Issue1433 update floor on ground

[annadellisola]

This PR addresses issue #1433

[jelgerjansen]

@annadellisola please update the reference results as well.

[annadellisola]

Changing the IDEAS.Buildings.Data.Constructions.FloorOnGround, the following reference results change:

IDEAS.Buildings.Components.Examples.RectangularZoneEmbedded

[annadellisola]

The new TSensor in IDEAS.Buildings.Components.Examples.RectangularZoneEmbedded was expected to be lower due to the change of the locGain position from 3 to 2. As a result of the different heat injection position, the temperature values of layer 4 (tile) are lower, while the temperature dynamic is faster due to the lower thermal capacity of the Tile compared to Cement.
In the plot below, simulation //1 is with the new FloorOnGround implementation, simulation //2 is the old reference:

On the other hand, the new temperature of layer 1 (cement) was expected to be higher. This is true for the first 6h, after which the reference temperature in simulation //2 becomes higher and the screed layer has been thermally activated. Moreover, the Q_flow injected in the floor heating (Q = (Tin - heatPortEmb.T)*min(G_t, G_max), where Tin is fixed equal to 30°C) is lower in simulation //1, compared to simulation //2.

Additionally, it seems strange that the new TSensor experiences an initial drop. However, this can be explained looking at the TRad and TAir. The TRad in the new simulation //1 is not compensating any more the drop of TAir.

[jelgerjansen]

@annadellisola as discussed online, I also updated the RadSlaCha parameters according to the used floor type. This gives the following results (compared to the old simulation, not your latest changes):

Similar to your initial changes, the room temperature decreases.