Heat Treatment - Shaft with Grooves 4

[mtipg]

Hello,

I am an EIT working on a prototype. I have a drawing attached in JPG format of the component I am interested in Heat Treating. I am currently thinking a Rc Hardness of 58-64 at a case depth of 0.060".

I am new to the whole heat treating process, so I have a few questions that I was wondering if anyone could enlighten me
on.

1. What type of Heat treatment is best for this component?
- The Ball Grooves will carry 1 1/8" steel ball bearings so they would be the critical section for wear resistance.
- I am currently thinking that carburizing would be appropriate as the grooves would not be treated proportionally through induction hardening. It this correct?

2. What would be the price difference in getting a Case depth of:

a) 0.060"
b) 0.030"

Big or small would be a suffient answer...

3. How much does a part distort in size/shape from treatment?
- How much extra material should be left on component prior to heat treatment? Our current plan is to Machine, heat treat, then grind material to appropriate size and tolerance. How much extra material should be present before heat treatment of component?

a) 0.001" sufficient?
b) 0.005" sufficient?

4. What is normally the approach for threaded parts?
- As you can see from the drawing attached, there is a threaded piece on the end of the shaft. Will this need to be
machined after heat treatment because of distortion? Is machining after treatment a plausible idea?

5. Not sure which material would be appropriate.
- We want a Rc Hardness of between 58 - 64, but we are not sure which material would be best. The materials I am currently considering are:

a) AISI 4340
b) AISI 3312
c) AISI 9310
d) AISI 8600
e) AISI 52100

Are all these metals easily heat treated? Maybe I am not on the right track here, if anyone has a better suggestion I am
all ears here as this is not my strong point.

Thank you for taking the time to look at my problem. I appreciate any insight anyone can provide and I look forward to hearing from you.

 

[mfgenggear]

mtipg

what qty?

The stock removal will depend the type of Heat treat
& depth of effective case.

for gas carburize
At .020 case .004 stock removal
At .030 case .005 stock removal
At .060 case .007 stock removal

deeper the case the more parts are apt to distort.
because of the longer carburize cycle.
there fore allow for more stock removal.

the case hardness must be held at 15N 90.2 @ .005 case depth. so when the stock is removed at a post heat treat operation it will be at the case hardness required.

My opinion AISI 9310 Steel is best for Carburize.
but it distorts bad. so precautions must be taken.

The deeper the case the cost goes up, from .030 to .060
the cost goes up becase of time & materials usage.
best bet is to request quotes with your Heat Treater.

always machine any parts attributes when it is possible after Heat Treat.

AISI 4340 Steel has less distortion. but tougher to machine post heat treat. because of the core hardness 50 HRc
This part may be a good for AISI 4340 due to it solid geometry.

Looking at the Drawing I have suggestions

Add centers for Inspection, Machining & grinding
add a groove behind the spline to allow shaping of the spline.
or allow a hob or cutter weep out onto the adjacent diameter.

the ball groove reqiure a redius for a ball nose end mill allowance & Jig grind operation if close tolerace.

this is a solid shaft so that will help with the distortion.
Allow a martempering operation for the quenching.
this will also minimize distortion.

request to hang the parts vertically during/at heat treat.

Take Care

 

[CoryPad]

1. The best surface treatment is the one that produces a hard surface that resists wear and plastic deformation. Case carburizing, induction hardening, nitriding, and other processes are used for parts like this. Each process has its advantages and disadvantages. To induction harden this part, you would need a shaped coil, which is more complicated, but still common.

2. Depends on the temperature, atmosphere, etc. In general, diffusion processes like carburizing are parabolic functions (time is proportional to depth squared), so doubling depth may more than double furnace time, which can more than double the cost.

3. Distortion is a function of temperature changes and geometry changes. In general, I would think a 0.001" allowance should work for most things, but you may need more for more complex shapes. This is hard to predict.

4. Usually threads are created before heat treatment, but you will need a localized tempering (induction works well) treatment on the threads to prevent brittleness.

