Electrical contacting on aluminium busbar

[Siwastaja]

Yeah, aluminium alloys are great for the mechanical sturdiness - low weight - low cost combo (as you could guess from the fact aeroplanes are made from it).

Almost all large-scale electrical power transmission is based on aluminium wiring and yes, even bus bars. It's a solved problem, although you need to know what you are doing and to have the process right to avoid quality issues. Copper is easier for the rest of us, but if you need to combine electricity transfer to mechanical support, aluminium's pretty high on the list!

[coppercone2]

Nah what made me weary is reading about it and seeing someyhing one bees dick from a full structure fire because of aluminum wiring. The oxide growth is insidious.

I feel bad for ypur train man, burning choo choo

[TomS_]

I can't remember the brand of grease we used beyond it being a slightly green/brownish grey and being in a tall squeeze bottle with a red lid.....

Try searching for "NO-OX-ID".

[duak]

I would recommend Alodine.  This is a chemical conversion process that leaves chromate on the aluminium surface that prevents the oxide layer from forming.  It was developed for the aircraft industry and is widely used.  We could get pieces done in Vancouver, Canada (an industrial backwater) over 40 years ago.

I see there is a Swiss company that might be helpful: https://www.bwb-group.com/en/Procedures_overview/Surface-treatments-for-aircraft.php

There are also dielectric greases and oils available to the electrical industry here that are applied to the contact areas before mechanical bonding.  These exclude air from the joint to prevent oxidation.

Cheers,

[damien22]

We will investigate all the solution proposed, many thanks for the feedback! 

[David Hess]

Would stainless helicoil thing be a good solution ? Unfortunately we cannot bolt as the bar lay flat on an insulator to which they are fixed and on the other side the PCB are mounted.

Helicoils would prevent the threads from bring ripped out of the relatively soft aluminum or copper but I have never heard of them being used for conductivity.

The PCB does contact with the bars so most of the current would go through there.

Be careful about crushing the printed circuit board material.

How much does a damn bus bar cost in the scheme of a fucking train even if you consider a order of magnitude cost increase? MTA is already going to fuck it up with unknown unpredictable means, don't make it worse. Train fires suck. They can burn in tunnels etc.

During World War 2, the copper bus bars from the substations and power houses along the Columbia river were replaced with silver bus bars and armed guards were assigned to protect them until they could be replaced with copper again.

[Gregg]

Aluminum bus bar is used extensively in large mains switchgear and even home breaker panels because it is cheap and light weight.  There are thousands of variations of alloy, conductivity and hardness available and a designer needs to choose wisely.  Most of the large current aluminum bus bars I have seen are tin plated at the factory to minimize the problems with connection conductivity.  There is nothing inherently wrong with properly designed and implemented aluminum bus bar installations; it is the branch circuit wiring that causes many of the aluminum wiring related problems and copper clad aluminum is just as bad.   
Quality commercial and industrial aluminum bus bar systems use bolted connections and are designed so that expansion and contraction will not stress the connections.  If you can't easily get a nut on the back side of your bus bar, look for PEM nuts or press fit nuts similar to these: https://www.ebay.com/i/173438526393
Helicoils are great for some purposes, but I personally don't think they are suitable for bus bar application.  In any case, as already mentioned, the bolt should not be considered as a conductor.

[kony]

Stock material in -T6 temper if it is some 60xx alloy or in -H18 if you go after highest conductivity (EN AW 1050), electroless nickel per ISO 4257 (at least 15um, ~10% P content) and fuck it. There is nothing nonstandard at it, even for aluminium alloys, and any good plating shop serving electronics industry will be capable of that. Don't forget to keep the bolted connections under preload, prefferably not just by the screw torque, but by some other means as well - best case disk spring washers. Especially if there is PCB pads in the stack-up of the fastener, the common laminates will creep badly over time!

Alodine or other conversion passivation is wrong here, so are dissimilar metals inserts.

Also be VERY wary of tin plating on busbars (regardless of base metals) - whiskers are a big deal here and can turn your gear into big plasma ball in short few years.

[MagicSmoker]

Stock material in -T6 temper...

Finally someone gets it!

The type of lock washer most commonly used in the US for high-reliability bus bar applications is called "Belleville." It is a conical in cross-section with radial ribbing or teeth; I'm not sure how popular Belleville washers are in the EU, though, so Nord-Lok are an excellent - albeit more expensive - alternative. I also agree that directly threading the bus bar - even with Helicoil inserts - is not ideal, but it can work just fine for small small ring terminals such as #6 (M3) through #10 (M5); once you go past 1/4" (M6), though, you really can't generate enough clamping force with threads in aluminum (unless the bus bar is unusually thick). I also agree that alodine is not an appropriate choice of surface finish as it's effectively an insulator.

[T3sl4co1l]

Ahh-- not quite!

Conical washers are indeed springy.  They're perfect for this: a short-throw, high-force spring.

