Whip Antenna Analysis

[brysonc]

Hello all,

I am running across a problem trying to analyze reactions in antenna brackets used to support whip antennas. The primary concern is our random vibration analysis. Many of the antennas have a spring base that keeps the antenna fairly "rigid" most of the time. However, when the antenna deflects radially, the spring causes it to "snap" back. I'm not sure what you would call it, but it's as if the spring experiences coil bind at the center of oscillation when it's damping out the recoil.

There are some ways that I have come up with to model this non-linearly, but that doesn't help me for my linear random vibration study. Simply measuring the stiffness of the spring and adding that to the model won't take into account the "shock" effect.

If this is unclear, I can try to elaborate. If anyone has any suggestions or ideas, please let me know as I am having trouble overcoming this obstacle.

Thanks,

Bryson

 

[btrueblood]

I'm not sure you can do much, if the mode exists in reality. I.e. if mechanical impacts (coil-to-coil contact) occur, then you either damp the motion, or live with the high frequencies generated by the impacts.

 

[brysonc]

I'm not trying to eliminate the motion, I'm trying to recreate the scenario using FEA tools. I would like to generate a similar scenario to use in my analysis of the antenna mounting bracket.

 

[rstupplebeen]

I would doubt a linear random vibration study could handle this. I think you need to send this to an explicit solver. Hopefully you have one and the mesh can be coarse. I hope this helps.

Rob Stupplebeen

 

[brysonc]

I do have a non-linear solver than I can use, however it won't allow me to use my input PSD curves to analyze random vibration response. I'm thinking the best method may be to accoustically measure the dominant natural frequency, then maybe create a spring-mass system to represent the antenna. I guess I would have to ignore the high frequency inputs as well as the shock loading, but that might be acceptable for the random vibration study (input PSD curves are between 5 and 500 Hz). Any thoughts?

 

[gwolf2]

Have you tried running it through the PSD analysis and comparing the resulting deflections with the linear portion of your spring's response? It doesn't solve the intelectual problem but you may get away with it if the driving power and frequencies are not enough to really get it going.

Failing that, real time dynamics with loadings derived from the psd. Yeuch! (but interesting). Sorry I can't help more - interesting question.

gwolf

 

[GregLocock]

I'd do it in matlab, skip the fea. The problem with psd data (I'm assuming at the mounting pt of the antenna) is that to simulate the non linear effect of the base spring snapping back into position you need to be in the time domain. Which you can't really do accurately from psd data.

You could synthesize a time history with the same spectrum as the psd, but you have to make some assumptions that will have big effects on fatigue life.







Cheers

Greg Locock


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

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

 

[kellnerp]

The PSD is at the base of the antenna right? But you are analyzing the whole antenna. The base spring limits the moment on the whip itself. The coils bind to provide a stable base for the antenna most of the time. Typically the base will "break away" when the antenna hits an obstruction. Perhaps you could run the PSD as if the base is solid and see how many times the moment on the base spring exceeds breakaway.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."