I was basing this design largely on the principals behind building a hotwire cutter. I should have clarified that my needs would be met by stamping the design in my original post. I apologize for any confusion I caused for all of you helping me with this.

I could potentially scale my stamping size down to 12" X 12" instead of 12" x 48" if that would make this project more feasible. I would really like to find a single power supply capable of heating the wire rather than multiple small power supplies.

 

Why not a conveyor of mesh, coming out of heater (furnace), and quickly stamp foam on mesh?

You might be able to stamp multiples of panels at a time. By time you reload, fresh hot mesh is there.

 

I was basing this design largely on the principals behind building a hotwire cutter.

I could potentially scale my stamping size down to 12" X 12" instead of 12" x 48" if that would make this project more feasible. I would really like to find a single power supply capable of heating the wire rather than multiple small power supplies.

Once you find the Va needed all you need is a transformer of that capacity size and a simple Triac controller to control the temp.
Many have used a dimmer style control on the primary, I have preferred to control the secondary.
Very simple to put together.
Max.

 

There is nowhere you can connect the source of current that will assure equal current in all wire segments.

That's what I thought, but for a chicken-wire array Spice disagrees :-

 

Greetings,

I am building a hotwire cutter that uses a grid of wires instead of a single wire to cut foam. The grid is approximately 12" wide and 48" long and will be constructed of 18g or 19g stainless steel wire. The grid will be divided into 1" squares. I know that a single wire foam cutter can be powered by a 12v power supply, but I am having trouble calculating the amount of power I will need to heat this grid to a temperature suitable for melting polystyrene. Any help will be greatly appreciated. Thanks!

Shouldn't use stainless.

 

I have used S.S. MIG welding wire in a single element configuration for a foam cutter?
Max.

 

That's what I thought, but for a chicken-wire array Spice disagrees

Apples and oranges. Your resistor array does not equate to and x-y square grid of equal-value resistors, because yours has no horizontal elements. If you feed all of the top and bottom sides with 0 ohm wire (copper instead of stainless or nichrome), then there is no current through the horizontal elements because all lateral nodes are equipotential.

If you feed the diagonal corners, then not all paths through the grid have the same number of elements. Hmmm... have to think about that one...

ak

 

Shouldn't use stainless.

Why? What is your first choice?

Here is a link to the Charles River model group. They are in Boston, USA, and a lot of the members are faculty at MIT or Harvard. I have found its site to be quite reliable for advice, and the group has been at it for a long time. Lots of good information there; recommendations for the cutting wire are at the end. Rene is 1st choice, aircraft safety wire (generally an annealed 302-304 alloy with a mil spec stainless) is second choice.

http://www.charlesriverrc.org/articles/foam_vac/mrcss_foamcutter.htm

@MaxHeadRoom : Which alloy?

John

 

I'm really not sure about the best way to measure the exact Va needed to heat this grid up to foam melting temps. What do you guys think about an overkill power supply with a Variac? Slowly raise the current until it starts to heat up?

 

I'm really not sure about the best way to measure the exact Va needed to heat this grid up to foam melting temps. What do you guys think about an overkill power supply with a Variac? Slowly raise the current until it starts to heat up?

Variacs do not provide isolation from the mains. Read the TOS.

John

 

Thanks for the replies tonight! Going to bed and brainstorming.

 

Hotwire, I don't believe you will be able to estimate the VA requirement for even heating on the grid. This would require an even current flow thru the grid. It can be worked out mathematically, but very hard to put into practice.

It might be possible to crosshatch the potential at some AC rate. But I'm with MAX. You're gonna have to play with some current and see how well the grid heats up with current in one direction first. This might give you an idea of what you're up against.
I would have advised against the SS too, but after reading John's link, I like the aircraft wire.

John, I couldn't make sense of your explanation link. Could you provide a working example. I couldn't find any.
Can one take a 12" by 48" wire grid, put a potential across 12" sides and get an even current flow thru all grid elements?
The current in the 12" elements is equal to the current in the 48" elements? I have never put in a concrete heated floor, is this how it is done?
Can we defrost our windows this way?

I'm curious of your solution.

 

A quickie simulation of a 4 x 6 grid of 10 ohm resistors, driven at the diametrically opposed corners with 100 Vdc, shows that the currents in the various resistors varies by 5:1. So that's out. Based on this, I don't see any way of this working with the horizontal and vertical wires connected at each crossover node.

BR - my bathroom has a heated floor, and it is a single-layer serpentine nichrome wire.

ak

 

Working with resistor simulation is totally different, most have seen the results of coiled heating elements that have come in contact with an adjacent coil, the result is usually a hot spot at that location and eventual burn through.
The junction of the crossing will be a potential hot spot, IMO.
Max.

