Good high speed shaft material?

Making a shaft that will run 20K rpm in one of my R/C boats, it'll b 1/4" diameter with a keyway and threads on one end to hold gear. I've had 2 break right at the end of some threads but wil eliminate them as a possible stress point.
My first was made of 416 stainless and didn't last long at all, th
next was 1144 "stress relieved" and did better bu eventually broke too. I need something that can be turned and no warp and that is tough. A toolmaker buddy suggested 414 "pre-heat treat", what's the thinking here

  Quoted Text Here  
Holy shit, how much HP are you asking of that shaft? Make a bigger shaft.

TG,
If the stressproof shaft (SAE 1144) broke, odds are that the pre heat-treated SAE 4140 will fail also. The physical properties of the two are quite similar except for their machinability.
If you can make the shaft larger in diameter. You can even use an aluminum tube of the right dimension!
If that is not possible you will need to remove any stress risers at the end which keeps breaking. Is a keyway really necessary? Would loctite do? How about a collet-type connection which would eliminate threads and keyways.
Wolfgang
Tom Gardner (nospam) wrote:
  Quoted Text Here  

  Quoted Text Here  
What is affixed to the threaded/keywayed end that keeps breaking off?
i

Terry Keeley wrote:
  Quoted Text Here  
snip
A toolmaker buddy suggested 4140
  Quoted Text Here  
4140 is good. I would have said 4130.
Watch your stress risers, and if the part is failing due to fatigue from vibration, balance the shaft. 6-4 AL-V titanium might work, though it would need to be a third larger for the same stiffness.

Thanks for the replys. The shaft is in the gearbox shown here
http://gallery.intlwaters.com/displayimage.php?albump&pos
The end that breaks can't be seen, but it has a collet system tha
attaches to the drive shaft, that's the end that breaks. I' estimate the motor puts out about 3hp at 30K rpm, the reduction i 1.5:1 so the shaft turns 20K. The plastic gear is held with a key
I would definetly like a bigger shaft but it means bigger bearings i
the unit which would be a real PITA. Hoping a re-design and bette material does the job

Terry Keeley wrote:
  Quoted Text Here  
The gearbox photo doesn't come up -- just a picture of a (very nice) RC boat.
How long is the shaft that's breaking, and how long is the drive shaft? If they're long enough to be resonant at 333.3Hz or it's close multiples (667, 1000) then they'll really whip around; no amount of stress relief would help you then. For that matter the motor power pulses will still be coming at 400Hz, so you have to worry about those resonances, too. Solving that would involve making the shafts more torsionally stiff, to move the resonances up.
If that's not it then the suggestion to balance everything (your shaft _and_ the drive shaft) would be good.

Can the shaft be made a little bigger between the bearings? You don't want stress risers so if possible make a smooth transition to maybe 2 mm greater diameter.
Dan
Terry Keeley wrote:
  Quoted Text Here  

On Fri, 28 Jul 2006 04:40:26 GMT, snipped-for-privacy@spam.invalid (Terry Keeley) wrote:
  Quoted Text Here  
Greetings Terry, You can also use 4340 in the heat treated condition. I think that's what your toolmaker buddy is talking about when he says pre-heat treat. I.E. the material is heat treated before machining. The stuff machines nicely, though it is tough because of the heat treated condition. I would rather machine it than 4340 in the annealed condition because of the much better finish attainable just turning with carbide. Since the shaft broke at the threads it looks like you have a stress riser there. What does the thread relief look like? And how about the threads themselves? Are they smooth? Ideally, your thread relief should be cut with a full radius tool. And the threads need to have as large as possible the radii in the roots and on the crests. A good finish is very important. Fine scratches can lead to cracking. So remove them with 600 or finer wet or dry sandpaper. The keyway could also be a problem. How is the finish in the keyway? Does the keyway start at the end of the threads? Have you looked with a magnifier to determine exactly where the failure started? Cheers, Eric

On Fri, 28 Jul 2006 04:40:26 GMT, snipped-for-privacy@spam.invalid (Terry Keeley) wrote:
  Quoted Text Here  
===============You may be encountering "shaft whirl" which causes the shaft to vibrate or flex like a plucked guitar string at certain rpms. All steels are about equally "stiff" or resistant to bending [although some are much more to take a permanent set or bend] so switching materials if this is the case won't help. How long is your shaft [l/d ratio is critical parameter here]. Can you install an intermediate bearing? It won't have to handle much load, just enough to keep the shaft from starting to bend/twang.
Unka George (George McDuffee)
...and at the end of the fight is a tombstone white with the name of the late deceased, and the epitaph drear: “A Fool lies here, who tried to hustle the East.”
Rudyard Kipling The Naulahka, ch. 5, heading (1892).

Beat me to it, but I'd endorse that

F. George McDuffee wrote:
  Quoted Text Here  
That's why I was asking about the shaft that he's driving -- if _that's_ vibrating it could take out the gearbox shaft before it dies. And I would expect a 'driveshaft' to have a longer l/d ratio than a 'gearbox shaft'.

