Quench Cracks in 1045 Brine hardening

Hey all,

At our manufacturing plant we are experiencing about 20% of our tools cracking along a groove on the shank (picture below) of one of our tools. Our current process is to heat the 1045 steel tools to 1500F for 45 mins, then take them out and slowly enter them into the brine quench groove side first. The brine is at 10 brix and has an agitator. This process has been done for years and only now are the tools beginning to crack. The tools are then drawn at 500F for an hour and a half.

We have another tool that has the same shank geometry and we use the same process, and have never had any cracking.

Open to any ideas and suggestions on how to stop the cracking! Have tried many different ways of quenching but nothing is working.

SaltyPotato

Duff batch of steel?

Try floating an inch or so of vegetable oil on top of the quench bath.

Has the steel or steel supplier changed? Has the steel supplier changed rolling or other processes? Have you had the steel analyzed in past years and recently, or do you get an analysis from the steel supplier with each batch you buy? The recipe for 1045 is pretty simple, but some stray (unwanted) alloying elements can certainly cause cracking in a brine quench.

Has the tooling or method of creating the shank grooves changed?

Larry

Brine is the most severe quench in existence (ok liquid nitrogen excluded), can you back the salt level off to slow the quench?
Mark

Is the brine heated, or kept at ambient? Is your oven temperature calibration correct? And how are the parts held? If in a grid, the inner parts may be hotter at time of quenching than the outer.

Limy's idea of an oil float (to slow the initial quench) sounds interesting.

I believe your process can’t tolerate the variability in harden-ability that normally comes from any heat of 1045. In material chemistry terms a heat of 1045 can vary anywhere between 1043 to 1047. Any single heat has a great deal of variability with a large swing in jominy. A thin shaft like yours will have no problems getting 100% martensite at low end chemistry of 1045. You could think about 1040 or 1042.

I would recommend a polymer quench. This will better allow you to adjust your process as needed. You should not need to slowly slip your part into the water. As a matter of fact it is not a good idea as opposite end may cool below transformation temperature.
You are in a quandary because your process is not tuned for the chemistry you are receiving. Your past experience tells me you have a shaft needs a low end chemistry to survive the quench that you currently have. You have options but all are more work for yourself.

You don’t mention where the steel is coming from or how much you purchase. The more you purchase the easier the fix is going to be.

You need to know what the harden-ability if of each batch of material you receive and tune your quench to it. This will clear your cracking problem. I assume you are not buying in heat quantities where you could specify a chemistry range. That being the case you will need to know the jominy for each batch and tune your quench to it. That is where polymer quenchants come in. It’s much easier to change the severity of the quench with the correct polymer additive. You will also be able to rapidly lower the parts into the quench. Agitation of the water is good but must flow across the parts. vertical presentation is the way to go.

i have to say check the steel lot if things worked for years withought change in house then something else is at work. as a knife maker i woudl not want to do the water under oil thing. for our water quench steels parks 50 is as fast as water on first quench then as if by magic the cooling rate slowes a hair for less stress. what is the target hardness for the part its not like it is through hardening

L Vanice said:

Has the steel or steel supplier changed? Has the steel supplier changed rolling or other processes? Have you had the steel analyzed in past years and recently, or do you get an analysis from the steel supplier with each batch you buy? The recipe for 1045 is pretty simple, but some stray (unwanted) alloying elements can certainly cause cracking in a brine quench.

Has the tooling or method of creating the shank grooves changed?

Larry

Click to expand...

The steel supplier has not changed and neither has their process. We have each shipment of steel analyzed by the supplier, but due to this problem we are going to have the steel tested from an independent laboratory. The grooves are about .020" below the max allowable depth, but this small change doesnt seem enough to be causing this problem. Do you know what alloying elements can result in cracking becoming more common?

Milland said:

Is the brine heated, or kept at ambient? Is your oven temperature calibration correct? And how are the parts held? If in a grid, the inner parts may be hotter at time of quenching than the outer.

Limy's idea of an oil float (to slow the initial quench) sounds interesting.

Click to expand...

The brine is heated slightly to about 80F, the ovens that are used fluctuate between 1450 and 1550. The parts are held with tongs one at a time far away from the shank with the grooves that are cracking. Thanks for the reply!

make a single part with less depth and see what happens,internal stresses can find the weakest link and fracture.

SaltyPotato said:

The steel supplier has not changed and neither has their process. We have each shipment of steel analyzed by the supplier, but due to this problem we are going to have the steel tested from an independent laboratory. The grooves are about .020" below the max allowable depth, but this small change doesnt seem enough to be causing this problem. Do you know what alloying elements can result in cracking becoming more common?

Click to expand...

I am not a metallurgist, but I know that O1 tool steel cracks if water quenched. Comparing the typical compositions of W1 and O1 tool steels, I suspect the higher Mn, Cr and/or W are what makes the difference in quenching speed critical. But there may well be other stray elements that can cause problems.

The groove geometry is another possible issue, which is why I mentioned tooling change as a possible issue.

Larry

Is the groove sharp cornered? Needs a fillet.

I'd personally be suspicious of any part that is prone to cracking, even if it survives intact once in a while. Use a better grade of steel that doesn't require a severe quench. Or carburize 1018.

The groove has a .015" radius which we are changing to .030" but will not get this change in place for a couple of months. Looking into different variety of steel, we use 15B41 for other tools and has been working great so will probably switch over to this

It's always impressed me how little Boron is needed to make a "Boron steel".

Rickyb said:

I believe your process can’t tolerate the variability in harden-ability that normally comes from any heat of 1045. In material chemistry terms a heat of 1045 can vary anywhere between 1043 to 1047. Any single heat has a great deal of variability with a large swing in jominy. A thin shaft like yours will have no problems getting 100% martensite at low end chemistry of 1045. You could think about 1040 or 1042.

