Do the metal contacts of an electrical switch ever actually make contact with each other ?

Hello all,
Am I right in saying that when an electrical switch closes, and after it has completed the whole switch bounce / arcing routine, that it settles into a state where the metal contacts are considered to be closed, but that the two contacts are actually separated by an ultra thin (possibly at the molecular level) film of air?
If I use a mechanical analogy, where I have a film of oil between two metal plates. If I assume the metal plates are perfectly (theoretically) smooth, then, no matter how much force I put on the metal plates, I still can never hope to remove all of the oil from between the plates.  In other words, the metal plated are separated by a film of oil, albeit it a very thin one.     
Are mechanical switch contacts always separated by a film of air, which is a factor of contact resistance, and that this film is so thin that electrical currents  can pass through?


Any comments welcome,
thanks, Noel.

Interesting question! I would say that at surface level (or molecular level, if you like), the contacts themselves will appear uneven and pitted. There must be some degree of direct metal-to-metal contact taking place, in my view, but at microsocopic level it won't be over a uniform area due to the rough nature of both surfaces.Contacts are often springy, phosphor bronze 'self cleaning' and tend to wipe themselves clean of deposits (carbon, sooting etc) leaving a brighter, clean-looking (to the human eye) but very small area where contact resistance is lowest. A switch cleaner aerosol removes contaminant and reduces friction by filling in the roughness with a lube. I tend to think the contacts would force their way through that in order to make metal-to-metal contact.

I don't design switches but contact resistant is related more to the alloys, gold plating, etc. and mechanical design used in the switch contacts. I'm not so sure that air gets in the way of overall conductivity.

Alan W

I can't add much to what Alan has said.  I am sure there has to be some metal to metal contact for the switch to work.  I was given a contactor (read: power mains relay) recently and found all the contacts were covered by a very thin layer of black carbon gunk and were all reading a high resistance (from tens to thousands of ohms) when closed.  Even just wiping the gunk off with a finger improved things considerably, though I cleaned the contacts with a wire brush to restore them.  Contact cleaner works by physically displacing the gunk (being a liquid with initially high velocity and then subject to gravity) before it itself evaporates, though some of them do leave some residue (avoid them!).  Alan is correct in that some contacts are "self-cleaning" in that they are designed to rub against each other before coming to rest.  That also tends to remove the gunk.You may be thinking of hard disk heads?  They never actually contact the disk, but are designed to "float" a tiny distance (think microns) above the surface of the disk.  If they do contact the disk while it is moving they will normally be damaged and, more importantly, damage the magnetic coating on the disk and destroy the data.  This is termed a "hard disk crash".  One reason for going to Solid State Disks (SSDs). 
And similarly the reason why solid state switches are favoured in some applications - think Hall effect devices, opto isolators and the optically isolated power relays that you can get.

In the case of switches which have to carry relatively high current, you need to have low resistance between the contacts or you will have localized heating. In some cases, this could be enough to melt the metal which could cause the contacts to be permanently bonded together. This is most likely to happen as the contacts are closing if there is an arc generated. Most switches and relays have max current specs for make, break and carry with the carry current higher than either make or break.
The company I work for has hybrid relay circuits where a solid state device is used to shunt the current from the relay contacts as the relay is being closed or opened.  Does wonders for keeping the relay contacts from welding.

Maybe the structure of electrical contacts could be thought of as two half-spheres of sandpaper smashing together. Some of the grains of sand are bound to make direct contact with each other, so that's where the current will flow through.
At the other extreme of things,  about 20 years ago I wrote a short series about power generation traced from the gas- power stations through to the household mains socket, and in the power station contact arcing is a big problem. They use gas-filled circuit breakers to snuff out the arc.  They did have a 40,000 amp rating though...I put the article online with photos here

Even some smaller contacts use tricks to reduce arcing. I remember fitting a set of points to a car many years ago, where one contact was in the form of a ring rather than a disc. This was supposed to reduce arcing when the contacts opened, and so reduce wear too. I can't say there was a noticeable difference when it came to replacing them, compared to the conventionally designed plain disc contacts.

Thanks all for you comments.I am a mechanical engineer with more than a passing  interest in electronics and I have posted many questions in the EPE chatzone over the years.  On some occasions like this one, a question comes up in the mechanical engineering world that has a direct analogy in the electronics world. Imagine a closed cylinder with pressure, e.g a cylinder of butane or an aerosol.  Now imagine a metal object inside the cylinder that moves about freely, like the metal ball you frequently get in a pressurised paint aerosols, used to mix the contents, only for the purposes of this discussion, it would help to visualise the object as a flat object such as a small disk or cube inside the container, pressurised with air.So when the cube is at the center of the cylinder, the pressure acts equally on all surfaces, and the cube has no tendency to move in any particular direction (ignoring gravity).   If however, this cube was placed on the end of the cylinder (flat surface of cube in contact with flat surface of cylinder), both experience and theory tell me that the cube is still in equilibrium which means that the air pressure acts upon all of the cube surfaces including the one in contact with the cylinder.  Taking Alan W.'s theory in the case of an electrical switch that "there has to be some metal to metal contact", if that were the case with the cube and the cylinder then the face of the cube in contact with the cylinder would have a smaller area subjected to the air pressure, as some of this surface would be taken up with metal to metal contact. Now the forces acting to push the cube away from the cylinder end are lower (think smaller surface area)  and this would cause the cube to be held fast / securely to the end of the cylinder, but obviously this is not the case. Why ?
Noel Dillon

