Analyzing Vibration of Bicycle Frame as a Spring Mass System

[chandran]

I have a bicycle frame. I want to analyze this frame as a spring mass system.
What details i should know. I should know the frame's stiffness. ok let us say it as x. Then the centre of gravity of the frame. Let us say it is also known.

But now the frame can be pulled or pushed and then released from any point on it so that it vibrates. (unlike a simple spring mass system where the mass is separate and is pulled through some distance and released to induce vibration).

How can such a system's natural frequency or how dynamics can be done on this?

 

[FredGarvin]

You need to do a search on modal analysis, rap tests and the like. On something that large, you can simply place accelerometers at different locations and hit the frame to induce vibration. The accelerometers record the data. From that data you should be able to discern the various modes' frequencies (up to a point). If you use enough instrumentation you can then discern mode shapes.

This kind of analysis is done quite often. We do modal analysis on blades to discover what their excitations are so we can avoid them in the operating ranges of the engine.

 

[chandran]

fred,
What are the places the accelerometres should be kept? Any website
on this type of analysis

 

[FredGarvin]

Usually one has an idea of where the maximum deflections are going to occur (mid spans of beams, etc). These are the places you try to place the accelerometers.

First and foremost, the Society for Experimental Mechanics is a great place for information. http://www.sem.org/PUBS_ArtDownload.asp

An introduction to modal analysis:
http://www.sem.org/ArtDownLoad/msma98.pdf

Other articles of interest:
http://www.gmi.edu/~drussell/guitars/electric.html
http://www.sem.org/ArtDownLoad/msja00.pdf
http://www.sem.org/ArtDownLoad/msjf00.pdf

As a side note, I have done a lot of testing simply using a microphone to pick up the local vibrations. Again, this is a single blade, not an entire bike frame. This will not give you the mode shapes, but it will give you the mode frequencies. We then use the models to determine which modes are being excited. This way is a much easier test set up.

 

[joslopher]

Hi,
If I wanted to get the natural frequencies of the bicycle frame by means of FE-analysis, should it be performed as constrained or unconstrained ? Indeed, the wheels are not bounded to the ground, so please, could you help me?
Thank you very much

 

[AlephZero]

If you want to use the results for some purpose, the model should be constrained the same way as the "real life" situation.

If you want to correlate the model with test results, it's always a good idea to start with no constraints. That removes one source of uncertainty in the model, because "real world" constraints are never "perfectly rigid".

Assuming the bike is standing on the ground, the wheels are "bounded" to the ground by the weight of the bike, unless the vibration amplitude is so large they lose contact with the ground. But you would probably need to include the flexibility of the tires to get a good correlation with test measurements.

 

[joslopher]

Ok! Thank you very much!

 

[joslopher]

However, I am not sure about the assumption that the wheels are "bounded" to the ground by the weight of the bike...
Should I constraint also the positive z-direction on those nodes corresponding to the link to wheels because of the weight? But weight affects to all nodes of the structure...