Accept using a phototach and analyzer, what else is there to determine the resonant frequency of a structure or machine. A friend mentioned a variable speed motor strapped to the structure and thus inducing the renosance by varing the speed. Another mentioned the use of impact hammers that could assist in modal analysis and determanation of resonant frequencies.

Has any of you used a impact hammer with a accelerometer to determine modal shapes and resonance frequencies.

I would appreciate your imput regarding this

Hadrian

Both of these techniques are common and useful for finding resonant frequencies: the "Run up / Run Down test" where the variable shaft rate vibration is used to force the strucure at varying frequencies and the "impact hammer" or "bump test" where a hammer hit inputs a broadband frequency and one looks for amplified response from the structure.

If you are only interested in resonant frequencies, you can do a bump test wih a 2 x 4 instead of a calibrated hammer. This test is much faster and simpler to accomplish.

www.vibtech.com has alot of good technical papers on modal analysis and the use of calibrated hammers. If you have a specific question, post it here and I'll be happy to answer it.

Alan
www.dliengineering.com

I have used a calibrated hammer, fixed, and roving accelerometer to establish resonant modes within a machine through FRF testing. However, I had to have special software for the testing and a dual channel analyzer. If you have a resonant frequency that you have ID through your vibration readings. Then I would suggest setting you analyzer up for a impact test and place your accelerometer in the direction and bump test 180 away from the accelerometer placement of at the placement always in same direction. Monitor a time waveform or time domain which ever you like for a good decay rate. I do not know the equipment you are using, but maintain a frequency sensitive domain. I must use a uniform or at least that is what is has been named because it is more frequency accurate. Hope this helps! Oh, one other thing, softer materials used to bump with will yield lower frequencies, harder materials will excite higher frequencies. The two by four would probably have a range of 10,000 to 60,000 Cpm’s give or take.

Hadrian,
I use the IMI Sensors (PCB Piezotronics) Model 086C41 Impact Hammer.
I have used coastdowns, ODS, and other methods of determining resonance.
The modal hammer is the most reliable and easy to use method.
I roam the accelerometer and use a stationary point to impact because in most cases, I can't roam the hammer to all the locations.
This hammer uses 4 tips and I have been able to get good response using it on my machines.
I use a CSI 2120-2 with VibPro software to store the data for future reference. This also allows displaying the coherance to validate the data.
I also use Paintshop Pro to share the graphics.
By doing a screen capture, I can make doodles on the displayed graphs to explain more clearly what I am trying to show them.

Hadrian, going back to your initial posting you mention phototach. I do not use it for bump tests, only for running the motor you mentioned or running the rotor in the machine itself. The 0-90-180 dgree phase shifts for each resonance is the good support that an angle reading gives. The other responses above mention a lot of rather expensive methods (concerning resources). Just a plain bump test using any spectrum instrument, impacting with fist, foot and a wooden piece 4x4 inch up to a sack used as a pendulum works for many situations. Each type of exciter has its best range. The small added motor has the drawback of the force increasing with the speed squared. A "rotor" made by a ball from a ball bearing blown around with compressed air in an outer ring can be easily adjusted to fit each situation well.

A good approach is to imagine to mode shape of the resonance you want to locate and then excite (hit) at anti-nodes, i.e. maximum deflection points. You can soon survey a full group of resonances when you move the impact point and sensor so that you provoke all resonances and verify them placing the sensor at both nodes and anti-nodes and points in-between.

I recommend you to start with simple cases. Survey the desk in front of you, find all main resonances of the engine in next car outside. Then take a fan on soft rubber elements. Survey the rotor as next level. Learn step by step. You will soon find arguments to invest in more equipment and software. The more you do it, the better you are ready for the real work where the handling of resonances can save you a lot of money in the production. Good Luck Arne

I like coastdown tests, especially if the equipment is on a VFD. It has a little more 'ooh--ahhh' factor when everybody can see and feel it jumping around. But that only works if the natural frequency is less than running speed. If the natural frequency is close to or greater than running speed, you gotta bump it.

Yesterday I looked at three cooling tower motors all in a row that had the same natural frequency--1824 cpm, give or take a few. Made it really tough to run them at full speed. But since 100% from the drive only ran the motors at 1785, we had to bump it to see exactly where the frequency was. About the only 'calibrated' hammer I ever use is a dead blow hammer I got at Home Depot.

Arne, good to see your technical difficulties were resolved.

Hello Patrick, about the dead blow hammer,
- but first a Salute to the Pope and his great mission acomplished over the past 26 years. If each of us would take on a tiny tiny fraction of his approach to serve mankind regardless of shape, colour and origin, within the given boundaries of reality (!), we would at least not cause more havoc to the world. That is something to strive for. May he rest in peace.

The dead blow hammer has a tendency to "stick" to the structure that you hit. Adding a damping to smaller things. For that I like a plain rubber mallet a bit better or a careful "take-back" after the hit. Using a 3 lb normal small sledge and adding a door stop at the tip (of various height) gives you a nice platform to later add an accelerometer and use its output converted to force (F = mxa) for the day when you expand to a two channel collector. I have compared to modal hammers and it gives the same result, just 1/50th of the cost.

Arne -

Thank you for adding such a first class statement on these boards.

Terry O

If you need a source for quality impact hammers - consider these at: http://www.reliabilitydirect.com/impacthammer/hammerindex.htm

Hadrian,
In practical approach, to find the Natural Frequency basically you need to excite the structure to its resonating state. You can use the following tools
1. Sensor and Phototach, within machine operating speed, you can use the unbalance force (or other force at frequency 1.5x, 2x, 2.5x, 3x of TS). Limitation: uncontrollable amount of force and its frequency, limited natural frequency will be excited. Advantage: operational stiffness and damping is taken into account. If you know the amount of force (such as unbalance) used to excite it, dynamic modal parameter can be calculated.
2. Vibration Shaker, Impact Hammer and Sensor. This method can trigger resonance at different frequency range (depend on your shaker force, hammer tip). Limitation: for rotating equipment, damping ratio can be influenced by other parameters (load, speed, temp, lubricant, etc). This method is prefered for static structure modal analysis, such machine base, piping, support, etc.

A non supported and rotating shaft might behave different compared to when it is installed, supported and rotated in a machine.