Nord-Lock aren't conical or domed, as far as I know.  So they don't take up any spring force.  They do prevent unscrewing -- very important in high vibration (especially high shear) applications, but only useful when the bolt and base metal can be relied upon to not creep or stretch.  They're also rather expensive AFAIK, so are best reserved for these cases when they're needed.

There's also wave washers, though they tend to be softer than cone washers.

Split washers, by the way, are literally no good at all: as a spring, they crush at even less tension than wave washers do, and in tests, they perform as well, or poorer, than plain old flat washers!  Which is often worse than no washer at all, because they add another friction surface and can make the joint easier to move.

FR-4 can creep or flow under pressure, too (especially at high temperature).  If you're making high current bolted connections to PCB, consider using spring washers there, too!

Tim

[Mechatrommer]

Now there are two issues, one is the electrical contact, because of the oxidation of the material which is an issue for the electrical contact and another is the strength of the material to avoid the bar to be bend during manufacturing.

i think you over complicating things. i think oxidation will only form on exposed part, if there will be oxidation on the contact, it means you've made not a good contact so you may revise your assembly technique. if bending is an issue, why dont use thick wire and find suitable tap to it? people have done bolt nut tap with no issue at much larger current, i think its the de facto standard... if you really concern, just solder them, problem fixed... now where's that high voltage nutter? why dont he chime in to give some good and practical advice, safety-wise advice, rather than keep breaking somethings

[MagicSmoker]

...
Nord-Lock aren't conical or domed, as far as I know.  So they don't take up any spring force. ...

Nord-Locks develop tension in the bolt through cam action. More specifically, by interlocking wedges on opposing washer faces with the wedges arranged such that unscrewing the bolt increases the tension on it (at least for the first few degrees of rotation). However, you are correct that they aren't recommended for materials which will experience significant creep after tightening. So, soft metals like unhardened copper or aluminum, especially if they are less thick than the bolt diameter. Also not recommended for cost-sensitive applications, as they are, indeed, bloody expensive. I only used them for the 1/2" bolt holes on dual-hole 4/0 compression lugs; a case where, contrary to the Mythbuster's mantra, failure was not an option.

Split washers, by the way, are literally no good at all...

Also true, at least for bolts under, say, #10 (M5) in diameter. Split-ring lockwashers can be okay for non-critical use on 1/4" (M6) or larger bolts. I tend to prefer internal tooth lockwashers for general purpose application, myself.

FR-4 can creep or flow under pressure, too (especially at high temperature)....

Very, very true! In fact, I don't think any kind of washer can accommodate the amount of creep a PCB will experience even at temperatures well below the glass transition. Either use a screw terminal standoff (e.g. - Keystone Electronics 7787) or a soldered-in male quick-disconnect (e.g. - Keystone Electronics 1287) for high/moderate current wire to board terminations, respectively.

[David Hess]

The type of lock washer most commonly used in the US for high-reliability bus bar applications is called "Belleville." It is a conical in cross-section with radial ribbing or teeth.

I would love to use Belleville washers more for power transistor mounting but pricing and availability has always been a problem.

There's also wave washers, though they tend to be softer than cone washers.

When it matters, I end up stacking wave washers to get enough force.

[coppercone2]

how about lockwire?

I have been meaning to get lockwire pliers.

Pretty sure you can modify hex insert bolts to accommodate lockwire too. What I am thinking is to drill a small hole in the busbar, braze on a peg with a hole in it, then lockwire the washer to the peg.

[OM222O]

why are you trying to manufacture the bars with such weird dimensions? usually for longer runs, you would use a plate shape rather than a square (6x7 is almost a square) and you can make runs like this without a problem using copper:


this also gives you a lot more width for your connections. as you mentioned, you're struggling with a single M4 screw! I suggest using 10 screws in parallel next to each other per PCB as it provides a lot better contact with the PCB as well as the bus bar, makes it easier to manufacture the bars and is also safer in case one of the screws fails or isn't making proper contact, as well as reducing the load to 12A per screw rather than 120! there would be a ton of voltage drop and heating of the screw with 120A passing through it and it won't last long at all! just make sure to use a washer on both sides which ensures there is good grip and you will be fine. I attached an example of the PCB which you might want to use:


Obviously make the thick copper zones (light blue) on both sides for a more even distribution and also add some vias for stitching just to be on the safe side.
One single M4 screw is a terrible idea and you will have lots of issues with it.Also if you make your plates out of copper, you can add plenty of flux and burn it with a torch to get rid of the oxidization but I'm not sure if the same can be done for aluminium. copper also oxidizes a lot more slowly than aluminium (a few days compared to a few minutes!).If you can, drill the holes in the positions with a smaller bit, let's say M2.5 instead of M4 so you have the correct positions. then just before installing the PCB, drill and tap the holes to the correct dimension. This way you can make sure there is no oxidization during installation. I'm not an electrician but this issue has already been solved by others  I hope this answer helps.

[T3sl4co1l]

Lockwire serves the same purpose as Nord-Lock but with cheaper materials and way more manual labor (and possibly less retention torque, depending on the stretchiness and ultimate tensile limit of the wire, and give or take how long the wire is to the tie point?).

