On one of my bookshelves is a publication titled "Flame Straightening Technology for Welders", written by John P. Stewart. As I look at it, I see that my edition was printed in 1988. I honestly do not remember if I obtained this through AWS or elsewhere.
I haven't read it in years, though I do remember that it had alot of detailed how-to information on heat straightening for many different situations/configurations. And if I remember correctly, it had very little on metallurgical effects.
Any time you use heat straightening, you'll need to make sure that the temperatures utilized do not have a negative effect on the metallurgical structure of the material. Once you overheat, it's too late to go back and try again.
Mankenberg

I can't tell you where to find that book. I do have some comments on your heat straightening. I will assume you are fabricating structural steel per AISC & AWS D1.1.

1st- I would review the requirements of AISC Standard Specification for temperature limits. I believe you'll find 1200 deg F as your max for ordinary steel with 1100 deg F for high strength steels.

2nd- don't use water to cool the steel, not even as a spray mist. I know it works really well but you will cause adverse metallurgical changes because of rapid cooling. AISC prohibits using water for cooling.

3rd- Vee heats work best if they are 75% of the section depth or even more (you are correct on this). If you are heating a heavy section, lay out your heats on both sides of the web and heat both sides simultaneoulsy (2 people). Doing this will reduce/eliminate web distortion, also known as oil-canning.

4th- use smaller rosebud tips so that heating in a specific spot takes slightly longer. The reason for this is that large tips overheat the surface so that you think you have enough heat. In reality, the 'core' for the section has not reached a high enough temp for the full movement potential. The 'cooler' steel resists the stresses imparted by the hotter material.

5th- You might try standing your beams 'upside down' with webs vertical and supported only at the ends. This is not required but gravity will help and it is easier to heat both sides of the web and the flange. Use bracing to prevent beams from overturning. It works well to clamp angles across 2 beams at the center and quarter points and heat each beam alternately. Use catch blocks to prevent sudden sagging.

6th - Start the heats at the point (apex) of the Vee and progress towards the base of the triangle. Don't try to keep the entire area hot. Once you have thoroughly heated a segment (through it's thickness) move up. When you get to the "k" section of the beam, spend extra time allowing the heat to "soak" in but be careful not to overheat the adjacent web section (remember that the "k" area is the thickest part of the beam while the web next to it is the thinnest). Hold the torches further away. And spend extra time heating across the entire flange. You get the most movement from shrinkage of the flanges.

7th- Don't trust steel color for your temperature control. Use temp sticks. I like to tape a 1000 deg F and 1100 or 1150 deg stick together so that one motion checks for 2 temperatures. I find that 1000 degrees is not enough for good movement but the 1100 - 1150 is enough

8th- Be very careful with preloading. I have personal knowledge of 2 gentlemen working for another company who met their sudden demise this way. Some people like to jack apart the middle of the beams with the ends locked in place, prior to heating, to 'help' the cambering. This does work. The problem is, the heating causes expansion that dramatically increases the stress level/ preload. Essentially it's the same as making the mouse trap spring stronger after the trap is set. As I mentioned, the results can be catastrophic; the 2 people I mentioned were sitting in between 2 beams being heated when they 'snapped' together.
The same stresses can damage the beams so many customers require engineering calculations to limit extreme fiber stress to 1/3 Fy or less.

Heating in the above steps (1-7) should give you good results. We typically get 1/8" minimum of movement per vee heat.

Make sure you have a LARGE volume of fuel and oxygen available. Excessive drawdown chills the bottles, regulators, and hoses and reduces the gases supply. That leads to starved torches that overheat and blow out the cores and is unsafe.

Avoid oxidizing (pale blue) flame settings. The temp is higher, but the BTU content is lower. You will 'burn' and overheat the surface but never get the 'core' of the steel up to proper temperature.

With 2 people heating at the same time be extra careful. Both should hold the torches towards the same end of the beam (a right-hander and left-hander work well together) to avoid pointing flames toward each other. We use clear full face shields with a 3" wide strip cut from a dark green shield for a stripe near the top. This gives good visibility while protecting the eyes from temperature, infrared, and ultraviolet. (Yes, oxy-fuel flames generate some UV too).

My last suggestion is that in order not to 'rack up the work-hours' on heating operations, set up several pairs of beams to heat in succession. When the heating is done, have the people move on to something else while waiting for cooling. The tendency is to hang around the beams to see how they react, but that is like watching paint dry.

Hope this helps,
Chet Guilford

Dear Doug,

Recently the FHWA has been developing heat straightening repair techniques that have been used to repair damaged steel bridge girders.

After attending the FHWA seminar, and attempting the HSR practices I was involved in training a steel bridge crew to perform HSR. Recently, the crew performed a successful heat straightening repair following the FHWA guidelines. You can find the FHWA guidelines at: http://199.79.179.19/OLPFiles/FHWA/009569.pdf

This manual outlines the amount of restraint and documents how heat straightening should be done. From my experience I have found that:
a) you use a rose bud tip that allows the operator to control the heating while at the same time applying the proper pattern.
b) the proper use of restraint will speed us straightening (or cambering)
c) the amount of movement can only accurately be measured on-site as different materials will move different amounts.

Also, I add another step to the FHWA guidelines to check for any "hard spots" hardness testing before and after the application of heat in areas that were to be heated. I would also recommend that MT be used on critical connections after and heat straightening/bending.

Hope this helps.

Regards,
Sean

I have the same book that Mankenberg mentioned, only mine is a Jan 96 version of the same. Good illustrations and tricks to manage heat to minimize distortion. We mechanically camber like you, but the larger 36" deep beams are Grade 50 or A992 don't bend as easy and we end up turning them upside down and doing like CHG stated with blocked at the ends and heat at quarter points until the proper camber has been achieved. Under 600F we have used compressed air to cool the beam off faster (you can see your results quicker). Using water is a touchy subject even under 600F and I just leave it alone to stay out of trouble. Get more than one piece out and block them up and spend time heating them work on something else like CHG stated, it will make better use of your time. Glad you brought this topic up, it needs to be discussed.
John Wright

Just thought of something else to mention, I found that mechanically straightening pieces that were distorted from welding didn't work too well for our other plant. They mechanically straighted truck loads of steel and when it got to the jobsite to be erected it was crooked again, just like they never touched it. I guess all the stress from welding was built up from mechanically straightening and the road vibrations relieved it going down thw highway on the truck. We have always heat straighted here at this plant and in 35+ years never had to bring loads back to be straighted again (well, except for the time a trackhoe on the jobsite ran over a stack of W8x10's) I can vouch for 19 of those 35+ years.
John Wright