material question, 4140 vs 4340 hydrogen embrittlement

Parkerbender · Mar 22, 2017

We have been making a product that gets these 3/4" ball, m10 thread ball studs. They always break. Out of like 64 studs maybe three or four per batch break. Doesn't matter if they are sitting in a driveway or running them over the edge of a cliff, and they also break in seemingly random places. We started with 12l14 because this application should only see 500-700lbs. We have worked from there, through 1045, 1144, hardened 1144 to 50rc, 4140, 4140 @ 50rc, 4140 @ 40rc.. Everything breaks, basically the same rate. All of these get Inc plated, we are wondering of the weaker stuff broke because it was weak, and the harder stuff broke because of hydrogen embrittlement. If that were the case, will I encounter the same problem with 4340 normalized bar, or does the normalizing part kindof protect against the embrittlement problem? Otherwise are there better material finishes for steel that do not cause this problem but would hold up well under a car?

Getting really tired of making these parts, as they are wrapped into a package price, so I am usually remaking them for freeish. :-/

rklopp · Mar 22, 2017

Can you show us photos of fracture surfaces? Do the fracture surfaces look the same regardless of whether 12L14 or 4140? Do you bake after plating (assuming you meant Zinc, not Inc ;-))? If fracture surfaces are relatively smooth and flat, then H2 could be the issue. If fracture surfaces are rough, associated with lots of plastic deformation, and not flat, then H2 may not be the issue.

Mtndew · Mar 22, 2017

So they're not breaking in heat treat, but when the customer uses them?
My first instinct is to say a design problem, but if they are breaking in random places then that pretty much rules out that idea.
I don't know anything about Hydrogen embrittlement, hopefully others smarter than I can help in that area.

TDegenhart · Mar 22, 2017

First, lets establish what hydrogen embrittlement is and how to handle it. Hydrogen embrittlement only occurs with hardened steel in the hardness range of 44Rc and up. Ya, I know, +-a few points. It can happen when hardened steel is exposed to hydrogen, often from plating baths. The part needs to be under stress. The breakage will be random in place and time, but will be most likely in thread roots or other points of stress concentration. Parts that have exposed to hydrogen usually are baked for at least 2 hrs at 400-500F. These are general rules.

12L14 is not a candidate as it is too soft, as the unhardened 1045 and 1144. Its unlikely that 40Rc would be a problem, 50Rc is in the zone, but if you are experiencing the same type of failure across all the different alloys and treatment, hydrogen probably is not your issue.

You need to define the application a little further. Where are the ball studs used and is it a cyclic load or static? 3 or 4 per batch break, how? Are the screw stretched as in over tightening.

Your last comment seems to indicate that they are breaking in the shop before shipment. Is this the case? If so where are they found to be broken?

We need more details.

Tom

9100 · Mar 22, 2017

Some people do not believe in hydrogen embrittlement, but I can say from personal experience that it is real. Normally the hydrogen will come out by itself over a period of time but it can be accelerated by baking. The typical bake is well below the tempering temperature so it will not alter heat treat. Here is a link-

http://www.sciencedirect.com/science/article/pii/S0010938X0300180Xe

If you are chroming, you may also be introducing stress points due to cracking of the plating. I have never figured out how a layer added one atom at a time can produce a stressed layer, but it does.

Bill

9100 · Mar 22, 2017

Some people do not believe in hydrogen embrittlement, but I cas testify from personal experience that it is real. The hydrogen will work out eventually, but baking accelerates it. Here is a link-

Hydrogen embrittlement of high strength steel electroplated with zinc–cobalt alloys

Also, if you are chroming, cracks in the chrome may produce stress points. I have never figured out how a layer that is applied one atom at a time can go on stressed, but it does.

Bill

Dave G. · Mar 22, 2017

Sounds like it's time to send a couple representative broken ball studs to a metallurgical lab for a failure analysis. In addition to the broken stud, send a new unplated stud for reference. They will ask for: 1) Steel Grade, 2) Process route for the part, 3) Process route for the plating, 4) location of the failures (on the part), 5) service condition (what's the part doing when it fails), and maybe some additional questions. If you what us to help here - we'd probably need the same info, however - without testing the actual failed parts - we're all just guessing. A failure analysis from a qualified lab is expensive, but replacing parts that are constantly failing for unknown reasons is costing you time, money and your reputation.

Matt_Maguire · Mar 22, 2017

Start here & read till you’re tired. → Hydrogen Embrittlement in Plating

So yes it’s a possibility 4340 @ 320Bhn used to be used for landing gears & boy do they have a stack of rules when you want to electroplate it.

Good Luck,
Matt

Milland · Mar 22, 2017

Parts breaking in the range of materials you describe and with the variety of loads means (IMO) that there's a design issue. Can you post pictures of the part, break areas, and how the parts are used/loaded?

Parkerbender · Mar 22, 2017

Here is a pic of the last ones 4140 at low 40's... I think that my range was 38-42 or something.

This is clear zinc, all the ones before were yellow zinc.

They don't fail in the shop, and I have even grabbed several of them in a vise and beat the hell out of them with a hammer to see if I can get them to fracture, they just bend, slow and nice.

