Ultracapacitor Hot Wire Plastic Bender

The Dougster

As I fabricate the dashboard for the MOEPED 5, an electric bicycle
with ultracapacitor energy store, I require a tricky bend in Lexan.

I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
provide a suitable hot wire bend with a 14 gage stainless spoke as a
bending element. It seems I'd be able to control the energy input very
accurately by measuring pre-bend charge to, say 0.01 VDC.

I think I'll try it and see, and report here.

Doug Goncz
Replikon Research
Seven Corners, VA 22044-0394

Dave

be careful that you don't exceed their current rating by discharing too
quickly.

"The Dougster" <DGoncz@aol.com> wrote in message
news:1171534574.404434.255560@q2g2000cwa.googlegroups.com...
> As I fabricate the dashboard for the MOEPED 5, an electric bicycle
> with ultracapacitor energy store, I require a tricky bend in Lexan.
>
> I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> provide a suitable hot wire bend with a 14 gage stainless spoke as a
> bending element. It seems I'd be able to control the energy input very
> accurately by measuring pre-bend charge to, say 0.01 VDC.
>
> I think I'll try it and see, and report here.
>
> Doug Goncz
> Replikon Research
> Seven Corners, VA 22044-0394
>

The Dougster

On Feb 15, 9:36 am, I wrote:
> As I fabricate the dashboard for the MOEPED 5, an electric bicycle
> with ultracapacitor energy store, I require a tricky bend in Lexan.
>
> I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> provide a suitable hot wire bend with a 14 gage stainless spoke as a
> bending element. It seems I'd be able to control the energy input very
> accurately by measuring pre-bend charge to, say 0.01 VDC.
>
> I think I'll try it and see, and report here.

OK! I have a model in emachineshop software, an acrylic plate with
holes to mount the caps, a stainless spoke, some 3/4 x 1/8 inch
aluminum, and some screws. I need to drill the aluminum to serve as
bus bars, clamp the busses to the caps with some of the screws, and
add some tall screws as posts for the wire. The plate was made for the
rack trunk segment of the MOEPED's store, for use in the spring.

I wonder how to predict the temperature rise. With the same formula
that tells us how awfully inefficient but wonderfully effective our
resistance heating in this apartment is at this chilly time of year, I
suppose.... 1/13 the efficiency of a heat pump, IIRC.

Have you got a formula handy, anyone?

I think I need to measure the resistance of the cold wire, and that of
a wire warmed to a uniform temperature, but I'll probably just put the
energy in, let it out, and see what happens.

Here's a link to Lexan Bending in Google's Usenet archive:

http://groups.google.com/groups?q=Bending%Lexan

I suppose a hand compass will detect the current, which might go to
1000 A....

Doug

The Dougster

The resistance of the stainless spoke wire I bent to about 135 mm
length is, on the 20 ohm scale of my LCR meter, between 0.1 and 0.01
ohm. I just can't do better today.

Assuming this is Rmin, then with Emax 10 VDC, Pmax = E^2/R = 100/.01
or 100/.1, or between 1000 and 10,000 watts. That thing should heat up
pretty fast at full charge. At half charge, that's 250-2500 watts, at
1 VDC, it's 10-100 watts, like a soldering iron.

Now, what about energy?

Well, the frame came out holding 4 each PC 2500 caps. At 2700 F each
in series, that's 2700/4 = about 700 F. Energy = 1/2 * C * V^2 (V is
the E from above). so 1/2 * 700 * 10^2 = 350 * 100 = 35,000 joules, or
at 3/4 ft-lb per joule, around 25,000 ft-lb. The same as dropping or
lifting me at 250 pound (back in 2003), a distance of 100 feet. So
that's way too much. I think this is in the deer-slug range of energy.
I don't know guns.

How about at 1 volt? It works out to 1/100 of the above, the same as
jumping from one foot onto the Lexan part. Now I know this bulletproof
plastic can survive a one foot jump from me, in bending, but it's a
lot of stress, and it might yield at that input.

Remember we are talking heat, so there's that awful efficiency factor.
Carnot is like 60% so electricity must be like 6% efficient.

I think it's about right.

