Home' Future Building: The Australian Infrastructure Review : December 2010 Contents 88 futurebuilding DECEMBER 2010
THE DANGER OF PRESCRIPTIVE DESIGN
CRITERIA FOR WRSB
One of the enduring frustrations selling wire rope safety barrier
(WRSB) across Australia is the variation in state main road authority
design criteria. Some state authorities use deflection performance
as the design criteria and require the manufacturer to submit design
calculations to verify the proposed design. That is fine as WRSB are
structures that obey the laws of physics and as long as the calculations
account for all the variables (post spacing, length of fence, number
of ropes, etc) this is a reliable design method. However some of the
other states favour prescriptive design criteria. One problem with
prescriptive design methodologies is that the formulae has to make
sense over a broad range of design conditions. Another problem
is that the product performance required in prescriptive design
methodoloies is generally based on the lowest product performance
and this bias affects the capacity of higher technology products to be
sold in that market. This article discusses both issues and considers a
specific and current case study.
What is the danger of prescriptive specification?
The adjective "prescriptive" is defined as "the imposition or
enforcement of a rule or method". The problem is that the rule or
method gains unquestionable acceptance by long usage or the
passage of time -- it then takes on a life of being too prescriptive and
becomes dictatorial, narrow, rigid and dogmatic -- which can cause
Engineers to stop thinking and automatically refer to the specification
as if it is Holy writ.
What is a prescriptive specification for WRSB in this case?
The case to be studied is a median application on a highway, two
3.5m wide lanes each way with a 2.2m wide median between the
lanes. The 2008 AATD count was 5200 vpd east bound with 23%
heavy vehicles, and 5800 vpd west bound with 16% heavy vehicles.
It is proposed to locate the WRSB centrally in the 2.2m wide median.
The contract conditions were that the fence had to built at 2m post
spacing and had to be constructed with the only use of anchors to
be at either end of the fence, making the fence length 2046 metres.
That is, the fence was to constructed without intermediate anchors.
The site is both dead straight and dead flat for the full length of
the proposed fence. Referring to the prescriptive specification, the
maximum acceptable deflection in the fence is be 1.56 metres. This
1.56 metre measurement is calculated from the specification, being
1.3m (the "deflection" for 2m post spacing fence) multiplied by 1.2
(the correction factor if the fence length is greater than 350m), equals
1.56 metres deflection. That is, the authority is willing to accept 0.46
metre deflection into the oncoming traffic lanes.
Why is it a problem?
In terms of this particular design guideline all WRSB fence systems
are treated to be equal. This road authority is unconcerned if the
fence system uses 3 ropes or 4 ropes, if it uses weak posts or strong
posts, if it was tested to NCHRP350 TL3 (2000kg pick up truck) over
a short length or over a long length, or if it was tested at close post
spacing or wide post spacing. All systems are treated the same. All
systems are required to be built at fence lengths and at post spacing
predetermined by the design guidelines.
Are all WRSB equal in performance?
Any Engineer will confirm that several factors influence deflection
performance of a WRSB including
the length of the fence in a test (the shorter the test length the
• better the result),
the post spacing (the closer the post spacing the better the
the number of ropes (the greater the number of ropes the
• lower the deflection),
the post thickness and shape (some post designs are stiffer
and maintain interaction with the ropes longer), and
the work done by the ropes on the post (ropes interweaving
• between posts do more work on the posts than simple
straight rope alignments passing through the posts).
A performance specification takes all these factors into account.
So how would this fence be designed in most other states in Australia?
It all starts with the test results and the site geometry. There are 4
fence types available in this case for use, however we are advised that
2 are deemed unsuitable for use at this site for this median installation
purpose. That leaves 2 possible fence systems, the Brifen TL3 fence
(which we shall refer to as Fence A) and another fence which shall be
referred to as Fence B.
Links Archive Issue 1 July 2011 Navigation Previous Page Next Page