I would recommend a polymer quench. This will better allow you to adjust your process as needed. You should not need to slowly slip your part into the water. As a matter of fact it is not a good idea as opposite end may cool below transformation temperature.
You are in a quandary because your process is not tuned for the chemistry you are receiving. Your past experience tells me you have a shaft needs a low end chemistry to survive the quench that you currently have. You have options but all are more work for yourself.

You don’t mention where the steel is coming from or how much you purchase. The more you purchase the easier the fix is going to be.

You need to know what the harden-ability if of each batch of material you receive and tune your quench to it. This will clear your cracking problem. I assume you are not buying in heat quantities where you could specify a chemistry range. That being the case you will need to know the jominy for each batch and tune your quench to it. That is where polymer quenchants come in. It’s much easier to change the severity of the quench with the correct polymer additive. You will also be able to rapidly lower the parts into the quench. Agitation of the water is good but must flow across the parts. vertical presentation is the way to go.

Click to expand...

Hey RickyB, thanks for all the good info. Currently, we are unable to change the quenching medium as it is needed for other tools that we are hardening as well as being a 2,000 gallon tank. I am working on ways that do not require a slow dip, and have found that 4 quick dips into the brine and out, followed by being submerged in the brine til it is cool after. this has been yielding better results with the hardness dropping from 68 to 52 which is right where we need it.

I wish we were capable of tuning the operation to the exact batch of material but we use the brine for a variety of parts from different suppliers. We do currently use vertical presentation of our parts.

Salty potato,
Your hardness comment makes me a bit concerned. Let’s take a step further back.
The process for heat treating steel goes like this for all alloy and plain carbon steels. The intention is to obtain a 100% tempered martensitic microstructure when finished. This yields maximum strength, fatigue life and toughness in steel.

1. Heat the steel in a furnace to austenizing temp. 1450-1650F. At this point the microstructure is 100% Austenite.
2. Quench in a liquid medium. This means cooling fast enough that your part is about 60-65Rc. For 1045 this is seconds, not minutes. In this very short period of time the austenite converts its structure to untempered martensite.
3.heat the untempered martensite in a tempering furnace to obtain the final hardness you are looking for. Any where from 350-1000F depending on your final hardness. You should always be starting with quenched parts at around 60Rc

This is the process you need to follow. There really aren’t any exceptions.

Since you are tied to a brine quench you could try increasing the temp of the bath. Possibly follow a batch of larger parts. In other words, watch the bath temp. and plan your run accordingly. Slow the agitation or add a basket which will slow the heat transfer. In and out dipping does not sound like a good idea.

I tried to stress before that there is enormous variability in any give batch of steel. Since you are getting chemistry verification you have the opportunity to calculate the hardenability of that batch. The steel guys should be able to do it for you. You could then tune that heat treat batch. You will then start to understand how it will respond in your process. It will allow you to start making sense of what is happening. You may even go back to parts already processed to see how hardenability caused cracking.

By the way. A deeper groove is significant.

There was thread last year where we talked at length about the reasons for quench cracking. Differential cooling and solidification can result in stresses exceeding the ultimate strength of the material, which is what you have in your flutes. Toning down the quench is really your only option now. Maybe in the future consider using 1040.

Milland said:

It's always impressed me how little Boron is needed to make a "Boron steel".

Click to expand...

it seems to be the magic element (30ppm?). i wonder why its not used more often, except maybe in vehicles (or is it?).

maybe grasping at staws, but I once had a cracking problem with W1 and fixed it by stress relieving the part first. This was not production with new material, but it always left me with the idea that along with everything else mentioned, internal stresses in the material do play a part in whether it cracks. Where the grasping at straws comes in, 'is there going to be variance in internal stress'. May not, but its also not unimaginable that some production change was made. Anyway, might be something to try, if only to eliminate a factor

Mcgyver said:

maybe grasping at staws, but I once had a cracking problem with W1 and fixed it by stress relieving the part first. This was not production with new material, but it always left me with the idea that along with everything else mentioned, internal stresses in the material do play a part in whether it cracks. Where the grasping at straws comes in, 'is there going to be variance in internal stress'. May not, but its also not unimaginable that some production change was made. Anyway, might be something to try, if only to eliminate a factor

Click to expand...

I doubt that it made a difference. The austenizing furnace will do the stress relieving and then some. Quench cracks always occur in the quench tank or a short while after. When austenite converts to martensite there is somewhere near a 3% volume increase. If one area solidifies first and then the section next to it expands as it solidifies, its volume increase pushes on the sections near it. If they are solid sections, they are forced to strain to get out of the way. When this strain causes the stress to reach ultimate strength of the material it cracks. This phenomenon is usually associated with geometry changes like corners, grooves and section size transitions.

Not an expert by any means, but a lot of problems seem to occur from too long a wait between quench and temper.

Your process, which seems to involve no tempering, seems odd. What rickyb says makes a lot of sense to me. Use the temper process to set the hardness, and not quench timing.

There may BE some tempering in your process, when the heat from the core of the part comes out and reheats the outer areas, but that seems to be entirely too variable.

Someone once told me (a smart person with regard to metallurgy) that if you have "percentage cracking" your process is wrong. The issue is not so much that that you have "some" cracking, but really that you have some parts that the bad process does NOT crack.

He said that those non-cracked parts are NOT the good ones, but the BAD ones. They are more tolerant for some reason than the cracked ones, which represent what you should expect. They have not cracked YET, but still may be full of excess internal stresses which will affect the way they perform in the end use, maybe leading to cracking in use.

You need to change the process until you get NO cracked parts.