Doug

The Dougster

Follwing up the OP and 2 others with a safety question:

I am confused about charge and discharge of capacitors in series with
resistance.

Can these elements be reordered without affecting behavior?

That is, given a source, a capacitor, a resistance, a capacitor, and
the sink associated with the source in series, will the capicators
charge the same as in the configuration source, resistance, capacitor,
capacitor, sink?

I say this because I can put a charge/discharge switch on the bender
and even control it with a computer, but it the caps don't charge and
discharge the same with the hot wire link in the middle of the circuit
as they do with the link at the head of the circuit, there could be an
overcharge or reverse charge with an associated cyanogen release, and
this is an enclosed space with loved ones; it's my home.

I'd like to monitor all four caps but I only have one precision
voltmeter. I have a 4-D-input USB DAQ, but it's only 10-bits. That is
1024 levels of energy to pick from and I don't plan on dong 1024
trials, but I'd really like to have some faith that the voltage is
properly distributed between the caps. These are hugs caps and very
low resistances bus bars, but the link has some resistance, and there
are huge currents involved. The link is in between with two caps on
each side in series.

Is that safe?

Doug

CarlB

On Feb 18, 1:57 pm, "The Dougster" <DGo...@aol.com> wrote:
> Follwing up the OP and 2 others with a safety question:
>
> I am confused about charge and discharge of capacitors in series with
> resistance.
>
> Can these elements be reordered without affecting behavior?

Theoretically, yes, practically, one must take precautions.

> That is, given a source, a capacitor, a resistance, a capacitor, and
> the sink associated with the source in series, will the capicators
> charge the same as in the configuration source, resistance, capacitor,
> capacitor, sink?

DC Capacitors in series are iffy things. How the charge gets
distributed
depends on how voltage is distributed across them and this depends
typically on leakage currents. Leakage current is specified as a
"no more than" specification, so one cannot be sure what will happen.
Electrical engineers put discharge resistors across the capacitors in
order to control this. The resistance is in parallel with the
resistance
of the capacitor, and is chosen so that the total resistance is
predictable,
that is, the resistor is chosen to be much smaller than the minimum
leakage resistance.

Now the reason one cares about the voltage across capacitors
placed in series is that the usual reason for putting capacitors in
series is to allow a higher voltage to be applied to them. This is
only safe if the voltage is equally distributed, hence the bleeding
resistors. If one capacitor bleeds more than the other, it can put
an overvoltage on the other, and it can explode in a cloud of
sparks, burning paper, flying aluminum, and PCB laden oil
smoke.

As far as discharging the capacitors too quickly, this usually turns
out to be a problem for some designs where an AC voltage is placed
on a capacitor. For example, a capacitor used in a AC to DC old
fashioned voltage regulator, but with the "pulsating DC" being
driven too low. For the problem of heating up a wire for cutting
plastic, I would think that the capacitor is to be discharged much
too slowly to give this problem. In addition, the specification has
two parts, a maximum current and a maximum power. I doubt
that you will touch the maximum power spec., but if you short
circuit a capacitor it is possible to hit the maximum current
spec. This could be quite exciting.

The measurement you have made of the resistance of the spoke
is inadequate for a number of reasons. First, your VOM doesn't
go low enough. What you need to do is figure out what kind
of metal is in your spoke (stainless steel I would bet), and then
calculate the resistance using a careful measurement of the
diameter and a sloppy measurement of the length. Second,
in this project the electrical resistance of your wires or bus
bars (especially aluminum) and the internal resistance of the
capacitor could dominate. You have to look up or calculate
the specs.

> Is that safe?

All in all, I think that what you are doing is ill advised. You
should get someone who understands the Art of Electronics
to do this for you.

Lexan is a thermoplastic that can be easily formed if it is
properly heated. I would be inclined to think about a wooden
form for it, and then to put it in an oven. If it's too big for
a standard oven, then talk to someone who has a kiln.
If you have friends active in model aircraft they will likely
have experience in thermoplastics.

