response modification for non building structures

huangyhg

response modification for non building structures
hi all-
i'm having some dialog with colleagues regarding the appropriate r value to use per ibc or asce 7.  specifically i'm looking at free-standing large diameter fiberglass (frp) tanks.  we get into situations where these things need to be designed with seismic loads "per asce 7" or "ibc 2003".  my question is:  what does r really measure?  our material has more flexibility than steel, but it far less ductile?  energy absorption in the material itself would be damaging (cause cracking), but the flexibility of the structure would permit significant displacement to occur without damage.
referring to asce 7, table 9.14.5.1.1, it would appear we are stuck with r=1.5 (other material).  
any further insights out there?  thanks
rd78
r is intended to encompass a couple of things, namely redundancy and ductility.  looking only at ductility, you can view this simplistically by envisioning a stress/strain graph with two separate and distinct behaviors.  one should be fully elastic and the other should be pefectly elastic-perfectly plastic.  now, look at the difference between the two graphical relationships in the plastic region, the difference between the elastic and perfectly plastic graphs is the value "r".
keep in mind that "r" defined in the codes is arbitrarily chosen as to be consistent with code requirements.  that is, code requirements are life safety and are a minimum level of effort for the engineer or designer.  not wanting to be less conservative, the r values are chosen arbitrarily and are on the conservative side, thus they will not exactly match the simplistic definition i posed above.  
in addition, the codes must deal with systems of structures not just materials (as presented above) and so the r factor indirectly accounts for redundancy or the lack of redundancy.
so using the above definition one can see that a large mass supported by a single column of concrete will have a smaller r value than the same mass supported by a single column of steel.  concrete typically will have less ductility and or energy absorbtion characteristics and will be more susceptible to failure.  so code developers would like to see this type of structure designed for higer forces (elastically) though they may not be achieved in practice.  steel can tolerate more movement without degradation and so has a lower r value.  
as for redundancy you can see this to by looking at frames vs single supports for the same materials.  frames will typically have higher r values than will single supports.
though it doesn't directly answer your question on free-standing tanks, i hope this is useful background for your situation.  i think you can extrapolate this information to fiberglass vs steel tanks.
regards,
qshake
eng-tips forums:real solutions for real problems really quick.
note that with a steel tank, you don't have to guess- you'll find "r" given in a steel tank standard, specifically, awwa d100.  so my first thought is, are there any standards relating to the design of the tank in question?  or any for even similar (non-metallic) materials?  (or even any "proposed" standards?)
the only standard we know about is asme rtp-1, which is pretty outdated with respect to seismic.  it's undergoing revision, but i think the writers are struggling to properly address this.  thanks.