White layer removal after Nitriding 3

[724napier]

Hello all,

I am looking for some expert guidance on some spline couplers I had made. To make a long story short, I had two spline couplers made that will be used for ground testing of a turbine engine. The torque on the adapters is relatively low, however they will be spinning at high rpm. My problem is that when I received the adapters back from nitriding, the vendor that did the nitriding sand/bead blasted them (most likely to remove the white layer). They were given specific instructions to nitride and remove the white layer in accordance with the aerospace nitriding specification.

I am unsure of whether or not this will usable now, due to gritty surface finish. Is this a common practice and do any of you have recommendations on how to proceed from here? I have attached pictures of the adapters. I was thinking of having the teeth of the splines polished to improve fatigue life. I hate to scrap these parts as a lot has gone into making them. The specs on them are as follows.

Material: Nitralloy 135M
Standard two stage gas nitriding process

 

[CoryPad]

You could use isotropic surface finishing to reduce surface roughness. This is used on hardened parts like gears. Contact REM:

http://www.remchem.com/

 

[mfgenggear]

724
or hand buff & polish, since there is only 2 parts.
what is micro finish requirement? 63 micro is standard
The heat treat supplier should have used vapor blasting/honing.
that appears to be aluminum oxide blasting.
maybe to heavy of a grit size.

it may look worse than it is. however make sure the splines are to print.
I forbid my suppliers to blast the splines with aluminum oxide. or we do it in house.

Mfgenggear

 

[tbuelna]

As CoryPad suggested you can use ISF to improve the surface roughness of the tooth flanks.

However, one thing that caught my eye was the type of spline teeth you used for these high speed couplings. They appear to be fillet root, side fit type teeth. The high-speed output shafts of turboshaft engines typically use a major diameter fit spline with crowned external teeth on each end. This permits some misalignment capability between the engine and load interfaces connected by the drive shaft.

One issue with using side fit spline teeth for this application is that the ends of the shaft are not tightly constrained in the radial direction, which means they can shuffle around quite a bit due to dynamic unbalance forces or shaft whirl. Using a major diameter fit spline at each end of the quill shaft provides a radial constraint at these locations, and this helps with the dynamic situation in a long slender high-speed shaft.

Just one other suggestion. From the photos it looks like this steel coupling is fairly heavy and is cantilever mounted at one end. So if you haven't done so already, it would be a good idea to have it dynamically balanced prior to installing it.

 

[724napier]

Thanks Guys,

I will look into the isotropic surface finishing process. Do you guys know how much material this removes from the surface?

Tbuelna,

These will be dynamically balance before being used and you are correct in that they are fillet root side fit. Do you think this will cause problems during use? They are only going to be used for a short period of time for testing and will only be used in a test rig.

 

[mfgenggear]

724
in opinion ISF for splines is over kill. but be that said good luck.

 

[CoryPad]

The material removal can be as low as 0.0001 inches (2.5 micrometers). Here is a nice article on this subject regarding actual parts:

http://www.gearsolutions.com/articl...ce-of-isotropic-superfinished-precision-gears

 

[724napier]

mfgenggear,

Would you just recommend hand polishing?

 

[mfgenggear]

724

do you have any in house bench capabilities?
if so check with your team and see if they have a small buffing tools that can
reach in between the spline and lightly polish.
give them Gauges if available to prevent OS condition.

Problem with ISF if the rocks(stones) are to large it can not get between the teeth
so it will not polish the entire tooth profiles.
hand polish would be easy and fast.
we have a machine that is called Almco. it would do a similar process as ISF.
it deburrs and polished the entire part automatically.

Good Luck

 

[tbuelna]

724napier-

The fillet root, side fit spline might present a problem in terms of loads on the fwd engine output shaft bearing due to shaft dynamics. It would help if you could provide more details of your test stand drivetrain, or what model of turboshaft engine you are testing.

A fillet root, side fit spline does not provide the same radial constraint that a major diameter fit spline does. So each end of the drive shaft will be somewhat free to whirl around within the female spline, and this constraint condition will affect the critical speed of the shaft.

The photos of the female spline coupling show an overhung arrangement with a 3-hole(?) flange having minimal e/d and that does not seem to provide much stiffness. Is there enough clamping friction from the 3-bolt(?) flange to prevent fretting at this interface? I would assume that this coupling is designed in accordance with the engine OEMs ICD, right? But the torque capability of the spline does not seem to be balanced with the torque capability of the adjacent bolt flange.

