Resurrection of a CFEI 100 KVA Induction Furnace

[awemawson]

As some of you may know it has been my intention to get my 100 KVA induction furnace back into a working state. It was all set up and functioning before I moved house nearly 13 years ago, but since then it has been tucked away in the fire proof room I built specially for it when the workshop was erected. Other things have taken priority and it's never been re-commissioned.

Back in Bromley before I moved, I powered it from a 100 KVA Agrecko diesel generator sat at the bottom of my garden, and as here I have a large enough incoming 3 phase supply (160 amps per phase) I sold the generator to fund building works.

Now, so many years later I am loath to slap it across the incoming mains, as if things have deteriorated over the time in storage it could cause considerable nuisance to the local community when I trip the local over head supply. Hence buying and re-building another generator described here :

https://madmodder.net/index.php/topic,12989.0.html

Initially I had intended to house the generator in a 'lean too' extension that we call the stable on the opposite side of my workshop from the fire proof room, but this has the disadvantage of being quite a way away, hence long expensive cable, and a bit remote to shut down in an emergency. Also the floor slopes somewhat and is rather rough concrete.

Investigating small containers / lorry backs / freezer rooms / concrete sectional garages to place in a bit of ground actually next to the fire proof room they all were discounted for one reason or another - mainly cost.

I then had the bright idea to buy a three axle 'plant trailer' allowing me to push the generator into place on this bit of ground, weather proofing it by building a steel sheeted  frame over it - trailer described in this thread :

https://madmodder.net/index.php/topic,13086.0.html

The shine came off that idea when I costed the metal work for the cover and it was coming out at over £600, so this morning I started investigating the cost of cabling (massive cross section) as I reasoned that with a bit of ingenuity I could extend the Emergency Cut Off circuit into the foundry itself.

Now I knew that I had some (short I thought) flexible 5 core 35 mm '07RN-F' umbilical cords already made off with massive male and female 'Commando' plugs and sockets from the original installation but had no idea how long they were. If I bought some more I could perhaps wire the generator across the back wall of the main workshop. Now this stuff is  £26 per metre so I didn't want to buy more than needed.

Having wrestled the coils from whence they were stored (on top of the CEIF driver box - oh boy are they heavy!) It turns out that I have a 10 metre and an 18 metre length - a really excellent find as it's enough without buying more 

My friend Graham (Seadog) came the other day and we were discussing the furnace and looking at pictures that I'd taken back in 2005 of it's internal electronics, and to my absolute HORROR I noticed that there is a 3.6 volt Lithium battery in a holder on the main board that presumably retains some needed parameters, and is now well over 15 years old  I have no idea what is RAM  retained or how the parameters are set, so this could be a show stopper as I doubt information is around on these rather rare animals.

Needless to say we shot out at a great rate of knots, moved things about so doors could be opened, and AMAZINGLY the battery STILL measures 3.6 volts 

So the show is still on - phew

[awemawson]

I think that the next thing to do is get the chiller working - this keeps the water cooling jacket cool for the internal high power electronics in the blue box, and also that surrounds the drive cables for the actual induction coil that is wrapped around the crucible.

I still have an access issue in that space - things have to go, so if anyone wants the huge BOC TIG welder (red box) and the Bridgeport chip tray (on top of the chiller) they can have them for free if they come and collect.

[mattinker]

Hi Andrew,
Looking forward to this thread! Maybe you'll have a pour scheduled around Christmas! Knowing you, you'll have it up and running well before then but I'd love to see it in operation!

Cheers, Matthew

[awemawson]

Matthew good to have you along - oh believe me I too would LOVE to see it working after all these years 

It's a totally illogical thing to do - I got the Induction Furnace because my neighbour at the last place built a house in the bottom of his garden so I ended up with someones kitchen next to my oil burning smelly foundry.

I moved so that I don't have such problems being a bit more isolated here (as you know) so logically I should really have abandoned this furnace and made another oil fired one but the pig headed bit of me is determined to get it going again.

[mattinker]

It's a totally illogical thing to do. What has logic got to do with all this? More fun!

[awemawson]

I put the BOC Transtig  350 on Facebook Marketplace for £1 and not surprisingly had about 20 replies in the first 10 minutes and it was collected within an hour of listing by a very genuine local lad setting up his workshop.

Result 

I can now get at the front of the Flowcool chiller to investigate it AND welder chappie is a happy bunny so a win / win situation all round.

