multiple surfaces occuring on one plane

[macpavee]

Hi

I am modeling a vaulted room within another, effectively nested polyhedra.

each zone (inner and outer) comprises three layers, basement, room, atic.

the central room and it's attic are a single space.

however the outer attic then cuts through the inner space with a new ceiling -unwanted.

then if i cut a hole into the attic of the outer zone to allow the inner to sit unobstructed, apache sim regards this as an external hole and wont allow.
the work around appears to be a permanent window open 100% of the time. -which seems crude, is this the only way? -subtracting rooms doesn't seem to work.

although if i combine relevant rooms through the model this does seem to work to produce a simple model, however i lose some of the distinctions between the zones (eg.upper and lower) useful for more indepth thermal analysis

in addition
although the floor doesn't need cutting out, there now appears to be four surfaces which need to be assigned their own construction type. the basement ceiling of both inner and outer, and the floor of both inner and outer.
but i wish to have a different construction for inner and outer.

i cant understand how the simulation now views this construction, once set.

does the simulation add each 'overlapping' floor type together thus giving me a false U value?

is there a better way to achieve this?

[Complex Potential]

It sounds as though you have constructed the inner room as an object inside the outer room. Normally this would create a collision error unless it has been suppressed.

If this is what you have done I would suggest trying to rebuild with the outer zones "wrapping around" the inner zones. You sometimes need to draw the outer zone in two halves to achieve this.

I'm honestly having a tough time picturing what's going on by your description only. Perhaps you could post a link to an image that might make it clearer?

[macpavee]

yeah, multiple rooms within rooms.

dont mind all the glass!
the top is meant to be an inlet but from what i gather i cant leave it open so i have included window openings.

i see what your saying with the 'wraparound' rebuild and can recreate this design easily enough i think. just build the central stack (which is meant to be thermal mass) and a one/eighth part of the room (and perimeter) as a cut through section, then copy and rotate the part eight times. rather than two halves.
if that actually makes sense either...
but such a rebuild will also grant me a lot more control over the analysis too

nevertheless, the model above has been working on the apache simulation now i changed holes into permanently open windows. but i am not sure how to 'supress'
however i still dont know how the sim is reading the multiple floor construction. also the water as thermal mass which is meant to be the central pillar is'nt working at all as it should. will this new model build fix that?

[Complex Potential]

As a general rule, each external surface or internal surface between zones should only allow a single construction type to be applied to it. If you are able to assign different constructions to the same surface from different sides then it suggests the model is not interacting in the way it should and the main culprit I would say is the method you have used to construct. You could try using the "rebuild" command to see if that consolidates things, although that could create more problems if the model geometry is a bit iffy (so make a copy first!)

By rebuilding in the way I previously suggested you should find that the models behaves in a more logical way.

BTW, a permanently open external hole can be modelled as a "window" with 100% open macroflow profile and a g value of 1.0.

The central pillar water thing is a totally different issue and is probably best discussed in your apache forum post. However, I would suggest getting the model geometry working correctly before attempting to tackle that issue.

[PCully]

Just a thought but when you say you can assign two constructions to the surface, I assume you don't mean it's an internal ceiling and you can assign it either as the ceiling of the room below and the floor of the room above? in that case whichever route you go it will replace the previous assignment.

Just in case it's as simple as that?

If not and it's behaving erratically in terms of these assignments then I agree you likely have some issues with the adjacency not being identified properly and you should review the geometry.

as always (and already noted) it's very important to get your geometry correct and robust before moving to analysis, I know you are always eager to get into the detail of the assessments you are running and no doubt have a time constraint on you but if you spend the time to understand how the ModelIT geometry works and how to create good quality models then you will achieve far more with the VE analysis packages. Especially where you are looking at creative solutions to problems

Phil

[macpavee]

Thanks Phil & CP

It's good to know how IES is thinking and working. i.e. that it is re-assigning the construction by the last entry. although in addition to the ceiling/floor effect, because i had one room within the other, both with floors, i wasnt sure how IES was assigning the construction. I'll assume the same process in that the last assigned is the one used. However, out of curiosity, as one floor (the outer) overlapped the inner may there still be an overlapping of the construction?

regardless of this, I have now rebuilt the model in the manner advised, which in fact took no time comparably. plus i have a lot more options for the simulations in which i can keep a great number of rooms and finely monitor thermal changes, or merge/combine rooms and make a simpler model for quicker simulations. WIN WIN.

however, on the down side, i dont seem to be getting responses on the water as thermal mass problem in the apache forum!
would love to see IES embrace passive solar fundamentals more readily.

[Complex Potential]

Hi Mac

Glad to hear that you have sorted the geometry. Overlapping and intersecting geomoetries is a common issue and not always easy to spot. IES has an automatic check for intersections as a command which you can try but I like looking at the constructions assigned to each surface; if you look carefully at the internal surfaces you may spot one that IES is defining as external which is a dead giveaway.

The responses you get on this forum are mostly from other users (and it is quite a small community) so a lack of response in the short term is sometimes to be expected and doesn't really mean anything about how IES view certain issues. However, I will pop over to the apache forum now and attempt to get the ball rolling on that one.