Newton Manual
Newton Manual
Table of Contents
1 INTRODUCTION ............................................................................................................................................. 4
2 INSTALLATION ............................................................................................................................................... 5
2.1 LICENSING INFORMATION ..................................................................................................................................6
2.2 NEWTON UPDATES ..........................................................................................................................................9
3 OVERVIEW .................................................................................................................................................. 10
3.1 TERMINOLOGY ..............................................................................................................................................10
3.2 UNITS OF MEASURE .......................................................................................................................................10
3.3 MAIN NEWTON TOOLBAR ...............................................................................................................................11
3.4 MAIN MENUS...............................................................................................................................................12
3.4.1 File Menu .............................................................................................................................................12
3.4.2 Solver Menu .........................................................................................................................................13
3.4.3 Display Menu .......................................................................................................................................14
3.4.4 Help Menu............................................................................................................................................17
3.4.5 Library Menu ........................................................................................................................................17
3.5 USER PREFERENCES WINDOW..........................................................................................................................18
3.5.1 Rendering Preferences .........................................................................................................................18
3.5.2 System Preferences ..............................................................................................................................20
4 PRE-PROCESSING ........................................................................................................................................ 21
4.1 GENERAL SIMULATION VARIABLES ....................................................................................................................21
4.2 MATERIAL PROPERTIES ...................................................................................................................................26
4.3 IMPORTING GEOMETRY ..................................................................................................................................29
4.3.1 Layer Control ........................................................................................................................................29
4.3.2 Layer Surface Velocity Control .............................................................................................................32
4.3.3 Layer Deactivation ...............................................................................................................................33
4.3.4 Layer Movement and Cyclic Motion Control ........................................................................................33
4.3.5 Layer Friction Settings ..........................................................................................................................34
4.3.6 Layer Triangle Control ..........................................................................................................................35
4.3.7 Graphics Rendering Window ...............................................................................................................37
4.3.8 Killboxes ...............................................................................................................................................39
4.3.9 Layer Movement Profiles .....................................................................................................................40
4.4 BELT GEOMETRY ...........................................................................................................................................46
4.4.1 Creating a 3-Roll or 5-Roll Conveyor Belt .............................................................................................47
4.4.2 Creating an Apron Feeder ....................................................................................................................53
4.4.3 Creating a Bucket Elevator...................................................................................................................54
4.4.4 Creating a Chain Feeder .......................................................................................................................57
4.4.5 Creating an En Masse Feeder ..............................................................................................................59
4.5 MATERIAL GENERATION .................................................................................................................................62
4.5.1 Newton Particles ..................................................................................................................................62
4.5.2 Particle Generation ..............................................................................................................................63
4.5.3 Generating Particle Chains ..................................................................................................................74
4.6 PARTICLE SET LIBRARY ....................................................................................................................................78
4.6.1 Size Distribution Curves ........................................................................................................................79
4.6.2 Group List .............................................................................................................................................81
4.6.3 Merge/Export Libraries ........................................................................................................................83
4.7 CLUSTER CREATOR .........................................................................................................................................85
4.7.1 Main Window ......................................................................................................................................85
4.7.2 Transform Cluster Window ..................................................................................................................88
2
Newton Discrete Element Modeling Software - Manual
3
Newton Discrete Element Modeling Software - Manual
1 Introduction
Thank you for your interest in our Newton Discrete Element Modeling software. This manual will discuss
how to use the Newton software, and describes the meanings of all input and output data.
This manual assumes that the reader is familiar with the basic principles of the Discrete Element Method
(DEM). It is not meant to be a tutorial on DEM, nor is it intended to be an instruction manual on how to
design conveyor belt transfer chutes.
If you have questions about the Newton software that are not addressed in this manual, direct them to
info@actek.com and we will respond as soon as possible.
4
Newton Discrete Element Modeling Software - Manual
2 Installation
The Newton software consists of two parts. These are the software installation CD-ROM and a Sentinel
USB hardware dongle. Newton is supported under Microsoft Windows XP, Vista, and Windows 7.
To install Newton insert the CD-ROM into your computer. The following items on the Newton CD-ROM
must be installed:
5
Newton Discrete Element Modeling Software - Manual
6
Newton Discrete Element Modeling Software - Manual
Note that if you have purchased the professional version, you may run up to 3 instances of Newton on
your network using one license. Only 1 of those instances of Newton can be running a simulation, but
the other 2 instances may be performing pre and post processing, such as setting up simulations,
viewing playback files, or creating animation files. Newton automatically detects the USB license
through your computer network. If you have purchased the Basic version, only 1 instance of Newton
can be running per license.
7
Newton Discrete Element Modeling Software - Manual
To view the settings for your USB license, open a web browser on the machine that contains the license,
and type http://localhost:1947 in the URL bar. This will open the Sentinel Admin Control Center, shown
in Figure 2.1.1.
8
Newton Discrete Element Modeling Software - Manual
9
Newton Discrete Element Modeling Software - Manual
3 Overview
3.1 Terminology
This section defines and describes some important terms that are used throughout this manual.
DEM: An acronym for the Discrete Element Method, a simulation method which records and tracks the
position, orientation, force, velocity, etc. of each individual particle using a discrete time-stepping
scheme.
Sphere: A sphere is the base unit of material in Newton. All material is created using spheres.
Cluster: This is a typical unit of material in Newton. Groups of spheres are arranged to form rigid
particles called clusters. A cluster can be composed of any number of spheres.
Particle: This is a general term for a piece of material, either a sphere or a cluster.
Time-Step: Newton uses an explicit time stepping scheme. This means that at a single instant in time,
Newton calculates the position, orientation, and velocity of each particle. Then the simulation ‘steps
through’ a specific length of time (usually on the order of a few microseconds) and repeats all
calculations. The time-step is defined as this length of time.
Triangle: Newton imports geometry from CAD files in the form of triangles. Each layer in the CAD
model contains hundreds or thousands of triangles. The term ‘triangle’ refers to one of these triangles
on a specific layer of the geometry.
Runtime: This is the length of real time that is required for Newton to run a simulation. Runtimes can
vary from a few hours to a few weeks, depending primarily on the number and size of the particles. This
can also be referred to as computation time.
Neighbor: For each particle, Newton maintains a list of ‘Nearest Neighbor’ particles that could
potentially be in contact with the particle. These neighbor particles are checked for contact at every
time step. The neighbor list is periodically updated.
Cell: The simulation domain is discretized into cubic cells. The location of each sphere, cluster, and
triangle includes a reference to which cell it is in. These cells are used to track particle and triangle
movement and contact.
Rendering Window: Many of the input pages in Newton contain an area where the particles and/or
chute geometry is rendered using DirectX. These renderings can be panned, orbited, and zoomed. The
sections of the window where these renderings appear are called Rendering Windows.
10
Newton Discrete Element Modeling Software - Manual
11
Newton Discrete Element Modeling Software - Manual
12
Newton Discrete Element Modeling Software - Manual
Save Playback Frame: If a Newton playback file is open, this command will save a Newton playback
frame for the current simulation frame.
Playback Tools: This submenu contains more commands for working with playback files.
Reduce Playback File: Opens a command window to modify and reduce the size of a playback
file. (See Section 6.1.3)
Merge Playback Files: Opens a command window to merge multiple playback files together.
(See Section 6.1.1)
Export Playback File to DXF: Exports the current playback frame to a .DXF CAD file. (See Section
6.1.2)
Preferences: Opens the window to modify user preferences. (See Section 3.6)
Exit: Closes all files and exits Newton.
13
Newton Discrete Element Modeling Software - Manual
14
Newton Discrete Element Modeling Software - Manual
15
Newton Discrete Element Modeling Software - Manual
Surfaces: This submenu contains commands for setting the global view settings for the layers in the
simulation. The layers can be rendered as solid, transparent, wireframe, or outer-lines-only. This
submenu is only visible in playback mode or when a simulation is running. This also contains the
command to show normal vectors for all triangles.
16
Newton Discrete Element Modeling Software - Manual
17
Newton Discrete Element Modeling Software - Manual
18
Newton Discrete Element Modeling Software - Manual
Custom > Max Scale Color: This is identical to the previous setting, except it applies to particles
that are higher than the color scale range (i.e. if the velocity scale is set from 4.0-8.0 m/s,
consider a particle with velocity 10.2 m/s)
Show Min/Max Colors In Color Scale: If this box is checked, then Newton will render an
additional triangle on both sides of the color scale that show the colors of particles and surfaces
that are outside the scale range.
Color Triangles < Min Wear: By default, when you switch a layer to Work, the entire layer is
colored blue (as shown in the left figure below). As the wear summary progresses, the triangles
that have a lot of particle impact will be colored green or yellow or red.
However, if you uncheck this checkbox, then triangles that have wear values less than the
minimum scale setting will be colored according to the layer color (as shown in the right figure
below). If the minimum scale setting is exactly 0, then all triangles with wear values less than
0.5% of the max scale value will be ignored instead.
19
Newton Discrete Element Modeling Software - Manual
20
Newton Discrete Element Modeling Software - Manual
4 Pre-Processing
21
Newton Discrete Element Modeling Software - Manual
A critical time ratio of 0.17-0.19 should be sufficient in most cases. But if your simulation has a
significant number of stagnant or very slow moving particles, or you are recording wear on any surfaces
(see Section 6.4.5), it will help to lower this value to 0.15-0.16.
22
Newton Discrete Element Modeling Software - Manual
Maximum Allowable Particle Overlap: DEM modeling operates by allowing spheres to slightly overlap
each other in the simulation. This region specifies how much sphere overlap will be allowed. AC-Tek
recommends leaving this at 5.0 %.
Maximum Expected Velocity: Newton uses this value when calculating the size of the variable time
step. You should input an estimate of the maximum velocity that you expect the particles to achieve as
they fall through the transfer chute. If your simulation has particles moving at 5 m/s maximum, but you
set this value at 10 m/s, then Newton will calculate time steps that are smaller than necessary and your
simulation will take much longer to run. If you are not sure what the maximum velocity will be, you can
run a short simulation using the default value of 5 m/s with just a handful of particles to observe the
max velocity and then reset the simulation using the observed value. It is sufficient to set this value to
within +/- 1.0 m/s of the true maximum velocity for the simulation.
Kill Particle Velocity: This specifies the maximum velocity any particle can achieve before Newton
removes it from the simulation. Occasionally a particle may become pinched between two moving
surfaces in the simulation. When the particle manages to wiggle free, it may escape with a very high,
unrealistic velocity. Newton will ‘kill’ any particle in the simulation with a velocity above this criterion
before it collides with other particles and disturbs the flow in an unrealistic way.
Sphere Size Ratio: At every time step, the simulation checks to see which particles are in contact with
which other particles. It is obvious that a particle at the top of a transfer chute is not going to be in
contact with a particle at the bottom of the transfer chute. So ideally, the simulation should not even
bother to check if those two particles are in contact. The most efficient way to reduce the number of
checks the simulation must perform is to divide up the simulation domain into cubic cells. A particle
that is inside one of these cells is only compared to other particles that are located inside that cell, or
one of the 26 adjacent cells. The sphere size ratio specifies how large to make the cells. A value of 1.1
means that the X, Y, and Z lengths of the cell size will be set 10% larger than the size of the largest
sphere in the simulation.
If this value is increased, the required memory to run the simulation will decrease, but runtime will
increase. This value should only be increased if you encounter system ‘out of memory’ exceptions. In
most cases, 3 to 4 GB of system RAM is sufficient to run Newton. AC-Tek recommends using a value of
1.10.
Directional Collision Ratio: While a simulation is running, Newton maintains a list of ‘Nearest Neighbor’
particles that are in close proximity to each particle in the simulation. At every time step, each particle
is checked for contact against its nearest neighbors. This setting inversely scales how often these
nearest neighbor lists are updated. So Newton will update these lists twice as often if this value is set to
2.0, compared to a value of 1.0.
Consider Case A: A particle is moving toward a stationary particle with some maximum velocity V. After
one time step, the distance between the particles has decreased by X.
Consider Case B: Two particles are moving directly toward each other, each with the same maximum
velocity V as in Case A. After one time step, the distance between those two particles has decreased by
2*X, twice the distance as in Case A.
Case A represents a situation where the Directional Collision Ratio should be 1.0, because one particle is
stationary and the other is moving with maximum velocity.
Case B represents a situation where the Directional Collision Ratio should be 2.0, because the distance
between the two particles is decreasing twice as fast as in Case A. Therefore the list of nearest
neighbors should be updated twice as often to maintain simulation stability.
23
Newton Discrete Element Modeling Software - Manual
If complete simulation stability is desired, then the value of the directional collision ratio should be set
to 2.0, because this will ensure that any situation similar to Case B will always be stable. But generally,
when modeling conveyor belt transfer chutes, all material flows in approximately the same direction, so
it is very rare that two particles would travel directly toward each other at maximum velocity. Therefore
this value can usually be set in the range 1.25-1.75 and the simulation will remain stable. AC-Tek
recommends using a value between 1.25-1.50 for most cases, and using values of 1.5-2.0 only if there
are locations in the chute geometry where the particles could travel toward each other at maximum
velocity.
