2

For a project that I am doing currently I will need an electromagnet whose total weight is less than 2kg, but can exert a force of at least 420 N on a ferromagnetic material 0.03 m (3 cm) away from the electromagnet.

The force is going to be: [(number of turns × current)2 × magnetic constant(4 × pi × 10-7) × cross sectional area] / 2 × (distance from the object you are trying to attract)2. Therefore the changes that will have the most effect are the number of turns and the current supplied. I am stuck after this point, because I don't know what components would work best. Just to make it clear, the only constraint here is that the total mass should be less than 2 kg.

So what do I want?

Well, I would be really grateful if someone could elaborate on how to choose the components and what to buy to make this possible. Or even better, tell me the components (especially the battery) I need to get in order to make this work. PS: I am a beginner, so I would really appreciate your help :)

Transistor
  • 168,990
  • 12
  • 186
  • 385
Sreenandan Balasubramaniam
  • 53
  • 6
  • There are some additional constraints. The equation is only correct if the 3 cm air gaps between magnetic core and ferromagnetic object are the main magnetic resistances in the magnetic circuit of the electromagnet. So the core has to be large enough and should have a high permeability. – Lars Hankeln Jan 29 '22 at 16:25
  • 1
    Perhaps a *magnetic chuck* would work better for your application. – Andrew Morton Jan 29 '22 at 17:56
  • 1
    Oh, I just thought: using aluminium wire instead of copper could give you a substantial weight saving. – Andrew Morton Jan 29 '22 at 18:51
  • @AndrewMorton - Maybe, maybe not. Since the resistance of aluminum is greater than copper, the heat dissipated in the coil will be greater, too. And this is going to be a problem. See my answer. – WhatRoughBeast Jan 29 '22 at 20:48
  • Can the magnet be cooled to superconducting temperatures? – Andrew Morton Jan 29 '22 at 20:53
  • What number of turns x current did you calculate? – user253751 Jan 29 '22 at 23:35

1 Answers1

5

Let's take gap length as fixed. Then the force is proportional to $$n^2 i^2 A$$.

Here is a handy-dandy force calculator. Let's start by assuming that the core is a square with the same dimensions as the gap. Then the area A will be $$A =.03^2 = .0009$$ or just about .001. Just as a starting point, let's assume 10 amps of current. Play around with the calculator and you find you'll need about 2500 turns.

How long is the wire? Well, if the core is .03 meters on a side, the length of one turn will be .12 meters, for a total wire length of 300 meters, or just shy of 1000 feet. From here you can see that to keep the weight under 4.4 pounds (2 kg) the wire gauge must be about #19, and the resistance will be about 8 ohms. Now things get interesting. How much power will the coil dissipate? $$P = i^2R = 800 watts$$ Also, the voltage needed will be iR, or 80 volts.

Let's say you use a higher gauge wire with 1/2 the diameter. Then the same weight of wire will produce 4 times the length and 4 times the number of turns. So you can cut the current by a factor of 4. Let's say 10,000 turns (4000 ft) of #25 wire at 2.5 amps. This will have a resistance of about 66 ohms and will dissipate about 400 watts, so that's a win. On the other hand, the voltage required will be about 165 volts. So it's clear that going to finer wire means less power, but more voltage required. Batteries are probably out, since I'm going to guess that stacking, say, 7 car batteries to get 84 volts at 10 amps is not in your plans. Of course, you can do that if you want.

Let's say you double the dimensions of the gap, and keep the current the same. Now you have 4 times the area, and can get by with 1/2 the number of turns. However, now you have twice the ohms per turn, so the wire resistance, power and voltage remain the same.

Given your equation, you're going to need a lot of wire and you're going to dissipate a lot of power. This means that you're going to be in real danger of overheating your wire insulation. This will be made worse by the fact that your coil will need multiple layers. How many? Let's say your coil is 3 inches (8 cm) long. Looking at the wire table, with a wire diameter of about .04 inches (remember, you do need insulation) that works out to about 75 turns per layer (3/.04), so you'll need 33 layers. The layers at the center of the winding will have a lot of trouble getting rid of heat. Also, since the winding will be about 4 inch thick (3 cm + 2 x 33 x .04 in) it's going to take a lot more wire to get the number of turns you need. So the power and voltage numbers will go up quite a bit. You can run the numbers yourself, but it's going to be a LOT more.

