Making a control Loop for an induction heater

Hello everyone! I'm new here and this will be my first post so hopefully I can get some help,

I'm having a hard time wrapping my head making some sort of control for an induction heater we have on a line here.

I have the internal temperature of the part and a set point so I can generate some error but how would I convert the error in temperature into a useable output for the induction heater power supply?

the power supply can range anywhere from 0 kw to 25 kw output power.

if the induction heater has to be precise then best to use a PID loop via a linear output temp controller.

If it does not have to be so precise then you could use hysteresis/ on off control with a SSR output type controller. This would suit most applications.

you would need a thermocouple or PT100 to provide temp feedback, depending on the precision needed and the application. if you want simplicity, go with a thermocouple.

I think a bit more information is required. Assuming you are going to use a thermocouple for the temperature a lot will depend on the power supply, is the power supply existing & as you mention 0-25kw how is that controlled ?.
Is it linear or just on/off, this will determine how you control it. for example if linear, i.e. power is controlled by 4-20ma or 0-10v then a temperature controller with an analogue output, however, if on/off then controller with PWM, as liam stated, if on/off & direct on-line then an SSR.

Something else to considder is this: Does the temperature continue to rise after you switch off the power? If that is the case then you may need to model the system to make it more accurate and/or alter where you measure the temperature.

Welcome to the forum and good luck with your project.

Nick

Also worth pointing that when going down the route of SSR's leave an on off relay on the power side to be able to isolate the supply to the heater should the SSRs short.

Welcome to the PLCTalk forum community.

The induction heaters I am familiar with are used to heat up small workpieces for short amounts of time to very high temperatures, and are generally running at a pre-set power level rather than attempting to do closed-loop temperature control.

Are you re-controlling an induction heater control unit ? What are the workpieces ? How do you sense their temperature ?

The more context you can provide the better. Many industrial heating processes have off-the-shelf solutions, but maybe this is one where a PLC-based controller is appropriate.

The Power supply is controlled linearly with 0V meaning 0 Kw and 10V meaning 25 Kw.

And to add some context. the induction heater is heating up a stabilizer bar for a car to bond a bushing. The internal temperature of the bushing needs to be 170 degrees C + for a good bond. The temperature does rise slightly after the induction unit is turned off but its not an issue.

the way we have it now, when we set it up manually, we make a relationship between the internal temperature of the bar and and external using a pyrometer. for example, change the power settings until we get a cycle time we like, and then measure what the external temperature is when the internal temperature reaches 170 degrees C. The external temperature is now the setpoint for the PLC. The PLC with look for the external temperature to be at the setpoint, and then turn off the induction unit.

What I'm trying to do is to give the PLC the information of the internal temperature, external temperature and a set amount of time. The PLC will then command the heaters to heat the bar by adjusting the kw setting on the power supplies until the internal temp is reached. The PLC will essentially record the cycle it just did in an array. and then when you go to auto, the PLC will playback that calibrated array and the power supplies will run the same power settings as when it was being calibrated.

Hopefully this provided some more information on the setup I'm working with.

right now I made a hysteresis control program that will turn the heater on to a set kw rating if the internal temperature is below what it should be at a given point in time, and turn off if its above. I haven't tested it yet but i have a feeling it could work. probably not the most ideal way but I'm not too familiar with PID control yet which i think would be the proper way to do things.

What controls hardware are we using here (Rockwell, Siemens, etc)?

Is there a constant, or at least linear, relationship between internal and external temperature? For example, you may find that there is always a 35° difference, 10% difference, etc. If this is the case then you could probably get away with only dealing with one temp reading or the other, reducing complexity and cost. Even if there are, say, five different models you typically run and this relationship is different for each, this data could be stored in a recipe structure and modified fairly easily.

We are using Allen Bradley PLC's.

And ideally there is a constant difference between external and internal temperature but it is not reliable at all. we have pyrometers to measure the external temperature but they become loose and move and the tolerances on the bars we make aren't amazing so there's quite a bit of variation when it comes to relating the internal to external temperature.

Drewster1123 said:

We are using Allen Bradley PLC's.

And ideally there is a constant difference between external and internal temperature but it is not reliable at all. we have pyrometers to measure the external temperature but they become loose and move and the tolerances on the bars we make aren't amazing so there's quite a bit of variation when it comes to relating the internal to external temperature.

Click to expand...

Either make the relationship between external and internal reliably constant, or reject using the external temperature and only use the internal temperature for your feedback. Unreliable sensing is mostly useless...

In the process, why would you not use maximum power to achieve the desired temperature in the shortest time?

The couple commercial induction heating shops I've been in used a lot infrared (IR) temperature sensing. An outfit the sells induction heating controls mentions both thermocouples (T/C)and IR sensing
https://ultraflexpower.com/learn-ab...ntrollers-for-induction-heating-applications/

Both T/C and IR sense surface temperature, unless the thermocouple is embedded somehow, like in a drilled hole. The temperature at the bonding joint is not necessarily surface temperature.

Some research is necessary to determine the relationship between surface temperature during heating and the bond joint temperature. I would not expect to find that information on-line because not only it highly application specific, but likely to be considered proprietary, given the cost, time and effort to determine that relationship.

They mention (bottom of the web page) the issue with a thermocouple being affected by the magnetic fields and recommending placement outside the field, which limits the functionality of the T/C.

They fail to mention the two major deals with IR sensing

1. emissivity.

IR reads the surface temperature which emits IR energy. Different materials emit different amounts of IR - shiny metals emit only 10% of what an apple, most plastics or a painted surface emits. The IR sensor (or what reads the signal) has to be 'calibrated' for the product's emissivity, because huge errors can result, emissivity being a proportional factor.

A classic work-around is to paint the surface black in order to get near 100% IR emission.

2. target size, FOV

IR sensors have a field of view (FOV), in effect, a circle at distance from the sensor that is the surface area 'seen' by the sensor. However that FOV registers 50% of the emitted energy, the remaining 50% is outside the FOV and if there is a major difference the 50% outside the FOV will contribute to the reported temperature value.

What that means is that the FOV should be half the diameter of the object that is being viewed, with the spillover outside the FOV still being on the heated object.

Opinion
Given that there are experienced vendors dedicated to the field induction heating, I wonder if buying an induction designed to do what you need it to do, consistently and reliably, might not be a better approach than re-inventing the wheel.