before 20% redistribution)
The simpler FE model gives 434 peak, but 311 kNm/m using the
integral strip.
This model exposes peak centroidal results at the column heads
and even when averaged out, the design moments are still high.
The more sophisticated FE model gives 297 peak, but 229
kNm/m using the integral strip.
This model inherently deals with moments at the column face.
The extreme peaks are not generated, and when averaged out
the design values compare well with the tried and tested code
method values.
Comparing FE with the Code Approach
Design Sagging Moments Comparison
The code approach suggests a design sagging
moment of 120 kNm/m
(Before redistribution this was 763/877 * 120 =
104 kNm/m)
The simpler and the more sophisticated FE models
give results of 111 and 108 kNm/m respectively.
The above is possibly of concern. Both FE model
results are smaller than the code value.
But we can make another adjustment to account
for this.
Set the positive moment factor
to 1.1.
This increases all sagging
moments by 10% - it is not
moment re-distribution, but
provides some allowance for
pattern loading.
The maximum sagging moment
is now 122 kNm/m.
This is close to the 120 kNm/m
derived by the code approach.
You may want to consider
increasing the positive moment
factor to as much as 1.2.
Sagging Moment Comparison
Conclusion
For this regular model, by using the option to
include column and wall sections combined
with the integral strip feature we have
generated comparable design moments in FE
when compared with traditional code
idealisation.
Comparing FE with the Code Approach
Orion Training
Orion Training
FE Analysis and Design Process for
Flat Slabs
Copyright CSC(UK) Ltd. 2006
Flat Slabs FE Analysis and
Design Process
So far we have achieved comparable design
moments between FE and the traditional
code idealisation.
However, there are other design criteria
such as punching shear or deflections that
may govern the design.
We will now consider the entire analysis and
design process.
Flat Slabs FE Analysis and
Design Process
The most up to date source of reference on
this topic is:
BROOKER, O. how to design reinforced
concrete flat slabs using Finite Element
Analysis The Concrete Centre
Flat Slabs FE Analysis and
Design Process
The following flowchart is based on
Figure 1 of the previously mentioned
publication and illustrates the design
process as it would be applied using
Orion.
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Initial sizing
Flat Slabs FE Analysis
You must select an appropriate slab
thickness.
FE analysis does not equate to thinner slabs.
Excessive deflection would be an indicator that
the slab is too thin.
Orion will be unable to determine the area of
steel required.
For an appropriate thickness refer to:
C H Goodchild, Economic Concrete Frame
Elements, British Cement Association
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No Linear
Analysis?
Flat Slabs FE Analysis
Is a linear analysis adequate for
checking serviceability?
The most comprehensive source of
reference on this topic is:
Concrete Society Technical Report No. 58,
Deflections in concrete slabs and beams
Flat Slabs FE Analysis
Technical Report (TR)58
Suggests the following methods for assessing
compliance with deflection criteria:
span/effective depth ratios
linear analysis
sophisticated non-linear solutions
No matter how sophisticated the method you
are always dealing with an estimate.
Calculated and measured deflections may differ
by up to 30%
Flat Slabs FE Analysis
TR 58 - span/effective depth ratios
Where normal limits are applicable then
span/effective depth ratios are perfectly
adequate.
These ratios will often give more economic
solutions.
Flat Slabs FE Analysis
TR 58 linear elastic analysis
Orion currently uses linear elastic analysis.
This is the simplest analytical method
discussed in the report. (So it should be the
simplest for you to use).
It should only be used to show that deflection
is not critical, rather than for providing true
deflection estimates.
For this method a long term elastic modulus
should be used.
Flat Slabs FE Analysis
TR 58 non-linear analysis
Non-linear is currently beyond scope of Orion.
Potentially it lets you control deflection and reduce slab
depth by adding more reinforcement.
TR58 notes that if you do this, you may start adding
steel before even starting to reduce the depth.
Consider a 300 thick slab that you want to reduce to
275. The relative stiffness of these slabs is something
in the region of 300
3
/275
3
= 1.3. A small (8%)
reduction in depth leads to a significant (30%)
reduction in stiffness and hence to a significant
addition of reinforcement.
Flat Slabs FE Analysis
TR 58 non-linear analysis
As slab thicknesses are reduced other design
issues such as punching shear can become
more critical.
