Resurrection of a CFEI 100 KVA Induction Furnace

[awemawson]

At last the trench is dug 

A friend of a friend came round this morning and kindly dug my trench - quite a big excavation initially to get a good angle to access the end of the 45 degree hole that I'd bored some days ago through the concrete. And of course it HAD to emerge just where years ago there had been a 6x6 fence post concreted in. Once the remnants of it's concreting in had been broken away life got a lot easier - the trench was brought up to about a spade depth and continued to the complicated corner where all the water pipes are currently exposed.

The corner of the concrete slab will need cutting off to bring the cable into the stable at a more benign angle and another friend is hopefully coming this afternoon to help me with that, and actually pull the cable though the hole.

[awemawson]

My friends came this afternoon and between us we got the cable pulled through the hole in the concrete using a bit of waste pipe as a sleeve to protect its sheath.

About 4 metres extra was pulled through to allow the bend of the cable to turn 180 degrees for jointing in the resin joint. Once that joint has been made and tested the excess cable will be pulled back into the trench and into the stable.

The corner of the concrete by the stable was cut off and again the  cable sheathed in a bit of waste pipe to protect it as the excess length is pulled back.

So, a resin joint to make off, a 125 amp wall mounted plug to make off, and two relatively short flexible umbilical cables to be made and apart from back filling the trench we are there 

[awemawson]

To more easily tackle the resin joint I decided to make an assembly jig to hold the two cables firmly, and pointing concentrically at each other. A couple of plasma cut steel plates mounting four suitably spaced cable cleats elevated on bits of 4x2 and screwed to a board.

It's just as well that I did this as assembly was a bit of a pain. I came across an issue that has never occurred to me before - the two ends of a cable are DIFFERENT. In this case looking at one end the cores read clockwise L1, L2, L3, N BUT the other end reads (again clockwise) L1, N, L3, L2. Now usually this doesn't matter a jot, but in this case where four heavy cores are being crimped in a tight space, and are separated by a cross shaped piece of insulation, two of the cores have to cross over no longer leaving a space for your cross shaped insulator  Had I realised this earlier I could have threaded the last cable starting at the other end, and the cores would have all been flowing the same - but too late for that. A bit of ingenuity re-manufacturing the cross shaped insulation and forming the cores with a mallet to ensure adequate spacing, and an assembly was formed that was JUST compact enough to go in the plastic shell ready for pouring the resin. Before the actual pour I did an insulation test at 1000 volts and got well over 500 M ohms.

It is now probably already cured, but I will leave it overnight just to be sure as manipulating it into place and pulling back the excess four metres of cable is likely to put a bit of strain on it - certainly it will need a further insulation test before powering up.

. . . hopefully that's one more milestone ticked off the list 


 

[modeng200023]

There is always something trying to trip you up 
John

[awemawson]

John, if life was simple it wouldn't be fun 

[russ57]

are you sure the electrons can go down the wire in the wrong direction?

[hermetic]

I would leave the jig on whilst you get the cable into place, if thats possible. I once had to attend one in the middle of a field, and wondered how I would find it, don't you worry bout that said Mr farmer, you just stand there and watch yonder, and toddled off to switch on the power in his deep litter hut, After a few moments, a little column of steam could be seen rising from the ground, and it was promptly dug up, to find that the cable had not been laying flat on the bottom of the trench when the joint was made, and as a result the joint had cracked across the resin, I was able to cut 6" off both cables and still have enough to remake the joint. It has given no further trouble, but I must admit those jointing kits are not my idea of fun! This one was the type which used four little line tap type connectors with plastic boots to cover them.
Phil

[awemawson]

Well Phil great minds seldom differ - that's EXACTLY what I did !

Oh - Boy what a day - lots done but long and intensive.

While I was waiting for a friend to help me push the four metres back down the hole I made some careful measurements and cut 2.4 metres off the 10 metre flexible conductor, and fitted a further 125 amp fitting to convert it into the 'short umbilical'.

Then my friend turned up, and just as Phil suggested I kept the cable jig on the cable until the last minute as we pushed, pulled and cajoled the cable back down that hole. Once this was done I cleated either side of the resin joint fixing it firmly to the wall inside the foundry. Then I was able to pull the excess cable into the stable and cleat it to the wall at a height that hopefully will stop it being knocked ready for the end  wall mounted 125 amp plug to be fitted

[awemawson]

Then it was a case of making off the SWA into the male wall mounting 125 amp plug. Having had a problem at the other end I had this time plasma cut some large spanners ready to fit the gland.

