Galvanic Corrosion

Working on LARC XVs an ali hulled exU/S military vehicle this issue is a dayly problem.! Dave(sod) ,Joseph and others have the chemistry and methodology. Heat, lube and gentle violence. Most threads on the LARCs are helicoiled from new. A good paint system will be good defence in future, but for now one trick not mentioned is EDM. take out the middle and then unwind the spring(thread). The process is very accurate. Then fit helicoils. Noel.

Turn fitting over and build a dam out of plasticine or some thing around the screw head. Fill with penetrating fluid and leave for a few days. Remove and dry off the surface fluid and gently heat the surrounding metal. Then try an impact driver, (electric or air), set to its' lowest force setting, and applied in alternate directions a few seconds each way?

Left hand drill bits are cheap, quick and easy.

Thanks for the further suggestions

We have Domestic matters which must now take priority over what I was hoping might be a quick & easy job; so it will be at least a couple of months before I can get on with this.

Whether I succeed, or end-up breaking something, I will of course report back.

MichaelG.

Hi MichaelG, if you have a soldering iron that gives off a lot of heat and can hold a tip a suitable size to place onto the screw heads without it touching the aluminium parts, it may be a way of putting heat into the screw. You might have to repeat the operation a few times to allow everything to cool down between each heating period. Another way might be to hold a red hot poker (a large pointy diameter piece of bar) but you might have to have two or three irons in the fire, as they will loose heat quite quickly. The idea being the heated up bar cramped up tight in a vice which just drops out when it cools, because whenever you heat anything it will expand in all directions it can, but if it is restricted like a screw in a hole it will expand more in it's length, but will still contract in all directions upon cooling. Apologies if any of this is trying to teach Granny how to suck eggs.

Another crazy idea I've just thought of, is wait for a hot sunny day and focus the sun onto the screw with a powerful magnifying glass, like you used to do to your school mates who were too naive to know it would be burning hot.

Regards Nick.

The actual corrosion product will be Aluminium Hydroxide - which is a sort of jelly in water. The combined water can be lost in a dry atmosephere or which heat, leaving Aluminium Oxide. The only 'solvent' you might have is a water-borne alkali, such as Sodium Hydroxide, (or Potassium ditto), and for a more gentle effect, sodium carbonate (in the UK = washing soda, other household names are available around the empire). But, as you sumise, Alkali will also corrode the aluminum itself.

This process - using aluminum foil and washing soda - is a good way to remove staining from real silver or electroplate. The aluminum will fizz gently - more in warm water, and it changes the sulphide on the silver back to silver itself, and gives off Hydrogen Sulphide ( which can be detected in the smell of the process.)

I am reminded of a ditty which circulated in the Banbury Alcan factory:

Aluminum does not rust,
It crumbles to a greyish dust,
And most of that which you can see
Consists of Al two O three.

Cheers, Tim

Caustic soda was supposed to dissolve the aledged Aluminium hydroxide blocking car radiators with engines sporting the latest 1970's joke, aluminium heads with a steel block when most antifreeze was methanol based.

That didn't work either.

Oddly the sellers of greenhouses worked out the solution this problem out years ago - aluminium nuts and bolts.

As indicated yesterday, I am obliged to let this lie for a couple of months, and will be packing the structure away carefully.

Meanwhile: here are a couple of snapshots

.

.

One of the screws removed from the pole-attachment cap …

The white substance on the thread is the Aluminium Oxide, and the quantity looks trivial; but this screw was extremely stiff to remove, even after using PlusGas. The majority of the corrosion is, I think, still adhered to the female thread, but [not possessing the appropriate Go/No-Go gauge] I cannot really check.

.

.

The remaining problem [one of two instances] … Photographed upside-down for convenience

.

Please keep the suggestions coming … I will report back in due course.

MichaelG.

You didn't mention it was a hex head set screw. I was thinking smaller Phillips head type of thing. Brute force and ignorance on the end of a long ring spanner with a bit of impact should do the job. Or mechanised brute force and ignorance in the form of an electric or air impact driver with an impact socket on it. These make quick work of such jobs and don't twist the parent assembly etc too much.

A fresh bolt makes a handy go/no go gauge but it would be a good idea to run a tap down the female thread anyway to clear the oxide powder etc out of the thread. Or make a redneck thread chaser by putting a longitudinal hacksaw slot part way through the end of a bolt.

