4140 cracking issue

[MOblacksmith0530]

I am going to post this I think I know what the issue was but would like some other input.
I had some little giant dies made and they were supposed to be 4140. I did a water pour quench on the face of the dies and before they were even completely cool they cracked badly. The rough cross section of the dies was 2 by 3 by 2.5 inches. I did normalize them after they were machined. They cracked in multiple directions both across the short and the long direction. I have made many hammer heads over the years out of 4140 and have never lost a hammer to cracking using a water pour face and peen dipped quench with cross sections as big as these dies were. I even made a larger hand hammer out of 4140 since this has happened and did a full quench on it and it didn't crack. When I was doing some grinding on the "4140" dies the sparking said it was more than 40 points carbon from my experience looked like 60 to 80 points carbon to me so I think the steel supplier told a fib. However if anyone has experience that differs from this please jump in here and straighten me out. I also heat treated my tire hammer dies which are 4140 with a full water quench and then temper to purple and they have held up extremely well.

[jmccustomknives]

Sounds to me like A-2 or some other air hardening steel. It wouldn't be the first time steels got mixed up.

[JNewman]

4140 is supposed to be an oil hardening steel. I would suspect that with such a large block of steel the temperature differential with the water pour was just to great and it cracked. While the water pour is a traditional blacksmith heat treatment I don't believe it is a common technique for heat treatment shops.

[Frank Turley]

Normalizing 4140. Heat to 1600-1700F (full bright red bordering on orange). Cool in air.
Hardening. Heat to 1525-1625F (bright red). Quench in oil.
Tempering would be iffy, a matter of experimentation. Tempering temperatures vary from 400F to 1300F.

[MOblacksmith0530]

When I have oil hardened 4140 in the past it just doesn't get hard enough. I have dented oil hardened and 600 degree tempered 4140 with mild steel that had been hammered cold.
Thanks for the input gents.

[Frank Turley]

Did you have a large enough volume of oil? Did you stir the oil to agitate before quenching? Did you agitate the piece while the entire piece was submerged? Was it the right kind of oil?

[gearhartironwerks]

I have found that 4140 used for hammer/power hammer dies does not get hard if quenched in oil. The dies on my Say Mak purchased before Tom Clark passed away are 4140 and disappiontingly, did not hold up. I now have several "beater" hand hammers that were quenched in oil. For thicker sections, I think water is the way to go. I once read that Rob Gunter said that his super quench was acceptable for 4140. I now quench 1045 and 4140 in water w/no ...gasp...tempering...and it works! The power hammer dies have been in use daily for 5 yrs with little distortion as opposed to their original state.
John

[forgemaster]

We routinely quench 4140 from 870/860 deg C in water, and then temper back to 580 deg C and we will get pretty well spot on 302HB every time, having said that, we are quenching jobs in thickness over 60mm. I have some small rings to HT tonight that have a wall thickness of about 32mm and these I will oil quench, only cause water is too severe on the smaller sizes. It is hard to get 4140 up to hardness in oil most times, I would guess you did'nt get 4140. Hard to tell thought it may have cracked because of uneven cooling, having a largish body of steel behind it, and having the face water quenched.

Phil

[Spears]

I don’t know how many times I read heat treat posts and responses and say to myself “what the hell could these people be doing to achieve or not to achieve these results they talk about?”.

Heat treatment is a process. 1) get it hot. 2) quench it. 3) temper it.
Times, temperatures, quench media, checking methods, etc etc etc depending on what kind of resources you have available will produce results directly coinciding certain parameters getting met.

In other words, if you didn’t get results, some parameters didn’t get met.

4140 will harden in oil. Maybe not with the way some of you perform the process, but rather with a process where certain parameters are “met”.

Down on the farm, in the garage, and in an industrially equipped heat treating facility can all have shortcomings as far as results go. But to outright say “that doesn’t work” or “regardless of what written specifications say” is a belittlement to some people who have spent large percentages of their lifespan in the specific industry.

