.

Dynamic Airflow and Energy Modeling...

Using Computational Fluid Dynamics (CFD)

.

Flomerics
► Founded in 1988
► By a core team with extensive background
experience in commercial general-purpose CFD
► Offering FLOVENT to the architectural
engineering industry since 1989

Corporate Mission:
To put CFD technology into the hands of engineering designers
.

Global Sales & Support Infrastructure

UK (Ltd. HQ)
US: Santa Clara, CA
Austin, TX Germany, France, Italy
Marlborough, MA (US HQ)
Sweden

Japan*, Korea*

Taiwan*
China

India
Singapore

* Via Distributor
.

Why use CFD Modeling?

► The physics of air and heat movement are extremely
complex and cannot be predicted by simple guidelines
► 3D CFD simulation provides the ONLY way to predict
air and associated energy flow behavior prior to build
► Enables reduction of design margins,
reducing all construction and operating costs
‰ Capital
‰ Change orders
‰ Commissioning
‰ Energy
‰ Liability and health issues
.

Cost of Design Changes

Make Better Decisions Earlier

$1,000,000
The cost of repairing mistakes
increases by roughly an order
of magnitude at each stage!

$100,000

$10,000
$1,000

Conceptual Detailed During In Operation

Design Design construction
.

What is “Design-Class” Analysis?

► “Design-Class” analysis is drastically

different from traditional analysis…
– Focus on real engineering design
problems and the fastest and easiest
methods to solve them

– Aimed at expert and non-expert users

…the shortest and
– Bridging the gap between science easiest route!
and industry
.

Airflow Modeling with FLOVENT allows you to predict …

► Air Velocity ► Pressure
► Temperature ► Contaminant Dispersion

Internally and Externally

.

With FLOVENT CFD modeling

optimize a built environment for :
► Realistic Energy Requirements
► Air Quality
► Comfort Indices
► HVAC System Specifications
► Thermostat Locations and Scheduling
► Envelope Design (including green roofs)
► Environmental Interior Design
► Building Green Intelligently!
.

FLOVENT CFD simulation vs. Smoke Test

FLOVENT Simulation Smoke Test

Clear Cloudy

Colorful Expensive!
.

FLOVENT Applications
These are just a few
Office Environment and
examples of FLOVENT
Comfort Modeling
applications

Cleanrooms and
Contaminant Modeling

External Flow and

Dispersion Modeling

Data Centers and

Cooling
Optimization
 .