Since your coupling does not appear to include any provisions for lubrication of the spline connection, I don't think the surface roughness of the grit blasted internal teeth will present an issue.

 

[724napier]

tbuelna,

Thanks for the response. These will be used to test a T-58 turbine engine. It has a relatively low torque output. The coupling actually has four flanges with mounting holes. Given the thickness of the four flanges and hole diameters, I don't think there will be any issues as far as clamping force. The spline would likely fail much sooner than that of the mounting flanges. I am thinking that the bigger problem will be a wear issue within the teeth of the spline.

 

[Tmoose]

The fixture or arbor used to balance a part should center the part just as it will be centered in service. Likely some kind of arbor locating off the minor diameter will be suggested for these spline hubs. If in service the splined hub will be "centered" (poorly) by the side fit, the possible centering variation, and resulting UN-balance may need to be qualified, or at least understood.

 

[mfgenggear]

Balancing is tough when it requires it to be from a spline. but it is doable. depending how close of balance it requires.
if these parts where broached, and the spline P.D. was cut simultaneous with the minor diameter. then yes it would be easily
tooled from a cheap tapered lathed mandrel that the diameters where ground. with .0003-.0004 taper per inch.
I prefer this type of tooling instead of a major diameter fit. but that is my opinion.

it's also possible to locate on a tapered splined arbor, with one flank straight, and the other tapered .0003 per inch.
this will lock the spline in place. this type of tool needs to be ground.

the other option would to grind the diameter over the spline, and have them locate on that diameter
for balancing. the spline P.D. would have to be held to the ground diameter.

Good Luck

 

[tbuelna]

724napier-

Ah, the good old T-58 turboshaft. There are hundreds of these engines out there in many different configurations. You mentioned the coupling will be subjected to relatively low torque. As a guess, I'd say the power turbine max shaft torque of your engine is probably around 300-350 ft-lb. From your coupling photo, it did not appear to me that the torque capability of the 4 bolt flange and spline were balanced. It looks like the spline has a higher torque capability than the 4 bolt flange, even after the spline capability is de-rated for fretting/wear.

Since the 4 bolt flange will not use match drilled shear fit holes, the torque capability will need to be based on the clamped friction provided by the 4 fastener preloads. I would suggest using a static friction coefficient of 0.15 and a safety factor of at least 1.5 when calculating the torque capability of this connection. With this type of metal-metal connection you want to make sure you have sufficient clamped friction so there is never the possibility for any relative motion at the clamped interface, or fretting damage will result. The holes look to be sized for a 1/4" or 5/16" bolt, and the PCD looks to be around 3". Are these values close to what you used for this coupling design? Can you provide details of the spline and bolt flange?

 

[724napier]

tbuelna,

Thanks for the response and you are dead on about the max shaft torque output. We are snowed in at the moment. Once the weather clears I will post that information. To make sure I understand you in regards to the torque capability based on the clamped friction provided by the fastener preloads. Essentially, I can find the reactions at the fasteners, determine the shear force, then using friction determine the minimum clamp force required by the fasteners.

 

[tbuelna]

That's basically it if you want a simple approach. Fastener pitch radius x fastener axial preload x static coefficient of friction at the clamped interface x number of fasteners. But with this simplified approach you'll want to use a conservative coefficient of friction value and a generous factor of safety to ensure the connection will never experience slippage.

Lastly, I can't stress the importance of this issue enough. As was discussed, the flange/spline runout of the installed coupling, the balance of the coupling and shaft, and the shaft dynamic modes must all be evaluated for their effect on the power turbine shaft bearing system. These bearings don't have much tolerance/margin for loads beyond their design limits, and even a small amount of excess out-of-plane moment from your overhung coupling could cause problems with the bearings. If you haven't done so already, get yourself a copy of the OEM ICD or installation documents. And make sure your coupling, shaft, test stand drivetrain and even the engine attachments are all compatible with what the engine spec shows.

Good luck to you.
Terry

 

[Tmoose]

Figure 1A page 3 ( 4 ) here has info about the relative centering changes that will "burn up" the desired residually balanced condition.

http://www.irdbalancing.com/downloads/techpaper1balqualityreqmts.pdf

For instance, If a "perfectly balanced" part moved off center about 0.0006" the resulting unbalance would be the entire general purpose G 6.3 tolerance.

 

[tbuelna]

I use that spec to define the dynamic balance condition of shafts and gears in aircraft transmissions I design. Typically the rotating parts are balanced to grade 6.3. However there is also a general design rule that the max dynamic radial unbalance force at any bearing should be kept under 10lbf.