Your results in heat treatment are a reflection of “your” process, not “THE” process. Use what works for “YOU” because your shortcomings reflect on how “YOU” do your heat treatment, not how the world should do theirs. Spears.

[ThomasPowers]

The biggest HT problem most of us have is that we often really *don't* know what the metal is. we know what someone *told* us it is or *sold* us as a specific alloy; but we all know how things can slip up down the line.

If you follow a well defined process for what an alloy should use to get a specific result and don't get that result; suspect the alloy!

[gearhartironwerks]

thomas is right in his assesment. so, best guess is quench in oil first. if it's not getting hard, try water or super quench. ultimately, it's always better to buy a quantity of steel and do some destructive testing.
je

[MOblacksmith0530]

Thanks to all for the replies. Judging by the spark test performed post heat treat I am going to stay with my original conclusion, it was not 4140. I know what has worked for me over the years and will stick with what works. I will also be paying much more attention and spark test it before heat treatment to hopefully get a better idea what it is.

[ciladog]

Thanks to all for the replies. Judging by the spark test performed post heat treat I am going to stay with my original conclusion, it was not 4140. I know what has worked for me over the years and will stick with what works. I will also be paying much more attention and spark test it before heat treatment to hopefully get a better idea what it is.

So the carpenter marks out a piece of wood to cut and when the cut is not square he blames the saw or the piece of wood.

Heat treating metal alloys and tool steels is a science and there is no mystery or secret to getting good results. It’s like following a recipe to bake a cake. But, if you don’t have the equipment like an oven, your cake will likely come out under cooked or burnt.

It took years of development by steel and alloy manufacturers to develop what we have today and to think that someone on this forum knows better than they know how to heat treat these metals is laughable.

It is one thing to make and heat treat a punch from a piece of spring (that has a small cross section) and completely another to heat treat a block of alloy (with a large cross section). You may get lucky and hit that decalescent point with a magnet or watching the color in a propane or coal forge. But you I know you can’t “soak” that piece of metal at a precise temperature over several hours. For that you need a controlled environment and a forge is not a controlled environment.

The cake recipe for 4140 is a well-known standard with some slight variations.

1. Like with most alloys and tool steel there is a pre-heat stage to condition the metal for hardening. The metal is brought up to 1200 F slowly and held there for 10-15 minutes.

2. Then for 4140 the temperature is brought up to 1575 F, which is the austenizing temperature of 4140. When that temp is reached and the part is uniform in color, you start timing the soak time. 4140 unlike other alloys doesn’t need a long soak time. Some require hours of soak. But for 4140 it’s only 5 minutes per each inch of the smallest cross section. So in this case the smallest cross section is 2 inches so the soak time is 10 minutes at precisely 1575 F.

3. Quench in oil (not water) and enough oil to reduce the temperature to about 150 F quickly. That would be somewhere around 5 gallons of oil. If you don’t have enough oil you can’t get the core of the piece cooled and the residual heat will begin to anneal what you are trying to harden. Why oil and not water? Because water shocks the steel and cools it too quickly. There is a crystalline reaction going on and the carbon needs time to move through the structure as it cools.

4. When the piece reaches somewhere around 150 F (that is just about what you can hold in your hand) you should start the tempering cycle. Depending on how hard you want the piece (Rc 53-36) the temp is between 350-1000 F. Here again there is a soak time. 4140 needs to soak at your tempering temp for 2 hours per inch cross section. In this case for 4 hours.

So how do you hold your die at say 350 F for 4 hours? Your toaster oven or your kitchen oven is not going to do it. It will vary a good 50-75 F which is as low as 225-400 F. Well that might be acceptable
.
So having said all this, you take a piece of 4140, heat in a forge to who knows what temp for who knows how long, cool it in water, temper for who knows how long at what temp and it cracks and you want to blame it on the supplier sent you the wrong material? How long would that supplier be in business if he sent the wrong material?

You don’t have to spend a lot of money on a controlled heat treat environment. I bought a 14 amp pottery kiln at a flea market and hooked up a digital controller and power relay for under $100. But it is precise and accurate.

