PVC Induction Heating Coil Design (early thoughts)

[RickDelta]

Hello Physics Forums Community! : )

prelude ...

My quest is to develop an alternative to the conventional method of "primer and cement" bonding of typical pressure PVC pipe and fittings.
I want to effect a molecular PVC (polymer to polymer) bond only ....... without any chemical priming or cementing.
To achieve this inductively, I must insert an "susceptor" between the interfacial contact area of the joining pipe and the receiving socket of the fitting.

early concept thoughts ...

This would need be only a very low power RF inductive heater (1K watt max. @ 380 degrees F for 3 second max. bond fusing "on" time).
The susceptor would be a very thin stainless steel micro-mesh fabric screen (think: the inductively bonded sealing foil under a plastic pill bottles' cap bonded to the bottle top ..... but in a screen form).

The pics below shows this micro-mesh fabric screen melted into the "butt end" of a PVC pipe.
This was done with a standard electric flat iron.
This was done to contain the water filtration's TAC polymer beads fluidized within the pipe (media tank).
No primer or cements! .... creates an indestructible bond at no cost or curing time (typically 8 hours)!

This pipe "butt end" polymer melt bonding led me to think ...... could I melt the pipe to socket connection like this as well!

The pic below now shows this "butt end", melted on screen, to its 2" PVC pipe, inserted into a 2"x 4" PVC reducing coupler.
The dark area (running from 2 o'clock to 4 o'clock) is the incerted micro-mesh screen susceptor (normally would encircle the entire O.D. of the 2" pipe ..... but shown here for contrast)

Now, from the outside of the assembled pipe and fitting, I would attach the inductive "work coil" onto the fitting's socket ...... completely encircling the socket.

The "work coil's" design ...

The work coil is to be of a "split ring" design (because it has to place itself around the already assembled pipe and fittings)

My thoughts are to make my work coil based on a segmented motor "stator" framework (see pic below)

The work coil would be mounted in a plumbers "ProPress" like "Jaws" fashion, ..... opening and engaging around the PVC socket.
This portable induction heating system's circuit board would be attached to the split ring work coil frame (about the size of a pack of cigarettes)
The circuit board to be powered by a supercapacitor's discharge.

My initial question in all this ...

Instead of the typical single inductor work coil, can the design be of an equivalent "distributed" multi-inductor/multi-capacitor (mini-tank circuits) surrounding the susceptor? ie: every mini-tank circuit is of each segment on the stator.

If so, would this not keep the very high tank currents off the circuit board ...... and away from the much lower exciter current?

..... your thoughts on this?

 

[Tom.G]

before investing in any hardware.
I suggest you research the specific heat, thermal conductivity, softening temperature, viscosity versus temperature of PVC.

From that information you should be able to determine how much energy is needed to adequately soften the PVC.

Add to that the energy needed to heat the metal mesh.

Also be aware that some fittings are slightly tapered. This may or may not be a problem.

Then at least triple the above total to account for system inefficiencies and any minor 'oops' for things you overlooked. (for instance outdoors in Winter with a cold wind)

Once you get some realistic numbers, you may find that a super capacitor discharge is not adequate.

Please let us know what you find.

Cheers,
Tom

P.S. How will you assure the correct positioning of the mesh during joint assembly?

 

[RickDelta]

@Tom.G
Hi Tom! ...... I hear ya! : )

Your correct in that the socket is tapered ..... and the pipe is not.

Only a third of the bond in a chemically (primer/cement) fused joint is actually a polymer to polymer bond .... the remaining void (empty) area around the pipe ...... is just the cement itself.

The susceptor screen is lightly coated with a PVC sticky power (plastisol).
This aids in retaining the screen in place and the surface of the metal screen itself has a slight "teeth grab" to it ..... further helping it to remain fixed.

The plastisol melts out into the void areas enabling a full socket polymer to polymer bond

I'm hopeful the supercapacitor approach will work as the duty cycle is only 3 sec "ON" max and 60 sec. "OFF" (recover cycle).

....... another alternative approach for a susceptor I'm looking into is using Tuball 814 Matrix (graphene nanotubes in a PVC polymer base) in a paste or film strip.

https://tuball.com/about-tuball

 

[Baluncore]

This aids in retaining the screen in place and the surface of the metal screen itself has a slight "teeth grab" to it ...

How will you make the screen?

If it is a woven mesh, the cross-wires will not have good electrical contact, which will reduce the eddy currents. If the mesh has good electrical contact, it will not conform to the two surfaces being melded.

Teeth, imply that air will be trapped in the mesh and joint space as it melts, that air will be unable to escape.

