Hey folks, I have posed this question in a previous post and it was never really addressed. So, I will ask again. Does anyone have any experience with using a force hammer for impact tests? I have two available (Kistler 9722A200 & 9728A20000). I have recently gained some interest in natural frequencies and am led to believe that using force hammers can greatly increase the accuracy of finding these frequencies. Problem is, however, that I don't know how to use them (hammers). Can someone explain the process to me or point me in the direction of an easy "how to guide"? Just for some additional info, I do have a variable speed motor mounted to a test stand in my office. I have performed a bump test on this motor and have found the natural frequencies after performing the "Bump Test Equipment Off" function with my 2130. I have verified that these are indeed natural frequencies by inducing and unbalance state into the variable speed motor, setting the 2130 into the monitor mode, and slowly raising the speed until it matches the peaks found during the bump test. I feel extremely confident that this test is right. Now, I would like to learn how to do the same process with one of these hammers. I realize that the different tips give you the ability to induce force of differing fmaxes, and I do know how to set up the 2130 to accept the force hammer on channel A and the accel on channel B. I need to know what settings to use after this though. Can someone help??

This message has been edited. Last edited by: Billy,

You need a 2-channel analyzer. Hammer goes on channel A and accel on channel B. Enable 2-channel mode.

Set up an 'acquire spectrum' with Channel A set to 'SENSR' and Channel B as 'VELOC' Try 200 Hz, low cutoff of 3 Hz, 400 lines, normal averaging, 4 averages, Trigger Mode of 'PRE-TRG' (pre-trigger), Trigger Level of 1 (lb) on A Channel, 10% pre-trigger, Full Scale range of 20 on A, and 0 on B (will enable autoranging on channel B, the accel input).

That said, I am not sure how you will interpret the results. I have the "Advanced 2-Channel" firmware on my 2-channel 2120 which takes care of everything for me.... I've never tried using the hammer in the brute force fashion discussed above.

Good luck.

I tried a setup like you said Rusty, but get no results.I can't seem to get the analyzer to trigger.

Is it possible to check for a resonance with a modal hammer??

I'm using a 2130 in my application, not a 2120, but I believe that I have advanced two channel capabilities.

Bwze,

I have a 2120-2 and use the downloadable Advanced 2-channel program. It is not handy to give you screen by screen setup, but here are some general tips:

1) Use an impact or impluse test setup if available
2) Use Flat-top and not Hanning Window
3) Make sure both channels (A & B) have ICP power switched on
4) Connect hammer to Chan-A and accelerometer to Chan-B
5) Hammer sensitivity is typically about 0.001 mv/lb and accelerometer is typically 100 mv/g
6) Use acceleration units and not velocity or displacement, unless you have a good reason not to
7) Let both channels auto-range for best FS amplitudes -- hammer blows are typically 100 to 1000-lb peak and accelerometer is typically 1 to 50-g's
8) Use 400 to 800 line of resolution and set Fmax to 100 to 5000 Hz depending on structure, frequencies of interest, and hammer tip hardness
9) Use a Pre-trigger of 10% and trigger off Chan-A with a typical level of 2 to 10 lb
9A) The 2120 analyzer has predefined Force and Response window functions that help with better data quality, but are not essential
10) Hit structure with a clean bounce-back hit and allow analyzer to acquire data between each hit
11) Several hammer hits may be needed to allow analyzer to autorange and then average 4 good hits
12) Use Coherence function to indicate good data with COH over 0.8 or better
13) Use TF-mag peaks (g/lbf) to indicate natural frequency
14) Use TF-imag (g/lbf) to more accurately indicate natural frequency and to plot mode shape
15) Correct TF-imag data for accelerometer orientation
16) Use MS Excel to plot mode shapes
17) Test one or more easy structures to learn how technique works before looking silly on a real job!
Try CSI for info at www.compsys.com
Also other info: http://www.modalshop.com/resources.asp
http://www.modalanalysis.com/
Peter Avitable at University of Lowell has excellent training materials: peter_avitabile@uml.edu

Walt
w_f_strong [at] msn {dot} com

Walt / Rusty, thanks for the great info, but maybe I'm wrong in my assumption.

