induction heating

For all but constant section parts, evenly sized the full length, it's all a best guess on the part of the engineers who figure out how to make it work, then they experiment until they get it right.

I recently got some good heating results on a metal I beam by resistance heating down the length of the web for a test. The heating was uneven just at the very ends where current density spread out, but the rest of the beam had little temperature gradients. Ironically, I thought it would work good, the ee who was running the numbers for me didn't think it would, funny how that work sometimes. Having a 250 amp DC power supply helped a lot. Less than two volts across a part that was passing over 200 amps. Would have cranked the current up more, as the supply was good for 350amps, but the circuit breaker wouldn't let me go any higher, so I stopped and called it good.

This site is a beginning. We are using an Intellifire 250. It certainly does heat things up fast when they are round.

Getting a magnetic field pulsing at 10 to 15 kHz plus running 450 to 600 volts seems scary. I look at the display and keep my distance once the switch is thrown. I was warned not to leave any loose metal articles near the cables. Randy

Good writting .. The beam wasnt insulated was it ? Heat loss over large area ...

I did the F-111 wing pivots ( yeah , all by myself ! ) . Many passes with TIG and.035% carbon wire . But induction heating w/ large cables over the 5 inch steel "arms" . Even had to heat them "down" hours after welding ....

Political ? I had to grind out most of the weld ! And that was before X-rays !!! Not after ! We never found any reason to grind out more , after vuing X-rays .

Magnetic fields heat iron by hysterisis , that will cause currents because the switching magnetic fields will force electrons . Its one or the other . If electrons are forced in iron , then the stationary mag fld will deflect the electron , but if the mag fld is moving , it will acelerate an elctron from rest . Its just a point of vu , if you are a stationary electron , or a mag fld , you see the other moving .

Plywood glue is cured by the "other" thing in an RF fld . The electrostatic force on wood fibers will rub them , heating them .

Rambling along .... HF MIG wire will stop , arc distance increases and noise goes down and power to weld goes down BIG ! You can MIG if the wire is not large diameter and almost touching the work . The reasons : MIG only has a low voltage , many arcs are 10 to 20 volts .

Stick welders can make a welding 'arc" anywhere in 5 to 25 volts for the amps is mantained , the voltage increases to maintain the amps. But MIG stops dead , for its voltage is clamped to 24 vdc . So pull the wire away a bit too much and you STOP welding ! Your MIG box would work better if it had a sophisticated elect control to speed up wire , for operator errors . You would maintain the power at the work . This is why i recommend to novices : Always speed up the wire too fast , then reduce if it dont weld . Too slow speed will baffle many ppl . Since MIG has a too high resistance in the small wire , always use the largest wire . I use .035 on ALL steel . ? Its intuitive to start on .030" steel with a small wire , it may be a "match" . It just complicates learning , covers up problems . But the complainst will be , its too fast to burn thru ! Yes but isnt that the main complaint always ?

Basics of welding : thin edges will need to burned back and form a round bulb edge before you can begin to build up and fill a hole , so burn the bulb quick and move fast , dont linger . The book says gring a perfect fit ???

OK , do it .

But ill burn a bulb and do a 2nd pass and its over in seconds ! Since you have CO2 coverage , you can do many passes over a MIG weld and get strong bond .

MIG is a quieter arc , very close , sometimes wire will touch and push ya a little , its a lower ,, blue arc . If you are too far away , quieter ,, whiter , brighter , you cant weld here .

Notice the split rollers on 151T 220vac HF MIG ? Better idea , less force to grab wire . i dont like small diameter rollers . Im workin on a wire feed device , no rollers , 3 cams push 3 arms , wire grippers can be very fault tolerant ! They dont bite the wire , but they dont need adjust force !

--Looks like a fun toy, but doesn't your electric power meter get a little whiplash when you turn that thing on? ;-)

The same appears to be the case when an electronic ckt shows a flyback , with a transistor pulling a coil to ground , then letting it go . The current thru the transistor looks like its wasted , a closed "switch" to gnd , only a coil slowing amps . But all those coils and transformers windings are actually very effecient and dont disapate what you'd think . and steel can be heated with 1000 watts to a high temp using induction . The waste can be very low in a 60 hz system and modern switchers are way up in the 80% and more eff .

I design switchers and theres some exciting things happening . I opened a cheap Chineese 400 watt PS and found a power transformer far smaller than the common IBMPC pow supplies . it was about 1 cubic inch . I had a 200w load avail' ... it worked ! it ran cool . The secret is in the wire , flat , perfectly formed soft copper , run at optimum freq of about 150-200 khz . All smaller switchers , as in Sony laptops only use 1 bipolar and it is always oscillating to nix the control needed to keep core from saturating . Switchers can be series or paralleld without problems . i got 10 ( thats 4000 watts ) i'll use for a 100% duty cycle MIG . Its much easier to safe a switcher , it can soft start , it can sense tiny overloads and shut down in mili seconds . Except for ICOM PS-125 , most switchers are protected , and its so inexpensive to do ! Im a Ham Radio Op and my 745PRO pow supply had a short and burned out the expensive hybrid circuit ( cause it is NOT protected ) But i have all the control circuits , so a LM431 and a few resistors and it will be running again . It was hype . They think MOSFETS are good !! There is nothing better than the old MJE1300x ( C2335 is Jap version ) . ICOM claimed they use 2 expensive MOSFETs !!

If you see a simple , single transistor switcher with no I.C. , only a zener and Opto ...or..... a LM431/Optocoupler , dont put it down , this is the future . They are more fault tolerant , simpler , cheaper , higher efficiency . You can even modulate and turn completly off , easier !

It is possible to build a small "lunch" box that runs cool , 100% duty cyle and 200 amps . and very low cost .

Since this is all square wave , all parts like diodes and transistors can be paralleld without problems . The transistor base drive can vary , no problem , as it wont harm the hotter transistor , it absorbs the base drive w/o much change , the gain is allready so low ! Or the collector characteristics can be mismatched and the power chnage will be so low compared to the load it drives , no problems .

Its called fault tolerant , if something changes , it wont burn up anything . At these hi frequencies , a tiny coil will ballast and increase the phase angle of conducting diodes to lower the peak current . Not so on the slower input diodes , they must take large over loads ! I use TRIACs to soft start these diodes . Or nix Triac/Diode with SCR's . . Everything is much easier in switch mode power supplies .

I would check out this book , its relatuvely old but aside from the power source types, the rest of the book is probably still quite current. I have found Abe Books a great way to find machining and engineering references.

James Crombie