guy tower analysis-linear and non-linear

huangyhg

guy tower analysis-linear and non-linear
i鈥檝e been tasked to do a structural analysis check on an existing guy mast, standing for 12years. a communications company wants to put up extra dishes. using a frame anlysis programme (prokon-south africa), i did a linear analysis first and checked all the members under axial stress, all checked ok. then i did a non-linear analysis and the programme tells me it can鈥檛 analyse because it is unstable due to long slender elements. now i assume the tower will buckle under my design windspeed (1:50 year storm) which i followed according to our codes. now how do i go about writing my report stating ,鈥?the tower you build which has been standing for  12 years is unstable鈥? i鈥檝e ran the frame anlysis under smaller winds and it stands but as soon it starts getting higher the structure is unstable.
any advice on how to approach my report.

i've been there before.  just re  
i'd be a little hesitant to take any action till i understood why the program was giving that message.  there are hand methods you could use to verify that there is an instability; i seem to recall that some of the structural engineering handbooks address this issue.  in my experience, when the program won't even analyze a model it's usually due to modeling issues and not the true behavior of the structure.  you might want to take a close look at the way the cables are modeled, and maybe try using a significant bending stiffness in them, and then run additional models incrementally reducing the bending stiffness.  good luck!
castigliano
i may be completle off-base, but isn't ok for some elements of the structure to buckle, so long as the structure remains standing.
what i mean is the guy on the down-wind side is probably going to buckle (the imposed compression from the wind load exceeds the pre-tension in the cable.  but can the cable on the up-wind side carry the required load ?  in which case the down-wind cable sags, and the tension in the up-wind cable increases.
if it can't work then you do indeed have a problem.  was the original tower designed for the same loads ?  maybe the required load now is higher than it was, maybe the original analysis leaves something to be desired.  if these are the loads required today, i think you should make the owners aware of your findings.  however they react to this information, i think you still have a duty to the general public (who live around the tower).
wr3x,
guyed tower analysis cannot be handled by a frame program.  there are specific analysis engines for these types of structures.  contact an engineer experienced in these types of structures.
vod
thanks ,all valuable remarks. the tower was a charity project and some guy basically build it in his backyard with no design done.
the anlysis program allows for catenary cables and takes the stiffness into account. if a member buckles it removes that   
i'm not sure what non-liner analysis you did for the program to fail, but why not do a simple euler buckling assessment of the tower. that is a linear solution and will tell you what multiple of the applied load you've used for the tower to buckle. if you get silly answers which shows that it fails now then you'll have made a mistake and no one will believe you.
but isn't the program just telling you that you need more elements in the structure for it to assess it correctly, and not that the structure fails?
corus
to me the two non-linear influences that make sense are that the elements went plastic or the structural deflections became unstable (deflections increased the load in the tower, just like an euler column).
i would look at your linear and nl results to see if you can see either (or both) of these things happening.
2nd, can't you reinforce the tower ... increase the pre-load, add cables, add struts (as both of these increase the compression load in the tower core) ?
3rd, charity or not, if this was in the us and something undesirable happened in the future, it'd be your a$$ on the line.