WhatRoughBeast
  • 59,978
  • 2
  • 37
  • 97
  • Thanks a lot for the info, I will do some more research and get back . – Sreenandan Balasubramaniam Jan 30 '22 at 03:42
  • Would it work if I use 10000 feet of #25 wire, and stack 13 of [these](https://www.amazon.in/Lithium-Rechargeable-Battery-Industrial-Equipment/dp/B085BGY3NV/ref=sr_1_6?crid=1PWBXAL6ARW79&keywords=12v%2Bbattery&qid=1643517043&s=electronics&sprefix=12v%2Bbattery%2Celectronics%2C287&sr=1-6&th=1) batteries? – Sreenandan Balasubramaniam Jan 30 '22 at 04:40
  • 1
    @SreenandanBalasubramaniam What length of time will the magnet be switched on? – Andrew Morton Jan 30 '22 at 15:14
  • 1
    @SreenandanBalasubramaniam - Well, you'll need 15, not 13, but I'd guess it would work. Of course, the batteries will only work for about 1/2 hour, and if you only have one charger it's going to take about 15 hours at 1 hr/battery to charge them. And you really, really, need to think about how you would wind your core. If you spin a form at 1 rps, it will take 10,000 seconds, or nearly 3 hours to build it up. How long have you got? If you try this, you MUST use "magnet wire". And you MUST build a temperature sensor into the middle of your coil - and pay attention to it in operation. – WhatRoughBeast Jan 30 '22 at 15:36
  • 1
    @SreenandanBalasubramaniam - And when winding a coil, you'll need to pay close attention ALL THE TIME. #25 wire is quite thin and not very strong, so any snag will likely break it. It's probably a two man job, one to do the winding and one to make sure the feed of wire is smooth. I'm guessing you'll need at least twice the length you think you do, due to the coil getting bigger and taking more wire per turn. Since weight is the limiting factor you won't get as many turns as you think you will, so you'll need to run it at higher current (and power). – WhatRoughBeast Jan 30 '22 at 15:44
  • 1
    @SreenandanBalasubramaniam - You really need to sit down and calculate exactly how many turns you'll get for 2 kg of wire when you take into account the effect of growing coil size. You cannot just ignore it. – WhatRoughBeast Jan 30 '22 at 15:46
  • 1
    @SreenandanBalasubramaniam - Also, when using lithium cells in series, you MUST NOT completely discharge the batteries. If one cell is weaker than the others and completely discharges, from then on the others will drive it in reverse, and this will destroy the battery. Being lithium, it will probably catch fire and may explode. In your potential setup, I would not run things more than 15 -20 minutes before recharging. It would be even better to put a cheap panel meter on each battery and monitor them while you are working. – WhatRoughBeast Jan 30 '22 at 16:04
  • 1
    @SreenandanBalasubramaniam - And one more thing. Your coil will have a lot of inductance, so when you try to disconnect the batteries you're going to draw an arc. If you're using a switch to do the job, that arc will tend to destroy the contacts inside the switch, so the switch may not last very long. Just a tip. – WhatRoughBeast Jan 30 '22 at 16:10
  • Thanks a lot for the help @WhatRoughBeast. It's enough for me right now if I can get it to run for 15-20 minutes, and yes I will sit down and do the calculations as soon as possible. About, the manpower and difficulty of the task, time isn't a constraint and I can get someone to help me. Thanks once again I really appreciate it. – Sreenandan Balasubramaniam Jan 31 '22 at 03:29
  • If you're going to run it for 15 - 20 minutes, make VERY sure you have a temperature sensor in the coil. And DON'T think you can use anything but magnet wire. Anything else will almost certainly melt the insulation (I'm assuming you can't afford 10,000 feet of teflon-insulated wire.) I can't stress this enough. – WhatRoughBeast Feb 03 '22 at 04:07