Opening up the plug it was revealed that there was no way to make off an SWA gland in it's 'bulkhead' and have enough free cable movement to pull the actual plug unit out to access the cable fixing screws - the whole thing slides into a shroud about 50 mm deep that completely obstructs the cable screws

The only solution the we could come up with was to saw the bulkhead horizontally through the 32 mm gland entry hole and remove temporarily that bit of the fitting. Then the cable and gland could go deeper in giving room to get the cable ends into their sockets AND fix their screws - the the gland and cable could be pulled back 50 mm and the gland seated in it's now halved hole, and the other 'half of the hole' re-assembled before the gland nut was tightened onto the banjo earth fitting and bulkhead. All rather fiddly and essential to get the cable length EXACT or it wouldn't fit. This took three assembly / disassembly sequences before it was right.

Eventually the cover could be re-fitted and that fortunately also serves to clamp the two halves of the bulkhead back together. Time to screw it to the wall . . .

. . . . arg . . . Houston we have a problem . . I'd assembled it 180 degrees out in the casing. The socket that mates with this plug has a spring loaded lid that needs to be away from the wall or otherwise there is no room for it. Not just a case of rotating in the housing as no way would this cable rotate 180 degrees in less then a few metres 

Fortunately a friend pointed out that as the plug / socket will effectively be permanently mated the lid is redundant - take it off. So I did 

At this point we did an electrical leakage tests at 1000 volt and continuity check - all good 

Time to make up the longer umbilical cable. This used the rest of the 10 metre flexible cable so 7.6 metres - I opened up and checked the already attached plug, cleaning up a bit of corrosion, and fitted a free socket to the other end. This umbilical runs from the wall plug up along the RSJ structure of the stable and down to the socket on the generator.

Gosh - fixed wiring FINISHED 

Time was getting on and domestic pressures were mounting, but I did manage a very brief test - again getting a bar glowing in the crucible. I ran everything up to about 60 KW and nothing went bang or got hot apart from the contents of the crucible.

More testing tomorrow 

I have managed to source a large three phase RCD unit that has selectable over current tripping (125 to 160 amps) and  selectable leakage current tripping (30 mA to several amps) so I will be fitting that to the generator for peace of mind when it arrives, and I also want to improve the earth bonding of the generator, the Furnace Driver, the new Chiller Unit and the particular Furnace Body that's in use. I'm intending to also earth bond the steel structure of the foundry building to these units and the steel structure of the workshop.

[awemawson]

Today was largely taken up with other domestic matters, but a run of surface drains that had become clogged with pea shingle afforded an opportunity to use the recovered gravel back filling the cable trench.

This is overkill, 'as dug' back fill would be perfectly adequate so long as sharp stones are removed, but as it was there the cable got bedded in the shingle, then a few inches of earth followed by a row of roof tiles (anti spade penetration!) more earth, then a warning tape before fully back filling to surface level with more earth.

No doubt it will settle, and the far end by the stable is still 'open' as there are water modifications to do before closing the trench. When that's done it can all be re-seeded, and a bit of normality regained!

[awemawson]

So let's try and actually MELT some metal 

First I carefully scraped out the crucible that is embedded in the inverting 'Furnace Body' and removed some iron slag and very small traces of copper slag. (Remember that I installed this crucible brand new many years ago, so this confirms that in the past I have melted both iron and copper with this beast). It's an awfully long time ago and I've had various furnaces over the years, so what was done in which gets a bit hazy !

Then, as I had no mould prepared, I selected a conventional crucible from my stock that would sit inverted on top of the furnace body, and when clamped and inverted would receive the molten metal - let's call this the 'drop crucible' (and hope we don't !!!!)

A test clamp, invert, release and all seems well so I charged the main crucible with a bit of copper scrap left over from the installation.

[awemawson]

So :-

A/ Start the generator and close it's isolator

B/ Turn on water stopcocks from Borehole, which automatically starts the borehole pump

C/ Power up the New Chiller Unit and check for water flow

D/ Turn on the massive main switch on the CFEI  and 'Ping' to measure resonant frequency

E/ Check Compressed Air is 'On' to Furnace Body

F/ Change CFEI Keyswitch from 'Test Frequency' to 'Normal'

G/ Wind the power control to minimum and press 'Chauffe' keeping fingers crossed!

H/ Slowly turn the power up to initially 67 KW then 75 KW listening to the generators 'note' getting harsher.

I/ Watch things starting to smoke then glow very hot.

Now I have no recollection of this happening previously, but a light saying 'Limit' comes on around 75 KW and no further increase seems to be allowed - it must be a pre-set parameter but as I say I have no recollection of it. In fact the display tops out at 77 KW.

So, watching the pot, things are getting extremely hot but no apparent melting happening - the long copper pipe had a dribble of molten metal down its side which formed a cut in the tube, but it's not forming a pool in the base of the crucible as I had expected.

