Need a book? Engineering books recommendations...

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

RE: Wood truss design criteria

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Nels,

It appears from your questions that you are following 10.1.3 of ANSI/TPI 1
which is entitled empirical analysis. The first paragraph says that "This
method has been developed based on many years of experience in wood truss
design and extensive investigation using PPSA on standard truss
configurations. The simplified method is considered to be conservative. This
method is intended for use by a truss design engineer experienced in the
design of wood trusses. The engineer shall use judgement with respect to the
applicability of the method for any given design."

Section 10.1.1 DESIGN MODEL and 10.1.2 MOMENT AND BUCKLING ANALYSIS is where
most of the industry operates these days. All the design methods that we are
aware of are proprietary and have test data behind the modeling being
performed.  There is a movement inside of TPI to remove the "simplified
method" from the standard, since it is not really representative of the
manner in which truss design is done today and can cause confusion. Most
design today is done using finite element analysis and is obviously far more
complex than the "simplified method."  This is the reason for differences in
uplift forces, chord forces, etc., that you may get from computer program
analysis and hand calculations. Yet, there is a contingent that says that
our industry needs to publish something for engineers to use that are not
intimately involved in the day to day truss design business even though it
is not perfect.

The answer to each of your questions is that this is the best wisdom
industry engineers have regarding how to undertake truss design outside of
using matrix methods analysis.

While that may not be the most satisfactory answer for you, you are
certainly welcome to participate in the ANSI/TPI consensus process, that is
going on currently, and ask all of these questions directly to the people
that are responsible for these judgements. They would be happy to provide
their insight and this would probably give you the confidence you need to
either decide to use this method or the confidence to decide not to.

You can start by calling TPI's Managing Director, Charlie Goehring at
608-833-5900 and discussing the ANSI/TPI consensus process and this standard
at length with him. He'll be able to help you with greater detail than can
be provided here.

Thanks for asking. Hope this begins to help.

Kirk Grundahl
Executive Director of WTCA
http://www.woodtruss.com




-----Original Message-----
From: NRoselund(--nospam--at)aol.com [mailto:NRoselund(--nospam--at)aol.com]
Sent: Wednesday, March 24, 1999 8:25 PM
To: seaint(--nospam--at)seaint.org
Subject: Wood truss design criteria


I'm reviewing the drawing for metal plate connected wood trusses.  This is
my
first project with this kind of truss.  The design submitted to me is based
on
the 1997 UBC, which requires ANSI/TPI 1-1995 as the design standard.

The rules of the design standard seem to defy structural mechanics.  Here's
my
summary of the design method for the top chord:
1)	The top chord, continuous from eaves to ridge over a single panel point,
is
analyzed as a continuous beam for flexure with compression.
2)	The positive moment analysis is based on assuming a simple span between
the
point of inflection and the pin-connected support (I concur with this).
3)	The compression stress used to determine the combined stress effect in
the
zone of maximum positive moment is based on the true length of the member
from
the panel point to the pin-connected support (I concur with this too).
4)	The analysis for negative moment at the panel point is based on 0.9x the
actual horizontal distance  between the panel point and the pin-connected
support; (how is the reduced span justified?)
5)	The allowable compression stress used to determine the combined stress at
the panel point does not take into account the unsupported length from the
point of inflection to the panel point (why not?).
6) 	In the combined stress computation at the panel point, the allowable
flexural stress is not required to be reduced by (1 minus the compression
stress ratio) as is required by formula 3.9-3 of the 1997 NDS.
7)	The connectors are designed to provide pinned connections with no moment
capacity assumed, so that I see no justification for assuming that the
connectors provide unaccounted-for continuity that justifies shorter-than-
actual spans for moment calculations, and increased member stability for
compression bucking stress allowable.

The result of these rules seems to be that excessively high stresses are
concealed by the calculation method; high stresses that would be determined
by
conventional structural mechanics.

Is there testing or rational analysis behind the ANSI/TPI Standard?  Has
anyone else questioned the standard?

Based on my requirement that the design of the trusses comply with the UBC,
it
appears that I am locked into accepting the design since it is based on the
UBC-specified standard.  But I'm not comfortable with it.

Nels Roselund
Structural Engineer