Static Analysis of a Composite Propeller

[aeromatt]

I am using FEMAP/NX NASTRAN to model a composite wind turbine propeller. The primary goal is to optimize the twist of the propeller blades at given rotational velocity. I will be applying shear, moment and centrifugal loads statically while adjusting the layup accordingly.

I am new to FEMAP and not sure if there are any unique features to help me model the loads correctly. I was planning to apply the loads with 10 bands of elements equally spaced from the center to the tip (to start anyway..)

Also, any considerations from those who would know how to do this really well with FEMAP would be greatly appreciated.

Thanks you

 

[rb1957]

personally, i'd optimise the twist of the blade with a CFD program (i'm guessing you want to maximise torque, instead of thrust like a typical airplane propeller), then optimise the structur.

 

[aeromatt]

thanks for the reply rb. That makes sense but I'm actually trying to use the twist to limit rotational speed.

 

[rb1957]

isn't that even more of a CFD problem ? it sounds like you're looking at the drag rise, or maybe stalling the blade ... hard to see how NASTRAN will help you ??

personally i'd link the blade pitch controller to a speed governor (decreasing the pitch should depower the blade).

 

[aeromatt]

we've out-sourced the CFD and know what type of deflection will work. we also do not have pitch control. the blade is fixed and hopefully it will be possible to increase drag due to passively induced twist at a given speed.

What do you think...?

 

[rb1957]

1st, the CFD should be telling you the aerodynamics of the blade. ok, it'll probaly consider a rigid blade but it could work with a flexible blade (as it sounds like these are the effects you're trying to utilise).

i guess you're trying to tailor the elatic axis of the blade so that the blade will deflect to depower the itself (like they've done with forward swept wings). very trcky design ... and with a fixed pitch blade ?? (mechanically simple, but diminished operating efficiency

anyways, your question was about loads .. your CFD should have given you pressures, which you can apply to each element ... interpolate onto nodes or element cg ... i suspect you'll probably need something like excel for this.

 

[aeromatt]

Thanks rb. I'll dig into the bulk file as you suggest.

I'm finding that this thing will need to be extremely flexible to achieve enough twist.

 

[rb1957]

it's much more about the position of the elastic axis ... you want the blade to pitch down (i think), so that'd mean having the elastic axis fwd of the center of lift (i think). i suspect you'll need lots of 45deg plies, and few 0/90 plies.

the other thing to consider is if you are making a flexible blade, and the CFD was done with a rigid blade (likely, if you'd don't have a structure) then how good is the CFD ?

 

[aeroafrix2]

May I ask how you are achieving your twist using the composite lay-up i.e. Stretch-twist or Bend-Twist coupling? And are you using the propeller-skin or an embedded box?

Besides what you are trying to do requires a fluid-structure interaction analysis where you would have to couple both your CFD analysis and your Static analysis.

Ansys has a publication on composite lay-up optimisation based on FSI analysis using a Wind Turbine as a case study. Unfortunately I cant lay my hands on the publication at the moment. It might be worth searching and reading through.

 

[aeromatt]

aeroa,

Thanks for the input. I will look for that ansys publication but feel free to let me know if you find it...

I do not think stretch twist is going to add much since the rpm, mass and length is very small.

I am not getting much bend twist from the aero loads but I am getting good twist from bending due to inertia. In other words, the eccentricity of the blade geometry is creating bending and twist.

Thanks

 

[btrueblood]

I'm not sure you want to twist and stall a rotating airfoil, the imbalanced and dynamic (time varying) loads from stalled 'foils might be too much for blades to handle. Twisting "down" (i.e. diminishing angle of attack) as speed increases is more typical, i.e. feathering the props to decrease the lift (and drag) as windspeeds get higher and higher. As rb indicates, it requires lift center to be aft of the elastic center, which can be pretty tricky for an airfoil shape.

The next problem you will encounter with very flexible vanes is flutter instability...whee!

(sorry if my terminology is backwards, but I'm thinking/writing here using fixed-wing conventions.)