EDEM Particle Size Distribution

Hi Ibrahim,

In the Creator tab, under Particle>Size Distribution, you can specify any size distribution you want using the user defined option (see image below). Just make sure the the mass percentages add up to 100%.

Best regards,

Renan

I have created 30 different particles shapes under one bulk material as shown. I would like to generate 20,000 particles and have each particle shape be randomly distributed among the 20,000 particles. Just like a hand full of sand. Do I just simply select random for each particle as shown below? Also, what is the recommendation for the min and max? I just want to double check my selections before running simulations because they take a long time. 

Thank you for your time and consideration

I have created 30 different particles shapes under one bulk material as shown. I would like to generate 20,000 particles and have each particle shape be randomly distributed among the 20,000 particles. Just like a hand full of sand. Do I just simply select random for each particle as shown below? Also, what is the recommendation for the min and max? I just want to double check my selections before running simulations because they take a long time. 

Thank you for your time and consideration

Hi Ibrahim,

With regard to size and run time if you scale by 'radius' the time-step has a linear scale with this.  So scale by 0.5 means simulation time-step with be reduced by 0.5 and simulations will take 2x longer, if you had the same number of particles.

However reducing size also typically increases the number of particles if you simulate the same volume, so you need to consider the total number of particles also.  We have a guide on estimating run time here - https://community.altair.com/community/en/estimating-edem-simulation-run-time?id=kb_article_view&sysparm_article=KB0116101

This distribution you show gives you a random distribution of sizes for that specific shape, there isn't a way in EDEM currently to automatically vary the shape, you have to specify the shapes required manually (as you seem to have 10 shapes shown in the screenshot) then vary the sizes of each specific shape.  

Regards

Stephen

Hi Ibrahim,

With regard to size and run time if you scale by 'radius' the time-step has a linear scale with this.  So scale by 0.5 means simulation time-step with be reduced by 0.5 and simulations will take 2x longer, if you had the same number of particles.

However reducing size also typically increases the number of particles if you simulate the same volume, so you need to consider the total number of particles also.  We have a guide on estimating run time here - https://community.altair.com/community/en/estimating-edem-simulation-run-time?id=kb_article_view&sysparm_article=KB0116101

This distribution you show gives you a random distribution of sizes for that specific shape, there isn't a way in EDEM currently to automatically vary the shape, you have to specify the shapes required manually (as you seem to have 10 shapes shown in the screenshot) then vary the sizes of each specific shape.  

Regards

Stephen

Thank you again for the help. So, just to double check, I am wanting to validate a simulation. In the simulation, the author states that the particle sizes D(10), D(50), and D(90) are 24, 48, and 72 micrometers. I created three bulk materials, each containing a particle with sizes 24, 48, and 72 micrometers. Now, I assume that I am supposed to select the "lognormal" size distribution option for each particle within the bulk material. But, how would you specify the D(10), D(50), and D(90) for each particle? The total number of particles to generate is 10,000. Here is an attachment to a pdf with details. I assume that I just simply mess with the "Particle Ration" section below? 

Thank you for your time and consideration 

Hi Ibrahim

Is there a reason why you would want to use the lognormal distribution and not the normal distribution? With the given d10, d50 and d90 values, you could calculate the mean and standard deviation of the distribution, and then use those values for your normal distribution. You would only require one material and one particle, and specify the distribution for that particle, instead of having three separate materials and particles.

Thanks,
Jerrin Job