Which material to use for bolts, washers and nuts? 4

[dakrause]

Greetings,
Please see the pictures below. The white pipe is part of a 15x20 foot tank made of carbon steel(fully painted). The bolts, washers and nuts are made of galvanized steel and the valve is made of brass. This corrosion appeared after the assembly was left out in the rain for just a few days. Galvanized bolts are used at other points, but do not show this problem.

Which material should be used for the bolts, washers and nuts if this is to withstand the elements for several years?
Thanks,
David

 

[Metaljon]

Corrosion is to be expected as you have set up a galvanic cell between copper and iron coated in zinc. You need to use a gasket to electrically insulate the different metals from one another. It is very important to use insulating materials with sufficient resistivity and that metallic contact between the different metals is avoided throughout the whole construction.

 

[TomDOT]

Yup. Mixing metals is a great way to set up galvanic cells and induce corrosion - particularly if you have more of a voltage difference. Are you by chance in an area where chlorides could accumulate on the surface? Nice salty water helps speed the galvanic corrosion along.

 

[dakrause]

Yes, this will be less than a mile from the ocean, so there may be some salt accumulation.

I understand that I should use non-conductive washers. Is this correct? Anything else?

 

[TomDOT]

Nonconductive washers, and prevent the bolts from touching the other metals.

 

[CoryPad]

What is flowing through these parts? You may have severe corrosion on the inside of the steel pipe if it is uncoated.

 

[racookpe1978]

You MUST also insulate the "inside" of the hole of the flange or the flange and bolt will touch each other at each "hole" in the flange. : There are "kits" available that include the specially-sized bolts, gaskets, hole inserts, washers, etc.

 

[rmw]

Repeating what the others are saying but in a different fashion, the two flanges have to be electrically separated (isolated) from each other, hence the need for the entire isolator "kit" that RAC mentions.

CoryPad brings up a good point too. Your need to do this may be more for what is inside the pipe to protect the less noble steel from the more noble brass. Your real corrosion problem may not be the bolts.

And, if you are only a mile from the sea, you are by definition a Marine environment.

rmw

 

[TVP]

David,

The others have clearly outlined what is required to make zinc-coated steel work with a brass flange. Another option would be to use brass fasteners. Here is a link with some information on standard brass grades (ASTM F468):

http://www.cookfasteners.ca/faqs_1.php

 

[dakrause]

CoryPad, (very dry) transformer oil will be flowing in these pipes, so internal corrosion is not expected to be a problem.

TVP, thanks, I would prefer brass fasteners to trying to electrically insulate galvanized fasteners. I will look into this.

So does this mean I should avoid using galvanized parts attached to the carbon steel main tank, including galvanized cable conduit and galvanized bolts?

 

[TomDOT]

Absolutely I would avoid galvanized parts attached to a carbon steel main tank. Not only will you be often setting up a galvanic cell (unless you isolate) - you will also accelerate that galvanic cell with the "large cathode/small anode" effect. Consider the difference in potential as your voltage in driving corrosion. The difference in surface area is your amps available per unit area being corroded.

If the carbon steel is painted well, it will minimize the effect. Your biggest problem with the initial photos you posted is the bare brass. While many grades of brass are relatively corrosion resistant, the zinc galvanizing will still sacrifice itself to protect the bare brass.

 

[Screwman1]

Monel bolts would be much better than the galv. ones that you have now.

 

[TomDOT]

Monel should only be 0.10 to .15V off from your brass. Zinc to brass is more like 0.8V - and as a bonus, the Monel would be your cathode.

In your materials selection, some of the major galvanic corrosion factors to consider are

1) Potential (voltage) difference
2) Time-of-wetness (rain, dew)
3) Chlorides (chlorides are awesome at driving corrosion)
4) Other conductive salts
5) Cathode/Anode area ratio.
6) Projected lifespan of the item

0.8V is a LOT - it's way too much for most outdoor environments. You have to start thinking about mitigation to break the electrochemical cell - isolation kits, painting, et cetera.

In your case, you have high time-of-wetness, high chlorides, high other salts, a long projected lifespan, and with the small galvanized parts, you have the wrong Cathode/Anode ratio and a large potential difference.

The first good report on this effect (including isolation) is the British Admiralty report on the iron-nailed copper sheathing on HMS Alarm - which only had an 0.5V difference. Similar Cathode/Anode area ratio (small iron nails) but higher chlorides, salts and time-of-wetness (saltwater immersion.) The report on the corrosion was issued just two years after Alarm launched.

The "isolation kit"? - the copper sheathing was wrapped in brown paper to prevent damage prior to installation. Some of the nail heads had brown paper trapped underneath: workers had simply nailed through the paper, and removed the bulk of it later. This made a noticeable difference in the corrosion rate.

 

[stevenal]

We use 304 stainless bolts and washers with brass nuts (to prevent galling) in similar work in a similar environment. Isolation might be desirable if you had an electrolyte in the pipe, but it is not necessary here.

 

[dakrause]

A follow-up post: I brought in a chemistry professor(onsite) and two corrosion specialist companies(remote) to solve this corrosion issue. A chronologic summary is as follows: the equipment finished testing and assembly in Mumbai, India(sorry, I didn't initially mention the current location of the equipment) on 2 August 2012. In August it was moved outside for less than a week, then it was brought back into the factory. The pictures above were taken at the end of September.

The investigation found some stainless/galvanized mismatch and some electrogalvanization instead of hot dipped galvanized, both of which were corrected.

But the fundamental problem(which was the cause of the corrosion in the pictures in the first post) is the extreme acidic nature of the monsoon rains in Mumbai. I wound up changing hundreds of nuts, bolts and screws before applying zinc appropriate primer and painting these parts.

The chemistry professor was saying that Mumbai rain results in a diluted version of Aqua Regia, the only substance capable of corroding gold. Smog leads to sulfuric acid in the rain, abundant lightning results in nitric acid in the rain, and salt from the sea combines with either of these to produce hydrochloric acid.

Here is another picture:

 

[TomDOT]

Thanks for the followup report! Being in the USA, I usually don't think about how much acid there can be in rain in more polluted areas of the globe. We have a fairly good handle on acid rain at this point, with the notable exception of some grandfathered coal power plants without much in the way of pollution mitigation.

 

[MJCronin]

You also may want to consider PTFE coated bolts and fasteners for your particular environment.

They offer superior corrosion resistance as compared to galvanized fasteners and are cheaper than a Monel type solution.

http://www.metcoat.com/ptfe-coating...ep549-4v47xy&gclid=CNDm_sCSyrQCFexAMgodlxYA4w

http://www.eng-tips.com/viewthread.cfm?qid=325662