Magnetic Core Material for Subsea Transformer

why?

As in for what possible practical purpose would it serve?

Why a subsea transformer?

Have you ever heard of subsea wellheads my friend? How do you think they are powered? Most new wells are in deep water out beyond the continental shelf. Far too deep for a platform. Umbilicals with power cables must be run to connect them to platforms closer to shore in shallower water. Some of those runs are fairly long (several miles to tens of miles or more), making low voltage high current interconnects prone to resistance losses, therefore you have to use high voltage AC and step it down when you get there.

As to why it would be exposed to seawater, that is proprietary and I cannot discuss it.

You mean they don't use sub-sea diesel engines ;^)

I think the trying to create a water-tight seal between laminations, isn't going to work. You're gonna have a water soaked core.

Back when I was at Hipotronics, the cooling issue was always on the back burner... Employee bonus bait, if ya know what I mean.

In this design there won't be laminations, solid barstock baby! and since it'll be submerged in water that is just a little above freezing and at pressures that will go a long way to prevent the formation of bubbles, I doubt cooling will be a big issue either.

solid core = bad efficiency.

Nickel 200 has low resistivity, so eddy currents will be horrendous, I²R power losses will exceed the cost of the can in a very short time, possibly enough voltage across the core to cause electrolysis or corrosion.

Get a premium quality core as small as possible and plan to run the thing *HOT*, seal it in a good stout SS can and flood the thing with high temp fluid (use "Coolanol Silicate Ester" even if it hydrolyzes is can't ignite under water). Deep water is a good approximation to the "infinite heat sink". By keeping the can small you can push the device to a 55C surface temp which should keep the thing clean.

In the long run the extra you spend on an efficient transformer in an expensive enclosure will more than pay for itself in energy savings.

I thought I posted this last night. Did anyone else notice some missing posts?

I wonder if the eddy currents in the core will lead to some localized electrolysis and corrosion, even in a corrosion resistant metal.

An iron core with a sacrificial anode might stand up to the salt water.

Hi,

should do the job but some problems may arise:

permeability is low, 500 to 700 (?),

corrosion by DC stray currents will start below 1 V,

corrosion by AC currents can be suppressed if using higher frequency, I don't think that this will be a 50/60 Hz device.

We operated 400Hz equipment with bare graphite electrodes below a conducting fluid and electrolysis started above 115 V (divided by two gaps in series of 3 mm each).

What are the hydrophone people doing?

Insulating the electrical part may bring bigger difficulties.

RHABE

Since this is a power transformer, one consideration is core saturation. core saturation of nickel is between 1/3 and 1/4 that of iron. This means that your core cross section will need to be 3-4 times that of an equivalent iron core. By the way, the core will still need to be laminated to avoid eddy current losses. You probably need to speak with a transformer design person, but I would recommend cobalt or a cobalt alloy. There are a number of iron alloys that may have adequate corrosion resistance. Alternately, you may be able to coat an iron core with corrosion resistant material. Most of the super metal cores are very costly.

On second thought, I would consider the use of a conventional power transformer in a water tight container filled with a dielectric fluid. I think the associated engineering problems will be much more manageable.

Unfortunately, because of the intended use, that won't be possible.

What about the less obvious problem of biological fouling? It would seem that it wouldn't take long for little critters would like to camp out on your core and change the cooling rate rather rapidly.

That is a concern, I'm thinking of plating the exterior of the core with something like copper to help cut down on fouling.

Hi,

I do not think this to be necessary as the small lovely creatures on your surfaces will be boiled off at temperatures of 60 to 70°C.

So unless you need clean surfaces no coating with copper needed.

Welderman's suggestion of using iron-cobalt (a little bit costly but very good) may be suitable without protection.

You should state what you need:

high permeability?

high magnetostriction?

low coercivity?

low high frequency loss by eddy current or by changing magnetisation?

Low sensitivity to stress and strain?

These are different and partially contradictory requirements.

RHABE

therein lies part of my problem RHABE, I'm a mechanical guy and I have a pretty good idea of what I want it to look like mechanically (it is actually quite a bit more complicated than a simple transformer, it does a couple other functions too which I can't get into unfortunately in an open forum like this), but designing a transformer is not my forte so I really don't have a feel for which parameters are important. we're going to have to bring in a consultant for that part of it I think.

Transformers are *very* simple devices heh-heh ;) AC voltage in, ac voltage out, some energy lost as heat and a tiny bit as sound. But seriously, if it does a more than that then it's not a transformer.

Generally, when designing transformers the only input the mech guys get to have, is where to put the lifting lugs on the *outside* of the can and *maybe* where to drill a hole for a bolt in the core clamps. {I know, I know, that sounds snotty doesn't it?}

But seriously, the mechanical factors in a transformer core are *almost* irrelevant. The laminations only have to be strong enough not to bend under their own weight, same goes for the conductors (I'm being metaphorical, because there is force on the conductors.) But cores are still wrapped with paper, plastic film, and adhesive tape.

Water always has been and always be the enemy of transformers & gunpowder.

Why not a ferrite core?

Not as good ferro-magnetically as soft iron or Si steel, but corrosion resistant and very high resistivity thereby reducing eddy current losses. Ferro-magnetic performance still quite good.

Another alternative could be to cast a core from epoxy loaded heavily with soft iron or Si steel powder.