The glossy surface of PVC does not have free bonds, as they are contaminated by the environment. Before you make the joint, you will need to re-activate the surfaces with a solvent that cleans contaminants from the surface.

 

[RickDelta]

@Baluncore
Hello! : )

"If it is a woven mesh, the cross-wires will not have good electrical contact, which will reduce the eddy currents" - Baluncore

Lets say, we make a wire mesh ...... where all of the continuous lengths of wire threads (going horizontal and vertical), don't even touch each other at any intersection!
........ your saying these wires cant be inductively heated??

 

[Baluncore]

........ your saying these wires cant be inductively heated??

There must be an area penetrated by flux, to generate the peripheral eddy current.
An orthogonal mesh, if it is woven with insulation between warp and weft, has virtually no area and so no significant eddy current or heating.

If, on the other hand, the mesh was made like chain mail, being thousands of small closed loops, then each planar loop would support an eddy current. Unfortunately, chain mail has a significant thickness, due to interlocking by loops on edge, so it will trap air in the joint. At the same time, the hot planar loops will not be in contact with the PVC, only those cold loops on edge will, so there will be no bonding. It will take a fancy magnetic field to heat all the loops.

Friction welding is used for assembly of long polymer pipelines. Maybe you can replace the spinning friction disc, with eddy currents, in small metal loops, that become embedded in the polymer.

 

[RickDelta]

@Baluncore
"There must be an area penetrated by flux, to generate the peripheral eddy current". - Baluncore

so, ...... your saying these metal "wires" can't be inductively heated??

 

[Baluncore]

so, ...... your saying these metal "wires" can't be inductively heated??

Not unless they are thin tapes.

 

[RickDelta]

@Baluncore
so, ..... then you cant inductively heat, lets say, ... a very thin sewing needle??

 

[Baluncore]

so, ..... then you cant inductively heat, lets say, ... a very thin sewing needle??

You can if the needle has a significant area compared to the area of the induction coil.

The laminations in a transformer are oriented so they will have minimum eddy current heating losses. Your mesh is similar.

 

[RickDelta]

 

[Baluncore]

That video shows a small coil with very close contact to the wire being heated. The diameter of the wire is greater than the separation between coil and wire.

You want to heat wires on the other side of the PVC, which will require a significantly greater coil size, and so coil area.

 

[RickDelta]

Agreed! : ) so, ...... we can inductively heat the wires?

 

[Baluncore]

Agreed! : ) so, ...... we can inductively heat the wires?

Not if they are far from your induction coils, which they must be, since the mesh is inside the PVC joint, and the induction coils are ranged around the outside of the pipe.

 

[RickDelta]

"Not if they are far from your induction coils ....."

You did read the proposed coil construction above ...... coils with flux concentrators ..... right?

 

[Baluncore]

You did read the proposed coil construction above ...... coils with flux concentrators ..... right?

No.
Induction heating coils have an air core, and operate between about 10 kHz and 30 MHz.

Induction coils cannot have a magnetic core, because the core would get hotter than the wires, and would screen the item that you wanted to heat.

What is a flux concentrator? How come it does not also get hot?

 

[RickDelta]

In induction heating, magnetic flux concentrators, also known as flux intensifiers or controllers, are materials used to focus and intensify the magnetic field generated by the induction coil, improving efficiency and allowing for more precise heating of specific areas of a workpiece.

https://www.inductoheat.com/an-objective-assessment-of-magnetic-flux-concentrators-2/#:~:text=March 26th, 2021. Magnetic flux concentrators (also,that of magnetic cores in power transformers.

Note: The segmented stator above are coil mounted "ferrite" cores ..... not iron laminations!

 

[RickDelta]

https://fluxtrol.com/magnetic-flux-control-in-induction-heating

 

[Baluncore]

It seems that a flux concentrator is actually a partial external magnetic core, designed to reduce interactions between multiple induction coils by screening.

Go ahead, design your coil geometry, include the flux concentrators, but don't expect to heat the thin wires, that are relatively far away on the other side of the PVC.

 

[RickDelta]

" ........ that are relatively far away on the other side of the PVC." - Baluncore

The gap is only 1/8 inch away ever!

 

[Baluncore]

The gap is only 1/8 inch away ever!

What is the thickness, t, of the wires?

 

[RickDelta]

.0065" (70 mesh)

 

[Baluncore]

0.125 / 0.0065 = 20

 

[RickDelta]

In my PVC "spin weld" prototype below .... the interfacial heating is largely restricted to just the immediate surfaces.
Its all done in less than 3/4 seconds!

