Automobile spring

[Simon Weedman]

Hi All. I RECENTLY forged a blade from 5160. The blade was annealed 3 times and then heated to 840°c for about 15 min. It was then quenched. For some reason the blade did not harden at all. The grain structure appears to be brittle. I spoke to the blade Smith who is teteaching me and he said that we may have cooked the steel. Is there any way to sort this out or should I bin it.. any advice is welcome.

[Buzzkill]

A few questions here to narrow down possibilities...

Was this new 5160 stock from a supplier or is this unknown/unverified spring steel that you are assuming is 5160?

When you say annealed 3 times, did you heat it up to critical and then bury it in insulating material to let it cool very slowly for several hours or a day, or did you heat to non-magnetic and let it air cool to black heat before repeating?  This first is annealing (of simple steel).  The second is normalizing.

What did you quench in?

You talk about grain structure.  That indicates breaking the steel.  If you broke it so you could see the grain structure, then I'm pretty sure you hardened it - at least somewhat.

In general though, many heats, especially high heats, can burn some of the carbon out of the material.  For the sake of discussion if you started with 60 points of carbon (which is what 5160 has) and lost 20 points due to a lot of high heats, then you're left with 40 points of carbon (or 0.4%).  That brings it down into the medium carbon steel range and it will still harden, but it may not harden well in oil and it may not skate a file any more. 

With a little more info we may be able to help figure out what happened.

[Simon Weedman]

Hi. The 5160 was a new piece of steel. The steel was forged first and then allowed to cool. The annealing was done in a temp controlled Kiln and allowed to cool down for 24 hours during each annealing cycle in the closed Kiln. Each annealing cycle was brought down in temperature over a week period for all three cycles which were in a closed Kiln. The heat treat was done at a temperature of 840°C for about 15 to 20 minutes and then quenched in a proper quenching oil. The person who did it is a South African guild member and explained to me that we had burned the steel in the forge. When I asked if it could be recovered he said he wasn't sure. As this was my 1st forged knife I wasn't too happy with the answer as he had done so many before. I started investigating who h is why I asked for help from anyone. Thank you again for the response.

[Steve Sells]

Pictues will help, but  be aware that annealing also gives large grain and doing that 3x is asking for monster grain growth, as does extended unnecessary soaking times where did you get this  information about hardening  I guarantee it was not in the  pinned heat treat threads here.

Also stating that you used " a proper quenchant" may not be true either, it all depends on what you used,  why wont you tell us ?

[ThomasPowers]

Annealing 3 times????  Totally wrong; Normalizing 3 times before hardening is suggested for many alloys to reduce grain size.  But Not annealing!

If it doesn't harden in that quench you can try a more aggressive quenchant, (after normalizing to reduce the annealed grain structure). You are sure that it's 5160?

Was there a lot of scale while forging it?  Usually you have to have quite an oxidizing fire and long times at temp to burn off all the carbon...What kind of forge?

[JHCC]

13 hours ago, SIMON WEEDMAN said:

The person who did it is a South African guild member and explained to me that we had burned the steel in the forge.

Did the blade come out of the forge at any point sparking and fizzing, sending off a shower of sparks when you hit it? Burning steel isn’t exactly something you’d fail to notice. 

[Simon Weedman]

10 hours ago, ThomasPowers said:

Annealing 3 times????  Totally wrong; Normalizing 3 times before hardening is suggested for many alloys to reduce grain size.  But Not annealing!

If it doesn't harden in that quench you can try a more aggressive quenchant, (after normalizing to reduce the annealed grain structure). You are sure that it's 5160?

Was there a lot of scale while forging it?  Usually you have to have quite an oxidizing fire and long times at temp to burn off all the carbon...What kind of forge?

There weren't sparks even while forging. It is a gas forge. I believe it was normalised three times not annealed, I got that wrong. Im new to this. It was then heat treated. There was some scale during forging but not copious amounts. 

[Buzzkill]

19 hours ago, SIMON WEEDMAN said:

The annealing was done in a temp controlled Kiln and allowed to cool down for 24 hours during each annealing cycle in the closed Kiln. Each annealing cycle was brought down in temperature over a week period for all three cycles which were in a closed Kiln.

That certainly sounds like annealing.  For normalizing you'd heat past critical and allow it to air cool to black heat then heat back up to a slightly cooler temperature than the previous cycle and repeat.

