Question-and-Answer Resource for the Building Energy Modeling Community
Get started with the Help page
Ask Your Question

Question for the Energyplus example file:AirflowNetwork_Multizone_HorizontalOpening

asked 2018-12-02 16:09:19 -0600

Haohan Sha gravatar image

updated 2018-12-02 16:09:53 -0600

Hello Everyone! I am confused by the description in the I/O reference and the example file of the horizontal openings. On the Page 1041 of the I/O reference, it is said:

Cannot do or restricted: Air circulation and/or air temperature stratification within a thermal zone. For example, you should not try to divide a high space, such as an atrium, into subzones separated by artificial horizontal surfaces that have cracks or openings with the expectation that AirflowNetwork will give you a realistic temperature in each subzone and/or a realistic air flow between subzones.

However, in the example file, AirflowNetwork_Multizone_HorizontalOpening, it creates a large space and divide it into two zones with a large horizontal opening in the middle. As it mentioned, if the AFN cannot calculate the temperature stratification simply, how can this example file get correct result?

edit retag flag offensive close merge delete

2 Answers

Sort by ยป oldest newest most voted

answered 2018-12-02 20:10:21 -0600

I believe the I/O reference is referring to a situation where a large space is divided up artificially into more than two fictitious subzones to try to capture stratification and circulation, not a situation in which there are two zones connected by a real opening. Dividing up a large space like that is something that many people have tried but few have been successful at. Note that the I/O ref says "within a thermal zone" in the first sentence of your quote.

edit flag offensive delete link more


I thought of this situation before. But I think the description is ambiguous. If two vertical spaces can be connected by a large horizontal opening and the AFN model can give a real result, there is no reason to say that AFN cannot calculate temperature stratification, because this two vertical spaces can have different temperature and pressure. I think that the reference's description means that if we divide a high space into some subzones, it must have some artificial floor slabs in the middle, and lead to wrong calculation. But I don't know whether I'm correct...anyway, thank you.

Haohan Sha gravatar imageHaohan Sha ( 2018-12-02 20:19:29 -0600 )edit

No, it's not ambiguous. There's no reason to expect that artificial subdivision of a space and application of AFN's pressure network will work, and it's not the same thing as two spaces separated by an opening. The pressure network fundamentally conserves mass. To capture air movement and temperature stratification in a large space additional physics are needed. That's why the answer always gets back to CFD, where the additional physics (e.g. conservation of momentum) are available.

Jason DeGraw gravatar imageJason DeGraw ( 2018-12-02 20:31:52 -0600 )edit

This explanation is more reasonable. I checked the RoomAir Model parts in the I/O reference, it said that the airflow in a large space would be more complex, and hence it cannot use the Artificial AFN model to simulate it. It means that AFN can be used to calculate the airflow between zones horizontally and vertically, as it mentioned in the engineering reference, but it cannot be used to simulate airflow inside a large space, and it needs more physics (Roomair or CFD) to simulate airflow patterns in a large high space. Great answer. Thanks a Lot.

Haohan Sha gravatar imageHaohan Sha ( 2018-12-02 20:42:05 -0600 )edit

By the way, in the part of RoomAirflow network model, if it needs to connect nodes in a single zone, linkage such as crack or horizontal openings, is required. If roomairflow network model can calculate temperature stratification in one zone,and it also does the similar things as mentioned above (adding openings between two nodes) why energyplus doesn't combine these two would be more simple... Just Kidding...

Haohan Sha gravatar imageHaohan Sha ( 2018-12-02 20:50:02 -0600 )edit

Well, the two models are combined in the sense that you can connect the RoomAir AFN to the "regular" AFN, but it's still going to be difficult to get good answers unless you know what the flow patterns are supposed to look like. We did this with a large space that we had CFD results for, in that case we could compare the two results and adjust the linkage flow components to get similar results. Without the CFD (or measurements or similar) then it comes down to an engineering judgement. Which isn't going to convince everyone.

Jason DeGraw gravatar imageJason DeGraw ( 2018-12-03 04:38:31 -0600 )edit

answered 2018-12-03 09:42:54 -0600

Haohan Sha gravatar image

OK This is a summary from Jason DeGraw's answer. The regular AFN model is based on pressure and temperature difference in different zones (see Engineering Reference in E+), and so it cannot be used to calculate temperature stratification in one thermal zone,(the restriction mentioned in I/O reference), because the theories of temperature stratification in one zone are more complex then the assumption of pressure model in AFN. For example, when dividing a large space into some subzones, you actually make a wrong assumption that the large space has a temperature stratification with the same pattern of the subzones you defined. And of course, that is incorrect. No one knows the real airflow pattern in a large zone without more advanced simulations or measurements.

If you want to know the true flow pattern in one single zone, it needs CFD or measurement, and then calibrate your Roomairmodel in the Energyplus. When you cannot use CFD or other methods, how to define the Roomair model is based on your engineering judgement. (be careful)

Thanks for Jason DeGraw's patient explanation.

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer


Question Tools

1 follower


Asked: 2018-12-02 16:09:19 -0600

Seen: 70 times

Last updated: Dec 03