It takes time; sometimes a lot of time to heat treat alloys. I just finished making and heat treating punches and dies for a Whitney 5 ton bench press out of A2 alloy. Started the heat treatment at 9:00 AM and finished at 3:30 PM.
Pre-heat at 1200 F for 10 minutes. Then soak at 1750 F for 40 minutes. Air cool for an hour to 150 F. Then temper at 600 F for 2 hours.

What a surprise it was when I followed the manufacturer’s specifications for heat treatment and the dies came out as they should. Maybe I should have tried quenching in cow manure.

An after thought, you are a better man than me if you can tell the difference between alloys with a spark test!

[Neil Blythin]

So the carpenter marks out a piece of wood to cut and when the cut is not square he blames the saw or the piece of wood.

If the carpenter is using a saw blade that's designed to cut particle board, but is actually working with hard maple - the problem may well be the saw (or the material), rather than his ability to make a cut. Or if you want to go with your 'recipe' analogy, if you substitute baking soda in place of flour, but otherwise follow the recipe to a tee, the cake will still fail.

One issue is most definitely that we don't always know what alloy we're working with.

[ciladog]

If the carpenter is using a saw blade that's designed to cut particle board, but is actually working with hard maple - the problem may well be the saw (or the material), rather than his ability to make a cut. Or if you want to go with your 'recipe' analogy, if you substitute baking soda in place of flour, but otherwise follow the recipe to a tee, the cake will still fail.

One issue is most definitely that we don't always know what alloy we're working with.

Neil, when it comes to woodworking you are not perching to the quire here. I have been a professional architectural woodworker for 40 years and I can assure you it is always the carpenter. Give me a sharp rock and I’ll cut you a square cut. :)

While it may be true that we don’t always know what alloy we are working with, my point is that companies (reputable companies) would not be in business for very long if you order 4140 and they ship you something else.

[ThomasPowers]

Yet almost everyone who has worked in the field has a story of reputable companies doing just that and many major corporations make it a practice of *testing* every piece as they come in and marking them with their code to be sure.

[JNewman]

So the carpenter marks out a piece of wood to cut and when the cut is not square he blames the saw or the piece of wood.

Heat treating metal alloys and tool steels is a science and there is no mystery or secret to getting good results. It’s like following a recipe to bake a cake. But, if you don’t have the equipment like an oven, your cake will likely come out under cooked or burnt.

It took years of development by steel and alloy manufacturers to develop what we have today and to think that someone on this forum knows better than they know how to heat treat these metals is laughable.

It is one thing to make and heat treat a punch from a piece of spring (that has a small cross section) and completely another to heat treat a block of alloy (with a large cross section). You may get lucky and hit that decalescent point with a magnet or watching the color in a propane or coal forge. But you I know you can’t “soak” that piece of metal at a precise temperature over several hours. For that you need a controlled environment and a forge is not a controlled environment.

The cake recipe for 4140 is a well-known standard with some slight variations.

1. Like with most alloys and tool steel there is a pre-heat stage to condition the metal for hardening. The metal is brought up to 1200 F slowly and held there for 10-15 minutes.

2. Then for 4140 the temperature is brought up to 1575 F, which is the austenizing temperature of 4140. When that temp is reached and the part is uniform in color, you start timing the soak time. 4140 unlike other alloys doesn’t need a long soak time. Some require hours of soak. But for 4140 it’s only 5 minutes per each inch of the smallest cross section. So in this case the smallest cross section is 2 inches so the soak time is 10 minutes at precisely 1575 F.

3. Quench in oil (not water) and enough oil to reduce the temperature to about 150 F quickly. That would be somewhere around 5 gallons of oil. If you don’t have enough oil you can’t get the core of the piece cooled and the residual heat will begin to anneal what you are trying to harden. Why oil and not water? Because water shocks the steel and cools it too quickly. There is a crystalline reaction going on and the carbon needs time to move through the structure as it cools.

4. When the piece reaches somewhere around 150 F (that is just about what you can hold in your hand) you should start the tempering cycle. Depending on how hard you want the piece (Rc 53-36) the temp is between 350-1000 F. Here again there is a soak time. 4140 needs to soak at your tempering temp for 2 hours per inch cross section. In this case for 4 hours.