Moving Cell Triangle Update: For each triangle in each layer of the chute geometry, Newton maintains
a list of cells that each triangle of each layer occupies. You may choose to run a simulation that uses
moving layers, such as a rotating hood, or a vibrating screen. A layer that has a surface velocity (like a
belt) is NOT considered a moving layer, because the triangles that compose the layer do not physically
move in the simulation, they are simply denoted as having the desired surface velocity.
If your simulation does use moving layers, like translation, rotation, or linear or rotational cyclic motion,
then Newton will periodically update the cells that the moving triangles occupy. This value specifies the
distance, in cell lengths, that any point on the triangle can move before Newton updates the list of cells
it occupies. This value cannot be set above 0.95, or particles may push through surfaces, resulting in a
nonsensical simulation. AC-Tek recommends leaving this at the default value of 0.50.
Preferred Thread Search Direction: If you are running a simulation using multiple processing threads,
then this setting specifies how to divide the particles between the threads. If the Preferred Thread
Search Direction is set to AUTO, then Newton will divide the particles along the axis which will result in
the fewest number of particles in the cleanup thread. Alternatively, you can choose the axis along which
you would like to divide the particles for processing.
Playback File Save Time Interval: This setting specifies how often to save simulation data for the
playback file and CSV file. With the default setting of 0.020 seconds, Newton will save 50 frames of data
for each second of simulation. Lowering this value will result in a higher playback frame rate, but also a
much larger data file.
PNG File Save Time: At the specified interval, Newton will capture a screenshot of the running
simulation. These snapshots provide an easy way to view the history of the simulation without having
to open the uncompressed data file while the simulation is running. The snapshot is taken based on the
current view settings, so be sure to position the geometry properly at the start of the simulation. If this
interval is set to zero or left blank, then Newton will only capture one screenshot at the start of the
simulation.
Save Triangle Force & Work Data: Newton is capable of showing the force and work applied by the
particles to the chute geometry. If this setting is set to YES, then Newton will record this force and
work. Note that Newton will not display force and work information while the simulation is running; the
data is used when making animation files during post-processing.
It is difficult to calculate the real force and work that will be applied to the chute during actual loading,
but Newton can provide a rough estimate of these values. Primarily the force and work data is used to
qualitatively compare one chute design to another to see which will result in higher force and work on
the chute.
Note that when analyzing triangle wear, Newton does this for an entire layer. Therefore to get the most
detailed work information about a wear plate, for example, put the wear plate on its own layer. If this
24
Newton Discrete Element Modeling Software - Manual
information is not required, then this setting should be set to NO, because recording this information
will result in a longer runtime and a larger file size.
Save Particle Force & Work Data: Newton is also capable of showing the force and work on the
particles themselves. If this setting is set to YES, Newton will record the force and work. As with Save
Triangle Force & Work Data, this information will not display while the simulation is running, and is
primarily used for qualitative comparison of one chute design to another.
If this information is not required, then this setting should be set to NO, because recording this
information will result in a longer runtime and a larger file size.
Use Restart File: Newton can periodically save the current simulation information to a Newton restart
file (.RST). In the event of a power failure, this enables you to restart your simulation close to where it
was, rather than starting over from the beginning, saving hours or days of computation time. AC-Tek
recommends always setting this to YES.
Restart Save Time Interval: This is where you specify how often to save a restart file. Note that this
value corresponds to seconds of simulation time, not seconds of computation time. If your simulation
runs very slow, taking hours of computation time to complete one second of simulation time, then you
may want to set this value to 0.5 seconds. Power failures are rare, and using a value of 2.0 seconds is
generally sufficient.
Overwrite Restart File: Newton can either overwrite the same restart file each time, or save a new
restart file at each restart save time interval. It is unnecessary to save multiple restart files, because you
essentially always want to restart the simulation using the most recent restart file.
Hard Drive Letter: This specifies the hard drive in your computer on which Newton will save the
simulation data.
Base Directory: This specifies the root directory on the specified hard drive to which Newton will save
the simulation data. To easily locate all your simulations, it is a good idea to use the same base directory
for all DEM simulations.
Client Name, Job ID, Chute ID, Revision Number, Friction Level: These five inputs specify the
subsequent folders in the Base Directory to which Newton will save the simulation data. There is no
reason that the folder designated as Client Name has to represent the name of the client. The same is
true for each of the other four folders. You need only remember that the folder tree for these folders
starts with the Base Directory, and each subsequent folder will be placed inside the previous folder (so
the Friction Level folder will contain no other folders, only the simulation data)
File Name: This specifies the base filename for each file that the simulation creates.
Comments: These are for your own notes about the simulation. These comments are not reprinted
anywhere, they appear only in the Comments box.
Save Default File Location: Clicking this button will save the current file location information in the user
preferences file as the default information to use.
25
Newton Discrete Element Modeling Software - Manual
26
Newton Discrete Element Modeling Software - Manual
Particle-Boundary Friction Coefficient: This specifies the coefficient of friction for particle-to-surface
contact.
Coefficient of Restitution: The coefficient of restitution describes the ratio of the velocity of a particle
after it collides with a rigid wall to the velocity of the particle before it collides with the wall. A
coefficient of 1.0 would imply that a particle bouncing off a rigid surface would lose no energy in the
collision, and a coefficient of 0.0 would imply that the particle does not bounce of the wall; it stops,
losing all its kinetic energy in the collision. Because of numerical concerns, the minimum value for this
input is 0.075. AC-Tek recommends using a coefficient of restitution between 0.10 and 0.30.
Rotational Damping: If a spherical particle is rolling along a surface, this value specifies how far it will
roll (in revolutions about its center of gravity) before it comes to rest. This is particularly useful for
forcing spherical particles, which tend to roll for a long time, to stop relatively quickly. However, in AC-
Tek’s opinion, spherical particles are of very little value in a DEM simulation because of their inability to
sustain a rotational moment.
The majority of particles generated in Newton are usually irregular clusters. The effect of rotational
damping on irregular clusters is very minor, and you can usually just leave this factor at its default value.
Because of numerical concerns, the minimum value for this coefficient is 0.10.
Use Ratchet Effect: This specifies whether to use the ratchet effect. The ratchet effect helps simulate
cohesiveness in the material. Normally, when two particles come into contact and overlap, Newton
creates a virtual spring between the two particles, forcing them apart. This force is directly proportional
to how far the particles overlap. With the ratchet effect, Newton will add a second spring between
these particles. This spring acts to pull the particles together, helping them stick to each other. The
magnitude of the ratchet spring force is scaled by the cohesion factor inputs. A factor of 1.0 would
create a cohesive force equal to the force that pushed the particles together. A factor of 0.25 would
create a cohesive force equal to 25% of the force that pushed the particles together.
Particle-Particle Cohesion Factor: The ratchet cohesion factor for inter-particle contact. For contact
between particles with different material properties, the average of the two factors is used. A blank
input represents a cohesion factor of 0. But if a particle with 0 cohesion factor contacts a particle with a
non-zero cohesion factor, the ratchet effect will still be applied, using the average of the two factors (i.e.
half of the non-zero cohesion factor).
Particle-Boundary Cohesion Factor: The ratchet cohesion factor for particle-boundary contact.
Use Liquid Bridge: This setting specifies whether to use liquid bridge effects. A liquid bridge is a
physical bridge of moisture that forms between two particles with high water content. When two
particles come very close, the bridge forms between them. The bridge applies a small force, pulling the
particles toward each other. After the particles move apart and become separated by a specific
distance, the bridge collapses, eliminating the pulling force. Figure 4.2.2 shows a liquid bridge between
two particles of different sizes.
27
Newton Discrete Element Modeling Software - Manual
28
Newton Discrete Element Modeling Software - Manual
29
Newton Discrete Element Modeling Software - Manual
If you right-click the combo box, three additional options are displayed:
Rename Layer: Renames the current layer.
Delete Layer: Deletes the current layer from the geometry.
Delete All Layers: Deletes all imported layers from the geometry.
Import CAD: This button allows you to import your chute geometry. After clicking this button, Newton
will prompt you to select a CAD file. Then it will bring up the input window shown in Figure 4.3.3. A
preview of the geometry will be shown in the rendering window. Both the existing geometry and the
new CAD file will be displayed. The existing geometry is colored gray, and the new geometry is colored
blue.
Newton supports the following CAD file types:
- STEP/STP (version AP203 and AP214)
- IGES/IGS (version 5.3)
- STL (Binary and ASCII formats)
- DXF (only reads surfaces that were drawn or converted to 3DFaces)
30
Newton Discrete Element Modeling Software - Manual
in inches, enter 25.4 in the Scale Factor text box, and Newton will scale the entire model by 25.4 about
the origin, converting your model to millimeters.
For STEP and IGES file types, Newton may be able to determine the units of the source CAD file and
convert to millimeters, so you might not need to enter a scaling factor.
Positive Vertical Axis: In Newton, the +Z axis is the upward vertical direction (i.e. gravity acts in the -Z
direction). If you are using a CAD program that has a different vertical axis, you can quickly change that
using the combo box. For example, if you are working in SolidWorks, you will probably want to select
the +Y axis.
Recalculate Normal Vectors: If this checkbox is checked, then Newton will recalculate the magnitude
and direction of the normal vector for every triangle. It is imperative that all the normal vectors are
correct when running a DEM simulation.
Almost all STEP and IGES files have proper normal vectors included in the file information, so when you
open a STEP or IGES file, this checkbox is unchecked by default. However, for STL and DXF files, the
normal vectors are often omitted or incorrect, so when opening an STL or DXF file, the checkbox is
checked by default.
Minimum Triangle Area: If your model contains a large number of very small, insignificant triangles,
Newton can ignore these triangles when importing the geometry (Simulations with thousands of
triangles have longer computation times than those with fewer triangles). Enter a value in the
Minimum Triangle Area (in mm2), and Newton will ignore any triangles with area smaller than this. The
minimum value is 0.001 mm2.
STEP/IGES Mesh Resolution: If you are importing a STEP or IGES file, this combo box will allow you to
select the triangulation resolution of curved surfaces like circles, arcs, and cylinders. Note that it is not
always necessary to have a very high resolution. Higher resolutions will triangulate circles and arcs with
better resolution by using more triangles, but will also increase memory usage and simulation run time.
Note: ALWAYS be sure to remove ALL unnecessary circular objects from the CAD file before import,
especially (pulleys, bolts, bolt holes, access door handles, motors, and any other non-essential
geometry).
Low: Use Low quality if your mesh contains no circles or arcs.
Med: Medium quality is sufficient for meshes with a few curved surfaces.
High: High quality can be used if the model contains several important curved surfaces
Fine: Fine quality should be used if the model contains very important continuous curved
surfaces (i.e. a smooth-curved hood or spoon).
Ultra: Ultra quality will ensure that a continuous curved surface is resolved to very fine detail.
Depending on the model, this may create a significant number of triangles and could impact
simulation speed.
31
Newton Discrete Element Modeling Software - Manual
Layer Name From File: This is the name of the object that was obtained from the CAD file.
Add to Layer: Specifies which layer the triangles from this object will be added to. For each unique
object name in the CAD file, Newton will automatically create a new layer with that name. This will
continue until the maximum import limit has been exceeded (5 layers for Newton Basic, 30 layers for
Newton Professional). Once the limit has been exceeded, all additional objects are placed on the final
layer (this can be changed using the combo box).
Overwrite: If checked, any existing geometry with the same layer name will be deleted and replaced
with the new geometry. If unchecked, then the new geometry will be added to the existing layer.
32
Newton Discrete Element Modeling Software - Manual
Stop Time: The time at which the surface velocity of the layer will stop. If a Deceleration Time has been
input, then the Stop Time represents the time at which the deceleration will begin. Therefore the layer
will come to a complete stop at Stop Time + Deceleration Time.
Deceleration Time: The length of time over which the surface velocity will decelerate from full velocity
to zero. Deceleration of the layer will start as soon as Stop Time has been reached.
Restart Time: The time at which the layer surface velocity will be restarted. If a Restart Acceleration
Time has been input, then the restart time is the time at which the layer acceleration will begin.
Restart Acceleration Time: The length of time over which the layer surface velocity accelerates from
zero to full speed. The restart acceleration period starts as soon as Restart Time is reached, therefore
the belt will reach full velocity at time Restart Time + Restart Acceleration Time.
33
Newton Discrete Element Modeling Software - Manual
i.e. (0, 0, 1), then when observing the geometry from above, the layer will rotate counter-clockwise if
the rotation rate is positive, and clockwise if the rotation rate is negative.
