How to heat the end of copper pipe

[marcophys]

I like the new software... the email feed was a good idea, and it coincided with an interesting problem I've come up with, that I thought I'd share:

A copper pipe:
14mm OD, 12mm ID, 45mm in length,
with a 2mm thick x 18mm OD shoulder.

The type that comes with a nut, for connecting to the bottom of a tap (often sold as a tap connector).

I'm thinking about baking a cement (at the shoulder end) to around 125deg C... think a 'bung' 5mm thick of cement.
I can't put it in the oven, because there is PVC sheathing at the other end.
The PVC can be pulled back from the copper tube... but it can't be removed entirely.

The shoulder is a bit of a bugger, because it stops me from just dropping the tube end into a close proximity circular resistance coil.

I could drop the end into a 19mm ID circular resistance coil... but the coil would be 2.5mm away from the tube I'm trying to heat.

Perhaps it could be a resistive cup... providing heat from the bottom (to the tube/cement face), and circular, around the tube diameter.
(Using a two part cement works, but clearly this is not an electrical engineering solution.)

I have in mind a scenario, where the tube is placed, and a button is pressed that triggers a timed dose of current, baking the cement.

Easy to say... but figuring it out is another matter.
:)

 

[Simon Bridge]

Use a different kind of bung?
What is stopping you from pushing the pipe through a coil from the other end - the end away from the shoulder?

 

[zoki85]

Perhaps it could be a resistive cup... providing heat from the bottom (to the tube/cement face), and circular, around the tube diameter.
(Using a two part cement works, but clearly this is not an electrical engineering solution.)

Why not?

 

[Baluncore]

How long will you have to maintain 125 deg C to cure the cement?
Why do you not braze, silver solder, or use a compression fitting?
Why not hammer in a wooden plug?
Do you really need to reinvent induction heating? http://en.wikipedia.org/wiki/Induction_heating

 

[Windadct]

Concur - Induction heater - induction range top will probably work.

 

[marcophys]

Use a different kind of bung?
What is stopping you from pushing the pipe through a coil from the other end - the end away from the shoulder?

It is possible, but messy, as there is a one meter length of PVC sheathed cable, that would first need to be passed through.

Perhaps it could be a resistive cup... providing heat from the bottom (to the tube/cement face), and circular, around the tube diameter.
(Using a two part cement works, but clearly this is not an electrical engineering solution.)

Why not?

If the 'Why not?' refers to a resistive cup... yes maybe it's feasible but what resistive sheet metal is available...I tried a search and nothing came up that I could see... any thoughts?

If the 'Why not?' refers to the two part cement... That's the current option... but it takes time to go fully off, is messy to mix... and the cement is being used elsewhere... and the cement would be immediately and consistently hard after the bake.

So this problem is worth looking at.

How long will you have to maintain 125 deg C to cure the cement?
Just till the water boils off.

Why do you not braze, silver solder, or use a compression fitting?
Why not hammer in a wooden plug?
None are applicable as the bung totally locks glass fibre.

Do you really need to reinvent induction heating? http://en.wikipedia.org/wiki/Induction_heating
This solution could have legs... see below:

Concur - Induction heater - induction range top will probably work.

The above image from Wiki highlights this usage.
What is the coil material?

The above image makes it look easy.
However, for the 'Top Hat' arrangement of the 'copper tube and shoulder'... we are only heating a minimum of 5mm (but a bit more could be heated).

Could a finer coil be used?
I note the above takes 15Kw however is a big lump of metal... my application is just 1mm wall copper tube.

Note: http://theinductionsite.com/proandcon.shtml
Paragraph 'The Cooking Vessels' states:

"The most obvious and famous drawback to induction cooking has already been mentioned: it only works with cooking vessels made of magnetic materials."

Does this fact 'preclude induction heating' for this application?

Hmmm!
Just popped out, then had a thought:
Might it be possible to use a thin wall ferrous tube, welded flat, to make a ferrous cup.
This would be induction heated, convecting the heat to the copper tube and cement face.

This cup could be plastered externally with the same cement to act as an insulator.

How does that compare to the 'resistive cup' idea (if that was at all feasible?

 

[Windadct]

The coil is the easy part - just copper - usually a tube and is liquid cooled ( there is coolant flowing through the coil)... is this for production or a one off? For production Inudction is very repeatable - for a one off, it may be more troube than it is worth, since a range top may not work. I worked with Induction heating companies and their systems typically are suitable for copper, however I did not realize the typical range top is not...probably due to the frequency used - so working with copper may be beyond the range of a standard range top - you could make a steel jacket that gets heated. Again an image of your workpiece would be helpful - 125C is really not too hot...

 

[marcophys]

The coil is the easy part - just copper - usually a tube and is liquid cooled ( there is coolant flowing through the coil)... is this for production or a one off? For production Inudction is very repeatable - for a one off, it may be more troube than it is worth, since a range top may not work. I worked with Induction heating companies and their systems typically are suitable for copper, however I did not realize the typical range top is not...probably due to the frequency used - so working with copper may be beyond the range of a standard range top - you could make a steel jacket that gets heated. Again an image of your workpiece would be helpful - 125C is really not too hot...