(the joint is in between the arrows!)

 

[RickDelta]

0.125 / 0.0065 = 20

....... ok

 

[Tom.G]

If your mesh has the overlaps in electrical contact, you might be able to get enough induced current to get resistance heating of the mesh.

You obviously have done a fair amount of research on the project. (more than most of us have!)

I'm curious how you can ensure 1/8" separation between the exciting coil and the mesh with different pipe diameters.
1) Are you assuming a different coil assembly for each pipe size?
2) What about buried pipes in a ditch?
3) How awkward would it be having to slide the coil assembly along a 20ft. pipe; say between floors in a building?

Possible workaround: make the coil assembly as a hinged item, like handcuffs.​

Please keep us updated an your progress and results, both success' and failures. We like to learn too!

Cheers,
Tom

 

[RickDelta]

I'm curious how you can ensure 1/8" separation between the exciting coil and the mesh with different pipe diameters.
1) Are you assuming a different coil assembly for each pipe size?
Yes!
2) What about buried pipes in a ditch?
Not an issue.
3) How awkward would it be having to slide the coil assembly along a 20ft. pipe; say between floors in a building?
Not an issue! ...... The work coil is a "split ring" design (read above).
Possible workaround: make the coil assembly as a hinged item, like handcuffs.
Agreed! ...... my thoughts as well.

In the above pic,
See the raised circular rings at the top of the pic? .... there would reside the "work coil tanks" (multiple tank inductors and capacitors) within that raised area only.
Notice that the rings are "split" (think: spring loaded hinged wooden clothes pin to open the "jaws" around the pipe)

 

[Baluncore]

I'm curious how you can ensure 1/8" separation between the exciting coil and the mesh with different pipe diameters.

The PVC pipe wall thickness is typically 0.125", the mesh is between the two walls where the joint will be made, so the mesh must be at least 0.125" from the coil.

 

[RickDelta]

The PVC pipe wall thickness is typically 0.125", the mesh is between the two walls where the joint will be made, so the mesh must be at least 0.125" from the coil.

Good Morning Baluncore! : )
I'm a bit lost in your question : (

Schedule 40 PVC pipe is approximately 1/8"( 0.125") in wall thickness (3/4" thru 4" pipe sizes).

so, yes! .... agreed!

 

[Baluncore]

Good Morning Baluncore! : )
I'm a bit lost in your question : (

Good evening RickDelta.
That is not a question, it is a reply to Tom.G .
It may be morning where you are. It is evening here, 2100 hours.

 

[RickDelta]

If your mesh has the overlaps in electrical contact, you might be able to get enough induced current to get resistance heating of the mesh.

You obviously have done a fair amount of research on the project. (more than most of us have!)

I'm curious how you can ensure 1/8" separation between the exciting coil and the mesh with different pipe diameters.
1) Are you assuming a different coil assembly for each pipe size?
2) What about buried pipes in a ditch?
3) How awkward would it be having to slide the coil assembly along a 20ft. pipe; say between floors in a building?

Possible workaround: make the coil assembly as a hinged item, like handcuffs.​

Please keep us updated an your progress and results, both success' and failures. We like to learn too!

Cheers,
Tom

@Tom.G
"If your mesh has the overlaps in electrical contact, you might be able to get enough induced current to get resistance heating of the mesh." - Tom.G

The manufacture of the micro-mesh screen is the next town over from me!
I met with the engineer there and he showed me that they can electronically weld the screens end to end and side to side without adding any additional metals or thickness!
I think this would meet your suggestion of an connected "overlap" now forming complete coils within the screen itself.

 

[Baluncore]

I think this would meet your suggestion of an connected "overlap" now forming complete coils within the screen itself.

I think you are using overlap in a different sense.

Where the individual wires cross, if they are to support eddy currents, and so be heated by the magnetic field, they must contact electrically. Unfortunately, that also reduces the compliance of the mesh to physically follow the joint surfaces.

Where the woven fabric screens are electrically welded, "end to end" and "side to side" without adding any additional metals or thickness, several smaller sheets are being joined to increase the total area. The cross wires will only be electrically connected where the sheets are welded at their edges. That will be too far apart to support eddy current heating.

PVC sheet is usually dielectrically welded, by clamping PVC between the plates of an RF capacitor. That employs the electrical field, not the magnetic field.
https://en.wikipedia.org/wiki/Radio-frequency_welding

 

[RickDelta]

" ......... several smaller sheets are being joined to increase the total area" - Baluncore

Don't understand what your saying here : (
Please explain in another way please.