Clearly there are cases where engineers are
reducing slab thicknesses at the expense of a
significant increase in reinforcement.
There is some debate as to whether this is in
fact an economic balance that should be
pursued in the majority of cases.
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Flat Slabs FE Analysis
If you are not going to check deflections
using span to effective depth ratios, then
as mentioned in TR 58, you need to
make an adjustment to the elastic
modulus.
Flat Slabs FE Analysis
TR 58 Adjustment of elastic modulus
Reduce E by half to conservatively allow for
cracking and shrinkage.
Reduce E to E
c,eff
to allow for creep
E
c,eff
= E / (1 + )
where = creep coefficient (from figure 7.1 of BS
8110 Part 2)
The above adjustments result in a long term
E value between 0.16 and 0.25 times the
short term value.
Flat Slabs FE Analysis
TR 58 Adjustment of elastic modulus
TR 58 also provides more rigorous formulae
for estimating the adjustment factors and
deflection (suitable for spreadsheets)
NOTE: These adjustment factors are aimed at
providing conservative deflection estimations
rather than showing likely compliance with
serviceability limits. Use of these factors on a
slab that satisfies span/effective depth limits
may thus predict deflections a little higher
than the suggested limits.
Flat Slabs FE Analysis
CIRIA Report 110 (2
nd
Edition)
This report also gives some guidance on adjustments
appropriate to allow for both creep and cracking.
In simple terms this suggests:
reducing E by half to two thirds to allow for creep
reducing I by half to allow for cracking.
Hence the stiffness (EI) should be adjusted (reduced) in
the range 0.16 to 0.25
The previously mentioned paper How to design
reinforced concrete flat slabs using Finite Element
Analysis suggests that 0.16 should be used for storage
and plant loads and 0.25 for office and residential loads
Flat Slabs FE Analysis
Applying TR58/CIRIA 110 Serviceability
Guidance in Orion.
This subject can potentially get quite
complex, but is there actually a problem to
solve? There does not seem to be any
reported problems with apparently well
designed slabs suffering from excessive
deflection.
Provided you have not completely ignored
traditional limits/guidance you are unlikely to
experience problems.
Flat Slabs FE Analysis
Applying TR58/CIRIA 110 Serviceability
Guidance.
The adjustments can be made in Orion via two
parameters, individually or in combination:
Slab Stiffness Multiplier, (SSM).
Elastic Section Modulus, (E).
How this is actually done is the Engineers
decision, but the following should be borne in
mind:
Flat Slabs FE Analysis
Adjusting SSM will affect the distribution of
moments
In general reducing slab stiffness will increase
moments thrown into columns
The code suggests consistent section
properties
Changing only E (for all members) will affect
displacements without affecting the distribution
of moments
Flat Slabs FE Analysis
Options available
1. All adjustment to slab (SSM only)
2. Some adjustment to columns and bigger
adjustment to slab (E + SSM)
3. Adjust all equally (E only)
Flat Slabs FE Analysis
Option 1. - All adjustment to slab
The short term elastic section modulus is unchanged.
The slab stiffness multiplier is reduced to between
0.16 and 0.25.
Flat Slabs FE Analysis
Option 2. - Some adjustment to cols and bigger
adjustment to slab
Adjust the elastic section modulus 50% reduction
applied below (Applied to All Members)
Flat Slabs FE Analysis
Option 2 (continued)
A further say 50% reduction is applied using the slab
stiffness multiplier
Total adjustment = 0.5 x 0.5 = 0.25 hence in
required range.
Flat Slabs FE Analysis
Option 3. Adjust All Equally
The slab stiffness multiplier is left at 1.0
The elastic section modulus is reduced to
between 0.16 and 0.25
Flat Slabs FE Analysis
Reduces
moment in
columns
Attracts
more
moment
into cols
Effect on
Col
Moments
Reduced
to 50%
Reduced
to 25%
SSM 50%
E 50%
Some adjustment
to cols and bigger
adjustment to slab
Adjust all equally
All adjustment to
slab
Options available
Increases
deflection
Reduced
to 25%
Reduced
to 25%
SSM 100%
E 25%
Reduces
deflection
Not
reduced
Reduced
to 25%
SSM 25%
E 100%
Effect on
Slab
Deflection
Column
stiffness
Slab
stiffness
Adjustment
Summary of Options
The highlighted option is probably favoured by most engineers because it complies with code
guidance that consistent properties should be used throughout, and it keeps transfer moments to
columns to a level that seems to be more in keeping with traditional engineering expectation. It is
the method used in the previously mentioned paper How to design reinforced concrete flat slabs
using Finite Element Analysis
Flat Slabs FE Analysis
A Final Point on Adjustment Factors!
