It’s modelled right or should i find the way to imput the constant gradient formula?
The paper you referenced used the Room Air Model feature of EnergyPlus, which allows the user to specify some kind of air temperature change within a zone instead of the default approach of having the entire zone's volume all at the same air temperature. There are many ways to do this, but the paper used two:
- Constant Gradient - Temperature changes from floor to ceiling within a zone at the same rate.
- Non-Dimensional Height - Temperature changes from floor to ceiling within a zone at a user-defined rated. User enters pairs of normalized zone heights with corresponding temperature differences from zone mean air temperature. This would handle the paper's following finding "In general, the air temperature increases gradually along with the height and the temperature gradient becomes fairly large in the region near the top of the atrium."
Note that the paper used temperature results from a CFD simulation (Figures 3-6 in the paper) done as a first step to know what the proper temperature values were to enter into the Room Air Model of EnergyPlus. If you want to repeat their approach, you will need also perform a CFD simulation.
I’m not getting any sun gain on the ground floor, its because of the air walls/floors?
My guess is that Rhino or Honeybee is using the same "Air Wall" approach as OpenStudio -- flagging a surface to be transparent ONLY for a daylighting simulation, not an EnergyPlus simulation. An "Air Wall" is actually 1" of gypsum board. If you have access to the EnergyPlus input file ending in .idf generated by Honeybee, check for the following:
- Find one upper-level zone using "Air Wall" for its floor surface
- Search for that zone's name in the IDF to find its floor surface
- Search for the Construction object with the same name referenced by the floor surface
- Search for the Material object(s) referenced by the Construction object and examine their heat transfer property inputs
If different stories are ok, should i use more zones (splitting each branch in 3 more zones)?
That seems unnecessary. The paper you mentioned recommends "**... the atrium space can be divided into two sections vertically according to air temperature stratification:
- From the floor surface to the plane surface 10m below the roof surface. In this region the air temperature varies little, generally from setpoint to 2$deg$C higher than the setpoint, because the solar radiation has little effect in this region.
- From the plane surface 10m below the roof surface to the roof surface. In this region the air temperature gradient is very big due to solar radiation effect.**"