info about TYPE 35

["Pietro Filippi" <P.Filippi@xxxxxxxxxx>]

Hi,
I am a new user of TRNSYS (version 15), so I am a beginner. I just
studied examples 5, 6 and 7.
I have a generic office room 5x5x3.1 m (LxWxH) in a side of a big
building, at 30 deg of azimuth. Its external wall is (almost) entirely
made of glazings.

- Simulation 1:

I use type 19 Wind-2 with:
Diffuse transmittance	0.21	dimensionless
Inside convection coefficient	28.8	kJ/hr.m^2.K
Transmittance	0.21	dimensionless
Loss coefficient	4.9	kJ/hr.m^2.K
Fraction striking illuminated surface	1	dimensionless


- simulation 2:

I use type 35b combined to type 19 Wind-1 (as in example n.7 but without
overhang). Type 35b has:
Number of glazings	3	dimensionless
Extinction	0.05	dimensionless
Refractive index	1.526	dimensionless
All other parameters and input are not varied respect to sim 1. 

- simulation 3:

I use type 35a in the place of type 35b. The output-input connections
between 35a and 18wind1 are:

Energy flow to room ----> Thermal energy gain
Solar radiation to room ----> Total solar through the window
Thermal energy to room        Beam rad through window 0
					Frac striking illuminated
surface 1

Considering the simulation output QCOOL I got that the Sim2's QCOOL is 5
times sim1's while sim3's QCOOL is 3.5 times sim1's!!!!! 
How is it possible?

What can be the most appropriate modelling setting for simulating this
kind of room?
Is sim1's model too simple or sim2's inadequate?

In sim 3 I considered type19 wind1 input "Thermal energy gain" as total
gain even if in type 35a output is specified that "thermal energy" is
the one from the ambient. So I connected its first input (Energy flow to
room) to the input "Thermal energy gain"  because it rappresents the
total. I left "Beam rad" unconnected and set to 0 cause it is just a
part of the total radiation through the window.

Is there anybody that can tell me where I am wrong?
I hope you can help me very soon! (my boss needs result...)
With regards 
Pietro Filippi