If you have input a custom CSV layer movement profile (see Section 4.3.9), and this layer includes
rotational motion, but the rotation axis is identical for all rotational movements, then you can input that
vector here and omit the vector in the CSV file.
Cycle Time: If cyclic motion is desired, then the time for one complete cycle is entered here. If using
linear cyclic motion, the cycle time is defined as the time for one complete back-and-forth motion. If
using rotational cyclic motion, the cycle time is defined as the time for one complete circle (or ellipse).
Note: Cyclic motion will override both linear translational motion and rotational motion. If this input
box is not blank or zero, Newton will assume there is cyclic motion and will ignore Linear Velocity and
Rotation Rate.
Note: Rotational cyclic motion will override linear cyclic motion; it is not possible to use both types of
cyclic motion simultaneously.
Cyclic Displacement: This creates linear cyclic motion. Enter the X, Y, and Z components of the cyclic
displacement. The layer will displace from its initial position by the values specified in these input boxes
and then back to its initial position over the cycle time.
If Moving Time is not a multiple of Cycle Time, then on the last cycle, the layer will stop moving mid-
cycle and freeze position for the rest of the simulation.
Cyclic Radius: This creates rotational cyclic motion, and is used primarily for simulating a vibrating
screen. The layer will maintain constant orientation, but the entire layer will orbit about its initial
position. The shape of the orbit can be circular or elliptic.
To create circular cyclic motion in the X-Y plane, specify the radius of orbit in both the X and Y input
boxes. To create elliptic cyclic motion in the X-Y plane with the semi-major axis in the Y direction,
specify the length of the semi-major axis in the Y input box and the length of the semi-minor axis in the
X input box.
To create circular or elliptic motion in the X-Z, or Y-Z planes, follow the same procedure described
above.
Note: If only 1 radius is entered in any of the three input box, the rotational cyclic motion has been
under-defined. In this case, Newton will assume circular orbit in the X-Y plane with the given radius
(even if the radius is entered in the Z input box)
Note: If radii are entered into all three input boxes, the rotational cyclic motion has been over-defined.
In this case, Newton will use only the X and Y input values to create elliptic motion in the X-Y plane. The
Z input value will be ignored.
34
Newton Discrete Element Modeling Software - Manual
Surface Liquid Bridge Multiplier: This input allows you scale the Liquid Bridge effects for each individual
layer. This increases or decreases the effects of the liquid bridge for particle-to-surface contact.
35
Newton Discrete Element Modeling Software - Manual
Note that although smaller triangles will give better resolution, it is unnecessary to mesh the layer into
triangles with average side lengths less than 50 mm, because the belt wear resolution will be sufficient
at this size (the default setting for this button is 100 mm). Also keep in mind that as you increase the
number of triangles, computation time for the simulation will increase slightly as well, especially if those
triangles are on moving layers (See above).
Consider the belt shown in Figure 4.3.6, which is composed of three faces (six triangles).
Figure 4.3.7 – The belt after its triangles have been equalized
Having equalized the triangles, we can now click the Divide Triangles button three or four times, until
we see that the triangles stop dividing. We’ll leave the smallest size of the triangles at the default value
of 100 mm. The resulting meshed belt is shown in Figure 4.3.8. This belt will record wear much more
accurately than would the unadjusted belt in Figure 4.3.6.
36
Newton Discrete Element Modeling Software - Manual
After meshing up the belt, right-click in the rendering window and click Triangle Normal Vectors. This
will render the normal vector for each triangle at the center of gravity the triangle. If you are planning
to view the Parallel or Perpendicular triangle forces (see Section 6.5.3), then you must have all the
normal vectors pointing inward and upward, toward the center of gravity of the belt, as shown in Figure
4.3.9.
37
Newton Discrete Element Modeling Software - Manual
38
Newton Discrete Element Modeling Software - Manual
Show UCS Icon: This enables or disables rendering of the UCS icon, which is rendered at the origin of
the model.
Surfaces – Solid: Renders the model using solid surfaces.
Surfaces – Wireframe: Renders the model using wireframe outlines.
Surfaces – Transparent: Renders the model using transparent surfaces.
Triangle Normal Vectors: Renders a vector at the center of gravity of each triangle that shows the
normal vector of the triangle.
Geometry: This opens a submenu that contains commands for showing or hiding specific geometry
settings.
Triangle Velocity Vectors: This will render a small vector at the center of gravity of each triangle,
showing the triangle velocity. This is useful for when setting up layer velocities to visually confirm
that the vector is correct.
Triangle Normal Vectors: This will render a small magenta vector at the center of gravity of each
triangle, showing the triangle normal vector.
Killboxes: This will show the Killboxes window. If the killbox window is open (see Section 4.3.8),
then the current selected killbox is shaded in magenta. The other killboxes are shaded in gray.
Material Generation Planes: When working in the Material Generator window, this option will
show or hide the drop location and drop shape of each generation point (see Section 4.5.2).
4.3.8 Killboxes
Newton allows you to draw killboxes into the simulation. When a particle in the simulation enters the
bounds of a killbox, Newton will remove it from the simulation permanently. Killboxes can be used to
keep a simulation looking clean. If there are a few scattered particles that fall off the conveyor belt (so
few that they are not a cause for concern and can be ignored), then you could draw a killbox on both
sides of the belt to remove any particles that fall off.
When viewing the model on the Geometry page before running a simulation, killboxes are rendered into
the model and are shown as gray transparent rectangular prisms. If the Killboxes window is open, the
current killbox that is being edited is shown in magenta to make it easy to tell which killbox is being
edited.
There are two ways to implement killboxes in Newton:
1. Input the killbox in Newton in the Killboxes window, shown in Figure 4.3.11. Specify a minimum
and maximum X, Y, and Z value to create a rectangular prism. You may also enter an activation
time and deactivation time for the killbox. If these are left blank, then the killbox will be active
for the entire simulation.
2. Draw the killbox into the geometry of the CAD file, and it must be on a layer named, KILL1, KILL2,
KILL3, up to KILL10 for each successive killbox. Newton recognizes the layer name and creates a
killbox from the layer. (This is an old method that was used before the Killboxes window was
created. AC-Tek recommends setting up killboxes via the Killboxes window instead).
Killboxes must be rectangular prisms. Newton looks at the minimum and maximum boundary in
the X, Y, and Z directions, so complex shapes cannot be killboxes. You can, however, use
39
Newton Discrete Element Modeling Software - Manual
multiple rectangular boxes, each on a different kill layer, to create a killbox with a complex
shape.
Note that if you import a killbox from the geometry, Newton will overwrite the existing killbox
data for that killbox. For example, importing a layer named ‘KILL8’ will overwrite the data for
Killbox 8 in the Killboxes window.
40
Newton Discrete Element Modeling Software - Manual
41
Newton Discrete Element Modeling Software - Manual
Rotation Vector boxes in Row 8. Newton will assume that Movements 3-7 will use vector (1, 0,
0) and Movements 8-12 would use vector (3, 1, 1). In fact, any rotations after Move 12 would
also use vector (3, 1, 1) unless a different vector is specified.
Note: You must specify the rotation vector on or before the first Movement where it is used. In
the example above, you could enter vector (1, 0, 0) into Row 1 or Row 2 and the LMP would be
unchanged.
Rotation Point: Recall from Section 4.3.4 that when specifying layer rotation, you must input a rotation
rate, a rotation vector, and a rotation point to which the rotation vector is attached. This applies to
layer movement profiles as well. For custom layer movement profiles, you must specify the rotation
point under Rotation Point on the Geometry input window. If this is left blank, Newton will assume the
rotation point is (0, 0, 0).
If the LMP translates the layer during the simulation, the rotation point is moved along with the layer, so
the point of rotation will remain fixed relative to the layer for the entire simulation.
42
Newton Discrete Element Modeling Software - Manual
Figure 4.3.13 – Bucket simulation using a custom movement profile (Time = 2.99 sec)
Move 1: At 3 seconds, the bucket moves in the Y-direction at 1.4 m/s, with no rotation. The bucket
begins moving along the length of the box to scoop up particles.
Figure 4.3.14 – Move 1: Bucket moves along the box, filling with material (T = 4.99 sec)
43
Newton Discrete Element Modeling Software - Manual
Move 2: At 5 seconds, there is no linear motion, and the bucket begins to rotate about the vector (1, 0,
0) at +20 degrees per second. The bucket rotates upward to help contain the material for when the
bucket is lifted out of the box. Note that the rotation follows the right-hand-rule.
The rotation point was specified as (0, 0, 0) at the start of the simulation. But during Move 2, the bucket
translated 2.8 meters along the Y-axis, so the rotation point was moved to (0, 2.8, 0), such that it
remains at the center of gravity of the bucket.
Figure 4.3.15 – Move 2: Bucket rotates upward at 20 degrees per second. (T = 5.99 sec)
Move 3: At 6 seconds, the bucket moves along the Z-axes at 1 m/s with no rotation. The bucket lifts
upward, out of the box.
44
Newton Discrete Element Modeling Software - Manual
Move 4: At 7 seconds, the bucket moves along the Y-axis at 0.7 m/s with no rotation. The bucket
moves forward, away from the box, in preparation to dump its material.
Figure 4.3.17 – Move 4: Bucket moves forward at 0.7 m/s. (T = 8.99 sec)
Move 5: At 9 seconds, the bucket has no linear movement and rotates at -45 degrees/sec about the
vector (1, 0, 0). The bucket rotates forward, dumping its material outside the box.
Note that during Moves 4 and 5, the bucket moved in the Z-direction by 1 meter, and in the Y-direction
by 1.4 meters. Therefore the rotation point of the bucket layer was moved from its position after Move
2 of (0, 2.8, 0) to its final position of (0, 4.2, 1), so that the bucket still rotates about its own center of
gravity.
Figure 4.3.18 – Move 5: Bucket rotates forward at 45 degrees to dump its material. (Time = 9.99 sec)
Move 6: The bucket has no linear movement and no rotation. Note that this final movement is
important. Because the CSV file does not include a ‘Move End Time’ it assumes that each movement
proceeds until the next movement section. Therefore when you want all motion for the layer to cease,
you have to end with a move section that specifies no movement (unless of course you never want the
layer to stop moving during the simulation).
Also note that the layer can have periods of time where it does not move. Simply insert movement
sections in the CSV file that have zero linear motion and zero rotation and the layer will stop moving at
the specified time, and will start moving at the next movement section that contains movement
instructions.
45
Newton Discrete Element Modeling Software - Manual
Figure 4.4.1 – Belt Geometry input page with 5-Roll Belt configuration selected
46
Newton Discrete Element Modeling Software - Manual
47
Newton Discrete Element Modeling Software - Manual
Optional Input
Dropbox (0-Yes, 1-No, 0.5-Split): A drop box is the primary means of delivering material to the feed belt
in the simulation. Material is created in layers, and these layers are dropped one at a time into the drop
box. Figure 4.4.2 shows a typical drop box.
48
Newton Discrete Element Modeling Software - Manual
Figure 4.4.3 – Split drop box with fine material in back and coarse lumps in front
Splitting the drop box is usually done when you want to drop two different material sets in the same
drop box. For instance, if you wanted to drop fine material in the back of the drop box, and then cover
that a layer of lumps, as in Figure 4.4.3, you can do this by splitting the drop box into two sections, and
then generating a different material set for each section.
Dropbox Length: The length of the drop box, in m. If left blank, Newton will use a default length of [2.0
* Total Width].
Dropbox Width: The width of the drop box, in m. If left blank, Newton will use a default width of [1.15
* Total Width].
Skirtboard Length: The length of the skirtboard attached to the drop box. If left blank, Newton will use
a default length of [1.0 * Total Width].
Tail Pulley Diameter: The diameter of the tail pulley, in m. If left blank, Newton will not create a tail
pulley.
Sub Pulley Diameter: The diameter of the snub pulley, in mm. If left blank, Newton will not create a
snub pulley. Note that the snub pulley is created only for appearance, and does not actually serve a
purpose in the simulation.
Snub Pulley Y-Position: The axial distance from the head pulley to the snub pulley, in m.
Snub Pulley Z Position: The vertical distance from the head pulley to the snub pulley, in m.
Snub Pulley Extension Length: The length of belt that is created behind the snub pulley, in m.
Head Pulley Cycle Time: The belt can be given cyclic motion if desired. The cyclic motion is linear, and
points in the direction of the belt axis. The head pulley cycle time is the time for one complete cycle (i.e.
one full back-and-forth motion).
Head Pulley Cycle Distance: The displacement along the belt that the head pulley will cycle. If the input
distance is positive, then Newton will move the head pulley forward to start the cycle. If the input
distance is negative, then Newton will move the head pulley backward to start the cycle.
Triangle Control
Coefficient of Friction – Belt: A default coefficient of friction was specified for particle-surface contact
on the Material Properties page. This input overrides that coefficient for this belt.
