Imagine you need to simulate a diesel fuel spray. This spray contains millions of tiny droplets. If you try to track each droplet one by one, your computer will need weeks or even months to finish the calculation. I remember facing this problem in my early CFD projects. It is a very common challenge.
The parcel concept solves this issue. Instead of tracking millions of individual particles, we group them into “parcels.” But what is a parcel? Each parcel represents many particles that share the same properties. This makes your simulation much faster while keeping the results accurate. (Note: If you are looking for a basic definition of DPM itself before reading about parcels, please check our Discrete Phase Model Basics blog).
In this guide, you will learn:
- What a DPM parcel is.
- Why the parcel model makes your simulations over 100 times faster.
- How to calculate the number of particles per parcel.
- Which parcel release method is right for your setup.
Figure 1: An illustration of the DPM Parcel concept. Instead of tracking each individual particle (small dots), the simulation tracks a single ‘parcel’ (the large blob) that represents a group of particles sharing the same properties.
What is a Parcel in DPM?
In simple terms, a parcel is a single computational group that we track in a simulation. This single parcel stands for many real particles that have the exact same properties.
Therefore, all the particles inside one parcel share the same:
- Size (diameter)
- Speed and direction (velocity)
- Location
- Temperature
- And other physical properties.
This parcel concept was first introduced by Dukowicz in 1980. It was created to solve the problem of simulating huge numbers of particles. Because all particles in the group are identical, the software only calculates the path of the parcel itself. It is assumed that every particle within that parcel behaves in the exact same way.
Interestingly, a parcel can represent a non-whole number of particles, like 308.04 particles! This is simply a mathematical method used by ANSYS Fluent to ensure the total mass flow rate remains perfectly accurate.
Figure 2: A simple diagram showing many small dots (particles) being grouped into a larger circle (parcel)
Why Use Parcels Instead of Individual Particles?
The main reason we use the parcel model is to save a huge amount of time and computer power.
Let’s look at a real example. A typical diesel fuel spray contains millions of tiny droplets. One study showed that tracking just 5,000 individual particles took 12 hours on a strong 8-CPU computer cluster. Now, imagine a simulation with 10 million particles. Tracking each one is practically impossible. The parcel concept fixes this. By grouping thousands of identical particles into a single parcel, the computer tracks just a few thousand parcels. This reduces the number of calculations drastically. Your simulation can run hundreds of times faster. Here is a simple table to show the difference:
| Feature | Tracking Individual Particles | Using DPM Parcels |
| Computational Cost | Extremely High | Low / Manageable |
| Simulation Time | Days / Weeks / Months | Hours / Days |
| Memory Needed | Very Large | Small |
| Feasibility for Large Problems | Often Impossible | Possible and Practical |
How to Calculate the Number of Particles Per Parcel
Now for a very practical question: how does the software know how many particles to put in each parcel? You can calculate the particles per parcel yourself to verify your setup.
The calculation is very simple:
Number of Particles per Parcel = (Mass of one Parcel) / (Mass of one Particle)
![\[ N_p = \frac{m_s \cdot \Delta T}{m_p} \]](https://cfdland.com/wp-content/ql-cache/quicklatex.com-5a73e9d7794c5bded679d19fe211424e_l3.png "Rendered by QuickLaTeX.com")
Step 1: Find the Mass of a Single Particle
First, we need to know how much one single particle “weighs.” Since DPM assumes particles are perfect spheres, we use the formula for the volume of a sphere and multiply it by the particle’s density.
Particle Mass = Particle Density * (π/6) * (Particle Diameter)³
Step 2: Find the Mass of a Single Parcel
Next, we calculate the total mass represented by one parcel. This depends on your total injection mass flow rate, how often particles are injected (time step), and how many injection streams you have.