Is it acceptable to use a modal hammer to confirm for sure what a natural frequency is instead of just using a soft/hard faced hammer? I am aware of the fact that using a modal hammer is correct for determining mode shapes, but (and this is just a conclusion that I'm drawing from information I've seen or heard) won't a modal hammer remove all doubt about whether a peak in a spectrum is a natural frequency or just background noise? The way I see it is that if you use a modal hammer, then the analyzer will only use the force that you have inputed by use of the hammer to determine natural frequencies. Is this correct, or is it only acceptable to use these hammers for ODS testing?

Bwze,

The accelerometer measures both vibrations from the hammer impact and from other sources at the same time. It cannot tell the difference. The force-hammer two-channel method allows you to use the Cohenence function to verify that the vibration at a particular frequency is from the hammer (high COH) or not (low COH from background vibrations).

If you are using a single channel analyzer and ordinary hammer, then you should at least measure the backgound vibration spectrum before impacting the structure. Since you own two modal/impact hammers and a 2130, you should learn how to use them correctly and then use them. If you get lost on an island without them, then use the single-channel coconut impact method! [just kidding]

Walt

Walt, after setting up the 2130 I was able to generate some data that lines up with the data I got from just hitting the motor with a regular hammer.

How do you determine coherence?

I have no option for a Flat-Top window, but I did use a Hanning. Seems to have worked out OK.

Bwze,

Coherence is computed from the Transfer Function based on 2-channel measurements. You'll have to consult the 2130 manual to see if and where that feature is located. I thought the CSI 2130 was supposed to be more powerful than the 2120, but I guess not! The Hanning window is not intended for impact data, but you can use it for a quick and dirty test.

Walt

If you don't buy the modal/ods package for the 2130, you don't get any of the nifty transfer functions. As far as I've been able to tell, you might as well hit it with a 2x4.

Hey lawrencep, I do have the ODS/Modal package for the 2130, but wasn't able to able to make heads or tails out of it. Although, I will be perfectly honest and say that I haven't done any research into it.

I will try to find out some more info tomorrow though.

Keep playing with the setup. I believe the different plot options will show you the coherence, etc.

Attached is a paper from Vibrant Technologies (ME'Scope).

This message has been edited. Last edited by: lawrencep,

28-_Experimental_Modal_Analysis.pdf (1,435 Kb, 24 downloads)

Bwze, just my opinion (and I may be alone in thinking this), but I would think you would want to at least read the manual for your meter and software, check CSI's technical support section on their web site, etc, etc, before you come on here asking folks to take their valuable time to tell you how to use your equipment.

You need to take some initiative, and that doesn't mean just "playing" with the setups on your meter.......

Sorry that I hadn't noticed this reply until today, but after reading and thinking about what was said, I felt the need express my feelings about it.

Excuse me, Rusty, for using this forum to try to find information pertaining to the field of vibration analysis. I had taken "initiative" to investigate different avenues into how to use this equipment, but had exhausted many efforts up to the point of posting this, to little or no avail. I have since, through the help I received here, been able to learn the correct way to use this device.

In retrospect, I use this forum to help find answers to questions that I'm unable to under normal circumstances due to the fact that I work alone and don't have easy access to someone with experience. I didn't realize that the folks who posted here considered it a waste of their, as you put it, valuable time to help answer, what I think, is a fairly non customary part of this field. If I am incorrect in my assumptions, then I feel that I have misunderstood the reasoning behind posting in this forum in the first place. Could it be possible that there is non published book of rules pertaining to what to and what not to post here? If this is the case, then maybe it should be placed somewhere that folks can have easy access to it so that they don't waste anyone's valuable time.

This message has been edited. Last edited by: Billy,

BWZE..

Rusty said it was "just his opinion", and from what I see.. several have responded to your question. Doesnt appear that your wasting everyones time.

Sometimes people have a bad hair day,I dont think there is a handbook on RULES on this forum, I have asked some questions before that some may have thought ridiculous. If there is it would be nice to see a forum rules posted on the front page. Might keep me from wasting someones time.