 

[RickDelta]

A novel susceptor concept for metal-mesh susceptors, designed to achieve uniform in-plane temperatures during induction heating, is documented.

Metal-mesh susceptors require much smaller amounts of magnetic energy for heating compared to bulk materials and, in some cases, the distance problem can be overcome by increasing the input power to the coil.

 

[Baluncore]

Don't understand what your saying here : (

What size are the sheets of mesh you will cut your patches of susceptor from?
How will you treat the edges of the cut patches of mesh?

 

[RickDelta]

"What size are the sheets of mesh you will cut your patches of susceptor from?"
1" wide x 7.85" (circumference) long is the susceptor size for a 2" pipe.

"How will you treat the edges of the cut patches of mesh?"
Don't know yet! Not sure if I have to do anything! I can have welded the ends and sides ..... or have every crossover tack welded.


I'm thinking the steel mesh of the susceptor will have an added strength attribute to the joint itself ! (think: steel rebar) : )

 

[Baluncore]

"What size are the sheets of mesh you will cut your patches of susceptor from?"
1" wide x 7.85" (circumference) long is the susceptor size for a 2" pipe.

I asked for the raw sheet size, not the cut susceptor size.

I'm thinking the steel mesh of the susceptor will have an added strength attribute to the joint itself ! (think: steel rebar) : )

You mean like concrete cancer?
Or corrosion, because there is an electrical connection between the internal fluid and the outside at the pipe at the joint?

 

[RickDelta]

"I asked for the raw sheet size, not the cut susceptor size."
What I saw in a walk thru of the manufacturing plant was metal screen coming off the loom around 8ft wide and ..... by how long? ...... a mile?? no clue!

" ..... because there is an electrical connection between the internal fluid and the outside at the pipe at the joint?"
How so? .... the susceptor is fully embedded within the polymer.

 

[Baluncore]

How so? .... the susceptor is fully embedded within the polymer.

Not so. The ends of the wires cannot be heated because they are not part of the eddy current circuit. They will be exposed to the environment, in and out.

 

[RickDelta]

"Not so. The ends of the wires cannot be heated because they are not part of the eddy current circuit"

....... and why cant they be heated? ... even if the flux couldn't reach the very tip tip ends ...... think : thermal conduction thru the steel! we only need around 385 degrees F !!

 

[RickDelta]

I just now center heated the metal susceptor screen , trust me ..... almost an instant hot at the ends!! .... I had to drop it on the table! : (

 

[Baluncore]

....... and why cant they be heated? ... even if the flux couldn't reach the very tip tip ends

The flux could reach there, but the end of a .0065" wire has insignificant area, so, insignificant heating. It requires a larger closed loop to generate heat from an eddy current.

...... think : thermal conduction thru the steel! we only need around 385 degrees F !!

The 0.0065" wire has insufficient section to conduct sufficient heat. Heat will be lost to the PVC before it gets to move along the wire. The heating time is insufficient.

I just now center heated the metal susceptor screen , trust me ..... almost an instant hot at the ends!! .... I had to drop it on the table! : (

It was not instant. The temperature sensing nerves in your fingers are slow. You would need to explain how you heated it. By induction? From a coil of what size? Spaced how far away?

If I were to reverse engineer your idea, I could make several improvements, but I would still expect to go broke before I got it to work reliably. I am going to stick with the PVC glue, and avoid the insurance and legal problems of faulty welds.

 

[RickDelta]

"If I were to reverse engineer your idea, I could make several improvements"

well ...... I'm sure we'd all would love to hear them! ( but haven't heard any as of yet) : (

Will you share these great ideas with us?

 

[RickDelta]

"Now is my way clear, now is the meaning plain:
Temptation shall not come in this kind again.
The last temptation is the greatest treason:
To do the right deed for the wrong reason."

- T.S. Eliot - Murder in the Cathedral

 

[RickDelta]

This is another possibility for the susceptor .......
Particle "Curie" bond temperature control

 

[RickDelta]

This looks to be the best method for the PVC susceptor :

A special property of magnetic materials is the presence of a Curie temperature, which marks the transition of the material from a ferromagnetic or ferrimagnetic state to a paramagnetic state. A material which is above its Curie temperature loses its magnetic properties, and can
no longer generate hysteresis heating. Therefore, if magnetic particulate susceptors can be designed which possessa Curie temperature and do not heat due to eddy current effects, they cannot be heated inductively above their Curie temperature.

This property allows the creation of specialized susceptors which “automatically” maintain a desired processing temperature without the need for external control.

In this situation all of the heating is due to hysteresis effects.