5. You need to specify a surface hardness (~ 60 HRC usually works well), and a case depth (effective case depth to 50 HRC works well). A case depth of 0.040" works for a myriad of parts. Of the materials you list, 9310 is really the only suitable one for carburizing. Materials missing from your list include 4320 and 8620, assuming you will use carburizing.

 

[tbuelna]

mtipg,

1) Carburizing will give you the best combination of case thickness and core strength. To achieve the optimum results, carburizing requires the proper steel alloy.

2) Getting a carburized case thickness of .030 or .060 inch is not much different in cost with regards to the heat treatment process itself. But you must be careful about specifying your case thickness. You should perform a contact stress analysis of your ball and groove contact. This contact analysis will tell you the depth of the max subsurface shear stress that occurs at the contact zone. You should have adequate case depth such that the location of the point of maximum subsurface shear stress lies well within the case layer.

3) Your part has substantial section thickness and symmetry, so it should not have much quench distortion. A .005 grind allowance should be adequate. You can copper mask any areas that don't require case hardening, or thin sections (ie. <.12 inch) that would be prone to through hardening and fracturing. Also, make sure to radius any sharp edges prior to carburizing to prevent through hardening.

4) Machine the threads after carburizing, in order to prevent through hardening and fracturing of the thread crests.

5) For a part like yours with high localized contact stresses, a carburized VIM-VAR 9310 (AMS 6265) would be best. Vacuum de-gassed steels always give substantially better fatigue life under contact stress, since they will not have gas porosity present in their structure which is one of the main causes of case spalls. 8620 and 4620 can also be carburized, but are not widely available as a vacuum de-gassed material.

E52100 is a through hardening alloy, not a case hardening alloy.

4340 can be case hardened by nitriding, but the resulting case will be much thinner than a carburized case.

Finally, your carburized case will not have a constant hardness through the thickness. To be specific, you should specify a hardness range at a certain case depth.

Good luck.
Terry

 

[mtipg]

To all;

Thank you for your responses I appreciate the insight.

To mfgenggear:

"Qty" is of no importance, just quantity of part setup as a default in my CAD program, sorry, should taken it out...

Also, I am not familiar with this term, "15N 90.2 @ .005 case depth", Sorry I wonder if you could break it down for me.

And, I was wondering what the "centers" are in your reference.

"Looking at the Drawing I have suggestions
Add "centers" for Inspection, Machining & grinding"

Thanks again.

 

[CoryPad]

mfgenggear is referring to the Rockwell 15N scale, where 90 HR15N is ~ 60 HRC. The goal is either to measure parts after machining to ensure adequate case depth, or add extra depth requirement to account for any subsequent machining after measurement.

 

[redpicker]

Most has already been said, but I'd like to add...

You seem to be set on carburizing, which I think would make the most sense unless you are considering a manufacturing run in the thousands of uints. With a large enough quantity, you could induction harden the ball grooves. If you go this route, you'd probably want to use a 4145/4150 or similar alloy. Higher carbon than 4340 to give a higher case hardness. Even then, you'd be looking at 53-58 HRC in the case. Processing costs could be reduced, though, and with proper lubrication, that might work. Otherwise, it's carburize.

Your section size on the end you need carbruized is 5.25" in diameter, rather large. You don't mention core property requirements, I'll assume you need somewhere around 28/36 HRC. With this size, I'd recommend 4815/4820. That should have enough hardenability to get you the needed core properties and fully harden the case. With 9310, you run into a problem that a high enough carbon content in the case to get you 58+ HRC will produce retained austenite, which will prevent you from achieving 58+ HRC without a sub-zero treatment.

You will need to protect the splined end from carburization so the splines can be machined after heat treatment. I'd blank out the splined end to the 3.375 diameter and machine splines and other features after HT. A good stop-off paint could be used to prevent carburization or a high temperature temper after carburization could be used so you could machine the case away from those areas it isn't wanted. This would be followed by another Q&T treatment to develop the case hardness. Same goes for the face of the ball groove section where you have those 3/8-16 threaded holes.

