Lorentz Force In Capacitors

[js2020]

TL;DR Summary

What happens to a capacitor during short circuit due to Lorentz force?

I am searching online for resources regarding studies done on the effect of the Lorentz force due to short circuit faults in capacitors. Although a DC-link capacitor only sees the ripple, there would be high current during a fault. Since F=(qE + JxB), I am curious what the effects of the high di/dt (which generates B and causes high J) would be. Does it reduce the lifetime? Induce microfractures in ceramic capacitors? Cause delamination? I've seen several papers on the effect of short circuit faults in bus bars but can't seem to find anything for capacitors.

Any other applications where the reliability of parts have been studied due to the effect of the Lorentz force is also very helpful.

 

[tech99]

Capacitors are rated with a ripple current, which might limit damage due to mechanical forces and heating.

 

[js2020]

Capacitors are rated with a ripple current, which might limit damage due to mechanical forces and heating.

Yes, I understand that they have a ripple current rating which limits heating. Forces would also be low if we're only considering ripple. But if you short your DC-link capacitors, you can have a huge fault current. Depending on the application, it can be so large that it rips the bus bars off of whatever they're attached to. Special attention has to be paid to the connector design to ensure they survive large fault currents. I've found papers where they study this for the actual bus bars as well. I'm interested in what's happening in the capacitors. The same is studied in overhead transmission lines. Surely the "parallel plates" must have a forces exerted on them too.

 

[berkeman]

Surely the "parallel plates" must have a forces exerted on them too.

I would think it would be similar to the situation with parallel conducting wires carrying currents in opposite directions. Have you tried doing any vector calculations?

 

[anorlunda]

I²R heating, more than forces can destroy a capacitor.

Also, http://bentoros.com/wp-content/uplo...n-Industrial-and-Commercial-Power-Systems.pdf says this:

A switching interrupter, especially when switching a capacitor circuit, may be observed to restrike two, or perhaps three times before complete interruption is achieved.

Each restrike causes switching transients which in turn can further damage the capacitor. The linked IEEE standard requires short circuits in capacitors to be considered only in the context of the external circuit connected. In other words,
the question can't be isolated to just the capacitor.

 

[js2020]

I would think it would be similar to the situation with parallel conducting wires carrying currents in opposite directions. Have you tried doing any vector calculations?

I think its similar to parallel conducting wires too. I've seen papers where they do this type of analysis for transmission lines to determine spacing between then so they can ensure the phases don't short from swinging into each other. Of course faults aren't the only reason the lines can swing into each other, but there are papers that consider this case.

I can do the vector calculations to predict the forces, but that will not help me with determining things like how much force it takes to damage the capacitor. For instance, if there's a huge force outward from the outermost plates of a MLCC, can it damage the ceramic causing stress fractures which lead to other failures in the future. Or how much does it take to completely explode the capacitor.

 

[anorlunda]

Or how much does it take to completely explode the capacitor.

They do explode in certain circumstances. But I think the failures are more ascribed to overheating.

I recall a new installation of $2 million worth of capacitors at a scrap steel arc furnace plant in Veracruz Mexico. The furnace created so many transients and harmonics, that the whole city was having problems. The capacitors shunted the transients and harmonics to ground.

Each capacitor in the bank had its own fuse. After just a few months of service the whole bank of capacitors went BANG with no warning.

 

[js2020]

I²R heating, more than forces can destroy a capacitor.

I2R losses are not my concern. I understand that if the temperature is allowed to rise then it can destroy that capacitor. If the short circuit is short, the capacitor MAY not have time to overheat.

Each restrike causes switching transients which in turn can further damage the capacitor. The linked IEEE standard requires short circuits in capacitors to be considered only in the context of the external circuit connected. In other words,
the question can't be isolated to just the capacitor.

I appreciate the link. I'll read through it because I can always learn. There may not necessarily be restrikes, especially in the case of a solid state breaker. In the case of solid state breakers, it MAY be possible to interrupt a fault in less than one cycle. Also, it is possible to isolate just the capacitor in a lab experiment. You could charge a capacitor, remove the source, short the capacitor, and do this repeatedly to determine how the fault current effected it. Regardless of if the capacitor was isolated or not, it should still be effected by the same force in or out of the system (assuming the terminals didn't become damaged and cause it to fail by other mechanism). The system could be completely destroyed and you may still be able to measure if the capacitor was functional. If you could not decouple the system for design and analysis, it seems like it would be pretty hard to sell components that have a specific rating without analyzing it first in your full system.

The link you provide will provide me with more background information on the topic of short circuits in industrial power systems so I do appreciate that.

I was hoping for more experimental, design, or any type of articles/experience pertaining to individual components and short circuit analysis that has been performed on them in terms of how they are effected by the Lorentz (electromagnetic) force. I'm not interested in discussing how I2R losses may cause failure as well. Should I repost this under a mechanical forum? Maybe it does not fit in electrical like I thought it would.

 

[js2020]

They do explode in certain circumstances. But I think the failures are more ascribed to heating and thermal expansion.

I understand that they may be more prone to thermal failures. I'm not trying to determine what issue is most likely to cause a capacitor to fail. I'm interested in the very specific case of how short circuit faults effect them due to the Lorentz force their parallel plates are subject to.

