[gmx-users] time step in gromacs 3.3

[Qing Zhu]

Hi David

I tried to run 3.2.1 tpr with 3.3 using 3 fs as the time step, but the 
simulation still crashed after several thousand steps. I have no idea 
why there is huge differece between 3.2.1 tpr and 3.3 tpr, because all 
the initial conditions are the same except the gromacs version. Any 
suggestion?

Thanks
Qing Zhu

David wrote:

>On Fri, 2005-11-04 at 16:01 -0600, Qing Zhu wrote:
>  
>
>>Hi David,
>>
>>What I did is to run 3.2.1 tpr files with 3.2.1, and run 3.3 tpr files 
>>with 3.3.
>>    
>>
>Try 3.2.1 tpr with 3.3
>
>  
>
>>The results of gmxcheck have hundreds of pages, and I don't know what 
>>could I get from gmxcheck.
>>    
>>
>See whether the difference is in the tpr or in mdrun.
>If you have hundreds of page (pipe it through more) it may mean that you
>are using a different force field, different coordinate/velocities which
>means you can  not compare the simulations.
>
>
>  
>
>>Thanks
>>Qing Zhu
>>
>>
>>David wrote:
>>
>>    
>>
>>>On Fri, 2005-11-04 at 15:25 -0600, Qing Zhu wrote:
>>> 
>>>
>>>      
>>>
>>>>Hi, everyone
>>>>
>>>>I tried to run lipid membrane simulations by setting the time step to be 
>>>>3 fs using gromacs-3.3,  and my simulations always crashed after a short 
>>>>time unless I changed the time step back to 2 fs. But if I run the same 
>>>>lipid membrane simulations using gromacs-3.2.1, I can even use 4 fs as 
>>>>time step and the simulations runs fine. It is not a big problem for me, 
>>>>but I am just curious that why gromacs-3.3 is so critical about the time 
>>>>step. Any one has an idea?
>>>>   
>>>>
>>>>        
>>>>
>>>Did you try running your 3.2.1 tpr file with 3.3 ? 
>>>Did you compare 3.2.1 and 3.3 tpr files (using gmxcheck)? 
>>>
>>> 
>>>
>>>      
>>>
>>>>Thanks
>>>>Qing Zhu
>>>>_______________________________________________
>>>>gmx-users mailing list
>>>>gmx-users at gromacs.org
>>>>http://www.gromacs.org/mailman/listinfo/gmx-users
>>>>Please don't post (un)subscribe requests to the list. Use the 
>>>>www interface or send it to gmx-users-request at gromacs.org.
>>>>   
>>>>
>>>>        
>>>>
>>_______________________________________________
>>gmx-users mailing list
>>gmx-users at gromacs.org
>>http://www.gromacs.org/mailman/listinfo/gmx-users
>>Please don't post (un)subscribe requests to the list. Use the 
>>www interface or send it to gmx-users-request at gromacs.org.
>>    
>>