Posts by Paradigmnoia

    The higher velocity at the edges of the tube compared to the middle (that I tested) is almost certainly due to using an axial fan. It is consistent with velocity profiles of axial fans from several fan manufacturers. What is a bit inconsistent is that the higher velocity near the edges of the outlet tube persists 60 cm beyond the fan, at which point it should ‘normally’ have settled to laminar flow. I suspect that this particular fan is generating a persistent vortex flow in the tube. I tried a tube of 75 cm long on the inlet side, in addition to the outlet tube, but that made no appreciable difference to the outlet velocity profile.

    I have the current model of the same fan as Mizuno ordered, and expect it to arrive next week, (based on parcel tracking). However I am currently quite far from home, and will be for at least a week, so I won’t be able to do any anemometer traverses for maybe 10 days. Actually I am in Dawson City, Yukon Territory, Canada, and I kissed the sour toe at the Downtown Hotel about 1/2 an hour ago (Google that). Last night for Internet for a week...

    I'm not certain - but those 300 and 450 points look as though they may have been book total emissivity corrected according to empirically determined band emissivity (remember the authors of the paper did not admit any difference between the two!).

    The plot points look like the ones they gathered from emissivity reports listed, so I see no evidence of the emissivity having been adjusted.

    This is a good point - there is one more parameter here that needs to be controlled. The issue is whether the airflow restriction, and hence fan back pressure, is the same between the power / airflow fan calibration run and the runs using that info. Jed however says that for some of these measurements (R20 I guess) the air speed was independently measured. Presumably that matched the calculated speed from fan power - though we have no details of that check. As with most of this stuff, I don't expect it to be an issue, but I wish it could be ruled out, especially when historic calibration curves are used...

    I don’t see how historic calibrations can be used. (Maybe I misunderstand what you mean by historic.) The outlet was previously a different size and the fan operated at a different power. Even the fan was different at one point.

    You can tell that instantly. The fan would probably draw more power. Plus you would glance at the anemometer, see that the air speed was lower and either cancel the test or fix the problem.

    You check all parameters and all instruments, several times a day.

    The anemometer being hooked up all the time is a good idea. Which position of the outlet does it monitor?

    However, the impression from the papers is that the blower power is being used to calculate, via a semi-empirical formula, the air velocity based on prior testing. So an increase in blower power would be expected to represent an increase in blower speed.

    Why not? It works.

    I think I just explained why not above.

    Let’s say that the fan has a bunch of lint and hair wrapped around the shaft/bearing area, and the fan slows down 5% compared to normal free running. The RPM will drop 5%. The fan power consumption increases by 5% (wild guess). The calorimeter box interior temperature increases by a degree due to the reduced air flow.

    1) The pre-determined blower velocity-input power calibration formula increases the velocity result by 5%, which then translates to 5% more air volume, which translates to 5% excess heat, plus some more due to the higher air temperature from the box being measured at the outlet.

    2) The RPM to velocity calibration formula correctly decreases the velocity result by 5%, translating to a correct 5% less air volume, but the higher box interior temperature (due to slower moving air) balances the air flow drop and the correct output power is calculated.

    Note that my plot starts at 220 degree C instead of 0 degree C in the published plot 1 of the Lugano report.

    The Alumina emissivity values I used are the digitized ones from the MFMP.

    It is good to be consistently comparing to the same data.

    However, consider the low temperature range in this overlay plot containing multiple Al2O3 emissivity vs temperature data points. (Blue line is from Lugano Plot 1)


    Yes. Mizuno always checks for that, before every test. (Also check computer fans for that, and vacuum them out.) Measure the power to the fan and air flow rate to make sure they have not changed.

    You should not obsess too much about these issues. A calibration reveals all. A low power calibration will recover nearly all the heat in the stream of air. So, just apply the standard equations and see if output balances input, minus a tad. If it does you are good to go.

    If the fan slows down or starts to fail, you will see that, I guarantee. You can't miss it.