 

What do you get with planes of voltage, ak?

 

I didn't do planes, because the question in post #1 was about a criss-cross grid of resistive heater wires. A plane would be the limit of the same grid reduced to 0-dimensions elements, or sheet resistivity.

ak

 

Sorry....I meant the edges. All nodes of an edge or side.

Edit: You did it catty corner, right.

 

Indeed this project is more for burning the design into foam rather than cutting it. I would prefer to use a large power supply wired in series. The material I will be using will be stainless steel chicken wire if possible.

If you only want to emboss the foam, instead of cutting it, I think the simplest solution would be to use two arrays of wires: One of 11 wires, each 48 inches long and spaced 1 inch apart (the "long array") , and one of 47 wires, each 12 inches long (the "short array) also spaced an inch apart .

I'd have the heater arrays suspended above the workpiece, next to each other, and the workpiece mounted in a carrier detented to place it accurately below each heater.

In use:
1. move the carrier to the START position
2. load a foam block into the carrier
3. move the carrier to the first position
4. lower its heater
5. dwell
6. raise its heater
7. move the carrier the second position
7. lower its heater
8.dwell
9. raise its heater
10. return the carrier to the start position
11. unload the finished block
12. goto 2

20 AWG stainless steel wire has a resistance of about 0.406 ohms per foot at 25C, (the resistance depends on the alloy) so the long array would comprise, basically, eleven separate 1.62 ohm resistors which you could connect in series, in parallel, or in series-parallel in order to optimize the cost of the power supply for that array.

Addressing its power supply; if the array's wires were all connected in series their total resistance would be about 18 ohms at 25C, and if you needed 10 amperes to get the array to temp, that would require a supply capable of delivering 180 volts into an 18 ohm load, which is 1800 watts.

The short array would comprise 47 resistors, each one foot long, and each with a value of about 0.406 ohms at 25C, which you can connect in series, in parallel, or in series-parallel in order to optimize the cost of the power supply for that array.

Addressing that power supply; if the array's wires were all connected in series their total resistance would be about 19 ohms, and if you needed 10 amperes to get the array to temp, that would require a supply capable of delivering 190 volts into a 19 ohm load, which is 1900 watts.

Keep in mind that unless you have spring-loaded frames, or somesuch, for the heaters, the wires will sag when they get hot and the emboss won't be flat across the foam block. Nichrome has a much lower thermal coefficient of expansion than stainless steel, so you may want to rethink your choice of heater material, however you configure the heater.

Also, be aware that the arrays' wires will have a positive temperature coefficient of resistance, which means that as they heat up their resistance will increase, which means that both power supplies should be constant-current supplies with a compliance greater than or equal to the voltage across the array at its highest resistance in order to keep the temperature o9f the wires constant.

Nichrome wins again, in this arena, by sporting a tempco of about 400 ppm /C at 25C.

A final note: if the arrays' wires were enclosed within tight-fitting PTFE sleeves, then the incidence of styrene melt adhering to the heater would be reduced and , indeed, eliminated by a casual wipe between cycles.

 

Sorry....I meant the edges. All nodes of an edge or side.

If you drive all of the left side and return from all of the right side, you are driving all of the "rows" in parallel; very easy to compute the current required, and each row has the same current and heat (within the tolerance and uniformity of the wire). BUT there is no current through any of the "columns" because there is no potential difference between rows.

ak

 

Suppose I give up on the idea of uniformly heating this as a grid, and I make two separate stations to melt the pattern as EM Fields and others have suggested. I have been manipulating the variables over at jacobs-online nichrome wire application calculator for this project. Let's say I'm using 18g nichrome wire instead of the stainless that I originally proposed. If I want to heat a 4ft length of 18g nichrome wire with a resistance of 1.686 ohms to 700 fahrenheit, it will require 12.4v, 7.35 amps, 91.14w. If I multiply the voltage by 13 and calculate the Va, I end up with 161.2v, 1184.82w and 21.918 ohms. Will someone please chime in and let me know if this is correct? Also I would like to mention for clarification that I would be using 13 resistors instead of twelve so that the 1" squares would appear to have a top and bottom border.

If these specs are correct for the needed power supply, does anyone know a website that sells anything close to that? I have been shopping around and haven't seen anything that closely matches the 161.2 output voltage with 7.35 amps. Is this something that I would need to build?
Could I use a power supply that has equivalent or greater output voltage and amps and use rheostat in front of the transformer to regulate the supplied voltage? Thanks for all the insights you guys have provided. It has been very helpful.