Re-heat-treating 4140 6

[dmalicky]

We are making torsion bar springs (like an anti-roll bar) out of 5/16" diameter 4140, 36" long, and are on a tight time frame. At a safety factor of 1.3, we need a yield stress of 180,000 psi. They won't see more than a thousand loading cycles. They may see a little (20%) reverse torsion.

I have found 5/16" 4140 locally, in TGP Q&T form, Rc30, probably around 130,000 psi yield. We will need to have the bar re-heat-treated, probably using a 600F temper (achieves a yield of 195,000, per matweb).

My questions are...
What (if any) special procedures need to be taken to prepare the Rc30 bar for another heat treatment?
Are there any problems that are likely to occur from this double-heat treatment?

If this is too problematic, we can probably find annealed or normalized 5/16" rod and have it air shipped in, but it would push the timing.

Thanks,
David Malicky

 

[metengr]

dmalicky;
Check for any decarburization before using this bar in a spring application. If you have no decarburization, the 4140 alloy steel bar can be re-heat treated. In fact, for quench and temper heat treatment, it is recommended to have the 4140 normalized at 1600 deg F followed by reheating to 1550 deg F, and quenched in oil.

For your particular case, since the 4140 has been heat treated previously, I believe you could heat directly to between 1500-1600 deg F, oil quench and temper at 600 deg F., The only precaution is that for a spring application, the surface must be protected from decarburization during the second re-heat treatment.

 

[metman]

Special preparation:
1) Degrease
2) Consider cutting longer than necessary to drill a 1/8 diameter or larger hole transverse near one end. The heat treater can then hang the 36"+ rods vertically for heat and quench. This will minimize distortion. But check with your heat treeater first because they might have a basket that will provide for vertical processing.

I do not think there is any problem re-heat treating but wait for a few others to concur on this.

Jesus is THE life,
Leonard

 

[swall]

There are two things about this project that concern me. First of all, you are really pushing the limits of standard 4140 to get 180,000 psi yield strength out of it. I'm also concerned about the 600F temper.As a rough rule of thumb, tempering in this range is avoided because of the possibility of temper embrittlement.The Heat Treaters Guide section on ultra high strength steels shows 775F as the lowest tempering temperature recommended.This can get you a 220-220 ksi tensile strength, assuming that you quench out at Rc 53-55.In your situation, with 5/16" bar, you should be able to hit this.

 

[unclesyd]

I agree with swall the tempering temperature concerns me in several ways. One as stated above and the second is the TS #'s given for the small diameter rod look low.
If you quench from 1550°F and temper in agitated oil @ 600°F you should have a TS of 250 ksi; a YS of 228 ksi.

Our policy, if the part intricate, was to full anneal(1500°F;20°F/Hr to 1200°F) any 4140 prior to reheat treating.
If the part was not intricate we would Normalize (1600°F, Air Cool) prior to the Austinizing @ 1550°F and minimum tempering temperature of 1000°F. There were few times we tempered any lower.

One other point to ponder is we found 4140 to be a lot more notch sensitive when tempered below 900° F. This was especially true in a condition where there was a high rate of loading, shock.
4140 impact properties are not very good at 600°F. Again we believe that under actual conditions the values were lower than some published values.

 

[dmalicky]

Thanks to all for your helpful replies. I'm a mech. engr., not a metallurgist; let me know if/where this is incorrect.

On previous decarb, I am guessing this will be ok since they were turned and group to the 5/16" dia. I was wondering if they couldn't just be austenitized directly w/o a separate anneal operation. 1/8" hole--also good idea.

I didn't know much about temper embrittlement but from what I've learned it doesn't appear to be a problem for this particular application. From a web search... e.g.,

http://www.key-to-steel.com/ViewArticle.asp?ID=102

http://www.oaklandsteel.com/tool.html

http://www.tmtco.com/bar_stock.html

...I gather it primarily affects impact properties and it is reduced or eliminated via the moly addition in 41xx steels. And, our 4140 is for a torsion spring (and for a door), and thus should not see significant impact. Does the Heat Treaters Guide specifically mention 4140 for the 775F rec?

I am not seeing how 180,000psi is pushing 4140--please explain. Matweb shows 4140 w/ a 220,000psi yield at a 400F temper, still with 8% elongation (this in a 1" dia. bar). On a .53" bar, a Bethlehem steel graph (in Juvinall) shows 4140 going to 252,000psi yield at 400F temper. Not sure if these are normal or idealistic; their #s are ~12,000psi higher than matwebs 1/2" bar at 1000F. I'm assuming these yield stresses represent the average of a pure tensile load across the cross section; higher yields expected near the surface (at least for the 1" bar) and throughout a 5/16" bar.

But yes, we can raise the temper temp and still get our yield: assuming the .53" dia graph is a better guide, (180,000+12,000)psi yield corresponds to 825F temper. So we could spec a 775F temper and probably be fine.

We will have the heat treat farmed out. Should we spec the temper temp or the final hardness desired, or both? If hardness, I am intepolating Rc 45 from matwebs data (though again this is for a 1" bar; not sure how much less a 5/16" bar would need to get 180,000 yield).

Thanks,
David

 

[metengr]

The bar can be re-austenitized without an anneal if the original heat treated bar was turned down to the 5/16" diameter. However, as I pointed out earlier, decarb must be prevented during the second re-heat treatment cycle to avoid a significant reduction in fatigue strength for this material. Any decarb layer on the surface of the spring is asking for premature failure in service.