Isn't a tube within a tube the best way to handle this? No whippng. Octura used to make lots of different drive systems for hydros. The guys around here (San diego) always seem to need to run the thing up to nearly full throttle and toss the whole boat into the water. That has to put incredible stress on the shaft at the weakest point. Maybe you could add some rubber isolator or clutch to soften the shock

Thanks again guys, looks like I'll go with "heat treat-stres proof" 4140, or "normalized" as it's described in th Machinery's Handbook. This material has a tensile strength o 148,000 and yeild strength of 95,000 whereas the 1144 I used befor is rated at 96,750 & 58,000 respectively
The big help should be the elimination of the threads on the end tha
keeps breaking, I'm goin to fix the collet to the shaft wit set-screws instead. If this set-up breaks I'll have to go with bigger shaft..

  Quoted Text Here  
    With setscrews, you will have a bit more difficulty balancing the shaft -- but I'll bet that the keyway introduced enough of an imbalance at the speeds which you mentioned.
    If you *must* have a keyway -- can you cut *two* of them at 180 degrees -- and use two keys as well -- to maintain the balance?
    And another thing which *might* improve your reliability with the threaded end of the shaft. Look into having the threads rolled, instead of cut. This makes for a more rounded root of the thread, and the rolling process makes for greater strength beyond that of the stress riser problem.
    Good Luck,         DoN.

  Quoted Text Here  
If you want to gild the lily, make your new shaft from VascoMax C-300 or C-350 (Teledyne Vasco). These materials have high yield, tensile, toughness, ductility and impact strengths, as well as high fatigue and compressive strengths. Could be your breaking problems will come to an end.
It will require heat treatment, but VascoMax requires temperature no higher than about 925 degrees F for several hours to achieve maximum strength.. Distortion is limited or eliminated, and shrinkage is predictable.
Harold

TK,
My reference for 1144 steel (cold finished) states the following: Yp minimum: 100,000 psi TS: 125,000 psi Machineability: 83% of B1112
For 4140, heat treated and stress relieved: Yp: 120,000 psi TS: 140,000 psi Machineability of annealed 4140: 55% or B1112
Agreed that 4240 HT & SR has a slightly higher yield point, but in steels with a tensile strength of over 100 ksi this does not affect the resistance to fatigue failure (endurance limit) by much.
Consequently I stand with my statement above that if a shaft of 1144 steel failed then, other things being equal, a shaft of 4140 HT & SR will likely fail also.
The suggestion to attach the flexible shaft onto the gear box with a double-ended collet sounds eminently suitable to me and that is what I would do.
Wolfgang
Terry Keeley wrote:
  Quoted Text Here  

  Quoted Text Here  
On Fri, 28 Jul 2006 04:40:26 GMT, snipped-for-privacy@spam.invalid (Terry Keeley
  Quoted Text Here  
b
You can also use 4340 in the heat treated condition. I think that' what your toolmaker buddy is talking about when he says pre-hea treat. I.E. the material is heat treated before machining. The stuf machines nicely, though it is tough because of the heat treate condition. I would rather machine it than 4340 in the anneale condition because of the much better finish attainable just turnin with carbide. Since the shaft broke at the threads it looks like yo have a stress riser there. What does the thread relief look like? An how about the threads themselves? Are they smooth? Ideally, you thread relief should be cut with a full radius tool. And the thread need to have as large as possible the radii in the roots and on th crests. A good finish is very important. Fine scratches can lead t cracking. So remove them with 600 or finer wet or dry sandpaper. Th keyway could also be a problem. How is the finish in the keyway? Doe the keyway start at the end of the threads? Have you looked with magnifier to determine exactly where the failure started Cheers Eric[/quote:a51be0eec1
Sorry, I wasn't seeing all the posts looking through the DIY forum.
Yes, i tryed to make the threads as smooth as possible and ground small radius in my threading tool so the bottoms weren't sharp.
The keyway end isn't a problem..
Think eliminating the threads completely will help a bunch

On Sat, 29 Jul 2006 09:13:20 GMT, snipped-for-privacy@spam.invalid (Terry Keeley) wrote:
  Quoted Text Here  
==================Have you considered an alternative material such as carbon fiber reinforced tubing, possibly with adhesive bonded metal ends? Unka George (George McDuffee)
...and at the end of the fight is a tombstone white with the name of the late deceased, and the epitaph drear: “A Fool lies here, who tried to hustle the East.”
Rudyard Kipling The Naulahka, ch. 5, heading (1892).

  Quoted Text Here  
On Fri, 28 Jul 2006 04:40:26 GMT, snipped-for-privacy@spam.invalid (Terry Keeley
  Quoted Text Here  
b
You can also use 4340 in the heat treated condition. I think that' what your toolmaker buddy is talking about when he says pre-hea treat. I.E. the material is heat treated before machining. The stuf machines nicely, though it is tough because of the heat treate condition. I would rather machine it than 4340 in the anneale condition because of the much better finish attainable just turnin with carbide. Since the shaft broke at the threads it looks like yo have a stress riser there. What does the thread relief look like? An how about the threads themselves? Are they smooth? Ideally, you thread relief should be cut with a full radius tool. And the thread need to have as large as possible the radii in the roots and on th crests. A good finish is very important. Fine scratches can lead t cracking. So remove them with 600 or finer wet or dry sandpaper. Th keyway could also be a problem. How is the finish in the keyway? Doe the keyway start at the end of the threads? Have you looked with magnifier to determine exactly where the failure started Cheers Eric[/quote:7aae61f5b4
Sorry, I wasn't seeing all the posts looking through the DIY forum.
Yes, i tryed to make the threads as smooth as possible and ground small radius in my threading tool so the bottoms weren't sharp.
The keyway end isn't a problem..
Think eliminating the threads completely will help a bunch