Best regards, Hamidreza
=======================================================================
This is very similar to high current primary injection testing for circuit breakers. I have designed and/or worked on many AC and some DC test sets which would perform the test you require. The voltage you need for this current will determine what sort of transformer you need, and this will depend on the overall impedance of the load, which will be the terminal block itself and the wire used with it.
As others have posted, a toroidal transformer with one or two turns of heavy wire or copper bus is a good start. You can easily get 1000 amps continuous at about 1 volt from a 1 kVA toroid with two turns of #4-0 welding cable. You can use this to determine the impedance of the load, by applying the 1000 amps and reading the voltage.
Next, you will need to figure what it will take to get the 18,000 amperes you require. You will need 18 times the voltage you determined for 1000 amps, and the transformer will have to be able to put out 18 times the current. Most high ratio transformers have an internal impedance of 3% to
10%. A 10% impedance transformer will put out about 10 times its normal rating into a short. A 3% unit will give about 30x. The transformer can safely generate such high currents for a short duration, followed by a period of cooling, which is known as a duty cycle. This is usually based on the square of the overload current, and a 10x output would require a 1% duty cycle. One second ON, followed by about two minutes of cooling should be OK. For breaker testing, the 10x rating is usually for about 0.1 seconds (for instantaneous trip), and 0.5 seconds for short time at about 5x.
From my experience, you will probably need a test set that is rated at about 2000 amperes continuous with at least 10 volts output to get 18,000 amperes into any sort of wire and terminal block. We usually use bus bar for testing beyond about 5,000 amperes. A 20 kVA test set like this will weigh about 500 pounds and will be about the size of a small desk.
One of the suggestions to use a DC welder is a very good alternative. High power spot welders are available fairly cheaply on eBay and other surplus outlets. They use high frequency (about 1000-2000 Hz) to drive the output transformer and diodes, so it can be 20 or 30 times smaller and lighter than a 60 Hz AC tester. Also, DC eliminates most of the inductance limitations, although you need to monitor the current with a storage scope or a specially designed instrument to make sure the rise and fall time is not significant. For one second, it should be minimal.
The actual measurement poses another problem by adding more impedance (if you use a shunt or iron core CT). A Rogowski coil or Hall effect device will provide minimal extra burden but the circuitry becomes more complex.
There is more information on this technology on my website
formatting link
If you would like to purchase or rent a test set which will do just what you need, please see
formatting link
I'm willing to help if you want to make your own test set from surplus parts. I would need more complete specifications on accuracy and load configuration, however.
Good luck,
Paul