Parcel Mass = (Injection Mass Flow Rate * Particle Time Step Size) / Number of Streams
Step 3: Put It All Together (A Practical Example)
Given Inputs:
- Injection Mass Flow Rate: 0.0001 kg/s
- Particle Density: 1550 kg/m³
- Particle Diameter: 0.00001 m (or 10 microns)
- Particle Time Step Size: 0.0005 s
- Number of Streams: 200
Calculation:
Particle Mass=1550×6π×(0.00001)3=8.116×10−13 kg
Parcel Mass=2000.0001×0.0005=2.5×10−10 kg
Number of Particles per Parcel=8.116×10−132.5×10−10=308.04
As you can see, the result is 308.04 particles per parcel. It is perfectly normal and correct for this to be a fractional number. It is a mathematical method Fluent uses to ensure the mass in the simulation is perfectly conserved.
Figure 3: A comparison showing the input parameters in ANSYS Fluent and the resulting “Particle Number in Parcel” calculated by the software, matching our manual calculation.
Parcel Release Methods in ANSYS Fluent
When an unsteady simulation is run, you must define how the parcels are released into your domain over time. Four different Parcel Release Methods are provided for this purpose. Choosing the right one depends entirely on your specific project needs.
Figure 4: The Parcel Release Method options available in the ANSYS Fluent injection properties panel.
Here is a simple guide to each method:
| Release Method | How It Works | Best For |
| Standard | This is the simplest option. It releases a single parcel stream at each particle time step. | Basic or simple simulations where precise control over the number of parcels is not critical. |
| Constant Number | You specify the exact number of particles you want in each parcel. Fluent then automatically calculates how many parcels it needs to release to match your total mass flow rate. | Sprays and simulations with a Particle Size Distribution (PSD). This is a very common and recommended method for controlling the simulation’s statistical representation. |
| Constant Mass | You specify the total mass of each parcel. Fluent uses this to determine the number of particles in the parcel based on their diameter. | Discrete Element Method (DEM) simulations or cases where you need to ensure the parcel size does not become larger than the mesh cells. |
| Constant Diameter | You specify the diameter of the parcel itself (not the particles inside). This also helps control parcel size relative to the mesh. | Also recommended for DEM simulations and situations requiring strict control over the parcel’s spatial size. |
In summary, for most spray and general DPM simulations, the Constant Number method is a great starting point. For specialized simulations like DEM where parcel-mesh interaction is critical, Constant Mass or Constant Diameter are better choices.
Difference Between Particle Parcel and Stream
There is another confusion for users about the definition of a particle stream in DPM (see Fig. 3). The number of streams corresponds directly to the number of locations where parcels are injected. According to the discrete phase model formula, the relationship between stream mass flow rate, injection time interval, and the number of particles in a parcel is clearly defined. The number of particles in each parcel can be fractional, allowing for more precise simulations. However, this definition belongs to the unsteady (transient) simulation. For Steady DPM simulations, the concept slightly differs. Instead of using the number of particles in a parcel, ANSYS Fluent employs “strength,” which represents the number of particles in a parcel per second. This approach allows for continuous particle injection in steady-state simulations.
** It’s worth noting that Ansys Fluent’s user documentation often refers to parcels as “particles,” which can sometimes lead to confusion. Understanding these distinctions and relationships is key to effectively utilizing DPM in CFD simulations.
Figure 3: Number of particle stream in DPM Fluent
Conclusion
In this guide, we explained the parcel concept from the ground up.
You have learned:
- A parcel is a smart way to group identical particles into one unit.
- We use parcels to make complex simulations run much faster.
- How to calculate particles per parcel using a simple formula.
- How to choose the correct Parcel Release Method for your spray or DEM analysis.
The parcel model is not just a simple setting; it is a fundamental technique. By using parcels correctly, you can simulate fuel sprays, cyclone separators, and pneumatic transport without wasting computer memory.
Setting up these multiphase flows can be difficult. If you need professional assistance with your industrial projects, you can explore our DPM CFD Simulation trainings category. Also, if you want our experts to handle your setup directly, feel free to Order Your CFD Project.