Home · DEM Course · Module 4.5
DEM (Discrete Element Method) Course with BrixDEM

Module 4.5 · Full case: drum with lifters

🎬 7 lessons ⏱ 2 h course 📝 5 questions · pass with 4
🔒

This module is still locked

To unlock it, pass the Module 4 quiz with at least 70%.

Go to the quiz · Module 4
Lesson 1 of 7 · 0:28

The case and the question

Lesson transcript

The silo taught the full workflow. Now let's repeat it on a very different piece of equipment: a rotating drum with lifters, like in a mill. The geometry comes ready-made from the Sketch. The part rotates. And the answer I want is an engineering number: shaft torque and power. Every simulation starts with a question. What torque and power does the shaft need? And what changes if the speed drops by half? Once that is defined, it is clear what to measure.

Quiz · Module 4.5

Check what you learned

5 multiple-choice questions. Get at least 4 right to complete the module. You can try as many times as you like.

✓ You have already passed this module. You may retake the quiz to review.
1.How is the drum's critical speed calculated?
2.The drum was drawn with Z up, but default gravity points to negative Y. What happens if nobody fixes it?
3.Instant fill was asked for 2000 particles, but about 1400 entered. Why?
4.When rotation was reduced from 4 to 2 rad/s, what happened to torque and power?
5.Which quick check showed that the drum force measurement made sense?
Transcripts of every lesson in this module

1. The case and the question

The silo taught the full workflow. Now let's repeat it on a very different piece of equipment: a rotating drum with lifters, like in a mill. The geometry comes ready-made from the Sketch. The part rotates. And the answer I want is an engineering number: shaft torque and power. Every simulation starts with a question. What torque and power does the shaft need? And what changes if the speed drops by half? Once that is defined, it is clear what to measure.

2. Geometry in the Sketch

New project, clean screen. I am going to build a drum with lifters from scratch. I open the Sketch. In the Advanced tab, under Rotary drums, I pick Flights, the drum with lifters. A form opens with the dimensions. Length is the length: twenty centimeters. Outer diameter is the diameter: half a meter. Wall thickness is the shell thickness: one centimeter. Flight height is the height of each lifter: four centimeters. Base thickness is the lifter thickness: one centimeter. And Flight count is the number of lifters: eight. The taller the lifter, the more material it raises. I click Create. The part shows up in the model tree. In the File and History tab, Send sends the drum to the main window. Done, the drum is in BrixDEM. One important detail: the Sketch creates the drum axis at the origin, along X. So the rotation center will simply be zero, zero, zero.

3. Domain, materials and interactions

Now the domain, with one click on Fit to Geometry. The program leaves a small margin around the drum. In Materials, the plus button offers two types: Particle material and Wall material. I create the particle first. I name it Minerio. Radius of eight millimeters. Density of two thousand seven hundred. And Young's modulus of ten to the eighth pascals, well below the real value, so the time step does not get too small. That value is not measured data: it is a numerical choice. Now the wall material, Aco. Notice that the part is already linked to this material, because it is the only wall material. In the interactions, the first pair is Minerio with Minerio. I choose Hertz-Mindlin, restitution zero point five and friction zero point five. In a real case, these numbers would come from calibration. I also switch on rolling friction, with the CDT model and a value of zero point one, and save the interaction. Now the Auto button. It creates the missing pairs, with default values. Minerio with Aco appears. I select that pair, click Edit, switch on rolling as well, and save.

4. Drum motion

Now the motion. First, in Properties, I check Measure force and torque. That is what will give me the shaft power. And I click Apply Properties. In Motion, I choose Rotational. The drum turns around X. So the vector is four, zero, zero: four radians per second, about thirty-eight revolutions per minute. The center stays at zero, zero, zero. How does that compare with the critical speed? The critical speed is the square root of g over R, here about six radians per second. We are at two thirds of it: the material rides up on the lifters and falls in a cascade. In Motion program, I switch on the segments. I will keep a single segment, with a ramp, so the drum speeds up slowly. Then OK, and Apply Motion to save it.

