How hot would I have to get the file to anneal it?
The temp is dependent upon the steel. Here is a page with temps to give you an idea:
http://www.knives.com/heatreat.html
and some other links on heat treating:
http://www.zianet.com/ebear/metal/heattreat5.html#steel
http://www.newmex.com/ebear/metal/heattreat0.html
http://www.mintek.ac.za/Physmet/intermet/heatreat.htm
http://www.survivalistbooks.com/faq/metalfaq.htm
http://www.cancom.net/~hnilica/metals.html#springs
Also, would it be possible to get the file red-hot (to heat treat) with a propane torch or similar?
Yes, but it will be dependent upon the torch.
This is a "weed burner", it will do it without a forge:
with a homemade burner in place:
and in my forge:
itrade's forge design using the original burner:
http://www.bladeforums.com/forums/showthread.php?s=&threadid=253613&highlight=propane+torch
If you are looking for something smaller; you might want to try a
"one fire brick forge". You can also see one on
Chris Crawford's site
(follow the "shop tour" link; it's in the background of the forge close-up photo).
They are discussed in these threads:
http://www.bladeforums.com/forums/showthread.php?s=&threadid=258389&highlight=one+firebrick
http://www.bladeforums.com/forums/showthread.php?s=&threadid=222702&highlight=one+firebrick
http://www.bladeforums.com/forums/showthread.php?s=&threadid=144377&highlight=one+firebrick
Do a search of the forum for more info on it.
Another type of small forge:
http://www.geocities.com/zoellerforge/miniforge.html
Some other alternatives:
http://www.bladeforums.com/forums/showthread.php?s=&threadid=159786
http://www.primitiveways.com/pt-knives-1.html (scroll down the page to the heat sources part)
http://assaultweb.net/ubb/Archives/Archive-000009/HTML/20010716-13-000093.html (picture links don't work, but you get the idea.)
One more question, how can one tell is a file is just case-hardened?
I just do a spark test. From one of links above:
" The first test I will discuss is the spark test. This is a test that anyone can perform at home. The idea is simple: the spark stream given off during a grinding operation can be used to approximate the grade or alloy of a steel. The equipment used should be a grinder with a no-load speed of 9000 rpm and a wheel size of around 2.5 inches. A semi-darkened location is necessary.
The easiest way to learn the test is to observe the spark streams from various known grades and compare them with this text. As you grind, you will see lines called carrier lines. At the termination of the carrier lines, you will see small bursts called sprigs. Low carbon (1008) is a very simple stream with few bright sprigs. The higher the carbon content, the more numerous the carrier lines and sprigs.
Some alloying elements change the appearance of the test. Sulfur imparts a flame shaped, orange colored swelling on each carrier line. The higher the sulfur, the more numerous the swellings. A spear-point shape that is detached from the end of the carrier line identifies phosphorus. The higher the phosphorous content the more numerous the spear points. Nickel appears as a white rectangular-shaped block of light throughout the spark stream. Chromium appears as tint stars throughout the carrier lines, having a flowering or jacketing effect to the carbon burst. The presents of silicon and aluminum have a tendency to depress the carbon bursts. All said, the best thing to do is make a set of standards to use as a comparison."
There are "spark charts"
here.
Here are some more links that might help you out:
http://www.engnath.com/public/manframe.htm
http://hossom.com/tutorial/jonesy/
http://www.knivesby.com/knifemaking.html
http://octopus.freeyellow.com/knife.html
http://www.reil1.net/design.shtml
http://www.visi.com/~darus/foundry2/
http://beaumontmetalworks.com/sander.html
http://www.customknifedirectory.com/CKD_TutorialFrameset.htm
http://64.176.180.203/sitedirections.htm
http://www.backyardmetalcasting.com/index.html
Regards,
Greg