I left it running at 77 KW for a while, and went to inspect the generator which was running magnificently but nothing looking untoward - I took a picture of it's control panel showing 150 amps of load but it's come out very fuzzy - either I was shaking or the camera didn't focus!

The chiller unit using the bore hole water seems to be working fine - I think that the maximum return temperature from the furnace was 34 degrees C - it's picture DID come out OK!

So, ramping the power back down I shut everything off and inverted the furnace body to drop the result into the drop crucible.

When it had all cooled I could then examine the results - each pipe section had. like the long one, a melted bit down it's length that had turned the complete circumference into a 'C' shape.

My theory -and PLEASE COMMENT if you have other theories is as follows:

The furnace melt metals by inducing eddy currents into the charge by transformer action, the charge effectively forming a short circuited secondary winding. Here the pipes having broken the circle are no longer carrying the major current that they were.

I think, given a reasonably dense initial charge that will form a pool in the base, and absorb much of the power, further small stuff will melt into the pool quite satisfactorily

Time for more reading and experiments, but I'm being called away on 'other matters' at the moment.

[russ57]

Can you start with something dense, as you suggest, but maybe aluminium or even a small coil of copper wire?

-russ

[awemawson]

If at first you don't succeed, try, try again 

I've had a lump of what I'd thought was copper at the back of my bench for years - rescued from the scrap of my foundry when I moved all those years ago, so I thought I'd try and melt that - about 5 kg.

Well it was too large for the crucible on the inverting furnace body so I started cutting it in two on the band saw. I wasn't entirely surprised when the blade broke - copper can be gummy, it was over an inch thick and far from stable on the saw table. Never mind, finish it with the Angry Grinder with a 1 mm disk. Now soft copper and disk cutters don't go well together, so why was it cutting so well Because it's BRASS  It had only looked coppery as the surface had been de-zincified.

So now it fits in the pot I went through same rigmarole turning everything on, ran it up to 75 KW - took a decent picture this time of the generator panel (!) returned to the foundry and it was already practically melted. Another less than a minute and the zinc was boiling off, time to invert the furnace.

I got a little bit of brass splash as it went over - I was probably a bit too slow doing it - then the furnace driver gave me an error light. (Water pressure or temperature to / from furnace body but not sure which)

Everything turned off, drop crucible released, and a 2.5 kg dome shaped lump of brass was deposited on the floor.

It was at this stage things got a bit hectic as one of the brewers hoses that carry the chilled cables from the driver unit to the furnace body decided to burst and squirt coolant onto the floor.

Now this is not entirely a negative, as it will persuade me to replace them despite the not inconsiderable cost, and it is best for safety anyway.

So this escapade had proved that the furnace is fully working, albeit that I have a few issues to correct.

Quite impressive rate of melting - 2.5 kg of brass in about 5 minutes - actually probably a bit less as I'm taking that timing from time stamps on the photographs.

. . . off now to cost rolls of Brewers Hose 

[WeldingRod]

I think you will have problems melting Things that are only a few skin depths in one dimension.

I've seen that worm tracking thing back when I tried to build a RF shunt to handle a kilo amp using thin brass sheet.  Some kind of field concentrating thing going on that runs away gleefully with positive feedback.

Sent from my SM-G892A using Tapatalk

[awemawson]

Bowing to the inevitable I removed the cables / hose assemblies this morning, measured them up and I reckon in the current layout I can shorten them all by about one metre for a neater 'lay'.

Order placed with Anchor Pumps, who are the people who supplied the  ceramic seal kit for that Grundfos pump by co-incidence, being the best value that I could find.

While I was at it I also ordered some reinforced neoprene 1" bore hose to replace the reinforced PVC that I'd fitted, reckoning that it'll be a bit more durable and things will already be drained down so relatively easy to do.

Bits should arrive mid next week. I now need to work out how to clean the ends of the 70 mm CSA welding cable that passes up the hose to a standard that will let me braze it to the fittings. Acid dip probably. As I recall I made a 'cup' shape on the fitting that passes into the hose and  receives the cable, the hose being pulled a fair way down the cable and clamped to keep it away from the heat. Then the clamp is released, and the hose pulled onto the 3/4" body of the fitting and secured with two 'Band-It' stainless clamps.

[Sea.dog]

Phosphoric acid should do the trick. Or maybe Baker's fluid.

[AdeV]

Phosphoric acid should do the trick. Or maybe Baker's fluid.

Baker's fluid for Brewer's cables?

Suddenly I am hungry and thirsty!

