No Stress Peaks 1

[Apsu]

Does any one know of a method to design a gas turbine or compressor blade so that under a combined tension (CF) and gas bending load the traditional located stress peaks at LE,TE and Blade to Platform are avoided.

 

[GregLocock]

Well, obviously if you make it look like a tapered cone with a nice fillet at the base you can reduce the stress peaks down to as low as you like.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Apsu]

Thanks Greg, how about avoiding the fillets

 

[tbuelna]

Apsu,

Put the blade in compressive preload.

Regards,
Terry

 

[rb1957]

"avoided" ... no
"reduced" ... possibly

maybe terry's suggestion ... shot-peen (or Hot Isostatic Pressure) the surfaces, interference fit the blades into the hub. but this'll create different problems with the hot turbine blades (obviously ?)

but then i'm not sure how these suggestions would work with single crystal blades ??

bit surprised that your stress peaks are at teh L/E and T/E ... i'd've thought you had more of a problem with normal bending of the blade ... just my 2c

 

[Apsu]

Hi Terry
Your suggestion about compressive preload is interesting particularly if I consider a ceramic blade design as crack fracture growth would be minimised. I am not aware of any nozzle guide vane to turbine blade configuration that would offer the compressive blade scenario. The Turbine blades would have to rotate on the outside with the tips pointing inwards, whilst NGV remain static.

I need to think about a suitable configuration more .

Regards
Apsu

 

[Apsu]

Thanks RB1957

Yes, I agree a great reduction of kt can be achieved by traditional geometric adjustments that usually involve some form of fillet.

Applying a compressive stress by shot peening as you suggest, into an area prone to kt maybe a consideration, it would preload the region and reduce the kt effects.

If fillets can be avoided its also a manufacturing benefit as it reduces the steps to production by not only one, but removes the need to confirm quality control on the fillet consistency.

I get kt more in the trailing edge (and some LE)and root of the Divergent blade design. The loading is combined.