David M. Palmer

In article <1171534574.404434.255560@q2g2000cwa.googlegroups.com>, The
Dougster <DGoncz@aol.com> wrote:

> As I fabricate the dashboard for the MOEPED 5, an electric bicycle
> with ultracapacitor energy store, I require a tricky bend in Lexan.
>
> I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> provide a suitable hot wire bend with a 14 gage stainless spoke as a
> bending element. It seems I'd be able to control the energy input very
> accurately by measuring pre-bend charge to, say 0.01 VDC.

It seems like a hard way to do it.

You typically want to use a reasonably high resistance for a heating
element (Power = current * resistance), so a 14 gage spoke would look
more like a short circuit than a heater. Nichrome wire makes a good
resistance heater.

But you need enough energy (power times time) to raise a sufficient
amount of the lexan to hot enough in order to bend. Just getting the
wire hot for a few seconds won't heat the lexan enough.

Unless you need the ability to fabricate new dashboards when you're out
mopeding in the wilderness, it's probably easier to make a heater that
plugs into the wall (transformer isolated of course).

Google on 'bending lexan' for more ideas.

--
David M. Palmer dmpalmer@email.com (formerly @clark.net, @ematic.com)

The Dougster

On Feb 18, 4:56 pm, "Dave" <n...@nowhere.com> wrote:
> be careful that you don't exceed their current rating by discharing too
> quickly.
>
> "The Dougster" <DGo...@aol.com> wrote in message
>
> news:1171534574.404434.255560@q2g2000cwa.googlegroups.com...
>
>
>
> > As I fabricate the dashboard for the MOEPED 5, an electric bicycle
> > with ultracapacitor energy store, I require a tricky bend in Lexan.
>
> > I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> > provide a suitable hot wire bend with a 14 gage stainless spoke as a
> > bending element.

Yes, the acrylic mounting plate is designed to prevent accidental
shorts. I have looked for plastic covers caps for the M6 socket head
cap terminal screws to no avail and may dip some with a product sold
in hardware stores.

You can melt a screwdriver with one of these at full charge, but on a
short, they don't vent the cyanogen unless they are in series and
there is an imblance causing a severe reverse charge. See previous
post. They are reverse polarity protected to some extent.
Unfortunately, Maxwell no longer serves documentation on these caps.
Also 10V is less than arc voltage in air (that's why cars had 12 volt
systems for so long) so that helps some for safety.

Doug

John C. Polasek

On Tue, 20 Feb 2007 14:09:50 +0000 (UTC), "David M. Palmer"
<dmpalmer@email.com> wrote:

>In article <1171534574.404434.255560@q2g2000cwa.googlegroups.com>, The
>Dougster <DGoncz@aol.com> wrote:
>
>> As I fabricate the dashboard for the MOEPED 5, an electric bicycle
>> with ultracapacitor energy store, I require a tricky bend in Lexan.
>>
>> I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
>> provide a suitable hot wire bend with a 14 gage stainless spoke as a
>> bending element. It seems I'd be able to control the energy input very
>> accurately by measuring pre-bend charge to, say 0.01 VDC.
>
>It seems like a hard way to do it.
>
>You typically want to use a reasonably high resistance for a heating
>element (Power = current * resistance), so a 14 gage spoke would look
>more like a short circuit than a heater. Nichrome wire makes a good
>resistance heater.
>
>But you need enough energy (power times time) to raise a sufficient
>amount of the lexan to hot enough in order to bend. Just getting the
>wire hot for a few seconds won't heat the lexan enough.
>
>Unless you need the ability to fabricate new dashboards when you're out
>mopeding in the wilderness, it's probably easier to make a heater that
>plugs into the wall (transformer isolated of course).
>
>Google on 'bending lexan' for more ideas.
Doug, these are the things you must consider.
With 12,500F at 2.5 volts the energy stored is
E = .5CV^2 = 3.9x10^4 joules
You must take into account the time constant of the Lexan piece,
because the wire has to be hot long enough to get through the plastic
thickness. Suppose you estimate 3 seconds, then you can compute the
wattage in the wire:
W = E/3 seconds = 13 kilowatts!
Still supposing 3 seconds is OK, you can compute the resistance of the
wire and the busses:
RC = 3 seconds time constant
R = 3sec/C = 2.4x10^-4 ohms
This is 1/4th of a milliohm, nearly impossible to measure, but can be
calculated.
If you fool around with the parameters you may get to a happy
solution. I foresee possible burning of the Lexan and caution that
39,000 Joules is an unbelievable amount of energy, as evidenced by 3
seconds worth of 13 KW energy pouring out of a few simple capacitors.
This is a very dangerous experiment.
John Polasek

The Dougster

Hi, David.

