Inelastic Collapse in Configruation File

[Alche]

Hi Marcus ( You are Marcus right?)

Before I ask you, I appreciate for your program dynamod . it is really powerful tools for my work.

However, I have a big problem.. sorry about that..

I want to know a dynamics when particles drop on Hexagonal base plate.

first I made a base plate with Hexagonal structure using Matlab ( I set the particle coordinate) and changed -m 23 Funnel config file because I want to use the haxagonal base like funnels in -m 23.

So, I wrote "Static" every each particle. After that, I make a config file -m 0 -C 4 to get a particle positions and the particles will be dropped on that hexagonal base.

I put interactions (base , dropped particles and between them.) of course, I have 4 walls (x and z direction) and gravitation.

The problem is here.

when I start this simulation. it seems to be good but it was stopped suddenly even if the particle is flying ! or they are stopped all at once and the event count doesn't increase at all.

If I don't play the movie and just do it like

dynarun tmp1.xml -c 1000000 -o tmp2.xml

dynarun tmp1.xml -c 100000000 -o tmp3.xml

the results (the position ) are exactly same. (the flying particle is still flying)

Can you guess what the problem is? I will attach the config file..

Next question is I tried to make the dropped particle size lager. ( the base particle size was 1) , it looks like working but when the dropped particle size is 1.42 , it didn't work anymore.

Is there any rule for set the ratio between large particle and small one in binary system?

I'm really sorry I have a lot of questions.

I'll ask one more thing.

Could you let me know what is Global, Scheduler Type and Neighbourhoodrange ? what are they working?

Also I want to know when I try to do simulation using stepped potential,

the simulation does not finish even if a particle has velocity and levitated, it doesn't go on.

Any event count does not increase ..

Thank you

[toastedcrumpets]

Hi, It is Marcus. You have a lot of questions in your bug report, I've cleaned it up to the following issues:

1. My configuration file isn't working as it halts after a certain time.
   • The problem here is that your configuration is inelastic and with gravity falling onto a plate. Slowly the particles loose all of their energy and begin to bounce more and more with the bottom plate. This is known as inelastic collapse (see any relevant literature paper, e.g., http://arxiv.org/abs/cond-mat/9810009). I've made a video of your configuration file to see whats going on
   • DynamO does not have any method for dealing with inelastic collapse automatically enabled, so your simulation will eventually be stopped by this. Do you have a particular method for overcoming inelastic collapse in mind?
2. By increasing the particle size I eventually find that interactions are missed!
   • The problem here might be that you're not increasing the neighbourlist NeighbourhoodRange correctly, can you send me a configuration file that shows this problem with a size of 1.42?
3. Is there any rule restricting the diameter ratio between large particles and small ones in a binary system?
   • Nope, there is no special considerations for values of the diameters (or for any property in DynamO) that I'm aware of.
4. Could you let me know what Global, Scheduler Type and Neighbourhoodrange mean in the configuration file?
   • I've tried to improve the documentation (tutorial 3) on the website so that these properties are clearer. Please let me know if you're still having difficulties.

I'll close the issue as I don't think there's a bug here; however, feel free to reply with any other questions and I'll reopen the bug report if it turns out to be a software issue. Marcus

[Alche]

Hi Marcus.

I really thank you for your helping me.

Although I have not found any solution about my problems, now I know what I was wrong. I'm working on it to solve them.

I feel really sorry to you but I have few more questions T.T

These days, I reserch what is the different between Hard Sphere model and Stepped Potential (L - J potential) to viscosity.

First, What is the unit of the viscosity which I got from Plug in (-L KEnergy -L Viscosity).

I get viscosities from the plug in add ons using Green Kubo equation( I read the viscosity is the slope of off-diagonal matrix on your homepage),

and I have real viscositis about mono molecular system in our lab.

The results are almost same but the order parameter is 1e-2! If I make the result that is from MD simulation smaller than 100times, they are almost same.

and the unit of real viscosity is Ps*s. I don't think it is not accident but I'm not sure because I do not know the unit.

So if you know or have some information about that, can you let me know?

