-
#1520 by Mulletron on 01 Sep, 2015 23:33
-
Nice slideshow to get the gears turning.
It is nice to see there is considerable attention given to field propulsion techniques, but I don't get why EmDrive catches so much flack. EmDrive is a proposed method of field propulsion.
http://npo-astro.org/papers/Minami/Field%20Propulsion%20Systems%20for%20Space%20Travel%20%28H.pdf -
#1521 by CraigPichach on 01 Sep, 2015 23:43
-
Ironically if you see thrust in the down direction or even a null result at low power your actual thrust is probably greater as you are overcoming thermal buoyancy. At high power this becomes less complicated.
-
#1522 by deltaMass on 01 Sep, 2015 23:54
-
Iulian Berca reported downward thrust when he flipped the device.
-
#1523 by Silversheep2011 on 02 Sep, 2015 00:03
-
Suggestion for today:
Best practice for pivots and beams.
If our objective is to develop a DIY technique which allows us to have high precision
The modern analytical beam may provide us some useful “design features” that we can copy over to simpler designs. With some associated loss of precision
These modern analytical beams are capable of measuring down with confidence amounts being transferred to a precision of one tenth of a milligram or one ten-thousandth of a gram, ±0.0001 g. but are limited to 160-200 gram capacity ranges.
This a schematic of a well-known brand
Some features I could see being of help for DIY’s are
1. Truss type design to keep weight low and rigidly high
2. Precision certified weights
3. Fixed Not pivoted counterweights
4. Using Zero weights around centre of pivot axis
8. Knife edge balance
12. Light source going through glass graduation plate [note the small distance required say 500mm max]
14. Readout panel in front easy to read.
15. Air damping cylinder
A note from personal use: They are sensitive enough to detect the transference of oil or other matter from one's fingers, and rubbing the side of your nose was the worst. They exhibit ‘bounce around’ on readings with any heated samples over room temperature of say 50°c. So heat is going to be a major concern or destroyer of accuracy for EMdrive builders.
We would also go so far as to put a machined solid 1” steel plate underneath the Balance say 20kgs when it was sitting on a bench [you would see movement of someone walking on the floor a metre away]
refine... refine ...refine....
-
#1524 by Sledgecrushr on 02 Sep, 2015 00:11
-
I am very interested in seeing the results from your heat testing. I think it would be very important to heat the frustrum exactly in the same way that your magnetron does.
-
#1525 by SeeShells on 02 Sep, 2015 00:30
-
I had a 160mm ceramic promised from one of my old contacts in the semi industry but it fell through. So, the only place I could find one was in China. I ordered it and asked to please ship it faster (than a slow boat), they never replied. They processed my order and posted this in the shipping details.
Arriving Thu, Sep 24 - Fri, Oct 16
Shipped
Track package
99.5% High purity Round Aluminum Plate *1 Aluminum Ceramic Plate Setter Plate Disk crucible lid dia.160mm*7mm
I was bummed and even started looking in other places. Today it arrived, way ahead of schedule! I'm jazzed, as I'll be finally be able to put exact numbers to the frustum sizes and begin to construct it!
Good news today and also got in my wire mesh for the faraday cage. Tomorrow is going to be a good day.
Shell
-
#1526 by deltaMass on 02 Sep, 2015 01:07
-
These modern analytical beams are capable of measuring down with confidence amounts being transferred to a precision of one tenth of a milligram or one ten-thousandth of a gram, ±0.0001 g. but are limited to 160-200 gram capacity ranges.
The Mettler H10 will get you 10 ug, and it's about 50 years old. When it was introduced, it revolutionised chemistry labs' technique. -
#1527 by SeeShells on 02 Sep, 2015 02:27
-
Suggestion for today:
Great info!
We would also go so far as to put a machined solid 1” steel plate underneath the Balance say 20kgs when it was sitting on a bench [you would see movement of someone walking on the floor a metre away]
refine... refine ...refine....
I had to chuckle as one of the last machines I built the height measurement was so precise 1/4 um that when someone would walk next to it we could measure the deviation of the concrete and thereby the structure of the machine. It was impressive to think the floor upon which you stood could warp so much.