Another thing you can do is go carefully around that boss that the bolt screws into with two small hammers tapping simultaneously on opposite sides of the boss. Or hold the boss on a piece of steel or the vice jaw etc and tap opposite with a hammer or hammer and drift while moving the job around between blows . This tends to expand the metal of the outer thread and loosen it up. An old steam fitter's trick for getting apart threaded pipe joints that have been corroding together for countless eons. It will loosen rusted pipework that otherwise will not budge even with five foot of pipe on the Stilson wrench handle for leverage.

Edited By Hopper on 02/05/2022 10:17:03

Edited By Hopper on 02/05/2022 10:18:02

Hi, as regards using extensions on a spanner or an impact driver, I would suggest you use caution on bolts that have corroded in on ally items, I've seen this done and they do come out, but very often they bring the threads with them. If however there is enough metal where the hole is, it can be plugged with a lager piece and a new hole drilled and tapped, but if the damage isn't too bad a Helicoil repair may work.

Regards Nick.

.

True … but I did provide several links to the manufacturer’s website.

MichaelG.

Which showed delicate instruments with a multitude of small screws.

Oops … Just had another look at the page

[Edit:  __ first of the three links that I posted previously]

and realised that the cross-arm appears to include the pillars

.

… which would suggest that the instrument is screwed down onto it

MichaelG.

Edited By Michael Gilligan on 02/05/2022 14:05:36

I have had reasonable success with parts of that configuration using a portable spot welder. There needs to be access to clamp the jaws across opposite sides of the boss. After scraping to get contact, power is applied to heat it. The current only flows in the outer, so expansion tends to loosen the grip. With steel it takes maximum welder settings to get to near red heat as quickly as possible. Aluminium must obviously to be kept cooler, but expands more. If the parts move at all, it is then merely a case of working in penetration oil, and patience. The method can also facilitate oil penetration at lower temperatures, with just enough heat to result in a capillary gap.

The fast and targeted heating seems to increase the chance of success. The nearest alternative is fast heating with oxy-acetylene which is far more risky.

I'd certainly like to know more about this, maybe in a separate thread if you haven't covered it already.

I like the idea of electrical, but I'd expect it to be harder to do to Aluminium than steel. The problem with Aluminium is it's a much better conductor than steel, so much more current is needed to get the same heating effect.

W = I²R

Where W is heat in Watts, I is current in amps, and R is the resistance of the heated element in ohms. Note the amps have to be squared to get watts out of a low resistance. Aluminium also conducts heat away better than steel, again requiring more watts to get up to temperature. A welder powerful enough to loosen to a steel bolt might not be man enough for Aluminium.

Rather than connect directly to the Aluminium, it might be better to interpose a carbon block as a heating element. As Carbon has a higher resistivity than Aluminium, it would heat up with fewer amps, putting less stress on the welder. The disadvantage is the carbon block heats the corroded joint indirectly, reducing the value of rapid expansion.

Macolm's comment about oxy-acetylene being risky here is well made, and similar applies to most of the other methods that work well on steel. Aluminium being soft, chemically reactive and with a low melting point makes it hard to free up a corroded joint without causing serious damage.

Dave

Edited By SillyOldDuffer on 03/05/2022 11:21:59

Aluminium and Stainless definitely don’t mix unless separated by a suitable inert compound, I have a Ring doorbell and the aluminium body is held together with stainless screws, and you guessed it, it’s corroding because a doorbell unit is outside and obviously gets wet. This is the second Ring doorbell I have had, first one failed when water penetrated and did for the electrics, and would you believe it was out of guarantee period. Do manufacturers deliberately use dissimilar materials to promote premature failure and guarantee future sales or are they just blissfully unaware of the characteristics of the materials chosen to construct their products and the consequences of their choices. As an ex aircraft engineer I can vouch for the effectiveness of such products as Duralac etc. they do a very effective job of preventing dissimilar corrosion. Dave W

Despite intuitive doubts, I used this method to free very corroded hydraulic unions on Citroen rear suspension cylinders. You can see the design on this photo of a tester made from a such a cylinder. I agree that spot welder dissipation will be lower for aluminium, but with perhaps 5000 joules available over a couple of seconds, 10 grams can be heated by more than 300C. Then taking the differential expansion as 7 parts per million per degree C, we have 2 thou per inch (to mix units!), a very useful loosening of the parts. It does indeed work.