I would think that a 0.030" case would suffice as I would think at 0.020" wear the part would be worn out, regardless of case depth. The 15N 90.2 @ 0.005" refers to Rockwell 15N hardness of 90.2, Rockwell 15N being a superficial hardness test. 90.2 15N = 60 HRC. The superficial scale is typically used for case hardenss testing.

Doubling the case depth would typically quadruple the carbuirzing time. I'd assume it would have a similar effect on cost.

The "centers" is refering to center holes along the longitudinal axis, used to align the part for machining and inspections. If they are not acceptable in the final product, the part would probably be made a bit longer and the centers cut into these prolongations and removed after processing.

rp

 

[mfgenggear]

mtipg

Google center drill dimensions, sizes are from 00 & on (10)

centers are machine at both ends with a drill that has a combination 60 deg. included
angle and with a pilot diameter . it allows inspectors to verify runout, concentricity & the grinding dept to grind the parts between male centers that are also 60 deg incl.

It is for ease of manufacturing.

the case on your engineering drawing should specify
the effective case depth
the case hardness "at the surface"
the core hardness
the specification (example AMS2759/7) pull and review this document if you have access to it.
It has a good outline of what the parts should be heat treated.

the case hardness is verified with a part sectioned & verified with a superficial or micro hardness tester. at increments until the case depth hardness drops to 50 HRc, this is your effective case depth.

for manufacturing reference only
for example if there is .005 stock per surface on the diameters, (.010 total on diameter) the hardness as standard practice should be checked at .005" depth during the met lab check.
the 15N 90.2 is a superficial hardness measurement. similar to Rockwell hardness verification which the equivalent would be 60 HRc

I agree with tbuelna about the carburizing only the ball groove to .060 case. if carburized case is required on other areas (spline & bearing diameters) it should be .020 case depth (as an example only)
it's not uncommon to see parts with different case depths. However if the case depth is to deep it will cause brittle areas that have thin sections like splines to be hard all the way through
however for cost & feasibility it is usually better to use one case depth on all areas, if your stress calcs permit.

There is so much info that a good book & reference material is a mandatory reading.

 

[mfgenggear]

redpicker

sorry I was writting & working
ditto on the answers.

Take Care

 

[mtipg]

To Redpicker/Corypad;

Thank you for the posts.

I thought I would mention a bit about core requirements, as I realized I had left some information out. I am expecting a load of about 9,000-10,000 ft*lbs of torsion transmitted from the Spline to the Ball Grooves of the shaft, so the core would need to be tough enough to withstand this torsion. Typically if I wasn't worried about wear resistance, etc. on the ball grooves I would use AISI 4340 for the material, which is common material we use for shafts.


The insight provided in this forum and my own research has brought me to narrow my selection to two designs:

1. AISI 9310 Carburized to 58-64 Rockwell C-Scale Hardness @ a Case Depth of 0.060"

2. AISI 4340 Nitrided 58-64 Rockwell C-Scale Hardness @ a Case Depth of 0.020"


1. Advantages/Disadvantages of Option 1.

a) AISI 9310 untreated has a similar Yield to 4340 and therefore should have a comparable core toughness, suitable to the application.

b) However, using this steel seems like it would lead to a more expensive product as it would take longer to heat treat, and with the deeper case depth, it would be more prone to warping, meaning more grinding as well.


2. Advantages/Disadvantages of Option 2.

a)AISI 4340 has core properties which are ideal to the application and I am familiar with the material.

b)Nitriding has minimal distortion from heat treatment meaning less grinding, etc, maybe therefore less expense.

c) However, I am not sure if I am confident in the case depth of 0.020". From my calculations using the Hertz Contact theory I found the stresses below the surface to relatively high until about 0.045", plus an extra 0.015" for a safety margin, gives about 0.060". This has left me feeling relatively confident with a 0.060" case depth but a little shaky about 0.020".


These are my assumptions, do they seem valid? If so, I could maybe make an educated decision about where to go with this.

 

[mtipg]

To mfgenggear:

Thank you as well.

 

[unclesyd]

May i ask what goes in the stepped recess on the backside?
I was thinking a pilot bearing but the recess is very shallow.
Your call-out has pretty tight tolerances and also does this area need to have a hard case.

