I put the pedal to the metal last night. Mostly to see where the peak reportable pyrometer temperature was.
2249 C is the max, for the main display, but the average display can go higher for a short while.
Guess the real surface temperature if you can.
.
Display More
Using the total weight of the Lugano ECAT (452 gram) I did a calculation of the Durapot density in the following way
First the volumes of the Durapot casted volume and that of the internal rod with insert where both determined by measuring the volumes from the FEM model.
The results where :
Volume Durapot cast-----------------------1.611E-4 m^3
Volume internal rod with insert---------1.232E-5 m^3
Furthermore the total internal length of heater wire was determined to be about 120.36 cm.
Using for the AWG wire a weight of 12.5 g/m the weight of the internal heater wire is 15.045 gram
Using for the internal Alumina rod with insert a density of 3900 kg/m^3 the weight of the rod with insert is 48.048 gram
Thus the weight of the Durapot cast is 452 - 15.045 - 48.048 = 388.907 gram
Deviding this by the volume of the Durapot cast gives a density of the Durapot of 2.414 g/cm^3Indeed in the range you measured !
In the above calculation was an error because the heater wire consisted of two twisted wires.
However I used in the above calculation the weight of single wire.
This requires a correction
The weight of the Durapot cast now becomes 452 - 2 x 15.045 - 48.048 = 373.862 gram
Deviding this by the volume of 1.611E-4 of the Durapot cast gives a density of 2.321 gr/cm^3
Almost complete in agreement with what Para determined.
My experiments with alumina show a steady increase in conductivity above 500C, I didn't keep any data as I was just interested in knowing what temperature this phenomenon began in my fuel tubes, but I suspect the resistance decreases rapidly when you go past 900C.
My experiments with alumina show a steady increase in conductivity above 500C, I didn't keep any data as I was just interested in knowing what temperature this phenomenon began in my fuel tubes, but I suspect the resistance decreases rapidly when you go past 900C.
Previously my coils have run at higher voltages (up to nominally 120V Ac) than at present (max 78 V AC), so I am thinking that the ceramic electrical resistance probably isn’t the problem here. (The thermocouples are electrically isolated and the temperature is manually set by adjustment of voltage.)
I think it is most likely a sort of ceramic thermal conductivity runaway, that makes it increasingly difficult for the coil heat to escape, while the heater wire temperature keeps increasing even at a steady voltage once a critical threshold has been reached.
The apparent super low coil resistance is odd, though. Probably a bad sign. Will find out soon enough.
Previously my coils have run at higher voltages (up to nominally 120V Ac) than at present (max 78 V AC), so I am thinking that the ceramic electrical resistance probably isn’t the problem here. (The thermocouples are electrically isolated and the temperature is manually set by adjustment of voltage.)
They might be isolated, but there may be a capacitance between the isolated thermocouple and the non isolated electronics.
As such AC current may be flowing even if the thermocouples are isolated.
If the AC current causes an AC voltage on the input of the thermocouple amplifier, this voltage can, if not enough AC filtering is applied, be rectified by the on chip input protection diodes of the amplifier.
This causes an offset voltage and thus a misreading.
Such an AC current can also be caused by the magnetic field of the heater coil which is picked up by the single coil consisting of the thermocuple wires.
Thus even isolated thermocouple circuitry may be subject to misreadings.
Display MoreThey might be isolated, but there may be a capacitance between the isolated thermocouple and the non isolated electronics.
As such AC current may be flowing even if the thermocouples are isolated.
If the AC current causes an AC voltage on the input of the thermocouple amplifier, this voltage can, if not enough AC filtering is applied, be rectified by the on chip input protection diodes of the amplifier.
This causes an offset voltage and thus a misreading.
Such an AC current can also be caused by the magnetic field of the heater coil which is picked up by the single coil consisting of the thermocuple wires.
Thus even isolated thermocouple circuitry may be subject to misreadings.
That may be true, but I don’t think it is a problem here. If I was running a PID it could be come an issue. Since I physically pulled the input connection and the temperature didn’t change suddenly, but just began to drop normally, that should discount very much interference. Also the outside thermocouple is much closer to the coil than the internal thermocouple, so the outside T is much more likely to get interference. However, the local voltage gradients are going to be small since the coil is many wraps. The wrap to wrap voltage is probably something like 3 V over something like 3 mm, while the full applied voltage at 78 V peak is over the 6 cm length.
Where trouble starts is around 1310 C, where stable voltage, (only slightly more than required to hold the temperature at 1290 C), leads to a constantly increasing temperature (inside and out) and dialling the voltage back barely slows the increase down if caught in time. Only a severe reduction in voltage prevents the continued climb.
The extra low resistance measured just after cutting power could be due to some sort of galvanic effect from the hot Cylinder and coil wire inferring with the resistance measurement.