    Jed, why would one expect the fan (blower) power to change significantly during an experiment?

    I see the purpose of monitoring it, but I don’t see the point in using fan power for a proxy for air velocity.

    How much does the fan power fluctuate during a typical experiment?

    How much do those fluctuations affect the calculated flow rate?

    How would you monitor that? I think an anemometer is best.

    Obviously the anemometer is required. But the fan power is being used in an equation to determine air velocity based on the fan power and the velocity measurements taken at various fan input powers (which should be steady anyways during the experiment). The fan RPM at various anemometer traverse averages would be more indicative of the flow rate.

    The fan could struggle for some reason (lint build-up, bearings drying out, etc.) which would drive up the fan power consumption, which changes the air calculation to a positive velocity increase when it would actually be a velocity decrease. The fan RPM, on the other hand, would decrease in such a situation, correctly indicating an air flow decrease.

    Right. The mass of air varies mainly with the air temperature. Mizuno accounts for that in the spreadsheet. I took the average values for 1 hour and applied the textbook values for air at STP, ignoring temperature and other factors. My answer was close to the spreadsheet. So these other factors make little difference.

    Why is the fan RPM not monitored? Note the tachometer lead (yellow) dangling in the photos of the calorimeter. The fan RPM is a much better indication of the air flow state than the fan power.

    According to a fan manufacturer, when operating in the normal range for a fan (not being abused somehow), a fan will always move the same volume of air at the same RPM. Only the mass of the air moved changes (due to environmental conditions).

    The emissivity of stainless steel in the usual (cheaper) IR camera band (7-14 um) is terrible. Stainless steel is as reflective to IR as it is to visible light through much of the IR spectrum.

    A camera that is suitable for stainless steel will operate in short wave IR, ideally matched with a notch filter to a specific, thin wavelength band matched to a peak in short wave IR emission specific to stainless steel. If you want good results.

    Trying to use LWIR thermometry on stainless steel is trying to mathematically transform a mirror surface to a blackbody surface in order to work out the temperature. Not figuratively, but literally.

    Don't forget, as the Lugano paper stated, that the values of emissivity used for alumina in the iterative calculation were adjusted to make the results of that match the TC measurement. Since those values were not used anywhere else (the active test is at much higher temperatures) the IR tests during the dummy measurements are essentially irrelevant, giving no extra information. I guess you could work back from their incorrect method to the actual band emissivity of the alumina at those lower temperatures?


    Or zirconia emissivity due to the Pyro Paint 634-ZO?

    LDM ,

    I suggest asking the Professors how the Dummy was tested with IR, thermocouples and E Dots.

    Experience suggests that questions that are not asking for an opinion on the device or Rossi, or using blaming language, sometimes get answered. The trick is to ask a question that is strictly a mechanical or scientific question.

    (Note that the Kapton-TiO2 calibrated 0.95 emissivity stickers are only good to about 380 C, whereupon the glue fails.)

    Actually the boiling point of nickel will be much lower in a strong vacuum.

    Roughly 1800 C at 100 Pa, I think.

    I don’t follow where 8 windings under the Ribs comes from. There should be 3 x 9.5 wraps under the ribs, based on the original patent application drawing.





    I read seven at centre, two at 1 cm from centre, two at 2cm from centre, 2 at 3cm from centre.

    the last two are 3mm from the wall... you are not the only one who had difficulty interpreting this

    perhaps Jed needs to draw a diagram... for those new to velocity traverses.

    Typically traverses go across the diameter, from one side to the other.

    Starting from the center I think is what is throwing some people off.

    (It might be easy at the outlet, but it would take some extra finesse to do that through a hole in a pipe)

    If there is another dot close to the edge then I have no issue with the report averages, with the possible exception that a center dot might skew things a tiny bit higher.

    It is a bit hard to read some of the diagrams due to the resolution, BTW.

    My apologies for wasting time on this again.