Second, I want to know how the plug in (-L ViscosityE) probes the pressures. Does it depend on the diameter of the system?

I usually think the viscosity of L - J potential is bigger than hard sphere model but the result was different. Hardsphrer model's viscosity was almost 10 times bigger than L J potential.

So I guess the problem is come from a diameter because when I make a config file using the stepped potential, Dynamod sets the diameter simultaneously according to the first step.

(if I define the first step is 1, dynamod makes the diameter is 1) but the problem is if I want diameter is 1 and give L J potential, I should make first step is smaller than 1(0.8 or 0.7)

If I don't, and I make it 1, the potential hase all negative value. So I could not help making it 0.8.. That's why I want to know how Dynamod get pressure.

If it depends on the diameter, is there any way to fix the diameter is 1 ?

Third, Whan I insult stepped potential (don't care diameter) and e = 1(depth) and kbT = 1, the system should be stuck or clustering each others where distance is 1 . Is that right?

If it is right, my system doesn't do that even if it is under square well system.

I confirm that thermostatic Temperature is 1 and total momentum is zero, ( dynamod -m1 -T 1 -Z -r 1) However after simulationing them few second (under 10e3 event count) they don't think make a cluster

and fly feely. Moreover, the total momentum became bigger. Is there any problems that I make?

I feel sorry make you bothering..

[toastedcrumpets]

Hi again, You shouldn't use the KEnergy or Viscosity plugins any more. They've been replaced and are automatically enabled now.

To calculate the transport properties, just run a simulation to collect the output file, then run dynamo/bin/dynatransport output.xml.bz2 This will give you the viscosity and thermal conductivity on the command line. Please note, that this will use the full correlator to calculate these values, and this usually gives huge error values. I recommend that you read the book by Haile called "Molecular Dynamics Simulations" to find out what the units are of the output.

I'm not really sure I understand your other problems. I can say that there is a stepped Lennard-Jones potential available in DynamO, and that you can create it using dynamod mode 16. dynamod -m 16 Marcus

[Alche]

Thank you for your answer.

It was really helpful for me Thanks a lot.

And...

Can I ask you something ? to run Dynamod?

When I install the Dynamod, I have lots of problems if I use radeon graphic card...

If I use Nvidia graphic cars, the problem doesn't show up.

Moreover I want to buy a Mac pro and it has two graphic cards, one is intel HD and Nvidia graphic.

Do you know anyone runing Dynamod in Macpro?

If you know, I would really happy :)

Thanks

1. 10. 23. ¿ÀÈÄ 6:17¿¡ "Marcus Bannerman" notifications@github.com´ÔÀÌ ÀÛ¼º:

Hi again, You shouldn't use the KEnergy or Viscosity plugins any more. They've been replaced and are automatically enabled now.

To calculate the transport properties, just run a simulation to collect the output file, then run dynamo/bin/dynatransport output.xml.bz2 This will give you the viscosity and thermal conductivity on the command line. Please note, that this will use the full correlator to calculate these values, and this usually gives huge error values. I recommend that you read the book by Haile called "Molecular Dynamics Simulations" to find out what the units are of the output.

I'm not really sure I understand your other problems. I can say that there is a stepped Lennard-Jones potential available in DynamO, and that you can create it using dynamod mode 16. dynamod -m 16 Marcus

Reply to this email directly or view it on GitHubhttps://github.com/toastedcrumpets/DynamO/issues/13#issuecomment-9695441.

[toastedcrumpets]

Hi again, You're right, DynamO used to crash a lot on Ubuntu systems with Radeon graphics cards. It was due to a race condition between the rendering and simulation threads.

These crashes have been fixed but only in the development for DynamO 1.4. You can try using the master branch of DynamO from git and you should find it doesn't crash any more.

At the moment, DynamO will not compile on mac's. The problem is that I don't have a mac to compile and test it on. As far as I know, no-one has gotten it working on a Mac yet! Marcus

[Alche]

Thanks you.

It was bad news T.T

I just wanted mac but .....Ok T.T

Thank you again !

