Aluminum Honeycomb Data

[bavfean]

Hi, I am doing a FE test in lsdyna which involves aluminum honeycomb panels being impacted by an automobile transversely. To model this correctly i need the core properties for an .5" thick aluminum core (multiple foil gauges and cell sizes would be great, although any would do as I just need a baseline right now) I assume that the modulus in the W and L direction are essentially 0, As I cannot input 0 into dyna what recommendation would you have? also is it correct to assume that the compressive modulus will be the same with and w/o laminate? Lastly, is there any known poisson's ratio for the core?

Thanks in advance

Junior Engineer
Friedman Research Corp.

 

[GBor]

An internet search for Aluminum Honeycomb Properties will get you a long way. Here is a link to one such aluminum honeycomb:

http://www.plascore.com/honeycomb/pdf/3003.pdf

It was the first of about 100,000 hits. I tracked through the company's website to the Aluminum product data sheets. hope it helps.

 

[bavfean]

I am really looking for a stress-strain curve so that lsdyna can accurately predict the failure mode, in which a quick internet search proves hard to find

thanks

Junior Engineer
Friedman Research Corp.

 

[GBor]

I have to admit that I haven't used LSDyna, but the stress-strain curve for aluminum honeycomb should still be the stress-strain curve for aluminum. The question is typically more of how much aluminum is actually there and how is it distributed.

Honeycomb core is typically orthotropic. The deflection is determined by various Young's moduli along the given axes. Failure is still often determined from failure strengths cited by the material properties compared to some anticipated failure mode.

Perhaps a little more about your specific application would cause more people to jump in on this...

 

[bavfean]

Thanks for the advice GBor

A client of the company is planning on making a honeycomb fairing which can be attached to the back end of semi trailers to reduce drag and increase fuel efficiency. Our company specializes in automotive safety and was hired to see how the fairing would change the crash dynamics of a rear end collision with a car or another truck. The plan is to use LSDyna to run a simulation of a rear end impact with the honeycomb, to see weather the honeycomb will penetrate a laminated windshield and deform the header.

In order to run this test accurately I will need to input the data from a stress-strain curve from aluminum honeycomb loaded in compression (along the height of the cells).

It is my understanding that the SS curve of the honey comb is not the same as straight aluminum. (the curve will spike at yield then decrease to a horizontal slope for remaining duration of the load, as opposed to spiking at yield then increasing to a higher ultimate strength with straight aluminum)

It is proving difficult to find empirical data for the honeycomb. So if anyone has more insight into this problem it would be greatly appreciated.

Junior Engineer
Friedman Research Corp.

 

[IRstuff]

> Would the fairing even go down that low? Typically, a passenger car, or even a moderate sized SUV would barely make it to the bottom of the deck for a trailer, wouldn't it?

> Wouldn't you model the honeycomb as an array of boxes, rather than some sort of unitary material?

TTFN

FAQ731-376

 

[bavfean]

Very Good point about the height compatibility although larger SUVs, other large passenger vehicles, and commercial trucks do have the potential for windshield impact.

Also in LSdyna it is much easier to model a solid core with honeycomb designated properties (less modeling and meshing time as well as less computing power on the cray)

Junior Engineer
Friedman Research Corp.

 

[IRstuff]

Well, if it's the windshield, then it's really the rooftop vs. the fairing, and I'd guess that the fairing would win.

Another semi, on the other hand, might be a fairer comparison.

I assumed that you were trying to use a solid with fudged properties, but I just can't see that you'd get the same general behavior. Although, perhaps if you modeled it as an object with a skin with one property and the interior with a different property, you might get some similar behavior.


TTFN

FAQ731-376

 

[bavfean]

thats exactly what we are doing, the core with its properties and the skin with its own properties. The reason for using the honeycomb is because it has very poor properties when loaded transversely, as it would be in a rear impact, which would mean that the automobile header (hardened steel) would have little deformation

IRstuff, do you have any experience with LSdyna? would MAT_126 or MAT_026 be representative of the core?

Thanks

Junior Engineer
Friedman Research Corp.

 

[IRstuff]

Sorry, no...

TTFN

FAQ731-376