So how do you hold your die at say 350 F for 4 hours? Your toaster oven or your kitchen oven is not going to do it. It will vary a good 50-75 F which is as low as 225-400 F. Well that might be acceptable
.
So having said all this, you take a piece of 4140, heat in a forge to who knows what temp for who knows how long, cool it in water, temper for who knows how long at what temp and it cracks and you want to blame it on the supplier sent you the wrong material? How long would that supplier be in business if he sent the wrong material?

You don’t have to spend a lot of money on a controlled heat treat environment. I bought a 14 amp pottery kiln at a flea market and hooked up a digital controller and power relay for under $100. But it is precise and accurate.

It takes time; sometimes a lot of time to heat treat alloys. I just finished making and heat treating punches and dies for a Whitney 5 ton bench press out of A2 alloy. Started the heat treatment at 9:00 AM and finished at 3:30 PM.
Pre-heat at 1200 F for 10 minutes. Then soak at 1750 F for 40 minutes. Air cool for an hour to 150 F. Then temper at 600 F for 2 hours.

What a surprise it was when I followed the manufacturer’s specifications for heat treatment and the dies came out as they should. Maybe I should have tried quenching in cow manure.

An after thought, you are a better man than me if you can tell the difference between alloys with a spark test!

While I agree that heat treatment is a science it is also a science that often has too many variables to always get the right results without trial and error and fudge factors that are not always in the book. I agree that using the "Proper" heat treatment with proper process control is important for critical parts. Most of the work I do for customers goes to a commercial heat treater, the dies I plan on getting machined in the next couple of weeks spending a couple of thousand in materials and machining will go to the heat treater. If I machine or forge a reasonably simple tool for my use I will treat it myself. If someones life relies on the part (eyebolts, crane hooks,) it is definitely going to someone who is using a furnace with better process control than I have and has certified testing equipment to test the hardness.

If it were as cut and dried as you are stating, the heat treater that I often use would not need to check the hardness on a couple of the parts every time. He would not need to keep a process card for each part that I send him, but he does both of those things because the shape of the part, the decarb from the forging process and the batch of steel that I have made the part from can all affect the heat treatment. I spent a couple of hundred dollars on the ASM heat treaters guide which is a sort of bible for heat treating, and it does not give a "recipe" for heat treating 4140. It does give lots of charts and graphs for hardness and a cooling transformation diagram. While it does state that the quenching medium for 4140 is oil, several of the graphs show results for water hardening as well. The OP's die is small enough I would oil quench it, if it did not get hard enough then I would try water. According to the guide 2" round as quenched would be 49RC on the surface and 38RC in the center. The corners on his die would be harder and due to the short length the whole thing will probably be harder. Jumping to 4" round you will only get 36RC on the surface and 34RC in the core.

Following your recipe Phil could not get his 60mm+ up to 302HB according to the graph that would only get him around 290HB as quenched. I tend to pay a lot of attention to his posts as I believe he does know better than the much simplified heat treatment that the steel manufacturers publish for small time users. I had a problem with some scrapers that I make for a customer out of W1 50-100 at a time, I had just over 5% of them crack using the "book" heat treatment which was also the customers recommended ht, the heat treater then had over 10% crack. I asked on here and Grant recommended quenching them in oil. I quenched a couple in oil, tempered them and then machined them half way through with a carbide end mill and tested the hardness in the center. It was in spec so I have quenched them in oil ever since and have not lost another to cracking. This is one of the few customer jobs that I heat treat in house.

[r smith]

The OP's die is small enough i would oil quench it, if it did not get hard enough then I would try oil.

On the OP's die, are you saying that you would quench in oil? :wacko:

[JNewman]

On the OP's die, are you saying that you would quench in oil? :wacko:

Thanks for the proofread I will correct it. Yes I would use oil and if it did not get hard enough I would try water.

