Boundary condition for a seismic gap

5 posts / 0 new
Last post

I'm modeling a building that is next to another conditioned building
but they don't touch. There is a 1 foot gap between them that is
semi-enclosed (sheet metal around the sides and top (It is a seismic
gap). If the buildings were touching or had a REALLY small gap I'd
treat this surface as adiabatic but this case feels different.

I'm considering modelling this as a semi-conditioned space to take
partial credit (treat as a "room" with insulated walls and no
heating/cooling). Any other approaches that folks have used for this

Jason Quinn

JasonQuinn's picture
Joined: 2012-03-03
Reputation: 0

Your descriptions sound as if these are two existing buildings. Can you drop a data logger in there? (Real question: and get it back...)

Temperature measurements might help you decide if the gap should be modeled as an unconditioned room, or else model the building wall as an exterior surface but without direct solar exposure.

Without any measurements, try both and see how it matches your calibration data.

Despite the lack of view, are there any windows?

DSE Mobile

David Eldridge's picture
Joined: 2012-05-08
Reputation: 1

Hi Jason,

I would allow whether the associated envelopes of each building are insulated to influence my decision re: adiabatic.

Investigating further will probably take more work than simply modeling the situation as it is. Model a separate unconditioned space, but assign each building's envelope construction to the appropriate interior partitions, and assign an appropriate sheet metal based construction to the sides and top as exterior surfaces.

Consider infiltration and the tightness of this seismic gap along the way.

If these are huge walls (multistory), reality may be that heat transfer is significant along the perimeter (front/back/top) and insignificant at the "core" (center-bottom). Heat transfer then is something like a slab-on-grade on its side with the exposed perimeters composing much of the heat movement. You could potentially approximate this by handling the top floors and perimeter spaces differently from the ground floors and core spaces in terms of adiabatic partitions.


Nick-Caton's picture
Joined: 2011-09-30
Reputation: 605

I would tend towards modeling them as two separate building shells. Is the cavity between them vented?

Vikram Sami, LEED AP BD+C

Sami, Vikram's picture
Joined: 2011-09-30
Reputation: -1

The cavity between them is vented along the top.

These are existing buildings but putting a data logger in and pulling
out temperatures is not going to happen due to cost and there aren't
any windows. (I've seen that before too - made me smile.)

I've been asked to calculate the loads for the new HVAC with planned
significant envelope upgrades (they are actually going to insulate the
walls and put in double glazing). I've been running loads in both
eQuest and HAP and wanted to try and model the wall "right".

I'm going to try the easy stuff first: adiabatic, exterior wall with
no sun, and then play with another unconditioned space like Nick and
Vikram suggested.

Thank you for the great feedback - I don't think I would've thought of
just trying an exterior wall with no sun exposure.

Jason Quinn

JasonQuinn's picture
Joined: 2012-03-03
Reputation: 0