49
Newton Discrete Element Modeling Software - Manual
Coefficient of Friction – Dropbox: A default coefficient of friction was specified for particle-surface
contact on the Material Properties page. This input overrides that coefficient for this drop box and drop
box skirtboard.
Liquid Bridge Multiplier – Belt: A default Boundary Surface Tension Multiplier for the liquid bridge
effect was specified on the Material Properties page. This input overrides that multiplier for this belt.
Liquid Bridge Multiplier – Dropbox: A default Boundary Surface Tension Multiplier for the liquid bridge
effect was specified on the Material Properties page. This input overrides that multiplier for this drop
box and drop box skirtboard.
Layer Color – Belt: Double-click this cell to set the layer color for the belt.
Layer Color – Dropbox: Double-click this cell to set the layer color for the drop box.
Max Triangle Size: This allows you to specify the maximum triangle size for the belt. Enter the
maximum average side length in millimeters. Newton will first try to square up the belt triangles and
match the size to the triangles in the transition section. Next, Newton will subdivide large triangles into
smaller triangles until the average side length of each triangle is less than the size specified here.
Note that in the process of squaring up the triangles and matching them to the size of the transition, the
size might already be smaller than the value you specify here. For example, for a typical 1200 mm belt
with the default transition length of 2.4 m, the average side length of the triangles in the transition
section is about 105 mm. So even if you enter a larger value in the Max Triangle Size cell (say, 200 mm
or 300 mm), the trough section triangles will automatically be matched to the triangles in the transition
section, so all triangles will be 105 mm or smaller.
It is recommended that the minimum triangle size be no smaller than the smallest particle in the
simulation. Because of the nature of the triangle-particle contact in a DEM simulation, it is pointless to
make the triangles any smaller than this.
Figure 4.4.4 shows a top view of a belt with no maximum triangle size, and a second view of the same
belt with maximum triangle size of 105 mm.
Figure 4.4.4 – Top: Normal belt; Bottom: Belt with 100 mm max triangle size
# Triangles in Belt: This is a reference value that specifies the number of triangles that make up the
trough and transition sections of the belt. The triangles that make up the head pulley, tail pulley, snub
pulley, and drop box are not included. A standard pulley contains 72 triangles, and a standard drop box
contains 24 triangles.
50
Newton Discrete Element Modeling Software - Manual
This input helps when meshing up the belt using the Max Triangle Size input. When the number of
triangles that makes up the belt exceeds 5,000-10,000, the speed of the simulation will start to slow
down. A belt with 10,000 to 15,000 triangles is not unusual, but it is rarely necessary to have a belt with
20,000 triangles or more.
Figure 4.4.5 – 3-roll (left) and 5-roll (right) conveyor belts, front view
After you specify a belt speed, you can turn on the velocity vector for each triangle of the belt by
enabled the Velocity Vectors option in the right-click context menu (see Section 4.3.7).
51
Newton Discrete Element Modeling Software - Manual
52
Newton Discrete Element Modeling Software - Manual
Triangle Control
The inputs for the triangle control are analogous to those for the Belt Triangle Control in Section 4.4.1.
53
Newton Discrete Element Modeling Software - Manual
54
Newton Discrete Element Modeling Software - Manual
Triangle Control
The inputs for the triangle control are analogous to those for the Belt Triangle Control in Section 4.4.1.
55
Newton Discrete Element Modeling Software - Manual
You can specify a non-zero position for the bucket elevator by setting the X, Y, and Z positions in the
Required Head Pulley Input section. You can also specify a rotation angle to rotate the bucket elevator
about the vertical axis, or an incline angle to tilt the bucket elevator system.
Note: The bucket that is used must be meshed up into small triangles (50-200 mm, depending on the
overall size of the bucket). Consider Figure 4.4.10, which shows two identical bucket elevators. The left
elevator is using a bucket composed of 10 triangles, while the right elevator is using a bucket composed
of 1,100 triangles. Rendered at the C.G. of each triangle is a velocity vector, showing the magnitude and
direction of the velocity of the triangle.
Each triangle can only have one velocity. As shown by the elevator on the right, the velocity of the
bucket varies linearly based on the distance from the center of rotation. The blue vectors closer to the
head pulley are moving slower than the green/yellow vectors near the outside edge of the bucket. The
elevator on the left will not provide an accurate bucket elevator simulation because it cannot reproduce
this velocity gradient, while the elevator on the right can.
We recommended that your bucket contain between 200-1500 triangles (depending on the size and
shape of the bucket). The more triangles in the bucket, the longer the simulation time, but the more
accurate the velocity gradient across the surface of the bucket. Only small portion of the simulation
time is used to calculate the bucket elevator motion, but we still recommend a maximum of 2,000
triangles to comprise the bucket.
Figure 4.4.10 – An image showing the distribution of velocity vectors in a bucket elevator
56
Newton Discrete Element Modeling Software - Manual
57
Newton Discrete Element Modeling Software - Manual
Once this is completed, Newton will create a bucket elevator like the one shown in Figure 4.4.11.
58
Newton Discrete Element Modeling Software - Manual
59
Newton Discrete Element Modeling Software - Manual
Triangle Control
The inputs for the triangle control are analogous to those for the Belt Triangle Control in Section 4.4.1.
60
Newton Discrete Element Modeling Software - Manual
Figure 4.4.12 – An En Masse generated in Newton; a custom feed hopper has been imported
Note: The flight geometry that is imported must be meshed up into small triangles (50-200mm,
depending on the overall size of the flight). This is to ensure that the velocity gradient across the surface
of the bar will be properly resolved.
See the note about the Bucket Elevators in Section 4.4.3, as well as Figure 4.4.10. The same principle
applies here. The geometry of a flight is usually much smaller and simpler than a bucket, so we
recommend that the flight contain only 40-200 triangles, although this depends on this size and
complexity of the flight that is used.
61
Newton Discrete Element Modeling Software - Manual
62
Newton Discrete Element Modeling Software - Manual
63
Newton Discrete Element Modeling Software - Manual
64
Newton Discrete Element Modeling Software - Manual
Clusters: This method generates material by cluster count. If material is generated by cluster
count, then the tonnage must be specified as well, because the material is still dropped in layers,
according to the specified tonnage.
Particles: This method generates material by particle count (Recall that a particle is a single
sphere, and several particles make up one cluster). If material is generated by particle count,
then the tonnage must be specified as well, because the material is still dropped in layers,
according to the specified tonnage.
Generation Start Time: Specifies when the material will be released into the simulation. If the
generation method is Layer, Clusters, or Particles, then this time specifies when the first layer of
particles will be dropped. If the generation method is Block, then this time specifies when the block of
material will be released.
Acceleration Period: If the generation method is set to Layer, Newton can linearly accelerate the
tonnage from zero to full tonnage. This input is the length of time over which the tonnage is
accelerated. The acceleration period begins as soon as Release Time is reached.
Deceleration Period: If the generation method is set to Layer, Newton can linearly decelerate the
tonnage from full tonnage to zero. This input is the length of time over which the tonnage is
decelerated. The deceleration period is timed such that the tonnage reaches zero exactly as the
material generation stops.
Release Velocity: The initial velocity of each particle is specified here. If these boxes are left blank, the
initial velocity of each particle will be zero. The X, Y, and Z input boxes do NOT represent the X, Y, and Z
components of the velocity. Instead:
X = Initial particle velocity in the transverse (across the belt) direction
Y = Initial particle velocity in the axial (along the belt) direction
Z = Initial particle velocity in the vertical (normal to the belt) direction
The velocities are specified in this manner because when the belt is rotated or inclined, these velocities
are rotated and inclined to match, so that the user does not have to manually calculate the X, Y, and Z
components.
Random Min Velocity: Each particle can be given a random initial velocity. The X, Y, and Z input boxes
specify the transverse, axial, and normal components of the minimum random velocity in meters per
second. (See Release Velocity above)
Random Max Velocity: The X, Y, and Z input boxes specify the transverse, axial, and normal
components of the maximum random velocity in meters per second. (See Release Velocity above)
Random Rotation: Each particle can also be given an initial random rate of rotation. The rate of
rotation of each particle will be some random number between 0 degrees per second and the specified
value.
Custom or Preset (Drop Location): This specifies where the material will be generated. If this is set to
Custom, then you must specify the center point where the material will be generated as well as the
Drop Shape (see next). If this is set to Belt 1, Belt 2, or Belt 3, you need only select the Drop Shape.
Drop Shape: This specifies the shape of the material drop. If the drop location is Belt 1, Belt 2, or Belt
3, then the center position and size do not need to be specified. If the drop location is Custom, then you
need to specify the drop center point for the material. The size of the drop must also be input,
65
Newton Discrete Element Modeling Software - Manual
depending on the drop shape. Figure 4.5.7 shows the input boxes for each drop shape, as well as an
example of each drop shape.
66
Newton Discrete Element Modeling Software - Manual
Figure 4.5.9 – Using too small a minimum particle gap can result in uneven belt loading
When the Minimum Particle Gap is increased from 3 mm to 75 mm, the material flow is much more
steady and realistic, as shown in Figure 4.5.10.
67
Newton Discrete Element Modeling Software - Manual
liquid bridge will immediately form between the edge clusters and the dropbox surface, disturbing these
clusters as they drop. The Bounding Box Gap should always be at least 5 mm if using liquid bridge
cohesion.
Pattern Distribution: Newton usually generates material to be distributed evenly throughout each layer
(like in Figure 4.5.8). Alternatively, Newton can generate material such that the density of clusters in the
layer varies in the X and/or Y direction.
Linear Gap Factor: Specifies the maximum distance between particles at the sparse side of the layer.
Figure 4.5.11 shows each type of pattern distribution.
68
Newton Discrete Element Modeling Software - Manual
Increment Multiplier: When generating layers, Newton divides the generation area into a very fine grid
and steps through the grid searching for open spots to place clusters. By default, the size of this search
grid is 1/20 the radius of the smallest sphere. This is usually small enough to tightly pack the clusters
and large enough to result in a reasonable generation time. However, when generating particle sets
that contain clusters with large numbers of spheres (> 50), this default search grid size can result in
extremely long generation times. Entering a multiplier in this box will multiply the cell size of the search
grid, thereby creating a search grid with fewer cells. The material will generate faster, but it may not be
packed as tightly as possible.
P-Count/Quick Gen: When Newton generates clusters, it uses a random generator to create clusters
according to the probability distribution in the material set library. By default, Newton will generate
however many layers is required to obtain 10,000 random clusters. After 10,000 clusters have been
generated, Newton copies the layers that have been created thus far and repeats them for all remaining
clusters.
The idea is that after 10,000 clusters, there is sufficient randomness in the clusters and layers that from
a statistical point of view, it doesn’t matter whether the other hundreds of thousands of remaining
clusters are generated randomly (which takes time), or copied from the existing layers (which is
essentially instantaneous).
If the particle size is very small (such that each layer will contain close to 10,000 particles), then it may
be desirable to force Newton to generate more particles to get sufficient randomness before copying
and repeating the layers. Enter the number of particles to generate in the first box to force Newton to
do this.
If quick particle generation is desired, then Quick Gen can be enabled by selecting YES for the second
input box. This will discard Newton’s complex cluster generation algorithm and simply space the
particles based on the space taken up by the largest particle in the particle set. Enabling this option is
almost always unnecessary, because the computation time for a simulation is always orders of
magnitude larger than the particle generation time. Figure 4.5.13 shows the difference between
Newton’s cluster generation algorithm and the simple Quick Gen method.
69
Newton Discrete Element Modeling Software - Manual
To force Newton to create a certain number of clusters in each layer, enter the desired number of
clusters in the second input box.
By default, Newton checks to see if particles from sequential layers will overlap each other. If Newton
finds any particles that overlap, it removes them from the simulation so they don’t disturb the other
particles. If the density, tonnage, and minimum particle gap are set correctly, Newton should never find
overlapping particles anyway. To allow overlapping particles during material generation, set the third
input box to YES. This is not ever recommended.
Generate Particles: After all options have been selected, click the Generate Particles button. While the
Newton is generating the material, the text on this button will show current generation progress.
Recall that Newton generates 10,000 clusters and then copies those clusters and layers and repeats
them for the remaining material. While Newton is still generating these 10,000 clusters, the text on the
Generate Particles button will display the percentage of these 10,000 clusters that have been created,
as well as how many layers have been generated thus far. As soon as random generation of 10,000
clusters is complete, Newton will begin copying and repeating these layers. The text on the button will
then display number of generated clusters and layers. Figure 4.5.14 shows the Generate Particles
button for each case.
Figure 4.5.15 – Error shown when more than 5,000 layers are generated
When particle generation is complete, one of two popup windows will display. If all particles were
generated with no problems, the popup box in Figure 4.5.16 will display.