[r smith]

Glad to help :D

[ciladog]

JNewman

Science by definition is an exacting discipline. Given a set of procedures the results must be repeatable or it is not science leading to a given conclusion. So if I or anyone else takes a given alloy of a specific volume and heat treat exactly the same way we should get the exact same result. That is a given.

Is there a chance for variables? Sure there is. Is the alloy exactly the same? If it is certified and falls within the accepted analysis then no that is not a variable. What was the temp and volume of the oil quench? What was the core temp when it came out of the quench? What was the time lag between quenching and tempering? What were the ramp-up times? All of these are variables and will influence the outcome. But if you and I do exactly the same thing we will get the same result or it is not science.

Now you mention that you have some rather expensive dies going to a heat treatment company. Do you want them to experiment with your dies or do you want them to heat treat according to what has been done thousands of times and yielded an expected result thousands of times before?

I agree with you that Phil is an exceptional person and some of the things he does (makes) is mind boggling. But I don’t know enough about the 4140 he quenches in water to make any judgment. Is hot water the same as cold oil? I just don’t know.

Look, the point I’m trying to make is that for us dummies it is far better to follow what 4000 years of development and experimentation has produced then to try to reinvent the wheel
.
I have been told about a smith that took a piece of alloy that was so hard it could not be drilled. He took it and heated it up to critical quench it in water when it is oil hardening alloy and then proceeded to drill a hole in it like it was butter. Go figure.

[GregDP]

Look, the point I’m trying to make is that for us dummies it is far better to follow what 4000 years of development and experimentation has produced then to try to reinvent the wheel

I don't really have a horse in this race so I should probably keep my mouth shut per norm... But I've got a few pieces of what a friend gave me, he told me they were 4140. Someone told him that too where he got them. I found this thread informative given some folks experiences... but if everyone had this mentality would we have developed as much as we have over the past 4000 years? Following a recipe or making a square cut can be done as well as a few factors allow, and seems, material, tools and experience are the only important ones.. Get those about right and you can probably get darn good results.. at least thats what I pulled from this thread.. by the way, isn't oil used because of it's viscosity and low boiling point? I know you can buy quenching oil.. I guess thats the baseline, but with so many 'little factors' already mentioned, it's good to know what results some folks are getting and how. Thanks

[JNewman]

Ciladog I don't think we are that far apart and I agree with most of what you are saying. My first post made the point that a face pour is not a common "scientific" heat treatment. Apart from induction or flame hardening where the whole piece is not being heated, the common modern method of differential hardening is to harden the whole piece and then temper the area wanted softer to a higher temperature.

There is a range of acceptable chemistry for any alloy. 4140 is .38-.43% carbon .8-1.1% chromium if your bar has .38% carbon and .8 Chromium and mine has .43 and 1.1 they are both 4140 and we heat treat them the same we will likely get different results.

The expensive dies that I will be getting heat treated are going to be 4340 so i can get them harder in the heavy sections they will be. I suspect that there are lots of commercial heattreat shops quenching 4140 in water in heavy sections, and they know from experience and the graphs and charts in the heat treaters guide that that is necessary to get the required hardnesses. While I am paying them for their expensive equipment I am also paying them for their experience in heat treating. The 3x3x4 dies I will be getting hardened that will be 4140, I suspect will be hardened in polymer or water rather than oil because they will not harden because oil will not cool them quickly enough.

I agree we should be following the last 100 years of development and experimentation (1000years ago it was just alchemy/magic). I also agree process control is very important 1575 degrees f is the correct austenitizing temperature for 4140 and you will get better results the closer you are assuming you have the proper soak time which is often based on thickness. But what I am saying is that we should not be discounting experience of people like Phil who is using an electric heat treating furnace and keeping notes that tell him what his process should be for the heavier material.

As well it can be good to know some of the methods THAT MAY NOT GIVE AS GOOD RESULTS, but can be used if you don't have the proper equipment, money or time to do the job 100%. For shop built hammer tools out of 4140 and 4340 I have been using Phil's method of flashing the oil off to temper them. While I have a thermocouple and meter on my gas forge I know I am not perfect in my austenitizing temperature and the flashing is not perfect guage of the tempering temperature. It is usually good enough, I can do it relatively quickly and get back to work.