We require total dead and imposed deflection
under service load (i.e. unfactored)
However, in the FE Post Processor the results are
for G+Q*F (i.e. factored).
If you work within the suggested range of
adjustments (between 0.16 to 0.25), bear in mind
that you will need to de-factor the deflections
reported for G+Q*F
Alternatively, to account for this, the stiffness
adjustment factors could be increased again by
something between 1.4 and 1.6. Hence the range
for the overall stiffness adjustment might be 0.22
to 0.40.
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Remember that all columns and walls
have to be positioned on slab
boundaries in order to form proper
connections between the FE slab
plates and the columns/walls.
Flat Slab Model Generation
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Flat Slabs Running the Analysis
Building Analysis is
run to generate
lateral design
forces for columns
and walls.
The notional load
calculation will only be
valid if the storey
weight is determined
using the Un-
decomposed slab
loads.
After running a building analysis, gravity
design forces will not be correct. This will be
obvious from the Axial Load Comparison
Report.
To generate gravity design forces an FE
chasedown analysis is required and the FE
column/wall results merged with the Building
Analysis results.
Sequential finite element analysis in general
is covered elsewhere in this course.
Flat Slabs Running the Analysis
It should be emphasized that FE analysis only
deals with gravity loads.
In building analysis, sway load will not be
resisted by framing action between slabs and
columns.
Sway loads should therefore be resisted by walls
(ie braced flat slab construction).
Unbraced flat slab construction remains beyond
the current scope of Orion.
Flat Slabs Running the Analysis
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Flat Slabs Verification
Checks
On completion of the sequential finite element
analysis, the Axial Load Comparison Report
should be checked to validate the result.
The deflection contours should be inspected. Are
the peak values where you expect?
wL
2
/8 hand calcs could be used (with a bit of
thought and adjustment) to estimate the
difference between average hogging and
sagging moments for particular spans.
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Area of Steel Requirements
The Previous Verification Model will be used
to illustrate a possible reinforcing strategy.
Threshold Contours will initially be used to
determine area of steel requirements.
User Defined Contours can then be used to
convey the information to the detailer.
Flat Slabs Area of Steel Requirements
For Area of Steel Contours to be drawn
correctly you should first right click to setup
the Concrete Cover.
Flat Slabs Area of Steel Requirements
Threshold Contours
Ensure that the required
design area of steel is
selected, then click on the
Threshold Contours icon.
Right click to display the
Contour Setup
The maximum As-reqd will be
displayed in red.
The Set Nominal Steel button
can then be used to display a
lesser threshold.
Flat Slabs Area of Steel Requirements
Example - Using FE Analysis
Bottom Steel
The As(d)1-bot effect is selected
The maximum As-reqd for
sagging moments is
1166mm
2
/m
We noted earlier that we expect
reinforcement which provides
70-80% of this peak to be
adequate over most of the slab.
The threshold can be set by
selecting bars and spacings. In
this case T16 @ 250 provides
804mm
2
/m. (69%)
The areas where additional steel
is required are graphically
highlighted.
Example - Using FE Analysis
Bottom Steel
Detailing Information
The additional steel
required in the highlighted
zones is always less than
1166 804 = 362mm
2
/m.
In this case T12 @ 250
provides 452mm
2
/m.
i.e. lay in extra bars to
create T16 and T12
alternate bars at 125crs in
the highlighted zones.
With closer examination
you could reduce the extra
bars to T10s in patches
is it worth it?
Example - Using FE Analysis
Top Steel
Threshold Contours
By setting a threshold level
of 0 we can see the areas in
which at least some top steel
is required.
Note that with pattern
loading these areas would
increase slightly.
However, many would
regard it as normal/good
practice to provide some
level of reinforcement
throughout the top of the
slab.
