Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Another ASCE 7-05 Wind Load Question

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Thanks to Drew, and Gary

I adopt the same conservative perspective when it comes to design.
Unfortunately the majority of the time I am assessing "existing" of some
description: construction that never got approval (6 or more years old), or
manufactured structures for locations which don't match the "standard"
calculations. Wind loading is just about always the critical loading, so I
generally want to get it as low as possible.

The discussion of ASCE7-05 over the past few weeks, indicates that its
complexity may fill in some of the gaps in AS1170.2. However, users
criticising its complexity, cost and errors negates getting it simply to
satisfy an interest.

Back to the topographical. The z0 that Drew mentioned in ASCE7 seems similar
to a combination of AS1170.2 terrain category averaging and shielding
multipliers. My main interest are the top of hill topographical variables
that Terry Weatherby mentioned. I think ASCE7-02 uses k-factors (K) where
AS1170.2 uses multipliers (M), from a copy of a spreadsheet I found the
topographic factor appears to be K[zt]. {Square brackets for subscripts,
Pascal convention}. 

Other than some differences in approach, and therefore formula, ASCE7-02
appears to have the same limitations as AS1170.2. I was wondering if
ASCE7-05 more clearly deals with the approach to the hill/escarpment, and
whether it considers a venturi like effect of wind blowing along a valley.
The Adelaide Hills face zone is basically barred from development, so when
it comes to topography the buildings are in the undulating terrain of the
hills.

I guess that's where the single exposure parameter tends to make things
easier: only have to make one judgement. 3 guesses multiplied together
doesn't exactly make a better guess, but I think our code has always been
that way.

The z0 you mentioned does give me another perspective. To AS1170.2 I can
treat the issue in three stages, combine terrain (Mz,cat) and shielding
(Ms), consider this to provide the wind speed from an equivalent horizontal
plain approaching the hill. Then for stage 2, apply the topographical factor
(Mt). Then for stage 3, apply local shielding (Ms) again. Not a good idea to
apply Ms twice in practice, but an interesting experiment. Considered with
respect to the 1km averaging distance it probably won't make a difference to
the results: just my perception. Namely, have a random undulating 3D
surface, apply one multiplier removes one set of troughs and peaks, apply
another multiplier and another set of troughs and peaks disappear, until all
troughs and peaks removed and get equivalent flat surface with one building
in the middle of it.

Terrain and exposure classification leave something to be desired. Suburban
terrain may generate surface roughness when there are reasonable gaps
between buildings and roof tops. But when building density increases to
having a near continuous roof scape within the terrain averaging distance,
then assessing the terrain from the lowest point of the roof-scape seems
more realistic, which suggests the walls are fully shielded. Not necessarily
sensible for design. But potentially explains damage to roof cladding and
little damage to walls. That is to AS1170.2 at ground level have: TC3 with
shielding by other houses, but roof scape of TC2. Thus whilst increased
research maybe getting better assessments of roof pressure coefficients, it
is the reference wind speed which is under estimated.

Also to the Australian code it is permitted to adopt a higher terrain
roughness on an assumption of future development. Thus a suburban house gets
designed to TC3, when it is on the fringes of an expanding suburb even
though in the middle of new development and surrounded by terrain of TC2.
The code actually requires consideration of 8 cardinal directions reduced to
4 orthogonal directions of the building axis. Most designers simply make an
assessment for one direction, making a judgment of the worst case rather
than calculating it. The result is semi-rural development of new suburbs are
classed as TC3, but remain exposed to TC2 for a substantial period. A check
of the Bureau of Meteorology severe weather statistics indicates that much
of the annual storm damage is to buildings in rural and semi-rural towns.

A new building may find itself in the middle of undeveloped open terrain,
and a retained old building may find itself in the open terrain of a large
redevelopment demolition site. A design decision. The economic solution for
the builder and owner is TC3 (B), whilst the insurance industry which pays
for the storm damage may be more in favour of TC2 (C). Alternatively wind
breaks, and/or temporary shielding of suburban development to reduce the
cost of waste.

I already mentioned the seasonal variation of terrain in rural settings,
still another situation is the coastal terrain. If look at a Beaufort wind
speed chart for use at sea, the wave heights generated by ultimate limit
design wind speeds are relatively large. The commentary to AS1170.2,
suggests adopting TC2 (C) for strength, and TC1 (D) for serviceability
assessment. That is at low wind speeds the water surface is smooth and has
low surface roughness, at high wind speeds wave height increases and surface
roughness increases. Not certain of the validity of that argument. A
lighthouse on a rock, not experiencing the wind generating the waves around
it, doesn't seem quite correct. But perhaps it is.

Now Bill Allen is going to accuse me of making stuff complicated again. But
it is a matter of design, versus assessment, versus understanding. If the
code is too simple then too much constraint is placed on approval.
Simplifying the code for routine day to day design we can do ourselves.

Wind tunnel research to more accurately assess pressure coefficients over
the surface of the building, does seem wasteful, when the wind speed at the
site is little better than a guess. But don't have to use the coefficients
in the code.

Most products are available in ranges determined by a geometric progression.
If a change to a code is not properly calibrated then a minor change in load
can cause a jump to the next size of structural section and that can be a
huge increase in cost across an industry. So it is understandable that there
are objections about having to work twice as hard to prove the traditional
section is still suitable. 

Try the following:

1) plot span (L) against maximum bending moment (M=wL^2/8) for a simple
beam, plot different curves for different wind loads.
2) Then plot the section moment capacities for different structural
sections, creating bands that define the maximum range of suitability for
the sections.
3) Then plot the member moment capacities 
4) Identify the region in which most of your projects fall
5) Alternatively follow steps 1 to 4 and plot deflections
6) Study the curves
7) Then answer the question: why are you calculating point-values for each
and every project? If do not need to do the calculations for each and every
project, then what does it matter if the complexity of the code increases,
and the code is revised every year?

Then work out similar methods for more complex structural forms. For example
plot iso-moment curves for gable frames, on a graph of height versus span.
>From such curves can make an assessment of what design-solutions the code
writers are trying to push down in section size, and which they are trying
to push up. Where is the industry looking for more economy, and where are
regulators looking for more conservatism?

Code writers are trying to make rules for all structural forms no matter
what innovative ideas turn up in the future. Computers now allow all of us
to put common structural forms into iterative loops, and assess the
suitability and limitations of available materials. When things are graphed
in such form, it is easy to question what all the fuss is about. Increase
this parameter, drop another, and the structural solution remains unchanged.
Just a simple matter of where somebody wants to put the focus this year.


Regards
Conrad Harrison
B.Tech (mfg & mech), MIIE, gradTIEAust
mailto:sch.tectonic(--nospam--at)bigpond.com
Adelaide
South Australia




******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at: http://www.seaint.org/list_FAQ.asp
* 
*   This email was sent to you via Structural Engineers 
*   Association of Southern California (SEAOSC) server. To 
*   subscribe (no fee) or UnSubscribe, please go to:
*
*   http://www.seaint.org/sealist1.asp
*
*   Questions to seaint-ad(--nospam--at)seaint.org. Remember, any email you 
*   send to the list is public domain and may be re-posted 
*   without your permission. Make sure you visit our web 
*   site at: http://www.seaint.org 
******* ****** ****** ****** ******* ****** ****** ********