On Tue, Oct 30, 2012 at 5:40 PM, Marcus Bannerman notifications@github.comwrote:

Hi again, You're right, DynamO used to crash a lot on Ubuntu systems with Radeon graphics cards. It was due to a race condition between the rendering and simulation threads.

These crashes have been fixed but only in the development for DynamO 1.4. You can try using the master branch of DynamO from git and you should find it doesn't crash any more.

At the moment, DynamO will not compile on mac's. The problem is that I don't have a mac to compile and test it on. As far as I know, no-one has gotten it working on a Mac yet! Marcus

— Reply to this email directly or view it on GitHubhttps://github.com/toastedcrumpets/DynamO/issues/13#issuecomment-9898055.

[Alche]

Hi : -).

Can i ask something?

I want to uninstall dynamod but i do not know how to do that..

Could you let me know?

Thanks a lot

[toastedcrumpets]

If you installed dynamo using the ubuntu packages, you just need to run: sudo apt-get remove dynamo-tools dynamo-core dynamo-visualisation

If you installed it by compiling yourself, just remove the following files /usr/bin/dynamod /usr/bin/dynarun /usr/bin/dynahist_rw /usr/bin/dynatransport /usr/bin/dynavis

[Alche]

Thanks Mr. Marcus.

Now I'm installing Dynamod. Everything was fine until I got problem..

When I do

cd DynamO

make

it shows me like

• DynamO: Boost headers : yes
  • DynamO: Boost filesystem library : yes
  • DynamO: Boost program options library : yes
  • DynamO: Boost iostreams library : yes
  • Magnet: OpenCL libraries and headers (optional) : yes
  • Magnet: libavcodec (video encoding support) : yes
  • Coil: Gtkmm : yes
  • Coil: GLEW (v1.6+) : yes
  • Coil: GLUT : yes
  • Coil: libCwiid Wii-mote support (Optional) : yes

  • DynamO-Coil Integration : yes ...patience... ...found 1599 targets... [root@WonhyukJo DynamO]# ^C

it looks like good

So I made a config-file like

dynamod -m 1 -o tmp1.xml

and I did that

dynarun tmp1.xml -V

but it does not work!

---

Main(): unknown option -V Main(): Try using the debugging executable for more information on the

error.

Of course, I did debug but it didn't change anything..

That's why I asked you how to uninstall Dynamod.

Moreover if I put [dynarun --help]

There are no option about -V(Visualizer) !

Do you know why it is happened?

I 'm really sorry bothering you again.

On Mon, Nov 26, 2012 at 3:04 AM, Marcus Bannerman notifications@github.comwrote:

If you installed dynamo using the ubuntu packages, you just need to run: sudo apt-get remove dynamo-tools dynamo-core dynamo-visualisation

If you installed it by compiling yourself, just remove the following files /usr/bin/dynamod /usr/bin/dynarun /usr/bin/dynahist_rw /usr/bin/dynatransport /usr/bin/dynavis

— Reply to this email directly or view it on GitHubhttps://github.com/toastedcrumpets/DynamO/issues/13#issuecomment-10707386.

[toastedcrumpets]

Hi, Ah now I understand your problem. The new version of dynamo has the dynavis program instead of dynarun -V. I'm in the process of rewriting the documentation before the dynamo 1.4 release.

Try writing dynavis tmp1.xml instead of dynarun tmp1.xml -V In a month or so the new version of DynamO should be out and all of the documentation will have been updated. Hope this helps, Marcus

[Alche]

Oh..my god Mr. Marcus

Thanks a lot!

It works!!

2012. 11. 26. ¿ÀÈÄ 6:43¿¡ "Marcus Bannerman" notifications@github.com´ÔÀÌ ÀÛ¼º:

Hi, Ah now I understand your problem. The new version of dynamo has the dynavis program instead of dynarun -V. I'm in the process of rewriting the documentation before the dynamo 1.4 release.

Try writing dynavis tmp1.xml instead of dynarun tmp1.xml -V

Reply to this email directly or view it on GitHubhttps://github.com/toastedcrumpets/DynamO/issues/13#issuecomment-10709516.