The IR pyrometer average scan temperature method is workable, but needs more work to be effective. It is sensitive to the time period measuring at any one spot and effective coverage of the object. The Cylinder ends still need to be checked individually and their power calculated, so there is little actual time savings. It could be automated but I will pass on that for now.
On the other hand, with the Cylinder ends’ radiant and convective power added to the 5 sections, the Cylinder did manage a Lugano Style COP of 7.08 last night.
The Dummy at 450 C, proper emissivity, managed a COP of 0.934 including the unmeasured power controller consumption. The missing 6 W should be about right for the power controller. I will see if I can measure the controller consumption properly to verify.
Summary of Test 7 data. Not bad for OK test gear and a basic set up.
.
Display MoreThey might be isolated, but there may be a capacitance between the isolated thermocouple and the non isolated electronics.
As such AC current may be flowing even if the thermocouples are isolated.
If the AC current causes an AC voltage on the input of the thermocouple amplifier, this voltage can, if not enough AC filtering is applied, be rectified by the on chip input protection diodes of the amplifier.
This causes an offset voltage and thus a misreading.
Such an AC current can also be caused by the magnetic field of the heater coil which is picked up by the single coil consisting of the thermocuple wires.
Thus even isolated thermocouple circuitry may be subject to misreadings.
And Rossi's favourite Triac chopped ac waveform is of course exactly what is needed to make any such effect unusually large.
A good candidate for Ferrara error mechanism?
And alumina going conductive gives you a possible resistive error as well.
THH
That may be true, but I don’t think it is a problem here. If I was running a PID it could be come an issue. Since I physically pulled the input connection and the temperature didn’t change suddenly, but just began to drop normally, that should discount very much interference. Also the outside thermocouple is much closer to the coil than the internal thermocouple, so the outside T is much more likely to get interference. However, the local voltage gradients are going to be small since the coil is many wraps. The wrap to wrap voltage is probably something like 3 V over something like 3 mm, while the full applied voltage at 78 V peak is over the 6 cm length.
Where trouble starts is around 1310 C, where stable voltage, (only slightly more than required to hold the temperature at 1290 C), leads to a constantly increasing temperature (inside and out) and dialling the voltage back barely slows the increase down if caught in time. Only a severe reduction in voltage prevents the continued climb.
The extra low resistance measured just after cutting power could be due to some sort of galvanic effect from the hot Cylinder and coil wire inferring with the resistance measurement.
Well, I guess it was getting a bit noisy. From the look of the data, at 1033 C the Durapot became conductive enough to show cycling temperatures.
The hotter it gets, the worse it gets. The thermocouple data logger goes OL (9999) at >1370 C apparently.
Interestingly the temperature seems more critical than the proximity to the coil, although proximity is also important.
.
Display MoreWell, I guess it was getting a bit noisy. From the look of the data, at 1033 C the Durapot became conductive enough to show cycling temperatures.
The hotter it gets, the worse it gets. The thermocouple data logger goes OL (9999) at >1370 C apparently.
Interestingly the temperature seems more critical than the proximity to the coil, although proximity is also important.
.
Good troubleshooting !
Also good to check the thermocouple reading by pulling the connection.
My experience is that above 1000 degree C conductance is going to influence the thermocouple readings.
At 1300 degree (About the maximum temperature we operated our ovens) it became a headache.
However we managed by using analog prefiltering, digital post filtering, averaging, distributed proportional power control to stay within about +/- .1 degree C at 1300 degree C.
But that was after a lot of research and development.
Lugano power increase simulation
During the Lugano active run the testers used a power of about 800 watt for the first 5 active run periods and then increased the power to about 905 watt for the remaining 11 periods.
This stepwise increase in power resulted in an increase of the measured temperatures.
For area 5 (near the middle of the ribbed area) they made a plot of this increase of temperature versus time. This was shown in plot 5 of the Lugano report.
When there was no excess heat then the increase in temperature must be solely due to the increase of thermal power of the heating element.
In that case the increase of temperature versus time must follow the physical response of the ECAT to the increase in heating element power.
This response can be simulated by using a FEM model of the ECAT.
I did this in the past (post #10) but discovered recently that I had made a mistyping error and used the wrong density of Alumina in the previous simulation. (Used 390 kg/m^3 instead of 3900 kg/m^3 for the density of Alumina)
In the meantime we also found out (thanks to Para) that the density of the casted Durapot Alumina is much less then that of high density Alumina. and that the value is about 2300 kg/m^3.
While I have the intention to improve the accuracy of the current FEM model, I was already curious about the preliminary results.
So I decided to do a simulation with the old model first using fixed values for the physical variables.
I did that for the case the ECAT was made of high density Alumina with a density of 3900 kg/m^3 and also for the case the density was 2300 kg/m^3. The last case simulating that the ECAT was casted from Durapot.