CREATING A FLOVENT CFD MODEL

Design and Analysis of a Hospital Ward
For this demonstration, we will:
1) Start with a single Patient Room
2) Follow all key steps of a basic
modeling project
3) Expand to model the entire floor
of a hospital wing
4) Highlight FLOVENT features:
Ö Geometry
Ö Gridding
Ö Solving
Ö Post Processing
Ö Outputs
Ö Solar Radiation
Ö Transient Analysis
Ö Optimization
 .
Starting FLOVENT
• The FLOVENT Project Manager contains the primary menus (at the top), common short cut tool bars, assembly node tree,
component creation (icons in the upper-right), and module access (icons at the lower left).
• The useful Library Manager is shown with some of the resident libraries (clear folders) + personal libraries (Data Center, Wark).
 .
Geometry
• Our Hospital Ward model begins with the creation of 3-D geometry in the Drawing Board.
• First, the Enclosure “Smart-Part” is used to specify construction of the walls of a single Patient Room.
• Geometry can also be imported using the FloMCAD add-on CAD geometry translator.
 .
Properties
• We create a Hospital bed w/ patient assembly by modifying a person assembly from the library and putting him/her in a bed
modeled using simple rectangular “cuboids” (top right icons in the Drawing Board).
• The cuboid can be renamed and material, surface, and thermal properties can be specified for the part.
 .
Geometry
• We finish the 1st half of the Patient Room by including a Tray subassembly in the Hosp bed w/ patient assembly, a Stuffed Chair
assembly for a visitor, an Instrument cart, wall lights, closet, counter-cabinet, and TV.
 .
Libraries
• Since we will want to use hospital bed assemblies and stuffed chairs again for future models, we store them in a personal library.
• All properties of all assembly elements are retained when an assembly or part is saved in the library.
• Assemblies can also be saved directly to other folders in your computer.
 .
Geometry
st
• All the elements of the 1 half of the room are copied to the other half and a bathroom and curtain are added in-between.
• Copying is done with regular copy-and-paste, patterning, or creating a mirror image across a datum plane.
• Note that we changed the Drawing Board to single pane mode. The image can also be zoomed and rotated in each pane.
 .
Smart Parts
• Air supply diffusers are created in the ceiling using the Diffuser Smart Part and its attributes dialog window.
• Once the first one was created, it was copied. The 2nd one is located at the other side of the room.
• A single return grill was modeled using a Fixed Flow Smart Part and located in the ceiling near the 2nd closet.
 .
Web Tools
• In addition to Smart Parts, common assemblies can also be created using Web Parts found at User Support of www.flovent.com .
• Here, a grille diffuser is specified and saved to the same folder as the Hosp Bed w/ patient and Stuffed Chair assemblies.
• Note the other useful web pages found at the User Support section of www.flovent.com .
 .
Glazing
• A window is created by locating a hole in the Enclosure wall near the 1st bed and replacing the hole with glass.
• The material and radiative properties dialogs for the glass can be accessed through the Hole construction dialog.
• At this point, the geometry is adequate for modeling a hospital room with the door closed. Hosp room 1 can be saved to the library.
 .
Geometry
• We will go ahead and create a doorway by cutting a hole in the wall opposite the window.
• The door is swung open inward. This means that the TV needs to be located at the other side of the counter-cabinet.
• This open-door geometry for Hosp room 1 will be used for the rest of the software demonstration.
 .
Initial, Ambient & Boundary Conditions
• Before solving, we set the system attributes: physical orientation and initial, ambient and boundary conditions.
• These attributes are especially critical for the modeling of natural ventilation and “stack effect”.
 .
Contaminant Tracking
• Along with the modeling of air flow and heat transfer, contaminant concentrations can be included.
• FLOVENT comes with a number of fluid and particulate contaminants in its library. We will track small bacteria from a patient.
 .
Sources
• We will track the bacteria contaminant concentration from the patient nearest the door. A “Source” (red icons at right) is used to
specify the rate of release of the bacteria. This particular Source is 2-dimensional and located at the face of the patient.
• Sources can also be used to set heat generation and pressures at specific locations.
 .
Monitor Points
• We can also add “Monitor Points” to the model to provide temperature, pressure, airflow, and concentration data at specific points.
• Since we are primarily interested in the health and comfort of each patient, we place a Monitor Point near the upper chest of each.
• As you will see later in the demonstration, Monitor Points are useful for tracking solver convergence and design optimization.
 .
Gridding
• The solution grid can be set automatically or manually. Initially, we will try a “Medium” automatic setting.
• Each 3-dimensional “cell” region located between adjacent grid lines defines the local volume in which temperature, pressure and
concentration are calculated. The pressure difference between adjacent cells determines the local air flow rate.
 .
Sanity Check
• The Solve menu provides access to all of the solution controls which include multiple setting options for convergence criteria.
• This menu also includes a Sanity Check which finds any potential problems in the model before starting a solution.
 .
Sanity Check
• The Sanity Check revealed a potential problem with the solution grid being too course for proper diffuser modeling (6 Warnings).
• The model would still run since there are no Errors, however, the accuracy of the flow calculations could be diminished.
 .
Localized Grid
• Rather than add more grid lines to the entire model and make it very large and possibly unstable, we will use “Localized Gridding”
which limits grid lines associated with the part (or assembly) to the space it occupies or a little beyond it.
• In the Diffuser Smart Part attribute dialogs, the grid lines associated with the diffuser are forced to be denser.
 .
Gridding
• The grid is further refined by adding Localized Grid and additional grid constraints to the other diffuser and the 2 bed assemblies.
• The system grid settings were also changed manually to obtain a reasonably denser grid while still keeping the total cell count low.
 .
Solving
• After starting the solver, we see that the Warnings have disappeared.
• The solver progress is tracked in the Profiles window which shows the remaining residuals and Monitor Point temperatures.
• Note that after 1 minute, the solution is 40% converged and the Monitor Point temperatures are within 1 degree of equilibrium.
 .
Solving
• The convergence criteria for this model is reached in about 8.5 minutes.
• For reference, the computer used to create and solve this model had a single 2 GHz processor and 2GB of RAM.
 .
Post Processing
• In Flomotion, FLOVENT’s post processor, the geometry and data can be displayed in many ways. Along with the data presentation
features, Flomotion has many of the same image manipulation capabilities as other popular graphics software packages.
• Initially, our hospital room is viewed as from the hallway without any data displayed.
 .
Post Processing
• Before looking at thermodynamic data, we will set up our model for optimal visualization. The ceiling and 2 walls have been made