(That trailer went on eBay this morning so fingers crossed - Bridgeport chip tray still waiting for a new home)

[awemawson]

So having disposed of the BOC Transtig - all 200 kg of it - I could now get at the Flowcool chiller to try and evaluate how much it has suffered in storage.

Answer - there is quite a bit of rusting from condensation on the lower - perhaps 12 inches - of it, and quite an infestation of spiders evidenced by some impressive webs. So long as the actual motors / pumps / etc have survived it 'should' be OK but I want to do quite a bit of clean up before I introduce it to electrons at 415 volt three phase !

As I recall, (and I need to refresh myself on this) there is an internal water tank open to atmosphere (with a loose cover) that is cooled by an enormous 'fridge type sealed compressor. There is (I think) a heat exchanger immersed in the tank that holds the fluid that is externally circulated by a large multi-stage pump in the base of the unit. Air is drawn through large filters on the sides, passes over the refrigerator radiator and out to atmosphere through the top of the unit.

So on the rear there are four water ports:

A/ Flow and return for the chilled water that keeps the bits of the Induction Furnace cool
B/ Top Up mains water to allow for evaporation of the tank, controlled by a simple float valve (Like a Loo cistern)
C/ Overflow from tank in case float valve misbehaves.

The whole lot is controlled by a temperature controller and a bunch of relays, transformers and over load cut outs housed in a cabinet on the front.

I can't get too involved in going much further today due to other commitments, but I want to give it a good blow out with an airline, then set up an air blower into the electrics to try and reduce the moisture before feeding it power 



 

[awemawson]

This morning I rigged a long airline extension to blow off the majority of the cobwebs and rust flakes from the Chiller.

To my horror I found that I was spraying significant amounts of water with the air - the vertical drop of my air line system that I was using to connect the air hose extension hadn't been used since I installed my total air drying kit, and must have previously accumulated water. Blasting air to waste quickly disposed of it, reverting to delivering dry air letting me carry on.

I'm pleased to report that a basic electrical safety test passed. Good earth bonding to the chassis from the plug & lead, and no measurable leakage to ground for the connections to the compressor, pump, fan or system transformer. NB I was measuring to earth from the outputs of their respective contactors, so effectively directly to their windings.

Next hurdle to overcome is aquatic. The reservoir tank that holds the main heat exchanger needs filling for the beast to work. I have no mains water yet plumbed to this room, and the tank top is too close to the base of the control cabinet to pour in from a container - perhaps 2 inches - it could be done with a hose, funnel and bucket but the other issue is that I need to be able to drain the tank to move the chiller as working is done on it and around it, but as yet I've not identified a drain plug. I must have done in before moving it here, but for the life of me cannot remember how - maybe an external pump?.

The the flow and return pipes that come from the main pump need filling with water and making into a circulatory loop system. If I don't do this there is a probability of destroying the pump ceramic seals if run dry.


. . . still it's progress of a sort 

[awemawson]

So next I rigged up a loop for the flow and return and primed it with water, and then pumped several buckets full of water into the main tank - all good so far.

I then connected up to 415V, crossed my fingers and switched on. No dramas and an indication on the temperature control of current tank temperature. Still all good until the compressor - that BIG blue compressor lump switched on, and the MCB for the 415v tripped. OK it's only a 16A 'B' rated one. I upped it to 32 amp, again a 'B' rating as that's all I had - still trips. Without the compressor running the load is only 5.7 amps per phase.

Pushing my luck I installed (as a temporary measure) a 50 amp 'C' rated breaker being the only spare 'C' that I had to hand. Works fine - running current is about 15 amps per phase AND the temperature is dropping. It got down to 7.1 degrees when I heard the dripping of a significant leak 

Main pump seal is FUBAR so I started siphoning out the tank, but a fair amount is going to end up on the floor to evaporate 

[Will_D]

Hi Andrew,

You make my "little problems" seem totally insignificant!!

Keep up the great work, "Nil Illegitimum, Nil Carborundum"

Regards to the pigs!!

Cheers

Will, Malahide.ie

[awemawson]

Thanks for the encouragement Will, I’m not looking forward to rolling about on the floor in a confined wet space removing that pump this morning.

. . . But it has to be done!

