It appears that the following support cases should be used for seismic no friction:

L1+W+P1=SUS1 (friction multiplier =1)
L2+W+P1+T1 (friction multiplier =1)
L3+W+P1+T1 (friction multiplier =0)
L4+W+P1+T1+U1 (friction multiplier =0)
L5+W+P1=SUS2 (friction multiplier =0)
L6=L4-L3=+U1 no friction considered
L7=L1+L6 = SUS1+U1 (scalar) friction in sustained
L8=L5+L6= SUS2+U1 (scalar) no friction in sustained

Is there a consensus on this?

Also should the lateral no friction seismic support loads be more or less than the friction loads? I would expect more.

To my knowledge
generally wind/seismic/sway or imposed displacement will yield a higher load in non friction case

since the lateral frictional component in Shoe/+Y supports help to absorb the wind/seismic loads




so its always conservative to do the cases with non friction

Our experience is that with friction under seismic you tend to get higher loads on your supports, guides etc and without friction under seismic you tend to get higher loads on your anchors. Hence when we run AutoPIPE we run the model twice, once with friction under seismic and once with no friction under seismic and then compile all the results to get the highest loads. In CAESAR II we are still not confident how to set up the load cases to run the no friction under seismic case.

Itchy - Go to the "load case options" tab of the Static Load Case Editor. Scroll this dialog to the right so you can see the "Friction Multiplier" column. For the cases you don't want friction, set the multiplier to zero.

Rich's response is adequate for a static seismic (g load) evaluation. If you are running a response spectrum you will have to run dynamics twice - once pointing to an operating case without friction and then again pointing to an operating case with friction. When I say "pointing to" I am refering to the control Parameters setting "Static Case for Nonlinear Restraint Status" or some such phrase.

Hi Richard

But my understanding is that using the friction multiplier turns friction off for everything, not just seismic. I don't want to run no friction for the complete load case.

For the SUS+T1+U1 loadcase I want friction considered for SUS+T1 and no friction considered for +U1.

Sometimes the clients design criteria we work to request a 'no friction under seismic' analysis.

Then you need load cases setup like this:

1) W + P1 + T1 (with friction) this is your SUS+T1
2) U1 (without friction)
3) L1 + L2 (the combination of cases 1 and 2)

Hi Itchy
Friction multiplier zero will trun off friction for that particular load case not the entire system. Same load case you can make with friction without friction , just your load case will increase.

1) W+P1+T1 Friction multiplier one
2) W+P1+T1 Friction multiplier Zero

You can make combination as you like just by incresing number of load case.

Regards

Habib

Ok thanks Richard/Habib that leads back to setting up the loadcases in a similar manner to that shown above by MUL211.

I guess the difficult thing is, it is easy to play with the load cases but harder to actually know the implications of doing so. Likewise for APIPE, we can tick a box saying do not use friction on seismic, but really have no understanding as to how this is in fact applied.Time for a leap of faith I guess.

Think about what goes on in the solution - you're solving [K]{x}={f} for {x}. So for a given load case (W+T1+P1+U1), each of these components adds or inserts values into {f}, the load vector. Only after the construction of [K] and {f} is complete, does the solution for {x} begin. From this point forward, the solution doesn't know (or care) which values in {f} are from "W" or "T1" or "U1".

So in CAESAR II, what you do for a load case, you do for the whole load case. That is why I pulled "U1" out into its own load case.

If you're running another software package where the operation isn't clear, ask them how its done.

Many thanks Richard