Jetty Structure _ Piles in tension

[eXCELman]

Hello Everyone,
We (me and my other two colleages) have designed a jetty structure having 2.5 km long. Structure was composed of a combination of 2 ( battered only in transverse direction at some locations) piles connected with a RC cap beam. Longitudinally the structure is connected with 2 I girders and CIP slab on top of them ( used for access road) and oil pipe line itself.
Piles are , at max. up to 60 m in length and they are partially in air then sea water then in sand , then in clay and touching the bed rock. We had to define whole structure in our model simply because the oil pipeline was continuous almost 750 meter long and every 750 meter there was an "S" in the pipe line. We performed response spectrum analysis using Maximum credible earthquake parameters.
Since we have got some excessive pile end forces (in tension ) at some locations , piles had to be anchored to the bed rock. For this purpose, some rock anchors are designed to take care of this force. As you know this rock anchor construction at a 60 meter deep is a quite difficult task to be performed for contractor.
For the seismic effects, we made use of reduction factors only for moments.
Indeed in seismic actions, (if the ends of piles not anchored to the bed rock) some slippage will be occurring b/w piles and the surrounding soil (clay and sand) and this will cause dissipation of seismic energy by frictional forces around the piles. Eventually this effect will reduce seismic forces and also axial tensile forces.Bu this effect can not be modelled quiet precisely in any Structural analysis program.

What reasonable reduction factor can we assume to reduce our tensile forces at the tip of piles? or Should we make the analysis using different techniques ( such as time history analysis or push over analysis )

Is there anyone who might have faced such a problem before ?

Comments will be highly appreciated.

Thanks to all.

 

[BigH]

This is a bit beyond my reasoning at 0430am, but I would like to point out something regarding your piles. You've indicated that battered piles are used. Years ago there was some papers describing damage due to 1964 Nagata (spelling?) earthquake in Japan and they found out that with battered piles the pile cap takes a real s*&^ kicking. Try to keep this in mind. (It was an ASCE Journal article - look in the late 60s or early 70s). In Vancouver area they started going for only vertical piles - unless there has been a rethink on this the last number of years.

Good luck - I'll be watching this thread!

 

[eXCELman]

BigH

Many thanks for responding to our problem.

We had been aware of the danger you pointed out in your reply. But that was the only option in transverse direction that we can make use of to reduce the displacaments of piles longing 60 meter at most in sea water+sand+clay and ending at the bed rock. But there is a reasoning that some slipping action in the piles could absorb some seismic energy. But we could not determine how to reflect this into our calculations. We know that it is a very complicated behaviour and it should be solved in the time domain using time history analysis. For the time being, we are after some recommodations from anyone who has previously faced such a problem before. Current AASHTO or CALTRANS Seismic specifications does not tell about much of this behaviour.

What could be your suggestions under taht light.

Thanks.