70
Newton Discrete Element Modeling Software - Manual
Figure 4.5.17 – Popup window shown if there is layer overlap during material generation
This window displays the same information and warns that there is material overlap for one of the
material sets. The material is overlapping because Newton has calculated that in order to achieve the
desired tonnage, the layers must be dropped too close together. There are three ways to remove the
material overlap:
1. Increase the layer size: If material is being generated in a drop box on a belt, increase the length
and width of the drop box so that more material can be generated in each layer. If material is
being generated in a custom location, increase the size of the drop shape.
2. Decrease the particle gap: Decrease the size of the Min Particle Gap so that more particles can
fit into each layer.
3. Reduce the Tonnage: If it’s possible, reduce the tonnage of the simulation so less material will
be generated. Or use multiple drop locations.
71
Newton Discrete Element Modeling Software - Manual
After the material has been successfully generated, information about each material set can be viewed
by clicking the Output tab at the bottom of the particle generation window. This will change the view to
show the output information. Figure 4.5.18 shows the Output window.
72
Newton Discrete Element Modeling Software - Manual
Figure 4.5.19 – Three cluster generation planes. The blue shaded plane is the current Generation Point
73
Newton Discrete Element Modeling Software - Manual
74
Newton Discrete Element Modeling Software - Manual
Figure 4.5.21 – You can customize your chain links in the Cluster Creator window
A chain link particle set must contain two chain links. Generally, they are identical clusters that are
oriented perpendicular to each other. However, if you wanted, you could have two different links, or
you could orient them at 45 or 60 degrees to each other, rather than 90 degrees (for example).
75
Newton Discrete Element Modeling Software - Manual
76
Newton Discrete Element Modeling Software - Manual
77
Newton Discrete Element Modeling Software - Manual
Figure 4.6.1 – Six of the default Newton particle sets, shown at the same scale
Recall that the base unit of material in Newton is a Sphere, and that spheres are grouped together to
form Clusters. Material sets that use relatively large spheres (such as the Spheres Only 200 mm set in
Figure 4.6.1) can run as much as 100 times faster than particle sets using very small particles (such as the
20-60 mm set in Figure 4.6.1). There are two reasons for this. When using very large particles, Newton
can create far fewer particles to represent the same tonnage of material. Additionally, the size of the
time step that Newton calculates depends greatly on the size of the particle, and larger particles will
result in larger time steps, and therefore decreased runtime.
Of course, the smaller the particles in the material set, the more realistic the simulation becomes. You
must balance the desired particle size against the realistic time constraints of the simulation. For
instance, a simulation that models material flow at 1000 TPH using spheres of diameter 1 mm would
require several years of run time, but the same simulation using spheres of diameter 50 mm would run
in a few hours.
By applying effects like Liquid Bridge and Ratchet (see Section 4.2), large particles can be given the same
global flow properties as very fine material, allowing you to create a realistic simulation that will run in a
few days and give reliable results.
78
Newton Discrete Element Modeling Software - Manual
79
Newton Discrete Element Modeling Software - Manual
Regenerate: After modifying the input parameters for the grid described below, click this button and
Newton will overwrite the current particle set with a new particle set using the given parameters. The
Rosin-Rammler curve and the distribution bar graphs will be updated as well. Note that clicking this
button will redefine the current particle set and this action cannot be undone; if you accidentally delete
a particle set, you have to re-import your particle library file.
Number of Groups: This specifies how many different types of clusters you want for the particle set.
Maximum Size: The maximum cluster size for the particle set. Note that this is the maximum size of an
entire cluster, not a single sphere in that cluster.
Minimum Size: The minimum cluster size for the particle set.
Average Size: The average size of a cluster for the particle set.
Uniformity: The Rosin-Rammler curve is follows a Weibull statistical distribution. The uniformity is the
shape parameter k of this distribution. This parameter is used to determine the percent of each cluster
type that will appear in the particle set.
Size Distribution Factor: This specifies how to determine the size of each cluster type. Values less than
1 will create a particle set where most of the cluster types are larger than the average cluster size. A
value of 1.0 will create a particle set with cluster sizes that take normally-spaced values between the
Minimum and Maximum Size. Values larger than 1.0 will create a particle set where most of the cluster
types are smaller than the average cluster size. Typical values for this take the range of 0.5-4.0.
Size Ratio: This specifies the maximum ratio of sphere sizes among spheres that make up the clusters.
Note that this does not refer to the ratio of cluster sizes, but the spheres that make the clusters.
The Size Ratio should be set as close to 1.0 as possible (it cannot be set lower than 1.0). When a
simulation has spheres that are vastly different sizes, the number of nearest neighbor particles for each
particle increases; this significantly increases the runtime for the simulation. It is not recommended that
you use a size ratio any larger than 2.0, or else your simulation will have a drastically longer runtime.
Values in the range of 1.0-1.5 are typical.
Solids Ratio: This is the ratio of solid material to total volume occupied by a mass of particles. When
10,000 particles are dumped into a cylinder, there are gaps between the particles, and the total volume
occupied by the particles is larger than the sum of the volumes of each individual particle because of the
spaces between particles.
The Solids Ratio MUST be calculated carefully to ensure that Newton creates proper material tonnages.
For instance, consider a situation where this ratio should be 0.50 for a material set, but it is set at 0.75
instead. When Newton generates 1,000 TPH of material for a simulation, it will actually generate a
higher volumetric flowrate of material at a lower bulk density because the solid/void ratio incorrectly
indicates that there is very little space between particles (50% less space, to be exact, so the flowrate
would appear 50% higher). Proper calculation of the Solids Ratio is described in Section 6.2.
Plot Spacing: This value defines how Newton should display the particle set in the particle set rendering
area in the lower-left corner of Figure 4.6.2). Setting this to a larger value (such as 20) will render the
particles closer together, which is useful when analyzing very small particle sets. Typical values for this
range from 5-20. Note that this input does not affect the material set in any way, it is simply for
rendering the particles in this window while you are viewing and modifying the particle sets.
Rosin-Rammler Curve: This is a cumulative density function statistical curve which shows the
distribution of cluster sizes in the particle set. The red line represents the calculated curve based on the
80
Newton Discrete Element Modeling Software - Manual
input parameters, and each star on the curve represents a particle set. The abscissa of the graph is the
cluster size in millimeters, and the ordinate of the graph represents the percentage of total clusters in
the particle set.
The Rosin-Rammler curve is a form of a statistical distribution, the Weibull distribution. The curve
describes particle size distributions. The stars on the graph should fairly closely follow the Rosin-
Rammler curve.
Distribution Bar Graph: This bar graph displays information about the particle set. It shows the size of
each cluster along the abscissa. Three different bars can be displayed for each cluster, they can be
selected using the combo box at the top left of the bar graph. The bars show the percentage by weight
of each cluster, the percentage of total clusters for each cluster type, and the percentage of total
particles for each cluster type. This graph is useful for checking if any cluster types occupy too much or
too little of the total weight or total count of the particles.
Information Labels
In the top right of the Size Distribution window, there are four labels that give data about the entire
particle set.
81
Newton Discrete Element Modeling Software - Manual
Minimum Timestep: Gives the approximate minimum time step required for a simulation using this
particle set. This also takes into account the Maximum Expected Velocity and Maximum Allowable
Particle Overlap from the General Variables window.
Particles per 10m3: The approximate number of spheres that would be generated if 10 m3 of material
was generated using this particle set.
Sphere Size Ratio: The ratio of the radius of the largest sphere to the radius of the smallest sphere.
Ideally, this ratio should be as close to 1.0 as possible. Values in the range of 1.0-1.5 are typical. Values
above 2.0 will cause the simulation to run much slower.
Simulation Time Index: This dimensionless parameter is a composite of the three parameters listed
above. This gives the approximate relative simulation time for a simulation using this particle set.
Lower values are better. A particle set with a Simulation Time Index of 2,000 will take approximately
twice as long to compute as a simulation with an index of 1,000. Computation time is affected by
hundreds of other factors as well, so this index is only an approximation.
Cluster Grid
The grid in the center of the window lists all the cluster types for the particle set. The columns for the
grid are described below.
ID #: This is the ranking number of the cluster type in the particle set list. It has no significance other
than to help differentiate each cluster type. This can be changed by re-ordering the cluster types using
the Move Up and Move Down buttons at the top of the window.
Group Type: This is where you specify the cluster type. The input is a dropdown box from which you
can select any cluster from the particle and cluster library. When you click the Regenerate button,
Newton automatically generates cluster types, ignoring the types in your cluster library.
Cluster Size: This is the size of the cluster, measured as the longest line that can be drawn through the
particle in any direction. This parameter can be directly modified by entering a new value in the input
box.
Material Type: This species which Material Properties Set (A, B, or C) will be used for this cluster. If you
are using sets B or C, make sure you have defined at least the coefficients of friction, coefficient of
restitution, and rotational damping for those sets.
% By Weight: The percentage of the cluster type by mass of the total generated material for the particle
set.
Clusters per 10,000: Specifies how many clusters in a group of 10,000 should be of this type.
Sphere Radius: The radius of the first sphere in the cluster.
Spheres in Cluster: The number of spheres that make up the cluster type.
Time Step Index: The ratio of the approximate required time step for this cluster to the minimum time
step for the entire particle set (shown in the Minimum Timestep label at the top of the window). To
obtain the approximate required time step for this specific cluster, multiply the Time Step Index for this
cluster by the Minimum Timestep for the set.
Buttons
The buttons along the top of the window allow you to add and remove cluster types, and reposition and
sort them in the list. The function of each button is described below:
82
Newton Discrete Element Modeling Software - Manual
Insert Cluster: Adds a grid row for a new cluster at the current position.
Delete Cluster: Deletes the current selected cluster from the particle set.
Move Up: Moves the current selected cluster row up by one position.
Move Down: Moves the current selected cluster row down by one position.
= Clusters: Sets the cluster count per 10m3 for each cluster to the same number.
= Spheres: Sets the sphere count per 10m3 for each cluster to the same number.
Sort: Sorts the rows according to Cluster Size, with the largest cluster at the top of the list and the
smallest cluster at the bottom of the list.
Scale: Scales the cluster size of all cluster types by the scale value in the adjacent input box.
Refresh Set: This refreshes each cluster to ensure that all recent changes in the Cluster Creator and Size
Distribution windows has been accounted for.
Freeze Cluster Count: If this checkbox is checked, then when the Cluster Count per 10,000 or % by
Weight is changed for a cluster, that value will be frozen, preventing it from being redistributed when
other cluster counts are changed. Once this checkbox is unchecked, all cluster counts will be unfrozen.
Chain Set: Newton can create particle chains using particle sets that are composed of cluster links.
Checking this check box informs Newton that the current particle set is a Chain Link particle set.
83
Newton Discrete Element Modeling Software - Manual
The term Internal refers to the particle library file that is embedded into the current input file. The term
External refers to some external library file that can be opened for importing or exporting.
This window allows you to do three things:
1. Export clusters and particle sets from the current library to a blank library file.
2. Export clusters and particle sets from the current library to an existing library file.
3. Import clusters and particle sets from an external library file to the internal library.
The window contains four colored panels. The blue panels list the clusters contained in the internal
(left) and external (right) library files. The yellow panels list the particle sets contained in the internal
(left) and external (right) library files.
When a particle set is moved from one library to another, Newton automatically checks which clusters
belong to that particle set and ensures that those clusters are moved as well.
Exporting to a Blank Library: Highlight the desired clusters/particle sets and click the right-arrow button
to send them to the external library. Once you have moved all necessary clusters/particle sets, click the
Save Library button and specify where to save the library.
If you have already opened an external library file, then you cannot export to a blank library file. You
have to close the Merge/Export Library Files window and re-open it.
Exporting to an Existing Library: Click the Open Library button to open the desired external library file.
Newton will populate the right-hand panes with clusters and particle sets from the external library.
Highlight the desired clusters/particle sets and click the right-arrow button to send them to the external
library. Once you have added all necessary clusters/particle sets, click the Save Library button and
specify where to save the library. Note that you don’t have to save over the original external library file,
you can save it as a different library.
Import from an External Library: Click the Open Library button to open the desired external library file.
Newton will populate the right-hand panes with clusters and particle sets from the external library.
Highlight the desired clusters/particle set and click the left-arrow button to bring them into the internal
library. Once you have imported all necessary clusters/particle sets, just close the Merge/Export Library
Files window.
Note that when you modify the internal library file by importing clusters and particle sets, the internal
library is automatically updated; there is no need to ‘save’ the internal library. This also means that you
cannot undo an import. You have to go to the library and manually delete a particle set or cluster if you
import it accidentally.
Check for Possible Identical Clusters: If this checkbox is checked, then if Newton tries to import a
cluster with the same name as an existing cluster, Newton will compare the parameters of both clusters
to see if they match within 0.001% of each other. If so, then Newton will ask if you want to assume that
the clusters are identical. If you select Yes, then Newton will cancel importing that cluster and make
sure that all library files point to the existing cluster instead. If you select No, then Newton will prompt
you for a new cluster name. After giving a new cluster name, Newton will make sure that all particle
sets containing that cluster are updated with the new name.
