Push Rod Flex

[Robbobnob]

There is nothing wrong with the deformation of the push rods. Have they caused a failure... no.

As they are on the inside wheel on the rumble curbing, they will have significantly less load than the outer tyre, so loading will be decreased and allow more movement to be excited by the input frequency of the rumble curbs.

Any and all vibrating systems vibrate at the same frequency of the input function, however if this input is close to or equal to the natural frequency of the structure, then the amplitudes of the vibrations in the system will be increased, up to 200 times the size of the input vibrations.

That being said, it simply looks like fundamental bending wave is being induced in the structure. This has no detrimental effect of the structural stiffness or capability for the structure to perform its task.

[bill shoe]

I think the Wikipedia entry for "buckling" is lousy. It describes the mathematical description of buckling. It does not describe buckling itself. It was written by someone who confuses a mathematical description of something with the thing itself. As you were.

[marcush.]

the Caterham push rod certainly does not look like a one piece item -so according to Savage´s paper it is at least a risk of bonding crack propagation for them?

[Tim.Wright]

I think the main danger is failure of the carbon section, as thats where the bending moment will probably be the greatest.

That being said, it simply looks like fundamental bending wave is being induced in the structure. This has no detrimental effect of the structural stiffness or capability for the structure to perform its task.

I'd disagree on that. I'd suggest the part was pretty close to failure. Yes, even composite parts flex under load, but not to the degree that you can see it on TV. The composite used in the wishbones are of a very high modulus, so if there are deflections big enough to see with the naked eye, you can be pretty sure the stresses inside are massive.

The intersting this is that the outside wheel push rod seems ok for now.

Tim

[thisisatest]

The front brake duct is also shaking, as if it's resting against the pushrod.

[DaveW]

I appear to have caused some controversy with my use of the word "buckling". Here is an attempt to explain.

The first point I made was that the video was "An example of poor installation stiffness caused by poor design". The rationale for that was rather quaintly phrased "To get any significant change in push-rod length it has to buckle", meaning that compression deformation could, in the context, be neglected.

By buckling I meant that the observed bending deflection was, presumably, caused by changes in end load (i.e. the deflection was not in the nominal "direction" of the applied load). The bending deflection (whatever its cause) would certainly reduce the distance between the push rod attachment points, hence decreasing its in-line stiffness. Whilst it is possible that the deflection was caused by a constraint occurring just at one end of travel, it is equally likely that the reduced stiffness persisted throughout a lap. That was my main point.

Bill Shoe called the Wikipedia definition of buckling as "lousy". I didn't really have a problem with it, I'm afraid, unless "buckling" is taken to be synonymous with "buckling failure".

Some have slightly ridiculed my contention that the natural frequency of the push rod in bending would be expected to be "in the hundreds of Hz". Whilst I have experienced the phenomenon, I struggled to find a justification for my claim, until I hit on the idea that somebody may have published information for designing wind chimes. Indeed they have. I don't want to make too much of an issue with this, but it does suggest that, for example, a 300mm tube of 0.625 inches o.d. by 20 swg has, despite the glorious mixture of units, a free-free natural frequency of just over 1,000 Hz. This would, of coarse, increase with the ends pinned. Please use you own numbers, or, alternatively visit.

I also thought of the "singing saw" phemomenon (changes natural frequency with applied bending moment), and discovered this, which presented a relationship between natural frequency and buckling (that word again) load. I concluded that it was unlikely to be an issue in this case, but it is in accord my view that buckling is not synonymous with buckling failure.

Gary Savage's article was a good find, and he has already been quoted. I would like to add the following "Inhomogeneity in manufacture and the non-symmetric nature of the global design lead to an uneven distribution of load. Any variation in from one component to another will be manifest as a scatter in the buckling (that word again) loads."

[marcush.]

The front brake duct is also shaking, as if it's resting against the pushrod.

I thought about this as well .Question is why would these make contact ? does the upright twist the wishbones enough to exert a force on the pushrod or is it the other way round the bending pushrod pushing on the brakedrum?

[Tommy Cookers]

the rumble load frequency is around 200-250 Hz ?
this would produce a dramatic response in a pushrod of Fn in the range 70-400 Hz ?
(pushrods (aero section?) are rather long and have relatively heavy ends)

what interested me was how much (or little?) rumble load (at such a high frequency) was apparently transmitted by the tyre
presumably the existing load sensor data would show the engineers what is going on ?