Below is the copper tap connector tube with the nut removed.

Below shows the steel insert currently two part bonded.

I had forgotten that I already have a steel tube insert in the copper tube, to constrain the glass fibres.

Perhaps this steel insert could be first bonded with cement and induction heat baked... nice and quick.
Then the fibres and cement could be inserted, and the steel insert could be used again for the induction heating process.

Note: Just realized... it is stainless steel tubing to prevent corrosion... so I guess that's a no go. :(

So it may be back to the ferrous steel cup idea...

Oh yes... it would be for repetitive production.

 

[Windadct]

Not all SS is non-magnetic - seems like an easy thing to check. For production ( what volume) - production rate, quality - etc..all good factors for induction heat..but then the total investment needs to be looked at. I am thinking you could also just run current though the copper to heat it - the trick will be the process control. I checked e-bay - seems like a few items may work, I can not see this needing more than 1000W. - for more info try these guys - http://www.ambrell.com/aboutinduction.php - they are a client (as part of disclosure) but I think your requirements may be too low power so they probably do not have a product for you - just info.

 

[marcophys]

Thanks for the link.
The steel insert is 6mm deep.
Given it was ferrous, it would supply the heat exactly where it was needed.
The inox steel is non-ferrous.
I guess finding suitable ferrous inox steel tube, might be a bit difficult.

Everything has been designed to enable small scale assembly, as initial batches sizes will be to get the product directly into the hands of first takers.
Assuming successful reception, and with hard user data, a sales/production strategy would develop.

Given decent quantities, the ferrous steel ring could be made to the perfect size.

Looking on ebay at induction heaters, I note one (12v - 30v DC) is UK made for £45

http://www.ebay.co.uk/itm/Induction-Heater-Circuit-Solid-State-Tesla-Coil-Driver-Power-Resonator-/151420326474?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item23415ad24a

This one (12v - 48v DC) at £42 or offers (looks similar to the Chinese one below)
http://www.ebay.co.uk/itm/12-48V-DC-1000W-Induction-Heater-Flyback-Driver-Zero-Voltage-Switching-/261480898122?pt=LH_DefaultDomain_3&hash=item3ce179de4a

This (12v 24v 36v DC) Chinese one at £17 (!)
http://www.ebay.co.uk/itm/12V-24V-36V-Flyback-Driver-Board-Zero-Voltage-Switching-Inductive-Heater-/331166645290?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item4d1b11fc2a

The UK one appears much simpler, with less components.
Can you discern if it would do the trick?
It is certainly cheap enough to try, and simple enough to repair (by the looks of it).

They provide an example:
If you consider that using a 24V input and your system pulls 10A, Then there is potentially 240W of heating power in the system (ignoring losses)

It's all new to me, so it's not clear whether this system could be setup to induction heat a 'small steel band'.
Perhaps if the piece sat on a small steel disk... it would help?

 

[marcophys]

On further examination... the UK bod has included 3 pictures - the 2nd of which shows his setup to heat a steel nut.
I think I will email him.

 

[marcophys]

But... while waiting for that reply... what about directly running current through the copper (without successfully pulling the main trip switch :) ) ?

 

[Windadct]

It will take a good amount of current, but very low voltage, looking at the size and low temp 1000W may be a pretty good estimate. ... If someone has a welder try running ~ 200A (AC or DC - but AC may be a little more effective) across the part for a few seconds and see how warm it gets. Make 2 electrodes to clamp just above the shoulder (although ideally to the SS insert below the shoulder) - I would start with about 1/4 Circle each or a V clamp - even if the contact patch between the electrode and the copper is the higher resistance (where the heat will be generated - the heat will be localized, the copper will spread it pretty well.

For this cement - you mentioned 125C, how long does it need to be heated. ( what type of cement? - not Epoxy?) --can it be over cooked? What will keep the Cement in place as it is curing

The PVC is at the other end?

As another idea... put a metal plate on a hotplate, in the plate make 10 cups (for 10 parts at a time) that just fit in the part up to the shoulder? (may need a shelf to support the part to not tip over)

Can you post a pic of a part after this process? Showing Cement and PVC?

I used to do this type of projects for my job...

 

[Baluncore]

How long will you have to maintain 125 deg C to cure the cement?

Just till the water boils off.

Then why not apply the cement, place the unit in a sealed chamber and pull a partial vacuum. The water will boil off at the lower pressure without needing to increase the temperature.

 

[marcophys]

Ah yes... a welder... and I've got one of those :)
That's an interesting suggestion... thanks for that, and Baluncore for first raising induction heating... and thanks to everybody for offering suggestions.

Yes, I could cut two vee electrodes... out of what... mild steel?

The Cement
What will happen is to be seen.

In a 'tube within a tube' scenario, (say 10 OD, 12 ID) the cement will tend to eject itself (at cure point)... at which point; cutting the heat is fine, and the job is done.
Clearly we can envisage a vapour expansion, that when arrested, leaves the cement cured.
Thereafter, future heating causes no such ejection.