On Feb 20, 9:09 am, "David M. Palmer" <dmpal...@email.com> wrote:
> In article <1171534574.404434.255...@q2g2000cwa.googlegroups.com>,The
>
> Dougster<DGo...@aol.com> wrote:
> > As I fabricate the dashboard for the MOEPED 5, an electric bicycle
> > with ultracapacitor energy store, I require a tricky bend in Lexan.
>
> > I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> > provide a suitable hot wire bend with a 14 gage stainless spoke as a
> > bending element. It seems I'd be able to control the energy input very
> > accurately by measuring pre-bend charge to, say 0.01 VDC.
>
> It seems like a hard way to do it.
>
> You typically want to use a reasonably high resistance for a heating
> element (Power = current * resistance), so a 14 gage spoke would look
> more like a short circuit than a heater. Nichrome wire makes a good
> resistance heater.
>
> But you need enough energy (power times time) to raise a sufficient
> amount of the lexan to hot enough in order to bend. Just getting the
> wire hot for a few seconds won't heat the lexan enough.
>
> Unless you need the ability to fabricate new dashboards when you're out
> mopeding in the wilderness, it's probably easier to make a heater that
> plugs into the wall (transformer isolated of course).
>
> Google on 'bending lexan' for more ideas.
>
> --
> David M. Palmer dmpal...@email.com (formerly @clark.net, @ematic.com)

Actually one of the dual-use purposes of the MOEPED series of
experimental electric bicycles is combat welding, so, yes, we do need
the ability to fabricate dashboards out of doors. But I am mostly just
working with what I have. I have large capacitors. Variacs, I don't
got.

I gave a link to bending lexan in the fourth post to this thread, as I
read it today, dated Sun, Feb 18 2007 4:56 pm. Of course it was in
Groups, but you just click on Web to get web.

With its high chromium and nickel content, stainless steel approaches
NiChrome as heating element material. Better than coat hanger wire,
anyway, which I have tried already. The first experiments along this
line were timed charging of a motorcycle battery, followed by a short
into a bent to shape coat hanger wire. Not very subtle. I didn't even
use a hydrometer.

I like the repeatability of this setup. From stone cold to ready to
work in only a few seconds, with no thermostat. I know how benders
work but enjoy greatly seeing units of energy transformed. My sweat
energy into electricity, electricty into heat or light. It is
fascinating.

Doug

The Dougster

Hello, Carl.

On Feb 20, 9:09 am, CarlB <c...@brannenworks.com> wrote:
> On Feb 18, 1:57 pm, "The Dougster" <DGo...@aol.com> wrote:
>
> > Follwing up the OP and 2 others with a safety question:
>
> > I am confused about charge and discharge of capacitors in series with
> > resistance.
>
> > Can these elements be reordered without affecting behavior?
>
> Theoretically, yes, practically, one must take precautions.

Yes.

> > That is, given a source, a capacitor, a resistance, a capacitor, and
> > the sink associated with the source in series, will the capicators
> > charge the same as in the configuration source, resistance,
> > capacitor, capacitor, sink?
>
> DC Capacitors in series are iffy things. How the charge gets
> distributed
> depends on how voltage is distributed across them and this depends
> typically on leakage currents. Leakage current is specified as a
> "no more than" specification, so one cannot be sure what will happen.
> Electrical engineers put discharge resistors across the capacitors in
> order to control this. The resistance is in parallel with the
> resistance
> of the capacitor, and is chosen so that the total resistance is
> predictable,
> that is, the resistor is chosen to be much smaller than the minimum
> leakage resistance.