Ignore Clusters that have Already Been Copied: If this checkbox is checked, then Newton will maintain
a list of which clusters have already been imported/exported and will not try to import/export them
again. Note that if the Merge/Export Library Files window is closed and then re-opened, this list is
reset.
84
Newton Discrete Element Modeling Software - Manual
85
Newton Discrete Element Modeling Software - Manual
Import CSV: Imports a CSV file and saves it as a cluster. The CSV file must contain the sphere number,
radius, and X, Y, and Z positions for each sphere in the cluster. In order to get the formatting of the CSV
file correct, you should export one of the other clusters to a CSV file and use it as a template for creating
your own CSV cluster file.
Import DXF: Imports a DXF file and saves it as a cluster. The DXF file must contain only spheres. As with
a CSV file, it is helpful to first export a cluster to a DXF file and use that file as a template.
Below the cluster combo box, there are three buttons that allow you to create, rename, and delete
clusters.
Number of Spheres: In the input box adjacent to this label, enter the number of spheres that will
compose the cluster.
The grid below Number of Particles contains the radius and X, Y, and Z information for each sphere.
Modify these values to create your cluster.
The box below the grid contains information about the cluster, including volume, center of gravity, and
moment of inertia about each axis.
Lock Rotation: Check this checkbox if you want Newton to prevent this cluster from rotating in the
simulation. This option is not usually needed. If you are using a material set that includes multiple
cluster types that are composed of a single sphere, then it might be helpful to lock the rotation on one
of those spherical cluster types so that the material does not roll around.
86
Newton Discrete Element Modeling Software - Manual
Cluster Meshing
Checking the Mesh checkbox allows you to preview the meshed version of the cluster (see Section
5.3.3). We have a tutorial on YouTube at http://youtu.be/qv7RXFSSG_w which discusses how to set up
and use the meshes.
Meshing is a rendering effect that helps make the clusters look less like spheres and more like irregular
particles. The effect does not alter the calculations in any way, it is only a visual effect. Figure 4.7.3
shows a cluster with and without the mesh effect applied.
Figure 4.7.3 – A cluster without (left) and with (right) the mesh effect
Figure 4.7.4 – A cluster prepped for meshing; ignored spheres are shown in gray
Mesh: Check this box to preview the cluster as a mesh. The mesh is generating using all spheres except
those which are set to Ignore. By ignoring certain spheres, you can control the size and shape of the
mesh.
Show Ign: If this checkbox is checked, then Newton will color the ‘ignored’ spheres as gray in the
rendering window. This only applies to viewing the cluster in the Cluster Creator window.
87
Newton Discrete Element Modeling Software - Manual
Include Ign: If this checkbox is checked, then when generating the cluster mesh, Newton will include
the ‘ignored’ spheres. This only applies to previewing the cluster in the Cluster Creator. When
generating meshes during a simulation or in a playback file, Newton won’t include the ignored spheres.
This is useful for helping you quickly swap between looking at the mesh with and without the ignored
spheres.
Factor: The mesh factor determines how tightly the mesh will follow the curvature of the spheres.
Higher values will create a tighter mesh and lower values will create a larger, more blob-like mesh.
Quality: The mesh quality determines the resolution of the mesh. Higher resolutions will look
smoother, but will require more memory and will take longer to render.
Note: For all the clusters in the default particle library (including the new clusters), we have already
optimized the Mesh Factor and Mesh Quality in the cluster creator. Feel free to change those values
how you see fit, but the current settings will provide a good quality mesh that follows the contour of
each cluster very well.
88
Newton Discrete Element Modeling Software - Manual
5 Running a Simulation
89
Newton Discrete Element Modeling Software - Manual
- File Version Error: If the Newton file is from a newer version of Newton, it cannot be opened.
Check that you have the most current version of Newton by navigating to the Help menu and
clicking Check for Updates.
- File Not Found: Newton cannot find the specified file. The file must have been moved after being
placed in the queue.
- Layer Move Profile Error: The file to be simulated has custom CSV movement profiles, but there
was a problem with one or more of the movement profiles. Make sure to check the following (also
see Section 4.3.9):
o If your movement profile contains rotation, but you have not specified any custom rotation
vectors, check that each layer using rotation has a non-zero rotation vector specified in the
Geometry input page. See the Help window on the Layer Movement Profile window for
more information.
o If your movement profile contains rotation and you have specified custom rotation vectors
in the LMP, check that the first move that has a non-zero rotation rate also has a non-zero
rotation vector.
- # Clusters > License Type: The input file was created with a different version of Newton that allows
more clusters to be generated. The file cannot be run until the material is regenerated with the
proper number of clusters.
- Missing rotation vector: One or more of the layers has a Rotation Rate specified on the Geometry
page, but the corresponding Rotation Axis is a zero-length vector (0, 0, 0). To have rotation, a
vector about which to rotate the model must be defined.
When running simulations via the File Queuing window, you can still pause and stop the simulations
using the Pause and Stop commands in the Solver dropdown menu. Clicking the Stop command in this
dropdown menu is identical to clicking the Skip to Next Simulation button in the file queuing window.
90
Newton Discrete Element Modeling Software - Manual
91
Newton Discrete Element Modeling Software - Manual
The top left corner of the rendering window displays simulation time in seconds, and in percent of total
simulation completed. Below that are the run time and the estimated time to completion of the
simulation (ETC). Finally, below the run time are the current cluster and particle count.
If you are dropping material in layers, then when you generated material , Newton most likely told you
that you generated some 200,000 particles (for example). But when you run the simulation, you will
never see ‘Particles = 200,000’ in the top left corner, because only a fraction of those particles are
moving through the simulation at one time. It would be very inefficient to actually generate all 200,000
particles at once and throw them all into a hopper above the feed belt.
The top right corner contains the Color Scale. In most cases, this legend will show material velocity, but
it will change based on the material view settings (see Section 5.3).
Figure 5.2.3 – Rendering window context menu with Velocity Scale submenu open
Pan, Orbit, Zoom, Zoom Window, None, Preset Views, Show UCS Icon: See Section 4.3.7
Orbit C.G.: This allows you to choose the point about which the model will rotate. By default, Newton
rotates the model about the origin, but this can be changed to rotate about the center of gravity of any
layer.
Orbit Sensitivity: This allows you to choose how quickly the model rotates when you are in Orbit mode.
92
Newton Discrete Element Modeling Software - Manual
Orbit Commands: Click this opens the advanced orbit commands, which are displayed on the right side
of the rendering window (outlined in magenta in Figure 5.2.3). These commands allow you to rotate the
model based on the current camera view.
Input Box: Enter the specified rotation amount in this text box, in degrees.
Left/Right: The left/right arrows will rotate the model about the vertical axis (running vertically
up and down the screen).
Up/Down: The up/down arrows will rotate the model about the horizontal axis (running left and
right across the screen)
CW/CCW: The clockwise/counter-clockwise arrows will rotate the model about the axis which
points out of the screen (i.e. it rotates the model in a CW/CCW fashion).
Surfaces: See Section 3.4.3.
Particles: See Section 3.4.3.
Velocity Scale: This submenu allows you to quickly change the velocity scale for the particles without
having to open the Surface and Material Visualization window. Clicking one of the scales in this
submenu will also change the particle color mode to Color by Velocity.
Geometry: This submenu contains commands for showing or hiding extraneous geometry.
Triangle Normal Vectors: Enabling this option will render a magenta vector at the center of
gravity of each triangle. The direction of the vector represents the positive normal vector of the
triangle.
Killboxes: Enabling this option will render all killboxes in the rendering window. Active killboxes
will be shaded magenta, which inactive killboxes will be shaded gray.
Simulation Bounds: Enabling this option will render the min/max X, Y, and Z bounds for the
simulation.
SMV Window: Opens the Surface and Material Visualization (SMV) window (see Section 5.3).
93
Newton Discrete Element Modeling Software - Manual
Layer Grid
Hidden: Completely hides the layer from view.
Solid: Renders the layer as a solid surface of the chosen color.
Wireframe: Renders a wireframe view of the surface, showing every individual triangle on the layer.
Transparent: Renders the layer as transparent. The transparency level can be set with the slider bar at
the bottom of the Surface Style panel at the top left of of the SMV window.
Work: Renders the work done on the surface. Note that in order to use this setting, you must set Save
Triangle Force & Work Data to YES in the general simulation parameters window before running the
simulation. This option is not available while the simulation is running, it is used only during post-
processing.
94
Newton Discrete Element Modeling Software - Manual
Force: Shows the the force on the layer. Note that in order to use this setting, you must set Save
Triangle Force & Work Data to YES in the general simulation parameters window before running the
simulation. This option is not available while the simulation is running, it is used only during post-
processing.
Outer Line: Shows the major outer lines of the layer. This setting can be combined with any of the
other view settings. For instance, a layer can be set to transparent and then the outer lines can be
turned on to help give the layer clear definition.
Color: Allows you to choose the color for each layer.
The top left corner of the SMV window contains a section entitled Surface Style. The purpose of this
section is to enable you to quickly set layer view settings for the entire model. Each button in the
section will apply the view effect to every layer. Additionally, this section contains the transparency
slider bar, which controls the level of transparency for all layers.
Simple Transparency
The normal transparency mode that Newton employs cannot easily differentiate between the ‘front’
and ‘back’ of a triangle due to the numerical difficulties of properly determining the normal vectors for
an irregular multi-surfaced non-closed non-continuous geometry. This is why sometimes surfaces (most
notably hoods and spoons) are not colored or shaded properly. This effect is especially noticeable when
the outer lines are turned off.
Therefore Newton has a Simple Transparency mode, which essentially ignores the front and back sides
of the triangles and treats all triangles equally. The downside of this is that the transparency must
always be set in the range of 80-100% transparent. This means that Simple Transparency cannot be
used in an AVI that uses an interpolated transparency sweep from 0-100% transparent (i.e. sweeping
from solid to transparent surfaces). However, Simple Transparency looks much better, especially when
viewing a transparent geometry with outer lines turned off.
Simple Transparency is enabled by default. But in the Preferences window, you can specify whether
Newton should Default to Simple Transparency on startup (see Section 3.5.1). To switch Simple
Transparency on or off, use the checkbox below the transparency slider.
Figure 5.3.2 - A transparent geometry with (left) and without (right) Simple Transparency enabled.
95
Newton Discrete Element Modeling Software - Manual
96
Newton Discrete Element Modeling Software - Manual
Figure 5.3.3 – Transparent particle setting: lumps are red, chains are magenta
Velocity: The default coloring mode is velocity. This colors each cluster based on its X, Y, Z, or Σ
velocity. The Σ velocity is just the absolute particle speed. In order to view only the X, Y, or Z direction,
set the final column to X, Y, or Z. Note that in order to view the X, Y, or Z velocities, you must set Save
Particle Force & Work Data to YES in the general simulation parameters window before running the
simulation. Otherwise the playback file will only save the Σ particle speed (this helps keep the output
playback file size small).
Fixed-Position: When this radio button is selected, all clusters in the simulation are immediately
colored according to their current X, Y, or Z position. As the simulation progresses, the color of each
cluster remains the same despite changes in cluster position. This allows you to see how the material
flows through the simulation. It is particularly useful if you are simulating bin or hopper flow. Coloring
the particles by Fixed-Z allows us to see how the geometry in Figure 5.3.4 will empty and whether there
will be patches of stagnant particles around the lower edges.
97
Newton Discrete Element Modeling Software - Manual
Figure 5.3.4 – Fixed-Z particle coloring at the instant it is enabled (left) and 3 seconds later (right)
Elevation: Colors each cluster according to its instantaneous X, Y, or Z position. Unlike Fixed-Position,
this setting uses a gradiated color scale, and the color of each cluster will change as the particles move
through the simulation.
Dimension: Colors each cluster according to its dimension in the X, Y, Z, or Σ direction. The Σ direction is
just the maximum cluster size along any axis. The X, Y, and Z dimensions are the maximum cluster size
in the X, Y, and Z directions, relative to how the cluster was originally defined in the Cluster Creator (see
Section 4.7). You will almost always use the Σ direction because this will always be the longest cluster
length.
Mass: Colors each cluster according to its mass. This setting has no associated direction, so the direction
column is grayed out.
Force: This colors each cluster according to the X, Y, Z, or Σ normal force applied to it. In order to view
particle force data, you must set Save Particle Force & Work Data to YES in the general simulation
parameters window before running the simulation.
Relative Velocity: This mode is used when you want to view the cluster velocity relative to some other
velocity (i.e. a conveyor belt). Typically you would use this when you are viewing the clusters as they
discharge on a receiving belt.
