Prev: optical flow using block matching
Next: predci
From: Andrew on 28 Nov 2009 13:59
On Nov 28, 7:48 am, "Kiril " <kkirq...(a)gmail.com> wrote:
> Yes i understand now. And my last question is from there i can find this model or may be some help to start makinh it by myself in simulink.??

I don't understand. A fault in the rotor? This seems to be an
unlikely occurrence. How much overcurrent in the rotor would cause a
fault? And what would be the corresponding stator current? Wouldn't
the stator be more likely to fail? Locked rotor current in a typical
stator is about 600% FLA. This would correspond to the greatest rotor
current, at a slip of 100%. That current can be calculated. How many
seconds at 600% stator current will the stator windings last?
Probably not more that 180 seconds, even for a very large motor. Is
that enough time to damage the rotor? I am assuming that there is no
overcurrent protection in this circuit?
From: Frank W. on 28 Nov 2009 16:12
Andrew <andrewkgentile(a)gmail.com> wrote in message <0e0a6017-812c-4c39-ad00-ddb7f54ce4bc(a)u1g2000pre.googlegroups.com>...
> On Nov 28, 7:48?am, "Kiril " <kkirq...(a)gmail.com> wrote:
> > Yes i understand now. And my last question is from there i can find this model or may be some help to start makinh it by myself in simulink.??

Hi,
allow me some time to have a look in some of my collected sheets. I do not have a model at hand but I might give you some literature hints. I once had to deal with digital electronic that was supposed to detect the fault you mentioned - I was glad not to have to deal with the machine basics. because the majority of the texts were german I do not know if I find English literature about it.
I will contact you, Kiril, under your personal mail adress.

>
> I don't understand. A fault in the rotor? This seems to be an
> unlikely occurrence. How much overcurrent in the rotor would cause a
> fault? And what would be the corresponding stator current? Wouldn't
> the stator be more likely to fail? Locked rotor current in a typical
> stator is about 600% FLA. This would correspond to the greatest rotor
> current, at a slip of 100%. That current can be calculated. How many
> seconds at 600% stator current will the stator windings last?
> Probably not more that 180 seconds, even for a very large motor. Is
> that enough time to damage the rotor? I am assuming that there is no
> overcurrent protection in this circuit?
Yep,
such problems in the rotor are unlikely and nobody would like to detect such a fault in a 400VAC standard motor. But it is different with bigger and sometimes specially designed machines with high reliability. Generally, what you mentioned is true. The main problem is that a broken bar in the rotor for instance will lead to a inverse field which might only have slight effects - little more vibrations, little more oscillating torque which subsequently and slowly might damages the machine connected to the motor and reduce its life. In such, admittedly very rare cases, some customers want to have such rotor faults detected to allow early remedy. And be sure you will not recognise a broken bar by measuring the starting currrent. Of course, in the worst case when you have a loss of one phase of the rotor completely the machine will not accelerate but only draw the mentioned current besides
humming angrily. Then its too late and thats what some customers want to avoid and that maight be the reason that Kiril got his task.
Regards,
Frank
From: Kiril on 28 Nov 2009 16:56
> (1)The main problem is that a broken bar in the rotor for instance will lead to a inverse field which might only have slight effects - little more vibrations, little more oscillating torque which subsequently and slowly might damages the machine connected to the motor and reduce its life.
> (2) And be sure you will not recognise a broken bar by measuring the starting currrent.
Yes Frank u are right. If there is a broken bar there are more vibration and so on. And about second - it true too and it's my main objective. I want to detect broken bar only by measure stator current (with hall effect sensors). I try different kind of methods. For example MCSA ( motor current signature analysis), wavelet analysis, to use Park vector trajectory. And use neural network (radial) to find broken bar. This is my second project in appliction of AI in motor drives. I try to use AI to make life of AC longer. And i need some kind of tool to simulate broken bars in AC motor. I found only one article - about simulation in Modelica. I will be vary glad if u can send me some about broken bar simulation. If u interested of MCSA (using it to detect broken bar), write and i will send u some articles.
First  |  Prev  | 
Pages: 1 2
Prev: optical flow using block matching
Next: predci