If you look up the definition of "cold welding", especially in regards to being in space, you would likely come to the conclusion that the two metal parts generally don't actually make contact. Except in a pure vacuum, as in space, there will be a thin layer of oxidation between the two metal surfaces. Certainly it's possible for rough edges to scrape through the oxidation, but if the whole surface was purely metal to metal contact, the two switch pieces might chemically/mechanically fuse.That being said, the two contacts, when reasonably clean, are effectively connected, just separated by a low resistance resistor. The dirtier the contacts, the greater the resistance in that "resistor."The two surfaces don't need to be in true mechanical contact. They just need to be close enough such that the distance and resistance is low enough for the electrons, at the given voltage, can move through whatever medium that the "resister is made of, be it air, grime, or oxidation.

Thanks Duane, that's kind off what I was thinking.  I'm thinking that, when when the metal contacts come together, with or without oxidation, they will still be separated by a layer of air  (ultra thin, possibly at molecular level) and this layer is so thin that it offers almost no resistance to the movement of electrons.If the current levels are so high through the switch that the contacts melt and fuse together - then obviously the situation is different.    My question is really about whether the metal contacts actually make metal to metal contact, with no current involved and assuming no dirt or grime on the contacts. 
Thanks again for the comments.

@Alan W.:  I forgot to say thanks for the power generation article - I have printed it off and an looking forward to reading it!.
Regards, Noel Dillon 

I have actually seen cases in which the contacts "welded" to each other  in small groups and at one point it had welded so solid that it broke the conecting rivet when the switch was moved to the off position.  Turned out there was a bad capacitor in the circuit causing a higher voltage and amprage through the switch..  Yet because of secondary filter the rest of the system was not affiliated. 

David Ashton just posted an article about switch bounce this morning https://www.eeweb.com/profile/david-ashton/articles/switch-bouncing-around

@noel: "...Am
I right in saying that when an electrical switch closes, and after it has
completed the whole switch bounce / arcing routine, that it settles into a
state where the metal contacts are considered to be closed, but that the two
contacts are actually separated by an ultra thin (possibly at the molecular
level) film of air?..."This is a really interesting question -- the more I think about it, the more interesting it gets. For example, I'm currently working from home (we're having a lot of work done following a leak) -- my "desk" consists of a plank of wood supported by two cardboard box pillars.I'm sitting on a metal bar stool. My body is formed from atoms, as are the jeans I'm wearing and as is the bar stool. Each of these atoms has a nucleus at the center and electrons whizzing around -- using their electrons, the atoms comprising my body form links between each other; similarly for the atoms forming my jeans and the atoms forming the bar stool.Let's pretend I'm not wearing jeans -- that I'm sitting directly on the bar stool. I'm tempted to say that the atoms in my body don;t form bonds with the atoms in the bar stool -- I imagine the electrons forming mini "force fields" so the force fields associated with the electrons forming my body are sitting on/repelled by the force fields associated with the electrons forming the stool.The more I think about it, however, the more I realize that I don;t really have a clue (this is true of so many things) -- maybe some bonds do form between me and the stool.I know we've strayed from your original question, but maybe when we learn the answer to one thing, it will apply to other aspects of this problem -- Max

The key is whether or not the atoms are sharing electrons. If the atoms are sharing electrons, they are chemically connected. In the case of pants or skin on a chair, the electrons don't don't jump back and forth between pants atoms and chair atoms, simultaneously filling an electron space in both atoms, therefore the two materials aren't chemically connected / bonded. They don't chemically connect.You might say they that they mechanically connected, because the two materials are as close as is possible, and the pants or skin atoms will deform into irregularities of the chair surface. Still, electrons aren't shared.The "mini force field" isn't a bad analogy. Electrons in the pants/skin repel electrons in the chair.
In the case of the switch, on Earth, with a thin oxidation layer, the two metals aren't chemically bonded either. Electrons will move back and forth when there are more on one side than the other, but in a stable state, they won't be shared between atoms on both sides of the switch. Electrons will only pass when there are more on one side than the other.In space, or in a theoretical vacuum setting where there isn't any oxidation, the metal atoms will begin to share electrons, which is the chemical bond.Most likely, in a normal environment, you'll have a lot of ultra tiny pockets of air, and a lot of areas where the oxidized layers of the two switch sides will mechanically, but not chemically, connect, with no air molecules between.