 

[mtipg]

To unclesyd;

The stepped recess in the back of the spline is a mount for a thrust bearing.

 

[unclesyd]

Sorry, I should have been a little clearer in referring to the left side of your your section B-B.

 

[tbuelna]

mtipg,

If your contact analysis shows the point of max subsurface shear at .045 inch, then a .060 carburized case is definitely in order. So you answered your own question with your analysis.

As for what material to use, that will depend upon what compromises you are willing to make between cost and reliability. As I previously mentioned, a double vacuum melt (VIM-VAR) quality material will give the best reliability since it will not have issues with gas porosity. A part that has high localized contact loads like in your ball spline grooves, will usually fail by spalling of the case. The spall fractures usually initiate at weak points or defects (like gas porosity) in the case subsurface. Double vacuum melt carburizing steel grades are expensive and not widely available in many alloys. Air melt carburizing steel grades are widely available in many alloys and are less expensive, but are more likely to have gas porosity. The statistical fatigue life and reliability of a part made from a double vacuum melt quality steel will be better by a factor of at least 2 or 3 over the same part made from an air melt quality steel. That's why even commercial grade rolling element bearings always use high quality steels. The cleanliness and porosity-free nature of the steel won't improve load capacity, but it will significantly improve the statistical (ie. L10) adjusted operating life.

Hope that helps.
Terry

 

[CoryPad]

This is a correction regarding high-quality deoxidized steels like VIM and VAR - they have improved fatigue performance because they have fewer non-metallic inclusions, not because of gas porosity. Thermomechanical processing of steel from casting to final wrought product (bar, plate, wire, etc.) eliminates porosity, but retains inclusions. Vacuum processing eliminates oxygen that forms oxide inclusions.

 

[metengr]

I will also add another correction, in my dealings with subcase fatigue failures in gears and roller bearings for wind turbines, the cause (99.999% of the time) is usually a subsurface inclusion located near the case/core boundary, which corroborates the statement by CoryPad.

 

[tbuelna]

CoryPad,

Thanks for the correction.

I went and checked my bearing text and it concurred with what you state. Except that it suggested that the repeated hot rolling during the mill processing of wrought materials (or thermomechanical processing as you described it) does provide some mitigation of nonmetallic inclusions flaws. Apparently, the repeated mechanical working of the material breaks down the brittle nonmetallic inclusion structures (oxides, silicates, etc.) into many smaller inclusion "defects" and distributes them within the metal such that they pose less of a potential fracture nucleation point.

Regards,
Terry

 

[unclesyd]

I was just checking my data and noticed something that Bethlehem Steel did when carburizing steels. They offered to different procedures for either maximum core properties or maximum case properties. In both cases they used a procedure called double quench and temper to get the optimum results. About the only difference is that for case properties they tempered at 300F and for core properties they tempered at 450F.

 

[mfgenggear]

I like it when our customers specify AMS6265 Vacuum melt vs AMS6260. It has been my experience & observation that It has less distortion, better met lab results, & as tbuelna & everyone else has said it just makes better parts.

Every aircraft part I have made thus far has been
Optional to normalize, Harden temper for machining to refine the grain structure & reduce internal stresses.

It has been my observation parts do not distort as much
during carburize & harden if the above process is done.

Usually a stress relieve after heavy turning & milling is Mandatory. but these add cost so the application, the cost, & the importance of the parts have to justify it.

If the case depth is for example .010-.020 then I will hold it to .016-.022 this allows for stock removal & gives the heat treater enough point spread to work with. Then when it is post machined "normally ground" it will yield the final blue print requirement.

what I have observed with 9310 steel is depending on the geometry, for example if it's a ring or hollow shaft it will contract, go out of round, & taper, if it's solid part it will slightly expand. These observations will vary from part to part & also from heat treat processes & facilities.

So generally I stick with one vendor or heat treat facility.
The trick is to know the product & what to expect then allow for the unexpected. I am usually more conservative,
when I need to be.

Thanks & Take Care