Top Steel
Threshold Contours
The maximum As-reqd for
hogging moments is reported as
3032mm
2
/m
We noted earlier that hogging
moments build rapidly and that a
much lower general steel
provision will be adequate over
much of the slab.
In this case T12 @ 250 provides
452mm
2
/m. This is little more
than nominal reinforcement in a
300 thick slab.
The areas where additional steel is
required are graphically
highlighted.
We are only left with deciding what to do about the
reinforcement patches required over the column
heads.
This is where the integral strip feature really comes
into its own.
Bear in mind that the code approach expects this
patch to be twice as heavily reinforced in the middle
half as in the outer edges.
There are a number of ways you could tackle this,
but a good start is to look at the effect of doubling up
the nominal steel provision and then cut strips over
the top of that.
Flat Slabs Area of Steel Requirements
Top Steel
Threshold Contours
In this case T12 @ 125
provides 904mm
2
/m.
The areas where additional
steel is required are
graphically highlighted.
The zones are now quite
concentrated at the column
heads.
Top Steel
Strip Results
A 2m wide strip is cut over the
column heads.
Using the integral option the peak
As-reqd is determined as
2450mm
2
/m.
The additional steel required is
therefore 2450 452 =
1998mm
2
/m.
Adding T25 @ 250 provides an
additional 1960mm
2
/m, or
T32 @ 250 provides 3220mm
2
/m.
NOTE - the 2m wide strip is
narrower than the red zone it is
conservative to cut this narrower
strip and then provide the
additional reinforcement in a
wider zone.
Top Steel Detailing Information
In this slab we have provided T12s at 250 throughout.
The threshold contours show where additional patches are
required.
In more complex jobs you might define a number of standard
patch details (although you may find that the incremental areas
provided by standard bar diameters at set spacing mean that
this is not practical).
In this example we have 1 standard patch over internal
columns, we are going to make it 5m square centred on the
columns
In the middle 2.5m the added bars are perhaps T25 @ 250
In the 1.25m strips to each side the added bars are T12 @ 250
Flat Slabs Area of Steel Requirements
User Defined Area of
Steel contours can be
generated by right
clicking.
These can then be
printed to indicate the
patches to the detailer.
Steel Detailing Information
Flat Slabs Area of Steel Requirements
As (required) Contours for As(d)1-top
The contours can be
exported as a layer
to be displayed in
the Graphic Editor
Drawings containing
the contours can
then be exported to
dxf for loading into
a CAD program.
Export of User defined As (required) Contours
Flat Slabs Area of Steel Requirements
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Flat Slabs FE Analysis
Technical Report (TR)58
Suggests the following methods for assessing
compliance with deflection criteria:
span/effective depth ratios
linear analysis
sophisticated non-linear solutions
As already discussed the non-linear method is
beyond the scope of Orion.
Flat Slabs Deflections
For the validation model, it regularity lends
itself to checking via span to effective depth
ratios.
If the model were designed via either the
Yield Line method or the Code Approach,
deflection would be checked in this way.
However, for models with irregular geometries
or slabs with holes it may be more difficult to
apply this technique.
Flat Slabs Deflections
The Yield Line Method
For the validation model the span/eff depth check is
shown to be OK
The Code Approach
Because it determines the As required by a different
method to yield line, the resulting allowable span to
depth ratios will vary.
However, the Code Approach would also indicate the
span to depth ratio is OK for the validation model.
Flat Slabs Verification Model
Flat Slabs Deflections
Checking Deflection via Orions linear analysis
BS8110 doesnt set a definitive limit for deflection.
However, the code span-to-depth checks imply:
Total deflection limited to span/250
Deflection occurring after installing non-
structural items limited to span/500
For the validation model, taking the span as the
diagonal across a slab panel, we arrive at:
11.3m/250 = 45mm
Orions deflection estimate =52mm - an apparent fail.
However, traditional techniques using strip methods and
span/depth ratios have indicated the model is acceptable.
Bear in mind that multipliers are probably conservative
and that all deflection estimates are approximate.
Having used a multiplier of 0.25 in Orion (in effect 1/6
th
) ,
most engineers would be happy if deflection is close to or
does not exceed span/250.