@ Max Maxfield: thank you for you insights, and you haven't strayed from my original question - you are actually bang on topic. What you elude to, is exactly my question. Interestingly, my question originated from a mechanical engineering problem I was working on (internal forces on components within pressure vessels, etc.), but when it comes down to it, the answer appears to lie in the electrical world when you start talking about atoms, electrons, mini force fields, etc.!@ David Benson: Thank you for you detailed description - its certainly helping me understand a bit more about whats going on at the atomic level.  Ok, so I think I understand, there are regions where the two sides of the switch mechanically, but not chemically connect, i.e. the electrons on both sides of the switch are not shared and simply repel each other.  Since the electrons repel each other, what occupies the space between opposing electrons other than the opposing magnetic fields ? I'm guessing that this is the same as asking what occupies the region between the nucleus of any atom and and its electrons, other that magnetic fields? thanks again on and all - I'm learning a lot.Noel.

It is the electromagnetic force repelling the other atoms. A lot of things will be happening on a very small scale. The oxygen molecules that are stranded in tiny air pockets may combine with the metal (oxidation), or may just sit there waiting to be liberated.
Any tiny patches of exposed bare metal may be fusing. Other bits of crud may oxidize or corrode the metal. Some of it will just mush together without having a chemical reaction. What looks like two chunks of metal just slapping together becomes an entirely new, and very fascinating world at a micro level.

Thanks Duane and apologies for getting your name wrong in my previous post.
At this micro level, both metal contact surfaces will have a rough surface.  Do you think that as the two metal contacts close that it might be feasible that due to the roughness of the metal surfaces,  that a "thin layer" of air molecules / atoms get trapped between the metal contacts keeping the contacts completely separated by a very, very small amount?  It would explain why the cube (see my second post) is not held securely to the end of the cylinder.  My thinking (whether right or wrong) would be rather like the following:Imagine a very rough surface with deep divots and pockets.  Now imagine placing some balls on this surface that fall into these divots and pockets - this represents one of the metal contacts.  Now imagine placing a second layer of balls on top of the first layer of balls, these will come to rest in the spaces between the first layer of balls - these represent air molecules / atoms.  If I then push straight down on one of the balls in the second layer with my finger, I imagine that it would be very difficult to "squeeze" this ball out of position so that my finger could come in contact with the first layer of balls.  
Thanks again,Noel Dillon.

No worries about my name. You were close enough.I wouldn't describe the phenomena as a thin layer, as you might see with a lubricant. With a rough, on a micro scale, surface, I believe you would see more a case of sporadic trapped molecules. The deeper the pockets, and the more opportunities for closing off those pockets, the more molecules you will have trapped.Free air molecules are in constant relative motion. They aren't stuck to each other as in a fluid, but they do constantly bump into each other. Up to a point, most of the air will just squirt out the sides. The balls on your analogy have a lot of friction; not enough to destroy the analogy on a large scale, but enough to break the analogy once you're down to a small number of balls.
With increases in temperature and pressure, some properties of the air start to behave more like those of a fluid, (your ball analogy starts to work again) and the air will have more difficulty escaping. Again, though, I would expect pockets rather than a cohesive layer, however thin.
Also, as the air molecules pass by the surface of other materials, electrons can be stripped from one surface or the other, creating ions, which have a charge. This isn't the same as sharing electrons. With ionic transfer, one atom will end up with extra electrons, and another atom will have fewer. With a different charge, the free atom or molecule can stick (not be repelled) to a surface without a chemical bond. It's really just resting there, not bonded. It is stuck to the surface using electromagnetic force but, unlike a chemical bond, an ion can be swept away somewhat easily. Think of a magnet being ionic attraction vs. a weld being a chemical bond.If the oxygen ion starts sharing the electron - the same electron fills a spot on both atoms - oxidation has taken place, and it's now a chemical bond.In your cylinder and cube example, if there are no, or very few, air molecules between the cube and end wall of the cylinder, without gravity, the cube will stick to the side due to the motion of the free air molecules. On three sides, molecules are constantly bumping into the cube and bouncing off, but not on the side against the cylinder wall. It won't take much force to dislodge it, but in absence of a force, it will stay flat against the wall.

Thanks Duane, very clear explanation - much appreciated.  

Please excuse this late reply, I found this thread while searching on another topic and felt I should add my tuppence-worth.

Electrical conductors are able to conduct because of a shared "sea of electrons" which are not locally bound. In order for switch contacts to permit this kind of sharing, they have to be in metallic contact.

Switch contacts are nothing like perfectly smooth, even at the microscopic level. When contacts mate and are pushed together by spring pressure, the microscopic peaks on each face are squashed together and form an array of metal-to-metal contact points. It is only at these contact points that conduction occurs. The contact area is obvously a lot less than it appears at first sight.

Fluid films of any kind are pushed aside by the force applied over these small contact areas, and brittle surface films are broken up.

If you want another way to think of it, consider that switches designed to pass high current have much larger contact face areas, and so make more physical contact points. Or consider that contact resistance of any switch is much higher than the resistance of a solid piece, of the same cross-section as the contacts, and of the same metal.