Help with vibration analysis of a bracket assembly subjected to forced vibration

[C98hris]

I've been asked by a customer to perform vibration analysis on a simple bracket assembly that will be mounted to a drilling rig. The assembly will be subjected to forced harmonic vibration from the drill which I know the frequency and amplitude of. I need to establish if the vibration of drill will come close to the natural frequency of the bracket assembly. I'm new to vibration analysis and fairly new to FEA so I'm looking for some help to get started. I don't have access to a complex FEA package, but I do have a simple package that I can do some analysis with, including modal and dynamic response. I was also hoping to perform some basic hand calculations to make sure the FEA results are realistic. After reading into this its seem quite complicated to do. Can anyone give me an idea of what I need to get started? or what resources can I use that aren't too in depth?

Any help with this would be really appreciated.

Many Thanks,

Chris

 

[desertfox]

Hi
It might help if you can post a picture or drawing of the bracket in question with some dimensions

 

[C98hris]

Hi,

I've attached a very simplistic drawing. I've just used blocks to represent masses on the assembly.

Chris

 

[GregLocock]

1) Analyse the drill signal to get excitation frequencies (you've already got those)

2) perform hand calc on static stiffness of beacket

3)mesh FEA model, check that static stiffness agrees with (2)

4) add mass to FEA model at point where you checked stiffness. Run dynamic analysis to get list of mode frequencies and shapes

5) check that frequency of obvious mode agrees with root(k/m) where k is from (2)

6) add all the other masses rerun dynamic analsysis

7) check that no modal frequency is within x% of frequencies from (1)

I'd add that the probability of you being even close to the right answer is rather low without doing a proper experimental modal analysis of the bracket. you will learn a great deal.







Cheers

Greg Locock


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

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

 

[Tmoose]

Is the attachment a bolt thru each of the 2 holes 100 mm apart?
A closed form first cut might be as two separate beams each with one end "built in" and with a mass on the end.

http://faculty.uml.edu/pavitabile/22.403/web_downloads/Frequencies_of_Common_Systems.PDF

Establishing realistic boundary conditions for any calculation is often tricky.
I'd expect the real part to behave differently, possibly with a lower than cyphered first mode because the two bolt attachement's flexibility would be greater than " built in."
But the back plate might be curved so tightening the bolts pulls the plate against the main "drilling unit." So the plate would be preloaded and immobilized until the acceleration exceeds the preload, or something in between.

 

[desertfox]

Hi

Vibration is not something I really deal with however I wondered if you treat the black blocks as a solid and calculate its mass and then consider the stiffness of the two bolts acting like springs and calculate the natural frequency from the normal square root (k/m) etc.
Others please chip in I may be well off base here.

 

[izax1]

I think the approach of Tmoose is the best way to attack this. Dealing with boundary conditions will be tricky independent of you using FE or manual analysis. Do a study with bands of attachment stiffness, and calculate a band of natural frequencies and see if that gives you the required frequency. Your (simple) calculation should be at least 2 to 3 times above the lowest driving frequency (rig)

I'd say that this task will require some experience with mechanical vibrations.