I have worked with discharge resistors before, when the pack was in
the bicycle and in use. There's no reason not to use them now.
>
> Now the reason one cares about the voltage across capacitors
> placed in series is that the usual reason for putting capacitors in
> series is to allow a higher voltage to be applied to them. This is
> only safe if the voltage is equally distributed, hence the bleeding
> resistors. If one capacitor bleeds more than the other, it can put
> an overvoltage on the other, and it can explode in a cloud of
> sparks, burning paper, flying aluminum, and PCB laden oil
> smoke.

Also, the smaller capacitor charges and discharges more quickly,
right? So running a pack to 10 V might put 2.75 V on a cap that
happens to have aged, been underfilled, or otherwise is low on the
scale. I measure the voltages across each cap and predict the
probablity of excess. I like to keep that p < 0.0001. Population
standard deviation, it is.

> As far as discharging the capacitors too quickly, this usually turns
> out to be a problem for some designs where an AC voltage is placed
> on a capacitor. For example, a capacitor used in a AC to DC old
> fashioned voltage regulator, but with the "pulsating DC" being
> driven too low. For the problem of heating up a wire for cutting
> plastic, I would think that the capacitor is to be discharged much
> too slowly to give this problem. In addition, the specification has
> two parts, a maximum current and a maximum power. I doubt
> that you will touch the maximum power spec., but if you short
> circuit a capacitor it is possible to hit the maximum current
> spec. This could be quite exciting.

Thus the need for insulator caps even if they are just wood dowels
with 10 mm holes in them.

> The measurement you have made of the resistance of the spoke
> is inadequate for a number of reasons. First, your VOM doesn't
> go low enough. What you need to do is figure out what kind
> of metal is in your spoke (stainless steel I would bet), and then
> calculate the resistance using a careful measurement of the
> diameter and a sloppy measurement of the length. Second,
> in this project the electrical resistance of your wires or bus
> bars (especially aluminum) and the internal resistance of the
> capacitor could dominate. You have to look up or calculate
> the specs.

I agree. I just thought of something, though. How about a Wheatstone
bridge? Three bus bars, and one stainless wire. Measure voltages with
a small current, one that produces very little heating. The wire goes
on a leg, the voltmeter is the bridge.

> > Is that safe?
>
> All in all, I think that what you are doing is ill advised. You
> should get someone who understands the Art of Electronics
> to do this for you.

Hm.

> Lexan is a thermoplastic that can be easily formed if it is
> properly heated. I would be inclined to think about a wooden
> form for it, and then to put it in an oven. If it's too big for
> a standard oven, then talk to someone who has a kiln.
> If you have friends active in model aircraft they will likely
> have experience in thermoplastics.

Yes, but it only needs one little bend. It's not formed overall. I
bend it to shape in a brake I modified. Acutally, with a bar I
recently made for that brake, all this may be moot. It's a pan-bending
bar that happens to fit the dashboard pretty well.

Lexan is the closest thing we have to transparent aluminum.

Doug

The Dougster

Hi, folks, and so many thanks for the participation.

On Feb 15, 9:36 am, I wrote:

> I am wonder if five each 2500 F, 2.5 WVDC caps from that vehicle would
> provide a suitable hot wire bend with a 14 gage stainless spoke as a
> bending element. It seems I'd be able to control the energy input very
> accurately by measuring pre-bend charge to, say 0.01 VDC.

I think folks are missing that 0.01 VDC figure. I propose to *control*
the charge to +/- 0.01 V from a low of 0.01 V to a high of 10.0 V. I
would certainly not *start* my trials at 10V!

I will likely start at 0.1 V. That's 1350 * 1/100 = 13.5 joules.
Around 9 foot-pounds. A sharp hammer blow at best.

See, it's easy to be safe, and yes, I do know what can happen.

Each time you double the voltage you put four times the energy into
the system. Each factor of ten in voltage is one hundred times the
energy.

I've already used pliers to balance out-of-balance cells one by one
rather than disassemble the pack and rewire in parallel. One does not
do this with kJ in the cell. Yes, the pliers warmed. I made the
calculation and judgement call based on the voltage reading.

I have already ramped up a motorcycle battery and coat hanger wire to
melt yet not burn Lexan with repeated charges and discharges of
increasing duration. Just shy of that point, the material will bend
and not be melted. I think I can do this.

Doug