[awemawson]

Very good Ade 

Amazingly the replacement Brewers Hose and also the 1" bore chiller feed hose has arrived this afternoon - I only ordered it yesterday and it wasn't due until next Wednesday !

Now this is good, but I daren't start to re-make the hoses as I've got heavily involved in a bit of water diversion, re-using a redundant 20 mm feed that comes from the domestic garden / garage up to the barn and was crying out to be connected to the bore hole water system for garden watering. Intercepting the pipe down by the garage was easy as a manhole / stop cock marked it's run, albeit 4'6" deep (Remove stop cock, cap pipe, extend other part of pipe up wall to a tap) but to do this the water feed had to be turned off. BUT when I came to turn it back ON it was jammed closed (plastic tap) so although the house still has mains water, nowhere else (including the loo in my workshop or any of the animal troughs  ) does. And getting at that stuck stop cock involves digging up the patio. The other end by the workshop / borehole pump was dead easy, just inserting a 25 mm Tee fitting with branch reduced to 20 mm - and is now all live and awaiting back filling.

. . . so I daren't start on those hoses until this mess is cleared up 

[awemawson]

Thank goodness normal service is resumed   

A friend and I managed to unscrew and remove the entire top of the offending plastic stop cock by extending a special plumbing tap wrench to the depth of the line. Once the top was removed it was relatively easy to unscrew the jammed operating bits and re-assemble them. Once done the stop cock worked normally, but I really feel that plastic underground stop cocks are the pits, and best avoided. I wouldn't mind but they are not at all cheap!

So, on a roll we back filled the excavation for the new borehole sourced tap, and 'tracked' the ground with my JCB803 until reasonably firm. Raking the pea shingle back you'd not know we'd been there 

As far as that job is concerned there is just a 'Butler Sink' and associated hose reel gubbins to move over, screw the tap to the wall . . . . AND I CAN GET ON MAKING NEW HOSES 

[awemawson]

It's a squally wet day here today so no outside work - what a shame I'll have to get on indoors with those hoses 

First off, before anything is cut, I need to sort out the hose lengths - they are not all the same, but I'd made the originals to be the same resulting in them sitting at odd angles - set them up and measure the difference. Turns out that the two at the top (A&B) need to be six inches longer than those at the bottom (C&D) at the Furnace Driver end due to the A&B angled fittings having to clear the C&D ones.

Then I did a forensic dismembering of one end to remind me how I put them together - it seems that:

A/ I've crimped the 70 mm CSA stripped welding cable into a tubular lug, which I've brazed to a length of 3/4" Imperial copper pipe

B/ The copper pipe is brazed into the screw end fitting

C/ The Brewers Hose is pushed on to the 3/4" copper pipe

D/ Three 1/4" Band-It clamps have been fitted

E/ All is then served with black heat shrink tubing

I have all the bits (must check my box of Band-It bits though) bar the heat shrink which is coming on Tuesday

Now I noticed that in the act of pushing the Brewers Hose onto the 3/4" scalped the innards of the pipe leaving a bit of rubber pushed down the pipe, so I must take precautions when I do it this time.

[russ57]

Looks like there isn't a lot of room for water /coolant inside that hose...

Certainly hope the copper cleans up. It looks pretty manky...

So close and then someone orders a couple more meters of tunnel.

-russ

[awemawson]

Yes that was my reaction when I first opened it up just now Russ, but I think what has happened is that manipulating it to cut the rubber has scrunched the weave of the welding cable up making it fatter and more open weave.

The new hoses will be about one metre shorter so I can lop quite a bit off each end, but I did just now go and check that when twisted back there is room, and it should be fine 

[hermetic]

this could partly be the reaction of any sulphur in the rubber leaching out and attacking the copper, especially in the presence of warm water! Same reaction seen in early VIR cables, which was why they started using tinned copper wire in VIR.
Phil

[awemawson]

So I stripped out the cable from one old hose (the burst one, not that that matters much) and un-brazed the crimped lug. I'd hoped to do it with a propane torch but in the end had to use the big guns and dragged the O/A set out. The Portapak had a cutting torch on it but we managed!

OK time to get brave and cut some new hoses, but first I needed to check that the 16 metres had indeed been delivered. It just fitted stretching from the foundry to the end of the welding shop tied between two 1 CWT cast iron weights which not only let me check it's total length (marginally over 16 metres  ) but also mark out the two versions of two shorter by six inches.

Time to pull the copper wire into a length of hose. That went easily using the reel of yellow fibreglass duct pulling rod.

What didn't go easily was getting the 3/4" copper pipe with the lug brazed on into the new rubber pipe -far too tight. It's the lug stopping it, so I'm going to have to re-design the fixing but that's enough for today.