I would tend to agree that temper embrittlement (TE) is probably not as significant problem for your application. The addition of molybdenum is effective in reducing susceptibility to TE (optimum is 0.70%).

What you do not see in postings of mechanical property data is fracture toughness. Tempering at 600 deg F will enhance notch sensitivity because of reduced fracture toughness resulting in less tolerance for cracks, which could be a problem under torsional loads.

The 800 deg F temper would probably provide you with adequate tensile/yield strength, and most important toughness. Keep in mind that the fatigue strength for low alloy steel is about 50% of the ultimate tensile strength, so tempering at 800 deg F will provide you plenty of margin regarding fatigue strength.

I would specify a hardness range rather than tempering temperature. According to my reference material (ASM Handbook), 4140 alloy steel tempered at 800 deg F results in the following properties;

181 Ksi UTS
165 Ksi YS
13% Elongation
49% RA
370 HB


Tempering at 600 deg F;
225 Ksi UTS
208 Ksi YS
9% Elongation
43% RA
445 HB

The data above were obtained on .505 in dia rounds. You should specify a hardness range of 380-400 HB.

 

[unclesyd]

dmalickey,
Have you already purchased any quantities of the heat treated bar?
If the quantities are large enough there are people that will supply the bar as needed and all you have to do machine to size.

meteng,swall,

Take a look at the following website.
Checkout the bar and rod database.

http://www.steel.org/autosteel/bar_rod/index.htm

 

[swall]

dmalickey--one option I would like to throw out would be the use of Cr-Si wire for this application. We make small torsion rods like this and that is what we use. For 5/16 diameter, the Cr-Si wire will come already heat treated at 250-270 ksi. Then you just form it and stress relieve and don't have to heat treat. But, given your short time frame, probably not an option. My aversion to the use of 4140 above 180 ksi tensile comes from my aircraft landing gear days. At higher strength levels than 180 ksi, we would go with 4330, 4340, 300M or H11.

 

[strokersix]

dmalicky,

I'm sure you already thought of this but I'll post anyway. If you are weight constrained you may wish to consider alloy tubing instead of solid round to make better use of the material. If you are size constrained at 5/16 then you are probably stuck with the solid.

 

[dmalicky]

Thanks again for everyone's input. I am seeing 4140 may not be the ideal material for this app, given what I've read from your posts and in Metals Handbook (blue brittleness) last night at the library. I was about to write that the bars have been bought, bent, and are on their way to the heat treat shop, but apparently they were more than bent: gouged, hammered, bent at different angles... not usable (and not Show quality, which is what this first round of bars is for). So we are doing it twice instead of right, but are now wide open to material choices... within these requirements:
- 5/16" dia.
- Shipped ASAP, reasonable cost, and readily heat-treatable (i.e., off-the-shelf, not exotic)
- Each bar is ~36" long has a 90 deg bend on each end, through a 2" radius bend (total pre-bent length = ~46" --> less than 6' --> shippable overnight) A 90 deg 2" rad bend on 5/16 rod is +/- 7.8% strain. So for forming we would like to buy it in a low yield stress condition, and would like 15%+ (?) ductility.
- Higher yield strength: at least 200kpsi (depends on some unknowns and final adjustment procedure... this is to counter the weight of a ~100lb upward swinging door, swung via a 4 bar mechanism through (previously) 95 and perhaps 105 degrees of torsion bar twist--the last is the main unknown).
- Low risk of brittleness (we don't expect major impacts, but...).
- Fatigue not a major concern.
- Weight is not a factor

From a quick look, 4340 tempered between 400 – 800F may be a better choice: readily available, 198-270ksi yield, better toughness than 4140 at those yields. On toughness, Metals Handbook shows it to have better Izod impact at 400F temper (20J) than 800F temper (16J). 1000F temper gives 47J, but only 168ksi yield. On ductility, Metals Handbook shows 11% at 400F temper, 14% at 800F temper; but, my Bethlehem graph shows ~2% at 400F and ~4% at 800F—low #s and quite a discrepancy. Matweb seems to confirm the higher ductilities, but they don't list rod size for those #s—any one have experience or other sources at low rod sizes and a ~800F temper?

If ductility is a risk for high yields in 4340, 4140 does have this advantage--all my sources say it maintains 10+% ductility down to a 400F temper, and has 15J Izod at that temper (up from 9J at 600F). Maybe 4140 at 400F temper is an option?

I certainly like the Cr-Si idea w/ only a stress relieve from Swall (Thanks!)—My main question is, how hard are they to bend if they are 250ksi? Could you tell me more about it--material spec, where to buy it, min. bend radius, stress relief procedure?

Thanks again,
David

 

[swall]

Should be able to bend a 2t radius.Typical spec would be ASTM A401.Stress relief is 750F for an hour.Sources for this material would be wire warehouses, such as Gibbs.

 

[dmalicky]

Thanks, swall, for the info. Unfortunately I can't find anyone who will sell ~5/16" Cr-Si in anything less than a full coil (1500 lb). Tried Gibbs, Interwire, Seneca, AWS, Wiresource, Amstek, a few spring shops, and a number of metal suppliers. Interwire can probably supply us w/ cut pieces in August, so it may be a good long term sol'n. In the meantime 4340 is looking good. Thanks again, David