Involute spline distortion at heat treat 7

[MrSparkles]

Greatings all,

We have recently nitided a batch of parts with an involute spline that has been finish cut before hand. The splines were nearly perfect before nitride, now the splines will not fit into the mating part. Basic measurements, MOW, major diameter etc., still measure in tolerance but have grown slightly.

Is this a common occurance and is there a way of lapping/polishing the flanks just enough to compensate for this distortion so that the mating part will fit?

Thanks,

 

[gearcutter]

Yes, this is a common occurrence. Measures ought to be in place to help limit the distortion on critical parts and on non-critical parts we usually allow for the slight distortion by adding extra backlash to the parts.
I know of no lapping/polishing process you could use to recover the parts.
If the splines are external; you could try gear grinding to correct the problem.
If the splines are internal; we have successfully recovered small batch quantity parts by having the involute profiles wire cut.
With all these methods there is the chance that you may end up losing your effective case hardening depth, particularly with a nitrided case.
The process we use to help control distortion on critical parts is;

Rough turn/mill.
Stress relieve.
Through harden to 32-35 HRC.
Finish turn/mill.
Gear cut.
Nitride.


Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[mfgenggear]

MrSparkles

When Heat treating finished Gears or Splines you must hold the attributes tighter than what is required. or have some method of machining after Heat Treat.

I do this often.

Splines have to be lapped if there is no clearance.
if there is clearance then formed or pencil ground.
one tooth at a time.

gears can be automaticlly power honed or ground.
if the effective case depth permits.
and thats an other whole issue there.

for automotive this is an expensive proposition.
for Aircraft it is common.

You must tighten the M.O.W or M.B.W. to allow for contraction or expansion. Index error, or out of round.
and as Gear cutter said parts will stress relieve during
nitride if not properly stressed. And you have to know the
material & the nitride process you are working with.

 

[tbuelna]

Nitriding generally produces very little distortion, since the process temperatures are relatively low and no quench is required. Distortion during nitriding is usually due to lack of adequate stress relief during the pre-nitride heat treatment.

Dimensional change during nitride is very minor, I'd estimate less than 10 microinch/inch. Not enough to change the fit of a spline.

In addition to the steps that gearcutter outlined above, after rough machining followed by the stress relief, quench and tempering, the secondary (pre-nitride) machining should remove enough material to eliminate the decarburized outer material (usually about .030 inch). Also, there must be a final finishing operation performed after nitriding to remove the brittle surface "white layer" that forms during nitriding.

Finally, with gears and splines, you should carefully radius all external edges prior to nitriding, to prevent through hardening at these corners. The through hardened corners will end up fracturing in service.

Good luck.

 

[israelkk]

gearcutter mention using wire EDM to correct the distorsion. However, wire EDM produce hydrogen that may create hydrogem embrittlement. Taking into account the very high hardness of the case this process is very risky.

 

[gearcutter]

israelkk,

Thank you for pointing this out.
This is the first time I've heard of Wire EDM creating a Hydrogen Embrittlement issue.
I’m very keen to learn more.
I am aware of other processes that can create this problem (in hardened steels with high levels of alloying elements) such as hard chroming & some welding procedures.
Could you please explain the mechanisms involved that produce the hydrogen when using Wire EDM and also the way in which this process causes the hydrogen produced to penetrate the metal.
I suspect it must have something to do with the plasma stream created at the arc.




Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[israelkk]

gearcutter

I have encountered this problem with wire EDM after we manufactured an EDM custom machined helical spring made of Custom 455 H1000. The part was machined as hollow tube from a solid bar and them heat treated to H-1000. After heat treatment the helical gap was cut using wire EDM. First time the spring was loaded it was shuttered to many tiny pieces. Metallurgy inquiry revealed that the EDM process produce hydrogen that created the hydrogen embrittlement. The correction action was to heat treat after EDM machining the helical cut.

 

[CoryPad]

gearcutter,

The mechanism is essentially the same as electrochemical corrosion. A potential difference exists, an electrolyte is present, the metal is dissolved/eroded, and hydrogen is produced electrochemically, which can then enter the steel.

I agree with tbuelna that nitriding is a low-distortion process, so the corrective action is stress relief prior to nitriding.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[dinjin]

I am wondering if the parts could be broached to get rid of the white areas. No mention of size has been offered or the depth of the nitride. Before and after measurements need to be taken to know the extent of change on these parts and then they can be machined undersize or oversize before nitriding depending if they are internal or external splines. I assume the same thing happens with nitrided parts as happens with induction hardened parts. Our induction hardened gears shrink. I think we see more change on internal gears.

 

[mfgenggear]

Gentlemen

Yes nitride is theoretically a low distortion process.
however let me say this depending on the material used, nitride process used. The amount of machining prior to nitride the geometry of the Splines or Gears manufactured,
All will effect the parts during Nitride. If white layer is not allowed then there will be material removal by mechanical, chemical or both. Been there done that.

Even variation from heat treat suppliers will change your attributes.

Regardless if all precautions that are taken, there is additional stress relief or expansion of gears and Splines during Nitride.

For example a part made from 4340M steel nitrided by the AMS2753 : Liquid Salt Bath Ferritic Nitrocarburizing Non-Cyanide Bath. Parts expand during this process.
been there done that.

Parts nitrided to ams2759/6 gas nitride do distort very little, however cut your splines and gear to blue print size and I guaranty there is a chance you will be reworking these parts.Then what about the white layer? need to adjust size for white layer removal.

I been manufacturing aircraft gears for a few years I say that there is no one recipe for all Gears and Splines.

I have been Wire EDM gears and Splines for years and if the
EDM process is done carefully. 0005-.001 max recast layer
then the recast surfaced is removed from the gears and Splines then the parts are fine.

