Announcement: Wind Pressure Coefficients-Integrated Building Energy Modeling Workflow Launched!

[MichaelDonn]

Kia ora @Naga. No worries for the delay. My hope was that, as this was part of your thesis, then some outside assistance with Quality Assurance might be of assistance in progressing the thesis as well. I recognise that the system cannot in reality be a part of the actual research, it is a means to an end. However, I was concerned as much for the accuracy of your calculations as for the benefit to my students.

Thanks for the rapid response. I felt the need to re-look because I was hoping that the odd result would not have invalidated work already done, or caused a deep dive that was indicative of a thesis issue.

Look forward to you finding the time.

[Naga]

@MichaelDonn

Thank you for your thoughtful message, and I truly appreciate your concern regarding the reliability of the results.

I’d like to reassure you that the issue is not from the simulation workflow itself, but rather from the specific model inputs currently being used.

1. Window Operation Inputs Missing
The model appears to lack details regarding window operation schedules and control parameters. This omission can significantly influence the simulation outcome.

2. Wind Pressure Profiles
For most orientations in your current setup, the wind pressure values are negative. This implies airflow directionality away from the zone, contrary to what is typically assumed in standard modeling methods.

3. Discharge Coefficient Representation
While conventional methods often apply a constant discharge coefficient, our parametric approach uses orientation-specific values derived from CFD-based data. This distinction plays a crucial role in accurate air movement predictions.
(Conventional)

(Parametric approach)

4. Window Operation Assumptions
In the conventional model, windows are always open, which naturally leads to higher energy use. This contrasts with our model, which implements dynamic window operation based on indoor and outdoor conditions.

While ladybug-tools offers an intuitive GUI, its true strength lies in its capacity for detailed spatio-temporal analysis. This enables us to trace and interpret unexpected simulation results with precision. This analytical layer is always employed by us to ensure that every model is rigorously validated. Hence, I hope this addresses any concern regarding the integrity of the results within the thesis.

If you’d like to explore our methodology further, I highly recommend reviewing our recent publication article along with E+ Engineering Reference and Input Output Reference.

[MichaelDonn]

Kia ora @Naga

Thank you for your careful outlining of the differences that appear between your two models built into your demonstration script.

Apart from changing the geometry, and updating the version of LadyBug Tools, I was assuming that the two examples you showed in your workflow were only different in the application of the CFD estimates of wind pressures.

However, it would seem that there is more happening inside the calculations that you are performing than merely applying ventilation wind pressures.

I am sorry to keep bothering yoou, but, in order to understand this further, so I could describe the situation to my Masters students, I have dug into the actual differences between these two runs, and I am more deeply puzzled than before.

As I understand it, the model for the “Standard Simulation” is the exact same model that has been fed into your CFD analysis to create the. The difference between the Standard and the WPC Simulations can be seen in this screen capture

.

The base model that was fed into the Pressure Coefficient calculations is connected at the Blue boxes. Thus, it can be seen that for the standard simulation a basic AFN Airflow Network Component is added in (upper red dotted box, with base model connection at blue box). For the Integrated Simulation, the model is connected directly because the AFN AirfFlow Network description is part of the text string generated by your CFD calculations and connected at the red arrow.

Thus, there are two differences between the simulations

1. addition of a very basic Air Flow Network definition, with basic default values, to the Standard Simulation that is not in the WPC Integrated simulation.
2. addition of the airflow network definition with calculated wind pressures to the Integrated Simulation (arrowed entry) via an additional “text string”

So, to understand the two calculations better, I tried rearranging the inputs.

I tried

1. just connecting the additional WPC text string to the standard simulation - crash, object naming issues
   

   

2. bypassing the AFN definition for the Standard Simulation and connecting the WPC text string - outputs identical, as hoped for
   

   

3. running the standard simulation as first set up in the workflow
   

   

4. stripping out the wind pressures from the WPC text string file but leaving in the rest of your AFN text strings and connecting these to the add string input of the Standard Simulation
   

   

5. disconnecting the AFN definition but leaving your WPC added string attached also with the Wind Pressure Coefficient data excluded for the Standard Simulation
   

   

What is striking to me is that 4) is using all the setpoint settings that you described that should be different and the energy use index is still far bigger than for the version with your full wind pressure data accounted for. This still seems strange to me in that the HEATING energy is so different, and that this is not cured by the addition of your AirFlow Network definition without the associated Wind Pressures .

m