I'm curious about what you said regarding the grain structure being brittle.  Did you break a piece and look at the grain structure or were you making that determination some other way?  "Brittle" and "hardened" go hand in hand usually.  However, a large grain structure is far weaker and easier to break than a small, tight grain structure.  The only time I've been able to see grain structure without breaking steel is when I burned it.  To me it looks a little like cottage cheese as far as texture when that happens. If you saw that while forging then the steel is destined for the scrap bin.

19 hours ago, SIMON WEEDMAN said:

Hi. The 5160 was a new piece of steel.

The topic title is "Automobile spring."  A lot of automobile spring is (or at least was in the past) 5160 steel.  However, springs may be some other alloy, especially from newer vehicles.  Were you able to verify some way that the steel was in fact 5160?  I've forged some coil springs which I worked like they were 5160, but came to the conclusion they were probably not based on how they reacted to the heat treating process.

[Steve Sells]

On 11/19/2017 at 9:05 AM, SIMON WEEDMAN said:

The annealing was done in a temp controlled Kiln and allowed to cool down for 24 hours during each annealing cycle in the closed Kiln. Each annealing cycle was brought down in temperature over a week period for all three cycles which were in a closed Kiln.

8 hours ago, SIMON WEEDMAN said:

I believe it was normalised three times not annealed, I got that wrong. Im new to this.

 

you detailed a classic annealing then say that you didnt anneal, Please make up your mind, we cant help you, if you do not even know what you did.

[Latticino]

I just went through a 3 day Heat Treatment class with one of the acknowledged experts in the bladesmithing field (Kevin Cashen).  The process you have described does sound like it would result in lamellar annealing (with lamellar plates of iron carbides interspersed in a matrix of Pearlite) and quite large grain, as others have mentioned. 

Fortunately this is solvable by proper normalizing, especially for a hypoeutectoid steel like 5160.  Your blade needs to be "reset" to smaller grain size by bringing up to critical temperature (A3)  then air cooling to black (normalizing).  Then a cycle or two of stress relief at below austenite temperature (A1) and cooling to black should be helpful as well.  Only then should you austentize and transition to martensite (heat to above critical and quench).  Make sure you don't overheat for this (heat to 1550 and quench within 4 seconds of the start of transformation, recolescence, to avoid the pearlite "nose", see TTT diagram below ).  Note that steel goes non-magnetic at a lower temperature on a heating cycle (1414 degrees),  so you will have to count on a little overshoot if you are only using a magnet to see the phase change (better to rely on decolescence  if possible). 

5160 is a deep hardening steel and a relatively slow quenchant can be used (i.e. Parks AAA, heated canola oil...).  I would certainly avoid quenching in water or brine, and Parks 50 might also be too fast leading to cracking.

   

[JHCC]

This is the kind of thing that is simultaneously fascinating and headache-inducing.

[John01]

2 minutes ago, JHCC said:

This is the kind of thing that is simultaneously fascinating and headache-inducing.

I know right.  I haven't even tried to forge an s- hook yet and I an reading all this stuff like, wow you guys know alot haha.  Thanks for the education

[Steve Sells]

I posted all this in simple English in the sticky and they wont read it, you expect them to  try to understand this ?>  lol gotta love the internet

[JHCC]

37 minutes ago, Steve Sells said:

I posted all this in simple English in the sticky and they wont read it, you expect them to  try to understand this ?>  lol gotta love the internet

I've read that sticky a dozen times, and it's extremely useful. It's getting into the details of the metallurgy that makes my head hurt. The good news is, each time I read about, for example, how to read an iron carbon phase diagram, my head hurts a little less and I understand a little more.

[Steve Sells]

My teacher was a Govt test engineer, so it would be near imposable for me to not have picked up this stuff from him,  still it can be a head ache waiting for a place to happen. So its understandable why so many dont get it,   At the simplest stage of understanding:  heat it past magnetic, and then freeze it to lock the carbon in place.  From there we get into various levels of detail. but when people change the story or didnt pay attention there is no way we can know what happened

[ThomasPowers]

The Nose it's all about the nose!

One issue with not trying to understand what's actually happening is that doing by rote ties you to *1* alloy that the "rote" was based on.  With more skill you can test variations on a different alloy to see what works best; but such testing does take time and experience.   Trying to use the rote method on differing alloys leads to a bunch of messed up blades over time!

