Induction Heater Project Report
The project started after OSHE was approached by the project sponsor. The sponsor informed OSHE of their current project and requirements. The sponsor is in the process of restoring an early 1900s steam shovel. The sponsor informed OSHE that in order to repair the machine they will be performing pneumatic riveting, and need an induction heater designed to heat up the rivets. The project was determined to be feasible.
The goals of the design were determined with the sponsor and divided into functional requirements, and value added goals as follows:
|Value Added Goals
|Appropriate Coil – YesCoil dimensionsNumber of turns
|A flow meter interfaced with the control system to indicate cooling water is flowing – No
|Heats rivet to white hot – No
|A coil quick disconnect, such that multiple coil designs can be easily tested and adjusted – Yes
|Ceramic rivet collar for positioning – Yes
|Overheat/temperature detection – Yes
|Heater needs to be safely enclosed – Yes
|LCD to display pertinent information – Yes
|The system requires physical E-Stop – Yes
|Adjustable collar for different height rivets – No
|Rivet needs to reach temperature in <2 minutes – No
After testing, we were not able to push current into the coil and heat the rivet. There are two main problems: smoothing the rectified input AC using a 3 milli Farad, 75V capacitor and redoing the MOSFET drain soldering which might be shorting to the drain side.
The coil has been designed to theoretically appropriate specifications but remains to be tested. The ceramic collar is made, but the adjustment and load/unload mechanism is going to be integrated by the sponsor when they have the device on location. The system is equipped with two relays, one for the E-stop chain, and one for the power switch. The system has an arduino based controller that monitors the status of the E-stop and power switch and will output their status to an LCD display. There is an indicator LED to inform the operator the E-Stop is off and power is engaged. The arduino is equipped with a K-type thermocouple, which is rated for the temperatures up to 1800 degrees Celsius, and outputs the temperature to the LCD display.
The system has standard garden hose connects for the water cooling system. The system is equipped with a quick change copper compression fitting that allows for the quick redesign or exchange of the work coil. There is no flow meter integrated into the system due to supply issues, however ports are available on the arduino controller that will allow easy integration in the future.
The initial considerations were power requirements, power supply, budget, and enclosure requirements. Newton’s law of cooling and the specific heat capacity of steel were used to calculate the required power for this application. 2.6 kW was calculated as our target , which would allow for a fast heating time and appropriate temperature.
The second consideration was power supply requirements, and budget restrictions, and there were really two designs that could be adopted. The first design was a 15 kVA induction heater that employed a three phase 240 input and a more modest 3kW design that required 50V dc input. The second design was settled upon because of its comparable affordability at <$300 for components, and comparatively modest power requirements. This design would have to be integrated in an enclosure with all appropriate connections made for power, and cooling water.
The induction heater board design was adopted from the design cited below. The induction heater uses an input of 50V, and a maximum input of 50A. The board design was well documented, with an accessible component list. The board was printed and assembled, with the only change from the schematic being the addition of improved toroid rings to accommodate the additional power draw of the 3kW design.
There were also careful considerations made to the design of the coil for these rivets specifically. The ideal coil geometry was determined using this link and associated pdf file. The coil was designed with an additional air gap of 1 cm over the radius of the rivet, with 7.5mm air gap between coil turns, and more than 6 coil turns to completely enclose the rivet. In order to ensure the coil was physically assembled as close to these specifications as possible, a simple mechanism was printed using PLA at 60% infill. The mechanism looks like a large screw, and when printed, allows us to easily wind our copper tubing around it.
The enclosure consisted of a modified Dell server case. The internals were stripped and reconfigured to accommodate the induction heater. The case was modified to allow all pertinent connections to be made, and the safety and power relays integrated into the system. All of the logic and status equipment was run at the 12V logic level. A OTS 12v power supply Care was taken to ensure all connections were intact, all cables heat shrunk, and vulnerable connections were covered in dielectric grease (petroleum jelly) to prevent corrosion. All thru-holes cut in the surface of the case were grommeted, to limit dust and water intrusion. The components were laid out with ease of service in mind, with all fasteners being tapped when possible, and wires labeled and routed neatly.
The system was configured with an arduino nano, which monitored the status of relays, monitored the temperature of the crucible, and the power status. The thermocouple was designed around the MAX6675 thermocouple board with a k-type thermocouple. This board was easily sourceable and cheap. The thermocouple was driven by an arduino nano and displayed its temperature and relay status on the LCD Screen. This design is inspired by the design mentioned down below in the resources section. The arduino code was tweaked to accommodate for the relay and power switch status and display.
A table of links pointing to all open source resources you utilized
|3kW Induction Heater Build Guide (Includes Schematics, Gerber, and BOM)
|Arduino MAX6675 Thermocouple with LCD Screen