Determining Elevator Stiffness 1

[Taz99]

Hi All,

I need to determine the vertical bending stiffness and torsional stiffness for an elevator at a number of spanwise locations. We have target stiffnesses at each location that the design must meet/exceed.

However, I am not clear on how I can use the fine grid FE model that I have created of the elevator to predict the elevator stiffness.
Questions that come to mind are;
1. Constraints to apply to the model. I believe that I should constrain the hinges and actuators.
2. Applied Loads on the model. Point Load, Point Moment? etc. and where to apply these, midway between the spanwise locations where I need to measure the stiffness, at the spanwise locations?
3. What to measure. Displacement or rotation a certain nodes.
4. What equation to use to convert the displacement/rotation measured into a stiffness.

I have looked into the methodology that is used for wings to measure the stiffness via an FE model, but translating this to an elevator is not 100% clear to me.

If anyone could describe the method they have used to for an elevator or maybe similar structure, rudder, aileron, flap etc I would appreciate it so I can use this information as background to know that I am on the right track in terms of general industry practice for the further discussions that we will be having with the customer.

Thanks

 

[SWComposites]

Well, if it was me, I would
- calculate the bending stiffness by a hand calc at each section (add up EI of each piece of the cross section)
- calculate the torsional stiffness using the FEM, fix the hinge fittings, apply a unit load to one actuator fitting, get the rotation angle of the actuator fitting and using the length between the adjacent hinge fittings calculate the effective GJ (or something like that). Repeat for each actuator fitting.

But I would probably first ask the customer exactly how they want the torsional stiffness define and calculated. They may want the torsional stiffness between the actuator fittings rather than relative to the hinge fittings.

SW

 

[SWComposites]

or perhaps your CAD program can calculate EI for you if you can define E values for each material in the cross section.

 

[cpinz]

I agree with SW about the need to make some hand calculation:

- in this way you have something to use for a check of your fem analysis;
- to apply rigid constraints on elevator ( moving part of horizontal tail) means that the stabilizer (fix part of horizonatl tail) is rigid too. But this, in my opinion, is not true; then the vertical stiffness of elevator results lower (but close to real) than the case of rigid constraints.
- on these basis, the overall assembly (elevator+stabilizer) must be considered, which, at the end, results in a redundant structure.
Nevertheless this redundancy can be solved if the "true" constraints of the elevator , that are the hinges,are in a low number (say 2-3 for each elevator semispan). Applying methods like strain energy, virtual work, Castigliano or similar and considering the system stabilizer+elevator like a beam system (where the hinges become your constraints),and performing a freebody analysis, is it possible to calculate the displacements. I suppose that you have the geometry and dimensions of the components, then EJ and GJ can be easily found.
- about loads: there are many ways, and as SW said, it depends on what they exactly want.Loads can be applied evenly distributed, as lumped loading, etc.. Maybe the degree of accuracy in your analysis will be based on the purpose of this work: design, certification..?
- about the need to measure displacement/ rotation: in my opinion it makes sense only you a have real numbers to compare with, otherwise what is a fem analysis for ?

Other opinions or suggestion are of course welcomed...

Cpinz

 

[rb1957]

my 2c ...

you're trying to meet a spec reqmt ... how is the spec worded ?
1) is it a frequency reqmt ? (ie 1st mode > X Hz) >>> dynamic FEA
2) is it lbs.in^2 ? (ie EI and GJ) >>> geometry analysis at various x-sections
3) is it in/lbs ? unlikely but >>> apply the load specified to your model
4) is it in deflection under limit load ? >>> even more unlikly but easier to determine from your FEA

remember your FEA is an approximation of a stiffness model, your FEA is designed to determine internal loadpaths.

another approach (maybe) is to say "do we meet this requirement with the low stresses the FEA is showing ?" ... hint, these should be low stressed componenets, 'cause they need to be stiff.

confirm with GVT

 

[SWComposites]

I was assuming that the spec defined the requirement in terms of EI, GJ, as these are typically used in control surface aeroelsstic analyses and in control surface actuator performance analyses.

SW

 

[rb1957]

EI, GJ work well on a wing, or an H. Stab., but an elevator stiffness is highly dependent on the hinge positions.


EI is easily calc'd from cross-section geometry, but doesn't reflect the hinges.

 

[GregLocock]

When we specify a requirement in FEA world we define the boundary conditions and give typical load cases.

So it might be worth asking your customer for clarification. If they don't care then you'll need to have another think.



Cheers

Greg Locock


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

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

 

[Taz99]

The stiffness requirements are given is lb.in^2 at multiple points along the span implying and EI and GJ. These points are not aligned to hinges/actuators and are not evenly spaced either.

As rb1957 mentioned, I was wondering about the effect of the hinges on the elevator stiffness and how the hinges might make it different from the case for the wing.
For a wing, one FE method to derive EI I have seen documented is to constrain the root and apply a unit moment Mx at say rib 7 and then measure the rotational deflection around the longitudinal axis (x) at rib 7 and rib 6 then:
EI = Length_rib6_to_rib 7 / (Rotation_rib7 - Rotation_rib6)

I'd still be interested in hearing if anyone has a methodology to determine elevator stiffness that they have used, as mentioned by Greg, what boundary conditions were applied what load cases etc. or how you did you do it by hand. I would have thought that there must be a reasonably uniform way of doing this in industry.

Finally any gotcha's on elevator design would be interesting. First one from above,
1. Design will be stiffness driven.

 

[rb1957]

many ways to skin cats ... since I is a cross-section property, it is easy to calc from the geometry, or from by loading the elevator in pure bending (shear load through the centroid), then interprete the results as My/I (or fixing one rib an applying a monet at the other (but that sounds like a bunch of work)

 

[SWComposites]

using a FEM to determine EI in this case is akin to using a pile driver to install a picture hanger .....

 

[GregLocock]

On the other hand you could (I would) argue that in the olden days aircraft structures were designed so that they could be analysed by hand, now that we have FEA for free it is quite possible, or even likely that an efficient design would be inordinately difficult to analyse accurately by hand.

Cheers

Greg Locock


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

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