Designing structures with high resonant frequencies 1

[oharag]

Could anyone lead me to articles about designing system frames that have high resonant frequencies (above 100 Hz). I'm currently designing a system frame for a medical device that will be comprised of aluminum rectangular tubes welded together. Analysis (Algor) has shown the first mode of this system to be around 33 Hz, second mode around 40 Hz, etc... The excitation of the system is being driven by a motor/centrifuge system spinning at 2400 RPM (40 Hz). We have another exciting frequency at 80 Hz (a subcomponent of the centrifuge spinning twice as fast). The centrifuge chamber will be isolated from the system frame using sandwich mounts. Hopefully the centrifuge will be balanced before being placed into service, but during the centrifugal process we will have some slight imbalances brought on by the process we are performing.

I was wondering, again, if anyone could lead me to articles/tutorials that help in designing structures that are resistant to low frequency inputs. Sorry for my ignorance here. I'm new to this I realize you need stiff structures, and cross beams or plates attached to the structure will help in increasing the resonant frequencies. Just hoping some additional enlightenment could be directed my way. Thanks.

oharag

 

[GregLocock]

Sorry , I don't have any references

In order to meet your (rather low actually) frequency target you need, crudely, to make your structure 9 times stiffer for no additional weight.

That is rather a tall order. How big is this thing roughly?

How much do the gubbins weigh compared with the frame?

Which FE package are you using? Does it have an optimiser?



Cheers

Greg Locock

 

[EnglishMuffin]

Presumably, your basic aim is not necessarily to increase the natural frequency of the structure but to minimize the amount of vibration. You can do this in a number of ways.

1. Make sure the resonant frequencies are not close to the excitation frequencies (I am sure you realize this of course). It is worth noting in this regard that some machines such as turbines often operate above their critical speed (shaft resonance) and have to run through this at start up - you can get away with this as long as you have enough damping present.

2. Increase the dynamic stiffness of the structure.
This can be done either by increasing the static stiffness to mass ratio as Locock implies, or by increasing the damping.

3. Isolate the source of vibration (you obviiosly know about this too)

There are a few points to note :

1. It is important that your model has the right boundary conditions, such as floor mounts etc - these can make a big difference.
2. If the heaviest masses present are the things mounted on the frame, rather than the frame itself, then you will probably get higher frequencies (and dynamic stiffnesses) by making the frame out of steel rather than aluminum..
3. Look closely at the mode shapes that you are getting from the Algor analysis. There may be some obvious improvement you can make - shear (including torsional) deformations can be much improved by using triangular bracing, assuming you haven't done that already.
4. You have to put the maximum amount of material where it will do the most good. For example, if the elements of the framework do not see much torsion, their stiffness to mass ratio may be improved slightly by using correctly oriented I or H beams rather than tubes.
5. If all else fails, you can apply damping to the structure. If you can apply enough, it doesn't much matter what the resonant frequency is. You can use tuned absorbers, for example - these need be nothing more than chunks of metal attached to the structure with rubber springs, close to the position(s) of maximum deflection. You can often get improvements in dynamic stiffness of four to one by this method.

 

[Anthonyr]

Since you are designing the frame out of rectangular tubes, you can also add damping by filling the structure with a vibration absorbing grout. The tubes will have to be mitered with openings to allow the grout to travel. You will need two holes in the structure, one for filling and one for letting the air escape. Once the grout cures, the holes can be capped off. Philadelphia Resins has a product called "Chockfast Red" that can be considered.

 

[GregLocock]

Good suggestion. Sand or lead shot is commonly used for this

Cheers

Greg Locock

 

[drawoh]

oharag,

Do you have weight constraints?

It does not sound like you are designing a jet fighter or a racing car. If all your aluminium tubes became three or four sizes bigger, the natural frequency would increase. If the thing just sits in a lab somewhere, no one would ever know or care that your original idea would have weighed half as much.

A common mistake in frame design is to use only orthogonal components. Do you have diagonal members forming a structural triangle? The stiffness of a tube in tension or compression is about three orders of magnitude higher than its stiffness in bending.

Trusses are rigid.

JHG