Registered users: Bing [Bot], Google [Bot], Yahoo [Bot]
Who is online
In total there are 36 users online :: 3 registered, 0 hidden and 33 guests
Most users ever online was 218 on Wed Dec 07, 2016 6:58 pm
Registered users: Bing [Bot], Google [Bot], Yahoo [Bot] based on users active over the past 5 minutes
I am using aluminum and copper because they are conductive and not ferro or paramagnetic. Ferromagnetic materials will be attracted to the coil, saturate, and eventually give in to the repulsive forces of induction. wasted energy.
The idea is to use induction alone to repel the projectile.
Congrats on the SciFair win!
I wonder if you couldn't "blink" the camera "shutter" instead of the LED. Just put a fan in front of the camera lens (set the lens to "bulb" or to a long shutter time).
Need to figure out the RPM of the fan since that gives you the time base.
Without the fan a continuously lit LED gives a streak. With the fan acting as a high speed shutter it should break the streak up into dashes. The length and separation of the dashes should give you velocity versus time data from a single image.
I wonder how fast a typical fans spins? A 4 blade fan at 1000 RPM would be 15mS between gaps ("exposures"). At a projectile speed of 500 FPS that would be 7.5 feet.
A fan blade will spin a lot faster if it is flat so that it isn't actually trying to move any air.
As far as the fan goes, good idea, but I don't think anyone here quite understands the scale I am talking about.
I need to get around 10 data points in 3" of travel. the disk is up to 0.5" thick. I really NEED the speed of an LED, and the intensity of looking directly at it. I am going to end up pulsing at around 3khz, with a duty cycle of around 33%. I need to find a nice small battery to use, so as not to put too much weight on the projectile. probably a button cell.
I modified the schematic to place the cold side of the variac on Neutral. I included some ground points for the safety green ground wire to connect to and identified the hot electrical locations that include both the Plus and Minus side of the power supply. Noted the cold side of the oven transformer is tied to the transformer frame and is required to be electrically connected to safety ground by a green wire.
Noted the high current loop between the cap bank and work coil. It must be restrained mechanically as it tends to jump if permitted to move. It needs to be large gauge to handle the pulse current without exploding.
It appears to be too large to display here. It is in the album.
It is linked in this page;
http://www.spudfiles.com/forums/album_s ... pic_id=438
Good luck on the Science Fair.
Please note that if the capacitor bank is grounded on the minus lead, and the MWT was not grounded, the risk is very high the case of the transformer will have 1800 volts on it. If they are both grounded you will get smoke from the diodes. If the frame of the transformer is not grounded the high frame voltage may arc into the primary 120 volt side. The oven transformer iron frame must be grounded for safety. The cold side of the secondary is tied to it.
As a side note, I am ordering phototransistors today to detect position vs. time. put them in series and in series with a resistor. I will measure the voltage across the resistor to detect when the phototransistors are in shadow They have a base pin, so I will bias them slightly so I can tell what is happening if multiple phototransistors are in shadow at the same time. could I get away with a single resistor to bias all ~10 phototransistors?
does anyone have any ideas as to how to keep the disk from tumbling too much? preferably without a hole through the disk and a rod in the center of the coil.
Thanks, and happy Easter. I hope you all got fittings in your baskets.
Okay, I just finished my tests and I found some surprising results. The Copper disks performed much worse than the aluminum ones, and the thick disks performed much worse than the thinner ones. I can honestly say I'm stumped.
I'll attach the spreadsheet with all my data in it.
Any Ideas? I'd compensate for conductivity, do a regression, and blame mass, but I can't find the relative conductivities. Anyone who has worked substantially with FEA software, feel free to find the hoop conductivity of copper and aluminum disks 6,8, and 12 mm thick, 72mm diameter, with no center hole, at 1785hz. The skin effect in the corners should be interesting.
With the low amount of power you are using, I would blame it on mass.
Mass is the big stopper. You have a limited amount of energy. The ring will take the sharp risetime of the induced current and produce an apposing current in the ring near the surface (resulting counter magnetic current shields the thicker mass behind the surface from getting current) The thicker material does provide lower resistance, but due to the skin effect, the current penetration is limited in depth. The coupled energy is limited so thicker mass does not compensate in a linear fashion with force.
Here is more reading;
Note at 60 Hz the depth is less than 10 mm. At your impulse duration inverted to give frequency, you will find your pulse is rich in high order harmonics with a high fundamental. This gives a very limited penetration depth of the current in your projectile. Higher resistance materials such as Aluminum, have deeper penetration as the surface eddy currents are reduced. Copper is considerably more dense than aluminum, so for the same thickness, copper is considerably heavier.
At high frequencies, the depth is quite shallow for copper.
Frequency Skin depth (μm)
60 Hz 8470
10 kHz 660
100 kHz 210
1 MHz 66
10 MHz 21
Please note that this Wikipedia article seems to be in error between the two tables. Either 60 Hz is in nm, or the table should be nm. mm to um is 6 orders of magnitude, not 3 that the table indicates. I don't know which is correct without further study. I suspect the lower table is supposed to be nm instead of um.
Edit, another table online doesn't match the Wikipedia numbers and an online calculator does not match either. Nanometer is much closer to correct than Microns.
table is from http://www.microwaves101.com/encyclopedia/skindepth.cfm
At your typical pulse duration, most of your frequency current components are less than 1mm deep.
exactly. that is why I can't just take the bulk resistivity of copper, and calculate the DC resistance. current density decreases as 1/e^D/Ds
Do I need to account for the "skin" on top, sides, and through the center of the disk? or treat it like a slice of an infinitely thick plate (skin on bottom only)?
that is EXACTLY what I am trying to compensate for.
I would like to "cancel" the conductivity differences, to be left with just a change in mass. The 4hv thread mentioned that the disk receives about the same impulse, but different energy. I calculated momentum, and that still doesn't explain the differences in thickness(mass). they should theoretically (if same conductivity) have the same momentum, but I just ran it with fake numbers on that ti83 ballistic program, and it wasn't even close. I can't even find an explanation of how to get V from A and D...found one. It looks like that particular part of the TI program is wrong. But the momentum's are still different, and the thicker disks have less momentum.
Now I have to determine which would be applicable in a motor or transformer and formulate an explanation; KE or momentum. I am starting to wish I had taken physics and calculus. Anyone feel like helping me out? Force on disk seems to be constant-ish with electrical energy.
The simple uncomplicated answer to the less momentum is the principal all electric motors suffer from when stalled. They generate heat. The longer the motor is in a stall condition, the more heat and thus more energy loss exists. Maximum energy transfer to momentum happens when the projectile is moving near the speed the back EMF limits current in the primary. Stalled the current and losses are high. All induction motors are inefficient at stall speeds. A disk launcher is no exception. More mass = longer dwell in the stalled zone. Force is fairly constant ish as you noted.
Who is online
Registered users: Bing [Bot], Google [Bot], Yahoo [Bot]