Iron core transformer heating 1

I have some linear power supplies (6v 30A), which have iron core laminated transformers. When operated on the high side (126vac) of their normal input voltage operating range 115vac +/-10%, the transformer core will be at a temperature that is much higher than is charateristic of this power supply. Also, one or more of the 4 bolts holding the core laminations together will get very hot.
What could cause this ?

 

[electricpete]

Iron core heating always increases with applied voltage because the flux in the iron is dependent upon the applied voltage. As voltage is increased to the saturation point, the increase is even more dramatic.

 

[230842]

An additional comment to Electricpete post. Iron core losses, and therefore heating, increases proportional to the square of voltage (almost). That makes a 20 % overheating. Julian

 

[gordonl]

The bolt may be heating up because it is shorting the laminations of the core togrther providing part of a path for eddy currents. (Laminations would need to be shorted together at a second point as well) This would also add additional heating to the core.

 

[electricpete]

Good point by gordon.

It gives me another idea... probably a little farther out.
Shouldn't those bolts be non-ferrous so that there will be no hysterisis losses. That would rule out standard steel bolts and require aluminum or other special bolts.

An interesting side trip... in transformers there is often a great deal of effort to insulate the through-bolts from the core iron, and to ensure the core is grounded at single point only. In motor stators all of the through bolts are grounded and sometimes welded to the core (uninsulated from it). I believe the reason for this difference is that the motor core bolts are located on the outside of the stator (back-iron) where the flux is very low.

 

[prc]

In trf ,core losses- iron losses or no -load losses (all one and the same)increases with increase in perturn voltage ie applied voltage.It cannot be concluded as proportional to the square of the applied voltage because it depends on the flux density at which the trf is operating.Some times it can be several times square esp if working in saturated region.But since the trf is rated for 126 volts it should not heat up excessively,eventhough heat may be slightly more than at rated tap.
Regarding bolt heating look how the bolts are located-through the core or outside only or one inside and other through window with ends insulated.Any chance for the bolt to form aloop through window ?Then it can act as a live turn carrying current and hence heating itself.No advantage for non magnetic bolts as no flux will flow through it and no question of hysterisis loss etc.If bolts are out side the core no need to insulate at ends.

 

[electricpete]

prc

can you explain a little more why non-ferrous bolts don't provide an advantage? How can you be sure that no flux flows through the bolt? I'm assuming it's a core through bolt. I would assume that the flux would divide through core steel and the bolt in inverse proportion to their reluctance. This will of course channel most of flux though the iron, but it doesn't take nearly as much flux to heat up regular steel. And if the core approaches saturation it's reluctance will increase forcing more of the flux into other paths which may include the bolt.

And why do some manufacturer's caution against the use of ferrous bolts in the vicinity of low-voltage bushings of power plant stepup transformers (very high current).

 

[prc]

In case of through bolts in core, as it is correctly analysed ,the flux will go only through the low reluctance silicon steel and not through the bolt.Of course, if overfluxing occurs,then flux will overflow in to bolts .But 126 volts seems not sufficient for such a calamity .In olden days when through bolts were used in large capacity trfs it was always ferrous bolts only.

In case of high current flow ,it generates a strong magnetic field around it and this will induce eddy losses and hysterisis losses in magnetic material causing it to heat up.Hence the advice to use non ferrous bolts in such locations.

 

[electricpete]

PRC - thanks for the good info. I learn something every time I read your posts.

I can now see your point that in general we wouldn't need non-ferrous bolts in a design if core is operating normally below saturation.

In the present case where bolts are "very hot", there are only 3 possible explanations I can think of:
1 - Conduction of heat from the core (Infrared survey should be able to tell if this is the case based on which part is hotter.)
2 - loop current flow though the bolt as mentioned by gordon and amplified by prc.
3 - hysteresis heating of ferrous (steel) bolt.

I think item 3 is certainly possible. As you say it's only likely if the core is operating in saturation... but that is not outside the realm of possibility since we already know the core is abnormally hot.

 

[prc]

electricpete,
Thank you for your kind appreciation for my postings.Let me try to rise up to your expectations in future too.
We do not know the fluxdensity at which the trf is working.Normally it will be at 1.6 Tesla .Then at 126 volts it will be at 126/115* 1.6=1.75T ,quite safe for working.But if it was designed at 1.7 T ,then flux density at 126 V will be 1.86T,high too cause core over heating but not sufficient to cause heating of nearby bolts by hysterisis loss as flux will start overflow at a flux density of ~2T only.It was also reported that bolt heating is on only one bolt and not on all bolts.Hence I thought it was not a case of flux over flow.To understand flux conductivity ,if permeability of air for flux is 1,then for non ferrous it will be 3-6,steel 300-10 000,silicon steel 15000-60 000 (dependiniung on flux density).