 

[renormalize]

My quest is to develop an alternative to the conventional method of "primer and cement" bonding of typical pressure PVC pipe and fittings.
I want to effect a molecular PVC (polymer to polymer) bond only ....... without any chemical priming or cementing.
To achieve this inductively, I must insert an "susceptor" between the interfacial contact area of the joining pipe and the receiving socket of the fitting.

You've expounded at length in this thread regarding inductive/RF bonding of PVC pipes using a susceptor. Can you articulate your best-use case for this rather exotic and comparatively expensive approach to joining plastic tubes? How would it be superior to the current most-used alternative to pipe gluing, namely heat welding?

It's hard to beat the simplicity and efficiency of resistive heating, especially when you can buy the pipe welder shown above on Amazon for less than $50 US.

 

[RickDelta]

@renormalize

Hi! : )

"Can you articulate your best-use case for this rather exotic and comparatively expensive approach to joining plastic tubes?" - renormlize

Sure! : )

It's a special OEM product design (TAC media filtration system) i'm doing for the commercial hotel/motel & laundromat industry.
This system requires the assembly of large complex pressure PVC manifolds (elbows,"T"s, crosses, etc) in very close proximity to each other.

With the conventional "primer & cement" process, you can't prime and cement, lets say, 80 fittings all at the same time. You have to work each fitting one at a time, and then hold pressure on each joint for 30 seconds min. as the joint will walk itself apart if you don't!

Think of the fitting to fitting alignment nightmare of assembling all those joints to form a perfect manifold plane and orientation.
With this chemically (prime/cement) bonded joint, you cannot put the connections under test water pressure (120psi) for at least 12 hours of its full cure time!

With an inductive PVC susceptor bonding method, I create a 100% "polymer to polymer" joint the full length of the fittings socket (as opposed to only 1/3 of the socket depth bonding with the primer and glue).

With my proposed system, you fully assemble the entire 80 to 120 piece manifold all at once .... un-bonded!

Only then, do I inductivally bond the entire assembled manifold! (each bond only takes 3 seconds and can be fully pressure tested immediately!)

..... your heat "welding" suggestion suffers much the same attributes as the primer and cement unfortunately.

 

[RickDelta]

@renormalize

"How would it be superior to the current most-used alternative to pipe gluing, namely heat welding?" - renormalize

A great example of this is from another project i've design. "trenchless sewer lateral relining"

I always get a good laugh at my competitor's advertising of their "trenchless" or "no-dig" sewer lateral repair system ........ and the first thing that shows up at your home ....... is a backhoe!! : )

Their epoxy impregnated fiberglass sock sewer liner can't negotiate around the lawn trap underground! ...... so they have to dig a 4ft x 4ft hole in your front yard to cut out the entire sewer trap!

My system "threads" down a 2" "rigid" schedule 80 PVC pipe from the yard sewer vent, thru the trap and out to the townships sewer main sewer pipe (typically an 80 ft run) ..... in a Truly "No-Dig" system!

With my 2" PVC repair liner in place inside the existing 4" damaged sewer lateral, ....... I expand the new liner's diameter out to 4", .... lining the damaged lateral with solid PVC pipe.

The longest length of 2" PVC pipe you can get is 20ft , so, I needed to be able to join multiple lengths of pipe together.

I couldn't use a standard PVC pipe "coupler"..... as this would double the wall thickness at the joints. When I would expand the entire pipe from 2" to 4" inches, the joints would remain "necked down" at the couplers.

If I were to use your suggested "heat welding" to join the pipes butt ends , that process creates a large melt "bead" both on the outside and inside of the pipe! ...... again, doubling the pipes wall thickness.

The best solution looks to be an inductively heated Particle "Curie" bond susceptor. It creates a bonding as strong as the pipe itself and expands like there is no joint even there!

 

[Tom.G]

A case of "Nothing new under the Sun"?

Around 15 to 20 years ago I watched a city work crew "repair" a storm drain a few hundred feet long. The drain was a few feet in diameter and ran underneath a city street. It was all done thru a couple of manholes.

They pulled a steel cable thru the pipe to the next manhole using a small motorized vehicle.

The cable was then used to pull a flexible plastic liner thru the damaged pipe. The liner was somewhat smaller than the pipe, flattened, and wound on a reel on a separate trailer.

Next step was to pump very hot water into the plastic liner, thus expanding the plastic to line the pipe. They had a good-sized boiler on a separate truck.

There was apparently a another drain pipe emptying into the one just repaired/re-lined. So another small vehicle was sent into the drain to clear things up. This vehicle had a video camera and a blow torch, which made short work of clearing the connection.

Cheers,
Tom