The beauty of induction heating is that it might offer the potential for a controlled bake, to point of cure, and then stop.

However I believe that this may be less of a problem, in this 'compression scenario within the SS insert'.
My prediction is that the cement moisture will 'wick away' up the encapsulated fibres (that are being cemented).

If this prediction is correct... then the 'cup method' would work, as there'll be no need to watch out for ejection.

You know what?
Sod the welder... what about if I just blast it with a heat gun, and then finish it with a butane torch?

AND I've just received a post alert from Baluncore... which I've just read... it's fantastic :)

I think (at first glance) that its a complexity step too far for this project (vacuum boiling the water), but it's a great concept.
(Physics Forums is just too good :) )

Would it even work?
I can't think of how I might test the idea.
But... I also think that a chemical reaction takes place @ boil temperature, because the cement changes its structure.

So neat idea... but I think it's time to actually see what happens when heat is applied to the assembly.
The easiest way must be the heat gun and torch.

Thereafter, the production method might present itself.

 

[marcophys]

The cement didn't work.
The simple fact is that the wicking of water is so rapid, that the cement loses its consistency, almost immediately.

Unlike the bricklayer, who can wet his bricks (to prevent the same problem)... the same cannot be done here.
A brick absorbs moisture, yet remains dry (no free water).
Wet the components... and they just remain wet :D

There appears to be no way around this other than to use 2 part kit, that chemically cures itself.
In such case, capillary action becomes beneficial.

The problem arises from using standard cements.
It may be possible to use purely the binding elements, but this will be for long term research.

This avenue of exploration has not proved to be a failure... rather a successful elimination of a possible assembly material.
The methodology we discussed, is now embedded knowledge, and is sure to be beneficial at some point in the future.
Thanks to everybody for their contributions.
:)

 

[sophiecentaur]

An alternative method could be to use a split 'collar' (two thick C-shaped halves) of steel, fixed to some scissor tongs, to hold it and squeeze it in place. This could be heated and held round the neck of the tube where the cement its. You could apply several times if necessary. Very crude but, if you could fit a collar with enough thermal mass, it would work (like the original 'soldering iron' which has been replaced by the electric model). Smart use of insulation (outer and inner) could give the right range of temperatures and extend the heating period.

 

[Windadct]

As for the ejection issue - is there a primer available that helps this cement bond to steel? I asked about epoxy vs cement - ether way the reaciton tends to be exothermic - once it gets going you may not need to add much heat...

 

[marcophys]

An alternative method could be to use a split 'collar' (two thick C-shaped halves) of steel, fixed to some scissor tongs, to hold it and squeeze it in place. This could be heated and held round the neck of the tube where the cement its. You could apply several times if necessary. Very crude but, if you could fit a collar with enough thermal mass, it would work (like the original 'soldering iron' which has been replaced by the electric model). Smart use of insulation (outer and inner) could give the right range of temperatures and extend the heating period.

I had pictured this exactly as you state:)
But I think it is best achieved with symmetrically split bracket/tube affair.
Self pinching on C clamp forced close... a la modern plastic assemblies, only with more force.

This would solve numerous 'assembly manipulation' issues (read nightmares :) ) and would allow 'cement packing' throughout the bracket, thereby eliminating wicking problems associated with small quantities of cement (a fundamental problem as it stands).

To be honest... this has to be the ultimate goal, once the product has proven it's financial viability for such tooling.
Until then, I do fear that there is a considerable amount of 'assembly manipulation' ahead :(

Witness to that is the piece, right behind me now; quietly curing in it's two part bliss (for effectively 24hrs). :)

As for the ejection issue - is there a primer available that helps this cement bond to steel? I asked about epoxy vs cement - ether way the reaciton tends to be exothermic - once it gets going you may not need to add much heat...

I hear you... but chemical reactions do vary.
In the case of this cement - no 'self-perpetuating reaction' is achieved.
It 'kicks off' @ its temperature.
Cut the heat... and it stops.
That fact alone is quite remarkable, because normally you'd think it would keep going (at least a bit)... but no.
Open the oven door, and it stops.

My other chosen cement formula had no such issues... but both were water based, and...
... as my post-test post stated... there is a fundamental problem with this format... probably only solved through compression packing a la previous para above.

I will admit... I did try Cyanoacrylate adhesives (can you blame me :) )
They got me through 'proof of concept' :k

The problem is that, when they 'wick'... they just go off instantly, and nothing more gets through.
Deep penetration is so much more preferable... but how to achieve it?

Again, we are up against the laws of physics.
A cement based upon water... but you can't keep the water.
A glue hitting a few molecules of water, and it goes off.

Actually... talking of physics...
... If the piece was 100% dry... Cyanoacrylate adhesives would surely work.
Wow!
That is true no?

If there is 'no moisture present' a Cyanoacrylate adhesive cannot go off.

Presumably, one could then stick it in water... or even better... water and bicarbonate of soda!
I wonder if it would cure all the way through?

What do you think?