5. Filling and setup

Now the filling. In Insertion, I pick the Cylinder shape, which follows the drum axis. The cylinder goes from minus eight to plus eight centimeters in X, with a radius of seventeen centimeters. Careful here: it must not touch the lifters. A particle born inside the wall gets a huge force, and destabilizes the case. With Instant fill checked, the drum fills all at once, at the start, and Particle Flow becomes the number of particles: two thousand. The program prevents them from being born overlapping. So check afterwards how many actually came in. And don't forget: Add. Now the inlet exists. In the Setup, a new project comes with a fixed step. End time of five seconds, and in dt method I switch to Automatic, by Rayleigh, with a factor of zero point two. In Save frequency, Adaptive. Calculate Time Step Now confirms the step. Watch the gravity. The default points to negative Y, but my drum was drawn with Z up. So zero in Y, and minus nine point eight one in Z. This mistake is easy to make, and nothing warns you: the material simply falls the wrong way. In Thread Count I put four, and Verlet is already on. Finally, I save the project under another name, and everything is ready.

6. Running and looking inside

Before running, one detail: the drum is closed, and you cannot see the particles. In the view settings button, I lower the surface opacity and turn on the edges. Save and apply, and I can already see inside. Now, Run. First the drum fills. Look at the console: N shows how many particles came in. I asked for two thousand, and about one thousand four hundred came in. That is the program preventing them from being born overlapping. That is why you have to check. The particles fall and settle at the bottom of the drum. Meanwhile, the ramp brings the speed up to its final value. Here, the whole run takes about two minutes. It finished in a little over two minutes. I go back to the first frame and click Play, to see inside the drum. Look at what happens. At first, the whole bed rides up stuck to the wall, like a block. Then the lifters carry the material up, and it collapses in a cascade. That is exactly the regime you want in a mill. To read it better, I color by velocity. I open the palette, choose Velocity Magnitude, and limit the scale from zero to two meters per second. That way, colors are comparable between frames. The fast particles are in the cascade. The slow ones are in the block that rides up with the wall. The spot where they land at the bottom is where the wall will take the impact. Now the numbers. In Results, first the kinetic energy. It rises and levels off at about five joules. That plateau shows the drum has reached steady state. The overlap. An average of zero point one six percent of the radius, and a maximum close to two point six percent. The chosen stiffness is adequate. The maximum has a six percent peak at the start, during filling. Keep an eye on that kind of peak. And the power. Under By Geometry, I click Mechanical Power. The average torque is about seven point nine newton meters, and the power is thirty-one watts. The negative sign only indicates direction. Use the magnitude. Compute the average in steady state, not at the start-up peak. A quick check: the average vertical force on the drum is eighty newtons. The one thousand four hundred thirty-two particles weigh about eight point three kilograms, that is, eighty-one newtons. The numbers add up. Always do this kind of check.

7. Changing the speed

Now the engineering question: what happens if the speed drops by half? I clear the results, go back to Motion, and change it to two radians per second. I save, and run again. At one third of the critical speed, the material still rides up, but much less. The cascade is short, and little is thrown into the air. Compare with the previous speed. Finished. Now it is just a matter of comparing the numbers from the two cases. Here is the result. At four radians per second, the torque is seven point nine newton meters, and the power is thirty-one watts. At two radians per second, the torque rises to ten point seven, and the power drops to twenty-one watts. And the kinetic energy drops from five to one point five joules. The surprise is the torque: it went up at the lower speed. The fast drum throws material into the air, and that material no longer weighs on the wall. On the same part, power and torque can move in opposite directions. Three precautions. These parameters were not calibrated. The numbers show the trend, and Module 5 shows how to get reliable values. Always compare in the steady-state window. And check mass, energy and overlap before believing any result.