Need stiff light material (like Beryllium) 3

[dculp1]

Hello --

I have an application that requires a 8" diameter x 3" high disk with supporting ribs and other features. For proper performance the disk must be both stiff and light. The disk vibrates so fatigue is important. Beryllium seems perfect but is too expensive. Aluminum would be suitable but it is not stiff enough. Because of the part's complex shape carbon fiber is probably not suitable. In any case, a prototype must be produced quickly.

Thanks for any suggestions.

Don Culp

 

[CoryPad]

If Be is too expensive, how about AlBeMet?

http://www.berylliumproducts.com/WhyBeryllium.aspx

If not this, then there likely isn't an easy answer. Most metallic alloys have similar E/[ρ] ratios.

 

[unclesyd]

Have you looked at Aluminum-Lithium Alloys?

 

[TVP]

Are you saying that you cannot adequately stiffen the design with the supporting ribs such that Al can be used? Are you limited in the size, shape, and number of ribs that can be implemented? It is always best to obtain stiffness with the design (geometry) and select the material based on strength/toughness/corrosion/manufacturability/etc.

 

[RPstress]

If you follow unclesyd's advice and use Al-Lith (maybe 2050, 2090 or 8090), you'll get a useful 15–20% advantage in specific modulus.

Al-Lith and beryllium are the only structural metallics with a specific modulus much more than 10⁸ inches. (Al-Lith gives about 1.25x10⁸ inches, pure beryllium 6.3x10⁸ inches according to MIL-HDBK-5 properties). Brush Wellman do beryllium-aluminum products that might be a bit cheaper than pure Be. Specific modulus 2.8–3.8x10⁸ inches.

The only way I know of to get the sort of specific modulus given by beryllium is to use ultra-high modulus (UHM) carbon fibre composite. I'm not sure what that might cost. More ordinary carbon fibre with good design (just putting in fibres every which way should give about 1.24x10⁸ inches) should match or even just about exceed the specific modulus of Al-Lith and be vaguely competitive on cost. Some intermediate solution such as a high modulus (HM) carbon fibre might come out about right.

 

[dculp1]

TVP - Yes, I am optimizing the design (including ribs) with FEA. However, neither aluminum nor titanium is adequate for the given design constraints. AlBeMet shows reasonable results.

RPstress - Is there any carbon fiber that is similar to a filled plastic -- i.e., so that I can just take a block and start machining, rather than building up from layers? If so, where can I obtain such material?

Where can I obtain aluminum-lithium in the size that I need (about 8" dia x 3" thick)?

 

[unclesyd]

Is this perchance work on a bicycle frame?

Alcoa is the leader in Lithium-Aluminum Alloys so i would contact them.

Here is paper on one of the Lithium-Alloys produced by Sandvik.

http://www.smt.sandvik.com/sandvik/0140/internet/s003931.nsf/GenFrame2?OpenAgent&@l=http://

http://www.smt.sandvik.com/sandvik/...//www.smt.sandvik.com/sandvik/0140/internet/s

[/URL]
003931.nsf/index/BF6F7339793DE476C12571AB00340847?OpenDocume
nt@t=http%3A%2F%2F

http://www.smt.sandvik.com/sandvik/0140/i

nternet%2Fs003931.nsf%2FTopBannerWeb%3FReadForm

 

[EdStainless]

Look at high strength steels, 300M and such. But you have to be ready to make things very thin. Better specific stiffness than Al and Ti, but nothing like the specialty high mod grades.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[IRstuff]

Other choices might be engineering ceramics, aluminum nitride, silicon carbide, and silicon nitride, all of which have specific stiffness greater than 10⁸ inch

TTFN

FAQ731-376

 

[tbuelna]

dculp1,

CoryPad's suggestion of Al-Be alloys is worth considering, but you'll need to note a few things about that material.

Al-Be has magical properties for sure. It has a MoE similar to steel with a density less than aluminum. But the downside is that it is very difficult to fabricate and is ferociously expensive.

The most practical way to fabricate parts from Al-Be is by casting. But the parts must be investment cast in a controlled atmosphere, since the material is very reactive with oxygen. Al-Be casting alloys also cost about $300/lb (at least the last time I checked).

Some Al-Be bar, sheet or plate is available. But the sizes are pretty limited since these forms must usually be made by sintering Al-Be powders.

If you use an Al-Be casting, be sure to apply the appropriate knockdown factors to your fatigue analysis for casting defects.

Good luck.
Terry

 

[RPstress]

Re short carbon fiber reinforced plastic, if you managed to obtain 50% fibre and somehow got the manufacturer to use (say) Cytec K-1100 or MCA K13D2U fibre you might get a material with a specific modulus a little more than 3x10⁸ inches.

Victrex produce PEEK filled with carbon. You may not need an exotic plastic like PEEK, but if you're using wildly exotic fibre the cost of the plastic (maybe $50/lb) won't matter too much.

Dave Adkin at Victrex has been quite helpful in the past ( dadkin@victrex.com ).

Overall I wouldn't hold out too much hope of obtaining such ultra-high modulus fibre in a thermoplastic. You might be better off trying to make a filled thermoset yourself if you could persuade MCA or Cytec to supply some fiber. MCA at least list some chopped and milled UHM carbon (

http://www.mitsubishichemical.com/dialead.php

).

http://www.victrex.com/

http://www.mitsubishichemical.com/

http://www.cytec.com/

On the topic of beryllium, note that it's toxic. There was a bit of a fuss a few years back when we wanted to use Cu-Be bushes with just 2% Be. (This was ok in the '80s).

 

[lumenharold]

As noted elsewhere in this thread, Al-Be is expensive and beryllium is toxic. That being said it is primarily a hazard if you are going to modify the finished part since controlling the dust is the challenge. A qualified manufacturer will handle that for you. The material can be machined but few companies are willing to do so. We had optics mounts quoted in Beralcast and the cost was about two orders of magnitude higher than Al but was amazing in every other respect such as projected stiffness and weight. Depending on the design, the part can be cast to net shape.

http://www.mdatechnology.net/techprofile.aspx?id=403

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[dculp1]

I think AlBeMet may be my best option.

Thanks to all.
Don Culp