I would love to use Belleville washers more for power transistor mounting but pricing and availability has always been a problem.

That never made sense to me... they can potentially be produced in a single stroke like any other washer.  Is it just a quantity thing -- they're expensive because no one wants them because they're expensive?...

The ridged ones I can see being more expensive or specialty, I guess, but still.

'Spose this is a good situation to shop China...

When it matters, I end up stacking wave washers to get enough force.

Indeed, wave washers add in parallel when stacked.   (They add in series if you can keep them aligned peak-to-peak rather than peak-to-trough, but that would be hard without keying.)

Belleville also add in parallel when stacked in nested-cups direction, or in series when stacked oppositely, rim-to-rim.

Series meaning, more displacement but the same force; parallel, more force but same displacement.

Tim

[dmills]

Belleville is a trade name for a type of "Disk Spring" that may be more readily available under that name.
I tend to keep a box of M3/4/5 on the general mechanical bits shelf, they are useful.

FR4 flows, a lot, so directly bolted connections are not a great idea, but you can get press fit inserts from the likes of Erni and Wurth that make heavy current bolted connections to PCBs a far more viable thing.

Regards, Dan.

[coppercone2]

don't you want to drill a hole and solder in a insert to a PCB?

[damien22]

We are currently looking to have tin plated or zinc plated aluminium busbar, waiting for feedback from some suppliers.

Here is a picture of how it looks like and also explains the dimensional constraints.




http://imgur.com/a/bWYE9om

https://imgur.com/a/bWYE9om

(image tag not working ?)

[David Hess]

I would love to use Belleville washers more for power transistor mounting but pricing and availability has always been a problem.

That never made sense to me... they can potentially be produced in a single stroke like any other washer.  Is it just a quantity thing -- they're expensive because no one wants them because they're expensive?...

I have always assumed it is a quantity issue.  Wave washers have more applications than Belleville washers and people use split washers in place of both even when they should not.

[langwadt]

I would love to use Belleville washers more for power transistor mounting but pricing and availability has always been a problem.

That never made sense to me... they can potentially be produced in a single stroke like any other washer.  Is it just a quantity thing -- they're expensive because no one wants them because they're expensive?...

I have always assumed it is a quantity issue.  Wave washers have more applications than Belleville washers and people use split washers in place of both even when they should not.

random hit from google, https://www.amazon.com/Belleville-Washer-Stainless-Steel-Pack/dp/B01J2XR93G

$8 for 100 M4 washers isn't what I'd call expensive

[David Hess]

I have always assumed it is a quantity issue.  Wave washers have more applications than Belleville washers and people use split washers in place of both even when they should not.

random hit from google, https://www.amazon.com/Belleville-Washer-Stainless-Steel-Pack/dp/B01J2XR93G

$8 for 100 M4 washers isn't what I'd call expensive

Oh, they can be found *now* but go back 20 or even 10 years and it was a problem.  The same thing from McMaster Carr is 10 times the price.

And them someone in the field loses the washer during a repair and cannot find a replacement at even the local industrial hardware store.

[tooki]

Posting here after Months of unsuccessful research and I hope there gonna be some smart guys with good idea.

Surely, with all the heavy electrical industry here in Switzerland, this is a problem that's been solved over and over domestically. There must be some industry groups that you could meet up with and talk about it with, including finding a supplier. It seems absurd to me that, in a country known expressly for precision manufacturing, that you can't get what you need made.

[tooki]

read about the oxide growth on aluminum to get an idea of whats going on. finishing101.com has some insight (linked directly in one of my previous posts).

When I read that I wanted to no longer ever use aluminum and just pay the piper at 10x cost. fuck it. Not an option? its infrastructure essential to the function of a modern society!!!

its a train they are in service for like 70+ years sometimes, look at cuba or Switzerland. Why are you going cheap for that? people actually like trains. Its not a going obsolete in 2 years equipment. Transportation is important.

How much does a damn bus bar cost in the scheme of a fucking train even if you consider a order of magnitude cost increase? MTA is already going to fuck it up with unknown unpredictable means, don't make it worse. Train fires suck. They can burn in tunnels etc.

Bruh… take it down a notch, mmkay?

[tooki]

Stock material in -T6 temper...

Finally someone gets it!

The type of lock washer most commonly used in the US for high-reliability bus bar applications is called "Belleville." It is a conical in cross-section with radial ribbing or teeth; I'm not sure how popular Belleville washers are in the EU, though, so Nord-Lok are an excellent - albeit more expensive - alternative. I also agree that directly threading the bus bar - even with Helicoil inserts - is not ideal, but it can work just fine for small small ring terminals such as #6 (M3) through #10 (M5); once you go past 1/4" (M6), though, you really can't generate enough clamping force with threads in aluminum (unless the bus bar is unusually thick). I also agree that alodine is not an appropriate choice of surface finish as it's effectively an insulator.

Just a little bit of nationalist pedantry: Switzerland (where the OP is) is in Europe, but is not part of the EU.