 

[berkeman]

I did a Google search for failure mechanisms in capacitors from short circuits and got lots of good hits. Have you done such a search?

 

[js2020]

I did a Google search for failure mechanisms in capacitors from short circuits and got lots of good hits. Have you done such a search?

Yes I have. I searched something like "capacitor failure from Lorentz force", but nothing useful came up. A lot of stuff on capacitor failure mechanisms not related to the mechanical forces due to short circuit.

I just searched the same term you used and there are more hits for "short circuit" faults at least. I've looked through the first 10 pages and found one document that simply states "Thermal and electrical forces for support insulators Short circuit currents will produce high forces that act on the insulators. In rare cases, ultra-high forces can cause failure of the insulator. However there are a few methods that can limit the amount of fault current with help of PTC devices". The others are all referring to short circuits in the capacitors due to dielectric breakdown, water ingress in microcracks of MLCCs, etc. BUT, I'll look for "electromechanical capacitor faults" and see how that goes. Thanks for the suggestion

 

[tech99]

The electrical energy stored in the capacitor must be less than the maximum mechanical energy it is able to store.

 

[js2020]

The electrical energy stored in the capacitor must be less than the maximum mechanical energy it is able to store.

I think my question has to do with how quickly the electrical energy is released. Releasing it very rapidly (a short circuit) will create a high di/dt which creates a high B field. The Lorentz force is related to JxB so a very high B field (caused from the short circuit) will create a very strong force. The capacitor is obviously capable of storing the electrical and mechanical (whatever that is) energy during normal operation. Its about how quickly it's released. If there are never any short circuits, you would be fine. I'm lookimg for resources on studies that have been done on how they degrade (over time or rapidly) due to the electromechanical forces present specifically due to short circuit faults.

I may have misunderstood your answer because I'm not sure how to relate it to my question. I'm not referring to short circuit failures in the capacitor caused by overvoltage. Thats what would happen if you exceeded the energy storage capacity of the capacitor. I'm not referring to anything that would occur during normal steady state operation.

 

[Baluncore]

Fundamentally, a wire transmission line differs from a capacitor in that the wires are filamentary, while capacitor plates are wide areas of tape, wound on a common axis or fabricated as a multilayer stack.

The distance between capacitor plates is very much less than the width of the plate, and the parallel currents flow across the full width of the tape, so forces between the tapes are often cancelled.

I think you must first specify the type of capacitor construction, then analyse the forces by modelling, not by experimental measurement.

 

[js2020]

Fundamentally, a wire transmission line differs from a capacitor in that the wires are filamentary, while capacitor plates are wide areas of tape, wound on a common axis or fabricated as a multilayer stack.

The distance between capacitor plates is very much less than the width of the plate, and the parallel currents flow across the full width of the tape, so forces between the tapes are often cancelled.

Fundamentally, wouldn't they be the same? Wouldn't I use the same fundamental equations for Lorentz force? There may be some simplifications due to forces canceling, but I thought the fundamentals would still be the same. A MLCC capacitor is stacked whereas electrolytic and film would commonly be rolled. So maybe one would be more robust than the other to short circuit faults, but the physics behind the forces should still be the same.

I think you must first specify the type of capacitor construction, then analyse the forces by modelling, not by experimental measurement.

I will be modeling the forces. I was hoping to get more background and hopefully some pointers towards references to help with my literature survey because I haven't found much on this topic. My guess is that the capacitor manufacturers or researchers specializing in capacitors (dependent on the group) are more familiar with this. I just haven't found good resources yet.

My final application is not a capacitor but actually a PCB with planar layers. Current will flow in opposite directions and it may be subject to high short circuit currents. I'm interested in seeing how these high forces may affect the PCB integrity. I know that capacitor layers are even closer than PCB layers, but it seemed a lot closer in comparison than looking at bus bars and transmissions lines with even larger spacing and must different geometries. Again, I know the underlying physics will be the same. I'm just looking for resources dealing with this issue in general.

Thanks everyone for the discussion.

 

[Baluncore]

Fundamentally, wouldn't they be the same? Wouldn't I use the same fundamental equations for Lorentz force?

There is only one physics, but there are several topologies. A power transmission line tends to concentrate the currents into long and thin catenary wires. That maximises the forces on the wires.

My final application is not a capacitor but actually a PCB with planar layers. Current will flow in opposite directions and it may be subject to high short circuit currents. I'm interested in seeing how these high forces may affect the PCB integrity.

Long before a PCB track is forced off the surface by Lorentz, the track will fuse. Tracks are designed for a maximum current per width, so for higher current tracks, track bonding to the substrate will increase directly in proportion to current forces. Using wider tracks, or a wide ground plane, will distribute the current further, and so will reduce the multiplicative effect from both sides of the circuit.

The aspect ratio of a capacitor plate will decide the current density and so the resistive heating. The foil must be thick enough to have sufficient thermal capacity, but the resistance must limit the current to generate less heat than will melt the foil during an external short circuit, or dry out the dielectric when subjected to high ripple currents.

My guess is that the capacitor manufacturers or researchers specializing in capacitors (dependent on the group) are more familiar with this. I just haven't found good resources yet.

That design theory is well known by capacitor manufacturers. A manufacturer will analyse and optimise their own capacitors. It is most unlikely that a manufacturer will publish their advantageous IP.