invisible and the model has been rotated for a y-axis view. Note that the view has perspective.
• The appearance of each element can be changed, including graphical finishes which come with FLOVENT or can be imported.
 .
Post Processing
• To visualize thermodynamic data, a 2-dimensional color-coded map is displayed at any point along an axis.
• Both scalar and vector maps can be displayed for a large selection of calculated data sets.
• Here, a cross section of the room Temperature is shown 6.72 ft. from the near wall along the x-axis.
 .
Post Processing
• Geometry can be shown in wireframe, allowing visualization of the conduction temperature gradients through solid objects.
• The model can also be shown without perspective (orthogonal). Background colors can be changed for better contrast.
• The color scheme and range of the map can be adjusted to optimize the clarity of the data. Units can also be changed.
 .
Post Processing
• Points on a map can be labeled with scalar value and location. The labels remain until turned off, so they can be used for reference
if the map changes for any reason. Different scalar labels can be put on a map, e.g. a Velocity label can go on a Pressure map.
• The labels travel with the map if the map is moved along the axis. Only the current axis value changes when the map is moved.
 .
Post Processing
• Velocity vector maps can also be displayed along any axis.
• Here, the flow just below the diffusers is shown. The legend shows the vectors color coded for speed (magnitude).
• Note that a model image can be rotated, panned, and zoomed.
 .
Post Processing
• Vector maps are actually 3-dimensional. Each arrow starts in the selected plane and points in the direction of flow.
• Here, we see the edge of the map shown in the previous slide.
• The high-velocity vectors representing flow out of the diffusers are angled down at 30 degrees.
 .
Post Processing
• Here is another velocity vector plane located on the same plane as the Monitor Point for the patient nearest the door. This plane will
be used for examining other air flow related data such as indoor air quality (IAQ) and comfort indices.
• The appearance of the vectors can be adjusted for size, range, and resolution.
 .
Air Quality Analysis
• Flovent calculates 2 common air quality indices: LMA (Local Mean Age of air) and LACI (Local Air Change Index).
• This LMA map shows that the air from the nearest diffuser flows directly to the foot of the bed while the air at the head is not fresh.
 .
Comfort Analysis
• Flovent calculates 2 Fanger comfort indices: PMV (Predicted Mean Vote) and PPD (Predicted Percentage Dissatisfied).
• Here, PMV is shown. The air at the foot of the bed is cool while the radiant heat from the lights keeps the upper end comfortable.
• Note that Flovent also calculates Percent Saturation ∼ relative humidity.
 .
Contaminant Tracking
• Contaminant Concentration 1 (renamed Bacteria since the properties of small bacteria are attached to it) is shown here.
• The units are set to PPM and the upper end of the range is lowered manually for better visualization of low concentrations.
• Note that the transparency of the data map can be adjusted to allow visual reference of objects behind the plane.
 .
Post Processing
• Multiple data planes with velocity vectors and other scalars can be displayed simultaneously.
• Here, a 2nd y-axis plane is added to show how the bacteria will spread across the room. It indicates that the bacteria gets carried
over to the near wall where it can flow over the chair and possibly out the door.
 .
Post Processing
• In addition to 2-D data maps, Flovent can visualize 3-D flow using virtual streamline particles that emanate from flow sources.
• The shape, size and number of particles can be adjusted. The color of the particles can represent scalar values.
• Both 2-D and 3-D visualizations can be animated for greater effectiveness. See the Flovent_anim_Jan06 demo for examples.
 .
Outputs
• Any images created in Flomotion can be exported in common graphics formats.
 .
Outputs
• All data that can be displayed in Flomotion can also be displayed in the Tables module.
• Tables data can be directly exported to text or spreadsheet files for further analysis.
 .
Model Expansion
• Our Hospital Ward is easily created by copying or patterning the Patient Room for a total of 6 rooms and adding a Waiting Room
and a central Nurse’s Station using assemblies from the library. Additional ventilation and simple utility rooms are included.
• Assemblies are rotated, moved, and modified for the final layout. Boundaries are set and the model is regridded and then solved.
 .
Post Processing
• The same data maps shown for the single Patient Room model can be displayed for the Hospital Ward.
• Maps can be moved throughout the building along each axis to find areas requiring design improvement.
 .
Post Processing
• The entire floor of a building can be modeled for more accurate whole-building analysis.
• Velocity vector fields can be used in conjunction with other data maps to optimize the layout of open multi-zone areas.
 .
Post Processing
• The importance of using the proper air flow analysis tool is seen in this image of Bacteria dispersion throughout the Hospital Ward.
• Even with the Patient Room return vents located above the door, airborne bacteria is transported directly out into the central area of
the Hospital Ward. FLOVENT reveals this and is then used to design an optimized system BEFORE ductwork is installed.
 .
Sunlight
• Solar loading can also be included in a Flovent model, including the visualization of sunlight through a window (Solar Viz).
• Solar modeling can be done automatically using position, time and cloudiness, or the intensity and angles can be entered manually.
• When including solar radiation in a model, thermal radiation exchange factors should also be calculated.
 .
Solar Heating
• By incorporating a Temperature map and a Solar Viz map, the thermal impact of the sunlight can be seen.
 .
Transient Analysis
• Transient analysis is another feature of Flovent. Models can be solved according to a transient time map to examine the impact of
changing ambient conditions, such as sun position, or indoor activity more accurately.
• This feature is particularly useful for designing building envelopes, window treatments, and establishing HVAC schedules.
 .
Optimization
• Optimization tools are another feature of Flovent. To demonstrate optimization, we will use the previously shown case of PMV
comfort index but with the wall light turned off. It is now too cool near the Monitor Point and LMA is marginal (note the LMA label).
• The diffuser location and angle will be optimized for comfort without decreasing fresh air flow across the patient.
 .
Optimization
• In the Command Center module, ranges for diffuser location and angle are set for a Design of Experiment (DoE) which
automatically iterates to find Monitor Point results that are closest to the selected target values (lowest Cost Function).
• The best result was found in the 6th iteration with the diffuser moved 3 inches toward the door and the angle increased 4.5 degrees.
.

Summary of Real FLOVENT Value:

; More accurate environmental modeling

; Fewer change orders

; Faster time to occupancy

; More versatile commissioning

; Lower energy and health-related costs

; Happier occupants!
.

For more information, visit our website

www.flovent.com