[awemawson]

As anticipated, rolling about on the floor getting the pump out was not my most enjoyable experience, but, never mind, THE PUMP IS OUT 

Leaving a powerful fan blowing in the area overnight had done a remarkably good job evaporating most of the water. The plastic pump inlet coupler unscrewed with relatively little trouble but this was not so with the one on the outlet. I tried every trick in the book - gentle tapping, gentle warming, and increasing levels of violence, but it was obvious that it wasn't unscrewing without breaking (plastic remember). I suspect that some of the solvent glue used to assemble the pipework had inadvertently got into the thread of this fitting. Angle grinder and hack saw removed it allowing me to get on disconnecting the electrical connections and unbolting the motor base from the chiller chassis.  The four bolts were fortunately stainless, not galled solid, and not too tight - lucky as access underneath is challenging.

Having got the pump on the bench and now able to read it's label it turns out to be a Grunfos CH4-60. Options are replace (at least £500 with the VAT ) or get a seal kit (£143) Interestingly the one for Glycol cold water version is cheaper than the standard one 

I will have to replace a bit of the existing plumbing - working back from a stub of original 40 mm plastic pipe, the branch leads to a 1/2" BSP re-circulation valve (controls pressure) and to the pump that is standard 1" BSP so simple enough.

Time to go and get changed - covered in concrete dust  - yuk !

[awemawson]

With the spare parts not yet to hand I'm not going to dismantle the pump, however I wanted to make sure that I can when the time comes, so am applying PlusGas to all the various bolts and screws. All now move except the two short pan head screws holding the fan cover, hopefully a bit of soaking will loosen them but if necessary they can be drilled out.

Removing the 1" BSP plastic pipe screwed into the outlet revealed a curious mangled spring which is part of the ceramic seal assembly which explains why the leak was so gross - usually they just weep when they fail.

I turned off the old plastic joiner and threaded the pipe 1" BSP to raise the knuckle joint above the electrical connector box.

[awemawson]

Of course the inevitable happened, and one of the fan cover screws sheered off, and the other was so mangled as not to be considered worthy of being used.

Simple matter to mount up in the Bridgeport, drill out and re-tap M4 and replace with stainless button cap screws. It is possible that had I left the PlusGas a bit longer they might have loosened, but I think not, and I wanted access to the fan to be able to rotate the pump and get a better look at that curious displaced spring. It rotates with the pump shaft and I'm fairly certain that it is the one that presses the two halves of the 'mechanical' ceramic seal together to make a leak proof joint.

Hard to work out how it got into this state - I can only think that the pump has ingested something that got caught on it so hopefully when I dismantle the six stages of the pump there won't be more damage to contend with 

[Sea.dog]

It's good to see you're under way with this project Andrew. Looking forward to the inaugural firing up 

[awemawson]

Looking forward to the inaugural firing up 

You and me both Graham, but a long way to go yet I'm afraid !

[awemawson]

I've just been watching Martin Zero's video on YouTube exploring James Brindley's 'siphon tunnel' diverting water from the River Irwell  under itself to drive mine pumping equipment in the 1700's - fascinating video, but he used an innovative way of getting his camera into a dangerous place (strapped to a model radio controlled tank!) - and it set me thinking . . . .

I want to investigate inside the 'chilled water storage tank' that feeds the leaking pump - I have an endoscope camera - will it work .... 

Well the answer in NO - focal length is quite wrong and cannot see anything but close ups of individual elements of the finned radiators

However - just the job to take a proper look at that broken spring on the ceramic pump seal - so here is the video with the endoscope poked down the outlet port of the pump:

[hermetic]

I watched Martins video too, fascinating stuff! On the pump, looks like the spring has caught on the back of the seal spinner and got unwound, not a failure I have seen before, and I have done a few pump seals! There is usually a stainless washer at either end of the spring to stop it catching.
Phil

[awemawson]

The pump motor cowl and base plate were quite rusty so I decided as the seal parts still hadn't arrived, I'd grit blast them and give them the usual zinc rich primer and satin black top coat, to make them less offensive to the eye as they sat on my bench.

I've recently made two alterations to my grit blaster that have made it significantly easier to use:

 - Firstly I've fitted a long thin axial fan (ex photocopier) to blow across the cabinet window to remove the condensation from my breath that was a major obstacle to seeing what I am doing.

 - Secondly I've plumbed my 104 CFM huge 18 kW Hydrovane into my air system so that I can blast at full pressure continuously without having to pause every now and again for the smaller one to catch it's breath.

All went well and paint is now drying, but guess what arrived whilst I was blasting - yes, the seal kit 

So the next job is to dismantle the pump in a controlled fashion to keep the myriad of pieces in the correct order for re-assembly. Six stages each of which have a housing, an impeller, a gasket and two spacers.

