Building separation (seismic joint/gap)

[engOFFG]

Dear Experts,

I have 6 stories building (of total height 31m) , with big plan area
divided by several parts , each part of 100mx100m

between each 2 parts I calculated seismic joints, based on UBC97 , ACSE7-2010 , and IBC2012
results are similar, and give unreasonable gap width ( separation) closed to 140mm to 200mm.

for example(UBC);
for first part I get delta-s (joint displacement)from etabs from case SPECx( response spectrum ,x-dir.)
then get delta-m = delta-s X 0.7 X R
where R=5.5
then get the same for the second part

then apply code equation, and get very conservative separation width.

IS this way I followed OK?
any ideas to get reasonable value?

(see attached UBC97 equation)

 

[JAE]

Looks OK to me.

Your joints of 140mm to 200mm sound appropriate for a 6 story building - I'm assuming these numbers come from the top story.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[engOFFG]

Thanks JAE
but these joint of 140mm to 200mm (5.5" to 7.8") making big problem with architectural design , in addition to the high cost .
yes these are the max. values at roof level
these values look unreasonable to me

I am wondering , why design code allows us to design buildings based on reduced seismic force (by dividing seismic force by R value) ,then when we check joint width or story drift ,we have to multiply it by R , so we use the unreduced seismic force!!!!

why we have to use inelastic displacement (delta-m = 0.7xR x delta-s) not elastic one ???

 

[JAE]

why we have to use inelastic displacement (delta-m = 0.7xR x delta-s) not elastic one ???

This is because when you design for strength - the R reduces the seismic demand to a level that represents how your particular system can absorb energy with high distortions (even going inelastic)...the idea here is that the system will bend and distort a lot but not fall down - i.e. safety against collapse and save lives, not the structure.

For deflections - you don't use the lower value that is reduced by R because you will want to know the "actual" deflection that really occurs.

There are joint cover assemblies out there that deal with the larger widths required.
Here is one brand that goes to 12" (Floor Joint Assembly Link.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Agent666]

You need to remember that the square root sum of the squares method doesn't eliminate the possibility of pounding. If the building come together at their design drift they will still impact because you have not provided the full clearance. Unlikely if they have similar responses, but still possible.

Also your drift is based on a design level event. You need to realise that in reality earthquakes larger than this occur in real life. In New Zealand where I practice there is certainly a movement within the engineering community since our 2011 Christchurch earthquake series to provide clearances for the what is termed the Maximum Credible Earthquake (MCE). This is 1.5-1.8 times the design level earthquake. We had a number of instances where stairs in the like fell off support edges due to inadequate seating and seismic gaps and cases of pounding causing collapses.

It might seem like a larger clearance but in all likelihood it will be required when the mother of all earthquakes hits. There are plenty of proprietary joint that should be able to deal with those types of clearances.

 

[engOFFG]

Thank you JAE & Agent666