[forgemaster]

Thanks for that John, its nice to get some credit for things now and again,
The way I came to be using water for quenching 4140 and other steel rather than oil, was we always used oil as per the book, but I was always having problems getting required hardness in jobs over 15kgs in weight. One particular order was discs 52kg each 220dia x 125 thick we took them to 850/860C and quenched them in oil, and tempered them to 580 degC as per the book. (We always have a chemical cert supplied with the alloys we buy and we are diligent in recording and checking heat No's from the bars supplied against the material cert issused to avoid the types of problems being discussed here.) These discs we heat treated came out at varying hardness ranges for the same bar off steel and they even had varying hardnesses in the same disc depending on where we tested them. So I rang the supplier (who is also a heat treater, who HTs many 1000s of tonnes of steel evey year) and I asked them if they had had any probs with heat No, such and such. They get their metalurgist on the phone, he says "no problems from this end that heat No accounted for about 500 ton of bar and it all heat treated as per SOP" 302HB on the knocker. So I asked their procedure ,"we ran it up to 870deg C and quenched in water, then tempered to 610 deg C which gave 302HB exactly. "But but but isnt 4140 an oil hardening steel" says I. "Yes essentially but in larger sizes etc etc you will not get a quick enough quench to get the required hardness using oil". "Hence we quench in water, in fact our water has a small amount of polymer in it so it probably quenches a little bit quicker than straight water". Since then I have followed what that metalurgist told me to do and a couple of 100 tons of forging later we can still hit the required 302HB on the head every time. We do HT smaller solid forged rings of 8kg in oil, and they also turn out OK on hardness, but any thing over 2 inches thick 4140 now goes straight into the H2O. As to oil capacity we used to have about 2500 litres in our quench tank, that was circulated and ran through a heat exchanger to cool it.
Its not what we had drilled into us at tech when I was an apprentice but it works and thats the main thing. We have even quench En26 (nickle moly carbon of about .5% that is used for shear blades etc hammer dies etc) in water and it has come out OK, and EN26 was definitely an oil quenched steel so we were told as apprentices, but our steel supplier has quenched it in there water quench tub no probs and so did we, the job by the way weighed about 175KG a disc with about a 120mm dia hole in it 130mm thick.

I only do what the experts tell me to do.

Phil

[forgemaster]

This is info I have copied from one of our steel suppliers info sources (Interlloy) as to HT of 4140 (you can google interlloy 4140)
You will notice that they recommend quenching in oil water or polymer as required. One thing that is stated by most other steel suppliers is to avoid where possible tempering 4140 in the regions between approx 200degC and 450/475degC (392F to 887F) as 4140 and also 4340 have a tendancy to lose ductility and become temper brittle in these ranges. Interlloy recomends 550degC to 700degC anyway, does not mention lower temps.


Hardening
Heat to 840 oC - 875 oC, hold until temperature is uniform throughout the section, soak for 10 - 15 minutes per 25 mm
section, and quench in oil, water, or polymer as required.
*Temper immediately while still hand warm.
Nitriding
4140 hardened and tempered bar can also be successfully nitrided, giving a surface hardness of up to Rc 60. Nitriding is
carried out at 490 oC - 530 oC, followed by slow cooling (no quench) reducing the problem of distortion. Parts can therefore
be machined to near final size, leaving a grinding allowance only. The tensile strength of the core is usually not affected
since the nitriding temperature range is generally below the original tempering temperature employed.
Normalizing
Heat to 870 oC - 900 oC, hold until temperature is uniform throughout the section, soak for 10 - 15 minutes and cool in
still air.
Stress Relieving
Heat to 680 oC - 700 oC, hold until temperature is uniform throughout the section, soak for 1 hour per 25 mm section, and
cool in still air.
Tempering
Re-heat to 550 oC - 700 oC as required, hold until temperature is uniform throughout the section, soak for 1 hour per 25
mm of section, and cool in still air.