Sometimes they break around there, at the bottom of the threads, sometimes they break on the neck right before the ball.

Good to know that if we stay in the 30's that hydrogen should not be a problem...

Milland · Mar 22, 2017

Truck part? I'd like to see more of a radius at the thread to body junction, and is that a poly bushing to ball, or is the poly just a dust shield and it's a metal/ball swivel?

Matt_Maguire · Mar 22, 2017

Parkerbender said:
Here is a pic of the last ones 4140 at low 40's... I think that my range was 38-42 or something.

Good to know that if we stay in the 30's that hydrogen should not be a problem...View attachment 194430

It'd be good to have a macro picture of that busted part...

BTW for maximum toughness with 4140 = 1200F temper = 29Rc hardness, for 4340 = 1200F temper = 31Rc. They are still not safe from hydrogen embrittlement.

Matt

TDegenhart · Mar 22, 2017

Does the failure occur when the end product is moving or stationary? Is it something like swing link on motorcycle? Did you follow all the standard rules for fatigue issues by having smooth corner radii?

Tom

Dan from Oakland · Mar 22, 2017

Are you having the plating shop bake the parts after plating? I always thought it was standard practice to specify a post plating bake on any kind of stressed parts. It does not add much cost from my experience.

Dan

magneticanomaly · Mar 22, 2017

As well as I can tell from the not-very-good photo, it looks like fatigue failure. I suspect cut threads and-or sharp corners.

Auto steering linkage studs use a tapered shank and seat for a reason...threaded sections loaded laterally are not good in fatigue.

robert123 · Mar 22, 2017

I don't know if hydrogen embrittlement is your problem, but the most common zinc plating spec ASTM B633 has both pre-plate stress relief and post-plate hydrogen embrittlement relief baking requirements.

Parts having a tensile strength 1000MPa or higher (~HRC 32) that have been machined, ground, cold formed, or cold straightened should be heated to 375 F for a minimum of 3 hours before cleaning and plating. If you are machining pre-hardened material, you or your plater should be doing this.

Parts having a tensile strength if 1200MPa or higher (~HRC 39) should be heated to 375 F for a minimum of 3 hours within 4 hours after plating. Baking should be done before the application of the chromate.

An experienced plater once recommended baking at even lower temperatures than required in the specs (maybe HRC 32?) to avoid problems.

Oldwrench · Mar 22, 2017

I'll add that the hoop stress from welding a circular section makes it vulnerable to hydrogen embrittlement even if the part is low carbon steel. Stress raisers like sharp internal corners and thread runouts also seem to be aggravated by plating. We have standing orders that ALL plated parts be baked, BTDT.

Parkerbender · Mar 22, 2017

I apologize for the picture quality, it is my customer's customer's picture...

It is for a tie rod endlink. The socket is an Energy Suspension polyurethane bushing with a ball socket cavity.

I believe I have a .035" radius in the bottom there, though the break there appears to be a bit further from the shoulder... I could put a little bigger radius but the part that it goes through doesn't have a very big chamfer or bevel for it to fit into...

How would you make those threads other than cutting them? I am making these things in ~100 quantity, for $4.50.

I use Lincoln Industries to do my plating. I havnt ever heard of them doing baking or anything, but if you guys say it is common, maybe I will ask. They are pretty darn huge (do lots of harley, paccar, roush/Shelby, etc's parts) so if it is common I would assume they do I guess.

The factory part has studs that are very close to identical to this one... I am tempted to buy one at the local o'riley's and send it off to see what the hell Jeep is making them out of... Is it very expensive to have an analysis done, I have never done that before...

Pete F · Mar 23, 2017

Matt_Maguire said:
So yes it’s a possibility 4340 @ 320Bhn used to be used for landing gears & boy do they have a stack of rules when you want to electroplate it.

Good Luck,
Matt

Matt I thought that was still used. What is normally used now?

robert123 · Mar 23, 2017

Parkerbender said:
I apologize for the picture quality, it is my customer's customer's picture...
I use Lincoln Industries to do my plating. I havnt ever heard of them doing baking or anything, but if you guys say it is common, maybe I will ask. They are pretty darn huge (do lots of harley, paccar, roush/Shelby, etc's parts) so if it is common I would assume they do I guess.

If they are aware of the hardness, I would think they would be baking as needed. If you didn't tell them and they didn't ask, they probably ASSUMED that they were not hardened and in need of baking.

The factory part has studs that are very close to identical to this one... I am tempted to buy one at the local o'riley's and send it off to see what the hell Jeep is making them out of... Is it very expensive to have an analysis done, I have never done that before...

The larger local recycling place near here has an XRF gun and once shot a part for us to determine the composition and charged us $25. Somewhere that could do hardness testing also would probably be more helpful to you. Perhaps your parts are simply too hard? Maybe they need a lower core hardness with a higher case hardness to prevent wear?

material question, 4140 vs 4340 hydrogen embrittlement

Stainless

Diamond

Diamond

Diamond

Diamond

Diamond

Cast Iron

Stainless

Diamond

Stainless

Diamond

Stainless

Diamond

Titanium

Titanium

Stainless

Titanium

Stainless

Titanium

Stainless

Similar threads