1. Enter receiving belt X velocity into the Min grid cell.
2. Enter receiving belt Y velocity into the Max grid cell.
3. Enter receiving belt Z velocity into the Div grid cell.
4. Set the direction column to Σ.
Then Newton will take the velocity of each cluster and subtract the receiving belt velocity components,
showing you the net speed of each cluster relative to the receiving belt.
98
Newton Discrete Element Modeling Software - Manual
If you set the direction column to X, then Newton will take the X-velocity of each cluster and subtract
the X-velocity of the receiving belt that you entered into the Min grid cell. Then you will be viewing just
the X-velocity of each cluster relative to the receiving belt. The same is true if you set the direction
column to Y or Z.
In order to view Relative Velocity, you must set Save Particle Force & Work Data to YES in the general
simulation variables window before running the simulation.
Particle Work: Displays the work done on each cluster. As with work on chute surfaces, the main
reason for modeling particle work is not to determine the absolute work done on the particle, but rather
to determine the relative particle work from one chute design to another. Note that in order to use this
setting, you must to set Save Particle Force & Work Data to YES in the general simulation variables
window. This option is disabled while the simulation is running. It is only used during post-processing.
The grid cells next to the Work radio button are grayed out because all view settings for particle work
are controlled on the Work tab of the SMV window (see Section 6.4.5).
99
Newton Discrete Element Modeling Software - Manual
Figure 5.3.6 – Chute Geometry without (left) and with (right) Outline Tolerance
You can also choose one layer that will be excluded from the outline tolerance by entering the number
of that layer in the Exclude Layer input box. All other layers will use the outline tolerance.
100
Newton Discrete Element Modeling Software - Manual
Len+: The X, Y, and Z inputs correspond to the distance from the center of the clip box to the positive
length, width, and height faces, respectively.
Len-: The X, Y, and Z inputs correspond to the distance from the center of the clip box to the negative
length, width, and height faces, respectively.
Z-Ang: The rotation of the clip box about the vertical (Z) axis.
With the Len+ and Len- inputs, any of the six sides of the rotated clip plane can be independently
extended or retracted without having to mess with the XYZ position of the box.
In the first drop down box, select whether to turn the clip plane Off, to Show Data in Box, or to Hide
Data in Box. In the second drop down box, select whether to apply the clip effect to Particles, Triangles,
Lines, or All Objects.
If two or more clip planes have intersecting spaces, then any clip plane which is designated to Show
Data in Box will trump any plane that is designated to Hide Data in Box. Figure 5.3.7 shows two clip
plane boxes inserted into a model. The first box is set to hide all data from -1.5 to +1.5 in the X, Y, and Z
directions. The second box is set to show particles from -0.75 to +0.75 in the X, Y, and Z directions.
101
Newton Discrete Element Modeling Software - Manual
6 Post-Processing
102
Newton Discrete Element Modeling Software - Manual
103
Newton Discrete Element Modeling Software - Manual
What to Export
Choose whether to export the triangles (layers), clusters, or both. You can export the clusters as
spheres or as lines which show the position of each sphere. Exporting as spheres requires about twice
as much file space, but will look better when the file is opened in AutoCAD. Figure 6.1.4 shows a set of
spheres exported both as spheres and as lines.
You can export position tracers as well. Newton uses the currently-visible tracers, so before exporting,
make sure you enable position tracers and skip forward a few frames so that Newton generates the
tracers.
104
Newton Discrete Element Modeling Software - Manual
Figure 6.1.5 – File Error popup that displays when opening an incomplete simulation
This occurs because Newton detects incomplete simulation data in the data file, and this can be
repaired. The tool that repairs playback files can also be used to reduce playback files. If a playback file
size is too large, or if for any other reason you want to reduce a playback file, Newton allows you to do
this.
Open the File menu, expand the Playback Tools submenu, and click on Reduce Playback File. A window
like the one shown in Figure 6.1.6 will be opened.
105
Newton Discrete Element Modeling Software - Manual
Reduced Frame Interval: If you want to make the uncompressed file smaller by removing frames, then
enter an integer in this box. Entering 0 will retain all frames and will simply fix any file errors that
Newton finds. Entering 1 will remove every other frame, entering 2 will remove 2 out of 3 frames, etc.
Start Time: If you want to remove frames from the start or end of the uncompressed file, then enter
the desired starting time for the new uncompressed file in this input box.
End Time: In order to keep all frames after the Start Time, leave this box blank. Otherwise, enter the
simulation time for the last frame of the new file.
Specifically Exclude Work Information: During the pre-processing for the simulation, you may have set
Save Triangle Force & Work Data or Save Particle Force & Work Data to YES. This would result in a
much larger uncompressed file size. If this information is not necessary, then you can check this box and
Newton will remove that data from the file, decreasing the file size.
Compress Crunched File: Select whether Newton should compress the file to save space. Newton
playback files are always compressed by default, so we recommend leaving this option checked.
Reduce: When you have entered all the information, click this button and Newton will reduce or repair
the data file.
Batch Reduce: If you want to create a file queue to reduce multiple files, Newton allows you to do this.
Create a text file, and enter the complete file path for each file, one per line, as shown in Figure 6.1.7.
Once you have created this text file, click the Batch Reduce button and open that text file. Newton will
reduce/repair each file using same specified input parameters.
106
Newton Discrete Element Modeling Software - Manual
107
Newton Discrete Element Modeling Software - Manual
These input files contain cylinders with the specified radius. The size of the cylinder you use depends on
the size of the particles in your particle set. The goal is to have a simulation that uses 50,000 to 200,000
particles. In order to accurately calculate the solids ratio, the simulation must use a large number of
particles.
Once the input file is opened, navigate to the Size Distribution library. If your particle set is saved in the
default Newton particle library file, then select it from the drop down menu. Otherwise, open the
library file containing your particle set, then select your particle set from the drop down menu.
Figure 6.2.2 – Set the Solids Ratio for the particle set
Confirm all the parameters for your particle set, and then make sure that solids ratio is set to 0.55.
Next, open Material Generation window, where the material is generated for the simulation. Select
your particle set from the Material Group Set drop down box and click the Generate Particles button.
For more information about generating material for simulations, see Section 4.5.
After Newton has generated the particles, make sure that about 50,000 to 200,000 particles were
generated. If less than 50,000 particles were generated, then you should use a larger cylinder. If more
than 200,000 particles were generated, you may use a smaller cylinder to save computation time. But
the more particles that are dropped into the cylinder, the more accurate the calculated solids ratio will
be.
After the material has been generated, check the simulation input parameters in the General Variables
window. The default settings should work fine, but you may want to change the file name and file save
location. For more information about setting the parameters in this window, see Section 4.1. Save a
copy of the input file, and then run the simulation by opening the Solver menu, and clicking Run.
108
Newton Discrete Element Modeling Software - Manual
109
Newton Discrete Element Modeling Software - Manual
6.2.5. This window outputs the height of material in the cylinder (for reference only), as well as the
solids ratio.
110
Newton Discrete Element Modeling Software - Manual
Depending on the thickness of the slice used for estimation, the estimated area can differ quite
significantly (The areas of largest and smallest sections in Figure 6.2.6 differ by over 20%). Ideally, of
course, the slice would be 1 particle thick and would be integrated along the entire length of the belt,
giving the exact volume of material on the belt. It should also be noted that when the slice thickness
becomes very small (<= 200 mm), the gaps between the particles on the belt are clearly visible. On a
real conveyor, these ‘gaps’ would be mostly filled in with fines. However, Newton does not model the
finest of particles, rather it gives the entire material set the same ‘global flow properties’ as very fine
material has. The presence of these gaps under high friction conditions leads to the area calculation
being artificially high. This is one reason why AC-Tek recommends calculating the solids ratio using low
friction settings.
Consider Figures 6.2.7 through 6.2.10. These figures were exported from four Newton simulations in
which material was dropped onto a belt under two different friction settings, using Solid/Void ratios that
were calculated using free-flowing and high friction settings. The goal of these simulations was to
compare the estimated cross-sectional area of material on the belt in each case. In these figures, the
yellow area line represents the approximate cross-sectional area of material on the belt. The white area
line represents the theoretical cross-sectional area for the given tonnage and belt speed.
Figure 6.2.7 – Free-Flowing simulation using an SV ratio calculated under high friction conditions
Figure 6.2.8 – High friction simulation us an SV ratio calculated under high friction conditions
Figure 6.2.9 – Free-flowing simulation using an SV ratio calculated under low friction conditions
111
Newton Discrete Element Modeling Software - Manual
Figure 6.2.10 – High friction simulation using an SV ratio calculated under low friction conditions
It should be noted that although the percent-error calculation is +7.9% for the simulation in Figure 6.2.8,
when the white and yellow area lines are compared, they are so close together that the case could even
be made that they are essentially the same line. And when analyzing Figure 6.2.10, when the gaps
between the particles (particularly near the top of the material, i.e. the empty space) are considered,
the experimental and theoretical areas are not very different.
Additionally, for high-friction cases, there are large gaps between the particles due to the high
coefficient of friction between the particles. On a real conveyor, these gaps would be at least partially
filled by fine material, which Newton does not model. Newton gives the entire flow the same ‘global
flow properties’ to compensate for this. In the DEM simulation, when the high-friction material flows
into a chute and piles up, the particles are squeezed together and a lot of that space is eliminated. It is
at this point in the simulation that that it is imperative that there is enough material on the belt. When
in doubt, being slightly conservative is better, because if a certain transfer chute does not plug up under
a conservative case, then it shouldn’t plug up using a smaller solid/void ratio either (i.e. a solids ratio
that had been calculated using high friction settings, rather than AC-Tek’s recommend low friction
settings). And if the chute does plug up under a slightly conservative case, then it’s probably a good
indication that the chute needs to be adjusted in some way in any case.
Furthermore, in real conveyor engineering, there is often be a significant variability in the ‘calculated’
value of the bulk density (and in some case even the tonnage) that is used (possibly as much as 10-15%).
And the consistency and material properties of the ore on the belt can vary over the course of a day due
to weather as well (temperature, sunshine, relative humidity, etc.).
With all this in mind, whether the solids ratio is calculated using low friction or high friction settings is
left to the discretion of the user. AC-Tek recommends using low friction settings for the reasons stated
above.
112
Newton Discrete Element Modeling Software - Manual
113
Newton Discrete Element Modeling Software - Manual
114
Newton Discrete Element Modeling Software - Manual
way, you only have to create a key frames file once, and it can be used to create animations for all
simulations using that chute.
Save File: Saves the animation settings to a Newton key frames file (*.KFF). You should always save a
key frames file when you create an animation. If you made a small error in the key frames, it could ruin
the animation; instead of having to recreate the entire animation, you can open the key frames file,
make the small correction, and re-encode the animation.
Create AVI: This button creates the animation file. Clicking this button will open a file prompt, so you
can choose where to save the animation. After this prompt, another window will open, prompting you
which codec to select. This window is shown in Figure 6.4.2.
115
Newton Discrete Element Modeling Software - Manual
drag and resize the screen while the drop down box is set to Custom and Newton will automatically
change the size in the input boxes (See Section 6.4.2).
Frame Rate: Enter the desired frame rate (in frames per second) in the adjacent input box.
Frame Skipping: This feature allows you to skip frames in the video, creating an animation with fewer
frames, and therefore smaller file size. If you enter ‘1’, Newton will render every other frame. If you
enter ‘2’, Newton render 1 frame, then skip 2 frames, etc.
Frame skipping cannot be used if you are including triangle or particle work in the animation file. The
triangle and particle work is saved with each specific data frame, so skipping frames during creation of
the animation will also cause Newton to skip the work data in those frames, resulting in an incorrect
work summation value.
Show Particle Count: Check this box to have Newton display the number of particles and clusters in the
animation file. Note that if this box is checked, then the particle and cluster count will override the sub
comment for any key frames. I.e. the sub comments will not be displayed, and the cluster and particle
count will be displayed in its place.
Sine Interpolation: By default, when interpolating between camera views, Newton uses simple linear
interpolation. Checking this checkbox will force Newton to interpolate the views using a Sine function,
identical to how Newton interpolates between camera views if you click the preset view buttons.
Copy Key Frame: This button opens the window shown below, which allows you to quickly copy layer
settings from the current key frame to any other key frame. In the left panel, select the properties that
you want to copy from the current key frame. In the right panel, select the destination key frames.
116
Newton Discrete Element Modeling Software - Manual
Text Color/Background Color: These buttons allow you to specify the color of the screen text and the
background for the animation file.
117
Newton Discrete Element Modeling Software - Manual
118
Newton Discrete Element Modeling Software - Manual
119
Newton Discrete Element Modeling Software - Manual
The estimated values of work on particles and chute surfaces in Newton is intended to be used for
relative comparison purposes between chute designs and revisions only.
That is, these work values are best used to compare the wear that one chute design would experience
compared to another design. For instance, if Newton calculates 2,000 J of work on a wear plate using
chute design A, and only 1,200 J of work on the same wear plate using design B, then the user could
reasonable assume that chute design B, would reduce the wear on that wear plate by approximately
40%. The user should not assume that the exact values of work on the wear plates for chute designs A
and B are 200 kJ and 120 kJ, respectively.