-
#1528 by rfmwguy on 02 Sep, 2015 02:29
-
Id love to use a mettler, but I think the laser displacement sensor will suffice. It fires off to 40 mm +/- 10 mm and should get into micrometer range resolution. Have to study specs but think sample rate is 10 msec. Longer term project will be to connect to latop running labview. Think I'll set up to show displacement as a chart or o-scope display. Other sensors can be added, mainly environmentals. Be nice to standardize on a common display that others can use to run tests. Temp and humidity could be inputs as well as seismic reference if we wanted to get fancy. Funny other higher profile labs haven't done this...if I could, so should others with much more time and money.These modern analytical beams are capable of measuring down with confidence amounts being transferred to a precision of one tenth of a milligram or one ten-thousandth of a gram, ±0.0001 g. but are limited to 160-200 gram capacity ranges.
The Mettler H10 will get you 10 ug, and it's about 50 years old. When it was introduced, it revolutionised chemistry labs' technique. -
#1529 by Mezzenile on 02 Sep, 2015 02:49
-
An idea to get rid of the thermal imbalance:
Why not to use on the other side of the balance a thermally and geometrically repesentative mockup of the EMDrive thruster system (frustum + magnetron) with the objective to reproduce and so to cancel most of the thermal perturbations effects. We could try, using electrical heaters and control thermistors, to reproduce on the mockup by a preliminary calibration, the same profil of temperatures experienced by the active device. -
#1530 by deltaMass on 02 Sep, 2015 03:04
-
It's quite a lot of work to do that, and to get it right, but I have to applaud your thinking there.
-
#1531 by Mezzenile on 02 Sep, 2015 03:24
-
It's quite a lot of work to do that, and to get it right, but I have to applaud your thinking there.
An other idea could be instead of building and calibrate a mockup, to replace it by a second identical EMDrive and to play with the important constant of time of the thermal effects:
1- We turn on the two EMThrusters installed on the two side of the balance.
2- We wait the thermal steady state of the system and mass balance it to have an equilibrium.
3- We turn off one of the EMTruster and with a thermal balance condition which should stay stable for a while, we measure the displacement of the balance. -
#1532 by aero on 02 Sep, 2015 03:40
-
It's quite a lot of work to do that, and to get it right, but I have to applaud your thinking there.
An other idea could be instead of building and calibrate a mockup, to replace it by a second identical EMDrive and to play with the important constant of time of the thermal effects:
1- We turn on the two EMThrusters installed on the two side of the balance.
2- We wait the thermal steady state of the system and mass balance it to have an equilibrium.
3- We turn off one of the EMTruster and with a thermal balance condition which should stay stable for a while, we measure the displacement of the balance.
That's a great idea.
@SeeShells are you reading this? @Mezzenile has came up with a way to common mode out the thermal artefacts. And really, get right down to it, the two thrusters don't have to be "Identical," just close enough to subtract out the thermal effects by a significant factor. And any significant reduction in thermal will allow any real signal to rise toward the top!
Add: Of course it would really be best if the frustums were identical. Could your power splitter work with the magnetron at the center balance point and drive a frustum at each end of your balance, or is that to long a run for your co-ax? -
#1533 by Croppa on 02 Sep, 2015 03:56
-
Doc, this test was 0, 200, 0 and 100 mg. No power applied.NSF-1701 New Video - Static test of assembly today without electrodes in galinstan, which adds dampening and drag. When the galinstan was removed, there were wild movements of laser spot, as I noted when I first designed the test stand. Therefore I re-attached Doc's oil dampener, which greatly reduced vertical meanderings. The oil dampener addition and galinstan removal provided a laser spot displacement on the target of approximately the same amount. IOW, galinstan and no oil dampener and 500 mg weight added approximately equal no galinstan and oil dampener and 200 mg weight added. So, the drag/viscosity of galinstan is equal to about 300 mg...far more than I imagined.
Here's the video for detailed analysis:
Excellent. By the way, let's not forget that you also verified the importance of air currents in the motion of the beam, upon removal of the Galistan damping action. There are still posters that question whether air currents are responsible for such motions. As I understand it, you tested by adding 500 mg weight, hence there is no electromagnetic force involved in this test.
Hi all,
I chopped it up at 1 fps. Not sure what happens around the 800 second mark. I put a grid on the montage just to make it easier to compare sections. The xls file with the tracking data is attached.
As I mentioned the first time I did this, imageJ has the origin of the image at the top left so that's why the upwards movement of the laser produces a drop in the graph. I quite like this since it means frustum down=plot down. Easy enough to flip if people find it confusing.