[Caito]

Don't know much about materials. But.. isn't there a chance of a failure in the bearing? Or the hub itself.. why is it so obvious that it's the push-rod the on at fault?

[Jersey Tom]

Don't know much about materials. But.. isn't there a chance of a failure in the bearing? Or the hub itself.. why is it so obvious that it's the push-rod the on at fault?

Looked to me like it was pretty clearly the push rod taking a buckled shape. Could have the wheel bearing or hub fail and the pushrod would still be perfectly straight (or at least until the car crashes).

[riff_raff]

Sorry about being late to the discussion. I would also disagree that this is not structural buckling in the classic sense. A structure that is experiencing buckling would not oscillate.

Instead, it seems more like a case of a structure responding to an input force coinciding with a natural frequency mode. In this case the structure modal response to the forcing frequency is a simple reversing bending mode. All elastic structures have numerous free-body structural modes that occur at various frequencies. The structure's response at these critical frequencies can assume various shapes such as single or multiple amplitude S-bends, axial "chucking", torsional oscillation, etc. How energetically the structure responds to the forcing frequencies can also be greatly affected by factors such as dampening.

Having said all that, the additional stress produced by this particular bending type modal response will reduce the buckling limit of the pushrod structure. So it would be something to be avoided during operation.

[olefud]

Having said all that, the additional stress produced by this particular bending type modal response will reduce the buckling limit of the pushrod structure. So it would be something to be avoided during operation.

This would seem to be the most significant concern. The longitudinal load would change the natural frequency perhaps out of susceptibility to the exciting events. This may be why only the unloaded rod seems to be acting up. But if the maximum load was suddenly applied while the rod was at the extreme excursion, it could be a problem.

[DaveW]

I like your post, riff-raff, but I'm not totally sure that the following is quite correct:

Instead, it seems more like a case of a structure responding to an input force coinciding with a natural frequency mode. In this case the structure modal response to the forcing frequency is a simple reversing bending mode.

If the push rod is completely symmetrical, and the ends are perfectly pinned, then the push rod should not bend, not until the Euler compression load is exceeded, anyway.

Clearly, there is a problem. I don't know what it is, but suppose the outer bearing is constrained so that an end load also imparts a bending moment. Then the inner bearing will be loaded so as to react the end load, but will also have a component reacting the bending moment. This, starting a zero at the inner end an increasing to a maximum at the outer end, will cause the push rod to deflect, and the end load then would further deflect it, until an equilibrium is reached. I think the "shape" of the deflection would depend on the end contraints, rather than any of the push rod natural mode shapes.

Anyway, I have a general question. Clearly the clip is not in real time. Would anybody care to guess how much it has been slowed down?

[flyboy2160]

... until I hit on the idea that somebody may have published information for designing wind chimes. Indeed they have. I don't want to make too much of an issue with this, but it does suggest that, for example, a 300mm tube of 0.625 inches o.d. by 20 swg has, despite the glorious mixture of units, a free-free natural frequency of just over 1,000 Hz. This would, of coarse, increase with the ends pinned. )...

i couldn't get the calculator to open, but if the material used in those calcs is some stiff "ringing ' metal, their results won't apply to a composite pushrod.

even with mostly 0s in the layup to resist buckling and using very high modulus fibers, i can't see getting a natural frequency in the 1000s. the resin matrix in composites acts like a damper. in fact, this natural damping is used to advantage in 'special' applications.

[flyboy2160]

a follow up to my previous post:
a poor man's way to detect delaminations and voids ~> about an inch on a side in relatively thin composite structures is called the "coin tap" method. it consists of tapping a coin on the surface (some companies have 'official' metal disks for this) and comparing the resulting sounds from good and bad regions. once you calibrate your ear, this is surprisingly efficient in thin structures. it's obviously not as good as a real ultrasonic or laser vibration or xray inspection, but it's a good "quickie" look at seeing large delams.

anyway, the coin tap sound you get in composites is a low frequency, well damped 'clunk' sound.

if you tap a metal wind chime, you get a much higher frequency, ringing, 'ding' sound.