120
Newton Discrete Element Modeling Software - Manual
121
Newton Discrete Element Modeling Software - Manual
Work Units: These two panels allow you to set the work units for triangles and particles. Please note
that these work units should be considered work indexes, rather than strict work units, because the
work feature should be used on a comparison basis for analyzing multiple geometries within Newton.
Screen Labels: Checking the Show Max Work and Show Total Work will display a line of text in the
rendering window. The text that this line displays depends on whether triangle work or particle work is
being shown. If triangle work is being shown, then this text will display the maximum work on any single
triangle that is currently set to show work, as well as the total work done all triangles that are currently
set to show work. If particle work is being shown, then this line will display the maximum work on any
single particle, as well as the total work done on all particles. If the radio button J is selected, then this
text line will display work in Joules, and if J/sec is selected, then the text line will display the work
averaged over time. The position of this line of text can be set in the AVI Creator window (see Section
6.4.1).
Triangle Impact and Abrasion: The middle column of sections contains options for displaying triangle
work.
Triangle Rendering Mode: This option determines how Newton will render triangle work. If Triangular
is selected, then Newton will add up the work from each particle according to the triangle that it
impacts. All the work from that particle will be attributed to that triangle.
Each triangle contains 3 vertices; these are the A, B, and C ‘points’ of the triangle. Each vertex is shared
by multiple triangles, so it is called a ‘node’ rather than a vertex. If Nodal is selected, then Newton will
take the work done by a particle on a triangle, and divide that work between the three nodes of that
triangle. The work done by other triangles attached to the same nodes is also added to those nodes.
Nodal rendering mode is more accurate than Triangular, and we recommend always using Nodal.
Figure 6.4.7 shows the difference between Nodal and Triangular rendering modes.
122
Newton Discrete Element Modeling Software - Manual
3. Enter the start and end times for the wear summary in the text boxes in the Wear Summary
panel.
4. Click the Generate button. Newton will render the triangle/particle work in the rendering
window.
While the work summary is generating, you can change any of the settings in the Work tab of the SMV
window without disturbing the wear data. This means you can flip back and forth between work types
in the Work Type panel, change the min/max scales, change the work units, or the rendering mode.
Even though you can only view one type of work at a time, Newton keeps track of all work types
throughout the duration of the work summary (assuming you followed the steps above).
If, while a work summary is generating, you change a Layer Style to Work, the work summary will stop
and all work data will be reset. Similarly, if you set Particle Color Mode to Work while a work summary
is generating, the work summary will stop and all work data will be reset.
Notes about Work:
1. Newton calculates work in real-time, as the particles fall onto the triangles. So if you are
interested in the amount of work done on a layer from 0-10 seconds, you must set the slider bar
to 0 seconds, and then click the Play button to render the work. (Or you can use the Work
Summary panel, as described above). But if the slider bar is set to, say, 2.5 seconds, and then
you click Play, Newton will only render the work that is done from 2.5 seconds onward.
2. If you have generated some wear, and then you skip to a different part of the playback file and
you want to render the work there, you should click the Reset Work Data button to reset the
work values, or else Newton will continue to add the new work data to the values that were
already calculated.
If you forget to do this, Newton should display a text message in the rendering window that says
‘Warning: Work summation not reset!’
3. You cannot use Frame Skipping during playback or when creating an AVI if you are showing
particle or triangle work. This is because the work data is encoded into each frame of data. So if
frame skipping is enabled, then Newton will skip over frames and the work data from those
frames will not be added to the wear summary.
123
Newton Discrete Element Modeling Software - Manual
a. Newton reaches a key frame in which one of the Layer Styles is set to Work and the
previous key frame did not have that Layer Style set to Work.
b. Newton reaches a key frame in which the Particle Color Mode is set to Work and the
previous key frame did not have Particle Color Mode set to Work
c. Newton reaches a key frame where all layers are not set to Work and Particle Color
Mode is not set to Work.
124
Newton Discrete Element Modeling Software - Manual
125
Newton Discrete Element Modeling Software - Manual
126
Newton Discrete Element Modeling Software - Manual
127
Newton Discrete Element Modeling Software - Manual
force applied to any single triangle. Note that the total layer force is taken as the sum for all layers that
are currently showing force. Therefore you only want to set one Layer Style to Force at a time.
The position of this line of text can be set in the AVI Creator window (see Section 6.4.1).
Triangle Rendering Mode: This option determines how Newton will render triangle work. If Triangular
is selected, then Newton will add up the force from each particle according to the triangle that it acts
upon. All the force from that particle will be attributed to that triangle.
Each triangle contains 3 vertices; these are the A, B, and C ‘points’ of the triangle. Each vertex is shared
by multiple triangles, so it is called a ‘node’ rather than a vertex. If Nodal is selected, then Newton will
take the force applied to a triangle, and divide that force between the 3 nodes of that triangle. The
force on other triangles attached to the same nodes is also added to those nodes. Nodal rendering
mode is more accurate than Triangular, and we recommend always using Nodal.
Force Summary: This panel contains controls for generating a work summary.
1. Set the Layer Style to Force in the lower section of the SMV window for each layer for which
you want to show triangle forces. It is best to only view one layer at a time.
2. Enter the start and end times for the force summary in the text boxes in the Force Summary
panel.
3. Click the Generate button. Newton will render the triangle forces in the rendering window.
While the force summary is generating, you can change any of the settings in the Force tab of the SMV
window without disturbing the data. This means you can flip back and forth between force types in the
Force Type panel, change the min/max scales, change the filtering options, or the rendering mode. Even
though you can only view one force type at a time, Newton keeps track of both force types throughout
the duration of the force summary.
If, while a force summary is generating, you change a Layer Style to Force, the force summary will stop
and all work data will be reset.
Notes about Forces:
1. Newton calculates force in real-time, as the particles fall onto the triangles.
2. The force data is different from the work data discussed in Section 6.4.5. Whereas work data is a
summation of work over time, the force data is not a summation, so you can skip around to
different parts of the playback file without seeing a warning that alerts you to reset the force
data.
However, if you have filtering enabled, the filter depends on the past history of the forces, so
when you skip around in the file, the filtering will become inaccurate if the force data is not
reset.
128
Newton Discrete Element Modeling Software - Manual
Each of the force settings described in this section is saved with each key frame, with the exception of
the inputs in the Force Summary (the force summary is only for quickly generating a summary outside of
an animation). So at different points of an animation, you can show different force types or min/max
scales.
Notes about Forces in an AVI File:
1. When generating forces in an animation, Newton will reset the force data if any of the following
occur:
a. Newton reaches a key frame in which one of the Layer Styles is set to Force and the
previous key frame did not have that Layer Style set to Force.
b. Newton reaches a key frame where all layers are not set to Force.
129
Newton Discrete Element Modeling Software - Manual
130
Newton Discrete Element Modeling Software - Manual
Axis Control: Newton automatically sets the X and Y axes on the chart as the data is plotted, but you
can override the minimum and maximum values for either axis.
Data Labels: In the top right corner of the Plotting window are the data labels corresponding to each
data box. As the data is plotted, the labels list the following three things:
1. Total number of clusters that have entered the data box.
2. Total mass of material that has entered the data box.
3. Instantaneous rate of tonnage flow through the data box.
From top to bottom, the data labels represent data boxes 1 through 5, respectively. The colors on the
plot correspond to the data labels/boxes of the same color.
Saving Data from the Graph: If you right-click on the graph, you can copy the graph image, or you can
copy the data arrays for all 5 data boxes to the Windows clipboard.
131
Newton Discrete Element Modeling Software - Manual
Use the first column of combo boxes to select the layers from which you want to plot the work. Use the
second column of combo boxes to select the work type to plot for each line. To obtain meaningful work
data, the surfaces for which you record wear should be meshed up into small triangles that have an
average side length of 50-200 mm (see Section 4.3.6).
Based on the work type that is shown (Total vs. Instantaneous) and the Work Units that are selected
using the radio button, Newton will plot the following data for each line:
- Total Work (units in Joules): Plots the cumulative work performed on the entire layer over the
specified time interval.
- Total Work (units in Joules/sec): Plots the cumulative work performed on the entire layer
divided by the time elapsed (the average work).
- Instantaneous Work (units in Joules): Plots the discrete work performed on the entire layer for
each playback frame in the specified interval. The sum of these values would equal the
cumulative work performed on the entire layer over the specified interval (i.e. the sum would
equal Total Work, units in Joules).
- Instantaneous Work (Units in Joules/sec): Plots the discrete work performed on the entire layer
for each frame of the playback file divided by the frame time interval (i.e. this is the work that
would be performed if the rate of work for each data frame was constant for 1 entire second).
Saving Data from the Graph: If you right-click on the graph, you can copy the graph image, or you can
copy the data arrays for all 5 lines to the Windows clipboard.
132
Newton Discrete Element Modeling Software - Manual
Figure 6.5.3 – Plotting window with the Triangle Forces tab open
In order to plot triangle forces, you must have set Save Triangle Force & Work Data to YES in the
general simulation parameters window before running the simulation. If you did not, the controls on
this tab will be disabled and there will be a red label stating that triangle forces were not saved.
If a layer has a surface velocity, then Newton tracks the forces on the layer in the parallel direction
(aligned with the velocity vector), and in the transverse direction (perpendicular to the velocity vector,
across the triangle).
This feature is particularly useful for analyzing material centering on the receiving belt. If there are
significant transverse forces on the receiving belt layer, this may indicate poor material centering and
possible tracking problems with the belt.
- These forces are ONLY recorded if the layer has a surface velocity.
- This is a new feature since v2.12. So for playback files older than v2.12, Newton did not record
values for the parallel and perpendicular forces on any layer.
Preparing a Layer to Record Triangle Forces
In order to ensure that the forces are recorded properly, we HIGHLY recommend that you redraw the
belt in your CAD program using just three 3D faces to represent the top face of the belt (or five 3D faces
for a 5-roll belt). This will ensure that there are no duplicate belt triangles and that the normal vector for
each triangle is calculated properly by Newton. Often, when 3D solid belt geometries are exported to
another CAD format, the exporting method does a poor job ensuring that the belt is replicated properly
133
Newton Discrete Element Modeling Software - Manual
with correct-facing normal vectors. Then, after importing the belt, you should use the Equalize Triangles
and Divide Triangles buttons on the Geometry page to mesh the belt into small triangles with average
side lengths of ~50-150mm. Next, right-click in the rendering window and turn on Triangle Normal
Vectors. Verify that all the normal vectors on the belt are facing upward and inward, toward the center
of gravity of the belt (see Section 4.3.6).
Viewing Triangle Force Data
For triangles that have a surface velocity, Newton records the average parallel and perpendicular forces
applied to the triangle for each frame of data. The parallel force is the net force due to any contacting
particles acting on the triangle in the direction of the triangle's velocity vector. The perpendicular force
is the net force due to any contacting particles acting on the triangle in a direction perpendicular to the
triangle's velocity vector (i.e. across the belt).
When viewing parallel belt forces, a negative force indicates that the particles are exhibiting forces on
the belt in a direction opposite of the belt velocity. A positive force indicates that the particles are
exhibiting forces on the belt in the same direction as the belt velocity.
When viewing perpendicular belt forces, the direction of the force (i.e. positive or negative) depends on
the normal vector and velocity vector of each triangle. If you position the camera view so that you are
observing the belt from a top view, with the positive belt velocity pointing toward the top of the screen,
then a negative perpendicular force points to the left, and a positive perpendicular force points to the
right. So a negative force indicates that the material is pushing the belt to the left, and a positive force
indicates that the material is pushing the belt to the right. This assumes that the normal vector of each
triangle on the belt points upward and inward, toward the center of gravity of the belt, as was discussed
above.
Use the first column of dropdown boxes to select the layers for which you would like to plot the parallel
or perpendicular forces, and use the second column of dropdown boxes to select the type of force to
plot for the layer.
Saving Data from the Graph: If you right-click on the graph, you can copy the graph image, or you can
copy the data arrays for all 5 lines to the Windows clipboard.
134
Newton Discrete Element Modeling Software - Manual
7 Reference Papers
1
Kruse, D., Lemmon R. (2005), “Using the Discrete Element Method as an Everyday Design Tool”, Bulk
Solids Handling, 6/2005 pg. 358-367
2
David J. Kruse (2000), Applications and Advances of the Discrete Element Method in the Material
Handling Industry, Bulk Material Handling by Conveyor Belts III – 2000 SME Conference
3
Xiangjun Qiu & David J. Kruse (1997), Analysis of Flow of Ore Materials in a Conveyor Transfer Chute
Using the Discrete Element Method, Fourth U.S. National Congress on Computational Mechanics.
135