[Latticino]

1 hour ago, Steve Sells said:

I posted all this in simple English in the sticky and they wont read it, you expect them to  try to understand this ?>  lol gotta love the internet

Steve,

Yeah, I know  , but I just got off three pretty intense days of metallurgy for bladesmiths and thought some folks might be interested in not only the steps, but with some of the science backing them up.  I mean the OP did know the annealing temperature lower limit for 5160 (A3), so I assumed that he had done some research or had some training.  His propensity to soak too long at annealing temperatures (which most likely promoted grain growth) probably came from some industrial heat treatment book, and those formula may not be applicable for blade crossections.  I am also confused by his insistence that the blade didn't harden, but was brittle.  In my experience you don't get to see grain unless the steel has been hardened and broken (likely by not following good tempering procedures and trying to flex the blade).  He may have been evaluating the "hardness" of the blade using a file on the edge and run into surface decarb issues, or overcooked his tempering cycles.  Hard to evaluate with the limited information given.

[JHCC]

1 hour ago, Latticino said:

He may have been evaluating the "hardness" of the blade using a file on the edge and run into surface decarb issues

I actually had the reverse problem yesterday when testing a RR track bolt for hardenability: the file skated on the first stroke and bit on the second and subsequent strokes.

[Simon Weedman]

On 19/11/2017 at 4:17 PM, Steve Sells said:

Pictues will help, but  be aware that annealing also gives large grain and doing that 3x is asking for monster grain growth, as does extended unnecessary soaking times where did you get this  information about hardening  I guarantee it was not in the  pinned heat treat threads here.

Also stating that you used " a proper quenchant" may not be true either, it all depends on what you used,  why wont you tell us ?

Hi Steve. The process for the forging of the new bought stock 5160 was as follows. Forged and then allowed to cool. The Normalizing was done above non magnetic. Normalised again at just on non magnetic and then again at just below non magnetic. All of the above in a Temp congrolled Kiln. The quenching was done once the blade was heated to 843°C and then quenched in oil. The shadows had all gone and the transformation phase had occured before the quench. The time from Kiln to oil was less than 1.5 seconds. I don't know what oil it was but I do know it was a fast quench oil as I did ask. The blade was then annealed after hardening at 600, 400 and 250°C in the Kiln. I have been making knives now for about 2 months and therefore don' know all the terminology and I did get the normalising and annealing mixed up. My error. The blade was file tested and it was extremely soft. I had enough steel to remove a piece with a grinder and then snap it which allowed me to look at the grain through a HP Loop. The grain was visibly minute but yet still soft. Is there any further information you need.

[Steve Sells]

no matter what terms you use,  your detail of what you did before hardening  is classic annealing, then you state again you annealed after hardening in your last post.   Long slow cooling is annealing and it  is used to make a blade easier to grind. but it leaves large grains and totally removes all hardening  to the blade. why did you anneal after hardening then complain its soft ?  It very well should be after annealing

[ThomasPowers]

You draw temper after hardening; but that is to a quite low temp compared to annealing----generally less than 500 degF

Process usually goes: Forging to shape then:  Anneal to soften for grinding and drilling. Normalizing to refine grain, Hardening to harden and Tempering to toughen.

[Latticino]

23 minutes ago, Steve Sells said:

why did you anneal after hardening then complain its soft ?  It very well should be after annealing

I think he meant to temper it after hardening.  If his temperatures are accurate; 600 deg C for the first of several tempering cycles is clearly  too hot and could have spheroidally annealed (as you mentioned) or just overtempered, depending on how long the blade was held at this temperature and if the oven temperature varies around the setpoint any significant amount. 

Thomas is also correct.  I would suggest for the experimental stage of getting the correct temper that he start with a  tempering cycle at 350 deg. F for one hour, then test the blade.  If still too hard (Rockwell hardness dependent on anticipated use), then temper again at a slightly higher temperature (say 400 deg. F) for another hour. 

Mind the potential presence of a decarberized surface when testing the blade edge.  It can give a false indication of a soft blade, and is common with beginner bladesmiths, as they often work the blade at forging temperatures for longer than an experienced smith.  The more time at elevated temperatures, the greater chance of a thicker layer of decarb.

 

 

[Steve Sells]

It is possible  that part of his problem  is his using words he does not know the meaning of,  but he did list temperatures that are sub-crit annealing.

I suggest he learns the terms before using them again to prevent problem they are still in the pinned HT thread for learning, which apparently he still has chosen not to do. We can not have any communication of we do not speak the same language

 

[Frosty]

Once again the clear signs of "learning" off the internet. He needs a dictionary to start then a bladesmithing book. Reading material posted online by real bladesmiths isn't possible if you don't know enough to tell. You run into the phenomena of the more you read the less you know. "A man with one watch knows the time, a man with two is never sure."

Frosty The Lucky.