Flat Slabs Deflections
Note that in more complex layouts FE analysis graphically
identifies critical locations so that consideration of deflection
and perhaps cross checks using span/effective depth can be
more focused.
We have checked taking the span as
the diagonal across a slab panel.
Checks should also be made between
the horizontal and vertical spans.
More generally, check max deflection
occurring along a straight line
between any 2 columns.
52mm
37mm
37
mm
0mm
0mm
0mm 0mm
Flat Slabs Deflections
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
This is well covered in BS8110 and expanded upon in Ciria Report
110. The guidance need not be repeated within the context of
this presentation.
Various checks/limits and alternatives may be applied/employed
it is all logical but quite idealised/empirical in nature.
One way of checking the moment transfer requirement is to use
the design moment developed in the column (above and below)
and design the strip for that.
We will look at what you can do within the FE post-processor in a
moment.
However, it is worth emphasising that this is a potentially severe
limit that ought to be considered early in the design process
careful re-arrangement of the slab/column intersection might
prove to be important and/or simply more economic.
Flat Slabs Moment Transfer at Edges
For this simple model we would
need to look at additional patches:
One for the corners.
One for the edges.
In each case the code needs to be
referred to for guidance on a
suitable strip width (Cl. 3.7.4.2)
In this example a 600mm wide
strip is appropriate.
Cutting a 600mm wide strip as
shown here a steel area is
determined (960mm
2
/m)
However, note that the strip is not
covering the entire red zone (the
threshold based on the general
reinforcement provision)
Flat Slabs Moment Transfer at Edges
The objective here is to ensure that
all of the hogging moment in the
slab is transferred to the column.
By cutting a wider strip (1.8m) we
can capture the total steel
requirement - 777mm
2
/m over
1.8m = 1399mm
2
.
If we choose to concentrate this in
a 600mm wide strip we need to
provide 2331mm
2
/m within that
strip.
Therefore the additional steel
required within this strip is 2331
452 = 1879mm
2
/m
Hence, in this very narrow strip the
additional steel requirement might
involve (4 No) T25 bars at 250.
Flat Slabs Moment Transfer at Edges
Orion Training
Flat Slabs Flowchart
Initial sizing
Linear
Analysis?
Use long term E, or
check serviceability
using span to depth ratios
Build the model
Run Analysis
Verification checks
Determine As requirements
Check deflection
Check moment transfer at
edge and corner columns
Check punching shear
Beyond
Scope
of Orion
Yes
No
Punching Shear Checks
After running Punching Checks for each column, any additional
shear reinforcement requirement will be displayed graphically.
For technical information refer to the Punching Shear Checks
Chapter of the Engineers Handbook.
Orion Training
Summary
For the Simple Model
We have made comparisons and validated the methods.
That should give us confidence
Comparing FE with the Code Approach
Idealisation for a Regular Structure
An orthogonal system
of column strips and
middle strips is easy
to visualise.
Comparing FE with the Code Approach
But for an Irregular Structure?
No obvious orthogonal
system of strips.
A system of triangular
zones radiating from a
core zone that combines
the walls and the column
would be more rational.
Codes give no way to deal
with this in a simplified
fashion.
And we have still not
added holes or
concentrated loads.
Comparing FE with the Code Approach
Summary
For Irregular Structures
We cannot make the comparisons because the code will not
extend to this.
But there is no reason why the same principles do not apply.
Comparing FE with the Code Approach
The Buildings we are asked about:
Nobody shows us a nice regular layout.
In the layouts we see the engineer has usually tried and failed
to apply the codes strip idealisation in a way that he is
comfortable with.
The engineer is turning to FE as a more sophisticated analysis.
And IT IS a more sophisticated analysis, BUT it is not a design
methodology. Any FE analysis will yield results that need to be
used carefully. The recently published TR58 starts to offer good
guidance on deflection, there appears to be less authoritative
guidance on issues such as the determination of design
moments across column heads. Engineers we speak to are often
hoping that we/software can fill such voids.
In practice the design methods ought to be established by the
profession. We can only offer guidance and this presentation
summarises the most important aspects of this.
Comparing FE with the Code Approach
Orion Training
Orion Training
Flat Slabs
Meshing Tips & Techniques
Copyright CSC(UK) Ltd. 2006
We would suggest that by following this
sequence you will ultimately save time.