. . .but first I must study the instructions 

(Various plumbing bits yet to turn up so no huge hurry)

[awemawson]

Opening the seal instructions attached to the (vastly expensive) seal kit, there is no illustration of the specific seal used on the CH4-60 - there are illustrations of similar pumps but the ends of the spring have outward facing pegs on the end turns, whereas the kit and the pump that I have has a plain spring. So the location methods obviously differ for the bit that rotates.

It probably will be perfectly obvious when I dismantle it, but my experience of these pumps tell me it's best that you have the CORRECT drawing before bits fly in all directions (*) 

Grundfos 'are surprised' and are trying to find a copy of the correct illustration to email me so I won't pull it apart at the moment.

Meanwhile the paint has dried on the end cover and mounting foot so they have been temporarily re-assembled to get them out of the way until that drawing arrives. I must say the pump looks cosmetically far better than it did

(* My first introduction to Grundfos pumps was when one failed in one of my Launderettes, and I got a panic and embarrassed call from the engineer who did my maintenance to say he'd dismantled the pump to replace the seal and couldn't work out what went where. I was presented with a plastic tote box with an entirely dismantled CR8 (8 stage) pump and spent the next couple of hours working out what went where - a baptism of fire - or water you might say!)

[awemawson]

So casting all caution to the wind I started dismantling the pump.

The first stage casing was firmly stuck to the inlet casting, somebody has used a thread lock I think. Gentle heating and a tap or two eventually got it loose to the stage you see in the first picture.

The seal / spring assembly is now exposed and actually it turns out the original definitely had bent over ends of the springs going into locating holes, whereas the kit does not, so more investigation needed.

. . .but supper calls . . . !

[awemawson]

I got the old seals out. They are quite different from the replacements in the way the spring is compressed and rotated.

The original has the ends of the spring protruding and engaging in a hole in the other parts, with quite a compact arrangement to house the shaft O ring. The replacement has a plain spring and pressed metal parts that is considerably larger. I can see that it probably fits, and may be an upgrade but there is no mention of the difference in the instructions, nor does it mention why there is a black and a green version of both O rings. However the original had green so that's what's going back in!

(Google tells me green O rings are Hydrogenated Nitrile and are intended for high temperatures and stresses!)

Perhaps I should explain that the static part of the seal is sealed to the motor casting with a large thin O ring, and the rotating part of the seal is sealed to the shaft with a small fat O ring. A ring clip sits in a grove in the splined shaft holding the assembly together as the rest of the pump is assembled which gives the final compression to the spring.

Everything needs a good clean up before reassembly - there is a distinct 'tide mark' where it has dried out down to just below the shaft level, leaving a rust and debris conglomerate that needs to be removed from most of the impellers and the end housing to ensure that it doesn't get pumped round the system when it's back together - oh joy !

[awemawson]

It seems that ceramic / mechanical seals are a very complicated but fascinating subject. I came across this treatise from Grundfos that I found very informative. Amusingly it illustrates the very pump I'm working one on one of its plates.

A bit of reading for those of you self isolating as I am !


https://api.grundfos.com/literature/literature/Grundfosliterature-5768950.pdf

[nrml]

Whoever wrote that paper did a brilliant job of demystifying the science and engineering and presenting it in a very readable and easy to comprehend manner.

[awemawson]

Yes it certainly sorted out my understanding of rotary seals!

So now the big clean up begins. Frankly it's not surprising that the spring and seal failed - all inner stainless steel surfaces are covered in a light coating of deposited rust, as are the two end castings, and there is a 'tide line' of something very varnish like that has set rock hard.

I gently grit blasted the inner surfaces of the cast iron inlet and outlet ports, masking up the motor, seal area and shaft to avoid damaging them. It worked quite reasonably, but I had to hand scrape some as I didn't want the blasting pressure too high to avoid ingress of grit where it wasn't wanted.

I'd intended to continue gentle low pressure blasting of the stainless steel pump vanes and housings, but it seemed a shame to lose the polished surface even though I don't suppose it would interfere with the pump operation. Experimenting with a specialist corrosion remover that came with my Wire CNC EDM  machine it worked extremely well apart from the varnish like deposit, so at the moment all the vanes, housings and spacers are soaking in a washing up bowl of hot water and Altrans (by DEB) that is mainly phosphoric acid with a wetting agent added  -  I had to top up with neat phosphoric acid as the Altrans was insufficient to cover everything.

Probably leave them there for a few hours to cook in their own juice.