-
#1534 by Mezzenile on 02 Sep, 2015 04:33
-
That's a great idea.
A third configuration for a good thermal effect suppression with two EMThrusters used in reverse (double thrust sensitivity) and thermal behavior symmetrisation via addition of some mechanical and heater mockup parts. Cf Pdf attached.
-
#1535 by Croppa on 02 Sep, 2015 05:40
-
Folks - sorry to go AWOL after volunteering to do something, but my father has been visiting for some days. He's 81 - he doesn't need much looking after, but he does spend his days wandering through my house and garden pointing out chores which need doing, and organizing me to do them straight away...
I said I would do fourier analysis on the beam movements of rfmwguy's first experiment if someone would post the time-series, which was very soon done.
Of course, I omitted to mention that I would need the magnetron on/off audio signal to make much progress. With that I can match up the magnetron signal to the filtered beam signal and see what it tells us.
The rough analysis so far doesn't say much beyond what can be seen by eye in the plots posted,
R.
If you're still interested and need some numbers here is a file that you use to get the timing. I extracted the audio and downsampled the signal to 400 Hz to produce a manageable file. The first column is time and the second is amplitude (and a 3rd column when the amplitude is rectified so that all numbers are positive). The 5 min lower power run goes from around 264 seconds to 564 seconds and has 12 on/off cycles of 25 seconds each (10 or 11 seconds on).
I checked the timing of the extracted audio with the tracking movie that you have data from (based on the clicks of the light switch). It was off by 500 ms, so I added 500 ms of silence to the start of the audio before I downsampled. The times should closely match now.
-
#1536 by Mezzenile on 02 Sep, 2015 05:53
-
A summary of the 3 configurations proposed to reduce the thrust measurement error due to thermal effects : Cf attached Pdf
-
#1537 by Mezzenile on 02 Sep, 2015 07:08
-
3 Test configurations to mitigate thermal perturbations:
-
#1538 by kwertyops on 02 Sep, 2015 08:19
-
My tracking analysis - redundant to Croppa's of course. I've still got artifacting when the laser was swaying horizontally, but not quite as much. I've made rough measurements of the changes in pixels.
Still displaying stickiness, inconsistency in weight-to-distance movements..
Plus there's the issue of that small drop in the middle there that was unprompted and doesn't recover. Fulcrum slippage of some kind, maybe due to air currents. That kind of thing that seems very dangerous - if an event like that happened as the magnetron was turned on, that would very much give the appearance of downward thrust.
-
#1539 by glennfish on 02 Sep, 2015 10:35
-
Doc, this test was 0, 200, 0 and 100 mg. No power applied.NSF-1701 New Video - Static test of assembly today without electrodes in galinstan, which adds dampening and drag. When the galinstan was removed, there were wild movements of laser spot, as I noted when I first designed the test stand. Therefore I re-attached Doc's oil dampener, which greatly reduced vertical meanderings. The oil dampener addition and galinstan removal provided a laser spot displacement on the target of approximately the same amount. IOW, galinstan and no oil dampener and 500 mg weight added approximately equal no galinstan and oil dampener and 200 mg weight added. So, the drag/viscosity of galinstan is equal to about 300 mg...far more than I imagined.
Here's the video for detailed analysis:
Excellent. By the way, let's not forget that you also verified the importance of air currents in the motion of the beam, upon removal of the Galistan damping action. There are still posters that question whether air currents are responsible for such motions. As I understand it, you tested by adding 500 mg weight, hence there is no electromagnetic force involved in this test.
Hi all,
I chopped it up at 1 fps. Not sure what happens around the 800 second mark. I put a grid on the montage just to make it easier to compare sections. The xls file with the tracking data is attached.
As I mentioned the first time I did this, imageJ has the origin of the image at the top left so that's why the upwards movement of the laser produces a drop in the graph. I quite like this since it means frustum down=plot down. Easy enough to flip if people find it confusing.
Croppa, clean and simple as usual.
It takes a while for the oscillations to drop off. In the first weight drop (200mg), they never went to zero. It looks overall for any "reasonable" time frame, the deflection would be +- 3% of 200mg or a net limit resolution in this set up of 6 mg dropping to 2 or 3 over time.
If the delta is slower, induced oscillation may be less, but the maximum resolution in this settup appears to be 1-2 mg
edit: Which translates to about 10-20 micro newtons... I think.... if I can count zeros correctly. With oscillation +- 60 micro newtons.