Just like any other Manufacturing a carefully planned
process must be Taken. For example Copper Plate , Then Bake Parts at a Temperature "as not to effect part Integrity" to remove Hydrogen embrittlement.

just my 2 cents. thats what makes the world go a round.

 

[gearcutter]

Thanks israelkk & CoryPad and everyone else.
I'll be doing more research on the Hyrdrogen issue when Wire EDM.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[WinstonH]

Start making a note library, for different materials, Nitraloy, chrome-moly, and the different stainless groups, 17-4, 15-5 etc they do move differently, record before & after dim's, to obtain you own experience of fudge's of so much per in/in or mm/mm.

AFM. Abrasive flow Machining was the way I handled oversize externals,undersize internals, until one can get to predict quite well, in some cases you will literally, SCRAP the part dimensionally before Nitriding.

As stated previouly, lapping or honeing. Many Aerospace Company Spec's required the white layer to be removed, If done chemically, 2HR bake at 50°F below temping temp.

Cheers,

 

[MrSparkles]

Thank you all very much for your advice. White layer is not much of a concern here, we have been planning to have the splines finished before nitride. My biggest area of interest here is on lapping methods. The splines are very very close to being in tolerance and I would like to find a lapping process that will yield the most accurate results.

Cheers,

 

[mfgenggear]

Sparkles

This is for small quantity.

Use a bead blaster to remove small amount of stock, verify parts with ring go gage.

or

Depending on the size of your splines.
make from cast iron, recommend round stock

Saw several pieces long enough to over lap the length of your spline.

Machine all over make minor diameter smaller than your TIF. by .005-.020, best to match your go gage.

Shape Internal spline laps with full composite. Fixture will be required.
Make laps loose fit to parts.

Use standard lapping compound or equivalent.

If parts are small & are of low qty parts, use a shaping adapter from a bridge port type mill, or a similar machine.(Automatic spline lapping machines are getting hard to find.)
It is optional to hold the laps or the parts in the shaping adapter.

You will need fixture to hold parts.
Use a chuck or equivalent to hold your laps.
Or vice a versa

I recommend for a large quantity, source out to a gear manufacturing company with Automatic lapping.

 

[tbuelna]

It is usually required to remove the nitride white layer, since it is very brittle and prone to crack. Fortunately it is very thin and can easily be removed by chemical etch (ammonium persulfate?), abrasive blast, or during a final grind/hone. Copper plating can also be used as a mask to prevent hardening of selected surfaces. The copper plating and white layer will both be removed by the chemical etch.

Stock removal on nitrided surfaces should be carefully controlled, since the nitride case is usually fairly thin (<.020 inch) and the case hardness decreases very rapidly as you penetrate into the case. I would not recommend removing more than about 10 or 15 percent of the effective case during final grind. Nitride case depth is primarily a function of time in the nitride furnace.

Partially removing a nitrided case on a part with radial symmetry (like a splined shaft) is usually not a problem. But if your part is asymmetric, grinding away the case can cause distortion. The reason is that the nitriding process creates a residual compressive stress in the part surface. Mostly, this is a good thing for stuff like fatigue life. But machining away the case relieves this compressive pre-stress, and if done asymmetrically it will cause distortion in the part.

 

[gearHTr]

1. Citric acid will also work well to remove white layer.

2. If parts were not core hardened prior to machining and nitride, that may have contributed to the distortion.

 

[mfgenggear]

as well as phosphoric or metex acid

with my suppliers white layer and copper strip are run seprately.

GearHTr

That was a good point over looked. I also core harden prior to machining if the core hardness permits.
Star for you.

 

[tbuelna]

mfgenggear,

Couple questions:

Can you provide additional information on Metex etch process for white layer removal? A process spec maybe? The only reference I could find was from Metex vendor MacDermid, but it wasn't much help.

What is your(or your supplier's) preferred method for copper mask strip, since it is performed separate from white layer removal? I specify AMS 2418, cl.1 for masking, but it does not cover stripping. I usually rely on chem bath for white layer removal to also strip the copper. But if you have a better process I'd like to hear it.

Thanks in advance for your response. I'm a designer, not an M&P guy, so I appreciate your input.

 

[mfgenggear]

tbuela

I wanted to comment about your statement of breaking the edges prior to Nitride,
This is very good Advice indeed, On My projects it is Mandatory to break all Nitride edges.
The stuff is hard and brittle, and will chip instantly.

The amount of stock I leave is regulated by my customer’s specifications. And I must adhere to their requirements.
But I have found from personnel experience sometimes I can fudge the stock removal if
I hold the case depth to the high limit or even slightly over case. And hold the surface hardness requirement at the depth of the stock removal.

My vendors preferred method of copper stripping is ammonia & white layer removal is with phosphoric Acid then a hydrogen Embrittlement bake is always required there after.
In general there has been very little issues with this method.

Metex is a very aggressive in stock removal.
And care must be taken during its use. Parts can be deviated in a flash.
The procedures are my customer’s proprietary specifications. Not aware of any commercial specifications. However the manufacture should be able to advise on mixture and concentrations. Once set up it’s a matter of trial and error with test pieces.

I personally like phosphoric acid because I have experience with it and I can estimate approximately the amount of stock removal in minutes.

If the two stage floe process gas nitride is used AMS2759/6.
White layer on my projects is generally held between .0001-.0003 sometimes .0005 max, so generally with finished gears or splines this is the maximum stock removal.

With Ion/Plasma Nitriding (AMS2759/8) on my projects no white layer has been present, or .0001 maximum of gamma prime (Black Compound) &. or upto 0001 max epsilon compound (white layer).

Hope this helps,
Happy Holidays to All

 

[tbuelna]

mfgenggear,

Great information. Thanks!

Terry