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

RE: FEM and Cables-Reply

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Real cable is geometric nonlinear, and has a 6X6 stiffness matrix for 3D
cables. The stiffness matrix also can be formulated into 6x7 matrix with
a constraint for advanced applications. Standard input for elastic
catenary cable includes the following:
(1)	coordinates of two joints
(2)	unit weight
(3)	unstrained length (it can be < joint distance.)
(4)	material constants (E and thermal expansion coefficients)

Standard procedure does not ask users to assume a pretension. A good
cable program can analyze cable nets, cable-stayed roof and bridge,
cable truss, or other type of cable-supported structures. Dynamic
analysis will be more complex.



	----------
	From:  James_F_Fulton(--nospam--at)RohmHaas.Com
[SMTP:James_F_Fulton(--nospam--at)RohmHaas.Com]
	Sent:  Tuesday, April 06, 1999 9:21 AM
	To:  seaint(--nospam--at)seaint.org
	Subject:  FEM and Cables-Reply

	STAAD has pretensioned cable members which are invoked with the
MEMBER CABLE 
	command. You specify the cable pretension, but these forces must
be 
	statically admissible, i.e. the forces specified must correspond
to a real, 
	achievable equilibrium state in the system of cables. If they
are not, the 
	final forces in the cable results will be significantly
different than those 
	input in order for the cable system to be in static equilibrium.
For 
	symmetrical cable layouts, *equal* pretensioning forces in all
the cables is 
	statically admissible and hence final cable forces will all be
the same, but 
	somewhat less than specified if the cables are, for example,
connected to a 
	stack where elastic shortening takes place under the cable
forces. This is 
	one little glitch in the program since in the field the
pretensioning force 
	is measured (?) as the stack shortens and hence elastic
shortening of stack 
	is accounted for as pretensioning takes place. In STAAD, you
would have to 
	specify a somewhat larger cable pretension and adjust by trial
and error to 
	end up with a specific final pretensioning force.

	In the situation where the cables are not symmetrical, it is
(usually) 
	practically impossible to predict beforehand what a set of
statically 
	admissible cable forces will be.  To determine this, I first run
the problem 
	of an equal, unit temperature drop on all the cables. However,
MEMBER TRUSS 
	members have to be used to represent the cable response rather
than MEMBER 
	CABLES. This then gives the linear force dependence among the
truss members 
	(cables) for the particular geometry.  Then, I run the lateral
load (e.g. 
	wind pressures) together with the MEMBER CABLES. This requires
several runs, 
	incrementing the cable pretensions, but maintaining their
relative, 
	statically admissible set of forces as determined from the
temperature drop 
	analysis, until a desired design criterion is achieved. One such
criterion 
	might be that no cables go slack under the design lateral load.

	As is usually the case with the STAAD folks, none of this
necessary procedure 
	is even touched on in the STAAD user manual. My approach I
discovered on my 
	first unsymmetrical cable situation. Maybe there are other
approaches. If so 
	I would like to hear about them. Also, in the field, I would
like to know how 
	the erector goes about measuring a specified cable pretension,
particularly a 
	relatively small 1500 lb. to 2000 lb. required pretensioning
force for small 
	(3/8' - 1/2" diam) cables, which I encounter with exhaust
stacks.