TIe Rod in Compression 5

[audacious12345]

Hi,
I have a tie rod which is installed on attendant partition in compression. Beacuase of the angle at which its installed it has ltaeral load. I was wondering if any of you have some experince in testing of tie rod which is in compression and installed at some angle? What are the possible modes of failure? Other than compression and the lateral load can cause bending. Anything else? See attached for orientation of tie rod. I dont have any test data to validate its installation. Any help will be highly appreciated.
Thanks

 

[rb1957]

if this is a tie-rod (with spherical bearings ?) then it's a two force member, the force it reacts Has to be axial. the rod can react a lateral component Only if the other end is clamped (and i wouldn't rely on friction).

if it is clamped, then it's a compression rod with bending, could be a beam column (depending on P/Pcr).

 

[SWComposites]

Are the ends the typical clevises with pins connecting to the structure? How is the lateral load applied/reacted at the connection?

Failure modes: rod compression, rod bending, buckling, clevis compression/bending, pin shear, etc.

Are you going to test the tie rod?

 

[audacious12345]

Hi,
Thanks for prompt reply. Both Ends of the tie rod are attached using clevis fittings. See attached.

 

[rb1957]

this rod can only react axial load. instead of pinning the constrained end, set up a co-ord system along the rod and restrain in the axial direction only.

 

[SWComposites]

So the load in the tie rod can only be an axial load (2 force member; no bending), unless there is a sideways load/moment on the clevis (aligned with the axis of the pins), which isn't a very good idea. Please provide a balanced free-body drawing of the tie rod.

 

[rb1957]

to me it looks like the end fttgs have spherical bushings

 

[audacious12345]

Its just a regular tie rod. With Aluminium Tube and Steel Rod Ends. Although it has X, Y and Z loads acting on it (see very first attachment) beacuse of the nature of tie rod resultant will be always be axial load and it will be reacted as a compression. Therefore, I am assuming failure modes would be tie rod tube buckling or rod end bending at rod end and tie rod interface.

 

[tbuelna]

Actually, the tie rod can experience some bending due to the end moment frictions at the sliding interfaces of the spherical bearings. The end moment frictions can result from rotation of the joints under load, or from deflections in the loaded clevis brackets causing the spherical bearing joint axis to move out of alignment. These joint end moments are normally not an issue, but if your tie rod has a very small margin in buckling based on a purely axial force, they may be enough to cause it to fail in compression.

Either way, these friction end moments are fairly easy to calculate if you know the friction characteristics of the joint interface and the applied loads. And then these moments can be applied about each joint axis when analyzing the tie rod for buckling.

Regards,
Terry

 

[rb1957]

and bending and shear of the pins ...

but to have my gripe, your last post says you understand the load to be axial, but your 1st post says there is lateral load ... sigh

different design question, are you tightening the pin over the two lugs of the clevis, or are you clamping the pin on one of the clevis lugs ? probably doesn't matter with an interior component ...

 

[SWComposites]

aud -
what is your specific question(s)?
- do you want to know how to test a tie rod?
- do you want to know how to analyze a tie rod?
- do you want to know how to convince the FAA/EASA that the tie rod is good without any analysis or test data?
- or what?

if you have only an axial load, how would you get "rod end bending at rod end and tie rod interface"? what interface? between the tube and clevis?

 

[tbuelna]

SWC-

In theory, a tie rod with a spherical bearing joint clamped within a clevis bracket at each end (as shown) has an end constraint condition permitting a full 3 DOF's in rotation, and 0 DOF's in translation. If the joints were frictionless, it would be theoretically impossible for the tie rod body to experience any static loading condition other than pure tension or compression. But in reality, that's not the case since friction does exist.

Aircraft structures or mechanisms are always certified by analysis. An analysis using accepted methods and safety factors is all the regulatory bodies like the FAA or DOD normally care about. Testing is expensive and time consuming, so for most situations it is not normally done. The only exceptions being qualification and acceptance testing of sub-systems.

Testing only serves to validate the analysis models. The reason most testing is not of much actual value is due to the large number of variables that every test typically involves. To produce meaningful results from a test, the test procedure would likely need to be repeated hundreds or thousands of times to obtain an accurate representative statistical sampling.

Of course, sometimes analytical FoS's can be reduced if the analysis is later validated by a structural test. But for the most part, a conservative FoS is used for most structural analyses, and testing is avoided.