"My opinion" .. ask whatever you wish to ask, if you get responses then someone didnt think you were wasting their time, if you get no responses.. take the hint

On some topics, I have had offers to email instead of post on this forum questions that may seem "ho hum" to others. I have found people here to be very helpful, and they dont charge !!

I dont know anything about the hammer,if I did I would give ya my 2 cents, in past when I performed a bump test I whacked it with a 4x4.

I have actually followed this topic to copy and paste the answer in my archives folder in case I ever had the need to use one.

take care and have a good day

Mike

Thanks Mike, I understand it was "just his opinion" but didn't feel it was justified. I too have my opinions, but this is an open forum.
Anyway, I don't have any hard feelings against Rusty. I actually respect him very much and have had personal correspondence with him in the past. He helped me in the beginning of my short time in this field with just starting out.
I agree that folks do have bad days, and maybe I should have just let it pass, but I guess I let my feelings get the best of me.
Anyway, enough of this touchy feely stuff. No harm, no foul.

BWZE,
When you have the 2130 on the home screen do you have Adv. Analyze on the top right hand corner (F7 key)? If you do select this and go th manual analyze. Press Set Analyze Mode and go down to Impact. You then need to set your parameters and input setup (force hammer on A and accel on B) You can autorange but that takes a lot of extra hits, you can set a fixed range for both instead. After you take your averages plots will be displayed. If you don't see Spectrum, Coherence, and Phase go to plot setup and choose these plots to be displayed. Any peak in the spectrum will be a natural frequency if at that freqency you have a peak in the spectrum, good coherence and about a 90 degree phase shift. If you don't have all 3 it could be coming from something else. Remember that the machine needs to be down to do this test.

Thanks Jim, I figured this out after doing some playing around with the analyzer over the last few days in my spare time. I'm hoping to perform my first successful impact test with this force hammer over the next few days. I've got some questions that I'm trying to get some help on in another post that deal with the physical hammer setup. Oddly enough, the Kistler website doesn't have any information about the hammer other that specifications. It was a little late today to try to give them a call, so maybe I can do it tomorrow.

Hey Billy, I think that I could have helped you with this. Give me a call if you want.

Bwze, I don't think that expecting you to read the manuals is too much to ask... you have a 2130, 2 force hammers, the Advanced 2-channel package.... I can't believe there aren't some manuals for this somewhere. I guess I was struck by your statement that " I wasn't able to able to make heads or tails out of it.... I haven't done any research into it" when discussing the ODS/Modal package.

In less than 5 minutes, I found the following articles at CSI's web site (I just Googled 'Computational Systems')... perhaps they will help:

http://www.compsys.com/drknow/pubtnot.nsf/9f0e142090062...053fe93?OpenDocument

http://www.compsys.com/drknow/pubtnot.nsf/9f0e142090062...04f97f6?OpenDocument

http://www.compsys.com/drknow/pubtnot.nsf/9f0e142090062...054642f?OpenDocument

Perhaps you original question has been answered, but to my knowledge, there is nothing all that "special" about a so-called "modally tuned hammer"... I believe that is just marketing talk. An impact hammer simply has a force transducer mounted in the head, and usually a provision for mounting different striking tips. As pointed out, a 2x4 will work about as well if you are not using a 2-channel analysis package. A note here, when doing a "simple" impact test (single channel), I do NOT like to use a deadblow hammer is they can tend to double-hit. I prefer a hard rubber or plastic tip that will actually "bounce" off, and away, from the surface so I don't get a double hit.

This message has been edited. Last edited by: rustythevibeguy,

One can get good results for the simple bump test by using random impacts with a Hann(ing) window and with averaging. Double hits will not be a problem when using random impacts. Just beat on the beast without a particular rhythm (For me this is easy since I didn’t get the rhythm gene.). Besides this is a good method to relieve tension.

One must hit the same location or close to it in the same direction. Since the extra hits add energy to the test, one may not have to hit so hard each time. This generally results in more impact energy in the test signal. The results can be excellent, particularly with the averaging.

If one starts beating before starting the analyzer, there is no need to use a trigger to start the bump test.

The true modal test does have advantages in addition to the extra time it takes as a disadvantage.