Flat Slab Model Generation
1. Make slab insertion easy by being
organised with grids
Avoid closely spaced parallel grids use one
rather than two where possible.
Consider shrinking grids before placing slabs.
Flat Slab Model Generation
2. If the intention is to include column/wall
sections, avoid short distances from the
slab edge to the column face.
Short lengths of slab edge can cause
meshing problems.
Flat Slab Model Generation
2. Create larger regular slab panels, avoid
multiple slab panels between columns
and especially around columns.
Consider using the Pick Axis method of
slab insertion.
Flat Slab Model Generation
Other tips for
successful modelling
On Slab Properties, click
Insertion button.
Pick Axis method of slab insertion can be used as follows:
Change insertion
method to Pick Axis
If any axes are listed
in the box click
Delete All Axes
Other tips for
successful modelling
Press Ctrl and click on
axes to define boundary
1st click
2nd click
3rd click
4th click
Click on New Slab Panel
and the slab is drawn
But avoid defining highly irregular slab
outlines.
Minimise the number of slab edges.
Flat Slab Model Generation
Other tips for
successful modelling
3. As you insert slabs, (and before adding
holes and concentrated loads)
occasionally check the FE mesh can
be generated (using column section
modelling if it is your intention to do
so).
Flat Slab Model Generation
4. Repeat the previous step after adding
holes and concentrated loads.
Flat Slab Model Generation
5. You should now have one floor that
you know is going to mesh. Other
floors may be completely different, in
which case the above procedure
should be repeated. If other floors are
partial duplicates you are now sure
that the copied model works before
editing starts.
Flat Slab Model Generation
6. Duplicate Floors It is not essential to
model every floor genuine
duplicates can be dealt with as such
hence avoiding repetitive meshing etc
take full advantage of his.
Flat Slab Model Generation
What can be done if meshing fails to work?
Meshing difficulties can be caused by:
Short slab edges
Short slab spans between openings/edges
Loads within a column boundary
Using the column/ wall section modelling
option inherently introduces short slab
edges
Flat Slab Model Generation
Flat Slab Model Generation
Example 1:
Short slab
edge
Because walls not
meeting at 90 degrees,
Short slab edges created
at corner of wall.
Enlarged View of Wall Corner
Flat Slab Model Generation
No problem if column/wall sections not
included.
But fails to
mesh when
column/wall
sections
included.
Solution:
Edit grids to
ensure walls are
orthogonal.
Flat Slab Model Generation
Example 2: Another Short
slab edge
Flat Slab Model Generation
No problem if column/wall
sections not included.
Flat Slab Model Generation But problem when
column/wall sections
included
Flat Slab Model Generation
Solution:
Move column down
to remove gap
Example 3: Slab load within column
perimeter
Flat Slab Model Generation
Meshes OK if column/wall sections not
included
Flat Slab Model Generation
Error if column/wall sections included
Flat Slab Model Generation
Solutions:
1. Apply as a Column Nodal Load instead of a
Slab Load, or
2. Move the load outside the column perimeter
Flat Slab Model Generation
Orion Training
Orion Training
Flat Slabs
Punching Shear Tips & Techniques
Copyright CSC(UK) Ltd. 2006
As geometry becomes more complex,
it becomes increasingly difficult to find
the punching perimeter.
Potentially very frustrating if you have
meshed and analysed lots of floors
and then cannot do punching checks.
Punching Shear Checks
However
This frustration is completely avoidable.
You can check the perimeters can be
found before even attempting meshing.
Punching Shear Checks
This is done as follows:
Display the Punching Shear dialog
Click Hatch Floor Area (Initially with Inc. Col,
Wall, Beam Edges Unchecked).
Punching Shear Checks
If entire floor area is hatched then punching
perimeters can generally be determined for
all columns.
Punching Shear Checks
If entire floor area is NOT hatched
then punching perimeters can not be
determined in the un-hatched areas
Punching Shear Checks
But, there is a tool that may resolve this:
Check the box to Inc. Col, Wall, Beam Edges, and then
click Hatch Floor Area again.
If entire floor area is hatched then perimeters can be
determined.
Punching Shear Checks

Вам также может понравиться