 

[SWComposites]

tbuelna -

"Aircraft structures or mechanisms are always certified by analysis. " - true 99.9% of the time; however, cert by test is done in some cases (aircraft seats, some general aviation structures, and others)

"Testing is expensive and time consuming" - sometimes it is, sometimes full scale testing is cheaper than a full building block of tests to validate analysis methods.

"Testing only serves to validate the analysis models" - not true; see above

"To produce meaningful results from a test, the test procedure would likely need to be repeated hundreds or thousands of times to obtain an accurate representative statistical sampling" - this is nonsense; full scale airplane tests are not repeated hundreds of times; even coupon level material tests are not repeated thousands of times for a statistical sample.

"a conservative FoS is used for most structural analyses" - not always true either; the weight penalty for doing this for aircraft design is typically not acceptable; appropriate amounts of testing are used to reduce conservatisms and minimize weight.

 

[audacious12345]

SW Composites,

aud -
what is your specific question(s)?
- do you want to know how to test a tie rod?

Tie rod wont be tested independantly but it will be installed on partition and the partition will be tested.

- do you want to know how to analyze a tie rod?

Yes. I want to make sure it doesnt fail in testing. I thinks thats why I get paid to avoid test failures

- do you want to know how to convince the FAA/EASA that the tie rod is good without any analysis or test data?

No

- or what?

Just to cover myself to avoid tie rod failure during testing of partitions.

Thanks a lot.

 

[JJ22mW]

I would just analyze the tie rod as an axial member in compression (using Niu, Bruhn, your company manual or whatever), I've never once had a DER be concerned with effects of end moments due to friction in the spherical bushings, because its most likely negligible. And if you have a positive margin close to 0, either consider whether or not the analysis method or load you applied is conservative, or just size it up a little if youre worried about the test... Also make sure you check the rest of the load path as well (the bolt, fittings, attachments to airframe and partition, etc) and if you analyze everything correctly you should have no worries of test failures.

 

[tbuelna]

SWComposites,

I give you a star for your critique of my post. My comments were probably too general in nature. But I'm a mechanical systems guy, and not a structures specialist. For the aircraft project I'm currently working on, we must use a FoS of 1.5 for parts certified by analysis. But this FoS can be reduced to 1.25 if the analysis is later validated by testing.

With mechanical systems, we do pay particular attention to things like friction effects and how they can impact the total combined load a component might see. And we tend to use very conservative values for friction coefficients, since Mu can vary widely in service, and thus the resulting loads due to friction also can vary widely. In extreme cases, mechanisms with lots of high-friction joints in series can even experience "friction lock" under some conditions.

Indeed, for mechanical systems like an aero control surface linkage, most FAA FAR's or DOD MIL spec's require the actuator driving the linkage to have a 200% force margin by analysis in order to ensure that the device never is subject to a friction-induced stall or a jammed condition due to things like ice accumulation. So the linkage components (including tie rods) should be analyzed for this high level of input force.

As for your comment, "even coupon level material tests are not repeated thousands of times for a statistical sample", I would heartily disagree. The fatigue data presented in material properties documents like MIL-HDBK-5 are the cumulative, statistical result of huge numbers of coupon tests.

Finally, as for making a DER happy with your analysis, you should include every load effect regardless how minor it may seem. Once a DER signs off on your analysis, he is responsible for it for life. So naturally, he will nit-pick every single detail. And based on my experience, this would include accounting for even small joint friction end-moments in your buckling analysis.

Good luck to you.
Terry

 

[kellnerp]

Beacuase of the angle at which its installed it has lateral load.

If it is a tie rod, by definition it has spherical bearings on each end. The vectors add like this: Fx^2+Fy^2+Fz^2=Prod^2. There is no lateral load if you look at the vector sum of the three loads.

I was wondering if any of you have some experince in testing of tie rod which is in compression and installed at some angle? What are the possible modes of failure?

Primarily buckling in compression and tensile failure in tension. Also the end connections can be a source of failure.

Other than compression and the lateral load can cause bending.

There is no lateral load (bending) unless one or more ends binds up. The joints have range of motion limits, and the designer should have checked that there is no binding through the range of motion. Looking at the design it appears (and this is an educated guess) the ends would bend. A bent end would impose a bending force on the member which might be a buckling problem. This would have to do with analysis of a failure mode though.



TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."