The results of both simulations where scaled to about the same beginning temperature of the step as reported in the Lugano report and also to the final temperature.
Then both curves where combined with the result from plot 5 of the Lugano report.
The result is shown in the following figure.
The curve for the lower density of 2300 kg/m^3 is rising faster then that of dense alumina.
This is due to the fact that there is less mass in the same volume and therefore the thermal capacitance is less resulting in a shorter time constant.
Due to the lower mass density of 2300 kg/m^3 the thermal conductance should for this case also be lower and compensate largely for the lower thermal capacitance.
This should then result in the curve for the density of 2300 kg/m^3 being close to that of dense alumina.
However this is not the case in the shown simulation above since the thermal conductance of dense alumina was used. (Something to improve upon in future simulations)
As can be seen the rate of temperature increase due to increasing the heater power is much higher then the rate of increase reported.
That makes it not likely that this rate of increase is solely due to the increase of heater power combined with inflated temperatures.
LDM ,
You may have missed this a short while ago.
Summary:
"power supply was increased by slightly more than 100 W." - The power supply increase is actually 137.9 W if Joule heat is ignored. When calculating COP earlier in the report, the Joule heat was subtracted (see equation [26], page 21). This means the effective power supply increase is 131.6 W, which is significantly higher than the 100 W discussed.
"The effect of raising power input was an increase in power emission of about 700 W." - Using Net Production from the report Table 7, the increase was actually 664.2 W. This might seem a bit nit-picky, but consider that the impression the report gives is that 100 W extra was applied and 700 W extra came out (7 x input). Based on the reported values. 131.6 W extra was applied, and 664.2 W extra came out (4.8 x input).
Note: The Lugano Professors may not have added as much power as the device could take, or may have increased power incrementally over the 400 seconds. It actually looks like they turned up the juice a little bit 3 times (~110 seconds [start], ~230 seconds and ~ 400 seconds), based on the small plateaus. This could cause the rather wobbly temperature curve in the report (Plot 5, Lugano report).
Below is a plot where it takes 17 minutes to level out. The exterior temperature starts at about 85 degrees less than the Lugano device may have been in real temperatures.
I will do a better one maybe tonight.
.
.
.
L 1300 to L 1410 in 56 seconds, in a hurry.
L 1300 to L 1410 in 8 minutes bumping up to the known 1410 hold voltage and waiting for it.
.
.
.
.
.
LDM ,
You may have missed this a short while ago.
Sure I did !
Note: The Lugano Professors may not have added as much power as the device could take, or may have increased power incrementally over the 400 seconds. It actually looks like they turned up the juice a little bit 3 times (~110 seconds [start], ~230 seconds and ~ 400 seconds), based on the small plateaus. This could cause the rather wobbly temperature curve in the report (Plot 5, Lugano report).
While this is a possibility it might not have been the case.
They did not state that they increased the power gradually.
They also where surprised that in about 400 seconds they arrived at the new temperature
Note: The Lugano Professors may not have added as much power as the device could take, or may have increased power incrementally over the 400 seconds. It actually looks like they turned up the juice a little bit 3 times (~110 seconds [start], ~230 seconds and ~ 400 seconds), based on the small plateaus. This could cause the rather wobbly temperature curve in the report (Plot 5, Lugano report).
It should also not have caused the wobbly behaviour between changing the set points.
You can see this from the curves you published, They are much smoother.
The what it looks like three stage behavior might also have been caused by the different parts of the ECAT having different time constants so that we have three about exponentional time curves with different time constants interacting with each other
As far as your "Test 8 - Waiting for it" curve is concerned. I wonder why it took so long to reach te fimal temperature in your test.
Lugano was a mere 400 seconds.
My simulation for the Alumina 3900 kg/m^3 indicates about the same 400 seconds to reach the maximum value.
Timing the internal TC temp rise when going from 500 to 700 watt in the MFMP retest also gives about 400 seconds settling time.
So I wonder why in your test is was about 17 minutes.
I don't know what type of cast you used, but if it was the round one which was larger in diameter to get the convection and radiation about in line with the Lugano ECAT, then you have at least more thermal mass and thus a larger time constant.
Probably also more thermal mass since you have no internal reactor chamber ?
I am now working on an improved FEM model which has a dense internal reactor tube and a less dense outher Durapot casting.
Will hopefully soon start to do some preliminary tests to see if the static temperatures are about right. If that is the case then new transients tests are planned.
The long delay (17 minutes) is probably more of a personal call on when true steady state occurs. It takes a long time for the core to settle into its own equilibrium, and that drags on the outside temperature vs input.
Below is the inside vs outside during the Fast Ramp period:
.
.
It takes a long time for the core to settle into its own equilibrium, and ...
interesting that LENR deniers can quote such a fact without asking the question why a pure Ohmic heater of < 500 grams needs 17 minutes for reaching a steady state...
